The RNA-binding protein HuR binds to acetylcholinesterase transcripts and regulates their expression in differentiating skeletal muscle cells.

During myogenic differentiation, acetylcholinesterase (AChE) transcript levels are known to increase dramatically. Although this increase can be attributed in part to increased transcriptional activity, posttranscriptional mechanisms have also been implicated in the high levels of AChE mRNA in myotubes. In this study, we observed that transfection of a luciferase reporter construct containing the full-length AChE 3'-untranslated region (UTR) resulted in significantly higher (5-fold) luciferase activity in differentiated myotubes versus myoblasts. RNA-electrophoretic mobility shift assays (REMSAs) performed with a full-length AChE 3'-UTR probe and the AU-rich element revealed that the intensity of RNA-binding protein complexes increased as myogenic differentiation proceeded. Using several complementary approaches including supershift REMSA, mRNA-binding protein pull-down assays, and immunoprecipitation followed by reverse transcription-PCR, we found that the mRNA-stabilizing protein HuR interacts directly with AChE transcripts. Stable overexpression of HuR in C2C12 cells increased the expression of endogenous AChE transcripts as well as that of the luciferase reporter construct containing the AChE 3'-UTR. In vitro stability assays performed with protein extracts from these cells versus controls resulted in a slower rate of AChE mRNA decay. The down-regulation of HuR expression mediated through small interfering RNA further confirmed the role of HuR in the regulation of AChE mRNA levels. Taken together, these studies demonstrate that HuR interacts with the AChE 3'-UTR to regulate posttranscriptionally the expression of AChE mRNA during myogenic differentiation.

Neurotransmission at cholinergic synapses of the central and peripheral nervous systems is promptly terminated by the catalytic activity of the enzyme acetylcholinesterase (AChE) 1 (for review see Refs. [1][2][3][4]. In addition to this pivotal role in terminating synaptic transmission, several lines of evidence have shown that AChE assumes other functions necessary for the normal operation of the nervous system (for examples see Refs. [5][6][7]. Appropriate expression of AChE is not only important in the brain, but it is also critical for normal skeletal muscle activity as exemplified by the phenotype seen in patients suffering from a myasthenic syndrome linked to a reduction of AChE at the neuromuscular junction (8,9). Given the functional significance of AChE in the nervous system, it thus becomes important to identify the molecular and cellular mechanisms regulating its expression in developing tissues as well as in neurons and skeletal muscle cells placed under a variety of conditions known to markedly affect AChE expression.
AChE is encoded by a single gene that is alternatively spliced to produce catalytic subunits that can be assembled into different molecular forms. In skeletal muscle, the AChE T transcript is predominantly expressed (10 -13) resulting in the synthesis of a catalytic subunit that can associate with structural subunits encoded by separate genes (14). In particular, the asymmetric forms of AChE as well as AChE transcripts themselves are preferentially expressed at the level of the neuromuscular junction (14 -17). In mature muscle, AChE expression is tightly controlled by nerve-evoked electrical activity as well as by nerve-derived trophic factors (14, 18 -21). However, AChE is also known to be regulated during myogenic differentiation, at times preceding the occurrence of nerve-muscle interactions (22)(23)(24). Under these conditions, however, there is relatively little information available concerning the nature of the molecular mechanisms that control AChE expression.
In a recent study, we have shown that an increase in the transcriptional activity of the AChE gene could partially account for the initial increase in AChE transcript levels observed in differentiating muscle cells grown in culture (25). In this latter study, the contribution of E-and N-box motifs located within the first intronic region, along with their respective transcription factors, myogenin and GA-binding protein ␣ and -␤, was shown to be involved in mediating this transcrip-tional induction. More recent studies performed by others have also highlighted the contribution of specific 5Ј-regulatory regions in the AChE gene during myogenesis, including Sp1, Egr-1, and cAMP response elements (26,27). In our earlier study, however, we noted a clear discrepancy between the observed increase in transcription and the overall induction in AChE transcripts, thereby suggesting that transcription alone could not fully account for the increase in AChE expression seen at later stages of differentiation (25). In agreement with a previous report (23), our results indicated that posttranscriptional events also participate in regulating the abundance of AChE mRNAs in differentiating muscle cells (19). Despite the implication of posttranscriptional mechanisms, the nature of the specific trans-and cis-acting elements involved in these regulatory events remains unclear. With this in mind, we have therefore initiated a series of experiments in an attempt to characterize the posttranscriptional mechanisms that regulate AChE transcripts in differentiating muscle cells.

EXPERIMENTAL PROCEDURES
Cell Culture-Mouse C2C12 cells (ATCC, Manassas, VA) were cultured on culture dishes (6-well and 100-mm) coated with Matrigel (Collaborative Biomedical Products, Bedford, MA) in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 20% fetal bovine serum, 292 ng/ml L-glutamine, and 100 units/ml penicillin-streptomycin in a humidified chamber at 37°C with 5% CO 2 . Myogenic differentiation and fusion was induced by replacing the growth medium on confluent myoblasts with differentiation medium containing low serum (2% horse serum). Culture medium was changed every 48 h.
Plasmid Construction and Transfection Studies-The 3Ј-untranslated region (UTR) of the mouse ϳ2.4-kb AChE mRNA was isolated and subcloned into a Renilla luciferase reporter construct driven by the thymidine kinase promoter (phRG-TK from Promega, Madison, WI) as described previously (19,28). The shorter 3Ј-UTR was chosen for these studies because the first polyadenylation site is predominantly used in differentiating C2C12 cells (23). To generate the AChE 3Ј-UTR fragment without the AU-rich element (ARE), we employed a PCR-based protocol described elsewhere and used to delete the ARE from the GAP-43 mRNA (29). Briefly, the AChE 3Ј-UTR cDNA was used as a template, and two rounds of PCR amplification were first performed with the following primers: 5Ј-CGAGCCCCTAGCAGGGCTGGGATAT-AATACGACCGA-3Ј and the flanking primer 5Ј-AGGTCTCGGATCCT-TTATTGGCGGCCCAGAGGGGCGAAGG-3Ј (reaction 1); and the flanking primer 5Ј-AGGTCTCTCTAGACCCCTTGGGGAC CCCAG-3Ј and primer 5Ј-TATCCCAGCCCTGCTAGGGGCTCGGGCAGGGCGGCA-3Ј (reaction 2). To remove the internal ARE region (21 nucleotides), we performed a third round of PCR with the reaction 1 and reaction 2 PCR products using the two flanking primers listed above to yield a product containing the AChE 3Ј-UTR without the ARE region as confirmed by sequencing.
The mouse HuR coding region was amplified by PCR from mouse muscle cDNA using primers designed from the available mouse HuR sequence that spans the entire coding region. The resultant PCR product was subcloned into the pCI-neo vector (Promega) driven by the cytomegalovirus enhancer/promoter. The resulting vector (pHuR) product was sequenced and used to stably transfect C2C12 cells that were selected for G418 resistance resulting in a mouse HuR overexpressing C2C12 stable cell line pool.
Plasmid DNA was prepared using the Mega-Prep procedure (Qiagen, Mississauga, Ontario, Canada). DNA pellets were resuspended in 10 mM Tris-HCl, pH 8.5. Transfections were performed with the Lipofectamine reagent kit (Invitrogen) according to the manufacturer's instructions. Briefly, C2C12 myoblasts at 80% confluency were transfected with a total of 2 g of plasmid DNA (1.0 g of either the reporter construct or the parental vector and 1.0 g of the vector used for transfection efficiency). Transfection efficiency was monitored by cotransfecting the basic firefly luciferase construct (pGL3) or the basic LacZ reporter vector (30). Transfected myoblasts were collected 24 -48 h after transfection or were induced to differentiate and collected 48 -72 h later.
To determine reporter gene activity, transfected myoblasts and myotubes were washed with cold 1ϫ PBS, scraped, and lysed in 1ϫ reporter lysis buffer (Invitrogen) followed by several freeze-thaw cycles. The extracts were centrifuged (12,000 ϫ g for 10 min at 4°C), and the resulting supernatant was used for the luciferase assays (Renilla and firefly) or for ␤-galactosidase assays using the appropriate kits and according to the manufacturer's instructions (Invitrogen). Values obtained for the AChE reporter luciferase and control parental (phRG-TK) constructs were standardized to the values obtained with firefly luciferase or with ␤-galactosidase, thereby controlling for transfection efficiency.
HuR siRNA and Transfection Studies-The HuR normal and mutant siRNA duplexes (target sequence, 5Ј-AAAAGUCUGUUCAGCAG-CAUUGG-3Ј, and mutant sequence, 5Ј-AAAAGUCAAUUCAUCAG-CAAUGG-3Ј) were obtained from Dharmacon RNA Technologies. C2C12 myoblasts were transfected with siRNA duplexes designed to mouse HuR sequence using the RNAiFect transfection kit (Qiagen) according to the manufacturer's instructions. Transfected myoblasts were harvested within 24 -48 h, and RNA was extracted (see below) and used for reverse transcription (RT)-PCR (see below).
Adenoviral Construct and Infection Studies-An adenoviral construct containing the human HuD sequence (AdLS5a) was generated using the HuD sequence subcloned from the previously described pcHuD vector (31). AdLS5a contains the human HuD cDNA under the regulation of the cytomegalovirus promoter and bovine growth hormone polyadenylation sequence. In these viruses, the HuD expression cassette replaces the early region 1, and transcription is directed leftward, relative to the conventional human adenovirus serotype 5 map. A control virus, Ad-lacZ, contains the Escherichia coli ␤-galactosidase gene under the regulation of the murine cytomegalovirus promoter and simian virus 40 polyadenylation sequence, and transcription from the expression cassette is directed rightward. The early region 1-deleted, first generation adenoviral vectors used in these studies were constructed using a combination of conventional cloning techniques and RecAmediated recombination (32,33). Recombinant adenoviral helper viruses were grown and titered on 293 cells as described previously (34). 293 cells (35) were grown in monolayer in minimum essential medium supplemented with 100 units of penicillin/ml, 100 mg of streptomycin/ ml, 2.5 mg of fungizone/ml, and 10% fetal bovine serum (complete medium). All cell culture media and reagents were obtained from Invitrogen.
Confluent C2C12 myoblasts were infected with 9.5 ϫ 10 7 particles of the adenoviral construct in a volume of 100 l of culture medium, with a multiplicity of infection of 475. At this multiplicity of infection, almost all of the cells on the plate were infected. 2 The cells were placed in a humidified incubator at 37°C and 5% CO 2 for 1 h before the addition of a supplementary growth medium. During this period, the cell culture plates were agitated every 15 min to ensure that the HuD adenoviral construct was in contact with all of the cells in the dish. The culture medium was changed 24 h later, and myogenic differentiation was induced with a low serum culture medium. Infected myoblasts and myotubes were washed with 1ϫ PBS 24 -72 h following infection, and RNA was extracted and used for RT-PCR as described below.
RNA Extraction and RT-PCR-Total RNA was extracted from myoblasts and differentiating myotubes using 1.0 ml of TRIzol reagent (Invitrogen) according to the manufacturer's instructions and as described previously (25,28). Precipitated RNA was resuspended in RNase-free water and stored at Ϫ80°C until used. RNA from each sample was quantified using the Amersham Biosciences Gene Quant II RNA/DNA spectrophotometer and adjusted to a final concentration of 50 ng/l.
Reverse transcription of 4 l of RNA (200 ng) was performed using the Omniscript reverse transcription kit (Qiagen) as recommended by the manufacturer. Briefly, the RNA was reverse transcribed in a final volume of 20 l of RT master mix, 0.5 mM concentration of each dNTP, 10 units of RNase inhibitor (Applied Biosystems, Foster City, CA), 1 M random hexamers (Applied Biosystems), and 4 units of Omniscript reverse transcriptase. RT was carried out at 37°C for 60 min. Negative controls consisted of the same RT mixture in which the RNA was replaced with RNase-free water. All of the samples were prepared using the same RT master mix.
PCR was used to amplify the cDNAs corresponding to AChE, HuR, and S12 ribosomal protein (S12 hereafter) mRNAs. The primers for AChE, S12 (used as an internal control), and HuR (5Ј-primer CAGAG-GTCATCAAAGATGC and 3Ј-primer ATCCCACTCATGTGATCTAC) were synthesized based on available sequences and amplified products of 670, 368, and 394 bp, respectively (16, 17, 25, 36 -38). PCR was performed using HotStarTaq DNA polymerase (Qiagen) according to the manufacturer's instructions. All of the samples were prepared using the same PCR master mix. PCR cycling parameters consisted of an initial activation step at 95°C for 15 min followed by denaturation at 94°C for 1 min; annealing at 70°C (for AChE), 54°C (for S12), and 57.5°C (for HuR) for 1 min each; and extension at 70°C (for AChE) for 2 min and 72°C (for S12 and HuR) for 1 min followed by a 10-min elongation step at 72°C. PCR amplification was stopped during the linear range of amplification (16,17). Typically, the cycle numbers were 35 cycles for AChE, 24 cycles for S12, and 24 cycles for HuR. PCR products were visualized on ethidium bromide-stained 1.5% agarose gels and quantified using the fluorescent dye VistraGreen (Amersham Biosciences) in 1.5% agarose gels. Quantification was performed using the Storm PhosphorImager and the accompanying ImageQuant software (Amersham Biosciences). The values obtained for AChE and HuR were standardized relative to the corresponding level of S12 in the same sample.
In Vitro Stability Assay-Poly(A) ϩ RNA was isolated from a total RNA extract obtained from mouse brain using Oligotex mRNA mini kit (Qiagen) according to the manufacturer's instructions. Cytoplasmic protein fractions were prepared as described previously (39) from C2C12 cells stably transfected with the HuR cDNA (pHuR cells) or from cells stably transfected with the empty vector (pCI-neo) as control. Briefly, C2C12 cell pellets were homogenized in MOPS buffer (10 mM MOPS-NaOH, pH 7.2, 200 mM sodium chloride, 2.5 mM magnesium acetate) with 100 M dithiothreitol, 100 M phenylmethylsulfonyl fluoride, and 1 Complete Mini, EDTA-free Protease Inhibitor Cocktail tablet as per the manufacturer's recommendations (Roche Applied Science). Homogenates were centrifuged (10,000 ϫ g for 15 min at 4°C), and the resulting supernatants were stored until further use. After optimization of protein and RNA concentrations, 2.5 ng/l poly(A) ϩ RNA were incubated with 0.2 g/l protein extracts in MOPS buffer containing 1 mM ATP, 0.1 mM spermine, 2 mM dithiothreitol, and 1 unit/l SUPERasin (Ambion, Austin, TX) in a final volume of 180 l and incubated at 37°C. At intervals of 0, 15, and 30 min, aliquots of 40 l were removed, and the reaction stopped by adding 200 l of ice-cold phenol/chloroform. De-proteination and precipitation of the RNA was then carried out in the presence of yeast tRNA (10 g) as a carrier. AChE mRNA was detected by RT-PCR as described above.
In Vitro Transcription-cDNAs encoding for the AChE full-length 3Ј-UTR, the ARE region alone, and the AChE 3Ј-UTR minus the ARE were obtained by PCR amplification of the pGL3-3Ј-UTR plasmid as described previously (19,28). This cDNA was used as the template for the in vitro transcription of the 3Ј-UTR. In vitro transcription was performed with the T7 in vitro transcription system (Promega) as recommended by the manufacturer. Briefly 0.5 g of cDNA, 5 Ci of [␣-32 P]UTP, nucleotides, RNase inhibitor, and T7 polymerase were used to synthesize radiolabeled transcripts. The template PCR was digested with DNase I for 20 min at 37°C. The resulting radiolabeled RNA was purified using the RNase-free G-25 RNA purification column (Roche Applied Science). The integrity of the RNA was verified by gel electrophoresis. Biotin-labeled transcripts (AChE 3Ј-UTR with or without the ARE, and with the ARE region alone) were obtained using the same method except that 0.5 M bio-11-CTP (Sigma) was used instead of the radiolabeled UTP. Unlabeled probes were generated by the same method and used in competition assays.
Electrophoretic Mobility Shift Assay and Supershift-RNA-based electrophoretic mobility shift assays (REMSAs) and supershift assays were performed with total protein extracts obtained from myoblasts and myotubes. The cells were washed with 1ϫ PBS, scraped, and lysed in 500 l of homogenization buffer (0.3 M sucrose, 60 mM NaCl, 15 mM Tris, pH 8.0, 10 mM EDTA) supplemented with protease inhibitors (mini-complete protease inhibitor, Roche Applied Science) followed by several freeze-thaw cycles. The samples were centrifuged (15,000 ϫ g for 15 min at 4°C), and the resulting supernatants were stored at Ϫ80°C in 20 -100-g aliquots to be used for REMSAs and other assays. The total amount of protein present in the extracts was determined with the Coomassie protein assay kit (Pierce).
REMSAs were performed as described previously (28). In brief, 20 g of protein extract were incubated with 2 ϫ 10 5 cpm of 32 P-labeled AChE 3Ј-UTR or the ARE region in the presence of yeast tRNA (0.2 g/l) and heparin (2.5 mg/ml) for 20 min at room temperature. The mixture was separated by 5% native PAGE with 0.5ϫ Tris borate-EDTA running buffer. The gels were subsequently dried and exposed to film at Ϫ80°C. Competition assays were performed by incubating an excess of cold probe with the protein for 10 min prior to the incubation with the radiolabeled RNA. Supershift assays were performed by incubating the protein in the binding buffer with 0.5 g of mouse monoclonal antibody directed against HuR (Molecular Probes, Eugene, OR) or a nonspecific antibody at room temperature for 1 h prior to the addition of the radiolabeled transcripts.
Immunoprecipitation and AChE mRNA Analysis-Immunoprecipitation of HuR from myoblasts and myotubes was performed as previously described with some modifications (28,40). Myoblast and myotube protein extracts obtained as described above were precleared with 5 g of normal mouse IgG (Santa Cruz Biotechnology, Santa Cruz, CA) and 30 l of paramagnetic protein G Dynabeads (Dynal Biotech, Oslo, Norway) in 300 l of immunoprecipitation buffer (10 mM Tris-HCl, pH 7.5, 1% Igepal CA-630, 1% bovine serum albumin fraction V, 2 mM EDTA, 3 mM magnesium acetate, 100 M ZnCl 2 , 1 g/ml yeast tRNA) for 1 h at room temperature with constant mixing. The precleared protein extracts were then incubated with 2 g of a mouse monoclonal antibody directed to HuR (Molecular Probes) or normal mouse IgG (Santa Cruz Biotechnology) for 1 h at room temperature with constant mixing. The protein antibody solution was incubated with 30 l of prewashed paramagnetic protein G Dynabeads for 1 h at room temperature with constant mixing. The protein G Dynabeads were washed with immunoprecipitation buffer three times, and the protein and mRNA were extracted with 0.5 ml of TRIzol according to the manufacturer's instructions. The mRNA was precipitated in the presence of 10 g of RNase-free glycogen (GenHunter, Nashville, TN) and subsequently used for RT-PCR. The protein pellet was solubilized in 1ϫ PBS with 1% SDS and used for Western blotting (see below).
mRNA-binding Protein Pull-down Assay-Myoblast and myotube protein extracts (100 g) were incubated with 2 g of biotin-labeled AChE 3Ј-UTR with and without the ARE or with the ARE region alone, in 1ϫ binding buffer as used in REMSAs (20 mM Hepes, pH 7.5, 10 mM magnesium acetate, 50 mM potassium acetate, 1 mM dithiothreitol, 0.1 mM ZnCl 2 , 5% glycerol, 0.2 mg/ml yeast tRNA) for 30 min at room temperature with intermittent mixing. The biotin-labeled RNA and associated mRNA-binding proteins were incubated with 50 l of paramagnetic streptavidin Dynabeads (Dynal Biotech), prewashed in RNase-free buffers and 1ϫ binding buffer for 1 h at room temperature with constant mixing, and washed three times with 1ϫ binding buffer. The protein was solubilized in PBS and subjected to Western blot analysis for the detection of HuR.
Western Blot-For Western blotting, 25 g of myoblast and myotube protein extracts or proteins extracted from the mRNA-binding pulldown assays were denatured in SDS loading buffer and subjected to 3% SDS-PAGE. The separated proteins were transferred to nitrocellulose Trans-Blot transfer medium (Bio-Rad, Hercules, CA). Following transfer the membranes were blocked in either 10% nonfat skim milk or 10% bovine serum albumin in Tris-buffered saline and Tween 20 for 1 h at room temperature with constant agitation. The membranes were incubated with mouse monoclonal antibodies directed against HuR (1:1,000) or ␣-tubulin (1:5,000) (Sigma) followed by the appropriate secondary antibody (1:10,000 or 1:25,000). The proteins were revealed using the SuperSignal West Dura extended duration substrate enhanced chemiluminescence kit (Pierce).
Statistical Analysis-An analysis of variance was performed to evaluate the effect of the AChE 3Ј-UTR on luciferase activity in transfected C2C12 cells. The Fisher's least squares difference test was used to determine whether the differences seen between group means were significant. The level of significance was set at p Ͻ 0.05. Data are expressed as mean Ϯ S.E. throughout.

AChE mRNA Expression during Myogenic Differentiation-
In an initial series of experiments, we examined the levels of AChE mRNA during myogenic differentiation by RT-PCR. As shown in Fig. 1A, AChE mRNA levels are barely detectable in 50% confluent myoblasts and increase considerably as the cells become more confluent and begin to differentiate. Upon the second day of differentiation, AChE transcript levels increased by ϳ10-fold over the levels seen in 50% confluent myoblasts. The extent of this increase reached more than 50-fold in 4-day-old myotubes (Fig. 1B). This pattern of expression of AChE mRNA during myogenic differentiation is in agreement with earlier reports (23,25).
Previous studies have also shown that changes in the transcriptional activity of the AChE gene cannot solely account for the large increase in transcript levels seen in myotubes, therefore implicating posttranscriptional mechanisms such as those regulating transcript stability. Because elements located in the 3Ј-UTR are often implicated in the control of mRNA turnover (41, 42), we next sought to determine whether the AChE 3Ј-UTR plays a role in regulating the relative abundance of AChE transcripts in differentiating muscle cells. To this end, we engineered a luciferase reporter construct in which the endogenous 3Ј-UTR was replaced with the mouse AChE 3Ј-UTR terminating at the first polyadenylation signal (see Fig. 2A). The short 3Ј-UTR was chosen over the longer 3Ј-UTR because previous studies have demonstrated that the first polyadenylation signal is used preferentially over the second signal (23,43). As a result, there is significantly more of the ϳ2.4-kb AChE mRNA in C2C12 cells than the ϳ3.2-kb transcript. In addition, the short 3Ј-UTR was shown to be important in the regulation of AChE mRNA expression during neuronal differentiation of PC12 cells (28,44).
Transfection of this construct into myoblasts resulted in a low level of expression of the luciferase reporter (Fig. 2B). In 4-day-old myotubes, however, the activity level of the luciferase reporter increased nearly ϳ 5-fold. To show the specificity of this increase, we transfected a reporter construct containing the 3Ј-UTR of utrophin. As expected on the basis of recent studies showing a lack of posttranscriptional regulation of utrophin transcripts during myogenic differentiation (45,46), no change in reporter expression was observed between myoblasts and myotubes (data not shown). These results highlight the specificity of the increase in reporter expression obtained with the AChE 3Ј-UTR. The increased expression of the reporter gene under these conditions supports the notion that the 3Ј-UTR is involved in regulating AChE transcript levels during myogenic differentiation.
Pattern of Interactions between the AChE 3Ј-UTR and RNAbinding Proteins-We next examined the intensity and binding pattern of proteins found in myoblasts and myotubes that interact with the AChE 3Ј-UTR using REMSA. For these experiments, we used in vitro transcribed RNA that corresponded to either the full-length AChE 3Ј-UTR (FL) or to ARE. AREs are well characterized cis-acting elements found in UTRs that mediate transcript stability through interactions with destabilizing trans-acting factors such as AUF1 or with stabilizing trans-acting factors including those of the Hu family (41,47). REMSA performed with the FL probe revealed the presence of two major protein-RNA complexes (Fig. 3, indicated with black arrows) that formed in myoblasts and myotubes. Although there is no apparent change in the mobility of these bands, the binding intensity of both complexes is greatly increased in myotubes, suggesting that the abundance of mRNA-binding proteins is increased in myotubes and/or that their binding characteristics to the AChE 3Ј-UTR are affected. This apparent increase in binding activity in myotubes was observed in several independent experiments using different protein extracts (n ϭ 4). REMSA performed using the ARE probe also showed some important changes in the pattern of RNA-protein interactions in myoblasts versus myotubes (Fig. 3). In particular, there was a clear increase in the band intensity using proteins extracted from 4-day-old myotubes (Fig. 3, indicated with arrows). Here, again, the increased banding intensity obtained with differentiated myotubes was consistently observed (n ϭ 3). The specificity of the binding was demonstrated in competition experiments using a cold probe to the ARE (Fig. 3, ARE-C). Taken together, these results suggest that there are specific RNA-binding complexes that interact with the AChE 3Ј-UTR in which the ability to bind to the 3Ј-UTR increases according to the state of myogenic differentiation.
HuR Binds the AChE 3Ј-UTR-Our findings using REMSA illustrate the importance of ARE in the binding of cytoplasmic factors to the AChE 3Ј-UTR. Members of the ELAV-like family of mRNA-binding proteins, known to interact with ARE, have recently received an increasing amount of attention because of their ability to stabilize a variety of transcripts in cells placed under different conditions. In our studies, we focused on one family member, HuR, because it is expressed in skeletal muscle (48,49).
In an initial experiment, we performed supershift REMSA with an antibody directed to HuR and protein extracts from myoblasts and myotubes, to determine whether HuR was present in the protein complexes interacting with the AChE 3Ј-UTR. As shown in Fig. 4A, the addition of the HuR antibody to the RNA/protein extract mixture resulted in the supershift of an RNA-protein complex formed using myotube protein extracts. This supershift was mostly evident using the myotube extract. As a control, no supershift was apparent when a nonspecific antibody was used instead of the antibody directed against HuR.
The interaction between HuR and the AChE 3Ј-UTR was further confirmed using an mRNA-binding pull-down assay. In this assay, excess in vitro transcribed biotin-labeled AChE 3Ј-UTR probes (both FL and ARE) were incubated with protein extracts from myoblasts or myotubes and subsequently pulled out of solution along with any proteins that might bind to the probes with streptavidin-coated paramagnetic beads. The pulled-down proteins were then resolved by SDS-PAGE and used for Western blotting. A Western blot to detect the presence of HuR (Fig. 4B) showed that indeed HuR could be pulled down with both the AChE full-length 3Ј-UTR and the ARE in experiments using myotube extracts. As observed with the supershift REMSAs, the FL probe was not able to bind HuR proteins from myoblast extracts, although a small amount of HuR could be detected in pull-down assays using the ARE probe. To further examine the binding of HuR to the ARE, we performed an additional set of pull-down experiments using AChE 3Ј-UTR probes that included (Fig. 4C, ϩARE) or not (ϪARE) the ARE and myotube extracts. As shown in Fig. 4C, the ARE is essential for the binding of HuR to the AChE 3Ј-UTR because we observed a significant decrease in the binding of HuR to the mutant AChE 3Ј-UTR probe (ϪARE). In these assays, we confirmed that HuR was still present in the protein extract following incubation of the extract with the mutant AChE 3Ј-UTR probe that did not contain the ARE (data not shown). Together, the results of the supershift and pull-down assays clearly indicate that HuR expressed in myotubes can interact with the AChE 3Ј-UTR via the ARE.
In complementary experiments, we determined whether HuR could bind to endogenous AChE transcripts in living cells. For these assays, we immunoprecipitated HuR from myoblasts and myotubes, extracted total RNA, and performed RT-PCR to amplify AChE. As shown in Fig. 4D, AChE mRNAs were present in immunoprecipitates obtained from myotubes but not in those from myoblast extracts. As a control, AChE could not be amplified from RNA isolated following immunoprecipitation with normal mouse IgG (Fig. 4D) or protein G paramagnetic beads alone (data not shown).
HuR Increases the Expression of AChE mRNA-To further demonstrate that HuR has a direct effect on the level of AChE mRNA, we created stable C2C12 cells that overexpressed HuR (pHuR). As shown in Fig. 5A, endogenous HuR protein levels are increased by severalfold in HuR-overexpressing muscle cells. Stable transfections with the empty vector (pCI-neo), which was used as a control in these experiments, showed no effect on endogenous HuR levels. We next measured endogenous levels of AChE mRNA in these cells by RT-PCR. In undifferentiated myoblasts, we did not observe any significant changes in AChE mRNA levels between control and pHuR cells (data not shown, see "Discussion"). In 2-day differentiated myotubes, however, we observed an ϳ3-fold increase in the levels of AChE mRNA (Fig. 5, B and C). Similarly, siRNAs directed against HuR resulted, as seen previously (50), in a decrease in HuR that in our experiments was accompanied by an ϳ30% decrease in the levels of endogenous AChE transcripts. Finally, infection of C2C12 cells with an adenoviral vector encoding the

FIG. 4. The mRNA-binding protein HuR binds to the AChE 3-UTR in myotubes.
A, supershift REMSAs were performed with total protein extracts from myoblasts (MB) and myotubes (MT) and a monoclonal antibody directed to HuR (␣-HuR) or a nonspecific antibody (control (Ctl)). Representative autoradiograms demonstrating the interaction between in vitro transcribed FL RNA and protein complexes are shown. Arrow points to the supershifted RNA-protein complex obtained with the antibody directed to HuR and the myotube protein extract. Arrowhead points to the free probe. B, a Western blot for HuR performed with myoblast and myotube proteins pulled down by in vitro transcribed biotin-labeled FL or ARE transcripts demonstrates that HuR from these extracts can bind to the AChE 3Ј-UTR. C, a Western blot for HuR performed with myotube proteins pulled down using in vitro transcribed biotin-labeled AChE 3Ј-UTR transcripts with (ϩ) ARE or without (Ϫ) the ARE demonstrates that HuR binds to the AChE 3Ј-UTR via the ARE. D, example of an ethidium bromide-stained agarose gel displaying AChE PCR product obtained from immunoprecipitates isolated from myoblasts and myotubes using an antibody directed against HuR (H) or normal mouse IgG (I). Note the presence of a PCR product corresponding to AChE only in the immunoprecipitate obtained with the HuR antibody and myotube protein extract.
highly related HuD family member resulted in a further induction of AChE mRNA in myotubes (data not shown).
We next performed in vitro stability assays using protein extracts from control and pHuR cells and used poly(A) ϩ RNA from mouse brain to determine whether overexpression of HuR affected AChE mRNA stability. As shown in Fig. 6, A and B, the rate of AChE transcript decay was slower in the presence of protein extracts from the pHuR cells, thereby confirming the importance of posttranscriptional regulation in mediating the changes seen in steady-state AChE mRNA levels in these cells. Using these stably transfected cells, we also determined whether HuR, through its interaction with the AChE 3Ј-UTR, could increase the expression of the luciferase-AChE 3Ј-UTR reporter construct. For these experiments, the luciferase-AChE 3Ј-UTR reporter construct was transfected into control and pHuR stable C2C12 cells. As shown in Fig. 6C, expression of the reporter construct was increased by ϳ 2-3-fold in C2C12 cells that stably overexpressed HuR. In additional control experiments, we verified that the ARE plays an important role in this process because transfection of a reporter construct containing the mutant AChE 3Ј-UTR fragment without the ARE decreased significantly expression of the reporter. These results indicate that through its interaction with the ARE located in the 3Ј-UTR, HuR increases the expression of AChE transcripts in differentiating skeletal muscle cells.

DISCUSSION
In recent years, there has been considerable interest in characterizing the transcriptional and posttranscriptional mechanisms that regulate AChE expression during myogenic differentiation (23, 25-27, 51, 52). In contrast to the recent progress made in the elucidation of some of the transcriptional events (see the Introduction), there is essentially no information available concerning the nature of the specific trans-and cis-acting factors that participate in the posttranscriptional regulation of AChE in differentiating skeletal muscle cells. In the present study, we have therefore sought to begin the characterization of these posttranscriptional events that act to stimulate the production and maintenance of high levels of AChE transcripts during myogenic differentiation. Transfection experiments per-formed using a luciferase reporter construct in C2C12 cells showed the important contribution of the 3Ј-UTR in controlling AChE transcript levels. Furthermore, we found that the AChE 3Ј-UTR interacts with several distinct complexes using muscle protein extracts and that the intensity of the binding clearly increases during differentiation. Using a combination of different assays, we established that the stabilizing mRNA-binding protein HuR binds to the AChE 3Ј-UTR. Finally, we have shown a functional role for HuR in increasing AChE mRNA levels in muscle cells via stabilization through its 3Ј-UTR.
The Hu Family of Proteins Regulates AChE Expression at the Posttranscriptional Level-The role of posttranscriptional regulatory mechanisms has recently been implicated as an important component of the molecular events regulating AChE transcript levels (23,25). In recent work, we observed for example a discrepancy between the increase in transcription and the extent of AChE mRNA induction during muscle differentiation (25). Here, we show that the presence of the AChE 3Ј-UTR resulted in an increase in the expression of a reporter construct in differentiating muscle cells, thereby demonstrating that FIG. 5. AChE mRNA levels are modulated by the level of expression of HuR. A, example of a Western blot for HuR depicting endogenous HuR levels in the pCI-neo control (Ctl) and HuR overexpressing stable C2C12 cells (pHuR). B, example of an ethidium bromide-stained agarose gel displaying AChE and S12 PCR products obtained from two-day differentiated stable cells. C, quantification of AChE mRNA levels in the control and HuR-overexpressing myotubes expressed in arbitrary units following standardization to S12. The asterisk indicates a significant difference from control (p Ͻ 0.0004; n ϭ 3 independent experiments).

FIG. 6. HuR overexpression increases AChE mRNA stability.
In vitro stability assays were performed with protein extracts from pCI-neo control (Ctl) and HuR-overexpressing (pHuR) C2C12 cells and poly(A) ϩ RNA from mouse brain. A, example of ethidium bromidestained agarose gels displaying AChE and S12 PCR products following 0, 15, and 30 min of incubation with protein extracts. Note that there is a greater amount of AChE PCR product remaining after the 30-min incubation with the pHuR protein extract than with the control extract. B, quantitation of the PCR products remaining following incubations and expressed as a percentage of the initial amount found at time 0. C, control and pHuR cells were co-transfected with AChE 3Ј-UTR reporter or parental constructs and a firefly luciferase construct (to measure transfection efficiency). Luciferase activity obtained in control and pHuR cells was quantified and standardized to the activity obtained with the parental vector and the firefly luciferase construct and expressed in arbitrary units. The asterisk indicates a significant difference from control (p Ͻ 0.0081; n ϭ 2 independent experiments done in triplicate). posttranscriptional mechanisms indeed regulate the expression of AChE during myogenic differentiation through the 3Ј-UTR.
It is well established that transcript stability is governed by the presence of the 5Ј-cap, the 3Ј-poly(A) ϩ tail, and mRNA-binding proteins that bind to specific sequences or structures in the 3Ј-UTR. In skeletal muscle, recent studies have found that in response to different stimuli, the stability of certain transcripts, including vascular endothelial growth factor and cytochrome c oxidase (COX), is altered (49,53). During myogenic differentiation, transcripts encoding the myogenic factors myoD and myogenin display increased stability that results, at least in part, from the specific binding of the mRNA-binding protein HuR (50,54). The increase in the stability of these transcripts is important in dictating, in turn, the overall levels of proteins encoded by these mRNAs during myogenic differentiation.
HuR is a member of the ELAV (embryonic lethal abnormal vision)-like family of mRNA-binding proteins that includes the neuronally expressed HuB (HelN1), HuC, and HuD proteins (48). These proteins contain three RNA recognition motifs that are known to bind AREs: the AUUUA general sequence, U-rich sequences, and CU-rich motifs (55,56). The function of this family of mRNA-binding proteins is to stabilize transcripts by inhibiting the de-adenylation and targeting of the transcripts to the exosome or by blocking the activity of specific endonucleases that recognize the AREs (57-59). Using several complementary approaches, we have shown here that HuR can directly interact with AChE transcripts via their 3Ј-UTR. Additionally, overexpression of HuR in stably transfected C2C12 cells, as well as down-regulation of HuR via siRNA, clearly affected the endogenous levels of AChE mRNA through the 3Ј-UTR. In parallel experiments, we have also found that HuD, although not normally expressed in muscle, can increase AChE transcript levels in differentiated cells. This indicates that the Hu proteins share similar target preferences. Thus, AChE joins a growing family of transcripts that contain varying types of AREs that are stabilized by the Hu family of proteins in several distinct tissues (40,41,49,60,61).
Transcriptional and Posttranscriptional Regulation of AChE-Myogenic differentiation is a complex process involving the expression of several genes including those encoding proteins found at the neuromuscular junction (for review see Refs. [62][63][64][65]. Previous studies have established that the expression of genes encoding synaptic proteins in muscle is most often induced by myogenic regulatory factors such as myoD and myogenin (63). As mentioned above, recent studies have demonstrated that HuR binds to elements in the 3Ј-UTR of myoD and myogenin during myogenic differentiation, resulting in an increase in the stability of these transcripts (50,54). Thus, transcription factors that are key to the process of myogenic differentiation are themselves regulated posttranscriptionally by HuR (50,54).
In the present study, we found that AChE, for which transcriptional regulation during myogenic differentiation is known to involve myogenin and an E-box (25,26), is also a direct posttranscriptional target for the stabilizing effects of HuR. Together, the results of these and other studies show that AChE expression is regulated by increases in transcription as well as by increases in transcript stability during myogenic differentiation (28). In this context, our preliminary studies performed with pHuR stable C2 cells have shown that overexpression of HuR does increase the expression of an AChE promoter-reporter construct. 3 Accordingly, combined with ear-lier studies that focused on the role of HuR in myogenic differentiation, our findings further indicate therefore that HuR assumes a key regulatory function: (i) by controlling the stability of mRNAs encoding transcription factors known to modulate the transcriptional activity of the AChE gene during myogenic differentiation; and (ii) by also directly controlling the stability of AChE transcripts. Such dual roles for HuR in differentiating myogenic cells would confer to this RNA-binding protein a master regulatory function necessary to optimize gene expression via both transcriptional and posttranscriptional events (see further discussion in Refs. 60 and 66 -68).
Additional Factors Regulate AChE Expression Posttranscriptionally-On the basis of our studies, however, it appears that HuR is not the sole factor important for modulating the abundance of AChE transcripts in differentiating myotubes at the posttranscriptional level, as suggested by the 50 -100-fold increase in AChE mRNA versus the ϳ 3-fold increase caused by HuR. Indeed, we observed the formation of different protein complexes with both the full-length and the ARE probes using extracts from myoblasts and myotubes. The 3Ј-UTR of transcripts often behave in a fashion similar to 5Ј-regulatory regions of genes because several trans-acting factors with different properties may bind with different affinities to form active complexes that affect the stability of specific transcripts. For instance, the 3Ј-UTR of cyclooxygenase-2, GAP-43, and BC1 are all known to associate with a number of different mRNA-binding proteins (31, 61, 69 -71). In addition, the same cis-acting elements within the 3Ј-UTR, including the ARE, can also interact with different trans-acting factors that have contrary effects such as the predominantly destabilizing factors AUF1 and tristetraprolin and the stabilizing factor HuR (61,72,73). In some cases, the level of expression of the different factors and their binding affinity may determine the preferential level of interaction of one factor versus another that can result in dramatically opposite effects (74). Therefore, HuR may represent only one of the several RNA-binding proteins that can interact with AChE transcripts to modulate their expression during myogenic differentiation.
In addition to the myriad of trans-acting factors that are expressed in cells under specific conditions, there are other elements that also need to be considered because they may regulate the binding ability and specificity of these RNA-binding proteins. In this context, one intriguing finding in our study was the inability of HuR from myoblasts to bind exogenous AChE 3Ј-UTR, despite the presence of HuR in these cells (data not shown). In fact, RT-PCR assays and Western blotting experiments revealed that HuR is equally abundant in myoblasts and myotubes (data not shown) as was also observed by others (see also Ref. 50). This indicates therefore that additional regulatory events favor HuR-RNA interactions in myotubes. One possibility involves the cellular localization of HuR in myoblasts versus myotubes. Previous studies have demonstrated that HuR is predominantly localized in nuclei in myoblasts and that it shuttles out to the cytoplasm during myogenesis where it can interact with and stabilize specific transcripts (50,54). Because in our study we used total protein extracts containing equivalent amounts of HuR protein and found that HuR from myoblast extracts was unable to bind the AChE 3Ј-UTR, we speculate that there must be some modifications or other factors that affect the ability of HuR to bind to exogenous, in vitro transcribed, AChE transcripts or to those already present in the cytoplasm. The regulation of HuR binding could indeed be mediated by posttranslational modifications, such as phosphorylation or methylation. In this context, metabolic stress was recently shown to regulate the cytoplasmic localization of HuR and its binding through activation of the AMP-activated kinase (75). In addition, HuR has been shown recently to be methylated by the arginine methyltransferase CARM1 (co-activatorassociated arginine methyltransferase) in the hinge region adjacent to the novel nuclear localization signal, resulting in a nuclear-cytoplasmic shuttling ability and/or enhanced interactions with other proteins or transcripts (76,77). Therefore, although our study clearly highlights the important role of HuR in the regulation of AChE mRNA in differentiating myogenic cells, there are several aspects relevant for our complete understanding of the nature of these posttranscriptional regulatory events that are obviously complex and that will hence necessitate further experimentation.