Role of deadenylation and AUF1 binding in the pH-responsive stabilization of glutaminase mRNA
http://www.100md.com
《美国生理学杂志》
Department of Biochemistry and Molecular Biology, Colorado State University, Fort Collins, Colorado
ABSTRACT
During chronic metabolic acidosis, increased expression of renal glutaminase (GA) results from selective stabilization of the GA mRNA. This response is mediated by a direct repeat of an 8-base adenylate-uridylate (AU) sequence that binds -crystallin and functions as a pH response element (pH-RE). A tetracycline-responsive promoter system was developed in LLC-PK1-F+ cells to perform pulse-chase analysis of the turnover of a chimeric -globin (G) mRNA that contains 960 bp of the 3'-UTR of GA mRNA including the pH-RE. The G-GA mRNA exhibits a 14-fold increase in half-life when the LLC-PK1-F+ cells are transferred to acidic medium. RNase H cleavage and Northern blot analysis of the 3'-ends established that rapid deadenylation occurred concomitantly with the rapid decay of the G-GA mRNA in cells grown in normal medium. Stabilization of the G-GA mRNA in acidic medium is associated with a pronounced decrease in the rate of deadenylation. Mutation of the pH-RE within the G-GA mRNA blocked the pH-responsive stabilization, but not the rapid decay, whereas insertion of only a 29-bp segment containing the pH-RE was sufficient to produce both a rapid decay and a pH-responsive stabilization. Various kidney cells express multiple isoforms of AUF1, an AU-binding protein that enhances mRNA turnover. RNA gel-shift assays demonstrated that the recombinant p40 isoform of AUF1 binds to the pH-RE with high affinity and specificity. Thus AUF1 may mediate the rapid turnover of the GA mRNA, whereas increased binding of -crystallin during acidosis may inhibit degradation and result in selective stabilization.
renal ammoniagenesis; metabolic acidosis; posttranscriptional regulation
AS PART OF THE ADAPTIVE RESPONSE to metabolic acidosis, renal extraction and catabolism of plasma glutamine are rapidly increased (3). This adaptation results in a large increase in the renal synthesis of ammonium and bicarbonate ions. The ammonium ions are preferentially excreted in the urine where they provide an expendable cation that facilitates the excretion of acidic anions. In contrast, the bicarbonate is added to the renal venous blood to partially compensate for the systemic acidosis. Renal catabolism of glutamine is initiated by a phosphate-activated mitochondrial glutaminase (5). During chronic acidosis, increased catabolism of glutamine is sustained, in part, by an 8- to 20-fold increase in the level of the mitochondrial glutaminase that occurs solely within the proximal convoluted tubule (4, 36).
During acidosis, the increased synthesis of the renal glutaminase (GA) protein (29) occurs without increasing the rate of transcription of the GA mRNA (14, 15). Experiments that characterized the half-life of a chimeric -globin-glutaminase (G-GA) mRNA in LLC-PK1-F+ cells, a pH-responsive line of porcine proximal tubule-like cells (11), established that the 3'-untranslated region (3'-UTR) of the GA mRNA contains a pH-responsive instability element (12). RNA gel-shift analysis demonstrated the presence of a protein in rat renal cortical cytosolic extracts that binds to the 3'-UTR of the GA mRNA with high affinity and specificity (20). This binding interaction was mapped to a direct repeat of an 8-base AU sequence. Functional studies established that this sequence was both necessary and sufficient to impart a pH-responsive stabilization to a chimeric G mRNA (21). Subsequent studies identified -crystallin/NADPH quinone reductase as the primary protein in rat renal extracts that binds to the pH response element (pH-RE) (28).
The rapid degradation of mammalian mRNAs is frequently initiated by the binding of proteins (25) that recruit a poly(A)-specific ribonuclease (9) and the exosome (30) to remove the poly(A) tail and accomplish a rapid 3'5' exonucleolytic degradation, respectively. In the current study, a tetracycline-responsive promoter (23) was used to perform a pulse-chase analysis of the decay of newly synthesized G-GA mRNA. The resulting data demonstrate that the pH-responsive stabilization of the G-GA mRNA is associated with a proportional decrease in the rate of deadenylation. Additional studies established that recombinant AUF1, an AU-binding protein that enhances mRNA turnover (32), also binds to the pH-RE with high affinity and specificity. Therefore, the enhanced binding of -crystallin to this site during acidosis (18) may block AUF1 binding and prevent the AUF1-mediated deadenylation and degradation of the GA mRNA.
MATERIALS AND METHODS
Materials. Male Sprague-Dawley rats were purchased from Charles River. [-32P]dCTP (specific activity 3,000 Ci/mmol) was purchased from MP Biomedicals or Amersham Pharmacia Biotechnology. The animal protocols used in this study were received and approved by The University Animal Care and Use Committee as protocol 93-25DA-12. The oligolabeling kit was from Ambion. Restriction enzymes, RNase T1, T7 RNA polymerase, and yeast tRNA were acquired from Roche, New England Biolabs and MBI Fermentas. RNase H was purchased from Epicentre Technologies. GENECLEAN kits were obtained from Bio101. 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 Promega. DMEM/F12 medium, Geneticin (G418), and hygromycin B were products of Sigma and Mediatech. A polyclonal antibody vs. AUF1 was obtained from Upstate. An expression plasmid that encodes the p40 isoform of AUF1 was obtained from Dr. J. Wilusz. All other biochemicals were acquired from Sigma.
Synthesis of pTRE2 constructs. pTRE2 (Clontech) contains a tetracycline-responsive element (TRE) that includes seven direct repeats of a 42-bp sequence containing the tet operator positioned upstream of a minimal CMV promoter, a multicloning site, and the 3'-coding and 3'-untranslated regions of the rabbit -globin gene. pTRE2 was digested with MluI and SapI to remove the -globin sequence. A segment of pG-GA (12) containing the -globin coding region, 960 bp of the GA 3'-UTR, and a segment of the bovine growth hormone (bGH) 3'-untranslated region containing the polyadenylation site was PCR amplified with primers containing unique MluI and SapI restriction sites. This fragment was cloned into pCR-Script SK(+) at the SrfI site, amplified, and then ligated into the pTRE2 fragment to create pTRE2-G-GA. The pG and pG-GA(R2-I) plasmids (27) were digested with CelII and XbaI and the isolated fragments were inserted into pTRE2-G-GA that had been digested with CelII and XbaI. This procedure removed the GA 3'-UTR or replaced it with a 29-bp sequence containing the pH response element, respectively. The pG-mGA plasmid (19) was digested with NheI and NotI and the isolated fragment was inserted into pTRE2-G-GA that had been digested with NheI and NotI to produce a construct containing the GA 3'-untranslated region in which the pH response element was mutated to include GC bases.
Selection of stable cell lines. LLC-PK1-F+ cells (11) were grown in a 50:50 mixture of DMEM and Ham’s F-12 medium containing 5 mM glucose and 10% fetal bovine serum at 37°C in a 5% CO2 atmosphere. Normal medium (pH 7.4) contained 25 mM sodium bicarbonate, whereas acidic medium (pH 6.9) contained 10 mM sodium bicarbonate supplemented with 15 mM sodium chloride to maintain an equivalent osmolarity and sodium ion concentration. Clonal lines of LLC-PK1-F+ cells that stably express the tTA protein were produced by calcium phosphate transfection (1) of 3-day postsplit cells with the pTet-off plasmid and selection with medium containing 0.8 mg/ml G418. After 14–21 days, colonies were selected and expanded in medium containing 0.2 mg/ml G418. The expanded cell lines were grown for 3 days on six-well plates using medium containing 1 μg/ml doxycycline (Dox) and then transiently transfected with pTRE2-Luc. Approximately 16 h later, the medium was replaced with fresh medium with or without 1 μg/ml Dox and the cells were cultured for an additional 32 h. Cell extracts were prepared and assayed using the reagents contained in the Luciferase Assay System (Promega). The firefly luciferase activities obtained for the various stable Tet-Off clonal lines were standardized per microgram of protein (24). The fold-induction of the TRE2-Luc gene was calculated as the mean of the ratio of the specific luciferase activities measured in triplicate cultures grown in the absence and presence of Dox.
Clonal cell lines that stably express the various G-GA constructs were prepared by cotransfection of LLC-PK1-F+ cells that express the tTA protein with the appropriate TRE2 plasmid and pcDNA3.1/Hygro (Invitrogen). At 48 h posttransfection, selection medium containing 0.2 mg/ml G418 and 0.8 mg/ml hygromycin was added. After 14–21 days, individual colonies were isolated and expanded. The cell lines were tested for responsiveness by growing the cells for 48 h in the presence or absence of 0.5 μg/ml Dox, a tetracycline analog. Total RNA was harvested and analyzed by Northern blotting.
Half-life analysis. For pulse-chase analysis, the stably transfected LLC-PK1-F+ cells were split 1:10 and grown for 5–7 days in normal medium containing 0.2 mg/ml G418 and 0.2 mg/ml hygromycin B. The cells were then grown for 48 h in pH 7.4 or 6.9 medium containing antibiotics and 25 ng/ml Dox. The cells were washed two times with phosphate-buffered saline and incubated in normal or acidic medium with antibiotics but without Dox for 3 h to create a transcriptional pulse. Then, fresh medium containing antibiotics and 1 μg/ml Dox was added and total RNA was isolated at 0, 3, 6, and 9 h using the TRIzol reagent (Invitrogen). Control cells were grown for 48 h in the presence or absence of Dox. Alternatively, the stably transfected LLC-PK1-F+ cells were grown for 5–7 days and then maintained in normal or acidic medium minus Dox for 24 h. Subsequently, 1 μg/ml Dox was added to each plate and total RNA was isolated at 0, 3, 6, and 9 h post-Dox addition. Control plates containing no Dox were harvested at 0 and 9 h.
Northern blot analysis. A 507-bp fragment of rabbit -globin cDNA was excised by restricting pRSV-G (10) with HindIII and BglII. A 2.0-kb fragment of the 18 S ribosomal RNA cDNA from Acanthamoeba castellanii was obtained by restricting pAr2 (6) with HindIII and EcoRI. A 228-bp bGH fragment was excised from pCRScript-bGH with SphI. The fragments were separated on 1% agarose gels, excised, and purified using the GENECLEAN kit. The synthesis of oligolabeled cDNA probes and Northern blot analysis was performed as described previously (27). The blots were exposed to a PhosphorImager screen and quantified using Molecular Dynamics Software. The level of the chimeric -globin mRNA was divided by the level of the corresponding 18S rRNA to correct for errors in sample loading. The log of normalized data was then plotted vs. the time of addition of Dox.
RNase H treatment. A 26-nt deoxyoligonucleotide (GA oligo), which is complementary to the 3'-end of the GA sequence encoded by pTRE2-G-GA (bp 2514–2539 of the GA cDNA), was synthesized by Macromolecular Resources (Fort Collins, CO). The sequence of the GA oligo is 5'-GACCGACAGCTACACACCACAGAGAG-3'. A 30-mer of oligo dT deoxyoligonucleotide was used to degrade the poly(A) tail of the mRNA. RNA samples, which were stored in Formazol, were precipitated with 4 vol of ethanol and a final concentration of 0.2 M sodium chloride and resuspended in 25 μl of DEPC-treated water. RNase H treatment was performed as described previously (26). Briefly, 10 μg RNA, 1.0 μl containing 50 pmol of GA oligo, and 1.0 μl containing 50 pmol of oligo-dT (if needed) in 12.5 μl were incubated at 55°C for 1 min. Next, 1.5 μl of 10x RNase H buffer (0.5 M Tris, pH 7.4, 1.0 M NaCl, 20 mM MgCl2, and 10 mM dithiothreitol) and 1.0 μl of freshly diluted RNase H (1 U/μl) were added and the sample was incubated at 62°C for 15 min. Then, 25 μl of denaturing buffer (20 μl Formazol, 3 μl 10x MOPS, and 6 μl 37% formaldehyde solution) were added and the mixture was incubated at 55°C for 5 min and then transferred to ice. RNase H-treated RNAs were mixed with 10 μl of an RNase H gel loading buffer (50% glycerol, 1 mM EDTA, 0.25% bromophenol blue, 0.25% xylene cyanol, and 100 μg/ml ethidium bromide) and subjected to Northern blot analysis using a 1.2% agarose gel. The blot was hybridized with the bGH probe.
Western blotting. Cytosolic extracts of rat renal cortex and of various cell lines were prepared as described previously (19). The samples were separated by electrophoresis on a 10% SDS-polyacrylamide gel, transferred to a nitrocellulose membrane, and incubated with a 1:1,000 dilution of AUF1 antibody. The membrane was then incubated with a 1:15,000 dilution of horseradish peroxidase (HRP)-conjugated secondary antibody. Images were developed with Dura reagent (Pierce Chemicals) and visualized by brief exposure to X-ray film.
RNA EMSA. 32P-labeled and -unlabeled GA(R2-I) and mutated forms of GA(R2-I) RNAs were synthesized as described previously (19). An aliquot containing 1–8 μg of rat renal cortical protein or 12.5–100 ng of recombinant AUF1 was preincubated for 10 min at room temperature in binding buffer containing 0.5% Nonidet P-40, 1 mM dithiothreiotol, 2 μg yeast t-RNA, 40 U of RNAsin, 10% glycerol, and 10–70 fmol of 32P-labeled RNA. The indicated excess of unlabeled RNA was added, and the reaction mixture was incubated at room temperature for 20 min. The samples were then loaded and subjected to electrophoresis for 2 h at 10 mA using a 90 mM Tris, 110 mM boric acid, and 2 mM EDTA running buffer. Gels were dried and exposed overnight to a PhosphorImager screen. In the antibody supershift experiments, the anti-AUF1 antibody was preincubated with the rat renal cytosolic extract for 10 min at room temperature and then for 1 h on ice before the binding buffer was added.
Statistical analyses were performed using a paired Student’s t-test.
RESULTS
Effect of medium pH on deadenylation of G-GA mRNA. Clonal cell lines that stably express the tTA protein, a tetracycline-responsive transcription factor, were isolated by transfection of LLC-PK1-F+ cells with pTet-off and selection with G418. Their Dox responsiveness was tested by transient expression of pTRE2-Luc in cells that were maintained in the presence or absence of Dox. The 8C line of LLC-PK1-F+ cells showed the greatest induction of luciferase activity. With the use of 0.3 μg of pTRE2-Luc DNA, they produced a 200-fold difference in luciferase activity when grown in the presence and absence of 1 μg/ml Dox. Therefore, the 8C line of LLC-PK1-F+ cells was used for the subsequent experiments. Clonal cell lines that stably express pTRE2-G-GA were isolated by cotransfection of the 8C cells with pcDNA3.1/Hygro and selection with hygromycin. The isolated cell lines were tested for Dox-responsive G-GA expression by culturing the cells in the presence or absence of Dox and quantifying G-GA mRNA expression by Northern blot analysis. Cell lines were evaluated based on the level of G-GA mRNA expressed in the presence of Dox and the increased expression produced by removal of Dox. The selected 8C-2A cell line exhibited a 50-fold difference in the levels of G-GA mRNA when grown in the presence or absence of 0.5 μg/ml Dox (Fig. 1).
Previous experiments characterizing the half-lives of various G-GA chimeric mRNAs used the Rous sarcoma virus long-terminal repeat (RSV) promoter to drive transcription (20). The level of G-GA mRNA transcribed from the tetracycline-responsive promoter in the absence of Dox was far greater than the level of mRNA synthesized from the RSV promoter (Fig. 1). To create a level of G-GA mRNA similar to that produced with the RSV promoter, transcription was initially inhibited by adding 25 ng/ml Dox for 48 h. Dox was then removed from the selected cells to create a pulse of transcription and RNA was isolated at various times. The level of G-GA mRNA produced by a 3-h transcriptional pulse was comparable to the level of G-GA mRNA produced by the RSV promoter (Fig. 1). Therefore, a 3-h transcriptional pulse should provide a level of newly synthesized and adenylated G-GA mRNA that is sufficient to follow decay and deadenylation.
To perform a pulse-chase analysis, the selected cells were maintained in normal or acidic medium containing 25 ng/ml Dox for 48 h. A 3-h pulse was created by removing Dox and then chased for 0, 3, 6, and 9 h following addition of 1 μg/ml Dox. The half-life of G-GA mRNA in cells cultured in normal medium was 2.9 h. However, when the cells were maintained in acidic conditions, the G-GA mRNA was significantly stabilized and now decayed with a half-life of 40 h (Fig. 2). An oligonucleotide (GA oligo), which specifically hybridizes to the 3'-end of the GA sequence, was used along with RNase H to cleave the G-GA mRNA and to produce a short 3'-fragment (Poly A+) that contains the poly(A) tail (Fig. 3A). Oligo(dT) was also added to an aliquot of the zero-time RNA sample to remove the poly(A) tail from the G-GA mRNA, creating a 330-nt deadenylated RNA fragment (Poly A–). The 3'-fragments were detected by Northern blot analysis using a 32P-labeled cDNA probe that specifically hybridizes to the bGH sequence within the 3'-end of the G-GA mRNA.
The adenylated and the fully deadenylated RNA fragments were easily resolved, making it feasible to monitor the extent of deadenylation of the G-GA mRNA. The 3'-ends of the G-GA mRNAs harvested from cells maintained in normal medium appear to exhibit a synchronous or nonprocessive pattern of deadenylation (Fig. 3B). However, the 3'-fragments containing the shortened poly(A) tail decay at a rate similar to that observed with the full-length mRNA. If complete removal of the poly(A) is required to initiate decay, then a portion of the G-GA mRNA must undergo a more rapid and processive deadenylation to form a transient, but fully deadenylated, intermediate that is rapidly degraded and does not accumulate. As a result, the synchronous deadenylation of the G-GA mRNAs would appear to occur concomitantly with the decay of the body of the chimeric mRNA. When the cells were treated with acidic medium, the rates of synchronous deadenylation (formation of shorter 3'-fragments) and processive deadenylation (decay of 3'-fragments) of the G-GA mRNA were significantly inhibited. Therefore, the pH-responsive stabilization correlates with a decrease in both the rate and the extent of deadenylation of the G-GA mRNA.
Functional analysis of the pH-RE. As reported previously (20), the chimeric -globin mRNA lacking sequence from the GA 3'-UTR is extremely stable. However, the insertion of a 29-bp segment containing the pH-RE is sufficient to create a labile mRNA that decayed with a half-life of 3.0 h in LLC-PK1-F+ cells that were maintained in normal medium (Fig. 4). When transferred to acidic medium, the G-GA(R2-I) mRNA decayed with a half-life of 17 h. Thus this segment functions both as an instability element and as a pH-RE. In contrast, the G-mGA mRNA, which contains the GA 3'-UTR in which the AU-rich pH-RE was mutated to the Mut3 sequence shown in Fig. 8A, is degraded rapidly in cells that are grown in either normal or acidic medium (Fig. 5). Thus the 3'-UTR must contain multiple instability elements that contribute to the rapid turnover of the GA mRNA.
Binding of AUF1 to the pH-RE. AUF1 binds to various AU-rich elements (AREs) and enhances the mRNA degradation (32). Previous studies (19, 27) demonstrated that GA(R2-I), a 29-bp RNA segment that contains the pH-RE and surrounding sequence within the GA mRNA, forms a specific complex when incubated with a cytosolic extract of rat kidney cortex. Preincubation of the cytosolic extract with AUF1-specific antibodies results in the reproducible formation of a complex that exhibits a slightly slower electrophoretic mobility (Fig. 6A). No shift was observed when the cytosolic extract was preincubated with an equivalent amount of an unrelated antibody (data not shown). The observed shift suggests that AUF1 may bind to the GA(R2-I) RNA. Western blot analysis was performed to determine whether AUF1 is expressed in rat kidney cortex and in various proximal tubule cell lines (Fig. 6B). Jurkat cell extracts were used as a positive control (17). They express the four isoforms of AUF1 (p37, p40, p42, and p45) that are produced by alternative splicing of the initial AUF1 transcript (31). Rat kidney cortex contains significant levels of the p37, p40, and p42 isoforms of AUF1, whereas the LLC-PK1-F+ and WKPT (13) proximal tubule cells express differing amounts of the four isoforms.
The p40 AUF1 isoform is predominantly cytoplasmic (37) and hence was utilized in the following binding experiments. Increasing amounts of the purified recombinant p40 AUF1 form multiple complexes when incubated with the GA(R2-I) RNA (Fig. 7A). This pattern is characteristic of the fact that AUF1 forms dimers and larger oligomers on binding to AU-rich sequences (35). The observed binding of AUF1 is competed by increasing concentrations of unlabeled GA(R2-I) RNA, but not by a nonspecific RNA segment (pBSSK) that was transcribed from the multicloning site of pBluescript-SK(–) (Fig. 7B). Thus the observed binding interaction is specific. The complexes formed between AUF1 and the GA(R2-I) RNA are similar to those observed when AUF1 is incubated with a short RNA that contains the canonical AUUUA element (Fig. 8). AUF1 also binds to GA(R2-I) segments in which one of the two AU sequences is mutated to contain GC-nucleotides (Mut1 and Mut2). However, AUF1 does not bind to an GA(R2-I) segment in which both AU sequences are mutated (Mut3).
Scatchard analysis was performed to determine the dissociation constant (Kd) for AUF1 binding to the GA(R2-I) RNA. When 80 ng of recombinant AUF1 were incubated with 70 fmol of 32P-labeled R-2I RNA, approximately one-third of the labeled probe formed a complex (Fig. 9A). The intensity of the shifted band decreased as the amount of unlabeled R-2I RNA was increased. The relative intensities of the bands that correspond to the bound and free probe were quantified by PhosphorImager analysis. These data along with the knowledge of the concentration of total probe in each sample were then used to calculate the concentrations of bound and free probe. The resulting values were plotted in the form of a Scatchard plot (Fig. 9B). Linear regression analysis of the plotted data indicated that the recombinant AUF1 binds the GA(R2-I) RNA with a Kd of 2 μM.
DISCUSSION
A tetracycline-responsive expression (Tet-off) system (23) was used to characterize the potential role of deadenylation in the degradation of the GA mRNA. With this system, it is feasible to inhibit the synthesis of a single mRNA through the addition of a nontoxic level of Dox, instead of using 5–6 dichloro-1--ribofuranosylbenzimidazole (DRB), a general inhibitor of pol II transcription. Thus this approach avoids potential nonspecific effects that may be caused by the rapid decay of regulatory proteins or nucleases that participate in mRNA turnover. As a result, this protocol should also provide a more accurate quantification of the half-life of a mRNA.
When fully induced, the tetracycline-responsive promoter produced a significantly greater level of G-GA mRNA than observed in previous experiments that used an RSV promoter to drive the synthesis of the reporter mRNA (12, 19). To avoid expression of very high levels of G-GA mRNA that might exceed the amount of -crystallin and/or other factors needed to accomplish a pH-responsive stabilization, a pulse-chase protocol was used to characterize the turnover of the G-GA mRNA. With the use of this protocol, the levels of newly synthesized G-GA mRNA were similar to those produced constitutively from the RSV promoter. None of the cell lines selected for stable expression of the G-GA(R2-I) or G-mGA mRNA produced the same rapid and pronounced increase, during a 3-h pulse, as observed with the 8C-2A line that expresses the G-GA mRNA. As a result, it was not feasible to perform pulse-chase analysis with the truncated or mutated construct. However, the fully induced levels of the G-GA(R2-I) and G-mGA mRNAs were similar to that produced following a 3-h pulse of the G-GA mRNA (data not shown). Thus all of the half-life analyses were performed using a similar level of expression of the reporter mRNA.
Using the pulse-chase protocol, the G-GA mRNA was degraded with a half-life of 2.9 h in LLC-PK1-F+ cells that were maintained in normal medium and a 14-fold stabilization was observed when the cells were transferred to an acidic medium. In the previous study that used DRB to inhibit transcription, a half-life of 5.8 h and a 2.5-fold pH-responsive stabilization were observed for this construct (19). Therefore, data obtained using this less intrusive protocol both confirm and refine the previous conclusions that the 3'-UTR of the GA mRNA contains instability elements and confers a pH-responsive stabilization of the G-GA mRNA under acidic conditions.
The total RNAs isolated from the transcriptional pulse analysis were also used to determine the mechanism of decay of the G-GA mRNA. It was hypothesized that deadenylation would precede its decay as the 3'-UTR of GA contains AREs. Therefore, the isolated RNAs were incubated with a GA-specific oligonucleotide and digested with RNase H to produce a fragment of the 3'-UTR that was suitable for monitoring deadenylation (Fig. 3). Deadenylation of the G-GA mRNA appeared to be synchronous under both normal and acidic conditions. Synchronous deadenylation is consistent with a distributive or nonprocessive ribonucleolytic removal of the poly(A) tail. Synchronous deadenylation has been observed in mRNAs containing both class I or class III ARE (2). Because the GA pH-RE lacks the canonical AUUUA sequence, it would be classified as a class III ARE. Thus the observation of distributive deadenylation by the G-GA mRNA is consistent with the previous classification. However, when only synchronous deadenylation occurs, the body of the mRNA is not degraded until the mRNA is fully deadenylated. With the G-GA mRNA, the body of the mRNA is degraded concomitant with the synchronous formation of a shortened, but not fully deadenylated poly(A) tail. Thus some of the G-GA mRNA may undergo a more rapid, but processive, deadenylation to yield a fully deadenylated mRNA that is degraded as rapidly as it is produced. The latter process would determine the rate of turnover of the G-GA mRNA.
When the LLC-PK1-F+ cells were transferred to acidic medium, both the rate and extent of deadenylation were significantly inhibited. This suggests that the factors that normally bind to the instability elements of the GA mRNA are either modified or displaced by alternative factors. As a result, the processive component of the deadenylation must be inhibited while the rate of synchronous deadenylation is also reduced significantly. Such changes in deadenylation would be sufficient to stabilize the GA mRNA. The observed changes in decay rates that were quantified with the G-GA(R2-I) and G-mGA mRNAs confirm the previous conclusion that the pH-RE is both sufficient and necessary to produce a pH-responsive stabilization (19). However, the observations that the G-GA(R2-I) mRNA exhibits only a 5.5-fold stabilization, while the G-mGA mRNA is degraded with a half-life similar to the G-GA mRNA, strongly indicate that the 3'-UTR of the GA mRNA contains additional instability element(s) that may also participate in the pH-responsive stabilization. Additional experiments are needed to identify the remaining element(s) and the proteins that bind to these sequences.
ARE-mediated decay of mRNAs is usually proceeded by rapid deadenylation (2). AUF1 is a protein that binds to AREs within various mRNAs and enhances the rate of degradation (7, 33). Thus the observation that deadenylation occurs concomitantly with the rapid degradation of the G-GA mRNA suggested that AUF1 may participate in this process. This hypothesis is supported by the observations that treatment with anti-AUF1 antibodies produces a slight supershift of the GA(R2-I) RNA/protein complex formed with a cytosolic extract of rat kidney and that recombinant p40 AUF1 exhibits specific binding to the pH-RE within the GA(R2-I) RNA. Analysis using the mutated GA(R2-I) RNAs indicates that a single 8-base AU-element is sufficient for AUF-1 binding. The same elements also bind -crystallin with high affinity and specificity (28), suggesting that the two proteins may compete for binding. Unfortunately, the recombinant p40 AUF1 and the affinity-purified rat kidney -crystallin form complexes that have identical mobility on a nondenaturing polyacryalamide gel (data not shown). Thus it was not feasible to test this hypothesis using the current RNA gel-shift protocol. However, previous studies demonstrating that AUF1 and HuR, an RNA-stabilizing protein, compete for binding to multiple AREs (16) support the feasibility of this hypothesis.
Scatchard analysis demonstrated that the interaction between recombinant p40 AUF1 and the GA(R2-I) RNA has a Kd of 2 μM. By contrast, fluorescence anisotropy measurements indicated that recombinant p40 AUF1 binds to the ARE of TNF- mRNA with a 50-fold greater affinity (34). However, this same analysis demonstrated that phosphorylation of AUF1 affects the affinity and physical interaction between AUF1 and an ARE. With the use of Scatchard analysis, affinity-purified -crystallin was also shown to bind to the GA(R2-I) RNA with a Kd of 20–40 nM (18). However, recombinant -crystallin bound the GA(R2-I) RNA with an affinity similar to that of the recombinant p40 AUF1 (data not shown). Thus it will be important to determine whether the onset of acidosis alters the phosphorylation states of AUF1 and/or -crystallin and whether these changes affect their relative binding interaction with the pH-RE.
The combined observations support the following hypothesis for the mechanism by which changes in acid-base balance determine the stability of the GA mRNA. In normal acid-base balance, AUF1 preferentially binds to the pH-RE and recruits the protein complexes that account for the rapid removal of the poly(A) tail and promotes the subsequent degradation of the GA mRNA. The onset of acidosis is likely to activate a specific signaling mechanism (8, 22) that may result in covalent modification of -crystallin and/or AUF1 and the preferential binding of -crystallin to the pH-RE. As a result, the interaction of the deadenylase with the poly(A) tail is reduced. This results in the prolonged maintenance of the poly(A) tail and the increased stability of the GA mRNA. This hypothesis suggests numerous experiments that can be performed to further characterize the underlying mechanism and the regulation of this important adaptive response.
GRANTS
This work was supported by Public Health Service Grant DK-37124 from the National Institute of Diabetes and Digestive and Kidney Diseases.
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
Chen C and Okayama H. High-efficiency transformation of mammalian cells by plasmid DNA. Mol Cell Biol 7: 2745–2752, 1987.
Chen CY and Shyu AB. AU-rich elements: characterization and importance in mRNA degradation. Trends Biochem Sci 20: 465–470, 1995.
Curthoys NP and Gstraunthaler G. Mechanism of increased renal gene expression during metabolic acidosis. Am J Physiol Renal Physiol 281: F381–F390, 2001.
Curthoys NP and Lowry OH. The distribution of glutaminase isoenzymes in the various structures of the nephron in normal, acidotic, and alkalotic rat kidney. J Biol Chem 248: 162–168, 1973.
Curthoys NP and Watford M. Regulation of glutaminase activity and glutamine metabolism. Annu Rev Nutr 15: 133–159, 1995.
D’Alessio JM, Harris GH, Perna PJ, and Paule MR. Ribosomal ribonucleic acid repeat unit of Acanthamoeba castellanii: cloning and restriction endonuclease map. Biochemistry 20: 3822–3827, 1981.
DeMaria CT and Brewer G. AUF1 binding affinity to A+U-rich elements correlates with rapid mRNA degradation. J Biol Chem 271: 12179–12184, 1996.
Feifel E, Obexer P, Andratsch M, Euler S, Taylor L, Tang A, Wei Y, Schramek H, Curthoys NP, and Gstraunthaler G. p38 MAPK mediates acid-induced transcription of PEPCK in LLC-PK1-FBPase+ cells. Am J Physiol Renal Physiol 283: F678–F688, 2002.
Gao M, Fritz DT, Ford LP, and Wilusz J. Interaction between a poly(A)-specific ribonuclease and the 5' cap influences mRNA deadenylation rates in vitro. Mol Cell 5: 479–488, 2000.
Gorman C, Padmanabhan R, and Howard BH. High efficiency DNA-mediated transformation of primate cells. Science 221: 551–553, 1983.
Gstraunthaler G and Handler JS. Isolation, growth, and characterization of a gluconeogenic strain of renal cells. Am J Physiol Cell Physiol 252: C232–C238, 1987.
Hansen WR, Barsic-Tress N, Taylor L, and Curthoys NP. The 3'-UTR of rat renal glutaminase mRNA contains a pH-responsive stability element. Am J Physiol Renal Fluid Electrolyte Physiol 271: F126–F131, 1996.
Hopfer U, Jacobberger JW, Gruenert DC, Eckert RL, Jat PS, and Whitsett JA. Immortalization of epithelial cells. Am J Physiol Cell Physiol 270: C1–C11, 1996.
Hwang JJ, Perera S, Shapiro RA, and Curthoys NP. Mechanism of altered renal glutaminase gene expression in response to chronic acidosis. Biochemistry 30: 7522–7526, 1991.
Kaiser S and Curthoys NP. Effect of pH and bicarbonate on phosphoenolpyruvate carboxykinase and glutaminase mRNA levels in cultured renal epithelial cells. J Biol Chem 266: 9397–9402, 1991.
Lal A, Mazan-Mamczarz K, Kawai T, Yang X, Martindale JL, and Gorospe M. Concurrent versus individual binding of HuR and AUF1 to common labile target mRNAs. EMBO J 23: 3092–3102, 2004.
Lapucci A, Donnini M, Papucci L, Witort E, Tempestini A, Bevilacqua A, Nicolin A, Brewer G, Schiavone N, and Capaccioli S. AUF1 Is a bcl-2 A + U-rich element-binding protein involved in bcl-2 mRNA destabilization during apoptosis. J Biol Chem 277: 16139–16146, 2002.
Laterza OF and Curthoys NP. Effect of acidosis on the properties of the glutaminase mRNA pH-response element binding protein. J Am Soc Nephrol 11: 1583–1588, 2000.
Laterza OF and Curthoys NP. Specificity and functional analysis of the pH-responsive element within renal glutaminase mRNA. Am J Physiol Renal Physiol 278: F970–F977, 2000.
Laterza OF, Hansen WR, Taylor L, and Curthoys NP. Identification of an mRNA-binding protein and the specific elements that may mediate the pH-responsive induction of renal glutaminase mRNA. J Biol Chem 272: 22481–22488, 1997.
Laterza OF, Taylor L, Unnithan S, Nguyen L, and Curthoys NP. Mapping and functional analysis of an instability element in phosphoenolpyruvate carboxykinase mRNA. Am J Physiol Renal Physiol 279: F866–F873, 2000.
Li S, Sato S, Yang X, Preisig PA, and Alpern RJ. Pyk2 activation is integral to acid stimulation of sodium/hydrogen exchanger 3. J Clin Invest 114: 1782–1789, 2004.
Loflin PT, Chen CY, Xu N, and Shyu AB. Transcriptional pulsing approaches for analysis of mRNA turnover in mammalian cells. Methods 17: 11–20, 1999.
Lowry OH, Rosebrough NJ, Farr AL, and Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 193: 265–275, 1951.
Parker RSH. The enzymes and control of eukaryotic mRNA turnover. Nat Struct Mol Biol 11: 121–127, 2004.
Porter D and Curthoys NP. Use of thermostable and Escherichia coli RNase H in RNA mapping studies. Anal Biochem 247: 279–286, 1997.
Schroeder JM, Liu W, and Curthoys NP. pH-responsive stabilization of glutamate dehydrogenase mRNA in LLC-PK1-F+ cells. Am J Physiol Renal Physiol 285: F258–F265, 2003.
Tang A and Curthoys NP. Identification of -crystallin/NADPH:quinone reductase as a renal glutaminase mRNA pH response element-binding protein. J Biol Chem 276: 21375–21380, 2001.
Tong J, Harrison G, and Curthoys NP. The effect of metabolic acidosis on the synthesis and turnover of rat renal phosphate-dependent glutaminase. Biochem J 233: 139–144, 1986.
Van Hoof A and Parker R. The exosome: a proteasome for RNA Cell 99: 347–350, 1999.
Wagner BJ, DeMaria CT, Sun Y, Wilson GM, and Brewer G. Structure and genomic organization of the human AUF1 gene: alternative pre-mRNA splicing generates four protein isoforms. Genomics 48: 195–202, 1998.
Wilson GM and Brewer G. Identification and characterization of proteins binding A + U-rich elements. Methods 17: 74–83, 1999.
Wilson GM and Brewer G. The search for trans-acting factors controlling messenger RNA decay. Prog Nucleic Acid Res Mol Biol 62: 257–291, 1999.
Wilson GM, Lu J, Sutphen K, Suarez Y, Sinha S, Brewer B, Villanueva-Feliciano EC, Ysla RM, Charles S, and Brewer G. Phosphorylation of p40AUF1 regulates binding to A + U-rich mRNA-destabilizing elements and protein-induced changes in ribonucleoprotein structure. J Biol Chem 278: 33039–33048, 2003.
Wilson GM, Sun Y, Lu H, and Brewer G. Assembly of AUF1 oligomers on U-rich RNA targets by sequential dimer association. J Biol Chem 274: 33374–33381, 1999.
Wright PA and Knepper MA. Phosphate-dependent glutaminase activity in rat renal cortical and medullary tubule segments. Am J Physiol Renal Fluid Electrolyte Physiol 259: F961–F970, 1990.
Zhang W, Wagner BJ, Ehrenman K, Schaefer AW, DeMaria CT, Crater D, DeHaven K, Long L, and Brewer G. Purification, characterization, and cDNA cloning of an AU-rich element RNA-binding protein, AUF1. Mol Cell Biol 13: 7652–7665, 1993.(Jill M. Schroeder, Hend Ibrahim, Lynn Ta)
ABSTRACT
During chronic metabolic acidosis, increased expression of renal glutaminase (GA) results from selective stabilization of the GA mRNA. This response is mediated by a direct repeat of an 8-base adenylate-uridylate (AU) sequence that binds -crystallin and functions as a pH response element (pH-RE). A tetracycline-responsive promoter system was developed in LLC-PK1-F+ cells to perform pulse-chase analysis of the turnover of a chimeric -globin (G) mRNA that contains 960 bp of the 3'-UTR of GA mRNA including the pH-RE. The G-GA mRNA exhibits a 14-fold increase in half-life when the LLC-PK1-F+ cells are transferred to acidic medium. RNase H cleavage and Northern blot analysis of the 3'-ends established that rapid deadenylation occurred concomitantly with the rapid decay of the G-GA mRNA in cells grown in normal medium. Stabilization of the G-GA mRNA in acidic medium is associated with a pronounced decrease in the rate of deadenylation. Mutation of the pH-RE within the G-GA mRNA blocked the pH-responsive stabilization, but not the rapid decay, whereas insertion of only a 29-bp segment containing the pH-RE was sufficient to produce both a rapid decay and a pH-responsive stabilization. Various kidney cells express multiple isoforms of AUF1, an AU-binding protein that enhances mRNA turnover. RNA gel-shift assays demonstrated that the recombinant p40 isoform of AUF1 binds to the pH-RE with high affinity and specificity. Thus AUF1 may mediate the rapid turnover of the GA mRNA, whereas increased binding of -crystallin during acidosis may inhibit degradation and result in selective stabilization.
renal ammoniagenesis; metabolic acidosis; posttranscriptional regulation
AS PART OF THE ADAPTIVE RESPONSE to metabolic acidosis, renal extraction and catabolism of plasma glutamine are rapidly increased (3). This adaptation results in a large increase in the renal synthesis of ammonium and bicarbonate ions. The ammonium ions are preferentially excreted in the urine where they provide an expendable cation that facilitates the excretion of acidic anions. In contrast, the bicarbonate is added to the renal venous blood to partially compensate for the systemic acidosis. Renal catabolism of glutamine is initiated by a phosphate-activated mitochondrial glutaminase (5). During chronic acidosis, increased catabolism of glutamine is sustained, in part, by an 8- to 20-fold increase in the level of the mitochondrial glutaminase that occurs solely within the proximal convoluted tubule (4, 36).
During acidosis, the increased synthesis of the renal glutaminase (GA) protein (29) occurs without increasing the rate of transcription of the GA mRNA (14, 15). Experiments that characterized the half-life of a chimeric -globin-glutaminase (G-GA) mRNA in LLC-PK1-F+ cells, a pH-responsive line of porcine proximal tubule-like cells (11), established that the 3'-untranslated region (3'-UTR) of the GA mRNA contains a pH-responsive instability element (12). RNA gel-shift analysis demonstrated the presence of a protein in rat renal cortical cytosolic extracts that binds to the 3'-UTR of the GA mRNA with high affinity and specificity (20). This binding interaction was mapped to a direct repeat of an 8-base AU sequence. Functional studies established that this sequence was both necessary and sufficient to impart a pH-responsive stabilization to a chimeric G mRNA (21). Subsequent studies identified -crystallin/NADPH quinone reductase as the primary protein in rat renal extracts that binds to the pH response element (pH-RE) (28).
The rapid degradation of mammalian mRNAs is frequently initiated by the binding of proteins (25) that recruit a poly(A)-specific ribonuclease (9) and the exosome (30) to remove the poly(A) tail and accomplish a rapid 3'5' exonucleolytic degradation, respectively. In the current study, a tetracycline-responsive promoter (23) was used to perform a pulse-chase analysis of the decay of newly synthesized G-GA mRNA. The resulting data demonstrate that the pH-responsive stabilization of the G-GA mRNA is associated with a proportional decrease in the rate of deadenylation. Additional studies established that recombinant AUF1, an AU-binding protein that enhances mRNA turnover (32), also binds to the pH-RE with high affinity and specificity. Therefore, the enhanced binding of -crystallin to this site during acidosis (18) may block AUF1 binding and prevent the AUF1-mediated deadenylation and degradation of the GA mRNA.
MATERIALS AND METHODS
Materials. Male Sprague-Dawley rats were purchased from Charles River. [-32P]dCTP (specific activity 3,000 Ci/mmol) was purchased from MP Biomedicals or Amersham Pharmacia Biotechnology. The animal protocols used in this study were received and approved by The University Animal Care and Use Committee as protocol 93-25DA-12. The oligolabeling kit was from Ambion. Restriction enzymes, RNase T1, T7 RNA polymerase, and yeast tRNA were acquired from Roche, New England Biolabs and MBI Fermentas. RNase H was purchased from Epicentre Technologies. GENECLEAN kits were obtained from Bio101. 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 Promega. DMEM/F12 medium, Geneticin (G418), and hygromycin B were products of Sigma and Mediatech. A polyclonal antibody vs. AUF1 was obtained from Upstate. An expression plasmid that encodes the p40 isoform of AUF1 was obtained from Dr. J. Wilusz. All other biochemicals were acquired from Sigma.
Synthesis of pTRE2 constructs. pTRE2 (Clontech) contains a tetracycline-responsive element (TRE) that includes seven direct repeats of a 42-bp sequence containing the tet operator positioned upstream of a minimal CMV promoter, a multicloning site, and the 3'-coding and 3'-untranslated regions of the rabbit -globin gene. pTRE2 was digested with MluI and SapI to remove the -globin sequence. A segment of pG-GA (12) containing the -globin coding region, 960 bp of the GA 3'-UTR, and a segment of the bovine growth hormone (bGH) 3'-untranslated region containing the polyadenylation site was PCR amplified with primers containing unique MluI and SapI restriction sites. This fragment was cloned into pCR-Script SK(+) at the SrfI site, amplified, and then ligated into the pTRE2 fragment to create pTRE2-G-GA. The pG and pG-GA(R2-I) plasmids (27) were digested with CelII and XbaI and the isolated fragments were inserted into pTRE2-G-GA that had been digested with CelII and XbaI. This procedure removed the GA 3'-UTR or replaced it with a 29-bp sequence containing the pH response element, respectively. The pG-mGA plasmid (19) was digested with NheI and NotI and the isolated fragment was inserted into pTRE2-G-GA that had been digested with NheI and NotI to produce a construct containing the GA 3'-untranslated region in which the pH response element was mutated to include GC bases.
Selection of stable cell lines. LLC-PK1-F+ cells (11) were grown in a 50:50 mixture of DMEM and Ham’s F-12 medium containing 5 mM glucose and 10% fetal bovine serum at 37°C in a 5% CO2 atmosphere. Normal medium (pH 7.4) contained 25 mM sodium bicarbonate, whereas acidic medium (pH 6.9) contained 10 mM sodium bicarbonate supplemented with 15 mM sodium chloride to maintain an equivalent osmolarity and sodium ion concentration. Clonal lines of LLC-PK1-F+ cells that stably express the tTA protein were produced by calcium phosphate transfection (1) of 3-day postsplit cells with the pTet-off plasmid and selection with medium containing 0.8 mg/ml G418. After 14–21 days, colonies were selected and expanded in medium containing 0.2 mg/ml G418. The expanded cell lines were grown for 3 days on six-well plates using medium containing 1 μg/ml doxycycline (Dox) and then transiently transfected with pTRE2-Luc. Approximately 16 h later, the medium was replaced with fresh medium with or without 1 μg/ml Dox and the cells were cultured for an additional 32 h. Cell extracts were prepared and assayed using the reagents contained in the Luciferase Assay System (Promega). The firefly luciferase activities obtained for the various stable Tet-Off clonal lines were standardized per microgram of protein (24). The fold-induction of the TRE2-Luc gene was calculated as the mean of the ratio of the specific luciferase activities measured in triplicate cultures grown in the absence and presence of Dox.
Clonal cell lines that stably express the various G-GA constructs were prepared by cotransfection of LLC-PK1-F+ cells that express the tTA protein with the appropriate TRE2 plasmid and pcDNA3.1/Hygro (Invitrogen). At 48 h posttransfection, selection medium containing 0.2 mg/ml G418 and 0.8 mg/ml hygromycin was added. After 14–21 days, individual colonies were isolated and expanded. The cell lines were tested for responsiveness by growing the cells for 48 h in the presence or absence of 0.5 μg/ml Dox, a tetracycline analog. Total RNA was harvested and analyzed by Northern blotting.
Half-life analysis. For pulse-chase analysis, the stably transfected LLC-PK1-F+ cells were split 1:10 and grown for 5–7 days in normal medium containing 0.2 mg/ml G418 and 0.2 mg/ml hygromycin B. The cells were then grown for 48 h in pH 7.4 or 6.9 medium containing antibiotics and 25 ng/ml Dox. The cells were washed two times with phosphate-buffered saline and incubated in normal or acidic medium with antibiotics but without Dox for 3 h to create a transcriptional pulse. Then, fresh medium containing antibiotics and 1 μg/ml Dox was added and total RNA was isolated at 0, 3, 6, and 9 h using the TRIzol reagent (Invitrogen). Control cells were grown for 48 h in the presence or absence of Dox. Alternatively, the stably transfected LLC-PK1-F+ cells were grown for 5–7 days and then maintained in normal or acidic medium minus Dox for 24 h. Subsequently, 1 μg/ml Dox was added to each plate and total RNA was isolated at 0, 3, 6, and 9 h post-Dox addition. Control plates containing no Dox were harvested at 0 and 9 h.
Northern blot analysis. A 507-bp fragment of rabbit -globin cDNA was excised by restricting pRSV-G (10) with HindIII and BglII. A 2.0-kb fragment of the 18 S ribosomal RNA cDNA from Acanthamoeba castellanii was obtained by restricting pAr2 (6) with HindIII and EcoRI. A 228-bp bGH fragment was excised from pCRScript-bGH with SphI. The fragments were separated on 1% agarose gels, excised, and purified using the GENECLEAN kit. The synthesis of oligolabeled cDNA probes and Northern blot analysis was performed as described previously (27). The blots were exposed to a PhosphorImager screen and quantified using Molecular Dynamics Software. The level of the chimeric -globin mRNA was divided by the level of the corresponding 18S rRNA to correct for errors in sample loading. The log of normalized data was then plotted vs. the time of addition of Dox.
RNase H treatment. A 26-nt deoxyoligonucleotide (GA oligo), which is complementary to the 3'-end of the GA sequence encoded by pTRE2-G-GA (bp 2514–2539 of the GA cDNA), was synthesized by Macromolecular Resources (Fort Collins, CO). The sequence of the GA oligo is 5'-GACCGACAGCTACACACCACAGAGAG-3'. A 30-mer of oligo dT deoxyoligonucleotide was used to degrade the poly(A) tail of the mRNA. RNA samples, which were stored in Formazol, were precipitated with 4 vol of ethanol and a final concentration of 0.2 M sodium chloride and resuspended in 25 μl of DEPC-treated water. RNase H treatment was performed as described previously (26). Briefly, 10 μg RNA, 1.0 μl containing 50 pmol of GA oligo, and 1.0 μl containing 50 pmol of oligo-dT (if needed) in 12.5 μl were incubated at 55°C for 1 min. Next, 1.5 μl of 10x RNase H buffer (0.5 M Tris, pH 7.4, 1.0 M NaCl, 20 mM MgCl2, and 10 mM dithiothreitol) and 1.0 μl of freshly diluted RNase H (1 U/μl) were added and the sample was incubated at 62°C for 15 min. Then, 25 μl of denaturing buffer (20 μl Formazol, 3 μl 10x MOPS, and 6 μl 37% formaldehyde solution) were added and the mixture was incubated at 55°C for 5 min and then transferred to ice. RNase H-treated RNAs were mixed with 10 μl of an RNase H gel loading buffer (50% glycerol, 1 mM EDTA, 0.25% bromophenol blue, 0.25% xylene cyanol, and 100 μg/ml ethidium bromide) and subjected to Northern blot analysis using a 1.2% agarose gel. The blot was hybridized with the bGH probe.
Western blotting. Cytosolic extracts of rat renal cortex and of various cell lines were prepared as described previously (19). The samples were separated by electrophoresis on a 10% SDS-polyacrylamide gel, transferred to a nitrocellulose membrane, and incubated with a 1:1,000 dilution of AUF1 antibody. The membrane was then incubated with a 1:15,000 dilution of horseradish peroxidase (HRP)-conjugated secondary antibody. Images were developed with Dura reagent (Pierce Chemicals) and visualized by brief exposure to X-ray film.
RNA EMSA. 32P-labeled and -unlabeled GA(R2-I) and mutated forms of GA(R2-I) RNAs were synthesized as described previously (19). An aliquot containing 1–8 μg of rat renal cortical protein or 12.5–100 ng of recombinant AUF1 was preincubated for 10 min at room temperature in binding buffer containing 0.5% Nonidet P-40, 1 mM dithiothreiotol, 2 μg yeast t-RNA, 40 U of RNAsin, 10% glycerol, and 10–70 fmol of 32P-labeled RNA. The indicated excess of unlabeled RNA was added, and the reaction mixture was incubated at room temperature for 20 min. The samples were then loaded and subjected to electrophoresis for 2 h at 10 mA using a 90 mM Tris, 110 mM boric acid, and 2 mM EDTA running buffer. Gels were dried and exposed overnight to a PhosphorImager screen. In the antibody supershift experiments, the anti-AUF1 antibody was preincubated with the rat renal cytosolic extract for 10 min at room temperature and then for 1 h on ice before the binding buffer was added.
Statistical analyses were performed using a paired Student’s t-test.
RESULTS
Effect of medium pH on deadenylation of G-GA mRNA. Clonal cell lines that stably express the tTA protein, a tetracycline-responsive transcription factor, were isolated by transfection of LLC-PK1-F+ cells with pTet-off and selection with G418. Their Dox responsiveness was tested by transient expression of pTRE2-Luc in cells that were maintained in the presence or absence of Dox. The 8C line of LLC-PK1-F+ cells showed the greatest induction of luciferase activity. With the use of 0.3 μg of pTRE2-Luc DNA, they produced a 200-fold difference in luciferase activity when grown in the presence and absence of 1 μg/ml Dox. Therefore, the 8C line of LLC-PK1-F+ cells was used for the subsequent experiments. Clonal cell lines that stably express pTRE2-G-GA were isolated by cotransfection of the 8C cells with pcDNA3.1/Hygro and selection with hygromycin. The isolated cell lines were tested for Dox-responsive G-GA expression by culturing the cells in the presence or absence of Dox and quantifying G-GA mRNA expression by Northern blot analysis. Cell lines were evaluated based on the level of G-GA mRNA expressed in the presence of Dox and the increased expression produced by removal of Dox. The selected 8C-2A cell line exhibited a 50-fold difference in the levels of G-GA mRNA when grown in the presence or absence of 0.5 μg/ml Dox (Fig. 1).
Previous experiments characterizing the half-lives of various G-GA chimeric mRNAs used the Rous sarcoma virus long-terminal repeat (RSV) promoter to drive transcription (20). The level of G-GA mRNA transcribed from the tetracycline-responsive promoter in the absence of Dox was far greater than the level of mRNA synthesized from the RSV promoter (Fig. 1). To create a level of G-GA mRNA similar to that produced with the RSV promoter, transcription was initially inhibited by adding 25 ng/ml Dox for 48 h. Dox was then removed from the selected cells to create a pulse of transcription and RNA was isolated at various times. The level of G-GA mRNA produced by a 3-h transcriptional pulse was comparable to the level of G-GA mRNA produced by the RSV promoter (Fig. 1). Therefore, a 3-h transcriptional pulse should provide a level of newly synthesized and adenylated G-GA mRNA that is sufficient to follow decay and deadenylation.
To perform a pulse-chase analysis, the selected cells were maintained in normal or acidic medium containing 25 ng/ml Dox for 48 h. A 3-h pulse was created by removing Dox and then chased for 0, 3, 6, and 9 h following addition of 1 μg/ml Dox. The half-life of G-GA mRNA in cells cultured in normal medium was 2.9 h. However, when the cells were maintained in acidic conditions, the G-GA mRNA was significantly stabilized and now decayed with a half-life of 40 h (Fig. 2). An oligonucleotide (GA oligo), which specifically hybridizes to the 3'-end of the GA sequence, was used along with RNase H to cleave the G-GA mRNA and to produce a short 3'-fragment (Poly A+) that contains the poly(A) tail (Fig. 3A). Oligo(dT) was also added to an aliquot of the zero-time RNA sample to remove the poly(A) tail from the G-GA mRNA, creating a 330-nt deadenylated RNA fragment (Poly A–). The 3'-fragments were detected by Northern blot analysis using a 32P-labeled cDNA probe that specifically hybridizes to the bGH sequence within the 3'-end of the G-GA mRNA.
The adenylated and the fully deadenylated RNA fragments were easily resolved, making it feasible to monitor the extent of deadenylation of the G-GA mRNA. The 3'-ends of the G-GA mRNAs harvested from cells maintained in normal medium appear to exhibit a synchronous or nonprocessive pattern of deadenylation (Fig. 3B). However, the 3'-fragments containing the shortened poly(A) tail decay at a rate similar to that observed with the full-length mRNA. If complete removal of the poly(A) is required to initiate decay, then a portion of the G-GA mRNA must undergo a more rapid and processive deadenylation to form a transient, but fully deadenylated, intermediate that is rapidly degraded and does not accumulate. As a result, the synchronous deadenylation of the G-GA mRNAs would appear to occur concomitantly with the decay of the body of the chimeric mRNA. When the cells were treated with acidic medium, the rates of synchronous deadenylation (formation of shorter 3'-fragments) and processive deadenylation (decay of 3'-fragments) of the G-GA mRNA were significantly inhibited. Therefore, the pH-responsive stabilization correlates with a decrease in both the rate and the extent of deadenylation of the G-GA mRNA.
Functional analysis of the pH-RE. As reported previously (20), the chimeric -globin mRNA lacking sequence from the GA 3'-UTR is extremely stable. However, the insertion of a 29-bp segment containing the pH-RE is sufficient to create a labile mRNA that decayed with a half-life of 3.0 h in LLC-PK1-F+ cells that were maintained in normal medium (Fig. 4). When transferred to acidic medium, the G-GA(R2-I) mRNA decayed with a half-life of 17 h. Thus this segment functions both as an instability element and as a pH-RE. In contrast, the G-mGA mRNA, which contains the GA 3'-UTR in which the AU-rich pH-RE was mutated to the Mut3 sequence shown in Fig. 8A, is degraded rapidly in cells that are grown in either normal or acidic medium (Fig. 5). Thus the 3'-UTR must contain multiple instability elements that contribute to the rapid turnover of the GA mRNA.
Binding of AUF1 to the pH-RE. AUF1 binds to various AU-rich elements (AREs) and enhances the mRNA degradation (32). Previous studies (19, 27) demonstrated that GA(R2-I), a 29-bp RNA segment that contains the pH-RE and surrounding sequence within the GA mRNA, forms a specific complex when incubated with a cytosolic extract of rat kidney cortex. Preincubation of the cytosolic extract with AUF1-specific antibodies results in the reproducible formation of a complex that exhibits a slightly slower electrophoretic mobility (Fig. 6A). No shift was observed when the cytosolic extract was preincubated with an equivalent amount of an unrelated antibody (data not shown). The observed shift suggests that AUF1 may bind to the GA(R2-I) RNA. Western blot analysis was performed to determine whether AUF1 is expressed in rat kidney cortex and in various proximal tubule cell lines (Fig. 6B). Jurkat cell extracts were used as a positive control (17). They express the four isoforms of AUF1 (p37, p40, p42, and p45) that are produced by alternative splicing of the initial AUF1 transcript (31). Rat kidney cortex contains significant levels of the p37, p40, and p42 isoforms of AUF1, whereas the LLC-PK1-F+ and WKPT (13) proximal tubule cells express differing amounts of the four isoforms.
The p40 AUF1 isoform is predominantly cytoplasmic (37) and hence was utilized in the following binding experiments. Increasing amounts of the purified recombinant p40 AUF1 form multiple complexes when incubated with the GA(R2-I) RNA (Fig. 7A). This pattern is characteristic of the fact that AUF1 forms dimers and larger oligomers on binding to AU-rich sequences (35). The observed binding of AUF1 is competed by increasing concentrations of unlabeled GA(R2-I) RNA, but not by a nonspecific RNA segment (pBSSK) that was transcribed from the multicloning site of pBluescript-SK(–) (Fig. 7B). Thus the observed binding interaction is specific. The complexes formed between AUF1 and the GA(R2-I) RNA are similar to those observed when AUF1 is incubated with a short RNA that contains the canonical AUUUA element (Fig. 8). AUF1 also binds to GA(R2-I) segments in which one of the two AU sequences is mutated to contain GC-nucleotides (Mut1 and Mut2). However, AUF1 does not bind to an GA(R2-I) segment in which both AU sequences are mutated (Mut3).
Scatchard analysis was performed to determine the dissociation constant (Kd) for AUF1 binding to the GA(R2-I) RNA. When 80 ng of recombinant AUF1 were incubated with 70 fmol of 32P-labeled R-2I RNA, approximately one-third of the labeled probe formed a complex (Fig. 9A). The intensity of the shifted band decreased as the amount of unlabeled R-2I RNA was increased. The relative intensities of the bands that correspond to the bound and free probe were quantified by PhosphorImager analysis. These data along with the knowledge of the concentration of total probe in each sample were then used to calculate the concentrations of bound and free probe. The resulting values were plotted in the form of a Scatchard plot (Fig. 9B). Linear regression analysis of the plotted data indicated that the recombinant AUF1 binds the GA(R2-I) RNA with a Kd of 2 μM.
DISCUSSION
A tetracycline-responsive expression (Tet-off) system (23) was used to characterize the potential role of deadenylation in the degradation of the GA mRNA. With this system, it is feasible to inhibit the synthesis of a single mRNA through the addition of a nontoxic level of Dox, instead of using 5–6 dichloro-1--ribofuranosylbenzimidazole (DRB), a general inhibitor of pol II transcription. Thus this approach avoids potential nonspecific effects that may be caused by the rapid decay of regulatory proteins or nucleases that participate in mRNA turnover. As a result, this protocol should also provide a more accurate quantification of the half-life of a mRNA.
When fully induced, the tetracycline-responsive promoter produced a significantly greater level of G-GA mRNA than observed in previous experiments that used an RSV promoter to drive the synthesis of the reporter mRNA (12, 19). To avoid expression of very high levels of G-GA mRNA that might exceed the amount of -crystallin and/or other factors needed to accomplish a pH-responsive stabilization, a pulse-chase protocol was used to characterize the turnover of the G-GA mRNA. With the use of this protocol, the levels of newly synthesized G-GA mRNA were similar to those produced constitutively from the RSV promoter. None of the cell lines selected for stable expression of the G-GA(R2-I) or G-mGA mRNA produced the same rapid and pronounced increase, during a 3-h pulse, as observed with the 8C-2A line that expresses the G-GA mRNA. As a result, it was not feasible to perform pulse-chase analysis with the truncated or mutated construct. However, the fully induced levels of the G-GA(R2-I) and G-mGA mRNAs were similar to that produced following a 3-h pulse of the G-GA mRNA (data not shown). Thus all of the half-life analyses were performed using a similar level of expression of the reporter mRNA.
Using the pulse-chase protocol, the G-GA mRNA was degraded with a half-life of 2.9 h in LLC-PK1-F+ cells that were maintained in normal medium and a 14-fold stabilization was observed when the cells were transferred to an acidic medium. In the previous study that used DRB to inhibit transcription, a half-life of 5.8 h and a 2.5-fold pH-responsive stabilization were observed for this construct (19). Therefore, data obtained using this less intrusive protocol both confirm and refine the previous conclusions that the 3'-UTR of the GA mRNA contains instability elements and confers a pH-responsive stabilization of the G-GA mRNA under acidic conditions.
The total RNAs isolated from the transcriptional pulse analysis were also used to determine the mechanism of decay of the G-GA mRNA. It was hypothesized that deadenylation would precede its decay as the 3'-UTR of GA contains AREs. Therefore, the isolated RNAs were incubated with a GA-specific oligonucleotide and digested with RNase H to produce a fragment of the 3'-UTR that was suitable for monitoring deadenylation (Fig. 3). Deadenylation of the G-GA mRNA appeared to be synchronous under both normal and acidic conditions. Synchronous deadenylation is consistent with a distributive or nonprocessive ribonucleolytic removal of the poly(A) tail. Synchronous deadenylation has been observed in mRNAs containing both class I or class III ARE (2). Because the GA pH-RE lacks the canonical AUUUA sequence, it would be classified as a class III ARE. Thus the observation of distributive deadenylation by the G-GA mRNA is consistent with the previous classification. However, when only synchronous deadenylation occurs, the body of the mRNA is not degraded until the mRNA is fully deadenylated. With the G-GA mRNA, the body of the mRNA is degraded concomitant with the synchronous formation of a shortened, but not fully deadenylated poly(A) tail. Thus some of the G-GA mRNA may undergo a more rapid, but processive, deadenylation to yield a fully deadenylated mRNA that is degraded as rapidly as it is produced. The latter process would determine the rate of turnover of the G-GA mRNA.
When the LLC-PK1-F+ cells were transferred to acidic medium, both the rate and extent of deadenylation were significantly inhibited. This suggests that the factors that normally bind to the instability elements of the GA mRNA are either modified or displaced by alternative factors. As a result, the processive component of the deadenylation must be inhibited while the rate of synchronous deadenylation is also reduced significantly. Such changes in deadenylation would be sufficient to stabilize the GA mRNA. The observed changes in decay rates that were quantified with the G-GA(R2-I) and G-mGA mRNAs confirm the previous conclusion that the pH-RE is both sufficient and necessary to produce a pH-responsive stabilization (19). However, the observations that the G-GA(R2-I) mRNA exhibits only a 5.5-fold stabilization, while the G-mGA mRNA is degraded with a half-life similar to the G-GA mRNA, strongly indicate that the 3'-UTR of the GA mRNA contains additional instability element(s) that may also participate in the pH-responsive stabilization. Additional experiments are needed to identify the remaining element(s) and the proteins that bind to these sequences.
ARE-mediated decay of mRNAs is usually proceeded by rapid deadenylation (2). AUF1 is a protein that binds to AREs within various mRNAs and enhances the rate of degradation (7, 33). Thus the observation that deadenylation occurs concomitantly with the rapid degradation of the G-GA mRNA suggested that AUF1 may participate in this process. This hypothesis is supported by the observations that treatment with anti-AUF1 antibodies produces a slight supershift of the GA(R2-I) RNA/protein complex formed with a cytosolic extract of rat kidney and that recombinant p40 AUF1 exhibits specific binding to the pH-RE within the GA(R2-I) RNA. Analysis using the mutated GA(R2-I) RNAs indicates that a single 8-base AU-element is sufficient for AUF-1 binding. The same elements also bind -crystallin with high affinity and specificity (28), suggesting that the two proteins may compete for binding. Unfortunately, the recombinant p40 AUF1 and the affinity-purified rat kidney -crystallin form complexes that have identical mobility on a nondenaturing polyacryalamide gel (data not shown). Thus it was not feasible to test this hypothesis using the current RNA gel-shift protocol. However, previous studies demonstrating that AUF1 and HuR, an RNA-stabilizing protein, compete for binding to multiple AREs (16) support the feasibility of this hypothesis.
Scatchard analysis demonstrated that the interaction between recombinant p40 AUF1 and the GA(R2-I) RNA has a Kd of 2 μM. By contrast, fluorescence anisotropy measurements indicated that recombinant p40 AUF1 binds to the ARE of TNF- mRNA with a 50-fold greater affinity (34). However, this same analysis demonstrated that phosphorylation of AUF1 affects the affinity and physical interaction between AUF1 and an ARE. With the use of Scatchard analysis, affinity-purified -crystallin was also shown to bind to the GA(R2-I) RNA with a Kd of 20–40 nM (18). However, recombinant -crystallin bound the GA(R2-I) RNA with an affinity similar to that of the recombinant p40 AUF1 (data not shown). Thus it will be important to determine whether the onset of acidosis alters the phosphorylation states of AUF1 and/or -crystallin and whether these changes affect their relative binding interaction with the pH-RE.
The combined observations support the following hypothesis for the mechanism by which changes in acid-base balance determine the stability of the GA mRNA. In normal acid-base balance, AUF1 preferentially binds to the pH-RE and recruits the protein complexes that account for the rapid removal of the poly(A) tail and promotes the subsequent degradation of the GA mRNA. The onset of acidosis is likely to activate a specific signaling mechanism (8, 22) that may result in covalent modification of -crystallin and/or AUF1 and the preferential binding of -crystallin to the pH-RE. As a result, the interaction of the deadenylase with the poly(A) tail is reduced. This results in the prolonged maintenance of the poly(A) tail and the increased stability of the GA mRNA. This hypothesis suggests numerous experiments that can be performed to further characterize the underlying mechanism and the regulation of this important adaptive response.
GRANTS
This work was supported by Public Health Service Grant DK-37124 from the National Institute of Diabetes and Digestive and Kidney Diseases.
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
Chen C and Okayama H. High-efficiency transformation of mammalian cells by plasmid DNA. Mol Cell Biol 7: 2745–2752, 1987.
Chen CY and Shyu AB. AU-rich elements: characterization and importance in mRNA degradation. Trends Biochem Sci 20: 465–470, 1995.
Curthoys NP and Gstraunthaler G. Mechanism of increased renal gene expression during metabolic acidosis. Am J Physiol Renal Physiol 281: F381–F390, 2001.
Curthoys NP and Lowry OH. The distribution of glutaminase isoenzymes in the various structures of the nephron in normal, acidotic, and alkalotic rat kidney. J Biol Chem 248: 162–168, 1973.
Curthoys NP and Watford M. Regulation of glutaminase activity and glutamine metabolism. Annu Rev Nutr 15: 133–159, 1995.
D’Alessio JM, Harris GH, Perna PJ, and Paule MR. Ribosomal ribonucleic acid repeat unit of Acanthamoeba castellanii: cloning and restriction endonuclease map. Biochemistry 20: 3822–3827, 1981.
DeMaria CT and Brewer G. AUF1 binding affinity to A+U-rich elements correlates with rapid mRNA degradation. J Biol Chem 271: 12179–12184, 1996.
Feifel E, Obexer P, Andratsch M, Euler S, Taylor L, Tang A, Wei Y, Schramek H, Curthoys NP, and Gstraunthaler G. p38 MAPK mediates acid-induced transcription of PEPCK in LLC-PK1-FBPase+ cells. Am J Physiol Renal Physiol 283: F678–F688, 2002.
Gao M, Fritz DT, Ford LP, and Wilusz J. Interaction between a poly(A)-specific ribonuclease and the 5' cap influences mRNA deadenylation rates in vitro. Mol Cell 5: 479–488, 2000.
Gorman C, Padmanabhan R, and Howard BH. High efficiency DNA-mediated transformation of primate cells. Science 221: 551–553, 1983.
Gstraunthaler G and Handler JS. Isolation, growth, and characterization of a gluconeogenic strain of renal cells. Am J Physiol Cell Physiol 252: C232–C238, 1987.
Hansen WR, Barsic-Tress N, Taylor L, and Curthoys NP. The 3'-UTR of rat renal glutaminase mRNA contains a pH-responsive stability element. Am J Physiol Renal Fluid Electrolyte Physiol 271: F126–F131, 1996.
Hopfer U, Jacobberger JW, Gruenert DC, Eckert RL, Jat PS, and Whitsett JA. Immortalization of epithelial cells. Am J Physiol Cell Physiol 270: C1–C11, 1996.
Hwang JJ, Perera S, Shapiro RA, and Curthoys NP. Mechanism of altered renal glutaminase gene expression in response to chronic acidosis. Biochemistry 30: 7522–7526, 1991.
Kaiser S and Curthoys NP. Effect of pH and bicarbonate on phosphoenolpyruvate carboxykinase and glutaminase mRNA levels in cultured renal epithelial cells. J Biol Chem 266: 9397–9402, 1991.
Lal A, Mazan-Mamczarz K, Kawai T, Yang X, Martindale JL, and Gorospe M. Concurrent versus individual binding of HuR and AUF1 to common labile target mRNAs. EMBO J 23: 3092–3102, 2004.
Lapucci A, Donnini M, Papucci L, Witort E, Tempestini A, Bevilacqua A, Nicolin A, Brewer G, Schiavone N, and Capaccioli S. AUF1 Is a bcl-2 A + U-rich element-binding protein involved in bcl-2 mRNA destabilization during apoptosis. J Biol Chem 277: 16139–16146, 2002.
Laterza OF and Curthoys NP. Effect of acidosis on the properties of the glutaminase mRNA pH-response element binding protein. J Am Soc Nephrol 11: 1583–1588, 2000.
Laterza OF and Curthoys NP. Specificity and functional analysis of the pH-responsive element within renal glutaminase mRNA. Am J Physiol Renal Physiol 278: F970–F977, 2000.
Laterza OF, Hansen WR, Taylor L, and Curthoys NP. Identification of an mRNA-binding protein and the specific elements that may mediate the pH-responsive induction of renal glutaminase mRNA. J Biol Chem 272: 22481–22488, 1997.
Laterza OF, Taylor L, Unnithan S, Nguyen L, and Curthoys NP. Mapping and functional analysis of an instability element in phosphoenolpyruvate carboxykinase mRNA. Am J Physiol Renal Physiol 279: F866–F873, 2000.
Li S, Sato S, Yang X, Preisig PA, and Alpern RJ. Pyk2 activation is integral to acid stimulation of sodium/hydrogen exchanger 3. J Clin Invest 114: 1782–1789, 2004.
Loflin PT, Chen CY, Xu N, and Shyu AB. Transcriptional pulsing approaches for analysis of mRNA turnover in mammalian cells. Methods 17: 11–20, 1999.
Lowry OH, Rosebrough NJ, Farr AL, and Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 193: 265–275, 1951.
Parker RSH. The enzymes and control of eukaryotic mRNA turnover. Nat Struct Mol Biol 11: 121–127, 2004.
Porter D and Curthoys NP. Use of thermostable and Escherichia coli RNase H in RNA mapping studies. Anal Biochem 247: 279–286, 1997.
Schroeder JM, Liu W, and Curthoys NP. pH-responsive stabilization of glutamate dehydrogenase mRNA in LLC-PK1-F+ cells. Am J Physiol Renal Physiol 285: F258–F265, 2003.
Tang A and Curthoys NP. Identification of -crystallin/NADPH:quinone reductase as a renal glutaminase mRNA pH response element-binding protein. J Biol Chem 276: 21375–21380, 2001.
Tong J, Harrison G, and Curthoys NP. The effect of metabolic acidosis on the synthesis and turnover of rat renal phosphate-dependent glutaminase. Biochem J 233: 139–144, 1986.
Van Hoof A and Parker R. The exosome: a proteasome for RNA Cell 99: 347–350, 1999.
Wagner BJ, DeMaria CT, Sun Y, Wilson GM, and Brewer G. Structure and genomic organization of the human AUF1 gene: alternative pre-mRNA splicing generates four protein isoforms. Genomics 48: 195–202, 1998.
Wilson GM and Brewer G. Identification and characterization of proteins binding A + U-rich elements. Methods 17: 74–83, 1999.
Wilson GM and Brewer G. The search for trans-acting factors controlling messenger RNA decay. Prog Nucleic Acid Res Mol Biol 62: 257–291, 1999.
Wilson GM, Lu J, Sutphen K, Suarez Y, Sinha S, Brewer B, Villanueva-Feliciano EC, Ysla RM, Charles S, and Brewer G. Phosphorylation of p40AUF1 regulates binding to A + U-rich mRNA-destabilizing elements and protein-induced changes in ribonucleoprotein structure. J Biol Chem 278: 33039–33048, 2003.
Wilson GM, Sun Y, Lu H, and Brewer G. Assembly of AUF1 oligomers on U-rich RNA targets by sequential dimer association. J Biol Chem 274: 33374–33381, 1999.
Wright PA and Knepper MA. Phosphate-dependent glutaminase activity in rat renal cortical and medullary tubule segments. Am J Physiol Renal Fluid Electrolyte Physiol 259: F961–F970, 1990.
Zhang W, Wagner BJ, Ehrenman K, Schaefer AW, DeMaria CT, Crater D, DeHaven K, Long L, and Brewer G. Purification, characterization, and cDNA cloning of an AU-rich element RNA-binding protein, AUF1. Mol Cell Biol 13: 7652–7665, 1993.(Jill M. Schroeder, Hend Ibrahim, Lynn Ta)