In Vivo Studies of Translational Repression Mediated by the Granulocyte-Macrophage Colony-stimulating Factor AU-rich Element

Abstract

The AU-rich element (ARE) controls the turnover of many unstable mRNAs and their translation. The granulocyte-macrophage colony-stimulating factor (GM-CSF) ARE is known to be a destabilizing element, but its role in translation remains unclear. Here we studied in vivo the role of the GM-CSF ARE on the mRNA and protein expressions of an enhanced green fluorescent protein reporter gene. The GM-CSF ARE had a repressor effect on translation independently of its effect on mRNA levels. In the context of an internal ribosome entry site, the GM-CSF ARE still repressed translation but was no longer functional as a destabilizing element. Gel retardation assays showed that poly(A)-binding protein is displaced from the poly(A) tail when the ARE is present in the 3′-untranslated region. These data suggest that the GM-CSF ARE controls translation and mRNA decay by interfering with poly(A)-binding protein-mediated mRNA circularization. The AU-rich element (ARE) controls the turnover of many unstable mRNAs and their translation. The granulocyte-macrophage colony-stimulating factor (GM-CSF) ARE is known to be a destabilizing element, but its role in translation remains unclear. Here we studied in vivo the role of the GM-CSF ARE on the mRNA and protein expressions of an enhanced green fluorescent protein reporter gene. The GM-CSF ARE had a repressor effect on translation independently of its effect on mRNA levels. In the context of an internal ribosome entry site, the GM-CSF ARE still repressed translation but was no longer functional as a destabilizing element. Gel retardation assays showed that poly(A)-binding protein is displaced from the poly(A) tail when the ARE is present in the 3′-untranslated region. These data suggest that the GM-CSF ARE controls translation and mRNA decay by interfering with poly(A)-binding protein-mediated mRNA circularization. The mechanisms of post-transcriptional regulation at the level of mRNA degradation and translation are of importance in the control of gene expression, with the 3′-untranslated region (UTR) 1The abbreviations used are: UTR, untranslated region; ARE, adenosine uridine-rich element; ARE-BP, ARE-binding protein; EGFP, enhanced green fluorescent protein; RNP, ribonucleoprotein; PABP, poly(A)-binding protein; GM-CSF, granulocyte-macrophage colony-stimulating factor; ECMV, encephalomyocarditis virus; IRES, internal ribosome entry site; TIAR, T-cell restricted intracellular antigen-1-related protein. 1The abbreviations used are: UTR, untranslated region; ARE, adenosine uridine-rich element; ARE-BP, ARE-binding protein; EGFP, enhanced green fluorescent protein; RNP, ribonucleoprotein; PABP, poly(A)-binding protein; GM-CSF, granulocyte-macrophage colony-stimulating factor; ECMV, encephalomyocarditis virus; IRES, internal ribosome entry site; TIAR, T-cell restricted intracellular antigen-1-related protein. central in this process (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar). The adenosine uridine-rich element (ARE) is the most common regulatory determinant found in the 3′-UTR of many unstable mRNAs, and several types have been described (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar). The ARE induces rapid mRNA turnover by first triggering poly(A) shortening, then inducing both 5′-end decapping and RNA body degradation by either 5′ and 3′ exonuclease complexes (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, 2Mitchell P. Tollervey D. Curr. Opin. Genet. Dev. 2000; 10: 193-198Google Scholar, 3Chen C.Y. Gherzi R. Ong S.E. Chan E.L. Raijmakers R. Pruijn G.J Stoecklin G. Moroni C. Mann M. Karin M. Cell. 2001; 107: 451-464Google Scholar, 4Vasudevan S. Peltz S.W. Mol. Cell. 2001; 7: 1191-1200Google Scholar). Recent data also suggest involvement of the ubiquitin-dependent pathway in the control of ARE-mediated mRNA turnover (5Laroia G. Sarkar B. Schneider R.J. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 1842-1846Google Scholar). Whereas the ARE-mediated cap-dependent deadenylation model is largely accepted in yeast, it remains to be demonstrated in mammalian cells. For this reason, human homologs of the yeast decapping enzyme were cloned (6Van Dijk E. Cougot N. Meyer S. Babajko S. Wahle E. Seraphin B. EMBO J. 2002; 21: 6915-6924Google Scholar), and a mRNA-decapping activity was shown to be regulated by ARE in HeLa cells (7Gao M. Wilusz C.J. Peltz S.W. Wilusz J. EMBO J. 2001; 20: 1134-1143Google Scholar). The ARE is also a translational silencer (8Kruys V. Huez G. Biochimie (Paris). 1994; 76: 862-866Google Scholar, 9Jarzembowski J.A. Rajagopalan L.E. Shin H.C. Malter J.S. Nucleic Acids Res. 1999; 27: 3660-3666Google Scholar, 10Mukhopadhyay D. Houchen C.W. Kennedy S. Dieckgraefe B.K. Anant S. Mol. Cell. 2003; 11: 113-126Google Scholar). The main question is whether ARE works both as an mRNA destabilizer and a translational repressor. Both termini of an mRNA can associate via an interaction between the cap binding complex (eukaryotic initiation factor 4F) and the poly(A)-binding protein (PABP) (11Preiss T. Hentze M.W. Nature. 1998; 392: 516-520Google Scholar, 12Wells S.E. Hillner P.E. Vale R.D. Sachs A.B. Mol. Cell. 1998; 2: 135-140Google Scholar, 13Gallie D.R. Gene (Amst.). 1998; 216: 1-11Google Scholar). By facilitating the binding of these proteins on their substrates, this interaction protects the mRNA extremities from the degradation and potentiates the translation of capped mRNAs. Because a poly(A) tail is needed for both mRNA stabilization and efficient translation (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, 14Dreyfus M. Regnier P. Cell. 2002; 111: 611-613Google Scholar) and because factors associated with the cis-stabilizing element of the highly stable α-Globin mRNA cooperate with PABP (15Wang Z. Day N. Trifillis P. Kiledjian M. Mol. Cell. Biol. 1999; 19: 4552-4560Google Scholar), AREs may control mRNA decay and translation by targeting the PABP/poly(A) tail association. In this work we used the granulocyte-macrophage colony-stimulating factor (GM-CSF) ARE as a model. The GM-CSF is a hematopoietic growth factor that influences the commitment, differentiation, and functional activity of myelomonocytic cells (16Metcalf D. CIBA Found. Symp. 1997; 204: 40-50Google Scholar). This ARE belongs to class IIA and contains five reiterations of the AUUU motifs embedded in an AU-rich sequence. It was the first to be associated with the control of mRNA stability (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, 17Shaw G. Kamen R. Cell. 1986; 46: 659-667Google Scholar). The physiological importance of the ARE in the control of GM-CSF gene expression was demonstrated in transgenic mice bearing GM-CSF gene constructs lacking the ARE that presented severe hematopoietic disorders (18Houzet L. Morello D. Defrance P. Mercier P. Huez G. Kruys V. Blood. 2001; 98: 1281-1288Google Scholar). The control of translation by the GM-CSF ARE is much less documented. ARE-containing mRNAs are poorly translated, and the GM-CSF ARE inhibits translation in vitro (8Kruys V. Huez G. Biochimie (Paris). 1994; 76: 862-866Google Scholar, 9Jarzembowski J.A. Rajagopalan L.E. Shin H.C. Malter J.S. Nucleic Acids Res. 1999; 27: 3660-3666Google Scholar). To date many ARE-binding proteins (ARE-BPs) have been described (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, 19Guhaniyogi J. Brewer G. Gene (Amst.). 2001; 265: 11-23Google Scholar). These include the embryonic lethal abnormal visual system-like protein family including HuR, CUGBP2, a novel ARE-BP involved in cyclooxygenase-2 post-transcriptional regulation (10Mukhopadhyay D. Houchen C.W. Kennedy S. Dieckgraefe B.K. Anant S. Mol. Cell. 2003; 11: 113-126Google Scholar), the heterogeneous nuclear ribonucleoprotein D isoforms, T-cell restricted intracellular antigen-1/TIAR proteins, and tristetraprolin (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, 2Mitchell P. Tollervey D. Curr. Opin. Genet. Dev. 2000; 10: 193-198Google Scholar, 19Guhaniyogi J. Brewer G. Gene (Amst.). 2001; 265: 11-23Google Scholar, 20Mitchell P. Tollervey D. Curr. Opin. Cell Biol. 2001; 13: 320-325Google Scholar). Because they associate with the ARE, these proteins are thought to participate in ARE effects on either mRNA decay or translation. However, further investigations need to be made to understand their exact role in these mechanisms. We sought whether the GM-CSF ARE could control in vivo the translation of a reporter gene. Using an enhanced green fluorescent protein (EGFP) expression system, we demonstrate that the GM-CSF ARE represses translation independently of its mRNA-destabilizing effect. In a context of cap-independent translation, the GM-CSF ARE is no longer efficient as a destabilizing element, whereas it still functions as a translational repressor. Finally, we show that the GM-CSF ARE intervenes on PABP binding on the poly(A) tail. Plasmid Constructions—Plasmids pT7/T3α-18, pDsRed2-N1, pEGFP-C3, and pIRES2-EGFP were purchased from either Invitrogen or Ozyme (Saint Quentin, France). To generate plasmids pEGFP-B and pEGFP-A, the BglII- and BamHI-digested PCR fragments containing either the whole 3′-UTR from of J. of and or the GM-CSF ARE from and were in the from To generate the and the region was by and and on To generate the the whole including the for was from by the and the and its To generate were in the GM-CSF ARE of by as the and its To generate the region was from pIRES2-EGFP by with and and and on To generate plasmids and the from either or was by in the pIRES2-EGFP the constructs were by Cell Cell cells were in were by France). cells were with used as control for were from HeLa cells as described C. C.Y. N. N. A.B. Cell. 2000; Scholar). of RNA was the RNA from France). were as described C. C.Y. N. N. A.B. Cell. 2000; Scholar). The mRNA were by on an and was on as described C. C.Y. N. N. A.B. Cell. 2000; Scholar). were purchased as was from was from was from was by and and France). PABP was by of of of RNA vitro were to the of RNA and either the or or by A. as J. N. J.A. Dev. 2001; Scholar). RNA were by including in the of and Gel were as described C. C.Y. N. N. A.B. Cell. 2000; Scholar). The of was and were at of and RNA were for then RNA was for by the was in and of was or in the The was further for Finally, ribonucleoprotein complexes were in or of by HeLa cells were in by on and by an and the were and HeLa cells with reporter plasmids were used as the control to the or To between and a of was used for the in with the and we the with the that the and data could be the expression of the was by the of the and the of the cells from data and as an This expression was as of the of of of the of the The in was on the of the of cells and GM-CSF ARE represses translation of an mRNA but can no longer its HeLa cells were with of of the plasmids or and of as the In the of was at by of the was by as described in The expression and the of cells is shown as described in of in was HeLa cells were with of used as control for and of the plasmids or the control was by and as described in In the the of the cells. The the expression and the of cells as described in of The controls used for and and mRNA were the as in The GM-CSF ARE in the ARE effect in vivo we the as reporter gene and as a and to the expression of the To the was in HeLa cells were with plasmids containing either the whole 3′-UTR or the GM-CSF ARE of the region in the was used as shown in and when HeLa cells were by pEGFP-A, GM-CSF ARE the expression in whereas the of cells was by The level of mRNA was also when with mRNA of but the role of the GM-CSF ARE as a destabilizing element. expression in cells with with the of the mRNA to a effect of the GM-CSF ARE on in vivo vivo effects of the GM-CSF ARE on mRNA and protein HeLa cells were with of pEGFP-B or the expression was on the by and The mRNAs were on RNA by In the with the of cells also The of the expression and the the of the an of is In controls were by In controls were by mRNA of mRNAs were by as of of Because the of the region and the GM-CSF ARE in the was by the in expression in the cells could be for by the protein However, to demonstrate the ARE in the translational in cells by pEGFP-A, further investigations were controls and for expression were made from the of the The was by the and the whole region and was used as control for both and mRNA expression in cells. mRNA protein was by cells with this the was by the and the In this was used as control for expression and control for the cells EGFP, but a mRNA of the and the was by To the of the ARE-mediated translational we made the GM-CSF ARE that motifs and the AUUU of HeLa cells with the the expression the of cells and the level of its of with pEGFP-A, contains the functional ARE and These demonstrated that the translational was The ARE-mediated on translation from was less when with the on translation from It is that the in the GM-CSF ARE were to the GM-CSF ARE effects on To these cells were with the a fluorescent protein for shown in a of HeLa cells the protein in the the in expression could be for by a in the Finally, we the ARE-mediated translational by cells with of either or In cells by mRNA and protein expressions in to the of used for In this by either or when cells were with of pEGFP-A, whereas the mRNA In the in and the showed that the cap-dependent translation was by the GM-CSF ARE the of used for the these data demonstrate that translation was repressed by the GM-CSF ARE in vivo independently of its destabilizing effects on GM-CSF ARE the role of the ARE on cap-independent translation, we the effects of the GM-CSF ARE when translation was from the were by either the as the control by of the region in the or the and plasmids containing the GM-CSF ARE or its of the region in the pIRES2-EGFP we the described in and of we that was in the with the contains the we by the cells with of either or as for and that in cells by expression in to the of used for However, the level of was with of in the no was in By the in and it may be that the level of expression in cells by was much in cells with less when of This that the translation from the was much less efficient in the HeLa cells the cap-dependent translation and that was to the and of of fluorescent cells were with of contains the functional ARE, the expression level was that when the control with of and with of in cells with of and with of These data to a role of the GM-CSF ARE on translation. However, to further demonstrate this we of either or the and the expression level of both and mRNA as by and and and and showed that in cells with the the expression level and the of the cells were and from that translation, cap-dependent translation, was by the GM-CSF in the shown in was by the cells with the Here a of HeLa cells the protein in the that the in expression could be for by a in the of the cells. showed that GM-CSF ARE was no longer functional as a destabilizing element when translation was from the IRES, the level of the mRNAs was the whether the the GM-CSF ARE or its The GM-CSF ARE in PABP on the ARE-mediated mRNA decay by poly(A) (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, 2Mitchell P. Tollervey D. Curr. Opin. Genet. Dev. 2000; 10: 193-198Google Scholar, 19Guhaniyogi J. Brewer G. Gene (Amst.). 2001; 265: 11-23Google Scholar), the GM-CSF ARE can both and translation and the have been shown to participate in both (11Preiss T. Hentze M.W. Nature. 1998; 392: 516-520Google Scholar, 12Wells S.E. Hillner P.E. Vale R.D. Sachs A.B. Mol. Cell. 1998; 2: 135-140Google Scholar, 13Gallie D.R. Gene (Amst.). 1998; 216: 1-11Google Scholar, A. N. RNA 2001; 7: Scholar), we whether the ARE may in PABP binding on the poly(A) tail. We assays HeLa RNA and with a poly(A) tail and their and The the functional of the GM-CSF ARE, whereas the a GM-CSF ARE The RNA were in vitro the RNA and the and plasmids by A. as J. N. J.A. Dev. 2001; Scholar). In the RNA were in vitro as the and plasmids as we the RNA no complexes that no HeLa protein was to associate with the RNA we the ARE-containing RNA we the of complexes to the with the GM-CSF ARE To the ARE of the we and TIAR, (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, 2Mitchell P. Tollervey D. Curr. Opin. Genet. Dev. 2000; 10: 193-198Google Scholar, 19Guhaniyogi J. Brewer G. Gene (Amst.). 2001; 265: 11-23Google Scholar). complexes either with or in However, no was with the or when the RNA with of these These showed that proteins associated with the with the and that and TIAR, known but PABP, these we the RNA the RNA complexes and with the poly(A) tail no protein could associate with the of these complexes was with the the and and but complexes were when the PABP was to the binding in the RNA the complex was to that with the and that this the complex of the as shown by the at the of The and complexes were also and of the the Here containing or with the of their in and the of the complexes for the ARE when the RNA However, when the was to the the complexes were with the the RNA and in Because the complex was to the we the a The in are to in and the that the GM-CSF ARE in vitro on the between PABP and the poly(A) tail. In this we used as a to in vivo the effects of the ARE from the GM-CSF mRNA on the expression level of an reporter gene. The data and demonstrate the of the in in vivo the role of regulatory as AREs in the control of gene show that in mammalian the GM-CSF ARE cap-dependent translation in independently of its effect on mRNA that the ARE is a translational repressor (8Kruys V. Huez G. Biochimie (Paris). 1994; 76: 862-866Google Scholar). this ARE-mediated translational was also when translation was from an IRES, but in this the ARE was no longer functional on mRNA Finally, we found that the GM-CSF ARE intervenes in vitro in PABP binding with the poly(A) tail. ARE-mediated translational been in vitro in and (8Kruys V. Huez G. Biochimie (Paris). 1994; 76: 862-866Google Scholar, 9Jarzembowski J.A. Rajagopalan L.E. Shin H.C. Malter J.S. Nucleic Acids Res. 1999; 27: 3660-3666Google Scholar) but also in cells (10Mukhopadhyay D. Houchen C.W. Kennedy S. Dieckgraefe B.K. Anant S. Mol. Cell. 2003; 11: 113-126Google Scholar). these data and show that GM-CSF ARE represses translation in vivo in HeLa cells. in yeast, ARE-mediated mRNA decay in mammalian cells by poly(A) and is by factors with either the ARE or the poly(A) tail (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, 19Guhaniyogi J. Brewer G. Gene (Amst.). 2001; 265: 11-23Google Scholar). the PABP on a RNA its in vitro binding was when it the GM-CSF ARE in its 3′-UTR and This that the ARE PABP binding on the poly(A) tail. The by ARE on the PABP/poly(A) tail is and further but it is that factors associated with the GM-CSF ARE are the human as the model showed in vitro a interaction between a and the ARE L. M. A. M. S. J. Biol. 2002; Scholar). We complex when the and the ARE-containing However, of the complex was when the was to the binding This may suggest that the PABP is of the complex and may participate to its In this PABP been shown to with either the mRNA in J. Biol. 2002; Scholar) or the complex in HeLa cells G. R. Brewer G. Schneider R.J. 1999; Scholar) and to to a protein complex by the GM-CSF ARE as R. M. S. Res. 2003; Scholar). we can that the ARE, by binding either or the PABP, may to the PABP from the poly(A) tail. This could a model for the ARE both mRNA degradation and translational a model is of the model by Wilusz (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar) in the complexes between the PABP and the poly(A) tail. Both mRNA stability and translational on the of the PABP with the poly(A) tail D.R. Gene (Amst.). 1998; 216: 1-11Google Scholar, B. V. N. Mol. Cell. Biol. 2001; 21: Scholar). the of PABP from the poly(A) tail by the ARE may to the poly(A) to at the by the poly(A) (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar). In at the the of PABP from the poly(A) tail may the interaction between initiation factor and the the binding of the decapping complex poly(A) and the decapping P. Tollervey D. Curr. Opin. Cell Biol. 2001; 13: 320-325Google Scholar). Because can on both of the it may participate in the deadenylation and decapping In cells ARE-mediated translational by of the from the mRNA been for the but this effect remains to the and was for the or GM-CSF mRNAs M. S. N. M. B. S. C. Kruys V. M. P. EMBO J. 2000; 19: Scholar). This is by the ARE-BP T-cell restricted intracellular protein. However, showed that the ARE-containing mRNAs are associated with the and that the ARE destabilizing on translation M. S. N. M. B. S. C. Kruys V. M. P. EMBO J. 2000; 19: Scholar, G. G. Mol. Cell. Biol. 13: Scholar, N. Mol. Cell. Biol. Scholar, E. M. V. Mol. Cell. Biol. Scholar). In of these data and we that the ARE-mediated PABP from the poly(A) tail ribosome or on the ARE-containing mRNAs the of its the translational of these mRNAs is This model is in with the expression by in cells with the To the effect of the GM-CSF ARE on cap-independent translation, we plasmids containing the reporter gene the translational control of the and either the functional GM-CSF ARE or its We first that in HeLa the expression level from was with the expression with the pEGFP-B or This was to the expression with the of used in the showed that the GM-CSF ARE could also translation from an However, the ARE-mediated on translation was less on cap-dependent translation. Because the translation initiation is much less efficient from the from the we that this less ARE-mediated on translation is to the The ARE-mediated of the translation also of the initiation complex in this context both the binding of the initiation factors and on the and the interaction between PABP and A. N. RNA 2001; 7: Scholar). Because to be as for initiation of the translation as for cap-dependent translation, of PABP from the poly(A) tail may translation in both we that the main for the control of either or translation by as the ARE, is the tail may the of the ribosome on the the of the translational and the of the ribosome the translation. when we mRNA expression, the level was the whether the mRNA the GM-CSF ARE or the that the mRNA is stable it contains an This that the ARE no longer functions in mRNA turnover when translation is from the To this is the first of the functional of an ARE in vivo in no of translation was or no factor was several have that unstable mRNA and that an ARE-containing mRNA to be to be (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, 20Mitchell P. Tollervey D. Curr. Opin. Cell Biol. 2001; 13: 320-325Google Scholar, A.B. Cell. Scholar). expression of many had a effect on ARE-containing mRNA N. A.B. Mol. Cell. Biol. 2001; 21: for 19Guhaniyogi J. Brewer G. Gene (Amst.). 2001; 265: 11-23Google Scholar). of ARE-mediated mRNA decay because translation from an that the may be to at the 5′-end of the as This is in with data in mammalian the ARE decapping in HeLa (7Gao M. Wilusz C.J. Peltz S.W. Wilusz J. EMBO J. 2001; 20: 1134-1143Google Scholar), and in vitro the GM-CSF was to the translational by the GM-CSF ARE in the 3′-UTR J.A. Rajagopalan L.E. Shin H.C. Malter J.S. Nucleic Acids Res. 1999; 27: 3660-3666Google Scholar). this we with a containing both GM-CSF 5′ and and of expression by and and In the of the GM-CSF the ARE functions on either mRNA decay or translation. and J. This could be by the of However, in their J.A. Rajagopalan L.E. Shin H.C. Malter J.S. Nucleic Acids Res. 1999; 27: 3660-3666Google Scholar) show that the GM-CSF ARE could translation of a independently of its effect on the mRNA in vivo these in vitro we showed that expression by either or translation is by the GM-CSF ARE, whereas ARE-containing mRNAs in the cells Whereas the cap-dependent translation initiation of ARE-containing mRNAs by either or in the mRNA of translation by to mRNA stabilization (1Wilusz C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, N. Mol. Cell. Biol. Scholar, A.B. Cell. Scholar). Because the ARE is functional as a destabilizer when translation from an IRES, we that the is to the cap-dependent translation process C.J. Wormington M. Peltz S.W. Nat. Rev. Mol. Cell Biol. 2001; 2: 237-246Google Scholar, 20Mitchell P. Tollervey D. Curr. Opin. Cell Biol. 2001; 13: 320-325Google Scholar, and A.B. Cell. Scholar). data the of and Tollervey P. Tollervey D. Curr. Opin. Cell Biol. 2001; 13: 320-325Google Scholar), the initiation of cap-dependent translation could be a of for the ARE-mediated decay to could be that the cap is to the decapping complex cap-dependent ribosome an ARE is present with its associated the decapping may be highly because the decapping complex is the initiation factor tail is by a less efficient binding of the PABP on the poly(A) tail. translation is an IRES, the ARE no longer be to decapping mRNA body ribosome no longer on the cap and its associated but it still could the translation by in the tail association. this the stabilization of ARE-containing mRNAs by of translation may also be by a of the because are no longer on the R. J. Biol. 1994; Scholar). In mRNA decay may a in both yeast and mammalian decapping deadenylation and RNA body investigations need to be to the of the decapping in the of the ARE-mediated mRNA decay and to the regulatory factors involved in this We P. J. G. J. A. and J. L. for and J. J. and R. for is for