The AU-rich 3' untranslated region of human MDR1 mRNA is an inefficient mRNA destabilizer.

The human multidrug resistance gene MDR1 encodes a membrane-bound protein, referred to as P-glycoprotein, that acts as a pump to extrude toxins from cells. The 3' untranslated region (3'UTR) of the human MDR1 mRNA is very AU-rich (70%) and contains AU-rich sequences similar to those shown to confer rapid decay on c-myc, c-fos, and lymphokine mRNAs. We tested the ability of the MDR1 3'UTR to act as an mRNA destabilizing element in the human hepatoma cell line HepG2. The MDR1 mRNA has an intermediate half-life of 8 h in HepG2 cells compared to a half-life of 30 min for c-myc mRNA. The MDR1 mRNA half-life was prolonged to >20 h upon treatment with the protein synthesis inhibitor cycloheximide. We constructed expression vectors containing the human beta-globin coding region with the 3'UTR from either MDR1 or c-myc. The c-myc 3'UTR increased the decay of the chimeric mRNA, but the MDR1 3'UTR had no effect. We tested the ability of MDR1 3'UTR sequences to compete for interaction with AU-binding proteins in cell extracts; MDR1 RNA probes had a fivefold lower affinity for AU-binding proteins that interact with the c-myc AU-rich 3'UTR. Overall, our data suggest that the MDR1 3'UTR does not behave as an active destabilizing element in HepG2 cells.     

Destabilization of interleukin-6 mRNA requires a putative RNA stem-loop structure, an AU-rich element, and the RNA-binding protein AUF1

Interleukin-6 mRNA is unstable and degraded with a half-life of 30 min. Instability determinants can entirely be attributed to the 3' untranslated region. By grafting segments of this region to stable green fluorescent protein mRNA and subsequent scanning mutagenesis, we have identified two conserved elements, which together account for most of the instability. The first corresponds to a short noncanonical AU-rich element. The other, 80 nucleotides further 5', comprises a sequence predicted to form a stem-loop structure. Neither element alone was sufficient to confer full instability, suggesting that they might cooperate. Overexpression of myc-tagged AUF1 p37 and p42 isoforms as well as suppression of endogenous AUF1 by RNA interference stabilized interleukin-6 mRNA. Both effects required the AU-rich instability element. Similarly, the proteasome inhibitor MG132 stabilized interleukin-6 mRNA probably through an increase of AUF1 levels. The mRNA coimmunoprecipitated specifically with myc-tagged AUF1 p37 and p42 in cell extracts but only when the AU-rich instability element was present. These results indicate that AUF1 binds to the AU-rich element in vivo and promotes IL-6 mRNA degradation.     

Okadaic acid-mediated induction of the c-fos gene in estrogen receptor-negative human breast carcinoma cells utilized, in part, posttranscriptional mechanisms involving adenosine-uridine-rich elements.

Signal transduction via modulation of phosphorylation after selective inhibition of protein phosphatase (PP) 1 and/or PP2A appears to play a role in okadaic acid (OA)-mediated effects. Treatment of several estrogen receptor-negative human breast carcinoma (HBC) cells with 100 nM OA resulted in induction of c-fos, c-myc, and cyclin-dependent kinase inhibitor p21WAF1/CIP1 genes. Transfections of various luciferase reporter constructs in HBC cells revealed involvement of activator protein-1-dependent as well as -independent pathways in induction of the c-fos gene by OA. MDA-MB-468 HBC cells were stably transfected with plasmids expressing luciferase, chimeric luciferase- c-fos 3' untranslated region (3'UTR), or chimeric luciferase-p21WAF1/CIP 3'UTR mRNAs. Expression of chimeric luciferase-c-fos and luciferase-p21WAF1/CIP1 mRNAs was elevated by OA in several independent sublines. Actinomycin D chase experiments revealed an enhanced rate of decay of luciferase-c-fos mRNA, whereas treatment with OA caused approximately 3.5-fold enhanced stability of the chimeric luciferase-c-fos mRNA only. By transfecting different plasmids containing deletions of c-fos 3'UTR, OA-responsive sequences were mapped to an 86-nucleotide, AU-rich region. UV cross-linking experiments using HBC cell cytosolic proteins showed multiple complexes with the AU-rich region subfragments of c-fos, as well as c-myc and p21WAF1/CIP1 mRNAs. OA enhanced binding of a novel Mr approximately 75,000 protein present in the cytosolic extracts of HBC cells to the AU-rich RNA probes of all of the above three genes. Taken together, OA regulation of HBC cell gene expression involves the activator protein-1 pathway, as well as enhanced binding of a novel Mr approximately 75,000 protein to an AU-rich region of the 3'UTRs of the target genes.     

Regulation of urokinase receptor expression by p53: novel role in stabilization of uPAR mRNA

We found that p53-deficient (p53(-/-)) lung carcinoma (H1299) cells express robust levels of cell surface uPAR and uPAR mRNA. Expression of p53 protein in p53(-/-) cells suppressed basal and urokinase (uPA)-induced cell surface uPAR protein and increased uPAR mRNA degradation. Inhibition of p53 by RNA silencing in Beas2B human airway epithelial cells conversely increased basal as well as uPA-mediated uPAR expression and stabilized uPAR mRNA. Purified p53 protein specifically binds to the uPAR mRNA 3' untranslated region (3'UTR), and endogenous uPAR mRNA associates with p53. The p53 binding region involves a 37-nucleotide uPAR 3'UTR sequence, and insertion of the p53 binding sequence into beta-globin mRNA destabilized beta-globin mRNA. Inhibition of p53 expression in these cells reverses decay of chimeric beta-globin-uPAR mRNA. These observations demonstrate a novel regulatory role for p53 as a uPAR mRNA binding protein that down-regulates uPAR expression, destabilizes uPAR mRNA, and thereby contributes to the viability of human airway epithelial or lung carcinoma cells.     

Cutting edge: clustered AU-rich elements are the target of IL-10-mediated mRNA destabilization in mouse macrophages

In the present study we show that IL-10-mediated inhibition of inflammatory gene expression can be mediated by an AU-rich element (ARE) cluster present in the 3' untranslated region (3'UTR) of sensitive genes. A series of chloramphenicol acetyl transferase (CAT) reporter gene constructs were prepared in which different fragments from the IL-10-sensitive KC mRNA 3'UTR were placed downstream of the coding region of the reporter gene CAT. CAT mRNA containing the KC 3'UTR was markedly destabilized as compared with the control CAT mRNA, and the decay rate was further increased in cells stimulated with IL-10. The KC 3'UTR contains an ARE cluster and three isolated ARE motifs. The ARE cluster spanning nucleotides 378-399 appeared to be both necessary and sufficient to mediate sensitivity to IL-10 because a 116-nucleotide fragment that contains the cluster conferred sensitivity, while mutation of the sequence between positions 378 and 399 eliminated sensitivity. The destabilizing effect of IL-10 was relatively selective, as the stability of chimeric CAT mRNAs was not modulated in cells treated with IFN-gamma or IL-4.     

The nonamer UUAUUUAUU is the key AU-rich sequence motif that mediates mRNA degradation.

Labile mRNAs that encode cytokine and immediate-early gene products often contain AU-rich sequences within their 3' untranslated region (UTR). These AU-rich sequences appear to be key determinants of the short half-lives of these mRNAs, although the sequence features of these elements and the mechanism by which they target mRNAs for rapid decay have not been fully defined. We have examined the features of AU-rich elements (AREs) that are crucial for their function as determinants of mRNA instability in mammalian cells by testing the ability of various mutant c-fos AREs and synthetic AREs to direct rapid mRNA deadenylation and decay when inserted within the 3' UTR of the normally stable beta-globin mRNA. Evidence is presented that the pentamer AUUUA, which previously was suggested to be the minimal determinant of instability present in mammalian AREs, cannot direct rapid mRNA deadenylation and decay. Instead, the nonomer UUAUUUAUU is the elemental AU-rich sequence motif that destabilizes mRNA. Removal of one uridine residue from either end of the nonamer (UUAUUUAU or UAUUUAUU) results in a decrease of potency of the element, while removal of a uridine residue from both ends of the nonamer (UAUUUAU) eliminates detectable destabilizing activity. The inclusion of an additional uridine residue at both ends of the nonamer (UUUAUUUAUUU) does not further increase the efficacy of the element. Taken together, these findings suggest that the nonamer UUAUUUAUU is the minimal AU-rich motif that effectively destabilizes mRNA. Additional ARE potency is achieved by combining multiple copies of this nonamer in a single mRNA 3' UTR. Furthermore, analysis of poly(A) shortening rates for ARE-containing mRNAs reveals that the UUAUUUAUU sequence also accelerates mRNA deadenylation and suggests that the UUAUUUAUU motif targets mRNA for rapid deadenylation as an early step in the mRNA decay process.     

Translational regulation of human beta interferon mRNA: association of the 3'' AU-rich sequence with the poly(A) tail reduces translation efficiency in vitro.

The 3' AU-rich region of human beta-1 interferon (hu-IFN beta) mRNA was found to act as a translational inhibitory element. The translational regulation of this 3' AU-rich sequence and the effect of its association with the poly(A) tail were studied in cell-free rabbit reticulocyte lysate. A poly(A)-rich hu-IFN beta mRNA (110 A residues) served as an inefficient template for protein synthesis. However, translational efficiency was considerably improved when the poly(A) tract was shortened (11 A residues) or when the 3' AU-rich sequence was deleted, indicating that interaction between these two regions was responsible for the reduced translation of the poly(A)-rich hu-IFN beta mRNA. Differences in translational efficiency of the various hu-IFN beta mRNAs correlated well with their polysomal distribution. The poly(A)-rich hu-IFN beta mRNA failed to form large polysomes, while its counterpart bearing a short poly(A) tail was recruited more efficiently into large polysomes. The AU-rich sequence-binding activity was reduced when the RNA probe contained both the 3' AU-rich sequence and long poly(A) tail, supporting a physical association between these two regions. Further evidence for this interaction was achieved by RNase H protection assay. We suggest that the 3' AU-rich sequence may regulate the translation of hu-IFN beta mRNA by interacting with the poly(A) tail.     

Deadenylylation: a mechanism controlling c-fos mRNA decay

The c-fos proto-oncogene mRNA is extremely labile and is rapidly degraded within minutes after being transported to the cytoplasm of growth factor-stimulated fibroblasts. Analysis of the structural determinants controlling c-fos message decay reveals that this message contains at least two functionally independent elements that are responsible for its short half-life. One of these determinants is an AU-rich sequence present in the 3' untranslated region of the c-fos message, whereas the other determinant, which is structurally unrelated to the AU-rich element, is located within the c-fos protein-coding sequence. Both the c-fos AU-rich element and the coding region instability determinant appear to function by facilitating rapid removal of the c-fos poly(A) tail as a first step in the mRNA degradation process.     

Protein synthesis inhibition stabilizes urokinase-type plasminogen activator mRNA. Studies in vivo and in cell-free decay reactions

Inhibition of protein synthesis stabilizes a number of mRNAs, but little is known about the mechanism. To understand the relationship between protein synthesis and mRNA stability, we studied the degradation of calcitonin-induced urokinase-type plasminogen activator (uPA) mRNA in LLC-PK cells. uPA mRNA became highly stable by pretreatment with either cycloheximide or pactamycin, and the stabilizing effect of cycloheximide treatment was time dependent with the full effect exerted by 60 min. Stabilization was also observed with histone H4 mRNA but only partially with c-myc mRNA. To further analyze, we developed a cell-free decay reaction system based on post-mitochondrial supernatant (PMS). In this system, uPA mRNA was completely stable when fractions were obtained from cells pretreated with cycloheximide, but very unstable in control fractions, paralleling uPA mRNA stability in intact cells. However, in contrast to uPA mRNA and the in vivo observation, histone H4 mRNA was unstable whether or not the cells were pretreated with cycloheximide. These results suggest that inhibition of protein synthesis stabilizes mRNAs in at least two different ways in LLC-PK1 cells. When PMS from cycloheximide/calcitonin-treated cells was mixed with PMS from untreated cells, uPA mRNA was not destabilized. This suggests that a putative labile factor responsible for uPA mRNA degradation is not a soluble protein.     

Regulation of urokinase expression at the posttranscription level by lung epithelial cells

Urokinase-type plasminogen activator (uPA) is expressed by lung epithelial cells and regulates fibrin turnover and epithelial cell viability. PMA, LPS, and TNF-alpha, as well as uPA itself, induce uPA expression in lung epithelial cells. PMA, LPS, and TNF-alpha induce uPA expression through increased synthesis as well as stabilization of uPA mRNA, while uPA increases its own expression solely through uPA mRNA stabilization. The mechanism by which lung epithelial cells regulate uPA expression at the level of mRNA stability is unclear. To elucidate this process, we sought to characterize protein-uPA mRNA interactions that regulate uPA expression. Regulation of uPA at the level of mRNA stability involves the interaction of a ~40 kDa cytoplasmic-nuclear shuttling protein with a 66 nt uPA mRNA 3'UTR sequence. We purified the uPA mRNA 3'UTR binding protein and identified it as ribonucleotide reductase M2 (RRM2). We expressed recombinant RRM2 and confirmed its interaction with a specific 66 nt uPA 3'UTR sequence. Immunoprecipitation of cell lysates with anti-RRM2 antibody and RT-PCR for uPA mRNA confirmed that RRM2 binds to uPA mRNA. Treatment of Beas2B cells with uPA or LPS attenuated RRM2-endogenous uPA mRNA interactions, while overexpression of RRM2 inhibited uPA protein and mRNA expression through destabilization of uPA mRNA. LPS exposure of lung epithelial cells translocates RRM2 from the cytoplasm to the nucleus in a time-dependent manner, leading to stabilization of uPA mRNA. This newly recognized pathway could influence uPA expression and a broad range of uPA-dependent functions in lung epithelial cells in the context of lung inflammation and repair.     

The half-life of c-myc mRNA in growing and serum-stimulated cells: influence of the coding and 3'' untranslated regions and role of ribosome translocation.

c-myc mRNA contains at least two discrete sequence elements that account for its short half-life, one in the 3' untranslated region and the other in the carboxy-terminal coding region (coding-region determinant). To investigate the function of each determinant, one or both were fused in frame to portions of a gene encoding long-lived beta-globin mRNA. Each chimeric gene was stably transfected into HeLa and NIH 3T3 cells and was transcribed from a constitutive cytomegalovirus promoter or from a serum-regulated c-fos promoter, respectively. The steady-state levels of the chimeric mRNAs in exponentially growing HeLa cells were compared, and their half-lives were measured by two independent methods: (i) in actinomycin D-treated HeLa cells and (ii) after serum addition to starved 3T3 cells. By each method, mRNAs containing either instability determinant were less stable than beta-globin mRNA. mRNA containing only the c-myc 3' untranslated region was not significantly more stable than mRNA with both determinants. In a cell-free mRNA decay system containing polysomes from transfected HeLa cells, mRNA containing the coding-region determinant was destabilized by addition of a specific RNA competitor, whereas mRNA containing only the 3' untranslated region was unaffected. When a stop codon was inserted upstream of the coding-region determinant, the chimeric mRNA was stabilized approximately twofold. These and other data suggest that degradation involving the coding-region determinant occurs most efficiently when ribosomes are translating the determinant.