- Email: [email protected]
Degradation of mRNA in eukaryotes
Cell, Voi. 81, 179-188, April 21, 1995, Copyright© 1995 by Cell Press
Review
Degradation of mRNA in Eukaryotes
Clare A. Beelman* and Roy Parker*t *Department of Molecular and Cellular Biology tHoward Hughes Medical Institute University of Arizona Tucson, Arizona 85721
decay independently of deadenylation. This diversity of decay pathways, in addition to different rates of decay for individual mRNAs within one pathway, allows for a wide spectrum of mRNA half-lives and for their differential regulation.
mRNA turnover is important in determining the levels and regulation of gene expression. Recent results have defined several different, yet somewhat related, mechanisms by which eukaryotic mRNAs are degraded (Figure 1). One mRNA decay pathway is initiated by shortening of the poly(A) tail followed by decapping and 5' to 3' exonucleolytic degradation of the transcript. A variation of this pathway has been observed in which transcripts undergo 3' to 5' decay after poly(A) shortening. Decay of mRNAs can also initiate prior to shortening of the poly(A) tail. For example, some specific transcripts can be degraded via deadenylation-independent decapping followed by 5' to 3' degradation. In addition, mRNA decay can also begin by endonucleolytic cleavage in the transcript body. These pathways suggest a model of mRNA turnover in which all polyadenylated mRNAs are degraded by a"default" pathway initiated by poly(A) shortening. In addition, pathways limited to subsets of mRNAs exist in which specific sequences trigger
Deadenylation Can Be the First Step in mRNA Decay Several observations suggest that shortening of poly(A) tails is required for the decay of many eukaryotic mRNAs (for reviews see Decker and Parker, 1994; Bernstein and Ross, 1989; Peltz et al., 1991). Some of the key observations are as follows. First, in transcriptional pulse-chase experiments (wherein a regulatable promoter is used to produce a burst of synchronous transcription), the mammalian c-los mRNA and several yeast mRNAs do not decay until their poly(A) tails have been shortened (Shyu et al., 1991; Decker and Parker, 1993), This temporal correlation is significant because mutations within unstable mRNAs that decrease the rate of deadenylation or transfer of sequences capable of stimulating rapid deadenylation to a stable reporter mRNA correspondingly alter the time at which the transcript begins to decay (Shyu et al., 1991; Wilson and Triesman, 1988; Muhlrad and Parker, 1992; Decker and Parker, 1993; Chen et al., 1994). In addition,
DEADENYLATION-DEPENDENT DECAY
DEADENYLATION-INDEPENDENT DECAY
UM
m7GI AUG
POLY (A) SHORTENING
m7GI AUG
1
UAA
T/
DEADENLYATION- /
ENDONUCLEOL'gTIC
CLEAVAGE
JAAA LOt~,
I AAA ou~
I/gkA u~a
i
m7G[ AUG
/
5' TO 3' DECAY
?
DECAPPING ~AOG l i ~ ; /
u.
\ 3' TO 5' DECAY
IAAA~ 5' TO 3' DECAY
5'TO 3' DECAY ~;,oG
o"
I~ ou~o
Figure 1. mRNA Decay Pathwaysin Eukaryotes The left pathway depicts deadenylation-dependentdecay, which can lead to decapping and 5' to 3' exonucleolyticdecay or to 3' to 5' decay mediated by endonucleases,exonucleases,or both. All eukaryotic mRNAs may undergo decay via this deadenylation-dependentpathwayunless they are targeted for rapid deadenlyation-independentdecapping, by a nonsensecodon,or for endonucleolyticcleavage,by presenceof a cleavage site (right). After deadenylation-independentdecapping, mRNA is degraded in a 5' to 3' manner. Endonucleolyticcleavage of mRNA may serve as a one-step deadenylationthat leads to decapping and 5' to 3' decay.
Cell 180
intermediates in decay can be detected in yeast that accumulate after deadenylation and have oligo(A) tails, even at times when the population consists of a mixture of poly(A) and oligo(A) tails (Decker and Parker, 1993; Muhltad et al., 1994). The length to which a poly(A) tail must be shortened for subsequent decay may differ between mammals and yeast. A poly(A) tail of - 25-60 adenosine residues is sufficient to allow decay of some mammalian transcripts (Shyu et al., 1991; Chen et al., 1994), whereas yeast transcripts decay when the poly(A) is - 1 0 - 1 2 adenosine residues (Decker and Parker, 1993). Deadenylation to - 10-12 residues as seen in yeast might lead to the loss of the last poly(A)-binding protein associated with the transcript, and the loss of poly(A)-binding protein may trigger further decay events. It is possible that the disassociation of poly(A)binding protein may occur at a longer poly(A) tail length in mammalian cells.
Deadenylation Can Trigger Decapping and 5' to 3' Decay of mRNA One important question is how deadenylation leads to mRNA degradation. In yeast, observations indicate that deadenylation to an oligo(A) tail allows the mRNA to become a substrate for a decapping reaction within the first few nucleotides of the transcript, thereby exposing the transcript body to 5' to 3' exonucleolytic decay. The critical observations are that when 5' to 3' decay is blocked in yeast, either by deletion of the XRN1 gene, which encodes a major 5' to 3' exonuclease (Larimer and Stevens, 1990), or by the insertion of strong RNA secondary structures, mRNA fragments lacking the cap structure accumulate after deadenylation (Muhlrad et al., 1994, 1995). This is true for both unstable and stable mRNAs. In addition, several other mRNAs also accumulate in xrnl/t strains as transcripts that have, at most, short (A) tails and that lack the cap structure (Hsu and Stevens, 1993). These results suggest this pathway of degradation is a general mechanism of decay that acts on many yeast transcripts. Since poly(A) tails and cap structures are common features of eukaryotic transcripts, an appealing model is that mRNA decay by deadenylation-dependent decapping and 5' to 3' digestion is a conserved mechanism of m R NA tu rnover. Although there is to date no direct data for such a pathway in more complex eukaryotes, mRNAs lacking the cap structure are rapidly degraded in many eukaryotic cells (e.g., Drummond et al., 1985). In addition, enzymes that could catalyze the removal of the cap structure and subsequent 5' to 3' degradation of the transcript have been described in mammalian cells (e.g., Coutts and Brawerman, 1993). Finally, the oat phytochrome A mRNA undergoes a cleavage near the 5' terminus of the mRNA and is degraded, in part, in a 5' to 3' direction (Higgs and Colbert, 1994). Taken together, these data raise the possibility that deadenylation-dependent decapping followed by 5' to 3' exonucleolytic decay may be a conserved eukaryotic mRNA decay mechanism. Interactions between the 5' and 3' Termini of mRNAs The observation that deadenylation can lead to removal
of the 5'cap structure (Muhlrad et al., 1994, 1995) suggests an interaction between the 5' and 3' ends of an mRNA. Interactions between the transcript termini have been proposed to explain the effects that the poly(A) tail and 3' untranslated region (3'UTR) sequences can have on translation initiation (Munroe and Jacobson, 1990; for review see Jackson and Standart, 1990). A unifying hypothesis is that the same 5' to 3' interaction may mediate the rates of translation initiation and decapping. Although there is no direct biochemical evidence for such a 5' to 3' interaction, the abundance of circular polysomes seen in some electron micrographs suggests that the transcript termini can be in close proximity (e.g., Christensen et al., 1987). The nature of the 5' to 3' interaction may allow the poly(A) tail to inhibit decapping indirectly or directly. For example, because mRNAs may interact with the cytoskeleton through their poly(A) tails (e.g., Taneja et al., 1992), the putative 5' to 3' interaction may cause deadenyiation to alter the subcellular localization of an mRNA, thereby exposing it to a localized decapping activity. Alternatively, the poly(A) tail might inhibit decapping directly by forming or stabilizing an mRNP structure involving the 5' and 3' termini of the mRNA.
Deadenylation Can Also Lead to 3' to 5' Decay of mRNA Eukaryotic mRNAs can also be degraded in a 3' to 5' direction following deadenylation. For example, fragments of the yeast PGK1 mRNA shortened at the 3' end accumulate when the 5' to 3' decay pathway is blocked (Muhlrad and Parker, 1994; Muhlrad et al., 1995). Decay intermediates that are consistent with 3' to 5' decay following deadenylation have also been observed for the oat phytochrome A mRNA in vivo (Higgs and Colbert, 1994). It is not known whether exonucleases, endonucleases, or both are involved in 3'to 5'decay or how prevalent this decay pathway may be. However, it should not be expected that deadenylation necessarily exposes transcripts to 3' to 5' degradation since several mRNAs can be quite stable with essentially no poly(A) tail (e.g., Decker and Parker, 1993; Chert et al., 1994). An important point is that an individual transcript can simultaneously be a substrate for more than one mechanism of decay. For example, both the oat phytochrome A and yeast PGK1 transcripts undergo deadenylation-dependent 3' to 5' decay in addition to a 5' to 3' decay mechanism (Higgs and Colbert, 1994; Muhlrad et al., 1995). The presence of multiple overlapping pathways has several consequences. First, the observed half-life for an mRNA will be a summation of the rates of decay through each individual pathway. In addition, the mechanism by which an mRNA is degraded may change under different conditions, even without a significant alteration of decay rate (e.g., Muhlrad et al., 1995). The existence of multiple decay pathways also suggests that mutations that inactivate one pathway may not always have significant effects on mRNA stability, yet combinations of mutations inactivating different pathways may exhibit synergistic effects on stability and possibly viability. A system of redundant overlapping mechanisms of mRNA turnover is highly analogous to that
Review:rnRNADegradation 181
observed in Escherichia coil (for review see Belasco and Brawerman, 1993).
In deadenylation-dependent mRNA decay, differences in mRNA stability result from differences in the rates of poly(A) shortening and steps required for the subsequent decay of the deadenylated transcript. Thus, features of mRNAs that influence the rates of poly(A) shortening, decapping, and possibly 3' to 5' decay will affect mRNA stability. The rate of 5' to 3' digestion of the transcript body appears to be relativelyfast and therefore unlikely to contribute to differences in decay rates. This is based on the observation that intermediates of 5' to 3' decay are not observed unless the XRN1 nuclease has been deleted or a secondary structure blocks the nuclease (Decker and Parker, 1993; Hsu and Stevens, 1993; Muhlrad et al., 1994). Several sequence elements within mRNAs that promote rapid poly(A) shortening have been identified. These include sequences within the c-fos coding region and 3'UTR, which contains a prototypical AU-rich element (ARE) (Wilson and Triesman, 1988; Shyu et al., 1989; Wellington et al., 1993; Chen et al., 1994; Schiavi et al., 1994), and in the yeast MFA2 3'UTR (Muhlrad and Parker, 1992; Decker and Parker, 1993; Muhlrad et al., 1994). Interestingly, mutations in the c-los ARE and the MFA2 3'UTR can also slow degradation following deadenylation (Shyu et al., 1991; Muhlrad and Parker, 1992). Thus, at least for the MFA2 mRNA, 3'UTR sequences can also specify the rate of decapping. There is little information about the nucleases involved in deadenylation-dependent decay and how mRNA sequences modulate their rate of activity. The product of the XRN1 gene appears to be the primary nuclease responsible for 5'to 3' degradation of yeast mRNAs after decapping (Hsu and Stevens, 1993; Muhlrad et al., 1994). Decapping activities (Nuss et al., 1975; Kumagai et al., 1992; Coutts and Brawerman, 1993) and poly(A) nucleases (,~str0m et al., 1991, 1992; Sachs and Deardorff, 1992; Lowell et al., 1992) have been biochemically defined in cell extracts of yeast and mammals, although the in vivo roles of these particular activities remain to be established. Several proteins have been identified that interact with the ARE (Malter, 1989; Bohjanen et al., 1991; Vakalopoulou, 1991; Chen et al., 1992; Zhang et al., 1993; Katz et al., 1994) and therefore might modulate deadenylation rate. However, only one of these proteins, called the ARE/poly(U)binding factor, has been shown to stimulate decay in an in vitro system (Brewer, 1991). Progress in this area may be aided by the identification of proteins that bind to a small sequence, UUAUUUA(U/A)(U/A), determined to be important for a functional ARE to stimulate mRNA decay (Lagnado et al., 1994; Zubiaga et al., 1995). ~
(Muhlrad and Parker, 1994; Hagan et al., 1995). The degradation of mRNAs with nonsense codons is part of a process, termed mRNA surveillance, that ensures the rapid degradation of aberrant transcripts. These aberrant mRNAs contain early nonsense codons (Losson and Lacroute, 1979; Leeds et al., 1991; Peltz et al., 1993; Pulak and Anderson, 1993), unspliced introns (He et al., 1993), or extended 3'UTRs (Pulak and Anderson, 1993). mRNA surveillance occurs in many organisms and thus appears to be a conserved process. Mutations in related genes that inactivate this surveillance pathway have been identified in yeast and nematodes, named upland smg mutants, respectively (Hodgkin et al., 1989; Leeds et al., 1991, 1992). mRNA surveillance may exist, in part, to increase the fidelity of gene expression by degrading aberrant mRNAs that, if translated, would produce truncated proteins. This process would be relevant since truncated proteins can have dominant negative phenotypes. A striking example of this phenomenon is seen in Caenorhabditis elegans, where mutations in smg genes convert recessive nonsense mutations in the myosin gene uric-54 into dominant negatives (Pulak and Anderson, 1993). Since mRNA surveillance degrades unspliced and aberrantly processed transcripts, this decay mechanism might be most important in organisms with a large number of introns where processing errors might lead to truncated proteins. How a nonsense-containing transcript is recognized as aberrant and how mRNA decay is triggered are unclear. The simplest model is that premature translation termination sends a signal to expose the 5' end immediately to decapping by the same nucleases that degrade normal mRNAs. This hypothesis is supported by the observations that the sites of decapping are the same in deadenylationindependent and deadenylation-dependent decapping (cf. Muhlrad and Parker, 1994 with Muhlrad et al., 1995), and the XRN1 exonuclease performs 5' to 3' decay in both cases. The generation of the destabilizing signal requires both the failure to translate a significant part of the coding region and the presence of specific sequences downstream of the nonsense codon (Cheng et al., 1990; Peltz et al., 1993; Zhang et al., 1995). In a similar mechanism, transcripts with extended 3'UTRs may also be recognized and degraded because destabilizing sequences that are not normally in the mRNA are now included within the extended 3'UTR. A related and surprising observation is that early nonsense mRNAs can reduce the levels of fully spliced but nuclear-associated transcripts (Belgrader et al., 1994; Baserga et at., 1992). One possibility is that the first round of translation occurs while the transcripts are still nuclear associated and that premature termination during this round of translation can generate the signal to expose the 5' end to decapping.
Deadenylation-lndependent Decapping: mRNA Surveillance
Decay of EukaryoUc mRNA via Endonucleolytic Cleavage
mRNA decay can also be initiated by decapping and 5' to 3'decay of the transcript independent of poly(A) shortening. An example of this process is the degradation of the yeast PGK1 mRNA containing an early nonsense codon
Eukaryotic mRNAs can be degraded via endonucleolytic cleavage prior to deadenylation. Evidence for this mechanism comes from the analysis of transcripts such as mammalian 9E3, IGF2, transferrin receptor (TfR), and Xenopus
Determination of mRNA Decay Rates via Deadenylation-Dependent Decay
Cell 182
X/hbox2B mRNA where mRNA fragments are detected in
vivo that correspond to the 5' and 3' portion of the transcript and are consistent with internal cleavage within the 3'UTR (Stoekle and Hanafusa, 1989; Nielson and Christiansen, 1992; Binder et al., 1994; Brown et al., 1993). It appears that deadenylation to an oligo(A) tail length is not required for endonucleolytic cleavage since the 3' fragment of 9E3 and TfR mRNAs is polyadenylated(Stoekle and Hanafusa, 1989; Binder et al., 1994) and since cleavage of the Xlhbox2B mRNA is not affected by the adenylation state of the mRNA (Brown et al., 1993). However, it is possible that endonucleolytic cleavage of some mRNAs could be dependent on the length of the poly(A) tail. Endonucleolytic cleavages have also been defined in vitro for the albumin mRNA (Dompenciel et al., 1995) and in the coding region of the c-myc mRNA (Bernstein et al., 1992). Since there does not appear to be any similarity between the cleavage sites in these mRNAs, there may be a wide variety of endonucleases with different cleavage specificities. Since sequence-specific endonuclease target sites are likely to be limited to individual mRNAs or classes of mRNAs, their presence allows for specific control of the decay rate of these transcripts. In some cases, the rate of endonucleolytic cleavage is modulated by the activity of protective factors that bind at or near the cleavage site and compete with the endonuclease (Brown et al., 1993; Binder et al., 1994). For example, the binding of the iron response element-binding protein in the TfR 3'UTR in response to low intracellular iron concentrations inhibits the endonucleolytic cleavage of this mRNA (Binder et al., 1994). Therefore, some endonucleases may be constitutively active, and the accessibility of the cleavage site is regulated. In other cases, the endonuclease activity may be directly regulated. For example, the mammalian endonuclease RNase L is normally inactive and is only activated by oligomers of 2',5' phosphodiester-bonded adenylate residues, which are produced in response to the presence of doublestranded RNA (for discussion see Silverman, 1994). Although RNase L is important in mediating interferon responses (e.g., Hassel et al., 1993), it has yet to be established whether this enzyme normally degrades any cellular mRNAs.
Summary
Based on the above mechanisms of mRNA degradation, an integrated model of mRNA turnover can be proposed (Figure 1). In this model, all polyadenylated mRNAs would be degraded by the deadenylation-dependentpathway at some rate. In addition to this default pathway, another layer of complexity would come from degradation mechanisms specific to individual mRNAs or to classes of mRNAs. Such mRNA-specific mechanisms would include sequence-specific endonuclease cleavage and deadenylation-independent decapping. Thus, the overall decay rate of an individual transcript will be a function of its susceptibility to these turnover pathways. In addition, cisacting sequences that specify mRNA decay rate, as well
as regulatory inputs that control mRNA turnover, are likely to affect all the steps of these decay pathways. One important goal in future work will be to identify the gene products that are responsible for the nucleolytic events in these pathways and to delineate how specific mRNA features act to affect the function of these degradative activities. The identification of distinct mRNA decay pathways should allow genetic and biochemical approaches that can be designed to identify these gene products. A second important goal is to understand the nature of the interaction between the 5' and 3' termini, which may also be critical for efficient translation. References
/~str~m, J., ,~str~m,A., and Virtanen,A. (1991).In vitro deadenylation of mammalianmRNA by a HeLa cell 3' exonuclease.EMBOJ. 10, 3067-3071. /~str6m, J.,/~.strGm,A., and Virtanen,A. (1992). Propertiesof a HeLa cell 3' exonucleasespecificfor degradingpoly(A)tails of mammalian mRNA. J. Biol. Chem. 267, 18154-18159. Baserga,S. J., and Benz,E.J., Jr. (1992).I~-Globinnonsensemutation: deficientaccumulationof mRNAoccurs despitenormalcytoplasmic stability. Proc~Natl. Acad. Sci. USA 89, 2935--2939. Belasco,J., and Brawerman,G. (1993). Controlof MessengerRNA Stability (NewYork: AcademicPress). Belgrader,P., Cheng,J., Zhou,X., Stephenson,S., and Maquat,L. E. (1994). Mammaliannonsensecodonscan be cis effectorsof nuclear mRNA half-life.MoL Cell. Biol. 14, 8219-8228. Bernstein, P., and Ross, J. (1989). Poly(A), poly(A) binding protein and the regulationof mRNAstability.Trends Biochem.Sci. 14, 373377. Bernstein, P. L., Herrick, D. J., Prokipcak,R. D., and Ross,J. (1992). Controlof c-myc mRNAhalf-lifein vitro by a proteincapableof binding to a coding region stabilitydeterminant.GenesDev. 6, 642-654. Binder, R., Horowitz,J. A., Basilion,J. P., Koeller, D. M., Klausner, R. D., and Harford,J. B. (1994). Evidencethat the pathwayof transferrin receptormRNAdegradationinvolvesan endonucleolyticcleavage within the 3' UTR and does not involvepoly(A)tail shortening. EMBO J. 13, 1969-1980. Bohjanen, P. R., Bronislawa,P., June, C. H., Thompson,C. B., and Lindsten,T. (1991).An induciblecytoplasmicfactor(AU-B)bindsselectivelyto AUUUAmultimersin the 3' untranslatedregionof lymphokine mRNA. Mol. Cell. Biol. 11, 3288-3295. Brewer,G. (1991).An A+U-richelementRNA-bindingfactorregulates c-myc mRNAstabilityin vitro. Mol. Cell. Biol. 11, 2460-2466. Brown, B. D., Zipkin, I. D., and Harland, R. M. (1993). Sequencespecificendonucleolyticcleavageand protectionof mRNAin Xenopus and Drosophila. Genes Dev. 7, 1620-1631. Chen, C.-Y.A., You,Y., and Shyu,A.-B. (1992).Two cellularproteins bind specificallyto a purine-richsequencenecessaryfor the destabilizationfunctionof a c-fos proteincoding regiondeterminantof mRNA instability. Mol. Cell. Biol. 12, 5748-5757. Chen, C.-Y. A., Chen, T.-M., and Shyu,A.-B. (1994). Interplayof two functionallyand structurallydistinct domainsof the c-fos AU-richelement specifies its mRNA-destabilizingfunction. Mol. Cell. Biol. 14, 416-426. Cheng, J., FogeI-Petrovic,M., and Maquat, L~E. (1990). Translation to nearthe distal end of the penultimateexon is requiredfor normal levels of splicedtriosephosphateisomerasemRNA. Mol. Cell. Biol. 10, 5215-5225. Christensen, A. K., Kahn, L. E., and Bourne,C. M. (1987). Circular polysomespredominateon the roughendoplasmicreticulumof somatropes and mammotropesin the rat anteriorpituitary.Am. J. Anat. 178, 1-10. C0utts, M., and Brawerman,G. (1993). A 5' exoribonucleasefrom
Review: mRNA Degradation 183
cytoplasmic extracts of mouse sarcoma 180 ascites cells. Biochim. Biophys. Acta 1173, 57-62. Decker, C. J., and Parker, R. (1993). A turnover pathway for both stable and unstable mRNAs in yeast: evidence for a requirement for deadenylation. Genes Dev. 7, 1632-1643. Decker, C. J., and Parker, R. (1994). Mechanisms of mRNA degradation in eukaryotes. Trends Biochem. Sci. lg, 336-340. Dompenciel, R. E., Garnepudi, V. R., and Schoenberg, D. R. (1995). Purification and characterization of an estrogen-regulated Xenopus liver polysomal nuclease involved in the selective destabilization of albumin mRNA. J. Biol. Chem., 270, 6108-6118. Drummond, D. R., Armstrong, J., and Colman, A. (1985). The effect of capping and polyadenylation on the stability, movement and translation of synthetic messenger RNAs in Xenopus oocytes. Nucl. Acids Res. 13, 7375-7394. Hagan, K. W., Ruiz-Echervarda, M. J., Quan, Y., and Peltz, S. W. (1995). Characterization of cis-acting sequences and decay intermediates involved in nonsense-mediated mRNA turnover. Mol. Cell. Biol. 15, 809-823. Hassel, B. A., Zhou, A., Sotomayor, C., Maran, A., and Silverman, R. H. (1993). A dominant negative mutant of 2-5A-dependent RNase suppresses antiproliferative and antiviral effects of interferon. EMBO J. 12, 3297-3304. He, F., Peltz, S. W., Donahue, J. L., Rosbash, M., and Jacobson, A. (1993). Stabilization and ribosome association of unspliced premRNAs in a yeast upfl- mutant. Proc. Natl. Acad. Sci. USA 90, 70347038. Higgs, D. C., and Colbert, J. T. (1994). Oat phytochrome A mRNA degradation appears to occur via two distinct pathways. Plant Cell 6, 1007-1019. Hodgkin, J., Papp, A., Pulak, R., Am bros, V., and Anderson, P. (1989). A new kind of informational suppression in the nematode Caenorhabditis elegans. Genetics 123, 301-313. Hsu, C. L., and Stevens, A. (1993). Yeast cells lacking 5' to 3' exoribonuclease 1 contain mRNA species that are poly(A) deficient and partially lack the 5' cap structure. Mol. Cell. Biol. 13, 4826-4835. Jackson, R. J., and Standart, N. (1990). Do the poly(A) tail and 3' untranslated region control mRNA translation? Cell 62, 15-24. Katz, D. A., Theodorakis, N. G., Cleveland, D. W., Lindsten, T., and Thompson, C. B. (1994). AU-A, an RNA-binding activity distinct from hnRNP A1, is selective for AUUUA repeats and shuttles between the nucleus and the cytoplasm. Nucl. Acids Res. 22, 238-246. Kumagai, H., Kon, R., Hoshino, T., Aramaki, T., Nishikawa, M., Hirose, S., and Igarashi, K. (1992). Purification and properties of a decapping enzyme from rat liver cytosol. Biochim. Biophys. Acta 13, 45-51. Lagnado, C. A., Brown, C. L., and Goodall, G. J. (1994). AUUUA is not sufficient to promote poly(A) shortening and degradation of an mRNA: the functional sequence within AU-rich elements may be UUAUUUA(U/A)(U/A). Mol. Cell. Biol. 14, 7984-7995. Larimer, F. W., and Stevens, A. (1990). Disruption of the gene XRN1, coding for a 5'--3' exoribonuclease, restricts yeast cell growth. Gene 95, 85-90. Leeds, P., Peltz, S. W., Jacobson, A., and Culbertson, M. R. (1991). The product of the yeast UPF1 gene is required for rapid turnover of mRNAs containing a premature translational termination codon. Genes Dev. 5, 2303-2314. Leeds, P., Wood, J. M., Lee, B.-S., and Culbertson, M. R. (1992). Gene products that promote mRNA turnover in Saccharomyces cerevisiae. Mol. Cell. Biol. 12, 2165-2177. Losson, R., and Lacroute, F. (1979). Interference of nonsense mutations with eukaryotic messenger RNA stability. Proc. Natl. Acad. Sci. USA 76, 5134-5137. Lowell, J., Rudner, D., and Sachs, A. (1992). 3'-UTR-dependent deadenylation by the yeast poly(A) nuclease. Genes Dev. 6, 2088-2099~ Malter, J. S. (1989). Identification of an AUUUA-specific messenger RNA binding protein. Science 246, 664-666. Muhlrad, D., and Parker, R. (1992). Mutations affecting stability and deadenylation of the yeastMFA2transcript. Genes Dev. 6, 2100-2111.
Muhlrad, D., and Parker, R. (1994). Premature translational termination triggers mRNA decapping. Nature 340, 578-581. Muhlrad, D., Decker, C. J., and Parker, R. (1994). Deadenylation of the unstable mRNA encoded by the yeast MFA2 gene leads to decapping followed by 5' to 3' digestion of the transcript. Genes Dev. 8, 855866. Muhlrad, D., Decker, C. J., and Parker, R. (1995). Turnover mechanisms of the stable yeast PGK1 mRNA. Mol. Cell. Biol., 16, 2145-2156. Munroe, D., and Jacobson, A. (1990). mRNA poly(A) tail, a 3' enhancer of translational initiation. Mol. Cell. Biol. 10, 3441-3455. Nielson, F. C., and Christiansen, J. (1992). Endonucleolysis in the turnover of insulin-like growth factor II mRNA. J. Biol. Chem. 267, 19404-19411. Nuss, D. L., Furuichi, Y., Kock, G., and Shatkin, S. J. (1975). Detection in HeLa cell extracts of a 7-methyl guanosinb specific enzyme activity that cleaves mTGpppN m. Cell 6, 21-27. Peltz, S. W., Brewer, G., Bernstein, P., Hart, P., and Ross, J. (1991). Regulation of mRNA turnover in eukaryotic cells. CRC Crit. Rev. Euk. Gene Exp. 1, 99-126. Peltz, S. W., Brown, A. H., and Jacobson, A. (1993). mRNA destabilization triggered by premature translational termination depends on three mRNA sequence elements and at least one trans-acting factor. Genes Dev. 7, 1737-1754. Pulak, R., and Anderson, P. (1993). mRNA surveillance by the Caenorhabditis elegans smg genes. Genes Dev. 7, 1885-1897. Sachs, A. B., and Deardorff, J. (1992). Translation initiation requires the PAB-dependent poly(A) ribonuclease in yeast. Cell 70, 961-973. Schiavi, S. C., Wellington, C. L., Shyu, A.-B., Chen, C.-Y. A., Greenberg, M. E., and Belasco, J. G. (1994). Multiple elements in the c-fos protein coding region facilitate mRNA deadenylation and decay by a mechanism coupled to translation. J. Biol. Chem. 269, 34413448. Shyu, A.-B., Greenberg, M. E., and Belasco, J. G. (1989). The c-fos transcript is targeted for rapid decay by two distinct mRNA degradation pathways. Genes Dev. 3, 60-72. Shyu, A.-B., Belasco, J. G., and Greenberg, M. E. (1991). Two distinct destabilizing elements in the c-fos message trigger deadenylation as a first step in rapid mRNA decay. Genes Dev. 5, 221-234. Silverman, R. H. (1994). Fascination with 2-5A-dependent RNase: a unique enzyme that functLons in interferon action. J. Interferon Res. 14, 101-104. Stoekle, M. Y., and Hanafusa, H. (1989). Processing of 9E3 mRNA and regulation of its stability in normal and rous sarcoma virus-transformed cells. Mol. Cell. Biol. 9, 4738-4745. Taneja, K. L, Lifshitz, L. M., Fay, F. S., and Singer, R. H. (1992). Poly(A) RNA codistribution with microfilaments: evaluation by in situ hybridization and quantitative digital imaging microscopy. J. Cell Biol. 119, 1245-1260. Vakalopoulou, E., Schaack, J., and Shenk, T. (1991). A 32-kilodalton protein binds to AU-rich domains in the 3'untranslated region of rapidly degraded mRNAs. MoI. Cell. Biol. 11, 3355-3364. Wellington, C. L., Greenberg, M. E., and Belasco, J. G. (1993). The destabilizing elements in the coding region of c-fos mRNA are recognized as RNA. Mol. Cell. Biol. 13, 5034-5042. Wilson, T, and Triesman, R. (1988). Removal of poly(A) and consequent degradation of the c-fos mRNA facilitated by 3' AU-rich sequences. Nature 336, 396-399. Zhang, A., Ruiz-Echevarria, M. J., Quan, Y., and Peltz, S. W. (1995). Identification and characterization of a sequence motif involved in nonsense-mediated mRNA decay. Mol. Cell. Biol., 15, 2231-2244. Zhang, W., Wagner, B. J., Ehrenman, K., Schaefer, A. W., DeMaria, C. T., Crater, D., DeHaven, K., Long, L., and Brewer, G. (1993). Purification, characterization, and cDNA cloning of an AU-rich element RNA-binding protein, AUFI. Mol. Cell. Biol. 13, 7652-7665. Zubiaga, A. M., Belasco, J. G., and Greenberg, M. E. (1995). The nonamer UUAUUUAUU is the key AU-rich sequence motif that mediates mRNA degradation. Mol. Cell. Biol. 15, 2219-2230.
Review
Degradation of mRNA in Eukaryotes
Clare A. Beelman* and Roy Parker*t *Department of Molecular and Cellular Biology tHoward Hughes Medical Institute University of Arizona Tucson, Arizona 85721
decay independently of deadenylation. This diversity of decay pathways, in addition to different rates of decay for individual mRNAs within one pathway, allows for a wide spectrum of mRNA half-lives and for their differential regulation.
mRNA turnover is important in determining the levels and regulation of gene expression. Recent results have defined several different, yet somewhat related, mechanisms by which eukaryotic mRNAs are degraded (Figure 1). One mRNA decay pathway is initiated by shortening of the poly(A) tail followed by decapping and 5' to 3' exonucleolytic degradation of the transcript. A variation of this pathway has been observed in which transcripts undergo 3' to 5' decay after poly(A) shortening. Decay of mRNAs can also initiate prior to shortening of the poly(A) tail. For example, some specific transcripts can be degraded via deadenylation-independent decapping followed by 5' to 3' degradation. In addition, mRNA decay can also begin by endonucleolytic cleavage in the transcript body. These pathways suggest a model of mRNA turnover in which all polyadenylated mRNAs are degraded by a"default" pathway initiated by poly(A) shortening. In addition, pathways limited to subsets of mRNAs exist in which specific sequences trigger
Deadenylation Can Be the First Step in mRNA Decay Several observations suggest that shortening of poly(A) tails is required for the decay of many eukaryotic mRNAs (for reviews see Decker and Parker, 1994; Bernstein and Ross, 1989; Peltz et al., 1991). Some of the key observations are as follows. First, in transcriptional pulse-chase experiments (wherein a regulatable promoter is used to produce a burst of synchronous transcription), the mammalian c-los mRNA and several yeast mRNAs do not decay until their poly(A) tails have been shortened (Shyu et al., 1991; Decker and Parker, 1993), This temporal correlation is significant because mutations within unstable mRNAs that decrease the rate of deadenylation or transfer of sequences capable of stimulating rapid deadenylation to a stable reporter mRNA correspondingly alter the time at which the transcript begins to decay (Shyu et al., 1991; Wilson and Triesman, 1988; Muhlrad and Parker, 1992; Decker and Parker, 1993; Chen et al., 1994). In addition,
DEADENYLATION-DEPENDENT DECAY
DEADENYLATION-INDEPENDENT DECAY
UM
m7GI AUG
POLY (A) SHORTENING
m7GI AUG
1
UAA
T/
DEADENLYATION- /
ENDONUCLEOL'gTIC
CLEAVAGE
JAAA LOt~,
I AAA ou~
I/gkA u~a
i
m7G[ AUG
/
5' TO 3' DECAY
?
DECAPPING ~AOG l i ~ ; /
u.
\ 3' TO 5' DECAY
IAAA~ 5' TO 3' DECAY
5'TO 3' DECAY ~;,oG
o"
I~ ou~o
Figure 1. mRNA Decay Pathwaysin Eukaryotes The left pathway depicts deadenylation-dependentdecay, which can lead to decapping and 5' to 3' exonucleolyticdecay or to 3' to 5' decay mediated by endonucleases,exonucleases,or both. All eukaryotic mRNAs may undergo decay via this deadenylation-dependentpathwayunless they are targeted for rapid deadenlyation-independentdecapping, by a nonsensecodon,or for endonucleolyticcleavage,by presenceof a cleavage site (right). After deadenylation-independentdecapping, mRNA is degraded in a 5' to 3' manner. Endonucleolyticcleavage of mRNA may serve as a one-step deadenylationthat leads to decapping and 5' to 3' decay.
Cell 180
intermediates in decay can be detected in yeast that accumulate after deadenylation and have oligo(A) tails, even at times when the population consists of a mixture of poly(A) and oligo(A) tails (Decker and Parker, 1993; Muhltad et al., 1994). The length to which a poly(A) tail must be shortened for subsequent decay may differ between mammals and yeast. A poly(A) tail of - 25-60 adenosine residues is sufficient to allow decay of some mammalian transcripts (Shyu et al., 1991; Chen et al., 1994), whereas yeast transcripts decay when the poly(A) is - 1 0 - 1 2 adenosine residues (Decker and Parker, 1993). Deadenylation to - 10-12 residues as seen in yeast might lead to the loss of the last poly(A)-binding protein associated with the transcript, and the loss of poly(A)-binding protein may trigger further decay events. It is possible that the disassociation of poly(A)binding protein may occur at a longer poly(A) tail length in mammalian cells.
Deadenylation Can Trigger Decapping and 5' to 3' Decay of mRNA One important question is how deadenylation leads to mRNA degradation. In yeast, observations indicate that deadenylation to an oligo(A) tail allows the mRNA to become a substrate for a decapping reaction within the first few nucleotides of the transcript, thereby exposing the transcript body to 5' to 3' exonucleolytic decay. The critical observations are that when 5' to 3' decay is blocked in yeast, either by deletion of the XRN1 gene, which encodes a major 5' to 3' exonuclease (Larimer and Stevens, 1990), or by the insertion of strong RNA secondary structures, mRNA fragments lacking the cap structure accumulate after deadenylation (Muhlrad et al., 1994, 1995). This is true for both unstable and stable mRNAs. In addition, several other mRNAs also accumulate in xrnl/t strains as transcripts that have, at most, short (A) tails and that lack the cap structure (Hsu and Stevens, 1993). These results suggest this pathway of degradation is a general mechanism of decay that acts on many yeast transcripts. Since poly(A) tails and cap structures are common features of eukaryotic transcripts, an appealing model is that mRNA decay by deadenylation-dependent decapping and 5' to 3' digestion is a conserved mechanism of m R NA tu rnover. Although there is to date no direct data for such a pathway in more complex eukaryotes, mRNAs lacking the cap structure are rapidly degraded in many eukaryotic cells (e.g., Drummond et al., 1985). In addition, enzymes that could catalyze the removal of the cap structure and subsequent 5' to 3' degradation of the transcript have been described in mammalian cells (e.g., Coutts and Brawerman, 1993). Finally, the oat phytochrome A mRNA undergoes a cleavage near the 5' terminus of the mRNA and is degraded, in part, in a 5' to 3' direction (Higgs and Colbert, 1994). Taken together, these data raise the possibility that deadenylation-dependent decapping followed by 5' to 3' exonucleolytic decay may be a conserved eukaryotic mRNA decay mechanism. Interactions between the 5' and 3' Termini of mRNAs The observation that deadenylation can lead to removal
of the 5'cap structure (Muhlrad et al., 1994, 1995) suggests an interaction between the 5' and 3' ends of an mRNA. Interactions between the transcript termini have been proposed to explain the effects that the poly(A) tail and 3' untranslated region (3'UTR) sequences can have on translation initiation (Munroe and Jacobson, 1990; for review see Jackson and Standart, 1990). A unifying hypothesis is that the same 5' to 3' interaction may mediate the rates of translation initiation and decapping. Although there is no direct biochemical evidence for such a 5' to 3' interaction, the abundance of circular polysomes seen in some electron micrographs suggests that the transcript termini can be in close proximity (e.g., Christensen et al., 1987). The nature of the 5' to 3' interaction may allow the poly(A) tail to inhibit decapping indirectly or directly. For example, because mRNAs may interact with the cytoskeleton through their poly(A) tails (e.g., Taneja et al., 1992), the putative 5' to 3' interaction may cause deadenyiation to alter the subcellular localization of an mRNA, thereby exposing it to a localized decapping activity. Alternatively, the poly(A) tail might inhibit decapping directly by forming or stabilizing an mRNP structure involving the 5' and 3' termini of the mRNA.
Deadenylation Can Also Lead to 3' to 5' Decay of mRNA Eukaryotic mRNAs can also be degraded in a 3' to 5' direction following deadenylation. For example, fragments of the yeast PGK1 mRNA shortened at the 3' end accumulate when the 5' to 3' decay pathway is blocked (Muhlrad and Parker, 1994; Muhlrad et al., 1995). Decay intermediates that are consistent with 3' to 5' decay following deadenylation have also been observed for the oat phytochrome A mRNA in vivo (Higgs and Colbert, 1994). It is not known whether exonucleases, endonucleases, or both are involved in 3'to 5'decay or how prevalent this decay pathway may be. However, it should not be expected that deadenylation necessarily exposes transcripts to 3' to 5' degradation since several mRNAs can be quite stable with essentially no poly(A) tail (e.g., Decker and Parker, 1993; Chert et al., 1994). An important point is that an individual transcript can simultaneously be a substrate for more than one mechanism of decay. For example, both the oat phytochrome A and yeast PGK1 transcripts undergo deadenylation-dependent 3' to 5' decay in addition to a 5' to 3' decay mechanism (Higgs and Colbert, 1994; Muhlrad et al., 1995). The presence of multiple overlapping pathways has several consequences. First, the observed half-life for an mRNA will be a summation of the rates of decay through each individual pathway. In addition, the mechanism by which an mRNA is degraded may change under different conditions, even without a significant alteration of decay rate (e.g., Muhlrad et al., 1995). The existence of multiple decay pathways also suggests that mutations that inactivate one pathway may not always have significant effects on mRNA stability, yet combinations of mutations inactivating different pathways may exhibit synergistic effects on stability and possibly viability. A system of redundant overlapping mechanisms of mRNA turnover is highly analogous to that
Review:rnRNADegradation 181
observed in Escherichia coil (for review see Belasco and Brawerman, 1993).
In deadenylation-dependent mRNA decay, differences in mRNA stability result from differences in the rates of poly(A) shortening and steps required for the subsequent decay of the deadenylated transcript. Thus, features of mRNAs that influence the rates of poly(A) shortening, decapping, and possibly 3' to 5' decay will affect mRNA stability. The rate of 5' to 3' digestion of the transcript body appears to be relativelyfast and therefore unlikely to contribute to differences in decay rates. This is based on the observation that intermediates of 5' to 3' decay are not observed unless the XRN1 nuclease has been deleted or a secondary structure blocks the nuclease (Decker and Parker, 1993; Hsu and Stevens, 1993; Muhlrad et al., 1994). Several sequence elements within mRNAs that promote rapid poly(A) shortening have been identified. These include sequences within the c-fos coding region and 3'UTR, which contains a prototypical AU-rich element (ARE) (Wilson and Triesman, 1988; Shyu et al., 1989; Wellington et al., 1993; Chen et al., 1994; Schiavi et al., 1994), and in the yeast MFA2 3'UTR (Muhlrad and Parker, 1992; Decker and Parker, 1993; Muhlrad et al., 1994). Interestingly, mutations in the c-los ARE and the MFA2 3'UTR can also slow degradation following deadenylation (Shyu et al., 1991; Muhlrad and Parker, 1992). Thus, at least for the MFA2 mRNA, 3'UTR sequences can also specify the rate of decapping. There is little information about the nucleases involved in deadenylation-dependent decay and how mRNA sequences modulate their rate of activity. The product of the XRN1 gene appears to be the primary nuclease responsible for 5'to 3' degradation of yeast mRNAs after decapping (Hsu and Stevens, 1993; Muhlrad et al., 1994). Decapping activities (Nuss et al., 1975; Kumagai et al., 1992; Coutts and Brawerman, 1993) and poly(A) nucleases (,~str0m et al., 1991, 1992; Sachs and Deardorff, 1992; Lowell et al., 1992) have been biochemically defined in cell extracts of yeast and mammals, although the in vivo roles of these particular activities remain to be established. Several proteins have been identified that interact with the ARE (Malter, 1989; Bohjanen et al., 1991; Vakalopoulou, 1991; Chen et al., 1992; Zhang et al., 1993; Katz et al., 1994) and therefore might modulate deadenylation rate. However, only one of these proteins, called the ARE/poly(U)binding factor, has been shown to stimulate decay in an in vitro system (Brewer, 1991). Progress in this area may be aided by the identification of proteins that bind to a small sequence, UUAUUUA(U/A)(U/A), determined to be important for a functional ARE to stimulate mRNA decay (Lagnado et al., 1994; Zubiaga et al., 1995). ~
(Muhlrad and Parker, 1994; Hagan et al., 1995). The degradation of mRNAs with nonsense codons is part of a process, termed mRNA surveillance, that ensures the rapid degradation of aberrant transcripts. These aberrant mRNAs contain early nonsense codons (Losson and Lacroute, 1979; Leeds et al., 1991; Peltz et al., 1993; Pulak and Anderson, 1993), unspliced introns (He et al., 1993), or extended 3'UTRs (Pulak and Anderson, 1993). mRNA surveillance occurs in many organisms and thus appears to be a conserved process. Mutations in related genes that inactivate this surveillance pathway have been identified in yeast and nematodes, named upland smg mutants, respectively (Hodgkin et al., 1989; Leeds et al., 1991, 1992). mRNA surveillance may exist, in part, to increase the fidelity of gene expression by degrading aberrant mRNAs that, if translated, would produce truncated proteins. This process would be relevant since truncated proteins can have dominant negative phenotypes. A striking example of this phenomenon is seen in Caenorhabditis elegans, where mutations in smg genes convert recessive nonsense mutations in the myosin gene uric-54 into dominant negatives (Pulak and Anderson, 1993). Since mRNA surveillance degrades unspliced and aberrantly processed transcripts, this decay mechanism might be most important in organisms with a large number of introns where processing errors might lead to truncated proteins. How a nonsense-containing transcript is recognized as aberrant and how mRNA decay is triggered are unclear. The simplest model is that premature translation termination sends a signal to expose the 5' end immediately to decapping by the same nucleases that degrade normal mRNAs. This hypothesis is supported by the observations that the sites of decapping are the same in deadenylationindependent and deadenylation-dependent decapping (cf. Muhlrad and Parker, 1994 with Muhlrad et al., 1995), and the XRN1 exonuclease performs 5' to 3' decay in both cases. The generation of the destabilizing signal requires both the failure to translate a significant part of the coding region and the presence of specific sequences downstream of the nonsense codon (Cheng et al., 1990; Peltz et al., 1993; Zhang et al., 1995). In a similar mechanism, transcripts with extended 3'UTRs may also be recognized and degraded because destabilizing sequences that are not normally in the mRNA are now included within the extended 3'UTR. A related and surprising observation is that early nonsense mRNAs can reduce the levels of fully spliced but nuclear-associated transcripts (Belgrader et al., 1994; Baserga et at., 1992). One possibility is that the first round of translation occurs while the transcripts are still nuclear associated and that premature termination during this round of translation can generate the signal to expose the 5' end to decapping.
Deadenylation-lndependent Decapping: mRNA Surveillance
Decay of EukaryoUc mRNA via Endonucleolytic Cleavage
mRNA decay can also be initiated by decapping and 5' to 3'decay of the transcript independent of poly(A) shortening. An example of this process is the degradation of the yeast PGK1 mRNA containing an early nonsense codon
Eukaryotic mRNAs can be degraded via endonucleolytic cleavage prior to deadenylation. Evidence for this mechanism comes from the analysis of transcripts such as mammalian 9E3, IGF2, transferrin receptor (TfR), and Xenopus
Determination of mRNA Decay Rates via Deadenylation-Dependent Decay
Cell 182
X/hbox2B mRNA where mRNA fragments are detected in
vivo that correspond to the 5' and 3' portion of the transcript and are consistent with internal cleavage within the 3'UTR (Stoekle and Hanafusa, 1989; Nielson and Christiansen, 1992; Binder et al., 1994; Brown et al., 1993). It appears that deadenylation to an oligo(A) tail length is not required for endonucleolytic cleavage since the 3' fragment of 9E3 and TfR mRNAs is polyadenylated(Stoekle and Hanafusa, 1989; Binder et al., 1994) and since cleavage of the Xlhbox2B mRNA is not affected by the adenylation state of the mRNA (Brown et al., 1993). However, it is possible that endonucleolytic cleavage of some mRNAs could be dependent on the length of the poly(A) tail. Endonucleolytic cleavages have also been defined in vitro for the albumin mRNA (Dompenciel et al., 1995) and in the coding region of the c-myc mRNA (Bernstein et al., 1992). Since there does not appear to be any similarity between the cleavage sites in these mRNAs, there may be a wide variety of endonucleases with different cleavage specificities. Since sequence-specific endonuclease target sites are likely to be limited to individual mRNAs or classes of mRNAs, their presence allows for specific control of the decay rate of these transcripts. In some cases, the rate of endonucleolytic cleavage is modulated by the activity of protective factors that bind at or near the cleavage site and compete with the endonuclease (Brown et al., 1993; Binder et al., 1994). For example, the binding of the iron response element-binding protein in the TfR 3'UTR in response to low intracellular iron concentrations inhibits the endonucleolytic cleavage of this mRNA (Binder et al., 1994). Therefore, some endonucleases may be constitutively active, and the accessibility of the cleavage site is regulated. In other cases, the endonuclease activity may be directly regulated. For example, the mammalian endonuclease RNase L is normally inactive and is only activated by oligomers of 2',5' phosphodiester-bonded adenylate residues, which are produced in response to the presence of doublestranded RNA (for discussion see Silverman, 1994). Although RNase L is important in mediating interferon responses (e.g., Hassel et al., 1993), it has yet to be established whether this enzyme normally degrades any cellular mRNAs.
Summary
Based on the above mechanisms of mRNA degradation, an integrated model of mRNA turnover can be proposed (Figure 1). In this model, all polyadenylated mRNAs would be degraded by the deadenylation-dependentpathway at some rate. In addition to this default pathway, another layer of complexity would come from degradation mechanisms specific to individual mRNAs or to classes of mRNAs. Such mRNA-specific mechanisms would include sequence-specific endonuclease cleavage and deadenylation-independent decapping. Thus, the overall decay rate of an individual transcript will be a function of its susceptibility to these turnover pathways. In addition, cisacting sequences that specify mRNA decay rate, as well
as regulatory inputs that control mRNA turnover, are likely to affect all the steps of these decay pathways. One important goal in future work will be to identify the gene products that are responsible for the nucleolytic events in these pathways and to delineate how specific mRNA features act to affect the function of these degradative activities. The identification of distinct mRNA decay pathways should allow genetic and biochemical approaches that can be designed to identify these gene products. A second important goal is to understand the nature of the interaction between the 5' and 3' termini, which may also be critical for efficient translation. References
/~str~m, J., ,~str~m,A., and Virtanen,A. (1991).In vitro deadenylation of mammalianmRNA by a HeLa cell 3' exonuclease.EMBOJ. 10, 3067-3071. /~str6m, J.,/~.strGm,A., and Virtanen,A. (1992). Propertiesof a HeLa cell 3' exonucleasespecificfor degradingpoly(A)tails of mammalian mRNA. J. Biol. Chem. 267, 18154-18159. Baserga,S. J., and Benz,E.J., Jr. (1992).I~-Globinnonsensemutation: deficientaccumulationof mRNAoccurs despitenormalcytoplasmic stability. Proc~Natl. Acad. Sci. USA 89, 2935--2939. Belasco,J., and Brawerman,G. (1993). Controlof MessengerRNA Stability (NewYork: AcademicPress). Belgrader,P., Cheng,J., Zhou,X., Stephenson,S., and Maquat,L. E. (1994). Mammaliannonsensecodonscan be cis effectorsof nuclear mRNA half-life.MoL Cell. Biol. 14, 8219-8228. Bernstein, P., and Ross, J. (1989). Poly(A), poly(A) binding protein and the regulationof mRNAstability.Trends Biochem.Sci. 14, 373377. Bernstein, P. L., Herrick, D. J., Prokipcak,R. D., and Ross,J. (1992). Controlof c-myc mRNAhalf-lifein vitro by a proteincapableof binding to a coding region stabilitydeterminant.GenesDev. 6, 642-654. Binder, R., Horowitz,J. A., Basilion,J. P., Koeller, D. M., Klausner, R. D., and Harford,J. B. (1994). Evidencethat the pathwayof transferrin receptormRNAdegradationinvolvesan endonucleolyticcleavage within the 3' UTR and does not involvepoly(A)tail shortening. EMBO J. 13, 1969-1980. Bohjanen, P. R., Bronislawa,P., June, C. H., Thompson,C. B., and Lindsten,T. (1991).An induciblecytoplasmicfactor(AU-B)bindsselectivelyto AUUUAmultimersin the 3' untranslatedregionof lymphokine mRNA. Mol. Cell. Biol. 11, 3288-3295. Brewer,G. (1991).An A+U-richelementRNA-bindingfactorregulates c-myc mRNAstabilityin vitro. Mol. Cell. Biol. 11, 2460-2466. Brown, B. D., Zipkin, I. D., and Harland, R. M. (1993). Sequencespecificendonucleolyticcleavageand protectionof mRNAin Xenopus and Drosophila. Genes Dev. 7, 1620-1631. Chen, C.-Y.A., You,Y., and Shyu,A.-B. (1992).Two cellularproteins bind specificallyto a purine-richsequencenecessaryfor the destabilizationfunctionof a c-fos proteincoding regiondeterminantof mRNA instability. Mol. Cell. Biol. 12, 5748-5757. Chen, C.-Y. A., Chen, T.-M., and Shyu,A.-B. (1994). Interplayof two functionallyand structurallydistinct domainsof the c-fos AU-richelement specifies its mRNA-destabilizingfunction. Mol. Cell. Biol. 14, 416-426. Cheng, J., FogeI-Petrovic,M., and Maquat, L~E. (1990). Translation to nearthe distal end of the penultimateexon is requiredfor normal levels of splicedtriosephosphateisomerasemRNA. Mol. Cell. Biol. 10, 5215-5225. Christensen, A. K., Kahn, L. E., and Bourne,C. M. (1987). Circular polysomespredominateon the roughendoplasmicreticulumof somatropes and mammotropesin the rat anteriorpituitary.Am. J. Anat. 178, 1-10. C0utts, M., and Brawerman,G. (1993). A 5' exoribonucleasefrom
Review: mRNA Degradation 183
cytoplasmic extracts of mouse sarcoma 180 ascites cells. Biochim. Biophys. Acta 1173, 57-62. Decker, C. J., and Parker, R. (1993). A turnover pathway for both stable and unstable mRNAs in yeast: evidence for a requirement for deadenylation. Genes Dev. 7, 1632-1643. Decker, C. J., and Parker, R. (1994). Mechanisms of mRNA degradation in eukaryotes. Trends Biochem. Sci. lg, 336-340. Dompenciel, R. E., Garnepudi, V. R., and Schoenberg, D. R. (1995). Purification and characterization of an estrogen-regulated Xenopus liver polysomal nuclease involved in the selective destabilization of albumin mRNA. J. Biol. Chem., 270, 6108-6118. Drummond, D. R., Armstrong, J., and Colman, A. (1985). The effect of capping and polyadenylation on the stability, movement and translation of synthetic messenger RNAs in Xenopus oocytes. Nucl. Acids Res. 13, 7375-7394. Hagan, K. W., Ruiz-Echervarda, M. J., Quan, Y., and Peltz, S. W. (1995). Characterization of cis-acting sequences and decay intermediates involved in nonsense-mediated mRNA turnover. Mol. Cell. Biol. 15, 809-823. Hassel, B. A., Zhou, A., Sotomayor, C., Maran, A., and Silverman, R. H. (1993). A dominant negative mutant of 2-5A-dependent RNase suppresses antiproliferative and antiviral effects of interferon. EMBO J. 12, 3297-3304. He, F., Peltz, S. W., Donahue, J. L., Rosbash, M., and Jacobson, A. (1993). Stabilization and ribosome association of unspliced premRNAs in a yeast upfl- mutant. Proc. Natl. Acad. Sci. USA 90, 70347038. Higgs, D. C., and Colbert, J. T. (1994). Oat phytochrome A mRNA degradation appears to occur via two distinct pathways. Plant Cell 6, 1007-1019. Hodgkin, J., Papp, A., Pulak, R., Am bros, V., and Anderson, P. (1989). A new kind of informational suppression in the nematode Caenorhabditis elegans. Genetics 123, 301-313. Hsu, C. L., and Stevens, A. (1993). Yeast cells lacking 5' to 3' exoribonuclease 1 contain mRNA species that are poly(A) deficient and partially lack the 5' cap structure. Mol. Cell. Biol. 13, 4826-4835. Jackson, R. J., and Standart, N. (1990). Do the poly(A) tail and 3' untranslated region control mRNA translation? Cell 62, 15-24. Katz, D. A., Theodorakis, N. G., Cleveland, D. W., Lindsten, T., and Thompson, C. B. (1994). AU-A, an RNA-binding activity distinct from hnRNP A1, is selective for AUUUA repeats and shuttles between the nucleus and the cytoplasm. Nucl. Acids Res. 22, 238-246. Kumagai, H., Kon, R., Hoshino, T., Aramaki, T., Nishikawa, M., Hirose, S., and Igarashi, K. (1992). Purification and properties of a decapping enzyme from rat liver cytosol. Biochim. Biophys. Acta 13, 45-51. Lagnado, C. A., Brown, C. L., and Goodall, G. J. (1994). AUUUA is not sufficient to promote poly(A) shortening and degradation of an mRNA: the functional sequence within AU-rich elements may be UUAUUUA(U/A)(U/A). Mol. Cell. Biol. 14, 7984-7995. Larimer, F. W., and Stevens, A. (1990). Disruption of the gene XRN1, coding for a 5'--3' exoribonuclease, restricts yeast cell growth. Gene 95, 85-90. Leeds, P., Peltz, S. W., Jacobson, A., and Culbertson, M. R. (1991). The product of the yeast UPF1 gene is required for rapid turnover of mRNAs containing a premature translational termination codon. Genes Dev. 5, 2303-2314. Leeds, P., Wood, J. M., Lee, B.-S., and Culbertson, M. R. (1992). Gene products that promote mRNA turnover in Saccharomyces cerevisiae. Mol. Cell. Biol. 12, 2165-2177. Losson, R., and Lacroute, F. (1979). Interference of nonsense mutations with eukaryotic messenger RNA stability. Proc. Natl. Acad. Sci. USA 76, 5134-5137. Lowell, J., Rudner, D., and Sachs, A. (1992). 3'-UTR-dependent deadenylation by the yeast poly(A) nuclease. Genes Dev. 6, 2088-2099~ Malter, J. S. (1989). Identification of an AUUUA-specific messenger RNA binding protein. Science 246, 664-666. Muhlrad, D., and Parker, R. (1992). Mutations affecting stability and deadenylation of the yeastMFA2transcript. Genes Dev. 6, 2100-2111.
Muhlrad, D., and Parker, R. (1994). Premature translational termination triggers mRNA decapping. Nature 340, 578-581. Muhlrad, D., Decker, C. J., and Parker, R. (1994). Deadenylation of the unstable mRNA encoded by the yeast MFA2 gene leads to decapping followed by 5' to 3' digestion of the transcript. Genes Dev. 8, 855866. Muhlrad, D., Decker, C. J., and Parker, R. (1995). Turnover mechanisms of the stable yeast PGK1 mRNA. Mol. Cell. Biol., 16, 2145-2156. Munroe, D., and Jacobson, A. (1990). mRNA poly(A) tail, a 3' enhancer of translational initiation. Mol. Cell. Biol. 10, 3441-3455. Nielson, F. C., and Christiansen, J. (1992). Endonucleolysis in the turnover of insulin-like growth factor II mRNA. J. Biol. Chem. 267, 19404-19411. Nuss, D. L., Furuichi, Y., Kock, G., and Shatkin, S. J. (1975). Detection in HeLa cell extracts of a 7-methyl guanosinb specific enzyme activity that cleaves mTGpppN m. Cell 6, 21-27. Peltz, S. W., Brewer, G., Bernstein, P., Hart, P., and Ross, J. (1991). Regulation of mRNA turnover in eukaryotic cells. CRC Crit. Rev. Euk. Gene Exp. 1, 99-126. Peltz, S. W., Brown, A. H., and Jacobson, A. (1993). mRNA destabilization triggered by premature translational termination depends on three mRNA sequence elements and at least one trans-acting factor. Genes Dev. 7, 1737-1754. Pulak, R., and Anderson, P. (1993). mRNA surveillance by the Caenorhabditis elegans smg genes. Genes Dev. 7, 1885-1897. Sachs, A. B., and Deardorff, J. (1992). Translation initiation requires the PAB-dependent poly(A) ribonuclease in yeast. Cell 70, 961-973. Schiavi, S. C., Wellington, C. L., Shyu, A.-B., Chen, C.-Y. A., Greenberg, M. E., and Belasco, J. G. (1994). Multiple elements in the c-fos protein coding region facilitate mRNA deadenylation and decay by a mechanism coupled to translation. J. Biol. Chem. 269, 34413448. Shyu, A.-B., Greenberg, M. E., and Belasco, J. G. (1989). The c-fos transcript is targeted for rapid decay by two distinct mRNA degradation pathways. Genes Dev. 3, 60-72. Shyu, A.-B., Belasco, J. G., and Greenberg, M. E. (1991). Two distinct destabilizing elements in the c-fos message trigger deadenylation as a first step in rapid mRNA decay. Genes Dev. 5, 221-234. Silverman, R. H. (1994). Fascination with 2-5A-dependent RNase: a unique enzyme that functLons in interferon action. J. Interferon Res. 14, 101-104. Stoekle, M. Y., and Hanafusa, H. (1989). Processing of 9E3 mRNA and regulation of its stability in normal and rous sarcoma virus-transformed cells. Mol. Cell. Biol. 9, 4738-4745. Taneja, K. L, Lifshitz, L. M., Fay, F. S., and Singer, R. H. (1992). Poly(A) RNA codistribution with microfilaments: evaluation by in situ hybridization and quantitative digital imaging microscopy. J. Cell Biol. 119, 1245-1260. Vakalopoulou, E., Schaack, J., and Shenk, T. (1991). A 32-kilodalton protein binds to AU-rich domains in the 3'untranslated region of rapidly degraded mRNAs. MoI. Cell. Biol. 11, 3355-3364. Wellington, C. L., Greenberg, M. E., and Belasco, J. G. (1993). The destabilizing elements in the coding region of c-fos mRNA are recognized as RNA. Mol. Cell. Biol. 13, 5034-5042. Wilson, T, and Triesman, R. (1988). Removal of poly(A) and consequent degradation of the c-fos mRNA facilitated by 3' AU-rich sequences. Nature 336, 396-399. Zhang, A., Ruiz-Echevarria, M. J., Quan, Y., and Peltz, S. W. (1995). Identification and characterization of a sequence motif involved in nonsense-mediated mRNA decay. Mol. Cell. Biol., 15, 2231-2244. Zhang, W., Wagner, B. J., Ehrenman, K., Schaefer, A. W., DeMaria, C. T., Crater, D., DeHaven, K., Long, L., and Brewer, G. (1993). Purification, characterization, and cDNA cloning of an AU-rich element RNA-binding protein, AUFI. Mol. Cell. Biol. 13, 7652-7665. Zubiaga, A. M., Belasco, J. G., and Greenberg, M. E. (1995). The nonamer UUAUUUAUU is the key AU-rich sequence motif that mediates mRNA degradation. Mol. Cell. Biol. 15, 2219-2230.