Portal de Periódicos da CAPES

Post-transcriptional Regulation of RNase-L Expression Is Mediated by the 3′-Untranslated Region of Its mRNA

2007; Elsevier BV; Volume: 282; Issue: 11 Linguagem: Inglês

10.1074/jbc.m607939200

1083-351X

Xiaoling Li, Jesper B. Andersen, Heather J. Ezelle, Gerald M. Wilson, Bret A. Hassel,

RNA regulation and disease

RNase-L mediates critical cellular functions including antiviral, pro-apoptotic, and tumor suppressive activities; accordingly, its expression must be tightly regulated. Little is known about the control of RNASEL expression; therefore, we examined the potential regulatory role of a conserved 3′-untranslated region (3′-UTR) in its mRNA. The 3′-UTR mediated a potent decrease in the stability of RNase-L mRNA, and of a chimeric β-globin-3′-UTR reporter mRNA. AU-rich elements (AREs) are cis-acting regulatory regions that modulate mRNA stability. Eight AREs were identified in the RNase-L 3′-UTR, and deletion analysis identified positive and negative regulatory regions associated with distinct AREs. In particular, AREs 7 and 8 served a strong positive regulatory function. HuR is an ARE-binding protein that stabilizes ARE-containing mRNAs, and a predicted HuR binding site was identified in the region comprising AREs 7 and 8. Co-transfection of HuR and RNase-L enhanced RNase-L expression and mRNA stability in a manner that was dependent on this 3′-UTR region. Immunoprecipitation demonstrated that RNase-L mRNA associates with a HuR containing complex in intact cells. Activation of endogenous HuR by cell stress, or during myoblast differentiation, increased RNase-L expression, suggesting that RNase-L mRNA is a physiologic target for HuR. HuR-dependent regulation of RNase-L enhanced its antiviral activity demonstrating the functional significance of this regulation. These findings identify a novel mechanism of RNase-L regulation mediated by its 3′-UTR. RNase-L mediates critical cellular functions including antiviral, pro-apoptotic, and tumor suppressive activities; accordingly, its expression must be tightly regulated. Little is known about the control of RNASEL expression; therefore, we examined the potential regulatory role of a conserved 3′-untranslated region (3′-UTR) in its mRNA. The 3′-UTR mediated a potent decrease in the stability of RNase-L mRNA, and of a chimeric β-globin-3′-UTR reporter mRNA. AU-rich elements (AREs) are cis-acting regulatory regions that modulate mRNA stability. Eight AREs were identified in the RNase-L 3′-UTR, and deletion analysis identified positive and negative regulatory regions associated with distinct AREs. In particular, AREs 7 and 8 served a strong positive regulatory function. HuR is an ARE-binding protein that stabilizes ARE-containing mRNAs, and a predicted HuR binding site was identified in the region comprising AREs 7 and 8. Co-transfection of HuR and RNase-L enhanced RNase-L expression and mRNA stability in a manner that was dependent on this 3′-UTR region. Immunoprecipitation demonstrated that RNase-L mRNA associates with a HuR containing complex in intact cells. Activation of endogenous HuR by cell stress, or during myoblast differentiation, increased RNase-L expression, suggesting that RNase-L mRNA is a physiologic target for HuR. HuR-dependent regulation of RNase-L enhanced its antiviral activity demonstrating the functional significance of this regulation. These findings identify a novel mechanism of RNase-L regulation mediated by its 3′-UTR. The control of mRNA stability is a potent regulatory mechanism, as small changes in mRNA half-life result in dramatic changes in the mRNA available for translation into functional protein. Such post-transcriptional regulation provides a means to rapidly alter gene expression in response to diverse stimuli, and is frequently observed in genes encoding proteins with essential cellular functions such as proliferation and stress response. One of the most extensively studied RNA decay pathways involves AU-rich elements (AREs) 2The abbreviations used are: ARE, AU-rich element; IFN, interferon; 3′-UTR, 3′-untranslated region; 2–5A, 2′,5′-oligoadenylate; AREBP, ARE-binding protein; RNP, ribonucleoprotein; IP, immunoprecipitation; TTP, tristetraprolin; HuR, Hu RNA binding protein family member; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; qPCR, quantitative PCR; RT, reverse transcriptase; TIAR, TIA-1-related. present in the 3′-untranslated region (3′-UTR) of mRNAs (1Barreau C. Paillard L. Osborne H.B. Nucleic Acids Res. 2005; 33: 7138-7150Crossref PubMed Scopus (788) Google Scholar). The AUUUA sequence is a loose consensus that is found in many but not all AREs; the absence of a strict sequence motif likely reflects the importance of structural components in ARE function (2Fialcowitz E.J. Brewer B.Y. Keenan B.P. Wilson G.M. J. Biol. Chem. 2005; 280: 22406-22417Abstract Full Text Full Text PDF PubMed Scopus (66) Google Scholar). The influence of AREs on mRNA turnover occurs through an interaction with ARE-binding proteins (AREBPs) that can function to destabilize or stabilize target mRNAs by modulating access to the decay machinery. Specifically, AREs modulate deadenylation, and decapping, two critical steps in mRNA decay (3Ford L.P. Watson J. Keene J.D. Wilusz J. Genes Dev. 1999; 13: 188-201Crossref PubMed Scopus (219) Google Scholar, 4Gao M. Wilusz C.J. Peltz S.W. Wilusz J. EMBO J. 2001; 20: 1134-1143Crossref PubMed Scopus (135) Google Scholar); consistent with a link between mRNA stability and translation, some AREBPs function to regulate translation (5Lopez de Silanes I. Galban S. Martindale J.L. Yang X. Mazan-Mamczarz K. Indig F.E. Falco G. Zhan M. Gorospe M. Mol. Cell. Biol. 2005; 25: 9520-9531Crossref PubMed Scopus (191) Google Scholar, 6Mazan-Mamczarz K. Galban S. Lopez de Silanes I. Martindale J.L. Atasoy U. Keene J.D. Gorospe M. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 8354-8359Crossref PubMed Scopus (378) Google Scholar). Several AREBPs have been identified, and genetic manipulation of specific AREBPs in mice, such as HuR and tristetraprolin, resulted in dramatic and complex phenotypes that are associated with inflammation and stress response (7Katsanou V. Papadaki O. Milatos S. Blackshear P.J. Anderson P. Kollias G. Kontoyiannis D.L. Mol. Cell. 2005; 19: 777-789Abstract Full Text Full Text PDF PubMed Scopus (200) Google Scholar, 8Taylor G.A. Carballo E. Lee D.M. Lai W.S. Thompson M.J. Patel D.D. Schenkman D.I. Gilkeson G.S. Broxmeyer H.E. Haynes B.F. Blackshear P.J. Immunity. 1996; 4: 445-454Abstract Full Text Full Text PDF PubMed Scopus (655) Google Scholar). These phenotypes underscore the critical role of AREBPs in gene regulation in physiological and pathophysiological processes. It is estimated that 5–8% of the mRNAs in the human transcriptome contain putative ARE elements (9Bakheet T. Williams B.R. Khabar K.S. Nucleic Acids Res. 2006; 34: D111-D114Crossref PubMed Scopus (276) Google Scholar). The current challenge lies in determining the specific mRNA targets of AREBPs that are responsible for the phenotypes associated with ARE-mediated regulation. RNase-L is the terminal component of an RNA decay pathway known as the 2–5A system that derives its name from the 2′,5′-linked oligoadenylates (pppA(2′p5′A)n, n ≥ 2) that are required for its activation (10Silverman R.H. Biochemistry. 2003; 42: 1805-1812Crossref PubMed Scopus (140) Google Scholar). A family of 2′,5′-oligoadenylate synthetases are activated by double-stranded RNA to polymerize ATP into 2–5A. 2–5A, in turn, binds the latent RNase-L resulting in its dimerization and activation. Activated RNase-L cleaves single-stranded viral, ribosomal, and mRNAs with a preference for UU and UA sequences. In addition to its nuclease function, a role for RNase-L in translational regulation was recently reported (11Le Roy F. Salehzada T. Bisbal C. Dougherty J.P. Peltz S.W. Nat. Struct. Mol. Biol. 2005; 12: 505-512Crossref PubMed Scopus (55) Google Scholar). RNase-L activity is attenuated by the inactivation of 2–5A by cellular phosphatases and a 2′-phosphodiesterase (12Kubota K. Nakahara K. Ohtsuka T. Yoshida S. Kawaguchi J. Fujita Y. Ozeki Y. Hara A. Yoshimura C. Furukawa H. Haruyama H. Ichikawa K. Yamashita M. Matsuoka T. Iijima Y. J. Biol. Chem. 2004; 279: 37832-37841Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar), and by RLI, a protein inhibitor of RNase-L (13Bisbal C. Martinand C. Silhol M. Lebleu B. Salehzada T. J. Biol. Chem. 1995; 270: 13308-13317Abstract Full Text Full Text PDF PubMed Scopus (208) Google Scholar). The 2–5A system is an established mediator of interferon (IFN)-induced antiviral activity, and genetic approaches to manipulate RNase-L expression and activity revealed that it exerts potent antiproliferative, proapoptotic, and senescence-inducing activities independent of IFN treatment and virus infection (14Hassel B.A. Zhou A. Sotomayor C. Maran A. Silverman R.H. EMBO J. 1993; 12: 3297-3304Crossref PubMed Scopus (247) Google Scholar, 15Zhou A. Paranjape J. Brown T.L. Nie H. Naik S. Dong B. Chang A. Trapp B. Fairchild R. Colmenares C. Silverman R.H. EMBO J. 1997; 16: 6355-6363Crossref PubMed Scopus (457) Google Scholar, 16Castelli J.C. Hassel B.A. Wood K.A. Li X.L. Amemiya K. Dalakas M.C. Torrence P.F. Youle R.J. J. Exp. Med. 1997; 186: 967-972Crossref PubMed Scopus (228) Google Scholar). Consistent with a broader role for RNase-L as a natural constraint on cell proliferation, RNase-L was determined to function as a tumor suppressor by mapping of the hereditary prostate cancer-1 susceptibility allele (HPC1) to the RNASEL gene locus (17Carpten J. Nupponen N. Isaacs S. Sood R. Robbins C. Xu J. Faruque M. Moses T. Ewing C. Gillanders E. Hu P. Bujnovszky P. Makalowska I. Baffoe-Bonnie A. Faith D. Smith J. Stephan D. Wiley K. Brownstein M. Gildea D. Kelly B. Jenkins R. Hostetter G. Matikainen M. Schleutker J. Klinger K. Connors T. Xiang Y. Wang Z. De Marzo A. Papadopoulos N. Kallioniemi O.P. Burk R. Meyers D. Gronberg H. Meltzer P. Silverman R. Bailey-Wilson J. Walsh P. Isaacs W. Trent J. Nat. Genet. 2002; 30: 181-184Crossref PubMed Scopus (432) Google Scholar), and by the association of RNASEL mutations with the disease (18Casey G. Neville P.J. Plummer S.J. Xiang Y. Krumroy L.M. Klein E.A. Catalona W.J. Nupponen N. Carpten J.D. Trent J.M. Silverman R.H. Witte J.S. Nat. Genet. 2002; 32: 581-583Crossref PubMed Scopus (251) Google Scholar). Most recently, a novel γ-retrovirus was detected at a high frequency in prostate cancer patients that were homozygous for the Arg462 → Gln RNase-L mutation, suggesting a functional overlap between its antiviral and tumor suppressor activities (19Urisman A. Molinaro R.J. Fischer N. Plummer S.J. Casey G. Klein E.A. Malathi K. Magi-Galluzzi C. Tubbs R.R. Ganem D. Silverman R.H. Derisi J.L. PloS Pathog. 2006; 2: e25Crossref PubMed Scopus (496) Google Scholar). A few cellular mRNAs that are degraded in an RNase-L-dependent manner have been identified, but a direct substrate relationship with RNase-L, and a role in mediating RNase-L antiproliferative activities, has not been established for any candidate substrate (20Lewis J.A. Huq A. Najarro P. J. Biol. Chem. 1996; 271: 13184-13190Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar, 21Le Roy F. Bisbal C. Silhol M. Martinand C. Lebleu B. Salehzada T. J. Biol. Chem. 2001; 276: 48473-48482Abstract Full Text Full Text PDF PubMed Scopus (74) Google Scholar, 22Chandrasekaran K. Mehrabian Z. Li X.L. Hassel B. Biochem. Biophys. Res. Commun. 2004; 325: 18-23Crossref PubMed Scopus (25) Google Scholar). Microarray analysis of 2–5A-induced gene expression in prostate cancer cells (23Malathi K. Paranjape J.M. Bulanova E. Shim M. Guenther-Johnson J.M. Faber P.W. Eling T.E. Williams B.R. Silverman R.H. Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 14533-14538Crossref PubMed Scopus (90) Google Scholar) and human diploid fibroblasts 3J. B. Andersen, X. Li, and B. A. Hassel, unpublished data. identified a finite number of down-regulated genes that represent candidate RNase-L substrates, and determined that a significant number of mRNAs were up-regulated following RNase-L activation. These findings suggest that the biological activities of RNase-L involve an extensive reprogramming of one or more gene regulatory networks. As a mediator of the cellular response to microbial and antiproliferative stress stimuli, RNase-L must be rapidly activated then efficiently attenuated, to prevent the deleterious effects of its uncontrolled activity in cells. However, apart from activation by 2–5A, little is known about the regulation of RNase-L activity. 2′,5′-Oligoadenylate synthetase transcription is markedly induced by IFN, antiproliferative agents, or cell stress, providing a source of 2–5A in physiological conditions that require RNase-L activity. In contrast, RNase-L mRNA and protein are present at low basal levels in most cell types, and its enzyme activity is not associated with significant changes in RNase-L transcription (24Zhou A. Molinaro R.J. Malathi K. Silverman R.H. J. Interferon Cytokine Res. 2005; 25: 595-603Crossref PubMed Scopus (40) Google Scholar). These findings suggested that RNase-L expression may be rapidly and transiently regulated by post-transcriptional mechanisms. Specifically, the presence of a conserved, 1.75-kb 3′-UTR in the RNase-L transcript prompted us to examine the role of RNase-L mRNA stability in the regulation of its expression. The RNase-L 3′-UTR contained multiple candidate AREs, and sequential deletion of these elements identified positive and negative regulatory regions. RNP (ribonucleoprotein) immunoprecipitation revealed that the AREBP, HuR, binds to the RNase-L mRNA in intact cells, and functions to increase its mRNA half-life and expression in a manner that is dependent on the two 3′-terminal AREs. Activation of endogenous HuR in response to stress stimuli, or during myoblast differentiation, corresponded to an increase in RNase-L expression and HuR-RNase-L mRNA interaction, suggesting that HuR regulates RNase-L in physiological conditions. In addition, the HuR-dependent regulation of RNase-L expression enhanced its antiviral activity demonstrating the capacity of this regulation to impact RNase-L function. These findings provide the first description of the post-transcriptional regulation of RNase-L expression by its 3′-UTR and HuR, and identify a novel mechanism by which the activity of the 2–5A pathway is controlled. Cell Culture, Transfection, and Expression Constructs—293T, HeLa, 2fTGH, and C2C12 cells were grown in Dulbecco's modified Eagle's medium containing 10% fetal calf serum and 1× antibiotic-antimycotic (Invitrogen). Cells were maintained in a humidified atmosphere of 5% CO2, 95% balanced air at 37 °C. To induce differentiation C2C12 myoblasts cells were grown to 80–90% confluence, then transferred to Dulbecco's modified Eagle's medium containing 2% fetal calf serum and Insulin-Transferrin-Selenium-A (Invitrogen) for up to 5 days. Transfection experiments were carried out using Lipofectamine 2000 as directed by the supplier (Invitrogen). RNase-L-3′-UTR expression constructs and deletion mutants used for transfection included the previously described RNASEL cDNA (25Zhou A. Hassel B.A. Silverman R.H. Cell. 1993; 72: 753-765Abstract Full Text PDF PubMed Scopus (463) Google Scholar), and are described below. The HuR expression construct was a generous gift from Myriam Gorospe (NIA, National Institutes of Health). pTRE-r βG-3′-UTR construct was generated by sequential PCR amplification of the RNase-L 3′-UTR with primers that added flanking BglII sites, cloning of the PCR product into the pCR-blunt-Topo vector (Invitrogen), and subcloning of the BglII 3′-UTR insert downstream of the rabbit β-globin gene in the pTRER β-wt vector (2Fialcowitz E.J. Brewer B.Y. Keenan B.P. Wilson G.M. J. Biol. Chem. 2005; 280: 22406-22417Abstract Full Text Full Text PDF PubMed Scopus (66) Google Scholar). The pTRE-r βG-3′-UTR construct used in the experiments presented contained the poly(A) signals from the RNase-L 3′-UTR and the vector; however, the presence of an extra poly(A) signal did not alter β-globin-UTR expression or mRNA stability, as a construct that contained only the vector-derived poly(A) signal gave equivalent expression (data not shown). All clones were verified by sequencing. Cloning of the RNase-L 3′-UTR and Generation of 3′-UTR Mutants—The human RNase-L 3′-UTR was amplified from human genomic DNA using primers that flanked the stop codon and the predicted polyadenylation site: forward, 5′-GTCCCTGGCATCGTGTATTCCATA-3′; reverse, 5′-CTCTCACTACATACTAGGCCCACA-3′, and cloned in to pCR-blunt-Topo. The PCR product was subcloned into the unique SpeI site in the 3′-UTR of clone ZC5 in the pcDNA3.1 vector (25Zhou A. Hassel B.A. Silverman R.H. Cell. 1993; 72: 753-765Abstract Full Text PDF PubMed Scopus (463) Google Scholar) to generate the full-length RNase-L construct (ZC5+, Fig. 1). To remove the poly(A) signal from the ZC5+ construct, the plasmid was amplified using forward, 5′-CACCATGGAGAGCAGGGATCAT-3′, and reverse, 5′-GAGTAAGCTTCACAAATGGGC-3′, primers, and subcloned in the pcDNA3.1 vector. To remove the poly(A) signal from the pTRE-rβG-3′-UTR construct, the forward primer, 5′-AGATCTGGACTGATTTGCTGGAG-3′, and reverse primer, 5′-AGATCTGAGTAAGCTTCACAAATGGGC-3′, were used and the PCR product was subcloned into the BglII sites of the parental plasmid. The cloning junctions and 3′-UTR were verified by sequencing (University of Maryland, Baltimore Biopolymer Core). The primers used for PCR in Fig. 1B to demonstrate that the 3′-UTR is contiguous with the coding region, and that the predicted poly(A) site is used are listed according to the PCR products A–D shown in Fig. 1B. PCR products A, B, and D were: forward primer, 5′-CTACCAGAACACTGTGGGTGAT-3′; reverse primer for A, 5′-CATATGCAGCATTAGGGGTCAA-3′; reverse primer for B, 5′-GAATGAGATTCCTGGAACCCCT3′; reverse primer for D, 5′-CTCTCACTACATACTAGGCCCACA-3′. For the PCR product C: forward primer, 5′-CTGGCCCAAGATTATTCATACCTAGCACTTTATAAATTTATG-3′, and reverse primer, 5′-GCACCAGAAAAACGTAAGACAG-3′. RNase-L 3′-UTR deletion constructs were generated from a ZC5+ template by PCR with pfuHF DNA Polymerase. (Stratagene, La Jolla, CA). The forward primer for all the deletion constructs was ZC5F, 5′-CACCATGGAGAGCAGGGATCAT-3′; the reverse primers for these constructs were: Δ7–8, 5′-GCACCAGAAAAACGTAAGACAG-3′; Δ5–8, 5′-CTTCTTCAGACTCTGCCAAATG-3′; Δ4–8, 5′-TATGTTTTGGGCCTCATCTGGA-3′; Δ2–8, 5′-AGCTCACACTCTCTGAGTCTCA-3′; Δ1–8, 5′-CATATGCAGCATTAGGGGTCAA-3′. Internal deletions of AREs 2, 3, 7, 8, and the predicted HuR binding site were generated by PCR with ZC5+ as a template (QuikChange, Stratagene). The primers used for mutagenesis were: ΔARE2, forward primer, 5′-CTGGCCCAAGATTATTCATACCTAGCACTTTATAAATTTATG-3′, and reverse primer, 5′-CATAAATTTATAAAGTGCTAGGTATGAATAATCTTGGGCCAG-3′; ΔARE3, forward primer 5′-CCTAGCACTTTATAATGTGGTGTTATTGGTACC-3′, reverse primer, 5′-GGTACCAATAACACCACATTATAAAGTGCTAGG-3′; ΔARE2&3, forward primer, 5′-CTGGCCCAAGATTATTCATGTGGTGTTATTGGTACC-3′, reverse primer, 5′-GGTACCAATAACACCACATGAATAATCTTGGGCCAG-3′; ΔARE7, forward primer, 5′-GTATACATTACATCTGAGTCAAAACAATCCTTTAAGGTC-3′; reverse primer, 5′-GACCTTAAAGGATTGTTTTGACTCAGATGTAATGTATAC-3′; ΔARE8, forward primer, 5′-GTTGATTAGGAACAAAGGCTTAAAAAATAC-3′; reverse primer, 5′-GTATTTTTTAAGCCTTTGTTCCTAATCAAC-3′. ΔHuR forward primer, 5′-GTGGTGGTTGAGATGGAGCCAGTACCTTAGGTTCTTTCTG-3′; reverse primer, 5′-CAGAAAGAACCTAAGGTACTGGCTCCATCTCAACCACCAC-3′. Analysis of RNA Expression and Half-life by Quantitative Reverse Transcriptase (RT)-PCR—Total RNA was prepared using TRIzol reagent as directed by the supplier (Invitrogen). RNA was treated with DNase I (Promega, Madison, WI) at 37 °C for 30 min, followed by the addition of 1 μl of stop solution at 65 °C for 15 min prior to analysis by Northern blot or qPCR. For qPCR analysis, 2 μg of DNase I-treated total RNA was used for reverse transcription to synthesize first strand cDNA using SuperScriptaseII RT (Invitrogen). Aliquots of the first strand cDNA were used in a standard PCR mixture (Abgene, Epsom, Surrey, UK) or in real time PCR (qPCR) using SYBR Green as directed by the manufacturer (Bio-Rad). For all primer sets, reactions were first conducted using a temperature gradient to optimize annealing conditions, and the absence of primer dimers was confirmed by melting condition analysis. RT reactions carried out in the absence of RT served as a negative control, and to detect PCR products generated from contaminating genomic DNA. PCR using plasmid templates served as positive controls. The sequences of the primers for detection of RNase-L were as follows: human coding region: forward, 5′-GCTCATTTGTACTGCGTTATGC-3′, reverse, 5′-CATTTTCTCAAGGAAAAGGC-3′;3′-UTR (used for the analysis of β-globin-3′-UTR): forward, 5′-GCACTGAAGAGAGCATTTGGCAGA-3′, reverse, 5′-GAGCTCCTAGACTGGGTATGGGAA-3′; rpl13a: forward 5′-CTCAAGGTCGTGCGTCTG-3′, reverse, 5′-TGGCTTTCTCTTTCCTCTTCTC-3′; glyceraldehyde-3-phosphate dehydrogenase: forward, 5′-GAGTCAACGGATTTGGTCGT-3′, reverse, 5′-TTGATTTTGGAGGGATCTCG-3′; β-globin: forward, 5′-TGCATCTGTCCAGTGAGGAG-3′, reverse, 5′-AGCATTTGCAGAGGACAGGT-3′. To analyze the mRNA half-life of plasmid-encoded transcripts, cells were first transfected with the indicated plasmids, then, at 24 h post-transfection, the transfected cells were trypsinized and seeded into individual plates for treatment with 5 μg/ml actinomycin D (Sigma). In this manner, samples used to determine mRNA half-life were derived from a single transfection, thereby eliminating potential differences in expression associated with variable transfection efficiencies. Total RNA was harvested at the time points indicated in the figures. Specific mRNAs were quantified by qPCR. Each analysis represents at least two independent experiments, and within an experiment, all reactions were performed in triplicate, and the Ct values were converted to RNA concentration using a standard curve. mRNA values from each time point were normalized to the constitutively expressed ribosomal protein transcript, rpl13A. mRNA values were graphed on semi-log axes, and first order turnover rates were calculated by nonlinear regression of the percentage of mRNA remaining as a function of time after actinomycin D treatment. Western Blotting—Cell lysates were prepared using RIPA buffer (Upstate Biotechnology, Charlottesville, VA). The protein concentration in the lysates was determined by the Bradford microassay (Bio-Rad), and an equal amount of protein (30 μg/lane unless otherwise indicated in figure legends) was separated on 10% SDS-PAGE gels. Proteins were electrotransferred to Immobilon-P membrane (Millipore). The membranes were blocked in 5% nonfat milk TBST buffer (10 mm Tris, pH 8.0, 150 mm NaCl, 0.1% (v/v) Tween 20) for 1 h at room temperature, and then sequentially reacted with the primary antibody for 1 h in blocking buffer and horseradish peroxidase-conjugated secondary antibody (1:10,000 dilution; Sigma). The RNase-L monoclonal antibody (provided as a generous gift from Robert Silverman, The Cleveland Clinic Foundation) was used at a 1:2000 dilution. The α-actin (Sigma), neomycin-resistance (Upstate Biotechnology), and HuR (Santa Cruz) antibodies were used at 1:1000, 1:1000, and 1:500 dilutions, respectively. The immunoreactive complex was visualized using the Pierce SuperSignal chemiluminescent substrate (Pierce Biotechnology, Rockford, IL) and exposure to X-Omat AR film (Eastman Kodak Co.). RNP Immunoprecipitation—RNA-protein complexes were isolated from C2C12 myoblasts at day 3.5 of differentiation by lysing cells in PBL buffer (100 mm KCl, 5 mm MgCl2, 10 mm Hepes, pH 7.0, 0.5% Nonidet P-40) to preserve native RNA-protein interactions. Lysate supernatants were precleared for 1 h at 4 °C using 15 μg of IgG (Sigma) and 50 μl of protein A/G-Sepharose beads (Sigma) that had been swollen in NT2 buffer (50 mm Tris, pH 7.4, 150 mm NaCl, 1 mm MgCl2, and 0.05% Nonidet P-40) supplemented with 5% bovine serum albumin. Proteins that nonspecifically bound the protein-A/G-IgG complex were removed by centrifugation to generate the precleared lysate. For RNA immunoprecipitation, protein-A beads (100 μl) were first incubated (16 h, 4 °C) with 30 μg of IgG, goat anti-TIAR (Santa Cruz Biotechnology, Santa Cruz, CA), or anti-HuR (Upstate Biotechnology); beads were then washed 5 times in 1 ml of NT2 buffer before the lysate was added. Precleared cell lysate (3 mg) was then added to the protein-A/G-antibody complex and rotated for 2 h at 4°C. After washing 5 times in 1 ml of NT2 buffer, the complexes were treated sequentially with DNase I (5 min at 37 °C) then Proteinase K at 25 min at 55 °C. The RNA was extracted by phenol/chloroform and precipitated in 0.3 m sodium acetate (pH 5.5), EtOH at –20 °C for 24 h and resuspended in dH2O for analysis. The presence of specific mRNAs in the immunoprecipitated complex was determined by RT-PCR using gene-specific primers from the coding region of murine MyoD and RNase-L; the primers used for this analysis were: MyoD: forward, 5′-TACCCAAGGTGGAGATCCTG-3′, reverse 5′-CATCATGCCATCAGAGCAGT-3′; murine RNase-L: forward, 5′-CAATCGAGAAGTGGCTGTGA-3′, reverse, 5′-ATAGGATGCTGTGGGCAAAC-3′. Antiviral Assay—HeLa cells were co-transfected with RNase-L expression constructs, ZC5+ and pZeoHuR or empty vector control. At 24 h post-transfection, cells were reseeded into 96-well plates, and treated with IFNα (PBL Biomedical Laboratories, New Brunswick, NJ) at the indicated concentrations for 16 h, then challenged with encephalomyocarditis virus (multiplicity of infection ∼0.1) for 10 h. Cell viability was analyzed using the MTT assay (Promega, Madison, WI), and the percent protection was calculated using the formula: sample value – virus control (no IFN)/cell control (no virus) – virus control (no IFN). Each value is the mean of 8 identical wells ± S.D. RNase-L mRNA Contains a Long 3′-UTR with Putative AREs—The mature human RNase-L mRNA is 4.3 kb in size, with a coding region of just 2.2 kb and an extensive UTR of ∼2.1 kb (25Zhou A. Hassel B.A. Silverman R.H. Cell. 1993; 72: 753-765Abstract Full Text PDF PubMed Scopus (463) Google Scholar). Primer extension analysis identified a short, ∼0.37-kb 5′-UTR (27Zhou A. Nie H. Silverman R.H. Mamm. Genome. 2000; 11: 989-992Crossref PubMed Scopus (14) Google Scholar) indicating that the remaining 1.75 kb was 3′-UTR. Predicted secondary structures and poly(A) stretches in the 3′-UTR that resulted in RT pausing and oligo(dT) binding upstream of the poly(A) tail, precluded cloning of the 3′-UTR by conventional RT-based approaches. However, analysis of the human genome sequence revealed a predicted poly(A) site, and permitted PCR amplification of the 3′-UTR from genomic DNA to generate a full-length RNase-L cDNA designated ZC5+ (Fig. 1A). RT-PCR using overlapping forward and reverse primer pairs beginning in the coding region and continuing through the 3′-UTR to downstream of the predicted poly(A) site confirmed that the 3′-UTR is contiguous with the coding region, and that the predicted poly(A) site is used (Fig. 1B; NCBI accession code 810337). Specifically, identical PCR products of the expected sizes were generated from ZC5+ and from endogenous RNase-L mRNA when primers upstream of the predicted poly(A) site were used. However, PCR using a reverse primer downstream of the predicted poly(A) site produced a product from the ZC5+ construct that contained this genomic sequence, but not from the untransfected sample that contained only endogenous RNase-L mRNA (Fig. 1B, lane D). This finding is consistent with the use of the predicted poly(A) site by endogenous RNase-L, and indicated that the ZC5+ transcript was using the vector-derived poly(A) signal. Sequence analysis of the 3′-UTR revealed eight putative AREs with canonical AUUUA sequences that may serve as determinants of RNase-L mRNA stability. The extensive nature of the RNase-L 3′-UTR, and the presence of candidate AREs suggested that it may serve an important regulatory function. Cis-acting regulatory elements are predicted to be evolutionarily conserved, therefore we compared the RNase-L 3′-UTR sequence in human and mouse. The mouse RNase-L mRNA contained a 3′-UTR of 1774 bases from the stop codon to the predicted polyadenylation site, which is slightly longer than the 1736-base human sequence. Clustal W alignment of the mouse and human 3′-UTR revealed a modest sequence similarity across the entire 3′-UTR (57% identity); however, islands of high conservation that approach the 74% identity observed in the coding region, were also detected (28Chenna R. Sugawara H. Koike T. Lopez R. Gibson T.J. Higgins D.G. Thompson J.D. Nucleic Acids Res. 2003; 31: 3497-3500Crossref PubMed Scopus (4081) Google Scholar). In particular, a 3′-terminal region of 305 bases exhibited a striking 71% sequence identity between mouse and human, and contained two putative ARE elements in a uridine-rich (35% U) context that characterizes functional AREs. These findings identify a conserved 3′-UTR in the RNase-L mRNA that contains ARE elements and may function in the post-transcriptional regulation of RNase-L expression. The 3′-UTR Is a Negative Regulator of RNase-L Expression and mRNA Stability—The presence of a long 3′-UTR is frequently indicative of post-transcriptional regulation of gene expression via modulation of mRNA stability; therefore, we investigated if the RNase-L 3′-UTR modulated its expression. Transfection of 2fTGH human fibrosarcoma cells with an expression construct containing the RNase-L coding region and 200 bases of 3′-UTR (designated ZC5) resulted in a robust expression of RNase-L protein (Fig. 2A); a comparable strong expression was observed with a coding region construct, demonstrating the lack of regulatory elements in the 5′-most 200 bases of the 3′-UTR (not shown). In contrast, RNase-L expression was dramatically reduced following transfection with the full-length RNase-L ZC5+ construct that contained the 3′-UTR (Fig. 2A). This construct contained poly(A) signals from the RNase-L sequence and from the vector; however, the presence of an extra poly(A) signal did not alter RNase-L expr