3 (cid:1) -Untranslated Region of Phosphoenolpyruvate Carboxykinase mRNA Contains Multiple Instability Elements That Bind AUF1*

Phosphoenolpyruvate carboxykinase (PEPCK) is regulated solely by alterations in gene expression that in-volve changes in rates of PEPCK mRNA transcription and degradation. A tetracycline-responsive promoter system was used to quantify the half-life of various chimeric (cid:1) -globin-PEPCK ( (cid:1) G-PCK) mRNAs in LLC-PK 1 -F (cid:2) cells. The control (cid:1) G mRNA was extremely stable ( t 1 ⁄ 2 (cid:3) 5 days). However, (cid:1) G-PCK-1 mRNA, which contains the entire 3 (cid:1) -UTR of the PEPCK mRNA, was degraded with a half-life of 1.2 h. RNase H treatment indicated that rapid deadenylation occurred concomitant with degradation of the (cid:1) G-PCK-1 mRNA. Previous studies indicate that PCK-7, a 50-nucleotide segment at the 3 (cid:1) -end of the 3 (cid:1) UTR, binds an unidentified protein that may contribute to the rapid decay of the PEPCK mRNA. However, the chimeric (cid:1) G-PCK-7 mRNA has a half-life of 17 h. Inclusion of the adjacent PCK-6 segment, a 23-bp AU-rich region, produced the (cid:1) G-PCK-6/7 mRNA, which has a half-life of 3.6 h. The (cid:1) G-PCK-3 cleotides and ligating into pBlueScriptII-SK( (cid:4) ), which was previously digested with Asp718 and XbaI. sequences of the forward and reverse oligonucleotides were 5 (cid:2) -GTACCGTATGTTTAAATTATTTTTA-TACACTGCCCTTTCTTACCTTTCTTTACATAATTGAAATAGGTATC-CTGACCA-3 5 (cid:2) -CTAGTGGTCAGGATACCTATTTCAATTATGTA-AAGAAAGGTAAGAAAGGGCAGTGTATAAAAATAATTTAAACATA-CG-3 plasmid and SpeI. PCK-7 RNA was into three subfragments, PCK-8 (bp PCK-9 (bp 2587), and PCK-10 (bp 2567–2595). Double-stranded oligonucleotides encoding the PCK-8, PCK-9, and PCK-10 RNAs were cloned into pBlue-ScriptII-SK( (cid:4) ) that was previously digested with Asp718 and XbaI. The pBSSK-PCK-8, pBSSK-PCK-9, and pBSSK-PCK-10 templates were obtained by digesting the plasmids with BssHII, SacI, and XbaI. pBSSK-PCK-Mut-6, pBSSK-PCK-Mut-1, pBSSK-PCK-Mut-2, pB-SSK-PCK-Mut-3, and pBSSK-PCK-Mut-4 were also constructed by an- nealing double-stranded oligonucleotides and ligating them into pBlue-ScriptII-SK( (cid:4) ). The template for the pBSSK-PCK-Mut-6 construct was obtained by digesting the plamids with BssHII,

Phosphoenolpyruvate carboxykinase (PEPCK) 1 catalyzes the conversion of oxaloacetate to phosphoenolpyruvate, a rate-limiting step in gluconeogenesis. However, this activity is not regulated by allosteric mechanisms or by covalent modifications (1). Instead, it is regulated by mechanisms that affect the level of the PEPCK mRNA and subsequently determine the level of the PEPCK protein. In liver, transcription of the PEPCK gene is inhibited by insulin and is acti-vated by glucocorticoids, thyroid hormone, and glucagon, which acts through cAMP. Within the renal proximal tubule, parathyroid hormone and angiotensin II increase cAMP levels and cause an increased transcription of the PEPCK gene (2,3). The level of the PEPCK mRNA in rat kidney is also increased rapidly following the onset of metabolic acidosis (4). The latter adaptation is initiated within 1 h and reaches a 6-fold induction within 7 h. The 6-fold induced level of PEPCK mRNA is sustained in the kidneys of rats that are made chronically acidotic (4).
The time required for an mRNA to change from one steady state to another is usually proportional to its half-life (5). Thus, rapid induction of an mRNA is feasible only if the mRNA has a rapid turnover. Previous studies have shown that the PEPCK mRNA is degraded rapidly in liver and in hepatoma cells (6), in rat kidney cortex (1), and in LLC-PK 1 -F ϩ cells (7), a line of porcine proximal tubule-like cells selected for their ability to grow in the absence of glucose (8). Previous studies have also demonstrated that the half-life of the PEPCK mRNA is increased in liver in response to cAMP (9) and glucocorticoids (10). Increased stability may also contribute to the sustained induction of renal PEPCK mRNA during chronic acidosis (4). Therefore, selective mRNA stabilization may play an important role in the physiological regulation of PEPCK gene expression. However, the mechanisms that cause the rapid turnover and selective stabilization of the PEPCK mRNA are unknown.
The presence of a destabilizing cis-element within the 3Ј-UTR of the PEPCK mRNA was previously demonstrated using DRB, a polymerase II inhibitor, to determine the half-life of a chimeric ␤-globin-PEPCK mRNA expressed in LLC-PK 1 -F ϩcells (11). The parent ␤-globin (␤G) mRNA decayed with a half-life of Ͼ30 h, whereas the chimeric ␤G-PCK-1 mRNA, containing the complete 3Ј-UTR of PEPCK mRNA, decayed with a half-life of 5 h. RNA gel shift assays using a cytosolic extract of rat renal cortex identified two protein binding interactions, but only one exhibited properties characteristic of specific binding (12). The high affinity and specific interaction mapped to the PCK-7 segment of the 3Ј-UTR of the PEPCK mRNA, whereas the low affinity and nonspecific interaction mapped to the PCK-6 segment. However, functional studies using luciferase constructs suggested that both segments contribute to the rapid decay of the PEPCK mRNA (12).
In the present study, the contribution of individual segments of the 3Ј-UTR to the stability of the PEPCK mRNA were quantified by analyses of the half-lives of multiple ␤G-PCK reporter constructs expressed from a tetracycline (Tet)-regulated promoter. Direct binding assays demonstrated that AUF1 binds to each of the segments that function as instability elements. Mutational analyses identified specific sequences within the PCK-6 and PCK-7 segments that bind AUF1. The results demonstrated that multiple AUF1 binding sites contribute to the rapid turnover of PEPCK mRNA.

EXPERIMENTAL PROCEDURES
Materials-Male Sprague-Dawley rats were purchased from Charles River Breeding Laboratories. [␣-32 P]UTP and [␣-32 P]dCTP (3,000 Ci/ mmol) were purchased from MP Biochemicals. The oligo labeling kit was from Ambion. Restriction enzymes, RNase T1, T7 RNA polymerase, and yeast tRNA were acquired from Roche Applied Science, New England Biolabs, or MBI Fermentas. GENECLEAN kits were obtained from Bio101, Inc., and the PCRScript cloning kit was obtained from Stratagene. Micro Bio-spin columns and chemicals for acrylamide gels were purchased from Bio-Rad. RNAsin was obtained from Fischer. An RNA standard was purchased from Invitrogen. Dulbecco's modified Eagle's medium/F-12 base medium was purchased from Sigma. Hybridase TM Thermostable RNase H was purchased from Epicenter. Geneticin (G418) and hygromycin B were obtained from Mediatech. TRIzol® reagent was purchased from Invitrogen. Anti-AUF1 monoclonal antibody (5B9) was a gift from Dr. Gideon Dreyfuss. Polyclonal anti-AUF1 antibody was purchased from Upstate Biotechnology. Anti-CP1 and anti-CP2 antibodies were obtained from Dr. Stephen Liebhaber. Anti-HuR antibody was purchased from Santa Cruz Biotechnology. An expression plasmid that encodes the p40 isoform of AUF1 was obtained from Dr. Jeffrey Wilusz. All other biochemicals were purchased from Sigma.
UV Cross-linking of RNA-Protein Complexes-The cross-linking experiments were performed with minor modifications of the procedure of You et al. (13). The samples were prepared as described for the RNA gel shift experiments, except that, following RNase T1 digestion, they were transferred to a 96-well microtiter plate and exposed to shortwave (254 nm) radiation for 5 min in a UV Stratalinker 2400 (Stratagene). The samples were transferred to 1.5-ml microfuge tubes, and an equal volume of 2ϫ SDS sample buffer containing 10% (v/v) glycerol, 5% (v/v) ␤-mercaptoethanol, 3% (w/v) SDS, 184 mM Tris-Cl (pH 6.8), and 0.2% (w/v) bromphenol blue was added. The samples were heated in boiling water for 5 min and then resolved on a 10% SDS-polyacrylamide gel. The gel was dried and exposed to a PhosphorImager screen.
In Vitro Transcription-In vitro transcription of 32 P-labeled and unlabeled RNAs was performed as described previously (12), and the products were purified using Bio-Rad Micro Bio-Spin P30 columns. The 32 P-labeled RNAs were quantified by scintillation counting, and diethylpyrocarbonate-treated water was added to adjust the sample to the desired concentration. The absorbance of the unlabeled RNAs was measured at 260 nm, and the concentration of the transcripts was calculated using extinction coefficients determined from the nucleotide composition.
RNA Electrophoretic Mobility Shift Assay-An aliquot of rat renal cortical extract (12) containing ϳ1-8 g of protein (14) or 40 -400 ng of recombinant AUF1 was preincubated for 10 min at room temperature in binding buffer containing 0.5% Nonidet P-40, 1 mM dithiothreitol, 2 g of yeast tRNA, 40 units of RNAsin, 10% glycerol, and 10 -40 fmol of 32 P-labeled RNA. A 30-, 200-, 300-, or 500-fold excess of competitor RNA was added with the labeled RNA. The reaction mixture was incubated at room temperature for 20 min, and then 1 l of RNase T1 was added and the samples were incubated for 10 min at room temperature. The samples were separated on 5% native polyacrylamide gels, and the gels were dried and exposed overnight to a PhosphorImager screen. In the supershift experiments, the various antibodies were preincubated with the rat renal cytosolic extract for 20 min at room temperature and then added to the binding buffer.
Construction of Tet␤G-PCK Vectors-A tetracycline-regulated promoter system (15) was utilized to map the segments within the 3Ј-UTR of the PEPCK mRNA that function as instability elements. A chimeric ␤-globin/growth hormone gene was cloned into the pTRE2 plasmid to yield pTet␤G. This construct contains a Tet-responsive promoter, the coding region of rabbit ␤-globin (␤G) gene, and the 3Ј-nontranslated region and polyadenylation site of bovine growth hormone (bGH) cDNA. The 3Ј-UTR of the rat PEPCK cDNA (corresponding to nucleotides 2008 -2595) was PCR-amplified using forward and reverse oligonucleotide primers that introduced SpeI and XbaI restriction sites at the 5Јand 3Ј-ends, respectively. The sequence of the forward primer was 5Ј-GCACTAGTGGGCGAATTGGGTACTAG-3Ј, whereas that of the reverse primer was 5Ј-CGTGGTCTAGATACCTATTTCAATTATG-TAAAGAAAGG-3Ј, where the underlined sequences are the SpeI and XbaI sites, respectively. The amplified sequence was then cloned into the SrfI site of pPCRScript-SK(ϩ). The resulting plasmid pPCR-PCK-1 was then digested with SpeI and XbaI to yield a 604-bp fragment that was subsequently cloned into pTet␤G at the SpeI site to yield pT␤G-PCK-1. pT␤G-PCK-1 was digested with BssHII and EcoRV restriction enzymes to release a 224-bp 3Ј-fragment. The remaining plasmid was blunted and re-ligated to produce pT␤G-PCK-2. The 224-bp fragment was blunted and inserted into pTet␤G that was previously linearized with EcoRV to produce pT␤G-PCK-3. Double-stranded oligonucleotides containing the PCK-6, PCK-7, and PCK-6/7 sequences (12) were synthesized with SpeI and XbaI overhangs and inserted in pTet-␤G that was previously digested with SpeI to yield pT␤G-PCK-6, pT␤G-PCK-7, and pT␤G-PCK-6/7, respectively ( Fig. 1). Thus, all of the chimeric ␤G-PCK mRNAs retained the bGH 3Ј-UTR that is contained in the control ␤G-PCK mRNA.
Creation of T␤G-PCK Stable Cell Lines-LLC-PK 1 -F ϩ cells were obtained from Dr. Gerhard Gstraunthaler and cultured as previously described (8). Cell lines that stably express the T␤G-PCK-1, T␤G-PCK-2, T␤G-PCK-3, T␤G-PCK-6, T␤G-PCK-7, and T␤G-PCK-6/7 constructs were made by calcium phosphate cotransfection (16) of the pT␤G-PCK plasmid and pcDNA 3.1/Hygro (Invitrogen) into 8C cells, a line of LLC-PK 1 -F ϩ cells that overexpress the tTA protein. 2 At 24 h post-transfection, the medium was removed, the cells were washed twice with phosphate-buffered saline, and then fresh medium containing 0.2 mg/ml G418 was added. At 48 h post-transfection and every 2 days thereafter, fresh medium containing 0.2 mg/ml G418 and 0.6 mg/ml hygromycin B was added. After 14 -21 days, individual colonies were isolated and expanded. After the cells were split onto 10-cm plates, the hygromycin B concentration was reduced to 0.2 mg/ml. The cell lines were tested for Dox responsiveness by maintaining the cells for 48 h in medium in the presence or absence of 0.5 g/ml Dox. Total RNA was harvested from the cells and analyzed on a Northern blot.
Half-life Analysis-For pulse-chase analysis, cells expressing the ␤G-PCK-1 mRNA were initially grown in medium minus Dox and then transferred to medium containing 50 ng/ml Dox for 48 h. This level of Dox was determined to be sufficient to completely shut off transcription from the Tet promoter. The cells were then washed two times with phosphate-buffered saline and grown for 3 h in fresh medium minus Dox to create a pulse of ␤G-PCK-1 mRNA. Subsequently, 1 g/ml Dox was added to inhibit transcription. RNA samples were isolated at various times following re-addition of Dox and subjected to Northern analysis to follow the decay of the newly synthesized ␤G-PCK-1 mRNA. Alternatively, the various T␤G-PCK-expressing LLC-PK 1 -F ϩ cells were grown for 5-7 days in medium minus Dox. At time zero, 1 g/ml Dox was added to each plate. At 0, 3, 6, 9, and 12 h post-Dox treatment, RNA was isolated and subjected to Northern analysis.
RNase H Treatment-RNase H treatment (17) was performed to selectively cleave the 3Ј-ends of the chimeric ␤G-PCK-1 mRNAs that were isolated from the pulse-chase experiments. The 21-nt PCK oligo 5Ј-GCCCAAGATTTTTTTCTCCCC-3Ј that encodes the template strand of the PEPCK cDNA from bp 2199 -2219 was synthesized by Macromolecular Resources (Fort Collins, CO). A 30-mer of oligo(dT) was used to remove the poly(A) tail of the mRNA. The RNA samples were precipi-tated from Formazol by adjusting the final concentration of NaCl to 0.2 M and adding 4 volumes of ethanol. After 5 min at room temperature, the RNA was pelleted by centrifugation at 10,000 ϫ g for 10 min. The pellets were resuspended in 25 l of DEPC-treated water. The cleavage reaction was performed in a final volume of 12.5 l containing 10 g of RNA, 1.0 l of RNase H, and 50 pmol of the PCK oligo with or without addition of 50 pmol of oligo(dT). The mixture was incubated at 55°C for 60 s, and then 1.5 l of 10ϫ RNase H buffer (0.5 M Hepes, pH 7.4, 1.0 M NaCl, 20 mM MgCl 2 , and 10 mM dithiothreitol), prewarmed to 62°C, was added, and the reaction was incubated at 62°C for 15 min. An additional unit of RNase H (freshly diluted to 1 unit/l with 1ϫ RNase H buffer) was added, and the sample was incubated at 62°C for 15 min. Thereafter, 29 l of denaturing buffer (20 l of Formazol, 3 l of 10ϫ MOPS, and 6 l of 37% formaldehyde solution) was added to stop the reaction. The sample was incubated at 55°C for 5 min and then stored on ice. The RNase H-treated RNAs were subjected to Northern blot analysis as described previously (11), except that 10 l of a RNase H gel loading buffer (50% glycerol, 1 mM EDTA, 0.25% bromphenol blue, 0.25% xylene cyanol, and 25 l/ml of 10 mg/ml ethidium bromide) was added to the samples, and a 1.2% agarose gel containing 20 mM MOPS, pH 7.0, 1 mM EDTA, and 1 mM sodium acetate was used. The blot was hybridized with the bGH probe.
Northern Analyses-Total cellular RNA was isolated using the TRIzol® reagent, and the RNA concentration was determined by measuring the absorbance at 260 nm. A 507-bp fragment of rabbit ␤-globin cDNA was excised by restricting pRSV-␤G (18) with HindIII and BglII. A 2.0-kb fragment of the 18 S ribosomal RNA cDNA from Acanthamoeba castellanii was excised by restricting pAr2 (19) with HindIII and EcoRI. A 228-bp bGH fragment was excised from pPCRScript-bGH with SphI. The fragments were separated on a 1% agarose gel, excised, and purified using the GENECLEAN kit. The synthesis of oligo-labeled cDNA probes and Northern analysis were performed as described previously (11). The blots were exposed to a PhosphorImager screen, and the intensity of the resulting digital image of each band was quantified using Molecular Dynamics software. The level of the chimeric ␤-globin mRNA was divided by the corresponding level of 18 S rRNA to correct for errors in sample loading. For half-life studies, the log of normalized data were then plotted versus the time after the addition of doxycycline. The reported values are the mean Ϯ S.E. of data obtained from triplicate samples. The line representing the best fit of the data points was determined by a KaleidaGraph program that weights each data point based upon its standard deviation.

RESULTS
Half-life Analysis of ␤G-PCK-1 mRNA-A tetracycline-regulated promoter (Tet-off system) was used to determine the half-life of the ␤G-PCK-1 mRNA. Stably transformed cells were maintained in the presence of 50 ng/ml doxycycline (Dox) to suppress synthesis of the ␤G-PCK-1 mRNA. A transcriptional pulse was created by removing Dox from the medium for 3 h and was subsequently chased by adding 1 g/ml Dox to selectively inhibit transcription of the T␤G-PCK-1 mRNA. RNA samples were isolated from the cells at various intervals following initiation of the chase, and the levels of ␤G-PCK-1 mRNA and 18 S rRNA were quantified by Northern analysis (Fig. 2A). This analysis demonstrated that degradation of ␤G-PCK-1 mRNA is initiated after a brief lag (30 min) and then proceeds with a very rapid half-life (t1 ⁄2 ϭ 1.2 h).
Deadenylation of ␤G-PCK-1 mRNA-Rapid degradation of mammalian mRNAs is usually initiated by the binding of specific protein(s) to unique element(s) within the 3Ј-UTR (20,21). The RNA-binding protein(s) subsequently recruit a poly(A)specific ribonuclease (22) and the exosome (23) to remove the poly(A) tail and accomplish a rapid 3Ј 3 5Ј exonucleolytic degradation, respectively. RNase H treatment of the RNAs isolated from the ␤G-PCK-1 mRNA half-life analysis was performed to determine whether deadenylation precedes the decay of the PEPCK mRNA. Incubation of the RNA isolated immediately after the 3-h pulse (0-h sample) with a complementary oligonucleotide and oligo(dT), followed by treatment with RNase H, produced a 315-nt fragment (Fig. 2B). This fragment corresponded to the expected length of the fully deadenylated 3Ј-end of the ␤G-PCK-1 mRNA. Digestion of the same RNA sample in the presence of only the complementary oligonucleotide produced larger fragments containing ϳ500 nt. Thus, the newly synthesized ␤G-PCK-1 mRNA contained a significant poly(A) tail. Identical treatment of RNAs isolated from the later time points demonstrated that the ␤G-PCK-1 mRNA undergoes a rapid deadenylation that occurs concomitant with the decay of the mRNA.
Mapping of the Instability Elements within the 3Ј-UTR of PEPCK-Further half-life analyses were performed using cells grown in the absence of Dox to maximally induce expression of the chimeric PEPCK mRNAs. Selective decay of the reporter mRNA was subsequently initiated by the addition of 1 g/ml Dox. Northern blot analysis was performed using RNAs isolated from 8C cells that stably express the parent Tet␤G construct. This analysis determined that the control ␤G mRNA is extremely stable and decays with a half-life of ϳ5 days (Fig. 3). Northern analysis also indicated that the ␤G-PCK-1 mRNA was degraded with a half-life of 1.8 h, consistent with the previous pulse-chase analysis.
Previous RNA gel shift analysis indicates that a protein in cytosolic extracts of rat kidney cortex binds specifically to PCK-7, a 50-nt segment of the 3Ј-UTR of the PEPCK mRNA (12). To test whether the PCK-7 segment was sufficient to cause the rapid destabilization of the PEPCK mRNA, the turnover of the ␤G-PCK-7 mRNA was determined. Northern blot analyses revealed that the ␤G-PCK-7 mRNA was degraded with a half-life of 17 h (Fig. 3). Therefore, insertion of just the PCK-7 segment only partially destabilized the chimeric mRNA and was not sufficient to produce the rapid turnover observed with ␤G-PCK-1 mRNA. Thus, additional sequences within the 3Ј-UTR may be needed to constitute the complete instability element of the PEPCK mRNA.
Sequence analysis of the 3Ј-UTRs of human, rat, and mouse PEPCK mRNAs revealed the presence of a conserved 16-nt AU sequence that is located immediately upstream of the PCK-7 segment. This sequence is part of a 23-nt segment termed PCK-6 that binds a protein in a rat renal cytosolic extract with low affinity (12). Therefore, the PCK-6 segment was inserted into the parent Tet␤G vector to generate pT␤G-PCK-6. LLC- PK 1 -F ϩ cells were stably transfected with this construct, and half-life analyses were performed. The ␤G-PCK-6 mRNA decayed with a half-life of only 6 h (Fig. 3).
To determine whether the PCK-6 and PCK-7 segments act synergistically, cells that stably express pT␤G-PCK-6/7 were subjected to half-life analysis. Northern blot analysis demonstrated that the ␤G-PCK-6/7 mRNA decayed with a half-life of 3.6 h (Fig. 4). This half-life was significantly lower than that observed with either ␤G-PCK-7 or the ␤G-PCK-6 mRNA but was still greater than that observed with the full-length PCK-1 segment. Therefore, additional sequences within the 3Ј-UTR may be needed to constitute the complete instability element of the PEPCK mRNA. To test this hypothesis, cells expressing constructs containing longer segments of the 3Ј-UTR of the PEPCK mRNA were generated.
The PCK-3 segment contains 224 nt from the 3Ј-end of the PEPCK mRNA that includes the PCK-6/7 segment (Fig. 1). Half-life analysis performed using RNAs isolated from cells expressing the ␤G-PCK-3 mRNA established that this mRNA decayed with a half-life of 3.6 h, the same as observed with the ␤G-PCK-6/7 mRNA (Fig. 4). Thus, the PCK-6/7 segment contains all of the instability elements within the PCK-3 segment.
The half-life of the ␤G-PCK-2 mRNA that contains the 5Ј-end of the 3Ј-UTR of the PEPCK mRNA was also determined. Previous RNA gel shift assays with 32 P-labeled PCK-2 RNA and rat renal cytosolic extracts failed to demonstrate the formation of a RNA-protein complex (12). However, the ␤G-PCK-2 mRNA, when expressed in the porcine kidney cells, decayed with a half-life of 5.4 h (Fig. 4). Thus, the half-life analyses of the PCK-1, PCK-6, PCK-7, and PCK-2 constructs establish that the 3Ј-UTR of the PEPCK mRNA contains multiple instability elements that contribute to the rapid turnover that is observed with the full-length chimeric mRNA.
AUF1 Binds within the 3Ј-UTR of the PEPCK mRNA-UV cross-linking analysis was performed to identify the proteins within rat renal cytosolic extracts that bind within the PCK-6/7 segment. Proteins with apparent molecular mass values that range from 40 to 100 kDa were found to be associated with the PCK-6/7 RNA (Fig. 5A). Identical patterns of labeled proteins were observed when PCK-1 and PCK-7 were incubated with the renal cytosolic extract and subjected to UV cross-linking (data not shown). Antibodies to known RNA-binding proteins of this size and nucleotide specificity were used in gel supershift assays. Preincubation of the cytosolic extract with AUF1-specific antibodies blocked formation of the PCK-7 RNA-protein complexes (Fig. 5B). In contrast, antibodies against CP1 and CP2, two CU-binding proteins (24), or HuR, an AU-binding protein (25), had little effect on complex formation. A similar supershift was observed when the PCK-1, PCK-6, and PCK-6/7 RNAs were preincubated with anti-AUF1 antibodies (data not shown).
To characterize the potential AUF1 binding interactions, gel shift assays were performed using various PCK RNA segments. The recombinant 40-kDa isoform of AUF1 exhibits high affinity and specific binding to the PCK-1 RNA (Fig. 6). The recombinant AUF1 protein also forms identical complexes with the PCK-6 and PCK-7 RNAs. Interestingly, the 32 P-labeled PCK-2 RNA forms a weak complex with recombinant AUF1. However, AUF1 fails to form a complex with a 50-nt RNA transcribed from pBlueScript-SK(Ϫ) (data not shown).
A competition analysis was performed to demonstrate the specificity of AUF1 binding to the various PCK RNAs. The interaction observed with PCK-1 RNA was blocked by the addition of increasing amounts of unlabeled PCK-1 or PCK-7 RNAs but not by the addition of a 50-nt RNA transcribed from pBlueScript-SK(Ϫ) (Fig. 7A). A similar experiment demonstrated the specificity of AUF1 binding to the PCK-6/7 RNA. A 30 -500-fold excess of unlabeled PCK-6/7 RNAs (but not an unrelated RNA) effectively competed the binding (Fig. 7B). Although AUF1 binds to the PCK-2 RNA weakly, the observed binding is also highly specific (Fig. 7C). Therefore, AUF1 exhibits specific binding to each segment of the PEPCK 3Ј-UTR that contains an instability element.
Mutational Analyses-Mutational analyses were performed to identify the specific sequences within the PCK-6/7 RNA that bind AUF1. PCK-6 contains a UUAUUUUAU sequence (Fig.  8A) that is similar to the 9-nt AU-rich element in tumor necro-

FIG. 5. Identification of AUF1 as a protein that binds to the instability elements within the 3-UTR of the PEPCK mRNA.
A, a rat renal cytosolic (RRC) extract was incubated in the absence and presence of 32 P-labeled PCK-6/7 RNA. The samples were subjected to UV radiation and then separated by 10% SDS-PAGE. B, the RRC extract was preincubated in the absence or presence of the indicated antibodies. 32 P-labeled PCK-7 RNA was subsequently added, and the resulting complexes were resolved on a native polyacrylamide gel. The two gels were imaged on a PhosphorImager. sis factor-␣ mRNA that binds AUF1 (26). When this sequence in PCK-6 RNA was mutated, by introducing multiple GC nucleotides, binding of recombinant AUF1 was abolished (Fig. 9). PCK-7 contains a direct repeat of an 8-nt CU sequence separated by a single nucleotide, a potential secondary structure containing an 8-nt stem and an 11-nt loop, and a 22-nt stretch in which 18 of the nucleotides are A and U residues. To further map the AUF1 binding site, the PCK-7 RNA was divided into three subsegments: PCK-8 (which contains the CU repeats), PCK-9 RNA (which consists of the AU-rich region), and PCK-10 RNA (which forms the potential stem-loop structure) (Fig. 8B). When incubated with a rat renal cytosolic extract, the PCK-8 and PCK-10 RNAs (but not the PCK-9 RNA) form a complex (data not shown). Similarly, the recombinant AUF1 binds to only the PCK-8 and PCK-10 RNAs (Fig 10A). Therefore, the CU-rich region and/or the potential stem-loop structure might function as AUF1 binding sites. Three mutations of the PCK-7 RNA, termed PCK-7 mut-1, PCK-7 mut-2, and PCK-7 mut-3, were designed to test binding to the individual or combined CU repeats (Fig. 8C). A fourth mutant, PCK-7 mut-4, was designed to discriminate binding to the conserved sequence that constitutes the loop of the putative stem-loop structure. Recombinant AUF1 binds to PCK-7 RNA but not to any of the mutated constructs (Fig. 10B). Therefore, within the PCK-7 RNA, AUF1 binds to the CU repeats, to the conserved loop sequence, or to both elements. DISCUSSION Previous experiments using DRB, a specific inhibitor of RNA polymerase II, to measure the half-life of the single ␤G-PCK-1 mRNA in LLC-PK 1 -F ϩ cells demonstrated that the 3Ј-UTR of the PEPCK mRNA contains a cis-acting destabilizing element (11). In the current study, a Tet-regulated promoter system was used to quantify the in vivo decay rates and to map the location of multiple instability elements within various chimeric ␤G-PCK mRNAs. Using this approach, the half-life observed for the ␤G-PCK-1 mRNA was significantly less than that measured in the previous experiments. Because DRB inhibits transcription of all polymerase II genes, it may block the continued synthesis of proteins that function in the normal process of mRNA turnover. In contrast, with the Tet-regulated system, the transcription of the gene of interest is selectively inhibited by the addition of a subtoxic level of doxycycline. Therefore, the observed half-life of the ␤G-PCK-1 mRNA of 1.2-1.8 h is a more reliable measure of its turnover in LLC-PK 1 -F ϩ cells. The analysis performed with multiple ␤G-PCK constructs indicates that the PEPCK mRNA contains multiple instability elements that are located within the PCK-2, PCK-6, and PCK-7 segments of the 3Ј-UTR. These findings are significant, because the physiologically important adaptations in the levels of PEPCK mRNA that occur in response to various hormones and to changes in acid-base balance are made feasible by the rapid decay of the PEPCK mRNA.
The decay of mammalian mRNAs that contain AU-rich elements is frequently preceded by the removal of the poly(A) tail from the 3Ј-end of the mRNA (27). Deadenylation can proceed in either a synchronous or an asynchronous manner. Synchronous deadenylation results in the nonprocessive formation of a shortened poly(A) tail containing 30 -60 nucleotides. With the FIG. 6. Recombinant AUF1 binds to 32 P-labeled PCK-1, PCK-7, PCK-6, and PCK-2 RNAs. The indicated nanogram (ng) amounts of recombinant p40 isoform of human AUF1 were incubated with the 32 P-labeled RNAs. The samples were treated with RNase T1 to digest the unbound probe, and the resulting RNA-protein complexes were resolved on a native polyacrylamide gel and imaged on a PhosphorImager. The arrows designate the resulting RNA-protein complexes.

FIG. 7. Competition analysis of AUF1 binding to various PCK RNAs.
A, the 32 P-labeled PCK-1 RNA was incubated in the absence or presence of recombinant AUF1. The samples in lanes 3-10 also contained the indicated -fold excess (30-, 200-, or 500-fold) of an unlabeled competitor RNA. pBS is a nonspecific 50-nt RNA that was transcribed from pBlueScript-SK(Ϫ). The samples were treated with RNase T1 to digest the unbound probe, and the remaining complexes were resolved on a native polyacrylamide gel and imaged on a PhosphorImager. The analysis was repeated using 32 P-labeled PCK-6/7 (B) and PCK-2 RNAs (C). c-fos mRNA (27), this process is completed before the decay of the body of the mRNA is initiated. In contrast, asynchronous deadenylation of the granulocyte/macrophage colony-stimulating factor mRNA (27) results in the processive formation of a fully deadenylated mRNA that is then rapidly degraded. The PEPCK mRNA lacks a canonical AUUUA or UUAUUUAUU element but contains a single related sequence (UUAUUUUAU) within the PCK-6 segment of its 3Ј-UTR. RNase H assays indicated that the ␤G-PCK-1 mRNA undergoes a rapid and synchronous deadenylation that is complete within 2 h (Fig. 2B). However, ϳ50% of the newly synthesized ␤G-PCK-1 mRNA was already degraded within 2 h. Therefore, the degradation of the ␤G-PCK-1 mRNA may proceed by a combined mechanism in which a portion of the chimeric mRNA undergoes a rapid processive deadenylation that is complete within 30 min, whereas the remainder undergoes a slower (but synchronous) deadenylation.
The clonal lines expressing the remaining ␤G-PCK mRNAs produced only a low level of the chimeric mRNA during a 3-h transcriptional pulse. As a result, it was difficult to accurately quantify the decay of the other mRNAs using the pulse-chase protocol. Therefore, the half-lives of the various ␤G-PCK mRNAs containing shortened segments of the 3Ј-UTR of the PEPCK mRNA were determined by adding doxycycline to cells that were grown in the absence of doxycycline. A potential problem with this approach can occur when high levels of constitutive expression from the Tet-regulated promoter saturate the cellular decay machinery and cause an inefficient decay of the mRNA (15). However, constitutive expression of the ␤G-PCK-1 mRNA exceeded that of the other mRNAs, and the half-lives of the ␤G-PCK-1 mRNA determined by the two protocols were nearly identical. Hence, the half-lives obtained for the different ␤G-PCK constructs should accurately reflect the actual decay rates of the mRNAs. Without the pulse-chase protocol, it was not possible to characterize the mechanism of deadenylation of the various deletion constructs of the ␤G-PCK-1 mRNA.
Previous studies have reported that cytosolic extracts of rat hepatoma cells (6) and hepatocytes (28) contain proteins that bind to multiple segments of the 3Ј-UTR of PEPCK mRNA. Although the observed binding interactions were affected by FIG. 8. Nucleotide sequence of the mutated and deleted RNAs derived from the PCK-6 and PCK-7 segments. A, within the 22-nt PCK-6 segment, the nine AU nucleotides (uppercase letters) were mutated to produce the mut-6 RNA. B, the PCK-8, PCK-9, and PCK-10 RNAs correspond to portions of the PCK-7 segments that contain the CU repeats, the AU-rich region, and the stem-loop structure, respectively. C, the uppercase letters indicate nucleotides that were mutated within PCK-7 to produce the Mut-1, Mut-2, Mut-3, and Mut-4 RNAs. The mutated segments selectively alter the first, second, both CU regions, and the sequence that constitutes the loop of PCK-7, respectively.
FIG. 9. Binding of recombinant AUF1 to 32 P-labeled PCK-6 and Mut-6 RNAs. The 32 P-labeled PCK-6 and Mut-6 RNAs were incubated in the absence and presence of AUF1. The samples were treated with RNase T1 to digest the unbound probe, resolved on a native polyacrylamide gel, and imaged on a PhosphorImager. pretreatment of the cells with cAMP and insulin, neither study establishes the specificity of the observed interactions nor identifies the binding elements involved. The subsequent use of affinity chromatography identifies ferritin light chain as the protein in rat liver extracts that binds to the 3Ј-UTR of PEPCK mRNA (28). However, the physiological significance of this interaction remains to be confirmed. In the current study, AUF1, a known RNA-binding protein, was shown to bind with high affinity and specificity to PCK-2, PCK-6, and PCK-7, the same segments that contribute to the rapid turnover of the PEPCK mRNA.
AUF1 binds to the 3Ј-UTRs of many unstable RNAs, including c-myc (29), tumor necrosis factor-␣ (30), GM-CSF (31), and COX-2 (32). AUF1 was also identified as hnRNP D (33), a protein that shuttles between the nucleus and the cytoplasm (34). Four isoforms of AUF1 (p37, p40, p42, and p45) are produced by alternative splicing of the initial AUF1 transcript (33). The p40 AUF1 isoform is predominantly cytoplasmic and, hence, was utilized in the reported experiments. AUF1 has also been shown to stabilize interleukin-3 RNA in NIH 3T3 cells (35) and destabilize it in K562 cells (36). Thus, its effect on mRNA stability depends on the target AU-rich element and the auxiliary proteins expressed in a particular cell line (37).
Recruitment of ancillary factors by AUF1 on the target RNA is essential for the promotion of AU-rich element-mediated mRNA decay. Yeast two-hybrid (38) and coimmunoprecipitation experiments (39) have shown that AUF1 interacts with a number of RNA-binding proteins, including NSAP-1, NSEP-1, and IMP-2. Interestingly, NSEP-1 exhibits an endoribonuclease activity in vitro, suggesting a possible role of these proteins in AUF1-mediated decay (38). Other proteins, such as the poly(A)-binding protein, the translation initiation factor eIF4G, the heat shock protein Hsp70, and the cognate heat shock protein Hsc70, also physically interact with AUF1 (40) and may regulate its activity. Thus, it will be interesting to determine whether any of these proteins are also involved in decay of the PEPCK mRNA.
The high affinity protein binding observed with rat renal cytoplasmic extracts maps to the PCK-7 segment of the 3Ј-UTR of the PEPCK mRNA (12). Thus, it was surprising to find that the PCK-7 segment did not produce a strong destabilizing effect compared with the full-length 3Ј-UTR. Inclusion of the adjacent PCK-6 segment, which contains a UUAUUUUAU motif, resulted in a decay rate nearly comparable with that observed with the full-length 3Ј-UTR. Thus, binding of AUF1 at both sites may be necessary to recruit the proteins that mediate the rapid decay of the PEPCK mRNA. This hypothesis is supported by UV cross-linking analysis of the full-length 3Ј-UTR incubated with rat renal cytoplasmic extracts. This analysis revealed the presence of multiple RNA-protein complexes (Fig.  5A), suggesting that other renal proteins might be binding to this RNA. However, supershift analyses using antibodies against known RNA-binding proteins, including HuR, tristetraprolin, and Hsp70 (data not shown), failed to detect these proteins in the PCK-1 RNA-protein complexes. Identification of additional binding proteins involved in PEPCK mRNA turnover is currently under investigation.
Previous studies using a rat renal cytoplasmic extract failed to detect a complex with the PCK-2 RNA that constitutes the 5Ј-end of the 3Ј-UTR of PEPCK mRNA (12). However, this segment also binds recombinant AUF1 (Fig. 6) and exhibits a significant destabilizing effect when incorporated into the ␤-globin reporter mRNA. Sequence analysis revealed the presence of a stretch of alternating purine and pyrimidine nucleotides near the 5Ј-end of PCK-2. This region also contains a sequence that is highly conserved in the mouse, rat, and human PEPCK mRNAs (Fig. 1). AUF1 might bind within this conserved sequence. However, the affinity of the AUF1 binding observed with this segment is weak compared with that observed with the PCK-6/7 region. Thus, some other protein may bind to this segment and participate in the turnover of the PEPCK mRNA.
Sequence analysis of the mouse, rat, and human PEPCK mRNAs identified two fully conserved AU-rich sequences within the PCK-6/7 segments of their 3Ј-UTRs (Fig. 1). The first is a UUAAAUUAUUUUUAUACAC sequence that is located in the PCK-6 segment. Mutational analysis indicates that AUF1 binds to the AU-rich core within this conserved sequence (Fig. 9). The second is a CUUUACAUAAUUG sequence that is found within the PCK-7 segment. The latter sequence includes the entire loop of a stable stem-loop structure. The observation that AUF1 does not bind to the PCK-7 Mut-4 RNA suggests that it also recognizes the highly conserved sequence that constitutes the loop of the predicted stable stem-loop structure. The sequence of the PCK-7 Mut-4 RNA was carefully designed using RNAdraw software to ensure that the mutations introduced in the loop sequence would neither alter the ability to form the predicted stem-loop structure nor promote the formation of an alternative structure. The additional PCK-7 mutations also failed to form a complex with AUF1. The PCK-7 Mut-2 and -3 constructs contained muta-FIG. 10. Binding of recombinant AUF1 to various deleted and mutated forms of PCK-7 RNA. A, 32 P-labeled PCK-7 RNA, the deleted segments of PCK-7 RNA (PCK-8, PCK-9, and PCK-10), and an unrelated RNA transcribed from pBlueScript-SK(Ϫ) (pBS) were incubated in the absence (Ϫ) or presence (ϩ) of recombinant AUF1. B, 32 P-labeled PCK-7 RNA and the mutated segments of PCK-7 RNA (Mut-1, Mut-2, Mut-3, and Mut-4) were incubated with recombinant AUF1. All of the samples were treated with RNase T1 to digest the unbound probe, and the remaining complexes were resolved on a native polyacrylamide gel and imaged on a PhosphorImager.