Post‐transcriptional control of gene expression: effectors of mRNA decay

It is well established that the control of gene expression in both prokaryotes and eukaryotes occurs not only at the level of gene transcription but also at a number of post-transcriptional steps, including pre-mRNA splicing and mRNA transport (in the case of eukaryotes), mRNA translation and mRNA decay. It is also well established that mRNA decay, the last event in the continuum of post-transcriptional steps, can be influenced by each of the preceding steps. All of these topics, and especially mRNA decay, were discussed at the most recent FASEB meeting on 'Post-transcriptional control of gene expression: effectors of mRNA decay', which was held in Tucson, AZ, USA (July 6–11th, 2002), and brought together both eukaryotic and prokaryotic molecular biologists who, either directly or indirectly, study mRNA decay. Analysing mRNA degradation in prokaryotes has been particularly difficult because short cell division times necessitate short mRNA half-lives, which are not easy to measure. In fact, the inherent instability of prokaryotic mRNAs has been a major obstacle to the profitable industrial production of proteins in microorganisms. The continuous degradation and synthesis of prokaryotic mRNAs permit not only the metabolic changes that are required as cells grow and divide, but also rapid adaptation to new environments and conditions. Most ribonucleases involved in mRNA decay are also implicated in the processing of rRNAs or tRNAs. In bacteria, mRNAs can be degraded by a number of mechanisms that act independently but in parallel and that target different sites with different efficiencies. The accessibility of sites to degradation depends on several factors, including RNA higher-order structure, protection by translating ribosomes, and polyadenylation status (Régnier and Arraiano, 2000). There were several talks on RNase E, a single-strand-specific endoribonuclease that is critical for mRNA decay in E. coli(Fig. 1). RNase E is a large protein with an N-terminal catalytic domain and a C-terminal domain containing RNA binding sites as well as a protein 'scaffold' that interacts with polynucleotide phosphorylase (PNPase), enolase and RNA helicase B (RhlB) (Cohen and McDowall, 1997; Vanzo et al., 1998) to form a multienzyme complex called the RNA degradosome (Carpousis et al., 1994). George Mackie (University of British Columbia, Vancouver, Canada) showed that a degradosome-like complex can form spontaneously in vitro in the absence of all other cellular components (Coburn et al., 1999) and reported that the C-terminal domain of RNase E interacts with both the N-terminal and C-terminal domains of PNPase. Data presented revealed that RNase E is minimally a dimer and that RNase G, a homologue of RNase E, can efficiently form dimers, trimers and tetramers. RNase G is an endonuclease that has 36% amino acid sequence identity with the N-terminal catalytic domain of RNase E. RNase G was previously designated Caf A. RNase E and RNase G have broadly similar sequence specificity and they can co-exist. The control of rng expression remains poorly understood. In the absence of RNase E, RNase G does not restore the normal processing of many Rne-cleaved mRNAs but RNase G expression can confer viability on an rne deletion mutant strain (Lee et al., 2002). RNase E and mRNA decay in E. coli. The decay of a hypothetical polycistronic transcript encoding genes A, B and C is shown. RNA is simultaneously synthesized by DNA-dependent RNA polymerase (RNAP) and translated into protein by ribosomes. Boxes and lines represent, respectively, translated and untranslated regions, where the intergenic regions harbour elements controlling translation and mRNA decay. This polycistronic transcript is first attacked by RNase E. The resulting translationally arrested and fragmented sequences are designated as small (broken) boxes, where the order of decay is arbitrarily chosen as B followed by A and then C. Fragments are then targeted for polyadenylation at the 3′ end and degraded to NDPs and NMPs by PNPase and RNase II. (Adapted from Carpousis and Dreyfus, 2003.) Joel Belasco (New York University School of Medicine, New York, NY, USA) proposed a three-dimensional structure for the amino terminal domain of E. coli RNase E based on the results of site-directed mutagenesis. All surface residues that are essential for cell growth and feedback regulation of RNase E synthesis mapped to the end of this domain, whose sequence is very similar to that of the S1 family of RNA-binding domains. The arginine-rich RNA binding domain within the carboxy-terminal half of RNase E is not required for ribonuclease activity (Diwa et al., 2002). The mechanism of RNase E action was discussed by several speakers. Stanley Cohen (Stanford University, CA, USA) presented results indicating that the catalytic domain of RNase E binds selectively to 5′ monophosphate RNA termini but that cleavage site selection shows inherent 3′-to-5′ directionality (Feng et al., 2002). In contrast, RNase G seems to have a non-directional and distributive mode of action. Cohen also reported the results of a DNA microarray study of RNA decay (Bernstein et al., 2002), demonstrating that 80% of all mRNAs in E. coli have half-lives of between 3 and 8 min. Marc Dreyfus (CNRS/École Normale Supérieure, Paris, France) discussed two possibilities for RNase E binding to mRNA: (i) binding to the 5′ end followed by migration to internal cleavage sites in a so-called 5′ tethering mechanism (Mackie, 1998), or (ii) binding in a direct entry pathway. Data were presented showing that the direct entry pathway is favoured in the absence of translation, making it likely that this pathway is responsible for the lability of many untranslated mRNAs. A.J. Carpousis (CNRS, Toulouse, France) showed that elements within the non-catalytic region of RNase E that are involved in formation of the RNA degradosome are important for the decay of ribosome-free lacZ RNA (Leroy et al., 2002). Gadi Schuster (Technion-Israel Institute of Technology, Haifa, Israel) compared RNase E encoded by genes from cyanobacteria, chloroplasts of green alga, the Arabidopsis nucleus and E. coli and found that there are three conserved sequences (boxes I, II and III) in addition to the S1 RNA binding domain. Database analysis revealed that chloroplast and cyanobacterial RNase E-like proteins lack the C-terminal domain, suggesting the absence of a degradosome-like complex involving PNPase (Baginsky et al., 2001). Bob Simons (University of California, Los Angeles, CA, USA) described how RNA metabolism appears to undergo major changes when bacterial cells are subjected to growth at low temperatures. These effects include the documented induction of PNPase activity (Zangrossi et al., 2000; Beran and Simons, 2001; Mathy et al., 2001), and the possible assembly of a cold-shock RNA degradosome. These adaptations appear, in part, to reverse a cold-temperature-dependent accumulation of polyadenylated 3′ tails, which may otherwise be lethal to cells under those conditions. PNPase and RNase II are considered to be two main E. coli exoribonucleases that degrade mRNAs processively to mononucleotides in the 3′-to-5′ direction (reviewed in Régnier and Arraiano, 2000; Fig. 1). RNase II together with RNase R are prototypes of the RNR family of ribonucleases (Zuo and Deutscher, 2001). Double mutants deficient in either PNPase and RNase II or PNPase and RNase R are unviable (Donovan and Kushner, 1986; Chen et al., 1998). Cecilia Arraiano (ITQB – Universidade Nova de Lisboa, Oeiras, Portugal) reported that RNase R is a cold shock protein that is tightly regulated by temperature. Studies of an rnr mutant characterized by a cold-shock phenotype revealed the importance of RNase R in the adaptation of cells to cold shock conditions. RNase R levels did not change in minimal media, unlike those of RNase II, another member of the RNR family (Cairrão et al., 2001). RNase R was also shown to be involved in the maturation of SsrA/tmRNA, a non-coding small stable RNA involved in protein tagging and ribosome recycling in a process called trans-translation. It will be important to define the role of RNase R in the processing of other small, non-coding RNAs because they are far more abundant and important than initially imagined (Storz, 2002). Polyadenylation takes place not only in the nucleus and, in certain cases, the cytoplasm of eukaryotic cells but also in eubacteria, archaea and organelles (Dreyfus and Régnier, 2002). Bacterial mRNAs harbour relatively short poly(A) tails that in general promote RNA degradation. Polyadenylation can occur on mature mRNAs as well as on RNA fragments resulting from endo- or exonucleolytic degradation and on stable RNAs or their precursors. Murray Deutscher (University of Miami, FL, USA) described how polyadenylation plays a role in the degradation of defective tRNAs as an example of RNA quality control in E. coli. When precursors of tRNAs or rRNAs can not be converted to their mature forms because of, e.g. abnormally extended 3′ ends, then polyadenylation by poly(A) polymerase targets them for degradation by exonucleases that include PNPase and RNase R (Li et al., 2002). In contrast, the 3′ terminus of properly matured tRNA or rRNA is relatively resistant to the actions of either poly(A) polymerase or exonucleases. PNPase and RNase II are the primary enzymes involved in the removal of the poly(A) tails from mRNAs. Whereas RNase II is unable to degrade most 3′ terminal secondary structures, it very efficiently removes the oligo(A) tails that can be used as binding sites by PNPase and, by so doing, can protect mRNAs that end in a stable hairpin from degradation by PNPase (Marujo et al., 2000). Philippe Régnier (IBPC and Université Paris 7, Paris France) demonstrated that inhibition of the RNase E-dependent decay of E. coli rpsO mRNA by either RNase E deficiency or removal of RNase E cleavage sites results in preferential rpsO mRNA decay by a poly(A)-dependent pathway. Sidney Kushner (University of Georgia, Athens, GA, USA) reported that an increase in polyadenylation leads to stabilization of the pnp and rne transcripts, which in turn leads to increased levels of PNPase and RNase E as a consequence of alterations in autoregulation of the respective genes (Mohanty and Kushner, 2002). It appears that polyadenylation serves as a sensing mechanism by which cells adjust the levels of both RNase E and PNPase. Polyadenylation in E. coli is performed mainly by poly(A) polymerase but occasionally heteropolymeric poly(A) tails can be added by PNPase (Mohanty and Kushner, 2000). The mechanism of RNA degradation in chloroplasts consists of sequential events including endonucleolytic cleavage, the addition of poly(A)-rich sequences to the 3′-ends of the cleavage products and, exonucleolytic degradation primarily by PNPase. Gadi Schuster reported that both polyadenylation and exonucleolytic degradation in spinach chloroplasts are performed by PNPase and that there is no poly(A) polymerase homologue (Yehudai-Resheff et al., 2001). It is well known that there are mechanistic connections between mRNA translation and mRNA decay. However, the mechanism of translation termination has long been a puzzle. Yoshi Nakamura (Institute of Medical Science, University of Tokyo, Japan) described recent crystallographic evidence suggesting that the eukaryotic release factor (eRF) 1, the bacterial release factor (RF) 2, and the ribosome recycling factor (RRF) all mimic a tRNA shape, whereas biochemical and genetic evidence supports the idea of a tripeptide 'anticodon' in bacterial RF1 and RF2 (Nakamura and Ito, 2002). Very recent structures determined by cryo-electron microscopy show that RF2 has a conformation in the ribosome that is distinct from the crystal structure of RF2 in isolation. Furthermore, results from hydroxyl radical probings of RRF on the ribosome are not in agreement with the simple idea of RRF mimicking a tRNA in the ribosomal A site. These new data raise serious issues with the simple concept that there is mimicry in shape between proteins and RNA and therefore justify additional studies. In a keynote address, Roy Parker (University of Arizona, Tucson, AZ, USA) reviewed evidence that cytoplasmic mRNA in S. cerevisiae is normally degraded from both 5′ and 3′ ends after an essential step of deadenylation, which constitutes the rate-limiting step in decay (Fig. 2). Deadenylation is catalysed by one of two different enzyme complexes (reviewed in Decker and Parker, 2002). The major deadenylase is an ∼1 mDa complex comprised of Ccr4p, a member of the ExoIII/AP endonuclease family that serves as the catalytic subunit, as well as Pop2, Not1, Not2, Not3, Not4 and Not5. The minor deadenylation complex consists of Pan2p/Pan3p. Once the poly(A) tail is sufficiently shortened to preclude Pab1p binding, mRNA decay proceeds by either (i) Dcp1p/Dcp2p-mediated 5′ decapping and, subsequently, 5′-to-3′ exonucleolytic degradation by Xrn1p or (ii) 3′-to-5′ exonucleolytic degradation by the ∼300–400 kDa exosome complex, the core of which is composed of enzymes related to the family of RNase PH/polynucleotide phosphorylase proteins (reviewed in Mitchell and Tollervey, 2000; Butler, 2002). In related work, John McCarthy (UMIST, Manchester, UK) reported that the accessibility of the cap to Dcp1p in S. cerevisiae is not simply controlled by competition for cap binding with eukaryotic initiation factor (eIF) 4E (Ramirez et al., 2002), the latter of which can stabilize yeast mRNAs in vivo (Vilela et al., 2000). McCarthy also reported the characterization of S. pombe eIF4E2, a second eIF4E protein that is distinct from eIF4E1 (Ptushkina et al., 2001), as well as a new type of S. pombe decapping protein that consists of a histidine-triad (HIT) motif protein (Nhm1) that could potentially participate in a nuclear degradation pathway (Salehi et al., 2002). hnRNA and mRNA degradation in S. cerevisiae. Nuclear precursors to mRNA that fail to undergo efficient processing can be targeted for degradation in the 3′-to-5′ direction by the nuclear exosome. These transcripts can also be degraded to a lesser extent in the 5′-to-3′ direction by Rat1p after decapping by the Dcp1/Dcp2 complex. Both pathways can be viewed as a form of quality control. In the cytoplasm, the Ccr4p, a member of the exonuclease III family together with Pop2p, an RNase D family member, and Not1–5 are critical components of the major deadenylase that functions to degrade mRNA 3′ ends. A minor deadenylation complex consists of Pan2p/Pan3p. Deadenylase function normally precedes Dcp1p/Dcp2p decapping and Xrn1p 5′-to-3′ exonuclease activities, which work in sequence to degrade the transcript body. To a lesser extent, the transcript body can also be degraded from the 3′ end by the cytoplasm exosome. N, nucleus; C, cytoplasm; Cbc1p/Cbc2p, cap binding complex; Dcp1p/2p, decapping complex; Lsm, SM-like; Pab1, Poly (A) binding protein 1;>, greater than. Parker and other speakers noted that some aspects of decay pathways in yeast are conserved in higher eukaryotes while others are not. As one example, Ann-Bin Shyu (University of Texas Medical School, Houston, TX, USA) described how deadenylation is the first step in the decay of the mammalian mRNA for the proto-oncoprotein c-fos (Grosset et al., 2000). Translation of c-fos mRNA is proposed to break or rearrange the physical connection between proteins bound to the protein-coding-region determinant of instability (mCRD) and poly(A) binding protein PABP1, making the poly(A) tail susceptible to shortening. Results of co-immunoprecipitation experiments revealed that PABP1 interacts with mCRD-binding proteins UNR, NSAP1 and hnRNP D (also called AUF1). Furthermore, ectopic expression of UNR or NSAP1 blocked deadenylation, suggesting that an interaction between the poly(A) tail and the mCRD is required to trigger deadenylation. Shyu also reported that deadenylation is the first step in the decay of beta-globin mRNA that prematurely terminates translation in non-erythroid cells. This contrasts with the deadenylation-independent decapping followed by 5′-to-3′ exonucleolytic decay that typifies the pathway of nonsense-mediated mRNA decay (NMD) in S. cerevisiae (Muhlrad and Parker, 1994). It also contrasts with reports that nonsense-containing beta-globin mRNA in erythroid cells is first degraded at the 5′ end (Stevens et al., 2002 and references therein). Mike University, USA) reported the of the gene based primarily on of the encoded protein to the yeast motif domain. demonstrated that protein has decapping activity et al., et al., 2002). This is an important S. cerevisiae Dcp1p was to be the decapping particularly (i) decapping activity with Dcp1p in (ii) Dcp1p with decapping and Dcp1p from which been by has decapping activity (reviewed in Decker and Parker, 2002). to Dcp1p have been and data that can have decapping activity and Parker, 2002). and Medical School, also reported another important whereas mRNA decay is conserved between yeast and this relatively minor decay pathway in yeast appears to be the major decay pathway both in vitro and in vivo for mammalian mRNAs containing an et al., 2001; and 2001; et al., Fig. of mRNAs in mammalian cells. The decay of mRNAs is primarily from the mRNA 3′ end the exosome and, to a lesser extent, decapping and 5′-to-3′ decay of the mRNA body et al., 2001; and 2001; et al., 2002). It is not known mRNAs can be simultaneously degraded from each PABP1, Poly (A) binding Poly (A) were shown to control mRNA as well as translation in S. cerevisiae by The 3′ untranslated region of the translation initiation factor A mRNA an the of either mammalian or mammalian c-fos is in S. cerevisiae to confer and decapping to transcripts and 2001). Whereas a yeast homologue of the family of was found to the of mRNA and mRNA data that it also binds the of mRNA but than A of proteins involved in mRNA the the and which also with Pab1p and confer The also mRNA translation by an All at some pathways that control mRNA to be conserved between yeast and mammalian as by the pathway when leads to of mRNAs and 2001). The control of gene expression – at the cellular or level – is by mRNA mRNA translation or (University of UK) reported that binding proteins the translation of which a member of the family involved in induction and formation of et al., 2001). is on an mRNA 5′ experiments several binding proteins as of the and of RNA binding (University of USA) showed that of C. mRNA by the binding of to two 3′ elements is of a 5′ cap but on a poly(A) tail in a that does not (A) tail (reviewed in and 2002). which is essential for binds in vitro as well as in a also mRNA as by the level of mRNA that typifies C. which a within the domain. to this is not the of mRNA Furthermore, instability is in a suggesting that the mechanism of instability is related to (University of USA) reported on the mechanism by which the of which protein required for the of DNA into is with the of DNA (reviewed in and 2002). When DNA synthesis is mRNA degradation depends on mRNA The that is after the mRNA termination and interaction with the binding protein are also critical for mRNA decay in the absence of DNA A was proposed in which mRNA degradation removal of the followed by decay. University, USA) described how the induction of cell is with a rapid in mRNA with this mRNAs are in ribosomal complexes in which they have short poly(A) tails on and lack that the complex is involved in terminal of mRNA decay. the continuum of post-transcriptional as a of the quality of gene (University of discussed a number of related but distinct RNA decay pathways that degrade nuclear in S. cerevisiae as a of transcripts (Fig. 2). As one example, data were presented demonstrating (i) and which are components of the pre-mRNA factor are characterized by abnormally that are by the absence of either a core of the exosome complex, or the RNA helicase whereas (ii) absence of the exosome in defective in abnormally short polyadenylated transcripts that are for with other these data that the between the which can be viewed as a type of RNA and processing these to which can be viewed as a type of exosome to the cleavage and polyadenylation is regulated in a where degradation is favoured with growth on and processing to mRNAs is favoured with growth on other et al., 2002). that the decay of cytoplasmic mRNA in S. cerevisiae is also to the of appears to exosome activity in both the and cytoplasm of S. cerevisiae. As another example, data that the absence of results in at an abnormally so that a mechanism targets the transcripts for rapid decay first by and then by the core exosome. As a example, in so as to be defective in nuclear mRNA mRNA is in the absence of either or and pre-mRNA is in the absence of the complex than the complex, which functions in the nuclear Dcp1p and the complex to degrade There were several talks on another quality control mechanism that typifies all eukaryotes that have been but seems to have no in (University of Medical to the of at termination et al., 2002), reported that in S. cerevisiae termination an that is on translation but is either by of one of the or by tethering Pab1p within the 3′ of the termination These results were to (i) between a normal and a termination event are at the site of and (ii) an abnormally long 3′ is to trigger in S. cerevisiae. (University of Medical NY, USA) described how in mammalian cells targets mRNA bound by the complex of cap binding proteins that is primarily but not a of translation et al., 2001; Fig. mRNA bound by the major cytoplasmic cap binding eukaryotic initiation factor (eIF) mRNA is also bound by the complex of proteins as a consequence of splicing as well as the and et al., 2002). As mRNA was found to from these and other data were to that is to synthesized mRNA of the in nonsense-containing mRNA is in the nuclear or cytoplasmic of cells. for cytoplasmic in mammalian cells. takes is called the of translation et al., 2001; et al., 2002). The translation initiation complex consists of mRNA bound by at the 5′ at the 3′ poly(A) and an complex that is as a consequence of splicing at each The or one of two nuclear proteins involved in then another protein involved in that appears on the cytoplasmic of the nuclear unlike and is not as an protein but interacts with when translation terminates more than of an is cytoplasmic but also et al., 2002). then takes place at both mRNA ends so as to (i) decapping followed by 5′-to-3′ decay of the mRNA body from the 5′ end and (ii) deadenylation followed by 3′-to-5′ decay of the mRNA body from the 3′ end and It is not known mRNAs can be simultaneously degraded from each translation terminates than of an or of a it for most then translating are to the together with the proteins so that the mRNA is no susceptible to termination (University of Texas Houston, TX, USA) described mechanisms by which nonsense-containing mRNA is very in as a to transcripts as a consequence of DNA and normal In one the of is by a sequence within the and that can the of the and 2002). This sequence also appears to a role in the second which the of transcripts within et al., that at certain within certain mRNAs can nuclear In with this from the of University School of Medicine, USA) in of for in and a distinct pathway called alterations in splicing Whereas in some has been to a sequence change within a of splicing an splicing than an on in other has been to an of the to nuclear site (reviewed in 2002). reported the of RNA to the cellular level of either or of both and whereas of only et al., 2002). Furthermore, RNA demonstrated for in and et al., 2002). Data demonstrating that mRNA can be translated and that can and raise the of translation within the it could pathways of nuclear RNA decay. University, UK) the of nuclear structure and including the of transcription and sites of nuclear translation (reviewed in and 2002). A on the of nuclear translation that was by brought some et al., but example, Murray Deutscher reported that cell by the of et al., the of amino acid as reported by However, that the extent of cytoplasmic to levels also the of amino acid to levels et al., This was used to that is cytoplasmic than within (University of USA) presented data demonstrating that synthesized ribosomal not in et al., Furthermore, which is of S. cerevisiae in with the and is then from the by cytoplasmic et al., 2001), the of and as does mammalian and These data the of nuclear translation in as well as in yeast and mammalian cells. the translation occurs within the within the cytoplasm either or after mRNA the nuclear complex but release into the (University of Tokyo, Tokyo, Japan) described a new that the cleavage activity of the with the activity of the RNA helicase by the to a poly(A) which binds with PABP1 and and the helicase secondary and structures in their cleavage, mRNA sequence that a can be and can be C, A or can be targeted for efficient reported the rapid of genes by in cells and for that were either resistant to or a in and A number of and new genes was this than were a that This be very in genes when confer a change in phenotype within a short results that mRNA decay shows between prokaryotes and poly(A) tails of mRNAs major in the control of mRNA decay in both of with parallel is by similar for the complexes of 3′ to 5′ which in the case of eukaryotes form the exosome and in the case of can form the degradosome. As a example, transcripts undergo as a of quality control of the from which they will the importance of mRNA decay as a critical and used of gene expression in both prokaryotes and The in the for and A.J. Carpousis for on Fig. is by from is by from the

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.