mRNA Decay
Chapter
56
SIDNEY R. KUSHNER
Over the past 30 years most RNA research in Escherichia coli has focused on either the process of making mRNAs (transcription) or converting the encoded information into proteins (translation). In contrast, only limited attention has been given to understanding the processes by which mRNAs are functionally and chemically degraded. With half-lives varying from 40 s to 20 min, it seems apparent that the stability of each individual mRNA plays some role in the overall expression of its coding sequence. Thus, if a particular mRNA has a very short half-life, it can only be translated a limited number of times before it will be functionally inactivated. This result allows for another level of gene regulation beyond transcription initiation or translation initiation.
It should be noted that there can be both cis- and trans-acting factors that can affect mRNA decay. cis-acting factors can include special types of secondary or tertiary structures whose presence or absence can either retard or increase decay rates. Similarly, while most trans-acting factors will likely be nucleases that lead to phosphodiester bond cleavage, there will also be proteins whose interaction with a particular feature of an RNA molecule will lead to slower decay.
This chapter will consider a number of topics relating to mRNA decay. These topics include physical properties of mRNA molecules, determination of mRNA half-lives, genetics and biochemistry of mRNA decay proteins, regulation of mRNA decay, posttranscriptional modification of mRNAs, and biochemical models for mRNA decay. It will also evaluate, on the basis of existing data and some speculation, the strategies that E. coli has devised to efficiently degrade these short-lived species. This analysis will demonstrate that bacteria have multiple pathways for ensuring the effective decay of mRNAs and that mRNA decay is required for cell viability. Finally, important unanswered questions in mRNA decay will be highlighted. The recent monograph edited by Belasco and Brawerman (16) provides a more detailed review of mRNA decay in both prokaryotes and eukaryotes.
A unique feature in prokaryotes is the existence of both monocistronic and polycistronic transcripts. Thus, the lengths of mRNAs can vary from as few as 325 nucleotides (nt) for the lpp transcript to many kilobases for operons such as gal or lac. Furthermore, even though transcription termination is well documented in E. coli (see reference 134 for a recent review), there are, apparently, many instances of significant read-through transcription such that very large transcripts can be generated. One example apparently occurs in the recC-ptr-recB-recD region of the E. coli chromosome, where transcripts of over 11 kb arise from the recC promoter (S. R. Kushner, unpublished results).
Yet transcript length is not a good predictor of mRNA stability. For example, the lpp mRNA of only 325 nt has a measured half-life on the order of 15 to 17 min (74), the trxA transcript of 493 nt has a half-life on the order of 2 to 3 min (8), and the ompA mRNA (1,215 nt) has a half-life of 15 min (155). Early studies of mRNA stability showed that various portions of a polycistronic mRNA had different half-lives (18). These examples demonstrate that features of mRNA other than its actual length are the rate-limiting steps in their decay.
Since, as will be discussed later, there do not appear to be 5'-to-3' exonucleases in E. coli (41, 42, 43), nor is there any evidence for 5' capping, one would not necessarily assume that the 5' end of an mRNA would be important for stability. In addition, unlike many eukaryotic mRNAs, prokaryotic transcripts generally do not have extensive 5' untranslated regions (5' UTRs). However, the analysis of the ompA gene clearly indicates that a short UTR can dramatically affect mRNA stability. In this case the ompA gene has a half-life of approximately 15 to 17 min if the 133-nt 5' UTR is left intact (52, 155) and 3.6 to 4.4 min if it is removed (52). Moreover, if the 5' UTR is transferred to the 5' end of a short-lived message such as bla, the half-life of the chimeric mRNA increases from 3 to 4 min to more than 13 min (17). Subsequently, Emory et al. (52) showed that the 5' UTR of ompA exists in a highly ordered structure consisting of three stem-loops. Significantly, there is no single strandedness at the 5' end of the molecule. As few as four unpaired nucleotides at the 5' end of the transcript are sufficient to overcome the stabilization afforded by the 5' UTR (52).
Another example of a 5' UTR affecting stability, but in this case in the opposite direction, is that of the rne (RNase E) transcript, in which there is a 361-nt 5' UTR (34). In wild-type cells, the half-life of the rne mRNA is 2 to 3 min (69). In the absence of functional RNase E protein, the half-life increases to more than 40 min (69). The presence of the 5' UTR appears to be essential for this significant difference in stability (76; Q. Liu, S. D. Yancey, and S. R. Kushner, unpublished data).
These two examples indicate that the 5' UTRs can either stabilize the downstream transcript (ompA) or destabilize it (rne). In the case of ompA, it appears that stabilization is most likely caused by an unusual secondary structure that prevents access to the nuclease(s) normally initiating decay at the 5' terminus. With the rne transcript the opposite may be the case. It is known that the rne leader region contains one or more RNase E cleavage sites (Q. Liu and S. R. Kushner, unpublished results). In addition, there are an increasing number of examples of regulation of genes such as pnp and rnc by RNase III cleavage at stem-loop structures in their respective 5' UTRs (15, 123, 124). In these two cases, the 5' UTR seems to function independently of translation initiation, since the RNase E and RNase III cleavage sites and the 5' terminus of the ompA mRNA are far removed from the ribosome binding site. In some cases, however, the effect of the 5' UTR may also depend on translation initiation.
Many transcripts contain stem-loop structures in their 3' UTRs. Two sources of stem-loops are Rho-independent transcription termination events (103) and the presence of REP sequences (58, 72, 144). There are many cases where the presence of a stem-loop structure in the 3' UTR increases the stability of the upstream transcript (73, 108, 109). This increased stability is thought to arise from the inhibition of both RNase II and polynucleotide phosphorylase by (PNPase) RNA secondary structures (60, 65, 90, 96, 113). These secondary structures, in the presence of a protein (EIF, exonuclease-impeding factor), seem to slow the movement of PNPase in the 3'-to-5' direction (30). Another potentially important feature of the 3' end of the mRNA molecule is the existence of a poly(A) tail on many, if not all, mRNAs at some point during their lifetime in the bacteria (66, 114, 149). This subject will be discussed in more detail later in the section on Posttranscriptional Modification of mRNAs.
With the possible exception of the pyrimidine-adenosine (Pyr-A) RNase M cleavage site (25), it is not clear whether specific nucleotide sequences, comparable to either type I or type II DNA restriction enzyme recognition sites, for example, play a significant role in the decay of mRNAs. Thus, cleavage site specificity for endonucleases involved in mRNA decay most likely is very dependent on secondary and possibly tertiary structures. The apparent absence of traditional-type 5'-to-3' exonucleases—enzymes that degrade RNA 1 nt at a time—in E. coli makes the importance of 5' termini as discussed in the case of the ompA mRNA particularly intriguing. Possibly, there is an exoribonuclease that requires a 5' terminus to bind but cleaves the RNA at a distance from the binding site. The RecBCD enzyme is an example of a DNase that functions this way (46, 137).
Computer folding programs now permit the easy determination of possible RNA secondary structures. In the absence of translation, it has been possible to determine in vivo whether the computer-predicted stem-loop structures actually exist (52). Clearly, however, the very stable computer-predicted whole mRNA structures (for example, the calculated free energy of the complete 493-nt trxA transcript is –125 kcal [–523 kJ] [C. M. Arraiano and S. R. Kushner, unpublished results]) are unlikely to exist in vivo because of ribosomes on the mRNA. Since, however, ribosomes are not uniformly spaced on mRNAs, it is hard to predict the extent, nature, and significance of higher-order structures within the translated portions of mRNAs. These structures most likely are integral to the decay process, particularly for targeting endonucleolytic cleavages.
Finally, the idea that some mRNA cleavage reactions in E. coli could be autocatalytic is within the realm of possibility. Although there is no direct evidence to rule out this possibility, so far our laboratory has not found any evidence for it, although we have used many different in vitro-generated transcripts incubated for long times under conditions favoring autocatalysis (E. B. O’Hara and S. R. Kushner, unpublished results). If the phenomenon does exist in E. coli, it probably affects only a few mRNAs.
Most half-life determinations appearing in the literature measure the loss of either total pulse-labeled RNA or a particular RNA sequence. This approach results in the term chemical half-life because it has nothing to do with the actual functionality of a given mRNA molecule. These types of measurements are performed using two basic approaches. Usually, 3H-labeled uracil or uridine is used to pulse-label a logarithmically growing culture. Following the addition of rifampin to stop new transcription initiation, samples are taken at various times. The loss of radioactivity is then measured either by acid precipitation (to determine bulk half-lives) or by using RNA-DNA hybridization techniques (to determine specific mRNA half-lives).
In determining bulk RNA half-lives, a problem with the technique has been what to use as the baseline for 3H incorporation into stable RNA species. With wild-type E. coli, the published numbers are probably a reasonable estimate because by 60 min after the cessation of new transcription initiation, most of the remaining acid-insoluble radioactivity should only be associated with rRNA species. However, with many of the RNA decay mutants that have much slower mRNA decay rates (8), choosing an appropriate baseline is more difficult.
This is not a problem when determining chemical half-lives with hybridization techniques because the loss of a particular RNA species can be measured directly, either on an RNA-DNA dot blot or on Northern (RNA) blots. Hybridization techniques are not without difficulties, however. If the dot blot technique is used, the half-life being measured is simply the rate of loss of the particular portion of the transcript that hybridizes to the probe used. If the probe is a short oligonucleotide, the measurement obtained may not accurately measure the true chemical half-life, particularly if the probe happens to fall in a section of the transcript that is rapidly decayed relative to the rest of the message.
By using Northern analysis and measuring the decay of the full-length transcript, this problem can be avoided. Here too, however, there are difficulties associated with the technique. There can be transfer problems from the gel to the membrane. In addition, while species of up to 1,500 nt can be resolved clearly on polyacrylamide gels (8), the larger transcripts have to be separated on various types of agarose gels. The resolution of these systems is not particularly satisfying.
There are other technical problems in addition to those described above. For one, although adding rifampin is thought to specifically block transcription initiation, it may affect the cell in some other way that, in turn, indirectly affects mRNA decay. Second, in many E. coli strains, the mRNA decay curves frequently appear biphasic, with an initial rapid burst of decay followed by a much slower turnover of the remainder of the transcript. Until this type of decay curve is better understood, it is not clear what represents a true chemical half-life measurement.
A final problem associated with chemical decay studies is that measurements for the same mRNA may vary considerably between laboratories and even within laboratories. Thus, it is probably best to view chemical half-lives as diagnostics rather than absolute numbers. Comparisons do, however, appear to be valid if half-lives have been determined within the same experiment. A difference between half-lives of 4 and 5.5 min may then be significant if the two numbers were determined side by side in a number of experiments. If, however, these two numbers arose from two determinations done in different laboratories, the difference may not be statistically significant.
A functional half-life measures how long an mRNA remains sufficiently intact to produce a full-length protein. These measurements are thus far more difficult, requiring either an immunological (109) or enzymatic (18) assay for the protein product of a given mRNA. The technique is particularly suited for inducible transcripts such as gal (1) or lac.
Pedersen et al. (119) used a completely different approach to measure functional half-lives. They assumed that the potential to synthesize a particular protein was equal to the functional decay rate of the corresponding mRNA. Synthetic potential was determined using two-dimensional polyacrylamide gel electrophoresis of 35S-labeled proteins following inhibition of transcription with either rifampin or streptolydigin. In this fashion they determined half-lives ranging from 40 s for the ribosomal protein S1 mRNA to 20 min for an unidentified protein (119). Their data also suggested that mRNAs within a single transcriptional unit could have different functional half-lives.
Functional half-lives measure a completely different type of mRNA decay. All that is necessary to functionally inactivate an mRNA is a single phosphodiester bond cleavage (to remove the ribosome binding site, for example). A similar cleavage within the coding sequence could also lead to functional inactivation, but only if the truncated polypeptide produced is no longer biochemically active. It is thus possible that mRNAs with short functional half-lives can have long chemical half-lives. It is important when reading the mRNA decay literature to distinguish between chemical and functional half-life determinations.
As initially envisioned by Apirion in 1973 (4), mRNA decay involved the combined action of endo- and exonucleases to convert full-length transcripts into mononucleotides. However, at that time there were relatively few genetic or biochemical data to support this hypothesis. In the ensuing years considerable progress has been made in understanding the complex events involved in mRNA decay. Deutscher (41, 42, 43) has summarized the biochemical properties of the E. coli RNases on several occasions. I have chosen here to try to integrate the genetic and biochemical evidence that indicates a role for a particular protein in mRNA decay. Some enzymes have been included even if genetic data are lacking because of their interesting catalytic properties. In addition, some interesting genes are also discussed even though the relevant biochemistry is not currently complete.
As historically defined, exonucleases degrade RNA substrates 1 nt at a time from either the 3' or 5' terminus. However, it is possible that there are RNases that require a terminus to bind but cleave phosphodiester bonds at a distance from the terminus similar to the already-mentioned RecBCD DNase. As will also be noted, in E. coli there are examples of both nucleolytic decay, which releases mononucleotides (RNase II), and phosphorolytic cleavage, which generates mononucleotide diphosphates (PNPase and RNase PH).
RNase II (rnb).
RNase II is a 3'-to-5 exonuclease that is highly processive but is sensitive to secondary structure (113, 141). The enzyme requires both magnesium and potassium to activate it (65). The first rnb mutation was identified in 1976 (110). Some 7 years later the structural gene for RNase II (rnb) was cloned and shown to encode a 72,000-dalton protein (47). Overexpression of RNase II activity did not change mRNA decay rates (47). The rnb structural gene has recently been sequenced by Zilhao et al. with a predicted molecular mass of 67,583 daltons for RNase II (160).
While Apirion and coworkers asserted that RNase II was involved in mRNA decay (78, 84), it was not until 1986 that Donovan and Kushner (48) showed that RNase II was essential for cell viability in the absence of PNPase. Even more important was the discovery that in the absence of both RNase II and PNPase, cells accumulated small partially degraded mRNA species (48). Recently, RNase II has also been shown to participate in a limited way in tRNA processing (133).
PNPase (pnp).
In the presence of millimolar levels of Pi, PNPase catalyzes the processive 3'-to-5' phosphorolytic degradation of RNA to nucleotide 5'-diphosphates (64). Some years after the first pnp mutations were isolated by Reiner (132) in 1969, a role for PNPase in mRNA decay was clearly demonstrated by Donovan and Kushner (48). Early purification studies suggested that the enzyme consisted of two subunits of 86,000 and 48,000 daltons (122, 125). The structural gene for the 86,000-dalton subunit was cloned and sequenced (124, 131). The exact nature of the 48,000-dalton subunit remains unclear. Deutscher (43) has pointed out that phosphorolytic cleavage has the advantage of capturing the energy from phosphodiester bond cleavage in the form of a nucleotide diphosphate. Recently, Causton et al. (30) suggested that a protein called EIF might associate with stem-loop structures and prevent the movement of PNPase through the structure. In addition, results now indicate that PNPase is the primary enzyme involved in the degradation of poly(A) tails (114) (J. A. Chekanova and S. R. Kushner, unpublished data).
RNase PH (rph).
Working with PNPase-deficient extracts of E. coli, Deutscher et al. (44) identified a second phosphorolytic activity that differed in size and substrate specificity from PNPase. They called it RNase PH. The structural gene (rph) was cloned, sequenced, and shown to encode a 238-amino-acid protein with a molecular mass of 25,500 daltons (126, 127). Although its primary role appears to be in the maturation of tRNAs (40, 80), preliminary experiments suggest that it can also help degrade poly(A) tails (Chekanova and Kushner, unpublished).
Other Exonucleases.
Deutscher and his colleagues have closely examined both the genetic and biochemical aspects of tRNA maturation. While it seems unlikely that any of the specific tRNA maturation enzymes (RNase D, RNase BN, RNase T) described by Deutscher (43) are involved in mRNA decay, until they are specifically ruled out, one must presume that they might play a role. In addition, it should be noted that even in the absence of both RNase II and PNPase, mRNA decay is slowed but not completely inhibited (8). Thus, there must be at least one additional exonuclease that converts the oligonucleotides generated by endonucleolytic decay into mononucleotides.
Riboendonucleases are defined as enzymes that do not require a terminus to bind. As will be noted in the discussion that follows, the substrate specificity associated with the known E. coli enzymes suggests a significant role for higher-order structures rather than actual nucleotide sequence.
RNase I/RNase I* (rna).
RNase I was originally identified as a nonspecific endonuclease located in the periplasmic space of E. coli (107). The structural gene for RNase I (rna) has been cloned and sequenced (97, 159) and shown to encode a 27,000-dalton polypeptide. Meador et al. (97) demonstrated that RNase I could degrade mRNA as well as the four ribonucleotide homopolymers.
While RNase I (rna) is a potential candidate for an RNA decay enzyme, its location in the periplasmic space (107) and the absence of any demonstrable phenotype in rna deletion mutants (159) seemed to rule it out as a major player in mRNA decay. Hautala et al. (68) did note, however, that the rna-19 allele had a small effect on the expression of the Neurospora crassa catabolic dehydroquinase gene in E. coli. This effect suggested that RNase I might play some role in mRNA decay. Recently, Cannistraro and Kennell (24) reported an altered form of RNase I, called RNase I*, which is associated with the inner membrane. They have suggested that RNase I* has a small role in the cleavage of larger RNA oligonucleotides and a primary role in the degradation of small oligonucleotides (25).
RNase III (rnc).
RNase III was first identified as an endoribonuclease that cleaves double-stranded RNA molecules (135, 136). Since then, it has been extensively studied to determine its substrate specificity, as recently reviewed by Court (38). The structural gene for RNase III (rnc ) encodes a 25,000-dalton polypeptide. The enzyme seems to specifically recognize stem-loop structures and can cleave either on one or both sides of the stem, usually within an internal unpaired region, to yield a two-base 3' overhang (19, 92, 117, 140, 158). The exact substrate specificity of the enzyme has yet to be determined. By comparing 34 RNase III cleavage sites, Krinke and Wulff (85) found a consensus sequence of A/UNAGA/UGNNCA/UUNN within one arm of the stem. However, the conserved CUU/GAA base-paired sequence immediately adjacent to the unpaired region where cleavage occurs was not required for accuracy or selectivity (31). Apparently, then, some type of tertiary structure is also required for RNase III recognition.
Following the 1973 identification of the first rnc mutation (83), RNase III was subsequently shown to be involved in the processing of 30S rRNA (49). Although the major phenotype associated with the rnc-105 allele is the accumulation of unprocessed 30S rRNA transcripts, approximately 10% of all cellular proteins are either under- or overproduced in an rnc-105 mutant (59, 147). This observation suggests that RNase III plays a direct role in the half-lives of a discrete subset of E. coli mRNAs, even though the half-lives of the majority of cellular mRNAs are essentially unaffected (6, 150). The enzyme directly affects the mRNA stability of a number of genes, including rnc (15), pnp (123), dicB (54), and metY (130).
Recently, Takiff et al. (148) and Babitzke et al. (11) showed that the structural gene for RNase III could be deleted without any loss of cell viability.
RNase E (rne/ams).
A second major rRNA-processing endonuclease in E. coli is RNase E. Originally identified by Ghora and Apirion (57) as an enzyme involved in the processing of 5S rRNA from a 9S rRNA precursor (5, 7), the genetic proof of its involvement in mRNA decay has taken a more convoluted path. In 1977, Kuwano et al. (86) isolated a mutation called ams (altered mRNA stability) that led to inviability at elevated temperatures and two- to threefold increases in the chemical half-life of total E. coli pulse-labeled RNA. Most interesting was that while mRNA decay appeared to be altered, protein synthesis continued normally for several hours after shift to the nonpermissive temperature (115).
Subsequently, Arraiano et al. (8) showed that a combination of mutations in ams pnp rnb led to a significant increase in the half-life of total pulse-labeled RNA (11 to 15 min in the triple mutant versus 3 min in the wild type). Additionally, they showed that discrete decay intermediates of the trxA (thioredoxin) and cat (chloramphenicol acetyltransferase) mRNAs could be observed in such mutants following shift to the nonpermissive temperature. Chemical half-life measurements for many specific mRNAs showed increases from two to fourfold (8).
Further studies with the ams-1 pnp-7 rnb-500 triple mutant have shown that the decay intermediates observed with trxA (9) and pyrF (C. M. Arraiano, A. A. Cruz, and S. R. Kushner, submitted for publication) arise from multiple endonucleolytic cleavages, many of which leave the 5' terminus of both mRNAs intact. Interestingly, the absence of RNase E, RNase II, and PNPase was only sufficient to retard mRNA decay, not to completely block it.
The cloning and initial DNA sequencing of the ams gene (33, 34) assisted researchers in determining that ams and rne encoded the same protein (12, 99, 104, 151). Further studies by Casarégola et al. (29) and Cormack et al. (37) have shown that the original ams/rne clones were missing DNA sequences encoding more than 200 amino acids from the carboxy terminus of the product. The truncated clones were sufficient, however, to complement the ams-1 and rne-3071 mutations in vivo (33).
RNase E requires either Mg+ or Mn2+ and a monovalent cation for activity (102). rne encodes a 1,025-amino-acid protein that migrates on sodium dodecyl sulfate gels with an apparent molecular weight of 180,000. The protein is unusual because it contains over 25% charged amino acids and only five cysteine residues and is much larger than the majority of other characterized RNases.
The specificity of the enzyme has been somewhat controversial. In 1988 Mudd et al. first proposed a 10-nucleotide recognition sequence on the basis of the observations of Ghora and Apirion (57) and their own experience with processed bacteriophage T4 mRNAs (105). Subsequently, Ehretsmann et al. (50) suggested that the consensus recognition sequence was A/GAUU/AU, a subset of the 10-nt site. In addition, it appears that the sequence occurs in single-stranded regions and can be flanked by stem-loop structures. Recent work in the laboratory of S. Cohen, however, suggests that RNase E has few primary structural constraints other than a preference for cleaving 5' to an AU dinucleotide (88, 95).
Another question regarding RNase E is the nature of the active protein. The enzyme is very susceptible to proteolysis, and a number of truncated forms that are still enzymatically active have been detected. While the full-length protein has catalytic activity (37, 152), a protein missing several hundred amino acids at the carboxy terminus has similar specificity in vivo (33).
Recent evidence has shown that RNase E is associated with additional proteins in vivo (28, 128). Clearly identified has been PNPase, along with several other polypeptides.
RNase M.
RNase M has an observed molecular weight of 26,000. In 1989, Cannistraro and Kennell proposed it to be involved in the decay of lac mRNA (23). The enzyme appears to have a target specificity of Pyr-A and cleaves RNA to give 5'-OH termini (23). On the basis of tryptic fingerprints, the protein seems to resemble RNase I (97), but its activity was still present to some extent in a cell with an interrupted rna gene (143). Cannistraro and Kennell (25) have argued that RNase M is the primary endoribonuclease for mRNA degradation in E. coli. Presently, however, there is no genetic proof to support this conclusion.
Other Endonucleases.
There have been reports of two broad- specificity endonucleases, both called RNase R, that could also be involved in mRNA decay (41, 143). In addition, a recent paper by Alifano et al. (3) suggests some role for RNase P (rnp) in mRNA decay. As will be noted later, there may also be one or more additional endonucleases involved in mRNA that have yet to be characterized.
Poly(A) polymerase catalyzes the template-independent sequential addition of AMP to the 3'-terminal hydroxyl groups of RNA molecules by using ATP as a substrate. The enzyme was first identified in E. coli in 1962 (10) and purified and characterized in 1976 (129). It requires both Mg2+ and Mn2+ for optimal activity (26). The structural gene for poly(A) polymerase I (pcnB) has been cloned and sequenced (26, 91). It encodes a protein of approximately 55,000 daltons. A role of poly(A) polymerase in mRNA decay has now clearly been demonstrated (66, 114). This evidence will be discussed in more detail later.
E. coli apparently contains a second poly(A) polymerase (77) that accounts for the small amount of polyadenylation seen in pcnB mutants (114). The structural gene for this enzyme has not yet been cloned.
When Babitzke et al. (11) constructed an ams/rne pnp rnb rnc quadruple mutant and examined mRNA decay at the nonpermissive temperature, they found that eliminating RNase III led to more rapid mRNA turnover compared with the pnp rnb ams/rne triple mutant. This surprising result was interpreted to mean that the absence of RNase III leads to an increased synthesis of enzymes that can compensate for the loss of RNase E, RNase III, RNase II, and PNPase (11).
That mRNA decay continues relatively normally even when four major RNases are missing has led to the search for additional genes that affect mRNA decay. Although some mutations that affect mRNA decay also lead to cell inviability (rne/ams, pnp rnb), others do not (rnc). There has thus been no simple selection method for new mutants. Recently, a series of temperature-sensitive mutants were isolated in a screen for new pnp alleles (L.L. Granger, E.B. O’Hara, R.-F. Wang, F. V. Meffen, K. Armstrong, S. D. Yancey, and S. R. Kushner, submitted for publication). Analysis of these mutants has led to the identification of three new loci (called mrs for mRNA stability) that affect mRNA decay.
The best characterized of these alleles (mrsC) encodes a Mg2+-dependent ATPase (R.-F. Wang, M. Aldea, C.I. Bargmann, and S. R. Kushner, submitted for publication), whose gene has independently been identified as a filamentation locus (ftsH [154]), a protein involved in the transfer of proteins across the inner membrane (2) and high frequency of bacteriophage λ lysogenization (hflB [70]). Inactivating the MrsC/FtsH/HflB protein in the absence of RNase E (rne), RNase II (rnb), and PNPase (pnp) causes a rapid cessation of RNA synthesis and a dramatic increase in the stability of individual mRNAs such as trxA, cat, and kan (Granger et al., submitted). While the MrsC/FtsH/HflB protein has been shown to be membrane anchored through its amino terminus (153), its exact role in vivo is yet unknown. Several reports have suggested that the MrsC/FtsH/HflB protein functions as an ATP-dependent protease (14, 71). If this is the protein’s only catalytic activity, it suggests that proteolysis is necessary to activate an RNase involved in mRNA decay.
The mrsA gene is also required for cell viability as well as for mRNA decay (O’Hara and Kushner, unpublished). mRNA decay is dramatically impaired in mrsA rne-1 double mutants. Although Wang et al. showed that the gene encodes a 49,000-dalton protein containing a helix-turn-helix motif, its biochemical activity is still unknown (R.-F. Wang, R. Dickson, and S. R. Kushner, unpublished data). The mrsF gene encodes a 27,000-dalton protein that is part of the same mRNA decay pathway as rne (F. Zheng, R.-F. Wang, and S. R. Kushner, unpublished data).
It has generally been assumed that mRNA decay is a straightforward salvage pathway that is constitutively expressed in E. coli. The first evidence that this assumption was false was the discovery that RNase III regulates the levels of both PNPase and itself through cleavage of their respective mRNAs (15, 123, 124). In some cases, mRNAs actually decay faster in the absence of RNase III (11). RNase E regulates its own synthesis by dramatically changing its mRNA half-life (69, 76). There is 10- to 20-fold more RNase E protein in an rne-1 strain at the nonpermissive temperature than in a wild-type control (F. Zheng and S. R. Kushner, unpublished results). Apparently, then, the relative amounts of several RNases are controlled in part by ribonucleolytic digestion of the respective mRNAs. Controlling enzyme levels seems to be important because overproducing either RNase E or poly(A) polymerase I causes a loss of cell viability.
Besides the control of many of the various RNA decay enzymes by a variety of posttranscriptional mechanisms, the physical structure of each mRNA molecule at the 5' and 3' ends (as discussed above in the section on Structural Properties of mRNA Molecules) also serves to regulate decay. There are probably additional structures within the translated regions of mRNAs that are also important for decay.
Apparently, each mRNA has programmed into it susceptibility to several possible decay pathways. One reason to assume that secondary and tertiary structures are important is that decay rates for individual mRNAs are independent of transcript length. This fact rules out the idea that longer mRNAs have more possible target sites and therefore decay more quickly. Rather, it appears that each mRNA can be degraded by both primary and secondary pathways. If a primary pathway is inactivated, one or more secondary mechanisms fill in. In most cases, this is sufficient for the cell to continue to function. In special cases, the inability to decay mRNAs normally leads to specific deleterious effects, resulting in cell death.
The idea that mRNAs are modified after transcription first appeared in 1975 when two laboratories (106, 142) suggested that there were poly(A) tails on pulse-labeled RNA species in E. coli. Even though a poly(A) polymerase activity was identified in E. coli in 1962 (10), it was generally assumed that polyadenylation was unique to eukaryotes (13, 87). From 1975 through 1992, several laboratories published a variety of reports on the existence of poly(A) tails on the RNAs isolated from E. coli (27, 61, 79, 138, 142) and Bacillus subtilis (62). By isolating and sequencing a number of lpp cDNAs, Cao and Sarkar showed that there were short poly(A) tails on the lpp mRNA (26). The tails were attached either to the normal 3' end following the Rho-independent transcription termination site or to a slightly shortened mRNA species in which the 3' stem-loop had been removed (26).
Recent work by O’Hara et al. (114) has demonstrated that in wild-type E. coli the average poly(A) tail length is between 10 and 40 nt. In the absence of PNPase and RNase II, tail lengths increase to more than 100 nt. More than 90% of the poly(A) tails are eliminated in the absence of poly(A) polymerase I (114), the product of the pcnB (26) gene. By sequencing random cDNA clones made from polyadenylated E. coli RNA, it has been possible to show that a variety of mRNAs are polyadenylated (Chekanova and Kushner, unpublished). The mRNAs are polyadenylated both at the normal 3' ends and at 3' ends within the coding sequences, suggesting that mRNA intermediates arising from either premature transcription termination or decay can be polyadenylated (Chekanova and Kushner, unpublished).
The extent of polyadenylation during the various stages of E. coli growth is yet unknown. Cao and Sarkar (26) determined that in wild-type bacteria approximately 1.3% of the total pulse-labeled RNA was polyadenylated. The levels of polyadenylation increased to 6% in the absence of both PNPase and RNase II. Preliminary experiments suggest that the extent of polyadenylation does not increase significantly in the absence of additional RNases such as RNase PH (Chekanova and Kushner, unpublished) but is significantly affected by either poly(A) polymerase I levels (C. A. Ingle and S. R. Kushner, unpublished data) or mutations in the structural gene for σ 32 (O’Hara and Kushner, unpublished). It is unclear whether at any given time in the cell only a few of the mRNAs can be polyadenylated or alternatively whether there is another 3'-to-5' exonuclease involved in poly(A) degradation that has yet to be identified. Since Deutscher and his coworkers have looked exhaustively for RNases in their analysis of tRNA processing (80, 81), it would appear that unless such an enzyme has very special cofactor requirements or is localized to the inner membrane, most of the 3'-to-5' exonucleases have already been identified.
One possible explanation for the low levels of polyadenylation is that at any time most of the mRNAs are not susceptible. The few polyadenylated molecules that do occur arise during the translation process when a 70S ribosome has an associated poly(A) polymerase I protein. In fact, Hardy and Kurland (67) estimated that as many as 10 molecules of poly(A) polymerase could be associated with a single ribosome. Figure 1 shows a model illustrating this idea. In this scheme, poly(A) polymerase I would be associated with some, but not all, 70S ribosomes translating mRNAs. Supporting this scheme is that there are far more ribosomes (estimated at 45,000 per cell at a generation time of 30 min [20]) than poly(A) polymerase molecules (although there is no accurate estimate, on the basis of purification data presented by Cao and Sarkar [26] there probably are fewer than 1,000 molecules per cell). Thus, when a 70S ribosome with an associated poly(A) polymerase enzyme reaches the end of an mRNA, polyadenylation could take place (Fig. 1). Assuming that the 1.3% poly(A) tails observed by Cao and Sarkar (27) in wild-type E. coli is an accurate measurement, that would suggest that only between 1 in 50 and 1 in 100 ribosomes would be associated with poly(A) polymerase I. Once poly(A) addition has taken place, the mRNA is rapidly degraded, leading to its loss from the polysomal fraction. This idea is supported by the fact that less than 1% of the polyadenylated RNA is found in polysomes (Chekanova and Kushner, unpublished). However, a major problem with this hypothesis is how to explain the polyadenylation of structural RNAs such as RNA I (156), partially degraded mRNAs (Chekanova and Kushner, unpublished), and 16S and 23S rRNAs (Chekanova and Kushner, unpublished). Another problem is that estimates of the extent of polyadenylation are not very quantitative.
It is also possible that polyadenylation in vivo is unrelated to the in vitro-observed association of poly(A) polymerase I and ribosomes (67; Ingle and Kushner, unpublished). If polyadenylation occurred randomly, one might expect to find a higher proportion of abundant mRNAs in a cDNA library made from E. coli polyadenylated RNA. This has not been the case. From the almost 100 cDNAs sequenced to date, no specific mRNA has been found more than once, and both abundant and rare mRNAs have been identified (Chekanova and Kushner, unpublished). Additional work will be necessary to determine at what stage of an mRNA’s life cycle polyadenylation takes place.
As first hypothesized by Apirion (4), mRNA decay seems to proceed by a series of endo- and exonucleolytic steps. Since E. coli and, presumably, its close cousin (Salmonella typhimurium [official designation, Salmonella enterica serovar Typhimurium]) apparently do not have 5'-to-3' riboexonucleases as does Saccharomyces cerevisiae (145), 3'-to-5' exonucleases such as RNase II and PNPase are the primary means for generating mononucleotides which can either be recycled into new RNA molecules or be converted into deoxynucleotides by nucleotide reductase and then funneled into DNA synthesis. In this regard the involvement of a phosphorolytic activity has advantages because it generates mononucleotide diphosphates (45).
While Apirion’s original description of mRNA decay correctly identifies the importance of endo- and exonucleases in the overall process (4), it does not satisfactorily explain the more recent biochemical and genetic findings. In addition, it does not deal with the implications of coupled transcription and translation or how the presence of ribosomes on the mRNA molecules influences the decay process. It also does not adequately account for how mRNAs, most likely in considerable excess of the number of RNase molecules, are degraded in an orderly way. A final problem is that the model does not explain the almost 30-fold difference in mRNA half-lives that have been empirically measured. The following paragraphs describe alternative hypotheses that may help explain these particular problems.
The results of O’Hara et al. (66, 114) suggest that polyadenylation plays an important role in mRNA decay. In particular, the loss of poly(A) polymerase I activity leads to a more than 90% reduction in the number of poly(A) tails observed (114). This reduction in poly(A) tails is accompanied by an increase in the chemical half-lives of many individual mRNAs (66, 114). The increase in half-lives is most striking in cells that are also deficient in PNPase, RNase II, and RNase E. One explanation for these results is that the poly(A) tail serves as a target for a series of mRNA-degrading complexes (Fig. 2).
The data that support this model are as follows. (i) RNase E and PNPase have now been shown to be part of a multiprotein complex (28, 128). By a number of biochemical criteria, these two activities copurify. Although the definitive experiments designed to show the in vivo association of these two proteins have yet to be done, it seems likely that this association is real. Thus, if PNPase is bound at the 3' end of an mRNA molecule, it is likely that an RNase E molecule will also be present. (ii) A unique protein (EIF) appears to associate with stem-loop structures at the 3' end of mRNAs, thereby impeding the progress of PNPase through the secondary structures (30). Since many E. coli mRNAs have such stem-loops at their 3' ends, it may be necessary to bypass them. The presence of RNase E along with PNPase would permit the cleavage of the RNA upstream of the stem-loop. (iii) For mRNAs that are terminated using a Rho-independent mechanism, there will be a short single-stranded region at the 3' base of the stem. This structure is probably a less effective substrate for PNPase than is a 10 to 40-nt stretch of A residues. Once PNPase effectively initiates decay on the poly(A) tail, it may have less difficulty in working its way through the secondary structure, even in the presence of EIF.
Thus, with the poly(A) tail anchoring a decay complex at the 3' end of the molecule, RNase E could cleave the mRNA upstream of the stem-loop. If there is extensive secondary structure present upstream of the new 3' terminus generated by RNase E cleavage, a second round of decay on the remaining mRNA could be initiated by polyadenylating the new 3' terminus. This polyadenylation would permit the binding of another PNPase-RNase E complex. This model predicts discrete shortening of mRNAs at the 3' end, a result documented for three mRNAs (trxA [9], cat [101], and pyrF [Arraiano et al., submitted]).
Since mRNA decay is only inhibited but not completely blocked in pnp rnb rne multiple mutants (8), there apparently is a second mRNA decay pathway that is poly(A) dependent but does not utilize PNPase and RNase E (Fig. 2). Employing the functional equivalent of a eukaryotic poly(A) binding protein, a yet unidentified endo- or exonuclease could bind to the poly(A) binding protein and initiate decay. Support for this hypothesis comes from the very dramatic mRNA stabilization observed in a poly(A) polymerase I, RNase E, PNPase, RNase II multiple mutant (114). In such strains only poly(A)-independent decay (Fig. 3) would be functional.
If polyadenylation were a major mechanism mediating mRNA decay, one would expect a more dramatic alteration of mRNA decay in a pcnB single mutant. In fact, although such mutants grow more slowly than wild-type controls, the changes seen in the half-lives of individual mRNAs are no more dramatic than those found in an rne single mutant (8, 114). There are major changes in half-lives and mRNA decay patterns if both poly(A) polymerase I and RNase E are absent, but the effects vary from one mRNA to another (114). Thus, while lpp mRNA decay is almost completely inhibited, trxA degradation is only slowed slightly.
Since we have now shown that the trxA transcript is polyadenylated (E. B. O’Hara, J. A. Chekanova, and S. R. Kushner, unpublished results), the minimal effect on trxA degradation in a pcnB rne pnp rnb multiple mutant must result from the presence of an effective alternative decay mechanism that is not poly(A) dependent. This system, however, does not work on lpp. Thus, as noted in Fig. 3, RNase E, RNase M, RNase I*, and possibly RNase X, as well as other as yet undiscovered nucleases, must function independently of the need for anchoring at the 3' end.
Mackie (personal communication) has suggested that in fact RNase E is anchored at the 5' end of the mRNA and moves in a 5'-to-3' direction to degrade particular mRNAs (Fig. 3). Another possibility is that enzymes such as RNase I (24), RNase M (23), or RNase R (143) are primarily responsible for RNA decay in the absence of poly(A) polymerase I, RNase E, RNase II, and PNPase.
Although many studies have suggested that growth rates do not influence the stability of most transcripts (20, 35, 55, 56, 118), a direct relationship between growth and the stability of the ompA and cat mRNAs has been reported (98, 111). More recently, Meyer and Schottel (100) showed that cat transcript stability was growth rate independent but varied with growth medium composition. Nilsson et al. (111) also showed that the bla and lpp mRNAs were not growth medium dependent. Additional studies are apparently needed to determine how important growth rate and medium composition are to mRNA stability.
Since mRNAs were discovered in 1961 (21, 63, 75), it has been assumed that the process of translation would stabilize mRNAs by protecting them from nucleolytic attack. Several lines of evidence have served to support this view. If antibiotics are used to stall ribosomes on the mRNA, stability increases (39, 89, 139). When termination codons or feedback inhibition of ribosomal protein synthesis are used to reduce the number of ribosomes on an mRNA (36, 53, 112), stability decreases. Kennell and Reizman (82) also observed a correlation between translation efficiency and mRNA stability in the lac operon.
Conflicting evidence, however, has also been reported. Mackie (93) noted that translational feedback regulation of ribosomal protein S20 led to the stabilization of the mRNA. In addition, von Gabain et al. (155) showed no differences in the stability of translated and untranslated segments of both the bla and ompA transcripts. Results of experiments using mRNAs with mutations in the leader or proximal part of a gene to alter the protein output of the mRNA suggested no clear explanation. A correlation between translation efficiency and mRNA stability was found in some cases (32, 157) but not others (22, 51, 120).
McCormick et al. (94) reexamined the question of translation and mRNA stability. For the fusion mRNAs that they tested (a combination of ribosomal protein S10 and β-galactosidase), they concluded that as translation decreased, mRNA degradation increased. They noted, however, that the rate of protein synthesis after inhibition of transcription with rifampin may not reliably indicate the kinetics of mRNA decay (94). Recently, Peterson (121) has reviewed this complex subject in considerable detail. Taken together, it would appear that translation is probably an important factor in the decay of some mRNAs but unimportant for others.
Since at least two riboendonucleases involved in mRNA decay, RNase III and RNase E, are also intimately involved in rRNA processing, it is worth considering the interrelationship of these two processes. rRNAs and many tRNAs are transcribed from seven rRNA operons as polycistronic molecules. The intervention of RNase III, RNase E, RNase P, and many exoribonucleases (see reference 133 for a recent review of tRNA processing) converts the 30S transcripts into mature 5S, 16S, 23S, and tRNA species. However, the absence of RNase III does not lead to cell death even though normal 23S processing is markedly impeded. In contrast, the absence of RNase E does lead to the accumulation of 9S rRNA precursors and subsequent cell death (5, 7).
Examination of the probable roles of RNase E and RNase III in mRNA decay reveals a striking difference. While the absence of RNase III causes protein level changes in approximately 10% of the total cellular proteins, none of these changes seems to interfere with cell survival. Additionally, most of the cases of RNase III involvement in mRNA decay suggest cleavages within polycistronic mRNAs that help to generate monocistronic transcripts. These monocistronic mRNAs, in turn, generally have very different half-lives (chemical and functional) than when they are contained within the larger polycistronic message. For example, in the absence of RNase III the half-life of the pnp mRNA is at least fivefold longer and there is a concomitant increase in PNPase protein (124, 146). This type of result would explain the variation of protein expression in rnc mutants. Additionally, the backup processing mechanism for 23S rRNA is sufficiently efficient to permit the assembly of the needed number of functional 50S ribosomes.
On the other hand, inactivating RNase E leads to cell inviability (7, 115). In contrast to RNase III, its role in mRNA decay in many cases seems to involve cleavages within the coding sequences of mRNAs, leading to functional inactivation. It is tempting to suggest that this result means that normal mRNA decay is more critical for cell viability than rRNA processing. There are some problems with this explanation, however.
Failure to produce 5S rRNA could prevent the formation of functional 50S ribosomes. This seems unlikely, though, because protein synthesis continues normally for several hours after shift to the nonpermissive temperature in an rne mutant (115, 116). Since the cells are still dividing during this period but the RNase E protein is functionally inactive, it appears that there are enough new functional 50S ribosomes being generated to continue protein synthesis.
Another explanation for the lethality observed in RNase E mutants could be that the protein has another function beyond rRNA processing and mRNA decay. The isolation of the RNase E, using antibodies against a yeast myosin protein (29), indicates that the protein may play some structural role in the bacterium. It might be possible to isolate rne mutants that have normal RNase activity but still die at the nonpermissive temperature.
Overall, it appears that E. coli has a finite number of RNases that can both process structural RNAs and, under slightly different circumstances, degrade mRNAs. The assembly of functional ribosomes may also be relatively flexible, allowing altered forms of 16S, 23S, and 5S structural RNAs to be incorporated. In contrast, the accumulation of partially degraded mRNA species may lead to cell death by utilizing ribosomes to synthesize nonfunctional truncated proteins. This might lead to a shortage of ribosomes for translating certain bacterial proteins essential for cell viability.
There are still many unanswered questions regarding both the mechanism and importance of mRNA decay. The following are some of these.
(i) Are there additional RNases involved in mRNA decay? (ii) What is the exact role of polyadenylation in mRNA decay? (iii) What is the complete composition of the PNPase-RNase E multiprotein complex? (iv) Is there a second multiprotein complex involved in poly(A)-dependent mRNA decay? (v) Why is RNase E an essential enzyme? (vi) Since both RNase I* and MrsC have been found associated with the inner membrane, are bacterial membranes involved in mRNA decay? (vii) Is the inviability of mRNA decay mutants related to alterations in protein synthesis? These questions represent only a small part of the important issues yet to be resolved before mRNA decay in E. coli and S. typhimurium can be truly understood.
This work was supported in part by a grant from the National Institute of General Medical Sciences (GM28760) to S.R.K.
References
1. Achord, D., and D. Kennell. 1974. Metabolism of messenger RNA from the gal operon of Escherichia coli. J. Mol. Biol. 90:581–599.
2. Akiyama, Y., Y. Shirai, and K. Ito. 1994. Involvement of FtsH in protein assembly into and through the membrane. II. Dominant mutations affecting FtsH functions. J. Biol. Chem. 269:5225–5229.
3. Alifano, P., F. Rivellini, C. Piscitelli, C. M. Arraiano, C. B. Bruni, and M. S. Carlomagno. 1994. Ribonuclease E provides substrates for ribonuclease P-dependent processing of a polycistronic mRNA. Genes Dev. 8:3021–3031.
4. Apirion, D. 1973. Degradation of RNA in Escherichia coli: a hypothesis. Mol. Gen. Genet. 122:313–322.
5. Apirion, D. 1978. Isolation, genetic mapping and some characterization of a mutation in Escherichia coli that affects the processing of ribonucleic acid. Genetics 90:659–671.
6. Apirion, D., and D. Gitelman. 1980. Decay of RNA in RNA processing mutants of Escherichia coli. Mol. Gen. Genet. 177:139–154.
7. Apirion, D., and A. B. Lasser. 1978. A conditional lethal mutant of Escherichia coli which affects the processing of ribosomal RNA. J. Biol. Chem. 253:1738–1742.
8. Arraiano, C. M., S. D. Yancey, and S. R. Kushner. 1988. Stabilization of discrete mRNA breakdown products in ams pnp rnb multiple mutants of Escherichia coli K-12. J. Bacteriol. 170:4625–4633.
9. Arraiano, C. M., S. D. Yancey, and S. R. Kushner. 1993. Identification of endonucleolytic cleavage sites involved in decay of Escherichia coli trxA mRNA. J. Bacteriol. 175:1043–1052.
10. August, J., P. J. Ortiz, and J. Hurwitz. 1962. Ribonucleic acid-dependent ribonucleotide incorporation. I. Purification and properties of the enzyme. J. Biol. Chem. 237:3786–3793.
11. Babitzke, P., L. Granger, and S. R. Kushner. 1993. Analysis of mRNA decay and rRNA processing in Escherichia coli multiple mutants carrying a deletion in RNase III. J. Bacteriol. 175:229–239.
12. Babitzke, P., and S. R. Kushner. 1991. The Ams (altered mRNA stability) protein and ribonuclease E are encoded by the same structural gene of Escherichia coli. Proc. Natl. Acad. Sci. USA 88:1–5.
13. Baker, E. J. 1993. Control of poly(A) length, p. 367–417. In J. Belasco and G. Brawerman (ed.), Control of Messenger RNA Stability. Academic Press, Inc., New York.
14. Banuett, F., M. A. Hoyt, L. McFarlane, H. Echols, and I. Herskowitz. 1986. hflB, a new Escherichia coli locus regulating lysogeny and the level of bacteriophage lambda cII protein. J. Mol. Biol. 187:213–224.
15. Bardwell, J. C. A., P. Regnier, S.-M. Chen, Y. Nakamura, M. Grunberg-Manago, and D. L. Court. 1989. Autoregulation of RNase III operon by mRNA processing. EMBO J. 8:3401–3407.
16. Belasco, J., and G. Brawerman (ed.). 1993. Control of Messenger RNA Stability, p. 517. Academic Press, Inc., New York.
17. Belasco, J. G., G. Nilsson, A. von Gabain, and S. N. Cohen. 1986. The stability of E. coli gene transcripts is dependent on determinants localized to specific mRNA segments. Cell 46:245–251.
18. Blundell, M., E. Craig, and D. Kennell. 1972. Decay rates of different mRNAs in Escherichia coli and models of decay. Nature (London) New Biol. 238:46–49.
19. Bram, R. J., R. A. Young, and J. A. Steitz. 1980. The ribonuclease III site flanking 23S sequences in the 30S ribosomal precursor RNA of Escherichia coli. Cell 19:393–401.
20. Bremer, H., and P. P. Dennis. 1987. Modulation of chemical composition and other parameters of the cell by growth rate, p. 1527–1542. In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, vol. 2. American Society of Microbiology, Washington, D.C.
21. Brenner, S., F. Jacob, and M. Meselson. 1961. An unstable intermediate carrying information from genes to ribosomes for protein synthesis. Nature (London) 190:576–581.
22. Cannistraro, V. J., and D. Kennell. 1979. Escherichia coli lac operator mRNA affects translation initiation of β-galactosidase mRNA. Nature (London) 277:407–409.
23. Cannistraro, V. J., and D. Kennell. 1989. Purification and characterization of ribonuclease M and mRNA degradation in Escherichia coli. Eur. J. Biochem. 181:363–370.
24. Cannistraro, V. J., and D. Kennell. 1991. RNase I*, a form of RNase I, and mRNA degradation in Escherichia coli. J. Bacteriol. 173:4653–4659.
25. Cannistraro, V. J., and D. Kennell. 1993. The 5' ends of RNA oligonucleotides in Escherichia coli and mRNA degradation. Eur. J. Biochem. 213:285–293.
26. Cao, G.-J., and N. Sarkar. 1992. Identification of the gene for an Escherichia coli poly(A) polymerase. Proc. Natl. Acad. Sci. USA 89:10380–10384.
27. Cao, G.-J., and N. Sarkar. 1992. Poly(A) RNA in Escherichia coli: nucleotide sequence at the junction of the lpp transcript and the polyadenylate moiety. Proc. Natl. Acad. Sci. USA 89:7546–7550.
28. Carpousis, A. J., G. Van Houwe, C. Ehretsmann, and H. M. Krisch. 1994. Copurification of E. coli RNAase E and PNPase: evidence for a specific association between two enzymes important in RNA processing and degradation. Cell 76:889–900.
29. Casarégola, S., A. Jacq, D. Laoudj, G. McGurk, S. Margarson, M. Tempête, V. Norris, and I. B. Holland. 1992. Cloning and analysis of the entire Escherichia coli ams gene. ams is identical to hmp-1 and encodes a 114 kDa protein that migrates as a 180 kDa protein. J. Mol. Biol. 228:30–40.
30. Causton, H., P. B. Py, R. S. McLaren, and C. F. Higgins. 1994. mRNA degradation in Escherichia coli: a novel factor which impedes the exoribonucleolytic activity of PNPase at stem-loop structures. Mol. Microbiol. 14:731–741.
31. Chelladurai, B. S., H. Li, and A. W. Nicholson. 1991. A conserved sequence element in ribonuclease III processing signals is not required for accurate in vitro enzymatic cleavage. Nucleic Acids Res. 19:1759–1766.
32. Cho, K.-O., and C. Yanofsky. 1988. Sequence changes preceding a Shine-Dalgarno region influence trpE mRNA translation and decay. J. Mol. Biol. 204:51–60.
33. Claverie-Martin, F., M. R. Diaz-Torres, S. D. Yancey, and S. R. Kushner. 1989. Cloning of the altered mRNA stability (ams) gene of Escherichia coli K-12. J. Bacteriol. 171:5479–5486.
34. Claverie-Martin, F., M. R. Diaz-Torres, S. D. Yancey, and S. R. Kushner. 1991. Analysis of the altered mRNA stability (ams) gene from Escherichia coli: nucleotide sequence, transcription analysis, and homology of its product to MRP3, a mitochondrial ribosomal protein from Neurospora crassa. J. Biol. Chem. 266:2843–2851.
35. Coffman, R. L., T. E. Norris, and A. L. Koch. 1971. Chain elongation rate of messenger and polypeptides in slowly growing Escherichia coli. J. Mol. Biol. 60:1–19.
36. Cole, J. K., and M. Nomura. 1986. Changes in the half-life of the ribosomal protein messenger RNA caused by translational repression. J. Mol. Biol. 188:383–392.
37. Cormack, R. S., J. L. Genereaux, and G. A. Mackie. 1993. RNase E activity is conferred by a single polypeptide: overexpression, purification, and properties of the ams/rne/hmp1 gene product. Proc. Natl. Acad. Sci. USA 90:9006–9010.
38. Court, D. 1993. RNA processing and degradation by RNase III, p. 71–117. In J. Belasco and G. Brawerman (ed.), Control of Messenger RNA Stability. Academic Press, Inc., New York.
39. Cremer, K. J., L. Silengo, and D. Schlessinger. 1974. Polypeptide formation and polyribosomes in Escherichia coli treated with chloramphenicol. J. Bacteriol. 118:582–589.
40. Cudny, H., and M. P. Deutscher. 1988. 3' Processing of tRNA precursors in ribonuclease-deficient Escherichia coli. J. Biol. Chem. 263:1518–1523.
41. Deutscher, M. P. 1985. E. coli RNases: making sense of alphabet soup. Cell 40:731–732.
42. Deutscher, M. P. 1988. The metabolic role of RNases. Trends Biochem. Sci. 13:136–139.
43. Deutscher, M. P. 1993. Ribonuclease multiplicity, diversity and complexity. J. Biol. Chem. 268:13011–13014.
44. Deutscher, M. P., G. T. Marshall, and H. Cudny. 1988. RNase PH: a new phosphate-dependent nuclease distinct from polynucleotide phosphorylase. Proc. Natl. Acad. Sci. USA 85:4710–4714.
45. Deutscher, M. P., and N. B. Reuven. 1991. Enzymic basis for hydrolytic versus phosphorolytic mRNA degradation in Escherichia coli and Bacillus subtilis. Proc. Natl. Acad. Sci. USA 88:3277–3280.
46. Dixon, D. A., and S. C. Kowalczykowski. 1993. The recombination hotspot, Chi, is a regulatory sequence that acts by attenuating the nuclease activity of the Escherichia coli RecBCD enzyme. Cell 73:87–96.
47. Donovan, W. P., and S. R. Kushner. 1983. Amplification of ribonuclease II (rnb) activity in Escherichia coli K-12. Nucleic Acids Res. 11:265–275.
48. Donovan, W. P., and S. R. Kushner. 1986. Polynucleotide phosphorylase and ribonuclease II are required for cell viability and mRNA turnover in Escherichia coli K-12. Proc. Natl. Acad. Sci. USA 83:120–124.
49. Dunn, J. J., and F. W. Studier. 1973. T7 early RNAs and Escherichia coli ribosomal RNAs are cut from large precursor RNAs in vivo by ribonuclease III. Proc. Natl. Acad. Sci. USA 70:3296–3300.
50. Ehretsmann, C. P., A. J. Carpousis, and H. M. Krisch. 1992. Specificity of Escherichia coli endoribonuclease RNase E: in vivo and in vitro analysis of mutants in a bacteriophage T4 mRNA processing site. Genes Dev. 6:149–159.
51. Emory, S. A., and J. G. Belasco. 1990. The ompA 5' untranslated RNA segment functions in Escherichia coli as a growth-rate-regulated mRNA stabilizer whose activity is unrelated to translational efficiency. J. Bacteriol. 172:4472–4481.
52. Emory, S. A., P. Bouvet, and J. G. Belasco. 1992. A 5'-terminal stem-loop structure can stabilize mRNA in Escherichia coli. Genes Dev. 6:135–148.
53. Fallon, A. M., C. S. Jinks, G. D. Strycharz, and M. Nomura. 1979. Regulation of ribosomal protein synthesis in Escherichia coli by selective mRNA inactivation. Proc. Natl. Acad. Sci. USA 76:3411–3415.
54. Faubladier, M., K. Cam, and J. P. Bouche. 1990. Escherichia coli cell division inhibitor dicF-RNA of the dicB operon. J. Mol. Biol. 32:461–471.
55. Gausing, K. 1974. Ribosomal protein in E. coli: rate of synthesis and pool size at different growth rates. Mol. Gen. Genet. 129:61–75.
56. Gausing, K. 1977. Regulation of ribosome production in Escherichia coli: synthesis and stability of ribosomal RNA and of ribosomal protein messenger RNA at different growth rates. J. Mol. Biol. 115:335–354.
57. Ghora, B. K., and D. Apirion. 1978. Structural analysis and in vitro processing to p5 (5S) rRNA of a 9S RNA molecule isolated from an rne mutant of E. coli. Cell 15:1055–1066.
58. Gilson, E., J.-M. Clement, D. Brutlag, and M. Hofnung. 1984. A family of dispersed repetitive extragenic palindrome DNA sequences in E. coli. EMBO J. 3:1417–1422.
59. Gitelman, D. R., and D. Apirion. 1980. The synthesis of some proteins is affected in RNA processing mutants of Escherichia coli. Biochem. Biophys. Res. Commun. 96:1063–1070.
60. Godefroy-Colburn, T., and M. Grunberg-Manago. 1972. Polynucleotide phosphorylase, p. 533–574. In P. D. Boyer (ed.), The Enzymes, vol. 7. Academic Press, Inc., New York.
61. Gopalakrishna, Y., D. Langley, and N. Sarkar. 1981. Detection of high levels of polyadenylate-containing RNA in bacteria by the use of a single-step RNA isolation procedure. Nucleic Acids Res. 9:3545–3554.
62. Gopalakrishna, Y., and N. Sarkar. 1982. The synthesis of DNA complementary to polyadenylate-containing RNA from Bacillus subtilis. J. Biol. Chem. 257:2747–2750.
63. Gros, F., H. Hiatt, W. Gilbert, C. B. Kurland, R. W. Risebrough, and J. D. Watson. 1961. Unstable ribonucleic acid revealed by pulse labelling of Escherichia coli. Nature (London) 190:581–585.
64. Grunberg-Manago, M. 1963. Polynucleotide phosphorylase. Prog. Nucleic Acid Res. 1:93–133.
65. Gupta, R. S., T. Kasai, and D. Schlessinger. 1977. Purification and some novel properties of RNase II. J. Biol. Chem. 252:8945–8951.
66. Hajnsdorf, E., F. Braun, J. Haugel-Nielsen, and P. Regnier. 1995. Polyadenylylation destabilizes the rpsO mRNA of Escherichia coli. Proc. Natl. Acad. Sci. USA 92:3973–3977.
67. Hardy, S. J. S., and C. G. Kurland. 1966. The relationship of poly A polymerase and the ribosomes. Biochemistry 11:3676–3684.
68. Hautala, J. A., C. L. Bassett, N. H. Giles, and S. R. Kushner. 1979. Increased expression of a eukaryotic gene in Escherichia coli through stabilization of its messenger RNA. Proc. Natl. Acad. Sci. USA 76:5774–5778.
69. Henry, M., S. D. Yancey, and S. R. Kushner. 1992. Role of the heat shock response in stability of mRNA in Escherichia coli K-12. J. Bacteriol. 174:743–748.
70. Herman, C., T. Ogura, T. Tomoyasu, S. Hiraga, Y. Akiyama, K. Ito, R. Thomas, R. D’Ari, and P. Bouloc. 1993. Cell growth and λ phage development controlled by the same essential Escherichia coli gene, ftsH/hflB. Proc. Natl. Acad. Sci. USA 90:10861–10865.
71. Herman, C., D. Thevenet, R. D’Ari, and P. Bouloc. 1995. Degradation of σ32, the heat shock regulator in Escherichia coli, is governed by HflB. Proc. Natl. Acad. Sci. USA 92:3516–3520.
72. Higgins, C. F., G. F.-L. Ames, W. M. Barnes, J. M. Clement, and M. Hofnung. 1982. A novel intercistronic regulatory element of prokaryotic operons. Nature (London) 298:760–762.
73. Higgins, C. F., R. S. McLaren, and S. F. Newbury. 1988. Repetitive extragenic palindromic sequences, mRNA stability and gene expression: evolution by gene conversion?—a review. Gene 72:3–14.
74. Hirashima, A., G. Childs, and M. Inouye. 1973. Differential inhibitory effects of antibiotics on the biosynthesis of envelope proteins of Escherichia coli. J. Mol. Biol. 79:373–389.
75. Jacob, F., and J. Monod. 1961. Genetic regulatory mechanisms in the synthesis of proteins. J. Mol. Biol. 3:318–356.
76. Jain, C., and J. G. Belasco. 1995. RNase E autoregulates its synthesis by controlling the degradation rate of its own mRNA in Escherichia coli: unusual sensitivity of the rne transcript to RNase E activity. Genes Dev. 9:84–96.
77. Kalapos, M. P., G.-J. Cao, S. R. Kushner, and N. Sarkar. 1994. Identification of a second poly(A) polymerase in Escherichia coli. Biochem. Biophys. Res. Commun. 198:459–465.
78. Kaplan, R., and D. Apirion. 1974. The involvement of ribonuclease I, ribonuclease II, and polynucleotide phosphorylase in the degradation of stable ribonucleic acid during carbon starvation in E. coli. J. Biol. Chem. 249:149–151.
79. Karnik, P., Y. Gopalakrishna, and N. Sarkar. 1986. Construction of a cDNA library from polyadenylated RNA of Bacillus subtilis and determination of some 3' terminal sequences. Gene 49:161–165.
80. Kelly, K. O., and M. P. Deutscher. 1992. Characterization of Escherichia coli RNase PH. J. Biol. Chem. 267:17153–17158.
81. Kelly, K. O., N. B. Reuven, Z. Li, and M. P. Deutscher. 1992. RNase PH is essential for tRNA processing and viability in RNase-deficient Escherichia coli cells. J. Biol. Chem. 267:16015–16018.
82. Kennell, D., and H. Reizman. 1977. Transcription and translation initiation frequencies of the Escherichia coli lac operon. J. Mol. Biol. 114:1–21.
83. Kindler, P., T. V. Keil, and P. H. Hofschneider. 1973. Isolation and characterization of an RNase III deficient mutant of Escherichia coli. Mol. Gen. Genet. 126:53–69.
84. Kinscherf, T. G., and D. Apirion. 1975. Polynucleotide phosphorylase can participate in decay of mRNA in Escherichia coli in absence of ribonuclease II. Mol. Gen. Genet. 139:357–362.
85. Krinke, L., and D. L. Wulff. 1990. The cleavage specificity of RNase III. Nucleic Acids Res. 18:4809–4815.
86. Kuwano, M., M. Ono, H. Endo, K. Hori, K. Nakamura, Y. Hirota, and Y. Ohnishi. 1977. Gene affecting longevity of messenger RNA: a mutant of Escherichia coli with altered mRNA stability. Mol. Gen. Genet. 154:279–285.
87. Lewin, B. 1994. Genes V, chapter 10. Oxford University Press, Oxford.
88. Lin-Chao, S., T.-T. Wong, K. J. McDowall, and S. N. Cohen. 1994. Effects of nucleotide sequence on the specificity of rne-dependent and RNase E-mediated cleavages of RNA I encoded by the pBR322 plasmid. J. Biol. Chem. 269:10797–10803.
89. Lindahl, L., and J. Forchhammer. 1969. Evidence for reduced breakdown of messenger RNA during blocked transcription or translation in Escherichia coli. J. Mol. Biol. 43:593–606.
90. Littauer, U. Z., and H. Soreq. 1982. Polynucleotide phosphorylase, p. 517–553. In P. D. Boyer (ed.), The Enzymes, vol. 17. Academic Press, Inc., New York.
91. Liu, J., and J. S. Parkinson. 1989. Genetics and sequence analysis of the pcnB locus, an Escherichia coli gene involved in plasmid copy number control. J. Bacteriol. 171:1254–1261.
92. Lozeron, H. A., J. E. Dahlberg, and W. Szybalski. 1976. Processing of the major leftward mRNA of coliphage lambda. Virology 71:262–277.
93. Mackie, G. A. 1987. Posttranscriptional regulation of ribosomal protein S20 and stability of the S20 mRNA species. J. Bacteriol. 169:2697–2701.
94. McCormick, J. R., J. M. Zengel, and L. Lindahl. 1994. Correlation of translation efficiency with the decay of lacZ mRNA in Escherichia coli. J. Mol. Biol. 239:608–622.
95. McDowall, K. J., S. Lin-Chao, and S. N. Cohen. 1994. A + U content rather than a particular nucleotide order determines the specificity of RNase E cleavage. J. Biol. Chem. 269:10790–10796.
96. McLaren, R. S., S. F. Newbury, G. S. C. Dance, H. Causton, and C. F. Higgins. 1991. mRNA degradation by processive 3'-5' exonucleases in vitro and the implications for prokaryotic mRNA decay in vivo. J. Mol. Biol. 221:81–95.
97. Meador, J., III, B. Cannon, V. J. Cannistraro, and D. Kennell. 1990. Purification and characterization of Escherichia coli RNase I. Comparisons with RNase M. Eur. J. Biochem. 187:549–553.
98. Melefors, Ö., and A. von Gabain. 1988. Regulation of stability of E. coli ompA mRNA. Cell 52:893–901.
99. Melefors, Ö., and A. A. von Gabain. 1991. Genetic studies of cleavage-initiated mRNA decay and processing of ribosomal 9S RNA show that the Escherichia coli ams and rne loci are the same. Mol. Microbiol. 5:857–864.
100. Meyer, B. J., and J. L. Schottel. 1991. A novel transcriptional response by the cat gene during slow growth of Escherichia coli. J. Bacteriol. 173:3523–3530.
101. Meyer, B. J., and J. L. Schottel. 1992. Characterization of cat messenger RNA decay suggests that turnover occurs by endonucleolytic cleavage in a 3' to 5' direction. Mol. Microbiol. 6:1095–1104.
102. Misra, T. K., and D. Apirion. 1979. RNase E, an RNA processing enzyme from Escherichia coli. J. Biol. Chem. 254:11154–11159.
103. Mott, J. E., J. L. Galloway, and T. Platt. 1985. Maturation of Escherichia coli tryptophan operon mRNA: evidence for 3' exonucleolytic processing after rho-dependent termination. EMBO J. 4:1887–1891.
104. Mudd, E. A., A. J. Carpousis, and H. M. Krisch. 1990. Escherichia coli RNAse E has a role in the decay of bacteriophage T4 mRNA. Genes Dev. 4:873–881.
105. Mudd, E. A., P. Prentki, D. Belin, and H. M. Krisch. 1988. Processing of unstable bacteriophage T4 32 mRNAs into a stable species requires Escherichia coli ribonuclease E. EMBO J. 7:3601–3607.
106. Nakazato, H., S. Venkatesan, and M. Edmonds. 1975. Polyadenylic acid sequences in E. coli messenger RNA. Nature (London) 256:144–146.
107. Neu, H. C., and L. A. Heppel. 1964. Some observations on the "latent" ribonuclease of Escherichia coli. Biochemistry 51:1267–1274.
108. Newbury, S. F., N. H. Smith, and C. F. Higgins. 1987. Differential mRNA stability controls relative gene expression within a polycistronic operon. Cell 51:1131–1143.
109. Newbury, S. F., N. H. Smith, E. C. Robinson, I. D. Hiles, and C. F. Higgins. 1987. Stabilization of translationally active mRNA by prokaryotic REP sequences. Cell 48:297–310.
110. Nikolaev, N., V. Folsom, and D. Schlessinger. 1976. Escherichia coli mutants deficient in exoribonucleases. Biochem. Biophys. Res. Commun. 70:920–924.
111. Nilsson, G., J. G. Belasco, S. N. Cohen, and A. von Gabain. 1984. Growth-rate dependent regulation of mRNA stability in Escherichia coli. Nature (London) 312:75–77.
112. Nilsson, G., J. G. Belasco, S. N. Cohen, and A. von Gabain. 1987. Effect of premature termination of translation on mRNA stability depends on the site of ribosome release. Proc. Natl. Acad. Sci. USA 84:4890–4894.
113. Nossal, N. G., and M. F. Singer. 1968. The processive degradation of individual polynucleotide chains. J. Biol. Chem. 243:913–922.
114. O’Hara, E. B., J. A. Chekanova, C. A. Ingle, Z. R. Kushner, E. Peters, and S. R. Kushner. 1995. Polyadenylation helps regulate mRNA decay in Escherichia coli. Proc. Natl. Acad. Sci. USA 92:1807–1811.
115. Ono, M., and M. Kuwano. 1979. A conditional lethal mutation in an Escherichia coli strain with a longer chemical lifetime of mRNA. J. Mol. Biol. 129:343–357.
116. Ono, M., and M. Kuwano. 1980. Chromosomal location of a gene for chemical longevity of messenger ribonucleic acid in a temperature-sensitive mutant of Escherichia coli. J. Bacteriol. 142:325–326.
117. Paddock, G. V., K. Fukada, J. Abelson, and H. D. Robertson. 1976. Cleavage of T4 species I ribonucleic acid by Escherichia coli exonuclease III. Nucleic Acids Res. 5:1351–1371.
118. Pato, M. L., and K. von Meyenbrug. 1970. Residual RNA synthesis in Escherichia coli after inhibition of initiation of transcription by rifampicin. Cold Spring Harbor Symp. Quant. Biol. 35:497–504.
119. Pedersen, S., S. Reeh, and J. D. Friesen. 1978. Functional mRNA half-lives in E. coli. Mol. Gen. Genet. 166:329–336.
120. Peterson, C. 1987. The functional stability of the lacZ transcript is sensitive towards sequence alterations immediately downstream of the ribosome binding site. Mol. Gen. Genet. 209:179–187.
121. Peterson, C. 1993. Translation and mRNA stability in bacteria: a complex relationship, p. 117–148. In J. Belasco and G. Brawerman (ed.), Control of Messenger RNA Stability. Academic Press, Inc., New York.
122. Portier, C. 1975. Quatenary structure of polynucleotide phosphorylase from Escherichia coli: evidence of a complex between two types of polypeptide chains. Eur. J. Biochem. 55:573–582.
123. Portier, C., L. Dondon, M. Grunberg-Manago, and P. Regnier. 1987. The first step in the functional inactivation of the Escherichia coli polynucleotide phosphorylase messenger is ribonuclease III processing at the 5' end. EMBO J. 6:2165–2170.
124. Portier, C., and P. Regnier. 1984. Expression of the rpsO and pnp genes. Structural analysis of a DNA fragment carrying their control regions. Nucleic Acids Res. 12:6091–6102.
125. Portier, C., R. van Rapenbusch, M. N. Thang, and M. Grunberg-Manago. 1973. Quaternary structure of polynucleotide phosphorylase. Eur. J. Biochem. 40:77–87.
126. Poulsen, P., R. Bonekamp, and K. F. Jensen. 1984. Structure of the Escherichia coli operon and control of pyrE expression by a UTP modulated intercistronic attenuation. EMBO J. 3:1783–1790.
127. Poulsen, P., K. F. Jensen, P. Valentin-Hansen, P. Carlsson, and L. G. Lundberg. 1983. Nucleotide sequence of the Escherichia coli pyrE gene and the DNA in front of the protein-coding region. Eur. J. Biochem. 135:223–229.
128. Py, B., H. Causton, E. A. Mudd, and C. F. Higgins. 1994. A protein complex mediating mRNA degradation in Escherichia coli. Mol. Microbiol. 14:717–729.
129. Ramanarayanan, M., and P. R. Srinivasan. 1976. Further studies on the isolation and properties of polyriboadenylate polymerase from Escherichia coli PR7 (RNase I–pnp). J. Biol. Chem. 251:6274–6286.
130. Regnier, P., and M. Grunberg-Manago. 1989. Cleavage by RNase III in the transcripts of the metY-nus-infB operon of Escherichia coli releases the tRNA and initiates the decay of the downstream mRNA. J. Mol. Biol. 210:293–302.
131. Regnier, P., M. Grunberg-Manago, and C. Portier. 1987. Nucleotide sequence of the pnp gene of Escherichia coli encoding polynucleotide phosphorylase. J. Biol. Chem. 262:63–68.
132. Reiner, A. M. 1969. Characterization of polynucleotide phosphorylase mutants of Escherichia coli. J. Bacteriol. 97:1437–1443.
133. Reuven, N. B., and M. P. Deutscher. 1993. Multiple exoribonucleases are required for the 3' processing of Escherichia coli tRNA precursors in vivo. FASEB J. 7:143–148.
134. Richardson, J. P. 1993. Transcription termination. Crit. Rev. Biochem. Mol. Biol. 28:1–30.
135. Robertson, H. D., R. E. Webster, and N. D. Zinder. 1967. A nuclease specific for double-stranded RNA. Virology 12:718–719.
136. Robertson, H. D., R. E. Webster, and N. D. Zinder. 1968. Purification and properties of ribonuclease III from Escherichia coli. J. Biol. Chem. 243:82–91.
137. Rosamond, J., K. M. Telander, and S. Linn. 1979. Modulation of the action of the RecBC enzyme of Escherichia coli K12 by Ca++. J. Biol. Chem. 254:8646–8652.
138. Sarkar, N., D. Langley, and H. Paulus. 1978. Isolation and characterization of polyadenylate-containing RNA. Biochemistry 17:3468–3474.
139. Schneider, E., M. Blundell, and D. Kennell. 1978. Translation and mRNA decay. Mol. Gen. Genet. 160:121–129.
140. Schweitz, H., and J. P. Ebel. 1971. A study of the mechanism of action of Escherichia coli ribonuclease 3. Biochimie 5:585–593.
141. Singer, M. F., and G. Tolbert. 1965. Purification and properties of a potassium-activated phosphodiesterase (RNase II) from Escherichia coli. Biochemistry 4:1319–1330.
142. Srinivasan, P. R., M. Ramanarayanan, and E. Rabbani. 1975. Presence of polyriboadenylate sequences in pulse-labeled RNA of Escherichia coli. Proc. Natl. Acad. Sci. USA 72:2910–2914.
143. Srivastava, S. K., V. J. Cannistraro, and D. Kennell. 1992. Broad-specificity endoribonucleases and mRNA degradation in Escherichia coli. J. Bacteriol. 174:56–62.
144. Stern, M. J., G. F.-L. Ames, N. H. Smith, E. C. Robinson, and C. F. Higgins. 1984. Repetitive extragenic palindromic sequences: a major component of the bacterial genome. Cell 37:1015–1026.
145. Stevens, A. 1993. Eukaryotic nuclease and mRNA turnover, p. 449–471. In J. Belasco and G. Brawerman (ed.), Control of Messenger RNA Stability. Academic Press, Inc., New York.
146. Takata, R., T. Mukai, and K. Hori. 1985. Attenuation and processing of RNA from the rpsO-pnp transcription unit of Escherichia coli. Nucleic Acids Res. 13:7289–7297.
147. Takata, R., T. Mukai, and K. Hori. 1987. RNA processing by RNase III is involved in the synthesis of Escherichia coli polynucleotide phosphorylase. Mol. Gen. Genet. 209:28–32.
148. Takiff, H. E., T. Baker, T. Copeland, S.-M. Chen, and D. L. Court. 1992. Locating essential Escherichia coli genes by using mini-Tn10 transposons: the pdxJ operon. J. Bacteriol. 174:1544–1553.
149. Taljanidisz, J., P. Karnik, and N. Sarkar. 1987. Messenger ribonucleic acid for the lipoprotein of the Escherichia coli outer membrane is polyadenylated. J. Mol. Biol. 193:507–515.
150. Talkad, V., D. Achord, and D. Kennell. 1978. Altered mRNA metabolism in ribonuclease III-deficient strains of Escherichia coli. J. Bacteriol. 135:528–541.
151. Taraseviciene, L., A. Miczak, and D. Apirion. 1991. The gene specifying RNase E (rne) and a gene affecting mRNA stability (ams) are the same gene. Mol. Microbiol. 5:851–855.
152. Taraseviciene, L., S. Naureckiene, and B. E. Uhlin. 1994. Immunoaffinity purification of the Escherichia coli rne gene product. J. Biol. Chem. 269:12167–12172.
153. Tomoyasu, T., K. Yamanaka, K. Murata, T. Suzaki, P. Bouloc, A. Kato, H. Niki, S. Hiraga, and T. Ogura. 1993. Topology and subcellular localization of FtsH protein in Escherichia coli. J. Bacteriol. 175:1352–1357.
154. Tomoyasu, T., T. Yuki, S. Morimura, H. Mori, K. Yamanaka, J. Niki, S. Hiraga, and T. Ogura. 1993. The Escherichia coli FtsH protein is a prokaryotic member of a protein family of putative ATPases involved in membrane functions, cell cycle control, and gene expression. J. Bacteriol. 175:1344–1351.
155. von Gabain, A., J. G. Belasco, J. L. Schottel, A. C. Y. Chang, and S. N. Cohen. 1983. Decay of mRNA in Escherichia coli: investigation of the fate of specific segments of transcripts. Proc. Natl. Acad. Sci. USA 80:653–657.
156. Xu, F., S. Lin-Chao, and S. N. Cohen. 1993. The Escherichia coli pcnB gene promotes adenylylation of antisense RNAI of ColE1-type plasmids in vivo and degradation of RNAI decay intermediates. Proc. Natl. Acad. Sci. USA 90:6756–6760.
157. Yarchuk, O., N. Jacques, J. Guillerez, and M. Dreyfus. 1992. Interdependence of translation, transcription and mRNA degradation in the lacZ gene. J. Mol. Biol. 226:581–596.
158. Young, R. A., and J. A. Steitz. 1978. Complementary sequences 1700 nucleotides apart from a ribonuclease III cleavage site in Escherichia coli ribosomal precursor RNA. Proc. Natl. Acad. Sci. USA 75:3593–3597.
159. Zhu, L., T. Gangopadhyay, K. P. Padmanabha, and M. P. Deutscher. 1990. Escherichia coli rna gene encoding RNase I: cloning, overexpression, subcellular distribution of the enzyme, and use of an rna deletion to identify additional RNases. J. Bacteriol. 172:3146–3151.
160. Zilhao, R., L. Camelo, and C. M. Arraiano. 1993. DNA sequencing and expression of the gene rnb encoding Escherichia coli ribonuclease II. Mol. Microbiol. 8:43–51.
Chapter56
