The Stringent Response
Chapter
92
M. CASHEL, D. R. GENTRY, V. J. HERNANDEZ, and D. VINELLA
The stringent response is a pleiotropic physiological response elicited by a failure of the capacity for tRNA aminoacylation to keep up with the demands of protein synthesis. Experimentally, this response can be provoked either by limiting the availability of amino acids or by limiting the ability to aminoacylate tRNA even in the presence of abundant cognate amino acids. Many features of the stringent response behave as if they are mediated by accumulation of ppGpp, an abbreviation for a guanine nucleotide derivative of GDP that possessed a pyrophosphate group on the ribose 3' hydroxyl, i.e., guanosine 3',5'-bis(diphosphate). Often, accumulation of ppGpp is accompanied by parallel behavior of the equivalent analog of GTP, pppGpp. Collectively, pppGpp and ppGpp are called (p)ppGpp. The accumulation of (p)ppGpp can also be provoked by nutritional or other stress conditions in addition to a deficiency of aminoacyl-tRNA. This chapter will summarize our current understanding of the stringent response, more general effects on the regulation of (p)ppGpp metabolism, and regulatory effects thought to be mediated by (p)ppGpp.
Historically, the genetic evidence that the stringent response reflected the operation of a single mechanism arose from the ability of mutants at one locus in either Escherichia coli or Salmonella typhimurium (official designation, Salmonella enterica serovar Typhimurium) to alter the stringent response (6, 37, 443). This gene was originally called the RNA control (RC) gene because the first measured phenotypic feature of the stringent response was rapid inhibition of accumulation of stable RNA (rRNA and tRNA) accumulation but not DNA. The regulatory relationship between amino acid availability and RNA accumulation was termed stringent control and was of interest because there was no known relationship between amino acid and RNA metabolism (37, 408). The mutant response consisted of continued stable RNA accumulation that only slowly, but progressively, diminished during amino acid deprivation. The mutant RNA accumulation response to amino acid availability was considerably "relaxed" in comparison with wild-type behavior (443). Some inhibitors of protein synthesis could also relax the stringent dependence of RNA accumulation on amino acids in wild-type strains (246). Subsequently (121), the wild-type RCstr and mutant RCrel alleles were designated relA + and relA, respectively, and the mutant response was called a relaxed response.
Comparisons of otherwise isogenic relA mutant strains with wild-type counterparts during amino acid deprivation has led to an appreciation that regulation of stable RNA accumulation is but one of a large, and still growing, number of relA-dependent changes in cell physiology and gene expression. Studies with conditional aminoacyl-tRNA synthetase mutants revealed that the signal for the stringent response involved charging of tRNA rather than the free amino acids themselves (321, 322). Comparisons of the effects of relA mutants were extended to include limitations for other nutrients (such as sources of carbon, nitrogen, and phosphate) as well as changes in composition of growth media allowing different rates of steady-state growth (see below). These studies led to many examples of relA-independent phenotypic alterations that were confusing because they appeared to parallel those of amino acid-starved cells as well as phenotypes that were starvation stress specific.
The rapid onset of the stringent response was correlated with the accumulation of (p)ppGpp (54). Demonstrations that the mutant relA gene abolished (p)ppGpp accumulation during amino acid limitation raised the possibility that (p)ppGpp might be an effector of the stringent response and the conjecture that elongating ribosomes might synthesize (p)ppGpp in a relA-dependent idling reaction when movement was stalled for lack of cognate aminoacyl-tRNA; pppGpp and ppGpp were then called "magic spots" (58). Discovery and characterization of in vitro (p)ppGpp synthetic activity of the relA gene product, then called the stringent factor, defined a reaction neatly accounting for many features of relA function in vivo and therefore went far to validate the role of relA in the stringent response (183, 184, 185). Mutations, such as relC and probably relB, that apparently interfere with the mechanism of activation of relA (p)ppGpp synthetic activity have been found.
A second mechanism for regulating ppGpp accumulation was discovered as a response to energy source availability occurring by inhibition of the (p)ppGpp degradase encoded by the spoT gene (250, 440). Furthermore, strains mutant in relA, relC, and spoT have been exploited to demonstrate that the defect in (p)ppGpp metabolism was responsible for the stringent RNA control response whether or not a wild-type RelA protein was present (115).
The simple explanation that (p)ppGpp was the inhibitor that mediated the stringent response was quickly challenged. A major effort was made to correct for nucleotide pool-specific activity changes during the stringent response and accurately assess effects on RNA synthesis because ppGpp levels were found to inhibit nucleobase uptake (140, 267). Levels of (p)ppGpp were also found to be inversely correlated with balanced growth rates, as if (p)ppGpp might be an effector of growth rate control, again an early simple view that is still an unsettled question.
Early studies of the relA1 gene mutants led to a strong suspicion that an additional source of (p)ppGpp synthesis existed in addition to the relA-dependent mechanism (14, 119). The construction of a true null allele of relA confirmed this suspicion (303). Surprisingly, deletion of spoT along with deletion of relA apparently made cells devoid of (p)ppGpp (510), a state designated (p)ppGpp0. Although (p)ppGpp0 strains have themselves a pleiotropic phenotype, they seem useful for assessing regulatory contributions of (p)ppGpp. Manipulation of relA and spoT genes has also led to the ability to manipulate (p)ppGpp concentrations without nutritional stress (411, 414). These constructions can also provide evidence complementary to that of (p)ppGpp0 strains as to regulatory involvement of (p)ppGpp (143). They can also be exploited to isolate mutants with specifically altered (p)ppGpp regulatory responses (455).
There is hope that mutants allowing manipulation of (p)ppGpp metabolism can someday help to provide a basis for agreement among the disparate views regarding the operation of the stringent response as well as more general regulatory effects of (p)ppGpp. The literature contains many reports of the lack of a correlation between (p)ppGpp and the operation of one or another phenotypic feature of the stringent response, often suggesting alternate regulators or regulatory mechanisms. Subsequent work has revealed the basis for the anomaly only in some instances. In other instances, further work on genes or putative regulator compounds apparently has not been pursued.
Despite conflicting views, current evidence suggests to us that many of the regulatory effects observed during the stringent response are related to variations in (p)ppGpp concentration. Sources of support for this bias are recurrent observations that mutations that alter (p)ppGpp metabolism are found to alter the stringent response and vice versa. It is notable that mutants that convincingly disrupt this linkage have not been found so far. For reviews on various aspects of the stringent response and related topics, see references 39, 49, 56, 57, 59, 77, 116, 137, 145, 155, 206, 213, 292, 293, 313, 322, 331, 337, 338, 383, 494, and 525.
A schematic summary of the reactions involved in (p)ppGpp metabolism and the genes encoding these enzymes is presented in Fig. 1.
Mutants of the relA [(p)ppGpp synthetase I (PSI)] gene fail to accumulate (p)ppGpp during aminoacyl-tRNA limitation. Studies of relaxed mutants and characterization of the relA gene product have played a key role in defining the stringent response.
Characteristics of the In Vitro (p)ppGpp Synthetic Reaction.
A number of laboratories purified a source of ribosome-associated (p)ppGpp synthetic activity in vitro that was altered in relA mutants (70, 184, 185, 363). The mechanism of this reaction was elegantly defined by Haseltine and Block (184), using phage R17 RNA as message and aligning ribosomes in the presence of stringent factor, initiation factors, and fMet-tRNAMet at known translational initiation sites for phage coat protein (fMet Ala Ser) and synthetase (fMet Ser Lys) genes. Under these conditions, about half of the ribosomes were shown to be bound in the P-site by measuring puromycin release, thereby aligning Ala and Ser codons in the A (acceptor)-site. Measurements of in vitro (p)ppGpp synthesis indicated that this alignment of ribosomes was inactive by itself but was necessary to render uncharged tRNAAla or uncharged tRNASer active as a cofactor for the (p)ppGpp synthetic reaction. Charged Ala-tRNAAla and Ser-tRNASer species were not stimulatory when added alone or with the corresponding uncharged tRNA species; however, addition of charged tRNAAla and tRNASer accompanied with EF-Tu and EF-Ts abolished the stimulation of (p)ppGpp synthesis by the uncharged tRNA counterparts. Instead, uncharged tRNALys became able to stimulate (p)ppGpp synthesis. The clear interpretation made was that ternary complexes (of charged tRNA, EF-Tu, and GTP) were bound to the A-sites, displacing bound uncharged tRNA and allowing translocation to the next unique serine codon lacking a charged cognate tRNA. This event simultaneously abolishes the stimulatory activity of uncharged tRNAAla and tRNASer and changes the uncharged tRNA cofactor requirement to one requiring tRNALys.
The deduced requirements for the RelA-dependent (p)ppGpp synthetic reaction were for ribosomes paused during elongation at a "hungry codon" and for nonenzymatic binding of cognate uncharged tRNA at acceptor sites of such a codon. Initiation and elongation factors (IF1 to -3, EF-Tu, EF-Ts, and EF-G) were not directly required for the reactions apart from their roles in ribosomal alignment or realignments at hungry codons. Ribosomal movement per se is not thought to be required for the (p)ppGpp synthesis; it has been proposed that each cycle of synthesis of (p)ppGpp is accompanied by a round of uncharged tRNA release (382).
The stimulatory effect of uncharged tRNAPhe for (p)ppGpp synthesis in a poly(U)-containing ribosomal reaction can be abolished by periodic acid oxidation of the cis-glycols on the 3' terminus of the tRNA, suggesting that this region of the tRNA may be recognized by the ribosome or by the RelA protein (184). Initial reports (106, 386) that the ribothymidine-pseudo(U-C-G) region conserved in loop IV of tRNA is an important recognition site for (p)ppGpp synthesis because a ribothymidine-pseudo(U-C-G) tetranucleotide could substitute for uncharged tRNA in the reaction have been questioned (67). The mRNA requirements in the reaction are readily satisfied by synthetic polymers (70, 184, 345, 363) but not by binding of single codons (386).
The RelA-mediated synthetic reaction itself is an ATP:GTP(GDP) pyrophosphoryl group transfer of the β,γ-phosphates from the ATP donor to the ribose 3' hydroxyl of GDP or GTP acceptor nucleotides; limited donor specificity is extended only to dATP among the eight common ribo- and deoxyribonucleoside triphosphates, and acceptor specificity includes ITP but not pyrimidine nucleotides, deoxypurine nucleotides, or ATP (70, 276, 447, 450). The pyrophosphoryl group transfer to the ppGpp ribose 3' hydroxyl was localized enzymatically by digestion with zinc-activated yeast pyrophosphatase to yield pGp followed by treatment with 3'-specific phosphatase (450). This structural assignment for ppGpp has been verified by 13C nuclear magnetic resonance spectroscopy (373) and is consistent with both the resistance of ppGpp to periodate oxidation and its inability to complex with borate (60). After pyrophosphoryl transfer to GDP, labeled γ-phosphate of the ATP donor is found on the 3' β position of the product ppGpp*, as judged by release of labeled Pi and unlabeled ppGp after mild hydrolysis for 30 min at 37°C in either 1 M HCl or 0.3 M KOH (60, 447). The Km values for GTP and GDP are equal at about 0.5 mM and within the physiological range for cellular GTP, but not GDP, pool values (70, 362); this suggestion that pppGpp is the most likely in vivo product has been verified with assays of kinetics of the timing of appearance of pppGpp and ppGpp (62, 115, 233, 269, 498).
Given the complex reaction requirements for the RelA- dependent ribosomal reaction, it was important to localize (p)ppGpp synthetic activity to the RelA enzyme itself. This was accomplished by demonstrating that the (p)ppGpp synthetic activity of purified preparations of RelA can be activated by methanol even in the absence of ribosomes, mRNA, or tRNA (361, 456).
The Response to Variations of Charged and Uncharged tRNA In Vivo. Considerations of the properties of the in vitro ribosome-dependent (p)ppGpp synthetic reaction catalyzed by the RelA protein together with deductions from the ability of conditional tRNA synthetase mutants to provoke the stringent control response (322) led to plausible, but probably oversimplified, models of the cellular response to varying levels of charged and uncharged tRNA. Charged/uncharged tRNA ratios can be continuously sensed by the demands of the active population of mRNA codons for translation. If there is no demand, there is no response. For example, if cellular protein synthesis becomes dominated by an abundant mRNA species that does not code for a particular amino acid, as it does during phage R17 infection, then the stringent response is not provoked by starvation for that amino acid (496). Conversely, if aminoacylation of any tRNA cannot keep up with the demands of protein synthesis, then transient stalling of ribosomal elongation occurs. If the stalled ribosome bears the RelA protein, it can be thought to undergo cycles of cognate uncharged tRNA binding, release, and (p)ppGpp synthesis (382). When the codon exposed in the A-site of a paused ribosome is bound to a cognate aminoacyl-tRNA as an EF-Tu–GTP ternary complex, ribosomal translocation occurs (232) and quenches the idling reaction. The ribosomal idling reaction could also be quenched by binding and translocating with a noncognate tRNA whose anticodon is a close match to the hungry codon, by a frameshift caused by an aminoacyl-tRNA matching out of frame, or by disengagement of the ribosome from the mRNA. Early studies of relationships between tRNA charging, inhibition of protein synthesis, and ppGpp synthesis raised questions as to the validity of this simple model (365, 366; reviewed by Cozzone [77]). For example, charged/uncharged tRNA ratios may remain unexpectedly high despite amino acid starvation, isoacceptor species may show unexpected patterns of aminoacylation during normal growth as well as during amino acid starvation (518), and certain tRNA species can be protected from periodic acid oxidation by something distinct from amino acids whose occurrence depends on starvation as well as the state of the relA gene (517).
In a series of studies (393, 394) the cellular abundance of tRNATrp was manipulated by an isopropylthiogalactopyranoside (IPTG)-driven promoter for its structural gene, and the percent charging was independently modified by varying the concentration of tryptophan in the medium of cells that had a tryptophanyl-tRNA synthetase mutant with a lowered affinity for tryptophan. It was found that relA-dependent (p)ppGpp synthesis responds to the charged/uncharged tRNATrp ratio rather than the absolute concentration of either tRNA species. In contrast, the amplitude of the (p)ppGpp synthetic response was determined by the abundance of the uncharged tRNATrp when tryptophan was the limiting amino acid but not when limitation of a different amino acid elicited (p)ppGpp synthesis. Protein synthesis, measured by the incorporation of labeled tryptophan, lysine, or cysteine, was unexpectedly found to be inhibited by low-level increases in uncharged cognate tRNA without eliciting (p)ppGpp synthesis, whereas 5- to 10-fold excess of charged over uncharged tRNA was necessary to trigger relA-dependent (p)ppGpp synthesis (394). Arguing that uncharged tRNA also has relA-independent effects on protein synthesis, Goldman and Jakubowski (155) proposed a model in which small increases in uncharged tRNA might compete with charged tRNA ternary complex binding to the 30S component of the ribosomal A-site whereas uncharged tRNA in excess can also interact with the 50S component of the A-site and trigger the stringent response. Alternatively, release of uncharged tRNA from the E-site after translocation might be impaired by an excess of cognate uncharged tRNA (394). A notable conclusion from these studies was also that relaxed strains generally displayed lower rates of protein synthesis than an isogenic stringent strain without involvement of ppGpp because the difference was observed even at uncharged tRNA levels insufficient to trigger ppGpp synthesis. In all, these considerations suggest that it might be informative to explore more precisely defined features of 30S and 50S components of A- and E-site binding using ribosomes saturated with bound RelA protein (311). Recently it was found that RNA secondary structures can form with tRNA that distinguish between charged and uncharged tRNA and thereby regulate terminator readthrough (167). It seems plausible to consider the possibility of analogous structures responsible for uncharged tRNA interactions on the ribosome using the codon as the one binding site and a T-box analog on the ribosome for sensing the 3' end of uncharged tRNA (167). A lattice of this sort might confer an activating conformational change in the RelA protein.
Inhibitors of (p)ppGpp Synthesis In Vitro and In Vivo.
Effects of a variety of inhibitors on RelA-dependent (p)ppGpp synthetase as well as on accumulation of (p)ppGpp in cells during the stringent response have been tested. The methanol-activated reaction is sensitive selectively to tetracycline, chlortetracycline, and oxytetracycline at effective concentrations too low to be explained by their ability to chelate Mg2+ ions (421, 456). Cellular studies with conditional tRNA synthetase mutants confirm the tetracycline sensitivity of (p)ppGpp synthesis during the stringent response (233). The ribosome-dependent activity, but not the methanol-activated protein, is sensitive to thiostrepton (see discussion of relC). Neither reaction is sensitive to low concentrations of chloramphenicol or to concentrations of fusidic acid sufficient to block translocation in vitro (70, 184, 451).
In contrast, in vivo (p)ppGpp accumulation during the stringent response is inhibited by a variety of antibiotics, including chloramphenicol. An explanation for the difference between in vitro and in vivo sensitivities is that any source of reduction in cellular protein synthesis rates reduces the demand for aminoacyl-tRNA, which leads to restoration of normal ratios of charged to uncharged tRNA (see review in reference 59). It can be argued that a source of free amino acids under these conditions can arise from proteolysis, itself induced during the stringent response (Table 1).
Table 1Pleiotropic effects of the stringent responsea |
Transcription of relA appears to be regulated. The addition of chloramphenicol to reverse the stringent response invoked by serine hydroxamate brought about a sixfold reduction in relA mRNA measured by hybridization (31). A study employing lacZ as a reporter failed to provide evidence for transcriptional modulation, but this may have been due to the absence of additional upstream promoters now known to exist (G. Glaser, personal communication).
As the stringent response progresses past initial stages, severe inhibition of (p)ppGpp synthesis probably occurs, particularly in spoT mutants (115). The source of this inhibition is uncertain. A feedback effect due to (p)ppGpp itself is unlikely, judging from in vitro behavior of the RelA reaction (362, 447). Instead, the inhibition of protein synthesis that accompanies the stringent response (251, 343) as well as even (p)ppGpp induction without starvation (501) could reduce the demand for aminoacyl-tRNA (432).
Relaxed Phenotypes.
Relaxed mutants display many amino acid starvation-dependent phenotypes in addition to differences in RNA accumulation (6, 7) (Table 1). A large number of relaxed mutants with RNA accumulation behavior ranging from slightly to strongly relaxed were isolated and characterized by Fiil and Friesen (112, 113). Mutant isolation was accomplished by exploiting a still incompletely understood property of the original relA1 mutant that reversal of a 3-h amino acid starvation results in a long lag before growth resumes, allowing penicillin enrichment of nongrowing mutant cells. Relaxed mutants have increased permeability to glutamate (51) and increased sensitivity to numerous inhibitors (356, 444). Two different plate growth tests are available for relA mutants. Both tests are based on positive regulatory effects of ppGpp accumulation on amino acid biosynthesis (Ilv and His; Table 1). One test, the SMG plate test, involves mutant sensitivity to glucose minimal medium supplemented with serine, methionine, and glycine (SMG) (79); the second test, the AT plate test, involves sensitivity to 3-amino-1,2,4-triazole (AT), an inhibitor of histidine biosynthesis (401).
relA
and relA1 Gene Sequences.
The relA gene is located at 59.2 min and was first cloned in ColE1 plasmids, where it was noticed that its presence in multicopy conferred a 10-fold overexpression of RelA, elevated (p)ppGpp levels, and slow growth (72, 123). The relA gene is an interior gene within a larger operon and is flanked upstream by a tRNA methylase homolog (Glaser, personal communication) and downstream by the mazE and -F genes. The mazE and mazF genes have been called chpAI and chpAK, respectively, and deduced to be homologs of pem genes involved in stable maintenance of plasmid R100, a plasmid addiction system (295, 301; Glaser, personal communication). The relA structural gene itself consists of 743 codons which encode a protein of about 84 kDa. The relA open reading frame terminates with an amber codon, which when suppressed extends the protein by 27 amino acids, thereby inactivating it. This behavior is notable because it verifies, at the sequence level, early observations that a multicopy plasmid bearing a glutamine-inserting Su+ 7 gene inactivated the RelA protein (48, 516) by lengthening the protein (47).
The relA1 mutant allele consists of an amino terminal IS2 insertion between codons 85 and 86 (303). The mutant allele possesses weak residual activity, as indicated by multicopy plasmids bearing the relA1 mutant conferring a nearly wild type phenocopy with respect to ppGpp levels during the stringent response and to plate tests for relA function. The basis for this activity was dissected to reveal that the IS2 insertion sequence encoded a downstream ribosomal initiation site which allowed a protein fusion of nine amino acids encoded by IS2 to the downstream portion of the relA open reading frame, thereby generating two RelA protein fragments (α and β) capable of complementing ppGpp synthetic activity in trans (303). Plasmids expressing only the carboxy-terminal portion of the RelA protein in a relA + host interfere with ppGpp accumulation during the stringent response and reverse the AT resistance as if the fragment functions as a trans-dominant inhibitor that competes with RelA protein binding to ribosomes (Glaser, personal communication).
relA
Null Behavior.
An allele of relA consisting of a deletion and kan insertion (designated relA251) was constructed and recombined into the chromosome to provide a certain null allele (303). Characterization of ppGpp accumulation during the stringent response of cells bearing the relA251 allele showed no accumulation of ppGpp, as for relA1 strains, reinforcing previous estimates that the relA1 mutation is indeed a severely defective allele. An interesting feature of relaxed relA1 mutants is that basal levels of (p)ppGpp drop precipitously during the relaxed response to amino acid starvation, rather than remaining constant, suggesting that the relaxed response is more complex than simply the absence of the stringent response (253, 254, 418, 444). This effect on ppGpp basal levels can be mimicked by chloramphenicol addition to relA1 strains (134, 253, 254). Similar studies of ppGpp basal level changes in relA251 strains have not yet been done, but such studies could indicate whether residual relA-dependent activity is operative under these conditions. The weak residual ppGpp synthetic activity of the relA1 mutant in single copy can be amplified to give a demonstrable phenotypic effect only when (p)ppGpp degradation is severely compromised, as shown by prototrophy of a relA1 Δ spoT mutant (see below). The accumulation of ppGpp displayed by strains bearing a null allele in relA has provided an unequivocal confirmation of the existence of a second source of (p)ppGpp (303). Fehr and Richter have designated the relA-dependent source of (p)ppGpp as PSI and the alternative source as PSII (110).
Induction of RelA and Truncated RelA Proteins.
Expression of the relA gene, as well as an amino-terminal relA ' portion truncated at codon 455, driven by a Ptac promoter, allows induction of expression of either the full-length or truncated RelA' protein (414). In otherwise growing cells, induction of either protein with high levels of IPTG (200 μM) induces (p)ppGpp accumulation to levels at least as high as those seen during the stringent response. The induced full-length RelA protein responds to amino acid starvation with further enhanced ppGpp synthesis in a ribosome-dependent manner. Thus, a ribosomal protein mutation, such as relC (see below), that interferes with relA-dependent (p)ppGpp synthesis interferes with accumulation of ppGpp when the full-length RelA protein is induced. In contrast, the RelA' protein appears to possess constitutive (p)ppGpp synthetic activity that is independent of ribosomes and does not require uncharged tRNA for activation (418, 446). The metabolically lability of the truncated RelA' fragment has so far precluded assessing its ability to bind to ribosomes in vitro, so it remains uncertain whether it even binds to ribosomes. In summary, overexpression of portions of RelA suggests the existence of physically and functionally distinct domains for (p)ppGpp synthetic activity and for ribosomal binding.
The IPTG-inducible constructs just described have proven particularly useful as a means of assessing regulatory effects of (p)ppGpp accumulation without the usual accompaniment of nutritional stress. For this purpose, induction of the truncated RelA' protein seems preferable (414). Comparisons of growth inhibitory effects as a function of (p)ppGpp concentrations reveal a toxic effect that accompanies overexpression of the full-length RelA protein that is not shown for RelA' overexpression. Possibly, this is because only very low levels of RelA' accumulate because of its metabolic lability.
A question arises as to why (p)ppGpp accumulates at all when the full-length RelA protein accumulates to high levels in cells that are not overtly nutritionally limited for amino acids or aminoacyl-tRNA. One explanation proposed is in terms of the changes in fraction of ribosomes containing bound RelA (414). Immunological assays of the level of RelA in wild-type cells have led to estimates that the fraction of ribosomes bearing the enzyme is low, at about 1% (362). If it is assumed that the majority class ribosomes lacking RelA encounter hungry codons at the same frequency as the rare class of RelA-containing ribosomes, the increase in (p)ppGpp synthesis shown when RelA abundance increases 100-fold could be imagined to simply reflect the titration of ribosomes containing RelA. This explanation assumes that some fixed frequency of hungry codon encounters occurs even under nutritionally adequate conditions and that the response in terms of (p)ppGpp synthesis is simply amplified by this titration of ribosomes with RelA. More complex situations could easily be imagined. Studies in our laboratory (D. R. Gentry and M. Cashel, unpublished data) have reaffirmed an early suggestion (375) that RelA seems preferentially associated with 50S ribosomal subunits rather than 70S ribosomes; this raises possibilities that artificially increasing the abundance of RelA could have effects quite distinct from those anticipated as strictly limited to elongating ribosomes.
Mutations at the relB locus relax the stringent response about 10 min after an otherwise apparently normal onset provoked by amino acid starvation. Little is known of the details of the mechanism by which persistence of the stringent response is affected.
Several different selection procedures yield mutants in the relB gene (34 min). The first relB mutant (relB1) was isolated on the basis of its prolonged growth lag after reversal of an amino acid starvation too brief (20 to 30 min) to appreciably enrich for relA mutants (263, 264). A very strong, if not exclusive, selection for relB (relB2) mutants occurs when 5-fluorouracil (5FU) is added during a much longer (5-h) amino acid starvation followed by penicillin enrichment a generation after reversal (96). The omission of 5FU and imposition of ampicillin selection a generation after reversal in an otherwise similar protocol enriches for relA but not relB mutants (112). Finally, repeated carbon source starvation of a thymine auxotroph, coupled with thymine starvation, also yields relB (relB3) mutants (315, 316). These three alleles were subsequently renamed relB101 to -103 and transduced into a common background to compare phenotypes (95). These phenotypes consist of a "delayed relaxed" RNA accumulation response which occurs after a transient stringent response to amino acid starvation (95, 264). Growth is transiently inhibited upon reversal of amino acid starvation and is attributed to a presumed protein inhibitor recovered from ribosomal washes (95, 264). Curiously, starvation for phenylalanine does not provoke the delayed relaxed response. All three mutant alleles have a heightened sensitivity to chloramphenicol and sulfacetamide during normal growth. The resistance of relB mutants to 5FU during starvation is not seen during normal growth.
After a brief amino acid starvation, high-salt washes of ribosomes from relB101 strains inhibit in vitro translation of bacteriophage MS2 RNA but not poly(U)-dependent polyphenylalanine synthesis. The inhibitor also blocks the relA-mediated ribosome-dependent (p)ppGpp synthetic reaction (264). Amino acid starvation of relB101 provokes transient accumulation of ppGpp which parallels the transient stringent response, and the disappearance of ppGpp parallels the onset of the delayed relaxed response. A relB-mediated accumulation of an inhibitor of protein synthesis that interferes with ppGpp synthesis as well could account for some of the behavior observed. Effects of ppGpp accumulation could explain some of the selection procedures yielding relB mutants. The long growth lag after a short amino acid starvation exposure for the relB101 mutant could well reflect the persistence of the inhibitor. The transport of 5FU is probably blocked in stringent parental strains by inhibitory effects of ppGpp on pyrimidine transport (Upp; Table 1); the failure of relB102 mutants to accumulate ppGpp can be associated with their ability to take up 5FU and therefore undergo a prolonged growth lag which renders them penicillin resistant. However, the selection procedure for relB103 is not easily explained. Bech et al. (23) have localized the relB gene near the terminus of DNA replication as a 79-codon open reading frame that contains different missense mutations for each of the three relB alleles. Bech et al. (23) suspect that the RelB protein functions as a negative regulator of inhibitor formation, rather than being the inhibitor itself.
Three additional small open reading frames are located downstream of relB; these have been designated relD, relE, and relF even though there is no evidence that they function to affect the stringent response (see GenBank entry ECRELB, accession number X02405). The RelF sequence has been noted as showing a 40% homology to the cell killing factor hok encoded by the plasmid R1 parB system, and overexpression of RelF results in cell lesions that are similar to those accompanying overexpression of hok; however, parB suppresses the destabilizing effects of hok but not relF (145; reviewed in reference 146).
The ribosome dependence of (p)ppGpp synthesis by RelA suggests that there might be mutations in ribosomal components that result in a relaxed phenotype. This turns out to be the case. A mutation in rplK, the structural gene for ribosomal protein L11, was picked up by penicillin enrichment for cells with a long growth lag after prolonged amino acid starvation using a relA + diploid parental strain to avoid reisolation of relA mutants (121). The mutant allele was originally called relC. Ribosomes with the mutant form of RelC (L11) still bind RelA but have only 10% of normal (p)ppGpp synthetic activity in vitro (121). The L11 protein is also implicated in resistance to thiostrepton, an antibiotic that inhibits the ribosomal (p)ppGpp synthetic reaction in vitro (185, 447).
The relC mutants, like relaxed relA mutants (7), are leucine sensitive and show a long lag before growth after undergoing a shift from rich medium to minimal medium containing l-leucine (115). This feature was used to select for extragenic revertants showing rapid outgrowth characteristic of stringent strains which proved a source of spoT mutants, defective in (p)ppGpp degradation (see discussion of spoT).
The relX mutation was first described in 1978 by Pao and Gallant (356), who uncovered a recessive relX mutation that was located between fuc and relA and gave a 10-fold lowering of ppGpp basal levels in relA hosts during growth on glucose. Ordinarily relA mutations have little effect on ppGpp basal levels during exponential growth. Other phenotypes include a relA-independent limited viability after an abrupt shift from 28 to 42°C on prewarmed plates as well as a relA-dependent inviability when plated on minimal medium containing leucine (356). The relX mutant lowered the levels of ppGpp accumulation during a glucose-to-succinate downshift, partially in a relA + host and almost completely in a relA double mutant. A demonstration that relX did not affect rates of ppGpp degradation, measured after adding tetracycline, led Pao and Gallant to propose that relX functioned at the level of lowering ppGpp synthetic rates. Currently, relX can be viewed either as a regulator of residual PSI activity in the relaxed mutants or as a regulator of PSII activity, now believed to be encoded by spoT. The source of the relX mutation was traced to the parental strain CP78 used by Fiil and Friesen (113) to isolate relA mutants. Despite the presence of the relX mutation, temperature sensitivity is not shown in a relaxed mutant of the parental lineage, such as strain CP79, which is therefore deduced to contain an as yet unmapped suppressor of this defect.
The SpoT protein is proposed to be a bifunctional enzyme possessing both (p)ppGpp 3'-pyrophosphohydrolase activity as well as 3'-pyrophosphotransferase activity with GTP as the acceptor. Mutants of spoT have several characteristic defects in (p)ppGpp metabolism: (i) elevated ppGpp basal levels in balanced growth and slower growth rates, (ii) higher induced levels of ppGpp during the stringent response, (iii) slower rates of turnover of ppGpp when the stringent response is reversed, and (iv) a defect in the ability to accumulate pppGpp during the stringent response.
The spo Operon.
The spoT gene (82 min) encodes a 79-kDa protein and shares an operon with four other genes in the order gmk-rpoZ-spoT-spoU-recG. Before identities of these genes were established, the genes were called spoR-S-T-U-V, respectively. There are two (412) or three (142) promoters for the operon. The P1 promoter is located upstream of gmk, which encodes what appears to be the only guanylate kinase gene of E. coli (141). The P2 promoter is embedded in the carboxy-terminal coding region of gmk, while a third possible promoter is located upstream of P2. The rpoZ gene encodes the omega subunit of RNA polymerase, present in nearly stoichiometric amounts with purified enzyme but whose function remains unknown (142). Igarashi et al. (204) have reported that the presence or absence of omega determines stringent control of RNA polymerase transcription by ppGpp with in vitro "mixed template" assays. However, a strain with a deletion of rpoZ retains a normal stringent response with respect to RNA accumulation (144). The spoT gene translation initiation occurs at a UUG codon that occurs seven codons beyond the end of the RpoZ open reading frame (412), verified by determination of the protein sequence (143), and proceeds to give a 702-residue protein. The spoU gene, immediately downstream of spoT, has been recently included in a family of RNA methyltransferase genes (244). The translational start of recG, the last gene in the operon, occurs shortly beyond the end of spoU (231, 282). The recG gene product is involved in the resolution of Holliday structures during recombination (499). Apart from the coparticipation of gmk and spoT in guanine nucleotide metabolism, no common function explaining placement of genes in this operon is apparent to us. It is intriguing, however, that again a (putative) methyltransferase homolog should also exist in another operon involved with (p)ppGpp, namely, just upstream of relA.
Comparison of the deduced protein sequences of SpoT and RelA reveals widespread homology, except for about 50 amino-terminal residues which are not closely related, consisting of 31% identity and an additional 54% conservative replacements (302).
(p)ppGpp 3'-Pyrophosphohydrolase Activity.
The 79-kDa SpoT protein (412) is the major catalyst for ppGpp degradation. Highly purified enzyme removes the 3'-pyrophosphate residue of (p)ppGpp with (GTP)GDP and pyrophosphate as products; the enzyme requires Mn2+ and is further stimulated by Mg2+ (189, 190, 191, 448; Gentry and Cashel, unpublished data). The enzyme will degrade pppGpp and ppGpp with equal facility in vitro (192; Gentry and Cashel, unpublished data), and the conclusion that SpoT will degrade pppGpp has supporting evidence in vivo (431). The spoT-dependent 3'-pyrophosphohydrolase activity will be abbreviated here as (p)ppGppase. The SpoT protein is normally expressed at low levels. Multicopy plasmids have been constructed with different promoters and different translational initiation sites to achieve levels of spoT gene expression ranging from modest 10-fold elevation to very high overexpression. Reported Km values for (p)ppGpp substrates vary from about 0.05 to 0.5 mM (10, 192).
There are physiologically less active spoT-independent routes of ppGpp degradation. An et al. (10) reported that such alternative activities can be resolved from the SpoT enzyme in cell extracts. Also ppGpp degradation has been noted to occur in extracts from a spoT1 relA1 double mutant yielding ppGp, pGpp, and pGp products (189). One among these, ppGp, has been described by Pao and Gallant and called MSIII (357) and has been argued to constitute an alternative candidate to (p)ppGpp for a role in mediating the stringent response.
The occurrence of (p)ppGppase is widespread in bacteria (387); mutants of the spoT genes of both E. coli and S. typhimurium (250, 401) have impairments of ppGpp degradation, which normally occurs with a half-life of about 20 s. Rates of in vivo ppGpp decay are typically measured as first-order rates of ppGpp disappearance after adding low levels of chloramphenicol or tetracycline to cells or by reversing a stringent response provoked by amino acid starvation by restoring amino acid sufficiency. Strong spoT mutants were first noticed as failing to accumulate appreciable pppGpp during the stringent response as well as during steady-state growth; this phenotype, termed spotless, provided the mnemonic for the name of the gene (250). It was later shown that spoT mutants do not really completely lack pppGpp but instead display an initial burst of pppGpp accumulation during the stringent response, which largely disappears after a few minutes (62, 89, 115) (see discussion of gpp). The spoT mutants also accumulate higher than normal levels of ppGpp and display slower exponential growth in inverse proportion to steady-state ppGpp levels achieved (62, 89, 115, 195, 411).
A major role for spoT in ppGpp decay is deduced from the overaccumulation of ppGpp in mutants in vivo and defects in ppGppase activity in vitro. However, there are indications of a more complex role for spoT function than simply hydrolysis proportional to enzyme concentration. A pGA1 plasmid spoT gene that is wild type with respect to both promoters and the translational initiation site gives about 10-fold overexpression of the SpoT protein yet does not affect in vivo ppGpp degradation rates (10). In vitro (p)ppGppase activities have been reported as stimulated by ATP (191, 449), whereas cellular ppGpp degradation is inhibited by respiration uncouplers that implicate the integrity of the transmembrane proton gradient in a manner separable from effects on ATP levels (456). Permeabilized cells show a strong (4- to 10-fold) dependence of (p)ppGppase on charged tRNA, using conditional valS and gluS synthetase mutants (377). Richter has reported that uncharged tRNA gives a twofold inhibition of (p)ppGppase in vitro and that this inhibition is reversed by aminoacylation or periodic acid oxidation of tRNA (385). A charged tRNA requirement could explain the ATP-dependent, heat-stable, low-molecular-weight activator of (p)ppGppase reported by Sy (449). The putative involvement of uncharged tRNA as an inhibitor of (p)ppGppase raises the question of why ppGpp degradation rates are not inhibited following amino acid starvation or when interfering with tRNA aminoacylation gives high levels of uncharged tRNA. This effect of uncharged tRNA could well be obscured when ppGpp decay rates are routinely measured with chloramphenicol or by restoring amino acid sufficiency. Perhaps ppGpp degradation rates are inhibited as claimed from early studies with a conditional valyl-tRNA synthetase mutant (114). Early kinetic analyses of rates of ppGpp synthesis, rates of ppGpp decay, and accumulated levels of ppGpp seem internally consistent without invoking this complication (134).
Although the SpoT-(p)ppGppase reaction readily occurs with the soluble protein, the SpoT protein was first described as associated with a ribosomal fraction but active when released by high salt and further purified (190, 448). Later, SpoT was proposed to be membrane associated (449, 456). Recent results (Gentry and Cashel, unpublished data) suggest that the enzyme is neither ribosome or membrane associated but cytosolic instead, as claimed by An et al. (10).
Studies of inhibition of ppGpp degradation in vivo are generally poorly understood with the exception of effects of Mn2+ chelators on the reaction, which is Mn2+ dependent in vitro. Divalent ion chelators, such as picolinic acid or 1,10-phenanthroline, inhibit cellular ppGpp degradation, and these effects are reversed by Mn2+ addition (25, 220).
Carbon and energy source starvation of spoT + strains gives a phenocopy of spoT mutants with respect to impairments of ppGpp degradation (115, 134, 250, 440) by a still elusive mechanism. Attempts to explain this effect of energy source starvation as exerted by changes in energy charge by providing varying ratios of AMP, ADP, and ATP have not yielded inhibition of (p)ppGppase in vitro (Gentry and Cashel, unpublished data).
A variety of seemingly unrelated drugs interfere with ppGpp degradation in vivo: tetracycline (reversible by Mn2+ addition), chlortetracycline, thiostrepton (189, 384; but see reference 448), polymyxin B (75), gramicidin (75), and the morphine analog levallorphan (36). In addition, ppGpp degradation is impaired by osmotic shock (182), heat shock (135), inhibition of fatty acid synthesis (417), uncouplers of oxidative phosphorylation (456), and long-chain alcohols (305). Richter (384) has confirmed inhibition by chlortetracycline, thiostrepton, and levallorphan in vitro (but see references 448 and 449). Both levallorphan inhibition and thiostrepton inhibition in vitro require much higher concentrations of the drugs than are required in vivo, suggesting that the cellular inhibition mechanism is indirect. The one feature that we can imagine as shared by all of these treatments is that they can alter membranes, possibly revealing the presence of a membrane sensor-like factor that somehow responds with inhibition of SpoT-mediated (p)ppGppase. Also, it could be imagined that exhaustion of energy sources somehow affects Mn2+ availability, which, in turn, regulates (p)ppGppase activity. New metal transport mechanisms might be involved (482).
Deduced (p)ppGpp Synthetase Activity.
A genetic search for PSII, the source of (p)ppGpp synthesized in strains deleted for relA, was initiated by deleting the spoT gene with the expectation that very slow growth, or lethality, would ensue. However, the surprising finding was that deleting the spoT gene eliminated all detectable cellular (p)ppGpp (510). The effect of the deletion was judged not to be a polar effect on some downstream gene because ppGpp synthesis could be restored by a plasmid bearing only the spoT gene (510).
Another search for the source of relA-independent (p)ppGpp synthetic activity was carried out with a relA1 spoT + strain, with mutants selected as displaying increased reporter activity of a ribosomal rrnB P1 promoter-lac fusion (196). This search also revealed mutants in spoT which were initially thought to possess abnormally high (p)ppGppase activities and then realized to have lowered PSII activity. Two of these alleles were characterized by PCR analysis as consisting of different insertions in the central region of the spoT gene that had lowered but not absent ppGpp-forming activity. When Δ relA was substituted for the relA1 in the spoT mutant strains to eliminate residual PSI activity, both were able to synthesize ppGpp at approximately half-normal rates and grow on minimal glucose medium (196). Other mutants obtained in this selection behave as if they are devoid of PSII activity, like spoT deletions (V. J. Hernandez and M. Cashel, unpublished data).
The genetic evidence cited above can be interpreted as indicating that spoT encodes a (p)ppGpp synthetic activity, implying that the SpoT protein is bifunctional and capable of both synthesis and degradation. Alternatively, some other gene responsible for synthetic activity that is somehow repressed by deletion of spoT could exist. So far, it has not been possible to demonstrate (p)ppGpp synthesis with purified SpoT protein preparations (Gentry and Cashel, unpublished data), and therefore the proposal that the SpoT protein is a bifunctional enzyme is not firmly established.
The extensive homology between RelA and SpoT protein sequences (302) has been taken as supporting the interpretation of a bifunctional enzyme. Coincidentally, this argument raises the question that RelA might have heretofore unsuspected (p)ppGppase activity; RelA exhibits micro-reversibility but requires substrate levels unlikely to be achieved in cells (362, 447). Mapping of mutant spoT alleles defective in (p)ppGppase, but not synthetase, reveals clustering in the N-terminal portion of the gene (412). The behavior of multicopy subclones of the N-terminal region showing degradation but not synthetic activity is consistent with this localization (Gentry and Cashel, unpublished data). Short amino-terminal deletions of the spoT gene in multicopy seem to retain (p)ppGpp synthetic activity but not degrading activity (Gentry and Cashel, unpublished data). The portions of the gene with degradation activity and the portions with synthetic activity are physically overlapping fragments; it is possible that a single catalytic site can be altered to favor synthesis or degradation by flanking portions of the protein.
Selection of spoT Mutants.
Laffler and Gallant (250) noted that elevated ppGpp basal levels in the archetype spoT1 mutant were associated with slower steady-state growth rates and argued that this might have provided a positive selection for the archetype spontaneous relA1 mutation that lowered ppGpp basal levels. This situation would account for the simultaneous occurrence of both mutations in the same laboratory strain. The ability of relA alleles to partially suppress both the slow-growth and elevated ppGpp basal levels of spoT mutants has been exploited to derive new relA mutants (401). Conversely, both relA and relC mutants can be phenotypically suppressed by spoT mutants (115, 250, 401, 411). Sources of selection initially involved combinations of leucine resistance, rapid recovery after reversal of a prolonged amino acid starvation, and resistance to low concentrations of rifampin. More recently, selections for spoT mutants in relA hosts were based on resistance to AT, an inhibitor of histidine biosynthesis whose effects are overcome by overexpression of the his operon due to elevated ppGpp levels (401, 411). These observations suggest that the spoT gene product plays a significant role in maintaining basal levels of ppGpp during growth (see below). It follows that genetic instability might be expected in laboratory strains containing either a relA or a spoT mutation or any mutation that elevates ppGpp basal levels. Perhaps for this reason, it is not uncommon to find strains whose relA and spoT phenotypes deviate from pedigree listings as well as to find intermediate but balanced levels of relA and spoT defects in supposedly wild-type strains.
A series of spoT mutant alleles (spoT202 to -204) found to be nonviable in a relA + background was selected as AT resistant with progressively severe growth impairments and higher ppGpp basal levels (411). This set has been useful for obtaining systematic elevation of (p)ppGpp basal levels during steady-state growth (195, 520).
The guanosine pentaphosphatase (pppGpp γ-phosphohydrolase) encoded by the gpp gene at 84 min was discovered by Somerville and Ahmed (431) to be a major determinant of the abundance of pppGpp, relative to ppGpp, during the stringent response. Mutants at the gpp locus were identified by direct screening of nitrosoguanidine-mutagenized cells for pppGpp hyperabundance after amino acid starvation. Mutants typified by gpp-1 were characterized as differing from the wild type by showing a biochemical phenotype consisting of a three- to fourfold elevation of pppGpp to reach plateau levels roughly equivalent to those of ppGpp; since levels of ppGpp were themselves unchanged in the gpp-1 mutant, total (p)ppGpp abundance increased about twofold (25, 26, 431). Somerville and Ahmed proceeded to demonstrate that the mutant allele was recessive, that (p)ppGpp synthesis and pppGpp/ppGpp ratios were unaffected in ribosome-dependent (p)ppGpp synthetic reactions, and that relA1 was epistatic to the expression of the gpp mutant phenotype (431).
Mutant and parental cell S100 extracts were fractionated on DEAE-Sephadex columns and assayed for pppGppase activity in the presence and absence of ribosomes; five peaks were resolved, two of which were ribosome dependent. In the mutant, two peaks (one major and one minor) were missing; both displayed ribosome-independent activity in the parental strain (431). The ability to convert pppGpp to ppGpp in vitro has been independently demonstrated for a number of enzymes, including IF2, EF-Tu, the uncoupled activity of EF-G, and polyphosphate phosphatase (173, 234). The major activity fraction affected by gpp-1 gave ppGpp as a reaction product as well as the absence of hydrolytic activity toward GTP, ATP, UTP, or GDP (431). Later purification of the Gpp protein revealed that it also failed to hydrolyze pppApp but would attack pppGp (179). The Gpp protein will also hydrolyze oligoribonucleotides bearing guanosine 5'-triphosphate termini so long as they possess a 3'-phosphate, i.e., pppGpNp but not pppGpN is hydrolyzed. Hydrolytic activity toward otherwise "permissive" oligoribonucleotides diminishes rapidly as chain length increases (179).
Belitskii and Shakulov mapped the gpp gene more precisely to the region between trxA and rep and constructed a gpp::kan insertion and a deletion allele extending from gpp to ilvC (24). Sequencing has led to a deduced size of the Gpp protein of 49 kDa (234; M. Kalman, H. Murphy, and M. Cashel, unpublished data). The gpp-1 allele has been sequenced and found to consist of a missense mutation (Kalman et al., unpublished data). Overproduction of the Gpp protein results in abnormally low, often undetectable pppGpp levels during the stringent response (25, 26; Kalman et al., unpublished data). The gpp gene is the downstream member of a two-gene operon, with the upstream gene found to be rhlB, a DEAD box putative ATP-dependent RNA helicase (230).
Recently, a relationship between Gpp and polyphosphate degradation has emerged. The polyphosphate operon has been found to consist of two genes with polyphosphate kinase (ppk) upstream (2) and polyphosphate exonuclease (ppx) downstream (3). A kan element insertion in ppk results in measurably lowered, but not absent, Ppx activity in vitro, presumably due to polar effects of the insertion (78). Purification of the source of residual Ppx activity and comparison of peptide sequences of tryptic digestion fragments revealed that this source of activity was identical to Gpp (234). Deduced protein sequence comparisons reveal homology between Gpp and Ppx (381). Functional comparisons reveal that Gpp indeed possesses polyphosphate phosphatase activity as well as that purified Ppx possesses pppGpp γ-phosphohydrolase activity (234). These observations are suggestive of a some connection between (p)ppGpp metabolism and polyphosphate metabolism, but the relationship is still uncertain. The ppk::kan insertion mutant as well as a Δ ppk-x operon deletion do not affect the pppGpp/ppGpp ratios during the stringent response in a gpp + or Δ gpp host; overexpression of Ppx only marginally affects these ratios in a Δ gpp host (M. Cashel and H. Murphy, unpublished data).
The behavior of gpp mutants has provided information about (p)ppGpp metabolism. First, since pppGpp, not ppGpp, is the major synthetic product during the stringent response, the existence of some ppGpp in a Δ gpp mutant indicates that there are alternative routes to degrade pppGpp that are minor compared with the Gpp reaction; among possible sources are IF1 and EF-G (see above; 24, 431). Second, inhibiting SpoT-encoded (p)ppGppase in a Δ gpp mutant by carbon source starvation or by adding Mn2+ chelators inhibits decay of both pppGpp and ppGpp (24, 25). This in vivo demonstration is consistent with the in vitro behavior of the SpoT enzyme (see above). Third, if Δ relA Δ gpp mutants are starved for glucose, pppGpp is found to accumulate; this finding suggests that the PSII activity of SpoT involves a pyrophosphoryl transfer to GTP, like the RelA synthetic reaction (Cashel and Murphy, unpublished data), and is the basis of listing GTP as the substrate for this reaction in Fig. 1. These considerations indicate that (p)ppGpp metabolism does not involve obligatory conversion of pppGpp to ppGpp and therefore does not really represent a cycle.
Interesting effects of combining gpp mutations with spoT mutations have been found (431). While the gpp-1 allele is reported to be inviable in a spoT1 background, strains containing a much weaker spoT802 allele are viable; under these conditions, pppGpp and ppGpp accumulation behavior is consistent with effects of the spoT mutation on degradation of both pppGpp and ppGpp (431). In spoT + strains, the gpp-1 mutation shows no effect on growth or any other phenotype other than hyperaccumulation of pppGpp as discussed. In contrast, the gpp-1 spoT802 double mutant just mentioned grows slowly, with about twice the doubling time of either single mutant parent. Faster-growing revertants of the double mutant were isolated; about half of these were found to map with linkage to thyA as if they were relA mutants, while the remainder were intragenic revertants (431). Thus, the functional interrelationships established by suppressor activity for relA spoT and relC can very probably be extended to gpp as well.
It could be important to know whether hyperaccumulation of pppGpp has any unique regulatory effects not ascribable to ppGpp (see the section on transcription). It has been argued that this is unlikely since spoT mutants accumulating only ppGpp during all but the initial phase of the stringent response appear to show the same regulatory responses as do spoT + strains with low levels of pppGpp accompanying ppGpp accumulation, but this point has not been systematically explored.
There is a mutation distinct from gpp that phenotypically elevate levels of pppGpp. The sbaA (serine–branched chain amino acid metabolic control) mutation is pleiotropic, with elevated levels of both ppGpp and pppGpp; other properties include thermosensitive growth, RNA synthesis, a phosphotransferase pts carbohydrate permeation system defect, and extreme sensitivity to 2-ketobutyrate (81). This mutation was originally reported to map at 90 to 96 min but has since been identified as an allele of the rho gene (A. Danchin, personal communication).
Conversion of GDP to GTP by the protein encoded by ndk (nucleoside diphosphate kinase), as well as a variety of other enzymes, can be viewed as regenerating the principal substrate for pppGpp synthesis. A cold-sensitive S. typhimurium mutant that grows slowly at permissive temperatures (37°C) but not at all at 20°C was reported; it has very high basal levels of (p)ppGpp comprising half of GTP levels at permissive temperatures and (p)ppGpp levels equal to GTP levels at restrictive temperatures (149); levels of (p)ppGpp this high are achieved during a stringent response. Originally the mutant strain was selected as a cold sensitive and screened as 8-azaguanine resistant. It was later found to harbor two mutations (391). The ndk-1 mutation at 55 min resulted in altered nucleoside diphosphate kinase activity, verified in vitro, but it is not reported whether it has unique effects on (p)ppGpp levels. The other mutation maps in the relA region (60 min) but again has not been characterized with respect to (p)ppGpp in the absence of the ndk-1 mutation. In the absence of the second mutation, 8-azaguanine resistance was reported to be unreliable as an indicator of the existence of ndk-1, as if the putative relA mutant contributes to 8-azaguanine resistance. The ndk gene of E. coli has been identified at 54 min, sequenced, and overexpressed (156), but not examined with respect to findings for the S. typhimurium ndk gene.
It has been known for a long time that rifampin resistance (Rif r) mutations that reverse relA-dependent growth sensitivities (476) or alter sensitivity to ppGpp can be isolated (280, 281). Rif r mutations occur exclusively in the rpoB gene (216). One phenotype selected was reversal of the growth sensitivity of relA1 strains to serine, methionine, and glycine (79): about 5% of such mutants were found among spontaneous Rifr mutants (476). In E. coli B strains, an unusual relationship between the fractional amount of stable RNA to total RNA synthesis (rs/rt) as a function of (p)ppGpp concentration was encountered among Rifr mutants; about 8% of mutants were found hypersensitive to ppGpp (281). Two mutant alleles, designated rif-1 and rif-2, were studied in more detail and found to confer slow growth, a 10-fold reduction in basal ppGpp levels for a given growth rate, and a 20-fold reduction in rs /rt values as a function of ppGpp concentration. The growth rate of the mutants remained growth medium dependent and remained inversely correlated with ppGpp, although with a higher apparent growth sensitivity to increasing ppGpp levels. The authors interpreted this behavior as due to the operation of a feedback loop involving a mutant RNA polymerase that is hypersensitive to ppGpp, resulting in a lowered ribosomal content that, in turn, causes a lowering of ppGpp levels in order to relieve ppGpp inhibition of ribosome synthesis.
Single amino acid changes in the β subunit have been generated by suppression of rpoB amber mutations (325). Among these, three amino acid substitutions at positions 736 and 906 were reported to result in a relaxed RNA control phenotype (326). This result was interpreted as defining a target of transcriptional regulation during the stringent response and supported by competitive mixed-template transcriptional assays of promoter activity in the presence of added ppGpp (152). However, subsequent in vivo characterization of the suppressor mutants revealed that the ability of the cell to accumulate ppGpp, rather than the sensitivity of RNA polymerase to ppGpp, was defective (20). It is unclear why the ppGpp accumulation response should be altered; it could be that the polymerase mutants negatively affect isoleucine starvation sensitivity or relA responses. The in vitro assays with mutant polymerases were performed with added ppGpp, and therefore these results (152) cannot be easily explained by the in vivo characteristics of the rpoB mutants.
Tedin and Bremer (455) have isolated rpoB mutants among those conferring a modest resistance to the growth inhibitory effects of inducing (p)ppGpp with a PlacUV5 -relA gene fusion. These were screened by their effects on the activity of a specially constructed rrnB P1-lacZ fusion with a weakened Shine-Dalgarno sequence (195), selecting mutants that displayed increased reporter gene activity when ppGpp was induced. A single rifampin-sensitive (Rifs) mutant was characterized as possessing a twofold increased resistance of rs /rt values to ppGpp concentrations. Considering the data in Table 1 and the existence of multiple sites for ppGpp sensitivity at the transcriptional level as well as sites of metabolic inhibition, it is not surprising that high-level ppGpp resistance mutants are not easily isolated.
Our laboratory has isolated and partially characterized suppressors of two phenotypic features of (p)ppGpp-deficient strains (Hernandez and Cashel, unpublished data; M. Cashel, D. J. Jin, and H. Murphy, unpublished data). As a result of the (p)ppGpp deficiency, these strains are unable to grow on minimal medium, and 9 amino acids (Arg, Gly, His, Leu, Met, Phe, Ser, Thr, and Val) are found to be required are found to be required when omitted from an otherwise complete mixture of 19 amino acids (510). Isoleucine requirements cannot be scored in this test in the E. coli K-12 background because of valine sensitivity. It is unclear whether this behavior represents multiple amino acid auxotrophy or growth sensitivity; nevertheless, this behavior will be assumed to represent auxotrophy here. In addition, it was noted that these (p)ppGpp-deficient strains show poor survival after a long exposure to stationary phase.
One selection, for growth on minimal medium, resulted in 50 spontaneous mutations; all mapped in either rpoB, rpoC, or rpoD. The majority of these presumed suppressors of (p)ppGpp-dependent gene expression are roughly equally distributed between rpoB and rpoC, while only three were found in rpoD. About a third of the rpoB alleles are Rifr. A second selection, for long-term survival, led to finding that half of these survivors were prototrophs. Of 100 such prototrophs, all are found to map in rpoB or rpoC. When rpoBC genes are present in multicopy, survivors screened as either prototrophs or long-term survivors are enriched in rpoD mutations (Hernandez and Cashel, unpublished data). Some of the (p)ppGpp0 suppressor alleles are capable of reversing the behavior on plate tests normally observed for relA mutant strains, such as SMG sensitivity or AT sensitivity (see above), whether the host background is Δ relA or Δ relA Δ spoT. In addition, suppressor mutants in relA spoT + hosts differ in growth sensitivity to induction of (p)ppGpp with IPTG as if they are ppGpp hypersensitive or mildly ppGpp resistant.
Screening of the 17-member set of unique, sequenced Rifr alleles (216, 217) for the ability to confer prototrophy on a (p)ppGpp0 host reveals a general pattern: the Rifr alleles that increase termination efficiency in tests of rho-dependent and rho-independent terminators (214, 216, 217) rarely display prototrophy. In contrast, Rifr alleles that decrease termination efficiency in equivalent tests are commonly prototophic in such hosts. Finally, a prototrophic phenotype in (p)ppGpp0 hosts is also found among Rifr alleles that have no systematic effects on terminator efficiencies as measured above but do have effects on transcripts with box A-mediated antitermination (27, 271) that diminish antitermination efficiencies; i.e., they increase termination efficiency (215). This finding reveals that Rifr alleles that suppress putative (p)ppGpp-dependent functions can be sorted into two distinct classes.
The class of "auxotrophic" suppressors that decrease termination efficiency and pausing have been characterized as enhancing the affinity of RNA polymerase-elongating complexes for substrates and increasing the rate of elongation (214). It might be imagined that the ability of this class of Rifr alleles to suppress auxotrophy indicates that in the absence of (p)ppGpp there is increased pausing or polar effects on amino acid biosynthetic operons, thereby generating multiple requirements; the mutants diminish polar effects and become prototrophic. Difficulties with this interpretation are that long operons are not unique to amino acid biosynthetic genes and not all of the apparently required amino acids are encoded by long operons. Also, not all of the specific amino acids required in (p)ppGpp0 strains are regulated by attenuation, and tryptophan, which is regulated by attenuation, is not required.
The second class of "auxotrophic" suppressing Rifr alleles is not easily explained; this class has minimal effects on termination efficiency except for box A-modified transcripts. The effect of these alleles is to increase termination efficiency, the opposite of the class of suppressors just discussed. Apparently, opposite effects of termination efficiency on messenger and ribosomal transcripts appear to be able to confer prototrophy in the absence of ppGpp.
The Rifs (p)ppGpp0 suppressors mapping in rpoB and in rpoC, unlike Rifr alleles, have not been sequenced or defined with respect to their effects on termination or antitermination. However, a large number of mutants in rpoB and rpoC with altered termination or pausing properties have been reported and characterized by Landick and colleagues (64, 258), indicating that Rifs alleles can affect these processes.
As just mentioned, suppressors of putative ppGpp-dependent functions that map in the rpoD gene encoding σ 70 have also been encountered. These mutants have been sequenced as missense alleles in region 3 (Hernandez and Cashel, unpublished data), a domain widely conserved among sigma-related proteins (285). The function of this region is known only from properties of a deletion extending into the region for rpoH encoding σ 32; this lesion alters the competition with σ 70 for association with core RNA polymerase (523). The existence of (p)ppGpp0 suppressors in region 3 therefore raises the possibility that (p)ppGpp can alter gene expression by exerting effects on the competition between various sigma factors for core association or provide a clue to some other function of region 3.
While mutants in RNA polymerase subunit genes are often cited as providing strong evidence that ppGpp operates directly at the transcriptional level, these arguments are not completely convincing unless accompanied by in vitro demonstration of appropriate behavior of mutants with respect to (p)ppGpp effects in a pure system. For example, it seems plausible that multiple amino acid auxotrophy in (p)ppGpp0 strains is due to multiple genetic targets of (p)ppGpp-dependent gene expression. It follows that perhaps the only way to simultaneously alter responses of all sites with a single mutation would be to induce a global change in gene expression occasioned by an RNA polymerase mutation. An additional caveat applies to rpoD mutants with respect to the level at which ppGpp effects occur. Although rpoD mutant suppressors might seem to localize the (p)ppGpp effect to transcription initiation as opposed to elongation, this conclusion is not compelling because effects at the level of initiation and promoter clearance could be imagined to alter formation or assembly of antiterminating transcription complexes, with effects displayed long after σ 70 is released.
The kinetics of (p)ppGpp formation and breakdown during the stringent response have been extensively studied. Typically, low steady-state (p)ppGpp basal levels (10 to 30 pmol/A 450) characteristic of cells in rapid exponential growth begin to increase within a few seconds after amino acid starvation. The ppGpp levels continue to rise, peaking some 5 to 15 min later, and then fall to a new steady-state value that is 10- to 20-fold above the original basal level (54, 114, 289) to reach concentrations equivalent to those of GTP, which drops roughly in proportion to the increase in (p)ppGpp during this time, a total drop of about 50%. Detailed kinetic studies show that pppGpp is formed first from GTP and that ppGpp is derived from pppGpp, rather than the reverse. Rates of RelA-mediated pppGpp synthesis and ensuing ppGpp accumulation initially increase at least 10-fold and then decrease as the steady-state plateau is reached, as if a feedback regulation occurs. Depending on allelic states of the spoT and gpp genes, the abundance of pppGpp during the stringent response can vary from an excess of pppGpp relative to ppGpp in gpp mutants to an excess of ppGpp and barely detectable pppGpp in spoT mutants. Turnover of pppGpp occurs with a half-life of about 10 s for two reasons: hydrolysis to ppGpp is mediated largely by the Gpp enzyme, and hydrolysis to GTP is catalyzed by the SpoT protein. Degradation of ppGpp to GDP by SpoT occurs with a half-life of 20 to 30 s. Rapid (p)ppGpp decay rates are found early and late (30 min) in the stringent response if protein synthesis inhibitors are added or if the starvation is reversed by amino acid resupplementation (115, 134). Persistent rapid turnover of (p)ppGpp during the steady-state plateau indicates equivalent rapid rates of synthesis. A note of caution in these interpretations comes from the possibility that these measurements of ppGpp turnover in spoT + strains might be overestimates because they were performed with protein synthesis inhibitors or amino acid resupplementation, both of which might be expected to alter charged/uncharged tRNA ratios.
Deductions of synthetic rates from rates of accumulation together with measured decay rates after addition of protein synthesis inhibitors give (p)ppGpp turnover numbers approximating the rate of peptide bonds that would have been formed in growing cells (289, 336). The fate of energy expenditure accompanying (p)ppGpp turnover is unknown; it may be lost or coupled to other reactions forming high-energy compounds.
In strongly defective spoT mutants, ppGpp decays with a half-life that is 20 times slower than that in the wild type, yet only about twice as much ppGpp accumulation is observed to plateau during the stringent response (115). This finding indicates that (p)ppGpp synthesis is correspondingly much slower in steady-state starved spoT mutant cells. It has recently been suggested that this feedback mechanism is due to inhibition of a sizeable fraction of global mRNA synthesis by the stringent response (267; reviewed in reference 133). Sorensen et al. (432) propose that this results in a diminished frequency of hungry codon encounters that leads to inhibition of ppGpp synthesis. It has been long known that global rates of mRNA synthesis are inhibited during the stringent response (267; reviewed in reference 134). An analogous effect was observed in rifampin-treated cells to be correlated with effects of mRNA decay on inhibition of protein synthesis (107). The telltale disappearance of pppGpp characteristic of spoT mutants after initiating the stringent response can be an indication that rates of pppGpp synthesis are restricted while pppGpp conversion to ppGpp persists.
Attempts to provoke the stringent response or modify the stringent response with inhibitors have often revealed unexpectedly complex features of (p)ppGpp metabolism or side effects associated with a particular protocol or bacterial strain background. The preceding discussion has revealed interactions between a sizeable number of genes, identified by their suppressor activities, that can affect (p)ppGpp responses. Growth inhibition associated with increased (p)ppGpp basal levels comprises a positive selection for suppressor mutations; one result is that strain construction may result in phenotypic behavior unexpected from stated pedigrees. Regulation of (p)ppGpp levels is clearly governed by the availability of not only aminoacyl-tRNA but also energy sources, nitrogen sources, and possibly phosphate. The answers to the question of how to provoke the stringent response therefore depend on the sort of information desired.
With respect to (p)ppGpp involvement, there are three basically different conditions of provoking the stringent response. The classical comparison involves amino acid starvation of a relA + and a relA strain; the latter will give a relaxed response, with (p)ppGpp pools dropping below basal levels and aminoacyl-tRNA limitation without (p)ppGpp accumulation. The second comparison is with a (p)ppGpp0 strain. The third comparison is gratuitous induction of (p)ppGpp but without amino acid starvation. If verifying the phenotypic behavior of a given strain or recombinant is desired, then one can use simple screening methods involving thin-layer chromatographic analyses with extracts of phosphate-labeled cells resuspended from colonies (57).
Investigation of whether (p)ppGpp is involved in some process probably requires two parallel studies. The first involves use of (p)ppGpp0 strains. There are two sources of caution in such experiments. The first is that these strains generate suppressor mutations in RNA polymerase subunit genes because of their vulnerability to starvation and to stationary-phase exposures. The second is that these strains show multiple alterations in gene expression, as judged by two-dimensional protein gel analysis (221, 414). Although (p)ppGpp0 strains grow with near normal growth rates when not nutritionally limited, they show low growth yields as well as morphological changes indicating widespread lesions; therefore, the absence of a given regulatory phenomenon might occur merely because the strain is sick and may not be directly attributable to the absence of (p)ppGpp.
A second study, namely, assessing effects of IPTG-inducible (p)ppGpp accumulation in otherwise growing cells in the absence of a starvation or other sources of nutritional stress, is probably also required. This sort of experiment also warrants caution: the system is possibly too efficient, and excessive (p)ppGpp induction is also accompanied by lowering the GTP pools to levels less than the 40% drop normally seen during purine starvation of a purine auxotroph, which may trigger RNA turnover, derepression of purine biosynthetic enzymes, and other changes (131, 414). Studies with gram-positive organisms have led to proposals that changes in GTP pools alone may have pleiotropic effects on gene expression (342). Therefore, a range of inducing IPTG concentrations should be used.
If a given treatment of a wild-type strain results in (p)ppGpp accumulation, there are some recurrent patterns that may allow assignment of the lesion to one or another gene in (p)ppGpp metabolism. For example, if (p)ppGpp accumulation occurs in a relA + but not a relA strain, then the inference is that an alteration occurs either at the level of starvation for an amino acid or at the level of aminoacylation. Conversely, if (p)ppGpp fails to accumulate after amino acid starvation in a given strain, it is probably a relA mutant strain. If (p)ppGpp accumulation occurs in both relA + and relA strains, the mechanism is probably related to altering spoT-mediated (p)ppGppase activity. If pppGpp/ ppGpp ratios are lower than normal during the stringent response for a given strain, it is likely that the degradation defect is more severe. If pppGpp/ppGpp ratios are excessively high, inhibitory effects on Gpp seem likely.
Simple addition or removal of amino acids does not always yield the effects initially expected. For example, adding an excess of one-carbon metabolites including serine can result in an isoleucine starvation in relA strains (476); mutant analysis led to identification of 2-ketobutyrate as a putative alarmone capable of inhibiting the phosphoenolpyruvate:sugar phosphotransferase system and the presence of an unexpected starvation for aspartate in glucose-grown cells (82). Serine can serve as a carbon source; starvation for serine can block ppGpp degradation, leading to slow ppGpp accumulation in a relA strain (461). Simultaneous removal of several amino acids can result in other changes reminiscent of carbon source starvation, such as severe depletion of ribonucleoside triphosphate pools, but ppGpp accumulation remains typically relA dependent (for a review, see reference 133). In fact, the (p)ppGpp nucleotides were first particularly noticeable by the fact that they were more actively labeled with 32P during Casamino Acids starvation than virtually all other phosphorylated compounds (58). Amino acids whose synthesis requires sufficient high energy to comprise a metabolic drain during the derepression accompanying starvation include arginine, histidine, lysine, threonine, and the aromatic amino acids (414). Removal of an amino acid from a prototroph also results in transient derepression of the corresponding amino acid biosynthetic pathway. In some instances, supplementing a prototroph with several exogenous amino acids can cause metabolic imbalances leading to high ppGpp basal levels (133); more typically, amino acid supplementation speeds growth and lowers ppGpp basal levels. The effects of starvation can also be allele specific, with different effects dependent on accumulation or depletion of pathway-specific metabolites.
These considerations raise the challenge as to how amino acid starvation can be used to provoke the stringent response while minimally perturbing other functions. With uniformly 32P labeled cultures, accomplishing this by additions rather than removal is desirable. Valine addition and its reversal by isoleucine supplementation is useful, but only for E. coli K-12 strains, because it does not severely affect ATP levels. Pseudomonic acid is used to achieve starvation specific for isoleucine and may well be a generally useful way to induce a mild stringent response by a simple addition (201, 407). Serine hydroxamate inhibits tRNASer aminoacylation and is often used for studies of nucleotide and RNA accumulation responses (462) but requires high concentrations which are potentially troublesome because of the hydroxamate reactivity. Reversal of serine hydroxamate effects by adding a large excess of serine can, as just noted, trigger isoleucine starvation under some conditions. The use of temperature shifts of thermosensitive aminoacylation mutants has the clear advantage of no exogenous additions to growing cells and initially seemed ideal. However, temperature shifts can introduce complications because of responses to temperature itself, as well as because other nucleotides, like ppGp, accumulate under this condition (see discussion of heat shock and cold shock). Since a single protocol for provoking a "pure" stringent response has not yet been found, several ways probably should be attempted and compared.
Comparisons of relaxed and stringent strains during amino acid starvation reveal many differences (Table 1) and often lead to complications. An early problem in quantitating the extent of changes in RNA accumulation was that (p)ppGpp inhibited uptake of radioactive RNA precursors, giving overestimates of inhibition of RNA synthesis; in addition, inhibition of instantaneous global rates of RNA synthesis was threefold, greater than could be explained by even complete inhibition of rRNA and tRNA synthesis, which would give about a 50% reduction (133, 267, 330). Visualization of changes in gene expression during amino acid starvation from two-dimensional gel analyses of both proteins and RNA species reveals that about half of the proteins are affected, with equal portions increased and decreased in activity, while virtually all stable RNA species are inhibited (126, 205, 238, 379). Analogous studies during gratuitous (p)ppGpp induction or with (p)ppGpp0 strains reinforce estimates of the pleiotropic effects of ppGpp (221, 414). The limited number of entries in Table 1 relative to the number of proteins seemingly affected by (p)ppGpp suggests that many regulatory features associated with the stringent response have not yet been identified.
Evidence of unexpected involvement of the stringent response can come from finding aminoacyl-tRNA synthetase mutants as suppressors of a given phenomenon, such as mecillinam sensitivity (see below). Accumulation of (p)ppGpp accumulation and probably secondary involvement of the stringent response also can be occasioned by inducing other stress responses (477, 478). There is an example in S. typhimurium in which a new nucleotide, ApppGpp, perhaps a regulatory chimera, appears as a result of provoking ApppA accumulation together with (p)ppGpp accumulation by irradiating with broad spectra in the near-UV range, which is argued to provoke an oxidative stress response (245). Species of tRNA that contain 4-thiouridine, with an absorption maximum at 334 nm, can be internally cross-linked with a cytosine residue by sublethal doses of near-UV, resulting in a transient inhibition of growth (458). The modified tRNAs are poorly aminoacylated (514) but apparently functional in inducing relA-dependent ppGpp synthesis. The inhibition of protein synthesis is the combined effect of decreased aminoacylation and induction of the stringent response (374, 457). The resulting increase in ppGpp is expected to inhibit both cellular growth and the synthesis of new tRNA molecules. Nuv mutants lacking the ability to synthesize 4-thiouracil have been isolated in E. coli (404) and by a similar procedure in S. typhimurium (245). In E. coli B/r relA strains, this growth lag is four times shorter than in a relA + strain (457), allowing a direct selection of relA mutants (279). However, an E. coli K-12 relA mutant is only slightly better at recovering from this growth inhibition than a relA + strain (457).
There are reports in the literature of in vitro assays that implicate ppGpp, pppGpp, and even ppGp as mediating a given regulatory event; however, it is also commonplace to encounter conflicting reports. This is perhaps not surprising. With the exception of promoter mutants, there are no instances reported in which in vitro behavior is bolstered by parallel studies with (p)ppGpp-resistant or -sensitive mutant proteins. Were (p)ppGpp an antibiotic, few notions as to the mechanism of action would be seriously entertained without evidence arising from such mutants. Mutant promoters have been explored in an attempt to assess the basis of putative positive or negative (p)ppGpp effects on transcription initiation. While these have led to a growing appreciation of domains involved in promoter regulation, they have not yet led to identifying structural or functional elements that are both necessary and sufficient to account for stringent control in vivo. The large number of adjustments occurring during the stringent response raises the challenge of attempting to sort out those that are direct effects of tRNA deficiency from those that are due to (p)ppGpp accumulation, not to mention deciding between direct and indirect effects in both cases.
Amidst these uncertainties, it is possible to imagine physiological reasons for the spectrum of changes that occur during the stringent response in terms of how cells overcome a functional deficiency of aminoacyl-tRNA. Protein synthesis components and stable RNA already in excess over demands during starvation need not be synthesized. Formation of ribosomal proteins is restricted by a variety of mechanisms secondary to restricted rRNA synthesis (see chapter 90). While it might be expected that aminoacyl-tRNA synthetases would be under negative regulation during the stringent response, their regulation does not follow a simple pattern (17, 30, 313). With the absence of the need for RNA accumulation, it makes sense that transport and synthesis of RNA precursors might be limited, but the varying extents of inhibition observed are puzzling. Manipulated elevation of (p)ppGpp or ppGpp basal levels gives rise to growth inhibition in unstarved cells (411, 414), as if during the stringent response, curtailing growth by (p)ppGpp accumulation prevents complete exhaustion of aminoacyl-tRNA and thereby ensures reserves available to adapt to the stress (343, 446). With limited cell growth during the stringent response, it makes sense to see relA-dependent inhibition of phospholipid synthesis and fatty acid synthesis and effects on DNA replication (Table 1; see below). As mentioned, relaxed strains do become progressively more sick and show severe growth lags upon reversal of amino acid starvation after prolonged starvation. They also show increased permeability to a variety of added inhibitors and can leak intracellular pools (Table 1). The absence of the stringent response impairs adaptation to amino acid starvation. When (p)ppGpp is totally absent, impaired survival when cells are confronted with other sources of starvation stress becomes evident.
Positive stringent control is exerted on processes that can be viewed as overcoming the transient stress of amino acid starvation or as preparatory for surviving prolonged starvation. Thus, relA-dependent stimulation of intracellular proteolysis (349) can be considered as a way to provide a supply of missing amino acids from existing proteins. The coinducing effect of (p)ppGpp on synthesis of many amino acid biosynthetic enzymes is consistent with this view but does not apply to all amino acids. Transport is enhanced for branched-chain amino acids but apparently and perhaps surprisingly not for others (Table 1). The (p)ppGpp induction of expression of rpoS, encoding the stationary-phase-specific sigma factor, could be an example of contingency planning for long-term starvation (143). The subsequent effects of RpoS induction on gene expression are widespread and provide an example of effects indirectly related to (p)ppGpp abundance.
Inhibition of stable RNA accumulation was the first attribute of the stringent response characterized. Later, it was observed that the nucleotide ppGpp accumulated during the stringent response in a manner reciprocal to stable RNA accumulation (54, 58). It is still unresolved whether the mechanism of stable RNA inhibition by ppGpp involves changes in promoter selectivity by RNA polymerase, effects on transcription initiation, or effects on transcriptional elongation. For both rRNA and tRNA transcription, it appears that stringent control of RNA accumulation occurs at the level of synthesis rather than degradation, although degradation of rRNA occurs at very slow growth rates (139, 140; reviewed in reference 133) and in stationary phase, particularly in S. typhimurium (201). Although examples of inhibition of gene expression during the stringent response are by no means limited to stable RNA genes, here we will review only studies on rRNA and tRNA gene activity regulation and then turn to growth rate control. Regulation of ribosomal protein synthesis during the stringent response occurs as secondary to rRNA synthesis inhibition by a variety of mechanisms (see chapter 90).
There are seven rRNA operons (rrnA to -H) on the E. coli and S. typhimurium chromosomes, clustered near the origin of replication and in approximately the same locations, with all oriented in the same direction as procession of DNA replication (218, 269, 275). The expression and regulation of individual operons during the stringent response measured as cat fusion activity vary somewhat depending on chromosomal location; however, it is unlikely that the sevenfold rrn operon redundancy reflects dedicated functions of individual operons (74). The rrn operon transcripts contain tRNA genes in the spacer region between the 16S and 23S structural RNA genes as well as 5S rRNA genes; in some cases, additional tRNA genes are found beyond the 23S gene. A total of 14 tRNA genes are embedded in rrn operons (9, 275). All seven rrn operons contain tandem promoters; the upstream P1 and the downstream P2 promoters are about 120 bp apart. There is strong conservation of sequences in the leader as well as regions flanking the 16S and 23S mature sequences. The seven rrn operons are responsible for 50% of total transcription at fast growth rates. Despite the strength of the promoters in vivo, the sequences and the spacing (16 bp) of the –35 and –10 regions are not consensus (200), and the promoter properties are atypical in vitro (see below and chapter 90).
Regions flanking the two promoters themselves have been found to be responsible for regulatory effects on the long rrn transcripts but apparently without effects on their stringent control (161, 163). An upstream sequence, designated the upstream activating sequence or upstream activating region, is involved in P1 promoter activation and extends to about –160 upstream of the +1 RNA start site. This region shows extended contacts with RNA polymerase and contains AT-rich clusters. This region shows altered electrophoretic mobilities and activates the promoter in two ways: binding of Fis protein activation factor and binding of the α subunit of RNA polymerase to an otherwise factor-independent domain referred to as the UP element (129, 159, 274, 319, 329, 330, 333, 376, 398, 399, 521, 522). The kinetic effects of upstream activation on RNA polymerase transcription initiation are to increase the affinity for double-stranded DNA binding (KB ) (274, 520) and possibly (376) altering the isomerization constant (kf) (but see reference 521).
Fis-mediated stimulation of stable RNA synthesis is reported to be growth phase dependent; Fis protein concentrations increase early during outgrowth from stationary phase and then diminish in late log phase (334; reviewed in reference 39). Negative regulation of Fis transcription is reported as relA dependent at the transcription level; an autoregulatory circuit is also involved (335). Interestingly, fis deletion mutations do not decrease the activity of an rrnB P1-lacZ fusion containing Fis binding sites but decrease activity when Fis binding sites are absent (399), as if redundant mechanisms of regulation can compensate for the loss of Fis or, as has been suggested (265), the role of Fis-mediated activation is limited to early events during nutritional shift-up, not during exponential growth. A study of P1-P2 cat fusions in each of the seven ribosomal operons with wild-type upstream regions revealed a 1.5- to 2.5-fold lowering of activity in fis mutants, again less than what might be expected from in vitro rrnB P1 behavior (74). The region downstream of the rrn P2 contains a sequence homologous to lambda nut that when transcribed confers transcription antitermination, but again does not affect stringent control (27, 161, 271, 439) and is therefore beyond the focus of this review.
The promoter regions of both rrn and tRNA genes do show stringent control when fused to reporter genes (93, 150, 306, 353, 390). Early attempts to observe inhibitory effects of ppGpp on rRNA promoter activity in vitro with linear templates and pure transcription systems gave inconsistent results. Attempts to demonstrate "stringent control in vitro" resulted in devising special conditions to obtain the effect with poor reproducibility (148, 174, 183, 337, 352, 445, 457, 464, 467, 490, 481, 489). In addition, an equally unfruitful search was made for special factors that could yield the same desired result (28, 87, 317, 318, 351, 457, 468, 469). In a relatively limited number of studies, more consistent rRNA synthesis inhibition by added ppGpp was obtained in crude S30 systems (150, 306, 390), as was evidence of sensitivity of rrn transcription to inhibitors of DNA supercoiling (513). Some day, when the primary effects of (p)ppGpp are understood, it may be realized that many of these studies touched upon critical features of the mechanism. A trend from these studies as well as new observations will be discussed in the section on tRNA transcription. The credibility of this approach is strained by the uncertainty as to whether specially devised conditions in vitro which allow a match with apparent in vivo behavior truly reflect the operation of the mechanism or whether in vitro conditions for studying unusual promoter properties can be perturbed in such a fashion that they produce an artifact affected by ppGpp.
The persistent difficulties in obtaining clear evidence of ppGpp inhibition of initiation of rRNA transcription were seemingly explained by proposals that the effect of ppGpp was on the elongation phase of transcription instead. These studies demonstrated that with phage T7, phage T3, and rrnB templates, pausing at low substrate concentrations was exacerbated by the addition of ppGpp; in contrast, ppGpp addition did not display inhibition of elongation competitive with ribonucleoside 5'-triphosphates with synthetic polydeoxyribonucleic acid templates (236-238). In the case of rrnB transcripts, pausing occurred in the region of the P2 promoter as well as at a site beyond the box A sequence in the leader region, called tL (236). From these data, Kingston and Chamberlin proposed a model of "turnstile" regulation, with one of the turnstiles in the P2 promoter itself (236). This view was supported by finding that pausing in vitro at P2 was abolished when the sequence in the +1 position of the P2 transcript was changed from a C to a G or an A by deletions which also changed sequences downstream of +1 (235). The sequence change also increased P2 promoter strength (235), which supported the earlier contention that this pause might involve promoter occupancy at P2 (236). The turnstile mechanism predicts that the presence of P2, tL, or both contributes to stringent control.
The relative strengths of P1 and P2 in vivo and their contributions to regulation during the stringent response could be measured with these regions cloned in multicopy plasmids (163, 410, 412) or in lysogens (161) and matched with earlier studies detecting promoter-specific RNA chain termini (88, 288); multicopy plasmid constructs required the presence of strong ribosomal T1 and T2 terminators to stabilize antitermination effects and allowed visualization of uniquely sized RNA transcripts. It was found that deletion of the P2 promoter along with the downstream tL site did not reverse inhibition of P1 activity during the stringent response (163, 410). Recently, the tL region has been implicated as providing a RNA chaperonin-like function for rRNA maturation (494). At relatively fast growth rates, the P1 promoter was found to be three to five times more active than the P2 promoter, and P1 was preferentially inhibited during the stringent response (412). During carbon source starvation (409) and by systematically increasing the ppGpp basal level in exponentially growing cells harboring spoT mutations of increasing severity (411), very similar quantitative relationships between ppGpp abundance and chromosomal rrn P1 inhibition were shown. In addition, when (p)ppGpp was gratuitously induced with IPTG and total chromosomal rrn activity was assayed by primer extension, ppGpp was found sufficient to inhibit P1 activity (414).
Jensen and Pedersen (213) have proposed that ppGpp can inhibit transcription elongation, thereby sequestering RNA polymerase and reducing the availability of active enzyme for initiation. They point out that if the rrn P1 promoter is not saturated at high enzyme concentrations, unlike most mRNA promoters, then the indirect effect of ppGpp-dependent sequestration will be the apparently selective inhibition of rrn initiation activity. In vivo studies of the time of appearance of full-length transcripts by hybridization probes reveal that ppGpp induction during the stringent response does reduce the transcription elongation rates by half for lacZ and infB mRNA but not for an IPTG-induced, box A-containing truncated rrnB transcript (489, 491). The box A sequence conferring antitermination is important for the elongation inhibition observed because if the box A sequence is introduced in mRNA constructs, the inhibition of elongation is abolished (U. Vogel and K. F. Jensen, unpublished data).
The current passive control model raises the question of the nature of the mechanism by which ppGpp affects pausing. General acceptance of this model, as with other proposals, requires a compelling identification of the sites at which ppGpp has direct effects. Demonstration that ppGpp amplifies RNA pausing at low substrate concentrations was accompanied by evidence that ppGpp is not a competitive inhibitor of elongation on natural templates, while no inhibition at all is observed on synthetic alternating templates (238). This finding supports the view that specific sequences are required for ppGpp inhibition.
The passive control model of rrn promoter initiation by sequestering RNA polymerase initiation capacity raises several experimentally testable issues. For example, overproduction of functional RNA polymerase should compensate for the proposed effects of elongation inhibition and abolish stringent control of rrn transcripts. Furthermore, quantitative documentation is needed for the generalization that rrn promoters, and presumably tRNA promoters as well, are not saturated with RNA polymerase relative to mRNA promoters. Finally, it should be possible to mimic the effect in vitro. A mixed template transcription assay system has been devised (227, 228, 229) and might be reexamined with this in mind. There is a brief report that the in vivo activity ratios of rrn P1 and P2 promoters on supercoiled plasmid templates could be achieved in vitro at elevated RNA polymerase concentrations (151). In this case, ppGpp was found to inhibit initiation at P1 in multiround transcription assays.
There is a sequence feature of the rrn P1 and P2 promoters that seems related to stringent control in a manner not necessarily in conflict with the passive model of rRNA regulation, but again with conflicting reports. It has been mentioned that analysis of activities of the tandem promoters as they naturally occur revealed that P2 activity was only modestly regulated when ppGpp levels increased. The notion that P2 was a constitutive promoter was supported by similar regulatory behaviors for P1 and P2 in constructs in which they were separated from one another (412). The unregulated feature of the activity of the solitary P2 promoter and the regulated activity of the solitary P1 promoter were independently verified with respect to effects of the stringent response and growth rate (161, 512). The sequence differences between P1 and P2 promoters could be expected to reveal critical regulatory determinants of stringent control.
A much more generalized sequence comparison between promoters showing negative and positive stringent control was made long ago (465, 466) and revealed recurrent patterns in the region, called a discriminator, between the –10 Pribnow box and the +1 transcription start site. An update on such sequences for stable RNA gene promoters is presented in chapter 90. Negatively controlled promoters are more often GC rich in this region, while positively controlled promoters tend to be AT rich. Mutational alterations of discriminator region sequences have yielded changes in responsiveness to ppGpp, and genes found to differ in their apparently ppGpp-mediated regulation have been found to differ appropriately in their discriminator regions (see section on tRNA, below).
When the P2 discriminator sequence –7 to –4 GCAC was changed to the GCGC found in P1, the regulatory properties of P2 were converted to those of P1. Other sequences were judged to be also involved because an equivalent change in the discriminator of an unregulated promoter did not confer stringent regulation. New studies (130, 225) have produced very different conclusions: evidence is presented that the P2 promoter in isolation shows activation as well as stringent control and that differential regulation of paired promoters seems to occur by a promoter occlusion mechanism (1), regardless of whether the promoter pair is P1-P1 or P2-P2. The regulatory consequences of the natural P1-P2 arrangement of promoters are that substantial rrn P2 transcription occurs when the P1 promoter is inhibited. Detailed comparisons of the properties of the P2 promoter with those of P1 seem worth examining to see if position and promoter occlusion are the only distinguishing features of their differential regulation.
Mutants apparently selectively defective in stable RNA accumulation for reasons other than ppGpp accumulation might be expected to lead to identification of unique participants in stable RNA synthesis or regulation. Temperature-sensitive fructose 1,6-diphosphate aldolase (fda) mutants that accumulate high levels of fructose 1,6-diphosphate are correlated with inhibition of RNA accumulation as well as rrnB P1 promoter activity (34, 35, 273, 424). In such mutants, high levels of fructose 1,6-diphosphate were not found to greatly affect the activity of the growth rate-unresponsive rrnB P2, PlacUV5 , or Pspc promoter (423). It is unclear whether this observation indicates that the sugar diphosphate can uniquely and specifically inhibit rrnB P1 or that it mimics the effects of ppGpp when at abnormally high concentrations. Studies investigating the effects of fructose 1,6-diphosphate in ppGpp-resistant and -sensitive mutant hosts might distinguish between these possibilities.
Interestingly, apparently all of the seven rrn P1 promoters have embedded in their sequences a heat shock promoter recognition element such that the start site of the σ 32-directed transcript in vitro is identical to the σ 70 RNA start site (327). The possibility that the heat shock-promoted activity participates in stringent control, or growth rate control, of P1 activity was ruled out by using rpoH deletions. As yet, there is no evidence of σ 32-promoted rRNA transcription in vivo (74), but it seems a reasonable possibility, particularly under conditions in which σ 70 is inactivated (327).
The total complement of 78 tRNA genes in E. coli K-12 strains has been mapped (116, 243) as more disperse than rrn operons and with varying degrees of organizational complexity, including their occurrence in translated operons, clusters of only tRNA genes, and single transcriptional units. Regulation of expression of most tRNA genes not embedded in rrn operon transcripts shares stringent control features with regulation of the rrn operons (205). There are exceptions. An undermethylated form of ` shows uniquely active synthesis when cells are treated with trimethoprim (205) or methionine starved or when a leucine-starved hisT mutant is hypomethylated (80). The three 
transcripts from the PleuV promoter are under stringent control without a C1 metabolite deficiency (Vogel and Jensen, unpublished data). In multicopy, massive overproduction of these tRNAs occurs and leads to slow growth, two- to threefold inhibition of translational elongation rates, 90% hypomodification, and no increase in ppGpp despite the finding of only a 40% level of charging (495). These studies could reflect features found by Rojiani et al. (393, 394), but interpretations are complicated by probable saturation of modification capacity for both the species overexpressed as well as other tRNAs that share modification systems. The hisT1504 mutation has since been found to exert effects on modifications in the anticodon loops of tRNAs and give a 23% reduction in chain elongation rates, which, in turn, affects transcriptional attenuation of several amino acid biosynthetic operons (172). The 
species, but not 
I or II, shows patterns of aminoacylation during growth and starvation that is the reverse of the expected behavior (365).
Growth rate can affect tRNA abundance in a manner depending on whether the tRNA species recognizes major and minor codon preferences. A study in which growth rates were varied revealed that the minor codon cognate 
abundance decreased with increased growth rate whereas the abundance of a major codon cognate, increased (105). Studies of stringent control of promoters for these two species revealed the latter to be also under strong stringent control, using either the leuV or argT promoter, whereas the leuX promoter gave a milder response and gave no response at all when a promoter fragment was fused to lacZ and inserted into the λ attachment site (400). Another example of differential regulation may be found for metZ-promoted expression of two copies of 
with a GC-rich discriminator and metY-promoted synthesis of 
lacking a GC-rich discriminator (320).
At least some well-studied tRNA promoters seem to share structural and functional properties with rrn promoters, although there are exceptions. An example of an exception is that the 16-bp spacing between the –35 and –10 recognition regions found for P1 and P2 promoters is found only in a fraction of tRNA promoters (116, 243). Changing the spacer to the consensus 17 bp activates rrnB P1 as well as the leuV promoter (23, 127). Instances of AT-rich upstream activation are also found for tRNA promoters (41, 256, 257) as well as unusual DNA structures giving anomalous fragment migrations on gels (41), activating effects of supercoiling (111, 255), and evidence of upstream contacts with RNA polymerase (256, 257, 489). The –35 regions of rrn promoters and most tRNA promoters do not have the consensus sequence TTGACA, and changes to the consensus sequence can give about a 20-fold activation and loss of growth rate regulation (23, 127). A GC-rich discriminator is common among tRNA promoters; changing the –4 to –1 positions of the tyrT promoter from CCCC to TTAA abolishes the stringent regulation of transcription in vivo (257) and in vitro (465). Although transcriptional regulation of the thrU (tufB) operon is appreciated as more complex owing to Fis activation (39), evidence exists for a parallel dependence of this promoter on a GC-rich discriminator region (308, 309, 310).
Among tRNA promoters, the tyrT promoter has been the most extensively studied. Like rRNA promoters, it is salt sensitive. Assays of ppGpp sensitivity in vitro reveal optimal inhibition under intermediate salt conditions, giving weak residual activity (481, 489). Hamming et al. (174) noted that for the rrnE promoters, the transition between closed and open complexes was apparently slow and that the rate of decay of open to closed complexes, judged by single-stranded DNA challenges, was stimulated by high salt. The effect that they found for ppGpp was to lower the formation of challenge-resistant open complexes (174). A later study on rrnE promoters suggested that P1, but not P2, was inhibited by ppGpp and that equivalent inhibition was seen if ppGpp was added before or after open complex formation, suggesting that the mechanism of inhibition was to enhance the decay of open complex (228, 229). The tyrT promoter was demonstrated under low-salt conditions to be extremely sensitive to S1 nuclease cleavage in the –10 region; this sensitivity was lessened by a –10 mutation that lowered promoter activity, and it was suggested that promoter activity involved melting in the –10 region (102). Lamond (255) noted that the salt sensitivity of the tyrT promoter in vitro was diminished with supercoiled templates, even without upstream activation sequences, and was restored by relaxing plasmid templates with topoisomerase I.
For rrnB P1, Gourse (160) found that stable open (ternary) complexes, called P1 RPinit, were formed in low salt if ATP and CTP were added as initiating nucleotides to allow the predicted formation of a dinucleotide pppApC from the template which footprinted from about –60 to +20. The kinetics of formation of the complex were extremely rapid, and the enzyme:DNA stoichiometry of complex was 1:1 (160). It was later appreciated that a contribution to stability was provided by a slipping mechanism whereby the pppApC formed as a first phosphodiester slipped back to the –3 position on the template, then three more nucleotides were added to form pppApCpCpApC, and the ternary complex was trapped in stable form (38). Subsequent studies of Liermo and Gourse (274) show that kinetically, the formation of the closed complex is efficient and is more so when the upstream activation region is present; the KB was estimated as 2.5 × 10–9 M–1. In addition, the rapid decay of open complexes to closed complexes was shown to be highly reversible; decay of open complexes is diminished with supercoiled templates but still much more rapid than for typical promoters. Permanganate assays of open complexes with the same promoter indicated that melting was difficult; very low salt or supercoiling facilitated melting and decreased the effectiveness of ppGpp in inhibiting the formation of open complexes (347). Whereas rifampin inhibits melting in vitro if added prior to ATP and CTP, presumably by blocking RPinit formation, in vivo neither the drug nor ppGpp altered melting of P1; it was concluded by Ohlsen and Gralla (349) that ppGpp effects occur after closed complex formation. The authors demonstrated in vitro in the absence of substrates that two different, unusually stable closed complexes form as defined by their DNase I protection patterns: with supercoiled templates, protection is from –40 to +10, whereas with linear templates, protection occurs from –40 to –7. If ATP and CTP are added to linear DNA, closed complexes predominate, with no effect of added ppGpp. With supercoiled templates, where melting is facilitated, a small amount of inhibition of open complex formation is seen as if ppGpp can inhibit the transition from closed to open complexes but not the formation of closed complexes (348). It is noteworthy that again the open complex in this case is actually a ternary complex, having formed at least a dinucleotide product rather than more typical binary open complexes.
It seems possible that these salt-sensitive rrn P1 and tRNA promoters share the atypical feature of being difficult to melt, owing perhaps in part to a GC-rich discriminator region, and likely to isomerize back to a closed complex if promoter clearance does not occur. Low-salt conditions or supercoiling can facilitate melting and thereby RNA polymerase activity. It appears that ppGpp may inhibit some step after closed complex formation, but the effect is subtle and can easily be abolished if melting is facilitated. The paradox raised by the difficulty of melting versus the extremely strong activity of this class of promoters in vivo suggested to Ohlsen and Gralla that a very high KB must keep the promoter filled with enzyme even at very low enzyme concentrations (348). Promoter clearance rates are an important consideration for passive control effects mediated by variation in the high range of enzyme concentrations. Experiments addressing this point over a wide range of concentrations are needed.
Venetianer (483, 484), noting that his operon mRNA was more abundant during the stringent response, was the first to suggest that it made sense for cell economy to allow amino acid biosynthetic operon expression during the stringent response. Stimulation of his operon expression during the stringent response was subsequently demonstrated in vivo and mimicked in vitro by using S30 extracts with coupled transcription-translation activity (444). Effects on histidine synthetic enzymes were generalized to other amino acids by showing that relA mutants were hypersensitive to a variety of amino acid analogs (444). As mentioned earlier, two-dimensional gel analyses of proteins synthesized during the stringent response revealed that the fraction of proteins whose expression was increased was large and equal to the fraction of those displaying negative control (126, 238, 379). Of the identified proteins displaying positive control, many are involved in amino acid biosynthesis (Table 1). The complex amino acid requirements of (p)ppGpp0 strains are consistent with multiple requirements for (p)ppGpp in these processes. There is also substantial evidence of a positive regulatory role for ppGpp arising from genetic selections for expression of amino acid biosynthetic operons. In the case of his operon expression, this evidence is particularly strong (see below). In vitro evidence of the mechanisms involved in positive control has been elusive and limited to studies with crude S30 extracts capable of transcription coupled to translation. In the case of his regulation (5, 444), experiments in which transcription was uncoupled from translation indicated that control by ppGpp occurs at the level of transcription and is attenuator independent. The attenuator independence of regulation is confirmed by cellular studies (401, 418, 502) in which lacZ reporter activity differences ranged from 20- to 60-fold in parallel over the spectrum of the lowered ppGpp basal levels during the relaxed response to the high levels achieved during the stringent response. Evidence that the his promoter structure might itself be a determinant of ppGpp dependence comes from mutational analysis (388); however, mutants in the –10 region conferring activity in the absence of ppGpp brought the promoter sequence closer to consensus, and activation might be expected in any case. Similar studies of the positive control of lac expression are consistent with this result, including the increase in ppGpp independence of the strong promoter lacUV5 mutant (369, 370, 418). An AT-rich discriminator region between the –10 and +1 region has been implicated by correlation in examples of genes whose expression is thought to be subject to positive control (465, 466).
A potentially worrisome point is the level of attenuator-independent transcriptional activation that actually occurs. Early estimates of his mRNA levels (483, 484) were for the wild-type operon; mRNA levels are only inferred from levels of his enzyme or of β-galactosidase activity synthesized from his-lac fusions. In the case of thr (240) or arg (501), mRNA increased only 2.5- or 2-fold, respectively, following starvation for threonine or arginine. In the case of thr, starvation for arginine or histidine did not cause an increase in thr mRNA (240, 364). In S. typhimurium, positive control of his expression can be elicited by adding serine hydroxamate (418) or starvation for other amino acids (502). In contrast, in E. coli, estimates of his promoter activity by primer extension, using primers localized upstream of the attenuator region, do show relA-dependent stimulation by histidine starvation but not by starvation for serine or isoleucine or by induction of ppGpp without histidine starvation (Glaser, personal communication). This finding implies that ppGpp may be necessary for positive regulation but not sufficient because of lack of some component influenced by starvation for a specific amino acid. In general, there may be a need to recharacterize examples of putative positive control responses with gratuitous (p)ppGpp induction and effects of specific starvation protocols.
An example of positive control is the apparent participation of (p)ppGpp in the induction of the rpoS gene during the stringent response (143). Many genes involved in stationary-phase development in E. coli are dependent on the sigma factor, σ S, encoded by the rpoS gene (see chapter 93). Levels of σ S increase in late exponential and early stationary phase (261, 284, 298), when (p)ppGpp synthesis is presumably induced, making it a candidate for positive control. A dependence on ppGpp by rpoS is also suggested by the phenotype of (p)ppGpp0 strains, which have many of the features of rpoS mutants (143, 510). Immunological estimates of σ S abundance in (p)ppGpp0 strains indicate abnormally low levels of expression following starvation for glucose, amino acids, or phosphate as well as during entry into stationary phase after growth on a rich medium. However, these impairments of rpoS expression are not complete defects because levels of σ S approach those of a wild-type strain in a 24-h culture. The implication of these observations with (p)ppGpp0 strains, coupled with the stationary-phase vulnerability of these strains, is that the timing of expression of rpoS might be critical to normal stationary-phase development.
A direct dependence of rpoS expression on ppGpp was shown by a 25- to 50-fold induction of σ S within 10 min after (p)ppGpp was induced with IPTG under conditions not limited for nutrients (143). Transcription of rpoS-dependent genes, such as bolA (4, 5, 260), a morphogene participating in cell shape determination, and dps, encoding a DNA binding protein (8), is also induced under these conditions but not when (p)ppGpp accumulation occurs in a strain carrying a rpoS deletion. This finding indicates that under the conditions tested, (p)ppGpp lacks effects on these two genes other than those mediated through σ S induction (Gentry and Cashel, unpublished data). These responses typical of starvation and stationary phase can also be demonstrated during slow exponential growth in nutrient-rich media caused by spoT mutants defective in (p)ppGpp turnover (Gentry and Cashel, unpublished data). Such mutants have only modest elevations of ppGpp yet display high basal levels of σ S and σ S-dependent gene expression; in the case of bolA, expression increased proportionally with ppGpp levels.
These observations with positive control of rpoS expression by (p)ppGpp underscore the likelihood that some features of the stringent response are mediated indirectly by (p)ppGpp through effects on σ S induction. The overproduction of σ S itself might alter competition between σ 70 and core RNA polymerase or otherwise alter the genes normally expressed during exponential growth; overexpression of the Dps DNA binding protein also may have general consequences for gene expression. For genes governed by σ S, these possibilities can be easily examined by comparing one or another feature of the stringent response in hosts deleted for rpoS. It seems reasonable that σ S-independent genes selectively expressed during stationary phase might also contribute analogous indirect effects. rmf (ribosome modulation factor) is a particularly interesting candidate in this respect because of its possible alteration of ribosomal activity (512).
Besides Dps, another nonspecific DNA binding protein, H-NS, is reported to show an increase in abundance in stationary phase (435) but is not induced by IPTG induction of (p)ppGpp (Gentry and Cashel, unpublished data). Site-specific DNA binding proteins may also be involved in these circuits. It appears that (p)ppGpp negatively regulates the CspA cold shock protein (156, 221, 226), which has regulatory effects on H-NS and gyrA mediated by specific DNA binding (see discussion of cold shock). Identification of enhanced expression of integration host factor (IHF) during IPTG induction of (p)ppGpp led Aviv et al. (16) to discover that ppGpp activated the himA P4 promoter, transcribing the α subunit gene of IHF, in a σ S-dependent fashion. They also found that the himD (hip) promoter, transcribing the β subunit of IHF, was also activated but in this case independently of rpoS mutants. The induction of IHF expression by ppGpp is of added interest because IHF, along with σ S, contributes to activation of dps (9). Also, the gene encoding Fis is apparently negatively regulated at the transcriptional level by ppGpp in addition to autoregulation (335), and these events have been implicated in differential regulation of Fis during phases of growth (39). It is becoming increasingly evident that understanding growth phase-dependent regulation has the potential of yielding insights into the operation of the stringent response and the role played by (p)ppGpp.
The known responsiveness of (p)ppGpp regulation can, in turn, be used as a way of viewing starvation-dependent regulation. For example, Spector and Cubitt (436) isolated a number of lacZ fusions that are induced by carbon, nitrogen, and phosphate starvation. These loci also showed relA-dependent and subsequently rpoS-dependent expression (350). A possible explanation for this behavior is that the relA effect is due to (p)ppGpp stimulating rpoS expression, which leads to induction of expression of fusions at these loci. They also found that phosphate starvation responses were unaltered by a relA mutant, which is consistent with observations that (p)ppGpp0 strains show defective induction of σ S expression during phosphate starvation (148) and that ppGpp accumulation in response to phosphate starvation has a strong spoT-dependent component (Hernandez and Cashel, unpublished data). This response to phosphate starvation may be strain specific, for it was not found to any appreciable extent in earlier studies with other strains (187, 266).
The mechanism of activation of rpoS by ppGpp is not easily explained at the transcriptional level and has been suggested to be regulated posttranscriptionally (284). The rpoS gene is in an operon containing one other gene, called nlpD, for new lipoprotein D, and at least three promoters in the order P1-nlpD-P2-P3-P4-rpoS, where promoters P2 to P4 are embedded in nlpD (203, 261, 454). Using transcriptional fusions, Takayanagi et al. (454) have shown that all three promoters are induced, giving about threefold elevation of rpoS mRNA as cells enter stationary phase, and that the strongest promoter is P2. Results from fusions, direct mapping by primer extension, and RNase protection experiments in (p)ppGpp0 strains show that transcripts originating from P2 are lowered, as if P2 is ppGpp-dependent (Gentry and Cashel, unpublished data; R. Hengge-Aronis, personal communication). While σ S abundance is lowered about 10-fold by (p)ppGpp deficiency, levels of rpoS transcripts are lowered at most by 3-fold. After induction of ppGpp with IPTG, the 25- to 50-fold induction of σ S protein is not accompanied by as large an increase in levels of rpoS message (Gentry and Cashel, unpublished data). The metabolic stability of σ S is known to be increased 10-fold upon entry from rich media into stationary phase (454), but (p)ppGpp induction of σ S in exponential cells is not accompanied by stabilization of the protein (Gentry and Cashel, unpublished data). Taken together, these data imply that ppGpp affects rpoS transcription and translation. Lange and Hengge-Aronis (261) reported that the translation of rpoS is induced by osmotic shock. This observation can be taken to support the view that ppGpp regulates translation, since osmotic shock also induces ppGpp accumulation (182) and fails to induce σ S in the (p)ppGpp0 strain (Gentry and Cashel, unpublished data).
A stimulatory effect of ppGpp on posttranscriptional gene expression is not unprecedented. Williams and Rogers (501) found a similar effect on the expression of the arg operon: starvation for arginine increased transcription in a relA-independent manner; however, the levels of arg enzymes increased within 15 min in a stringent strain but were delayed for 2 h in a relaxed isogenic strain. When arginine was added back, the wild-type strain exhibited a pulse of arg enzyme synthesis whereas the relA1 mutant strain did not. Furthermore, if glucose starvation of the relaxed strain was superimposed on arginine starvation by addition of α-methylglucoside in order to restore ppGpp accumulation, then the relA1 strain regained the ability to induce the pulse of arg enzyme expression when resupplemented with arginine. These studies were predated by in vitro experiments using S30 extracts, in which it was found that ppGpp addition stimulated the expression of arg gene expression more than 10-fold while transcription increased by only 50% (524). When transcription and translation in this system were uncoupled, ppGpp failed to stimulate arg gene expression, as if translation must accompany transcription for ppGpp stimulation to occur.
It is plausible that enhanced expression during the stringent response for a portion of the many genes showing positive control is due to indirect, passive effects and that different mechanisms coexist (432, 433). Decay rates of a given mRNA species can show complex dependence on efficiencies of translational initiation that may in turn be affected by the competition of species of mRNA for ribosomes (299, 515). The stability of some mRNA species might be altered through some direct aspect of the stringent response itself, altered by some effect secondary to destabilization of other mRNAs, or regulated by the growth rate itself (327). A convincing experimental example involves inducing a long-lived lacZ mRNA construct and measuring its translation efficiency during the period of slow decay of ppGpp after starvation of a spoT mutant strain is reversed (432). It was found that β-galactosidase expression was enhanced under these conditions whereas the rate of translational elongation was unaffected.
In related studies of the effects of induction of lacZ during the stringent response, it was found that mRNA can become limiting and in turn limit overall residual protein synthesis by about 30%. It is argued that this is sufficient to have a significant effect on codon demand and constitute a feedback control on ppGpp synthetic activity (432), although this view is debatable.
An example of enhanced gene expression probably due to unique stability of a specific mRNA is the product of the ompA gene (187). This protein was expressed in such large amounts during the stringent response that it was called the stringent starvation protein (Ssp) (324, 379). Considerable confusion was generated when a protein copurifying with RNA polymerase was misidentified as Ssp and sequenced, and the corresponding gene was identified at 69.5 min (124, 211, 416). The sequence of the (new) ssp operon comprising sspA and sspB was determined, along with measurements of sspA and sspB expression during the stringent response and analysis of the regulatory effects of an sspA deletion (500). This SspA was not expressed during the stringent response like the original Ssp (OmpA) protein but was modestly responsive to slow growth as well as starvation for amino acids, phosphate, glucose, and ammonia in a relA-dependent but rpoS-independent manner (500). A deletion of the new sspA gene affects expression of a number of proteins, including H-NS (500). In contrast to F24.5 migration of the OmpA protein on two-dimensional gels (324), the new SspA protein migrates with an alphanumeric of D27.1 (498). Thus, evidently the new SspA protein exhibits but a modest response to amino acid starvation and is distinct from the OmpA protein originally identified as Ssp.
The strongest evidence of positive regulatory effects of ppGpp on amino acid biosynthetic operon expression has come from the widely used plate tests for identification and isolation of mutants affecting (p)ppGpp metabolism. These involve conditions giving a mild amino acid starvation stress and finding growth differences dependent on ppGpp accumulation responses. Plates containing serine, methionine, and glycine (or leucine) (SMG medium) can differentiate on the basis of defective ilv operon expression in relaxed strains of both E. coli and S. typhimurium (79, 401, 476). A medium containing the histidine antimetabolite AT inhibits growth of relaxed mutants due to defective his expression and is suitable for identifying relA mutations in a wide variety of organisms, including Bacillus subtilis (164, 401). Mutations in spoT can be isolated as suppressors of the growth defect of relA mutants on AT or SMG plates by virtue of their ability to raise basal levels of ppGpp, leading to increased basal expression of the his or ilv gene (401, 411, 444, 503). spoT mutants can be identified in a relA + background by using another histidine antimetabolite, 1,2,4-triazole-3-alanine (401). Suppressors of AT sensitivity of relA mutants other than spoT have been isolated in spoT merodiploid backgrounds and have been localized to gyrB, which can be explained by the effect of supercoiling on his expression, and to rpsB mutations, whose effect on his expression is not presently understood (459, 460).
There is a temptation to rank the various examples of positive control with respect to responses to ppGpp concentrations. For example, expression of the his operon is thought, from studies with spoT mutants and effects of histidine inhibitors, to respond maximally to slight elevations of ppGpp above basal levels (401, 444, 502) and be titrated by lower levels (418). The responses of isoleucine and valine also seem particularly sensitive to ppGpp, although specific effects of low levels of ppGpp on enzyme activities have not been measured (158, 475, 476). However, this ranking may be invalidated by the existence of amino acid specific contributions even in attenuator-deleted examples.
The literature contains evidence of correlations between the accumulation of (p)ppGpp and the operation of one or another feature of the stringent response, usually stable RNA synthesis. The literature also contains numerous examples of exceptions; some of these can be explained, and some cannot (see reviews in references 206, 337, and 425). Both sorts of exceptions will be summarized, with more attention given to those that are unresolved. Exceptional behavior sometimes is provoked by unusual starvation protocols, host backgrounds, temperature shifts, carbon source starvation, or extremely slow growth.
The levels of (p)ppGpp have been correlated with instantaneous rates of stable RNA synthesis. There are sources of error if these measurements involve isotope labeling arising from (p)ppGpp-mediated inhibition of labeled RNA precursor uptake (Table 1). The result is lower specific activities of the ribonucleoside triphosphate pools as ppGpp levels increase (133, 267, 330, 505). This effect is most severe for labeling with pyrimidines or pyrimidine nucleosides, moderately severe for guanosine labeling, and least severe for adenosine or Pi labeling (207). Thus, using incorporation of labeled uracil to measure rates of RNA accumulation gives an overestimate of effects of ppGpp. Precursor equilibration artifacts can be eliminated by measuring specific activity changes and integrating these over the time of incorporation (139, 140, 267). A way of circumventing measurements of nucleotide pool-specific activities is to normalize pulse-label incorporation for one class of RNA species to that of another species or to total pulse-labeled RNA, as in rs/rt values reflecting the fraction of stable RNA synthesis normalized to total RNA synthesis (403, 405, 406, 407). Measurement of RNA accumulation by chemical reactivity is another solution (128, 412). Measurement of rates of gene expression by reporter gene activity also eliminates this specific problem but may introduce others. Degradation of rRNA has been observed in stationary-phase and slowly growing cells (139, 140, 201, 339). Mild RNA degradation is reported during the stringent response (100, 330, 403), and severe RNA degradation occurs during carbon source starvation (312, 505).
When overall cellular transcription is dominated by activities of extremely strong phage T7 or λ promoters, the ensuing transcripts have been found to display stringent control whether the transcripts are of rrnB, phage T7 gene 10, total T7 phage DNA, or even plasmid DNA (511). These observations were interpreted as indicating that stringent control may act nonspecifically at some level that becomes rate limiting for very active transcription. One candidate might be steps involved purine nucleotide synthesis where (p)ppGpp exerts metabolic inhibition (Table 1). A key question is the quantitative extent to which this special high-strength transcription condition approximates the natural one in which overall cellular transcription is dominated by rRNA and tRNA synthesis.
Effects of inhibitors of (p)ppGpp accumulation have often proven difficult to interpret. Early studies of the inhibitory effects of rifampin led to the suggestions that RNA polymerase activity was the source of ppGpp synthesis (508) or that initiation of protein synthesis was directly involved (290; but see reference 478). Later experiments demonstrated that the basis for inhibition was instead requirements of the RelA-dependent reaction (107, 120, 289). Early experiments with trimethoprim, an inhibitor of fMet-tRNA formation, were initially interpreted in terms of a unique participation of protein synthesis initiation in the ppGpp synthetic reaction. Later studies revealed that trimethoprim inhibition of dihydrofolate reductase resulted in another dominant effect of C1 metabolism inhibition, namely, a starvation for methionine and glycine as well as RNA precursors (425, 426). An excess of one-carbon metabolites also causes problems for the cell (79). A cellular mechanism for sensing C1 metabolic stress is thought to be mediated by another likely signal nucleotide, 5-amino,4-imidazole carboxyamide riboside 5'-triphosphate (32). The existence of a folate stress regulon controlled by this nucleotide in E. coli has been disputed (392). Gramicidin and polymyxin were found to provoke ppGpp accumulation by inhibiting degradation (50). Cyclophosphamide inhibits protein synthesis by limiting aminoacylation of Leu-tRNA (117). Leu-tRNA charging behavior is itself peculiar even during overt leucine starvation (365). Addition of methanol, ethanol, or propanol to growing cells provokes relA-dependent ppGpp synthesis, whereas butanol and pentanol seem to impair ppGpp degradation and stimulate synthesis (305). As noted in the section on spoT, the (p)ppGppase mediated by this gene product is sensitive to a Mn2+ chelators such as tetracyclines.
We favor the view that (p)ppGpp is probably the major effector of the stringent response and conversely, when some experimental condition provokes changes in (p)ppGpp levels, regulatory consequences are likely. This is distinct from the view that if a regulatory event can be correlated with changes in (p)ppGpp levels under some conditions, then it can be regulated only by (p)ppGpp. For an obvious example, growth inhibition predictably accompanies manipulated elevations of (p)ppGpp without nutrient limitation (411, 414, 415) and steady-state growth rates achieved in a variety of ways can be correlated with (p)ppGpp levels (407), yet inhibition of growth can occur without participation of ppGpp.
There have been recurrent reports of differential effects of starvation for specific amino acids, resulting in different levels of (p)ppGpp obtained and different degrees of inhibition of RNA accumulation (reviewed in reference 101). Starvation for specific amino acids or the presence of groups of amino acids can have complicating secondary effects on other aspects of cellular metabolism. A role for codon usage dependence in the mRNA being translated that can confer insensitivity to starvation for specific amino acids was mentioned earlier (496). Starvation for most individual amino acids does, however, result in fairly uniform effects on RNA accumulation.
Although carbon and energy source transitions provoke ppGpp in both relaxed and stringent strains, it was noticed in early experiments that ppGpp levels apparently did not correlate well with effects on stable RNA accumulation (181, 504). These early studies utilized a change of carbon sources, which produced diauxic lags. Subsequently, the use of α-methylglucoside as a competitive inhibitor of glucose uptake has allowed more precise timing and quantitation (178). It was found that RNA accumulation abruptly stopped after glucose uptake was limited in both relaxed and stringent strains and that it did not correlate with variations in ppGpp levels observed. Later measurements of the contributions of synthesis and degradation to RNA accumulation revealed that marked RNA degradation occurs during glucose starvation; indeed, measurements of the instantaneous rRNA synthetic rates have largely restored the correlation thought to be anomalous (312, 407). Interestingly, a GTP derivative with a purine ring modification, called phantom spot (136), was shown to disappear simultaneously with activation of RNA degradation and was proposed to be an alternative regulator to ppGpp (136).
Anomalies in the correlation between rRNA accumulation responses and (p)ppGpp levels have been reported for three mutations affecting his operon expression that involve effects on several tRNA species: hisU, hisW, and hisT. Both hisU and hisW mutations are now known to be alleles in gyrB and gyrA, respectively (402). Sensitivity of his operon expression to changes in DNA gyrase activity are apparently mediated through the supercoiling dependence of the hisR promoter, leading to tRNAHis expression (111); the readthrough of the his attenuator is severalfold sensitive to even a 50% reduction in levels of tRNAHis (13, 49). The hisU mutation is apparently defective in its ability to activate rRNA degradation during carbon starvation, giving rise to accumulation of rRNA despite accumulation of (p)ppGpp (56, 84). The hisW333 mutation is also defective in RNA accumulation. The hisU1820 and the hisW333 mutations give phenotypes that are strikingly similar to one another (50) with lowered levels of several tRNA species, including tRNAHis, and impaired derepression of his and ilv expression (40, 49, 56, 85, 86, 270).
In S. typhimurium, hisT mutants are defective in tRNA pseudouridylate synthetase (76), leading to a defect in forming pseudouridine in the anticodon loop of a large number of tRNA species, including tRNAHis (422, 474). As discussed above, this defect results in altered attenuator readthrough and pleiotropic effects on several amino acid biosynthetic pathways. Spadaro et al. (434) discovered two anomalies with a stringent hisT hisD double mutant starved for histidine that did not occur after starvation for amino acids whose anticodon loops lacked pseudouridine. One anomaly is that (p)ppGpp did not accumulate. This anomaly could be explained as an impairment in mRNA-dependent ribosomal A-site binding of the hypomodified tRNA. However, the other anomaly, called stringency without ppGpp accumulation, was that while (p)ppGpp did not accumulate after histidine starvation, an abrupt cessation of RNA accumulation did occur in a stringent host but not in a relA mutant host. The explanation for this behavior is still unknown.
Molin et al. (312) uncovered yet another anomalous feature of the relationship between ppGpp and rRNA synthesis. Lowered but significant rates of rRNA synthesis persist long (30 min) after the onset of glucose starvation when ppGpp levels are quite high in both relaxed and stringent strains. Measurements of rrn P1 and P2 activities in vivo under these conditions reveal that while P1 activity is severely inhibited, P2 promoter activity persists (409). Thus, partial inhibition of rRNA synthesis can be accounted for by the ppGpp sensitivity of the P1 promoter, and the persistence of low but significant rates of rRNA synthesis can be accounted for by the ppGpp insensitivity of the P2 promoter (229, 409).
There is a report (122) that cellular levels of ppGpp do not correlate with rRNA accumulation in the transient period when stringent cells are subjected to a shift-up from acetate to glucose plus 19 amino acids; ppGpp levels fall and do not return to levels above background for 30 to 40 min, yet rRNA accumulation begins minutes after the shift and quickly reaches the new steady-state rate well before ppGpp. Subsequent measurements have demonstrated that the correlation with ppGpp was restored, with the rapidity of the response leading in part to adoption of the proposal that RNA polymerase partitioning is a mechanism of action of ppGpp (278, 407).
Peptidoglycan is a rigid cell wall macromolecule essential for resistance to osmotic pressure and in the progress of the cell cycle (see chapters 6 and 68). The last steps of peptidoglycan synthesis are catalyzed by several enzymes located in the inner membrane that are able to bind the β-lactam antibiotics (penicillins and cephalosporins) and are therefore called penicillin-binding proteins (PBPs). It is likely that a balance of synthesis and degradation, accomplished by enzymes called autolysins, must be required to maintain a constant cell wall thickness and to prevent cell lysis. Both hydrolysis (see below) and synthesis (208, 209, 210) have been reported to be under stringent control.
Nongrowing bacteria are more resistant than growing cells to the killing effects of penicillins. Amino acid-starved cells are penicillin tolerant if starvation is accompanied by the induction of ppGpp (157, 247, 367, 479). It was proposed that peptidoglycan synthesized in stringent cells is more resistant to autolysins (157). This proposal is supported by reports that amino acid starved or slowly growing stringent cells have a different peptidoglycan composition which is more resistant to hydrolysis in vitro (471, 472). However, modification of peptidoglycan composition was not detected by others (479), who proposed instead that when peptidoglycan synthesis stops, degradation stops as well. In this view, the inactivation of PBPs by β-lactam antibiotics might not inactivate the autolysins and thereby account for cell lysis. It is possible that both mechanisms can occur because certain inhibitors of cell wall synthesis are able to induce autolysis in stringent strains deprived for amino acids for long periods (470, 473).
Several conditional Lyt– mutants which become tolerant to β-lactam antibiotics, i.e., lysis insensitive, without amino acid starvation at restrictive temperatures have been described (168, 180, 249, 420). This phenotype is suppressed by chloramphenicol or a relA2 mutation (248); it was observed that transfer to high temperatures induces an increase in the (p)ppGpp levels (249). This behavior was interpreted to mean that the lyt gene products regulate the activity of the RelA protein. Determination of the identities of these putative regulators and whether they do contribute to (p)ppGpp regulation during normal growth will be awaited with great interest.
Unlike the case for wild-type E. coli strains, microscopic observation of (p)ppGpp0 strains shows that the bacterial population is heterogeneous in size, including small cells, elongated cells, and long filaments (510) reflecting partial inhibition of cell division. This phenomenon may be indirect, for example, caused by perturbation of the DNA replication or partitioning process. The (p)ppGpp0 strain phenotype could be due to a (p)ppGpp requirement for the expression of one or more genes essential for cell division, or a deficiency of (p)ppGpp could induce the synthesis of endogenous cell division inhibitors. In fact, there is growing evidence that (p)ppGpp participates in two processes: cell division and initiation of DNA replication (see next section).
Most of the results attributing a role for ppGpp in cell division come from studies of effects of mecillinam (291), an antibiotic of the penicillin family which specifically binds and inactivates PBP2 (437, 438). As a result of PBP2 inactivation, the cells become spherical, showing that PBP2 is required for the maintenance of the rod shape. PBP2 seems also to be essential for cell survival, since mecillinam is lethal for wild-type strains and a pbpA deletion is lethal for wild-type strains (212, 296, 346). The same conclusion comes from the inability to transduce a deletion of pbpA, the gene coding for PBP2, into wild-type strains (346). The lethality of PBP2 inactivation has recently been attributed to cell division inhibition (487). However, mecillinam-resistant (Mecr) mutants can be isolated (11, 494), and in some of these, the pbpA deletion is no longer lethal (486). Instead, the mutant cells divide and survive as spherical cells. Two such mutants, called lov-1 and lovB, have been shown to unexpectedly possess partially defective aminoacyl-tRNA synthetases, and their Mecr phenotype was shown to be relA dependent (42, 486). In a wild-type strain, decreasing the growth rate or IPTG induction of ppGpp also results in a Mecr phenotype(226, 486). Finally, cells with elevated ppGpp levels in a Δ relA strain carrying a spoT mutation also become Mecr during growth (D. Vinella and M. Cashel, unpublished data).
Vinella and D’Ari (485) uncovered an rpoB mutation which failed to exhibit mecillinam resistance on minimal media but was resistant when plated on media which caused mild amino acid starvation (43, 44, 486). The mutation was designated rpoB (Fts) because it also caused filamentation at high temperatures (485). Both phenotypes of the rpoB (Fts) strain are completely suppressed by increased ppGpp pools. Since mecillinam sensitivity could be altered by an rpoB mutation, we examined a set of rpoB mutations differing in apparent sensitivity to (p)ppGpp, mentioned above in the section on rpoBCD, for the ability to confer the Mecr phenotype to wild-type and (p)ppGpp0 strains; these included 17 sequenced Rifr mutations differing in the ability to suppress the multiple amino acid auxotrophy of (p)ppGpp0 strains, as well as several Rifs rpoB alleles with suppressing activity. Four of these mutations were found to confer mecillinam resistance in both hosts, and two mutations had this activity in a wild-type but not a (p)ppGpp0 strain, as if the latter reflects a weak requirement for ppGpp (Vinella and Cashel, unpublished data). While all six mutations giving mecillinam resistance also suppress multiple amino acid requirements of (p)ppGpp0 strains, not all auxotrophic suppressor alleles confer mecillinam resistance. Two rpoD alleles with (p)ppGpp0-suppressing activity, described in the section on rpoBCD, also conferred mecillinam resistance. It appears that correct processing of the cell cycle and the cell division process can require a signal involving ppGpp that might be in common with a signal for ppGpp effects on amino acid biosynthesis or sensitivity.
One explanation for this behavior is that PBP2 inactivation may cause a defect in some element needed for cell division and that ppGpp can compensate for this defect, possibly at the level of transcription. The ftsZ gene seems a likely candidate. The FtsZ protein appears to be the main target for cell division regulation and has been reported to be limiting for the process (138). It is suggested that a minimal number of FtsZ proteins are required to form the FtsZ ring essential for an early step in cell division (see chapter 102). The ftsZ gene is in a complex transcription unit along with the ftsQ and ftsA genes; all three are essential for cell division. The transcription of ftsZ is driven from at least five promoters (390, 519). At least two of the promoters are activated by slow growth, as judged by lacZ fusion behavior and Western blot (immunoblot) analysis (5, 389). At least one of the promoters is probably σ S dependent like bolA gene expression; bolA is a determinant of cell shape (4, 5, 260). Since expression of rpoS encoding σ S is induced by ppGpp (143), the possibility exists that ftsZ expression is induced by ppGpp through induction of σ S. The overproduction of FtsA, FtsQ, and FtsZ proteins (487) or only the FtsZ protein (D. Vinella, D. Joseleau-Petit, A. Jaffe, and R. D’Ari, unpublished data), does confer mecillinam resistance. This finding suggests that the effects of ppGpp might be solely explained by indirect induction of ftsZ expression. An interaction of RelA and ftsZ has been noted previously: overproduction of the RelA protein gives partial suppression of the phenotype of an ftsZ84 mutation (147); a similar suppression activity toward ftsZ84 is shown by the lov-1 and lovB mutations (Vinella et al., unpublished data).
Very recently, effects of IPTG induction of (p)ppGpp to levels approaching those achieved during the stringent response have been reported (415). It was found that septum formation was not blocked under these conditions; instead, residual cell division gave smaller cells approaching sizes of those of overnight stationary-phase cells.
In wild-type E. coli, DNA replication starts at a single site, called the origin or oriC, and proceeds bidirectionally until the two replication forks meet at the opposite terminus, or ter site, where several events prepare daughter chromosomes for their subsequent partition, which takes place just before cell division (see part IV). It is known that the rate of DNA replication is controlled by growth rate at the level of initiation, with initiation frequency proportional to growth rate. Initiation is the only step which requires de novo protein synthesis after each round of replication, as shown with protein synthesis inhibitors or by amino acid starvation. It has also been observed that the initiation mass, defined as the ratio of number of oriC copies per cell mass, is remarkably constant when replication starts and is relatively independent of the growth rate at all but very slow growth rates (99). Accordingly, several models suggesting that an initiation potential has to accumulate after each initiation event at a rate directly correlated with growth rate and thus to rates of protein synthesis have been proposed (99, 176, 371). The fairly constant initiation mass of a wild-type strain is affected in an isogenic (p)ppGpp0 strain which shows a constant decrease in initiation mass when growth rate increases (197). However, a similar relationship between initiation mass and growth rate has been reported for a wild-type strain, AB1155 (507).
There is abundant evidence that at least part of the de novo protein synthesis requirement for initiation is explained by a need for DnaA protein synthesis and that DnaA protein is a major determinant of the initiation mass (15, 187, 242, 283). oriC activation itself involves at least two components: (i) the binding of 20 to 40 molecules of DnaA protein to four 13-mer repeats (dnaA boxes) in oriC that contribute to opening the DNA helix in the course of assembly of the initiation complex (46), and (ii) the methylation of GATC sites in oriC and the adjoining mioC gene, which is implicated in coordination of the timing of initiation. In addition, oriC function on plasmids is stimulated by the transcription from promoters of genes flanking oriC (mioC and gid) that could, possibly indirectly, lead to R-loop formation and give rise to RNA primers for initiation of replication (see review in reference 525).
Evidence exists of (p)ppGpp involvement in two of the three oriC activation requirements. A study demonstrating activation of oriC function by concurrent transcription of the genes for gid and mioC on plasmids is accompanied by in vitro evidence of ppGpp inhibition of gid and mioC transcription (344). In addition, the abundance of in vivo transcripts terminating in the oriC region has been shown to be stimulated by chloramphenicol and inhibited by amino acid starvation of a stringent strain (395, 396), and mioC mRNA abundance was shown to display growth rate control (65). The abundance of DnaA protein was also shown to be subject to growth rate control (65), although this conclusion has been contested (175). Transcription of dnaA is from two promoters, and both transcripts were found to be growth rate dependent, subject to negative stringent control in a relA-dependent manner, and inhibited by elevated ppGpp in exponentially growing spoT mutants (66). Finally, when (p)ppGpp is gratuitously induced with IPTG to high levels, evidence of inhibition of replication initiation (415) is reported to accompany inhibition of growth (414).
Guzman et al. made the interesting observation that amino acid starvation inhibition of the replication initiation of either the bacterial chromosome or minichromosomes is relA dependent only if starvation is imposed for isoleucine, as opposed to other amino acids; if relA strains are starved for other amino acids, inhibition of initiation occurs (170). Guzman et al. deduced that this reflects the participation of a protein lacking isoleucine and have proceeded to isolate such a protein encoded by a gene at 73 min that is a putative calcium-binding protein (171).
There are also indications that (p)ppGpp regulatory effects might be exerted on origins of replication other than oriC. In an rnhA mutant lacking RNase HI activity, initiation of DNA replication occurs at sites other than oriC, called oriK sites (90); initiation activity from these origins continues in the presence of chloramphenicol and has been referred to as constitutively stable DNA replication (241, 492). It was proposed that certain long DNA-RNA hybrids are synthesized at several sites which might be used as a template for starting replication if they are stabilized by inactivation of RNase HI (492). This type of initiation is also relA dependent when cells are starved for tryptophan, suggesting that it might be under negative control by (p)ppGpp (492). RelA-dependent effects on replication of plasmid pBR322, with a ColE1 origin, were first described by Hecker et al. (188), verified (187, 194), and extended to include reports of similar regulation during amino acid starvation of replication from the λ origin (497) as well as orip15A, oripSC101, and oriV with instances of amino acid-specific starvation differences (200). Analogous studies with IPTG induction of (p)ppGpp have verified apparent inhibitory effects on replication due to plasmid-borne ColE1, oriC, λ origin, and oripSC101 but not oriV (193).
To summarize the last two sections, strong evidence has emerged that (p)ppGpp may be involved in bacterial cell cycle regulation in addition to its probable effects on other cellular components, perhaps allowing cells to reduce themselves to a minimal unit required for survival.
Relaxed and stringent strains have been repeatedly found to show differences in translational accuracy occurring during the starvation period. Evidence included finding a thermolabile β-galactosidase with lower specific activity in relaxed strains (124), finding a misincorporation of cysteine into flagellin specific for arginine starvation, and detecting charge variants on two-dimensional protein gels, a phenomenon called stuttering (238, 358, 359). By using various combinations of spoT and relC mutants, it could be shown that mistranslation was eliminated by the presence of ppGpp rather than some other property of the RelA protein (97, 238). Accumulation of mistranslated proteins during a prolonged starvation of a relaxed strain has been often speculated as possibly contributing to their much longer growth lag after reversal of starvation than in a stringent strain.
It was originally thought that this ability of ppGpp to reduce translational errors was due to a direct effect on the ribosome or protein synthesis. This notion has been challenged by in vivo evidence that leads to the conclusion that inhibitory effects of ppGpp induction on protein synthesis are indirect and operate at the level of reducing mRNA availability for protein synthesis (432). The system used involves measurements of overall and specific translation rates of preinduced, long lived lacZ mRNA during the period of slow decay of the mRNA after reversal of amino acid starvation of a spoT strain (432). The protein chain elongation rate was found to be unaffected by ppGpp under these conditions as well as with the wild-type strain during the stringent response. The measurements of mistranslated charge variants revealed them to persist in the presence of ppGpp before, but not after, the lacZ mRNA decayed. It is concluded that (p)ppGpp inhibits mRNA abundance because of (p)ppGpp-dependent pausing leading to protein synthesis inhibition. Limited protein synthesis allow higher levels of charged tRNA which, together with limited mRNA, result in a lower frequency of hungry codon encounters and reduce mistranslation (432). In the starved relaxed strain, mRNA levels remain high and mistranslation persists.
This interpretation is consistent with the finding that ppGpp reduces by 65% the synthesis of RNA-stimulated protein synthesis in an in vitro protein synthesis system (446). This measurement was a postscript addition to a study reporting 90% inhibition of protein synthesis by IPTG induction of (p)ppGpp (446).
When cells are deprived of a required amino acid, protein synthesis continues at reduced rates, allowing derepression of the appropriate biosynthetic operon, and net protein accumulation is minimal. Preexisting proteins are degraded to perhaps provide the missing amino acids. Intracellular proteolysis during the stringent response is thought to be governed in part by activation of an ATP-dependent protease (68) originally thought to be encoded by the lon gene, because of lon stimulation during heat shock (153, 218). The identity of the protease activated during the stringent response is, however, now uncertain (S. Gottesman, personal communication).
Inconsistent reports of differences in polysome lengths and stabilities shown by relaxed and stringent strains have appeared. The prevailing view is that starved relaxed strains have longer and more stable polysomes (77, 103, 294).
There have been a number of reports of a lack of correlation between (p)ppGpp levels and RNA control during abrupt temperature shifts. Shifts of E. coli from 23 to 40°C provoked an only partially relA-dependent, transient accumulation of ppGpp but not pppGpp, features reminiscent of an energy source starvation and spoT inhibition (135). Furthermore, RNA accumulation persisted for at least 20 min despite ppGpp levels nearly as high as those seen during the stringent response. Temperature downshifts from 40 to 20°C gave a rel-independent drop in ppGpp levels which were unresponsive to amino acid starvation, yet RNA accumulation was reduce fourfold in the stringent strain (354). The apparent rel-dependent regulation of RNA accumulation despite either high levels of ppGpp at 40°C or low levels of ppGpp at 20°C was taken to indicate that ppGpp is not the real effector (135, 425). Instead, a related nucleotide, ppGp, was proposed for this role in temperature transitions (354, 356, 357).
Subsequent studies challenged measurements of stable RNA synthesis rates, taking into account uptake barriers, degradation contributions, and relA-independent contributions to ppGpp synthesis (63, 151). Measurements of rs/rt changes suggested that rs /rt responded to a relA-independent increase in RNA chain elongation rates and an opposite and compensating effect from relA-dependent ppGpp synthesis (407). Analogous disparities were found between the concentrations of ppGpp required to give equivalent levels of RNA synthesis inhibition in steady-state-growth cells and those of valyl-tRNA synthetase mutant cells subjected to restrictive temperatures (114). This disparity was also resolved when contributions from degradation were evaluated along with instantaneous rates of rRNA synthesis (403, 405, 406, 407).
A moderate shift of a valyl-tRNA synthetase mutant from 29 to 33.5°C was found in a relA-dependent manner to activate transcription from heat shock promoters as well as to give increases in labeling of GroEL, DnaK, and other heat shock proteins ranging from two- to sevenfold (167). During the stringent response without temperature changes, heat shock proteins are reported as either mildly elevated or not elevated (167, 478). Furthermore, finding a normal heat shock response in a (p)ppGpp0 strain has ruled out either a necessary or sufficient role for (p)ppGpp (478).
VanBogelen and Neidhardt noted that apparently specific induction of heat shock proteins without temperature shifts can be provoked by some inhibitors of protein synthesis whereas other inhibitors repress heat shock proteins (478). A reciprocal behavior was shown by cold shock, induced by shifting temperatures from 37 to 10°C, which repressed heat shock proteins and induced of a set of cold shock proteins (224). Cold shock is also associated with inhibition of protein synthesis initiation, a lowering of (p)ppGpp levels, and a prolonged lag before slower growth resumes at the low temperature during which synthesis of many components of the transcriptional and translational machinery occurs (221; see review in reference 222). Studies with (p)ppGpp0 strains reveal that the major cold shock protein, CspA, is constitutively expressed along with other cold shock proteins at intermediate temperatures but without cold shock (221). In this state, cells also behave as if they are already cold adapted in that they grow without a lag period when cold shocked. The CspA protein itself may have multiple regulatory activities, including activation of transcription of hns and gyrA owing to recognition of DNA binding sites in the promoter regions of these genes (223, 262). In addition, three other CspA homolog genes have been found, and an RNA chaperonin-like function has been proposed for CspA (see review in reference 222).
Studies of the ability of various ribosomal inhibitors to elicit a cold shock-like or a heat shock-like response have revealed a simple classification (478). Inhibitors that are deduced to either block the A-site or fill it with aminoacyl-tRNA (chloramphenicol, erythromycin, fusidic acid, spiramycin, or tetracycline) give a cold shock-like response. Inhibitors deduced to leave the A-site empty (kanamycin, puromycin, or streptomycin) give a response like heat shock. These findings have led to the proposal of a ribosome sensor model in which translational impairments result in an altered physical state of the ribosome, or some derived ribosomal product, that comprises a signal linking the environmental stimulus to a response, which includes correcting the source of the translational impairment (478). While this model was discussed specifically with respect to heat and cold shock, it potentially applies to (p)ppGpp regulation as well. Inhibitors of protein synthesis initiation diminish (p)ppGpp accumulation responses, and this is usually explained by the effects on restoring tRNA charging and the subsequent effects on RelA activity. It now seems plausible that these experiments can be reinterpreted in terms of ribosome sensor function. An early experiment demonstrated that chloramphenicol and fusidic acid abolished (p)ppGpp accumulation during amino acid starvation of a stringent strain but that this effect was reversed when a conditional aminoacyl-tRNA synthetase mutant was shifted to restrictive temperatures (233). Since these two antibiotics are expected to inhibit (p)ppGpp accumulation by the ribosomal sensor model, this could mean that the ribosomes must be allowed partial function for their role as sensors to be operative. Of course, shifting the tRNA synthetase mutants to nonpermissive temperatures could have effects both from the temperature shift alone and from the charging defect that overcomes the effect of the drugs on total tRNA charging.
The question often arises as to whether the regulatory functions ascribed to (p)ppGpp are due instead to nucleotides structurally related to (p)ppGpp, such as phantom spot and ppGp, that have been suggested as better correlated with rRNA accumulation responses than (p)ppGpp. We have argued above that the conditions of temperature shifts and carbon source starvation that reveal these anomalies can be viewed with some suspicion. Nevertheless, the timely appearance of these unusual nucleotides raises the possibility that there is regulatory specialization among distinct but metabolically related nucleotides, including chimeric regulators, as apparently can occur in the case of simultaneously inducing the stringent response together with ApppA formation, discussed above. Alternatively, a family of multiple structurally related nucleotides might have common regulatory targets and modes of action. It seems unlikely but possible that the whole (p)ppGpp family of related nucleotides are simply pathway intermediates or breakdown products of an as yet undiscovered true effector whose lability allows visualization of only degradation products. Accurate measurement of cellular (p)ppGpp poses a challenge because of rapid metabolic responsiveness together with the chemical lability of the 3'-pyrophosphate group in both acid and alkali (60) as well as vulnerability to enzymatic degradation at acidic and neutral pH (83). While formic acid extraction and freeze-thaw cycles are still useful for measurements of (p)ppGpp with 32P-labeled cells (57), Lagosky and Chang (252) have devised what is claimed to be a more efficient lysis procedure that involves exposures to neutral pH during lysozyme lysis. Little and Bremer (277) have devised a method which uses formaldehyde fixation of cells followed by KOH lysis and high-pressure liquid chromatographic resolution of ppGpp, but the migration of pppGpp is uncertain and not reported in these assays. Bochner and Ames (33) have described a generalized search protocol for surveying nucleotides that uses acid extraction. Radioimmunoassays with high sensitivity have been developed (125, 289), but the problem of degradation during extraction remains.
Commercial sources of (p)ppGpp currently range from uncertain to nonexistent despite frequent catalog listings; it therefore may be necessary to prepare sizeable quantities of (p)ppGpp in vitro. Crude ribosomes can support (p)ppGpp synthesis in large preparative reactions without supplementation with mRNA or tRNA (55). Exploiting recent developments can facilitate (p)ppGpp preparation. For example, specific synthetic activities (per ribosome) are enhanced at least 50-fold by preparing ribosomes from strains in which the RelA protein is induced to high levels, and the relative yields of pppGpp and ppGpp can be influenced by using conditions and mutants leading to defective (p)ppGpp degradation.
Currently, there are three fundamentally different views of how growth rate control can be achieved. Two of these include a regulatory role for ppGpp. The promoter selectivity model involves a direct effect of (p)ppGpp on promoter activity. The passive growth rate control model, reviewed above and called a metabolic growth rate control model by the authors (213), involves indirect inhibitory effects on stable RNA promoters arising from lowering of free polymerase concentrations. Lowering of RNA polymerase concentrations is proposed as due to sequestering the enzyme at ppGpp-dependent pause sites on mRNA transcripts (213). A third view is called the ribosomal feedback model of growth rate control. It proposes an inhibitory role on stable RNA promoter activity by translating ribosomes that is argued to be independent of (p)ppGpp (reviewed in references 218 and 338). We shall briefly discuss how growth rate control has been defined and the consequences of changing growth rates in terms of cell composition (see references 239 and 323 and chapter 97).
The classical description of growth rate regulation employed S. typhimurium. Schaechter et al. (413) demonstrated that cell composition (RNA per cell, DNA per cell, mass per cell, and cell size) measured during balanced growth in different media varied in a systematic way as growth rates changed at 37°C. The growth conditions used to measure growth rate-regulated products bear careful consideration. Different (shallower) growth rate-dependent relationships were found when steady-state growth rates were restricted by the concentration of nutrients, i.e., dilution rates in chemostats, compared with when growth rates were limited by types of nutrients, i.e., unrestricted growth (413). At an identical growth rate achieved under restricted and unrestricted growth conditions, results in terms of bacterial cell composition may not be the same. Measurements of cell composition at a given unrestricted growth rate can be reproducibly made when growth is balanced, i.e., when levels of RNA, DNA, protein, etc., increase by the same factor over a given time interval (52, 413). The relationship found for different states of balanced growth was that RNA, DNA, and mass per cell increased exponentially with growth rate but with different slopes: RNA > mass > DNA (323, 413); cells growing in rich media are larger and contain much more RNA. For example, it is now appreciated (see chapter 97) that in going from a doubling time of 100 min to one of 24 min, an approximately 10-fold increase in the number of ribosomes per cell occurs with a proportionate increase in tRNA, maintaining a ratio of approximately 10 tRNA molecules per ribosome. Since there is relatively little change in the peptide chain elongation rate per ribosome, a high efficiency for ribosome function is found over this range of growth rates. Since the synthesis of ribosomes is limited by the synthesis of rRNA (see chapter 90), the increase in stable RNA with growth rate reflects a proportionate increase in the concentration of ribosomes with growth rate. When cells are shifted from a low to a high growth rate, RNA synthesis is quickly increased to approach the characteristics of the new medium, along with protein synthesis, while DNA synthesis and cell division continue at the slower rate for a longer time before shifting to the new rate (52, 239). When cells are shifted from a fast to a slow growth rate, RNA and protein syntheses are the first to show inhibition, which persists until the new slower rate is achieved (293).
Growth rate control has been viewed in two ways; both are consistent with the classical observations of growth rate-dependent changes of cellular content of RNA and DNA and cell mass. If the key elements of growth rate-dependent changes are considered in terms of the distribution of total gene expression, then models emerge in which (p)ppGpp is to be involved. According to this view, the adjustment in bulk content of RNA and protein per se is not indicative of growth rate regulation; rather, it is the changing array of mRNAs and the proteins they encode, together with the changing amounts of stable rRNA and tRNA relative to other cell components at different unrestricted growth rates, that reflects a growth rate-dependent regulation of macromolecular synthesis. This view includes the concept that the growth rate-dependent increase in stable RNA relative to other cell components occurs at the expense of bulk mRNA synthesis, i.e., due to a redistribution of a limiting amount of RNA polymerase over stable RNA and mRNA genes (69). Alternatively, if the key element of growth rate control is simply the empirical relationship between the abundance of stable RNA, cell mass, and DNA as a function of growth rate, then a model emerges that has been proposed as not involving ppGpp.
The levels of accumulation of some other cellular components respond to growth rate in the same, or nearly the same, manner as levels of rRNA and bulk tRNA and are therefore judged to be under growth rate control; examples are translation elongation proteins, but this feature applies especially to EF-Tu, whose concentration is maintained in a constant ratio with tRNA (125, 307; see chapter 97). As mentioned earlier, components displaying growth rate control are frequently, but not invariably, subject to stringent control (17, 30, 53, 313). An early review (326) called instances of growth rate regulation but not stringent control by the same name as the mechanism proposed by Jensen and Pedersen (213), namely, metabolic, to distinguish it from stringent control.
We shall first review evidence relating to ppGpp participation in growth rate control of rRNA. Neidhardt (321) first called attention to the fact that a relaxed mutant strain displayed normal growth rate regulation of stable RNA; the altered rRNA phenotype in the mutant was specific for amino acid deprivation. It was later shown that normal growth rate control of RNA accumulation in relaxed mutants was accompanied by normal regulation of ppGpp levels (266). Ryals et al., in the course of deriving a mathematical description of the physiology of cell growth, measured the instantaneous rate of stable RNA transcription as a fraction of total transcription, i.e., rs/rt values. They encountered a constant relation between rs/rt values and ppGpp levels, in both relaxed and stringent strains, for variations in steady-state growth rates as well as growth rate transitions, with or without chloramphenicol, for amino acid limitation, and for temperature shifts of thermolabile aminoacyl-tRNA synthetase mutants (405, 406, 407). This correlation led them to propose that there was no distinction between stringent control and growth rate control; both regulatory events were ascribed to effects of ppGpp (18, 407). As ppGpp levels extrapolate to vanishingly small values, rs/rt values were measured as approaching 1.0 and overall transcription was predicted to be virtually completely dominated by rRNA and tRNA synthesis (18). Recent studies revealed that rs/rt values measured with (p)ppGpp-deficient strains (197) instead were much lower, 0.6, indicating a discontinuity in the transition between barely detectable ppGpp and none.
Systematically increasing levels (18, 195) of ppGpp gives an apparently linear inverse relationship with rs/rt values until a ratio of about 0.5 is reached; thereafter, further increases in ppGpp are associated with rs/rt values asymptotically approaching 0.3. It was argued in early proposals that the constancy of this relation is not achieved by variations in transcriptional elongation rates, in accordance with several reports of measurements at the time; instead, there was proposed a ppGpp- dependent inhibition of total RNA polymerase activity as well as its partitioning with respect to promoter selectivity among promoter classes. Early efforts to characterize the relative amounts of active and inactive RNA polymerase in growing cells suggested that most (70 to 80%) was probably not engaged in RNA chain elongation (419) and that DNA concentration is not limiting (69). Recently, this interpretation has changed, in accordance with the proposal of Jensen and Pedersen (213) and the demonstrations that ppGpp does slow mRNA chain elongation rates by about 50% during the stringent response (432, 433, 489, 491; see section on rRNA transcription). An effect of trapping RNA polymerase on mRNA transcripts due to (p)ppGpp-mediated pausing was judged to at least partially, if not fully, account for what was previously thought to be inactive RNA polymerase (195). The contribution of this effect as a function of growth rate variation has yet to be quantitatively assessed. The anomalously low rs/rt value unaffected by growth rate in the absence of ppGpp is now interpreted as due to the absence of such pausing, which leads to increased mRNA synthesis, which in turn affects the competitive distribution of RNA polymerase between stable RNA and mRNA transcripts (195). The reason for the discontinuity of rs/rt values between conditions giving extremely low ppGpp levels and those in which ppGpp is absent remains puzzling.
A promoter selectivity model still exists and proposes that (p)ppGpp directly, or at least not passively, inhibits transcription initiation at stable RNA promoters while specifically stimulating initiation at positively controlled mRNA promoters. This hypothesis is based on the premise that RNA polymerase transcription initiation properties are altered such that changes in the synthesis of stable RNA relative to total RNA synthesis appear as a strict inverse correlation with ppGpp concentrations (see above). Genetic evidence does implicate ppGpp as an effector of transcription. Support for ppGpp as a negative effector of growth rate regulation of rrn P1 promoters came from studies employing a chromosomal rrnB P1-lacZ reporter gene, which was used to screen for phenotypic LacZ "up" mutants; mutants with a lower basal level of ppGpp during steady-state growth were found (196). The isolation of mutations in RNA polymerase subunit genes capable of suppressing either ppGpp toxicity accompanying gratuitous induction of ppGpp or defects associated with the loss of ppGpp synthesis activity is also supportive (see section on rpoBCD). There are biophysical indications of interactions between ppGpp and RNA polymerase. One involves changes in circular dichroism (509), and the other comes from evidence with a fluorescent derivative of ppGpp with a single ppGpp binding site that is apparently identical for either holoenzyme or core RNA polymerase (378). The latter observation implies that ppGpp binding itself is independent of the σ 70 subunit. The ability to select rpoD mutants as suppressors of putative ppGpp-dependent functions could reflect an altered function of the σ 70 subunit when core RNA polymerase is allosterically modified by ppGpp binding. In vitro studies have now confirmed that mutations in σ 70, which suppress the amino acid auxotrophy of (p)ppGpp0 strains, do alter transcription initiation properties of RNA polymerase, but effects of adding ppGpp with this purified system have not yet been seen (Hernandez and Cashel, unpublished data). Caveats discussed in the section on rpoBCD still apply. In addition, effects on promoter clearance mediated by RNA polymerase subunit mutations could be easily imagined to alter passive responses to enzyme availability. It would seem that ppGpp-dependent effects on RNA polymerase in vitro and appropriate behavior by mutant enzymes will be critical elements for deciding between promoter selectivity effects and passive control effects.
An early model of passive control of growth rate (292, 293) involved the redistribution of RNA polymerase away from stable RNA promoters as a consequence of widespread derepression events accompanying growth on poorer media. The current passive control model of Jensen and Pedersen (213) instead involves differences in promoter saturation shown by mRNA and stable RNA promoters. For growth rate control, mRNA elongation rates are envisioned as decreasing with decreasing growth rate as the key determinant because of sequestering RNA polymerase. If the average mRNA promoter is saturated at low concentrations of RNA polymerase, raising the concentration of enzyme is predicted to saturate their expression; however, stable RNA promoters, because of their extreme efficiency in promoter clearance, are never saturated with RNA polymerase, and therefore these activities will be responsive to a wide range of RNA polymerase concentrations (432, 433, 589, 591). The quantitative relationship between inhibition of elongation and varying ppGpp levels during the slow decay after reversal of a stringent response in a spoT mutant is very different from inhibition measured over a range of growth rates. In the former case, an increase of ppGpp of 100 pmol per cell absorbance unit, a major effect in terms of growth rate control, gives about a 5% increase of inhibition of elongation rates (432), whereas a 40% reduction in chain growth rate was observed over a fourfold change in growth rate, probably accompanied by an increase in about 50 pmol of ppGpp per cell absorbance unit but not measured in this experiment (488).
An exception to the invariant correlation between instantaneous synthesis rates of stable RNA synthesis and ppGpp levels has been noted under conditions of pyrimidine limitation, i.e., restricted growth (490). This finding was taken to support the Jensen-Pedersen model. In these studies, a partial pyrimidine auxotroph was provided with different concentrations of a source of pyrimidine to achieve variations in steady-state growth rates. Both ppGpp pool levels and instantaneous rates of rRNA synthesis were found to decrease at greater and greater growth limitations. It was proposed that pyrimidine limitation could therefore bypass the effects of ppGpp by directly lowering RNA polymerase chain elongation rates, reducing free RNA polymerase concentrations, and thus reducing stable RNA synthesis according to the prediction of the Jensen-Pedersen model. There may be alternative explanations for this behavior. The first concerns the matter of supply and demand for aminoacylated tRNA and recalls an early experiment (107), already mentioned, showing that rifampin treatment diminished the ppGpp response to starvation for an amino acid in parallel with protein synthesis inhibition. When the stringent response was imposed instead with a conditional tRNA synthetase mutant, the ppGpp response was rifampin resistant so long as mRNA was judged to be present. The interpretation of the early experiment was that inhibiting mRNA synthesis led to curtailing the demand for aminoacylated tRNA, which made the supply in excess, thereby reducing the effects of amino acid starvation on uncharged tRNA. A similar explanation could be that effects of pyrimidine limitation on mRNA chain growth limit protein synthesis; increased tRNA charging gives lower ppGpp responses. It could also be that pyrimidine limitation to the extent that it alters RNA chain elongation rates also affects initiation rates, especially those of rRNA genes. It has been observed in vitro that the formation of a stable rrnB P1-RNA polymerase complex requires both initiating nucleotides ATP and CTP (the 5' terminal sequence of the rrnB P1 transcript is pppApCpU); high concentrations of ATP alone do not provide for a stable complex (190). However, only four of the seven rrn P1 promoters share the sequence giving slipping, and promoters other than rrnB have not been checked for this activity. A reduced rRNA initiation rate together with a reduced RNA chain elongation rate could lead to a lowering of ppGpp, which rather than being considered an anomalous response could be viewed as the appropriate cellular response, namely, an attempt to increase stable RNA promoter initiation by relieving ppGpp inhibition in response to a limitation in ribosome concentration.
A theory of growth rate control of stable RNA synthesis proposed by Nomura and colleagues as not involving participation of ppGpp is termed ribosome feedback control. The first observation leading to this proposal was that increasing the gene dosage of an intact rrn operon with a multicopy plasmid did not increase rRNA expression (161, 162, 219); instead, the activities of chromosomal and plasmid rrn operons, measured as synthesis of tRNA species encoded in spacer regions, were reduced in an equivalent fashion. This behavior was reasonably interpreted as representing a feedback signal that was a result of sensing ribosome concentration and negatively regulated expression of all rrn gene copies in an equivalent fashion. Negative feedback did not occur if the rrn operon on the multicopy plasmid was rendered nonfunctional by internal deletions (164, 219) or for other reasons (see reviews in references 337 and 338). It was proposed that the source of negative ribosome feedback could be functional ribosomes not engaged in protein synthesis; the abundance of this class of ribosomes could neatly be a sensor of the balance between ribosome concentration and demand for protein synthesis. The possibility that free ribosomes are the feedback signal was tested by artificially increasing the abundance of free ribosomes by limiting growth and increasing protein synthesis by limiting the expression of IF2 in cells in which the sole source of IF2 was an IPTG-dependent expression system (71). The excess of free ribosomes generated by growth rate-limiting IF2 concentrations did not inhibit but rather stimulated rRNA synthesis; the implication is that ribosomes giving a negative feedback signal are translating ribosomes (71). The strain used to alter the copy number of rrn genes has been reevaluated with respect to ppGpp levels and rs/rt values and argued to not result in exceptional regulatory behavior with respect to ppGpp levels and rRNA synthesis (19). Instead, increasing the rrn gene copy number was found to increase ppGpp levels and to lower rs/rt values; it was argued that the mechanism of ribosomal feedback accompanying an excess or deficiency of functional translating ribosomes was explained by ppGpp playing the key role as the feedback regulatory component as a consequence of functional ribosomes.
Recently, effects of deleting four (rrnABGH) of the seven rrn operons in a single strain have been measured under a single growth condition, Luria broth containing glucose (73). Under this condition, the reduced rrn copy number resulted in a 20% inhibition of growth rate; rRNA content was completely compensated for by a 2.3-fold increase in the expression of the three residual rrn copies. This compensation occurred by increasing transcription initiation and, surprisingly, by increasing elongation rates as well. Measurements of ppGpp concentrations made during minimal medium growth were found to be the same as in wild-type strains. The mechanism involved was interpreted as an example of ribosomal feedback control (73). This conclusion is clearly consistent with the evidence available. However, a systematic exploration of effects of varying growth rates could solidify this conclusion.
Gaal et al. isolated a set of 112 rrnB P1 mutants in the –88 to +1 region and, after transfer to λ trp-lacZ fusions, inserted them in the chromosomal lambda attachment site. They characterized 50 with respect to growth rate control by using a protocol of restricted growth, i.e., varying the Casamino Acids concentration to achieve differences in growth rates (94, 127). Among these were found mutants defining the upstream activation region as well as mutants in which growth rate control was clearly altered; the latter class was found to activate promoter activity. These mutants abolishing growth control consisted of a mutant with a T-to-A change at at position 33 (T33A mutant) restoring the –35 consensus, a mutant restoring a consensus 17-bp spacing, and a double mutant (C1T C15G). When the double mutant was dissected, the C1T mutant was found to give growth rate invariance that did not change promoter activity at fast growth rates but was more active at low growth rates (21). Growth rate-insensitive promoter mutants have been recently been shown to nevertheless be under stringent control (225). A mutational analysis of the promoter, using the same trp-lacZ fusion reporter in a lysogen, led to conclusions similar to those for the rrnB P1 studies, namely, that changes that activated the promoter tended to confer growth rate independence and also included changes restoring the –35 consensus sequence and altering the discriminator region (225).
It could be argued that stable RNA promoter mutants with activities that are growth rate invariant, but still subject to stringent control, behave as if they are insensitive to the small changes in low ppGpp concentrations that occur in balanced growth but still sensitive to high ppGpp concentrations achieved during the stringent response. This could be examined by systematically varying (p)ppGpp levels. A possible example of naturally occurring promoters of this class might be those for ribosomal protein genes. For example, the promoter activity of the spc operon appears insensitive to growth rate (278) but still subject to stringent control (63, 91, 92). The counterargument is that these early measurements of stringent control are not valid because the abundance of ribosomal protein mRNA is generally regulated by attenuation and mRNA turnover secondary to interactions between mRNA and their encoded ribosomal proteins (108; see chapter 90).
The existence of (p)ppGpp0 strains might be expected to resolve the question of involvement of ppGpp in growth rate control; instead, parallel studies of such mutants have sharpened the differences in interpretations. Gaal and Gourse (128) have measured RNA/protein ratios, as well as their rrnB P1 reporter activities, in such strains as a function of growth rate and found them much the same as in wild-type strains. Hernandez and Bremer (197), using the same strains, found RNA/protein ratios very similar to those found by Gaal and Gourse but also found that rs/rt values and the rrnB P1-lacZ fusion growth rate were independent in the (p)ppGpp0 mutant. Therefore, the increase in rRNA content in the (p)ppGpp0 strain at fast growth rates was concluded to be accompanied by a parallel increase in total (mRNA) synthesis (197).
How can these models of growth rate control be experimentally reconciled? Suggestions relating to tests of the Jensen-Pedersen proposal have been discussed in the section on transcription. Readers of papers describing growth rate control of the two rrnB P1-lacZ reporter activities are exposed to reasonable arguments that validate one or the other fusion. Comparative studies of the two fusion constructs as well as possible effects of their placement sites on the chromosome (74) would be interesting. Arguments are also evident as to whether rRNA promoter reporter activities should be normalized to protein or some other parameter when one is studying growth rate control (197; see chapter 90). Normalization per cell, as in the classical definition, is complicated by the tendency of (p)ppGpp0 strains to alter cell morphology (510). We would like instead to consider the uncontested measurement, namely, the increase in accumulated RNA/protein ratios. Although the slope of this increase with increasing growth rate may be debatable, both observations indicate that faster growth does increase the concentration of stable RNA in the absence of (p)ppGpp. The alternative view is that RNA/protein ratios are meaningless with respect to measurements of growth rate control because total protein is largely determined by ribosomes synthesizing protein at nearly constant rates. However, a key question is raised by the behavior of the C1T, mutant which suggests the existence of (p)ppGpp-independent effects of growth rate on rRNA abundance. Is this accomplished by multiple metabolic effects, or is there a single mechanism? If there is a (p)ppGpp-independent mechanism with a single control point responsible for this accumulation, like relA for stringent control, then it should be possible to obtain single-step host mutants in which this mechanism is inoperative. Since strong evidence does exists for (p)ppGpp regulation of stable RNA synthesis, such mutants might best be sought in a (p)ppGpp0 host. If major reductions in rrn gene copy number are verified as resulting in growth rate-dependent hyperexpression of residual operons without variation in (p)ppGpp content, then the putative mutant phenotype might be further amplified in such hosts. Obviously, the isolation of mutations altering growth rate dependence in the absence of (p)ppGpp would be a major step in defining this mechanism. The growing complexity of features controlling the activity of the rrnB P1 promoter alone, not to mention poorly understood features of the P2 promoter and the conserved P1 P2 tandem promoter arrangement, enhances the expectation that more might be going on than is now apparent.
Experimental questions also arise regarding the identity of the RNA that contributes to the high levels of total pulse-labeled RNA in the (p)ppGpp0 strain that parallels the fraction hybridizing to a ribosomal DNA probe as a function of growth rate, i.e., the basis for the rs/rt discontinuity mentioned earlier. It is of interest to know whether it is a special class of gene transcripts activated in (p)ppGpp0 strains, such as ribosomal protein mRNA, as opposed to a general increase in global mRNA transcription or decrease in degradation.
Several views of the stringent response have been considered here. The most focused concerns differences between relA and relA + behavior and so deals with the effects of amino acid starvation. With less focus, the stringent response includes (p)ppGpp responses to limitations of other nutrients and concerns spoT and relA mutant effects. By any definition, it is evident that changes in (p)ppGpp concentrations do have many effects on cell physiology. It is equally clear that a detailed account of a mechanism for these effects is uncertain.
One reason why the question of how (p)ppGpp works is elusive is the large number of cellular networks that it affects. A response to one perturbation seems almost inevitably accompanied by others. Synthesis and regulation of (p)ppGpp are intimately involved with ribosome and protein synthesis, the main intersection of cellular activity. A clear correlation evident under one condition has been seen to almost predictably become a lack of correlation under another condition. In some instances, after further analysis, the lack of correlation may be disputed, and instead the original correlation is restored to the status of a possibility.
The reason for the absence of compelling interpretations for all proposals is perhaps deceptively simple. No one has found a way to make cells permeable to (p)ppGpp. This property of (p)ppGpp disallows using the successful approaches of the past to uncover mechanisms of action of antibiotics or cyclic AMP. A deficiency of (p)ppGpp cannot be supplemented with (p)ppGpp, and mutants resistant or sensitive to (p)ppGpp cannot easily be obtained by direct selections; instead, until recently, they have been fortuitously encountered. In retrospect, the reason for the uncertainty applying to all proposals of one or another regulatory target for (p)ppGpp arises very simply from the lack of such mutants.
It is therefore necessary to exploit properties of the relA and spoT genes coding for steps in (p)ppGpp metabolism to devise ways of manipulating the intracellular concentrations of (p)ppGpp. The gpp gene may soon be added to the list of genes included in this approach in order to manipulate ratios of pppGpp and ppGpp. Alternatively, the capacity to form (p)ppGpp can apparently be abolished by deleting sources of synthesis. With such strains, one can easily ask whether a given response persists; it is already evident that while this approach is useful, it is of limited value because the deficiency of (p)ppGpp itself comprises a very major perturbation. Ideally, the prospective use of the ability to manipulate (p)ppGpp is to isolate the desired mutants so that in vitro behavior can be validated, not so that more correlations of cellular behavior can be generated with mutants. Given the fitful history of the search for a mechanism of action for (p)ppGpp, the criteria for proof of a hypothesis have become more stringent.
Thus, our preferred answer to a multiple choice question on, for example, the mechanism of growth rate control currently would be "none of the above." The proviso would quickly be added that each hypothesis has features that probably contribute so that "all of the above" is an equally acceptable response. Although unproven, it also seems plausible that regulatory redundancy exists and that multiple sources for control coexist. The fact that there are no real answers despite over three decades of work on the stringent response poses an engaging and obviously enduring challenge.
References
1. Adhya, S., and M. Gottesman. 1982. Promoter occlusion: transcription through a promoter may inhibit its activity. Cell 29:939–944.
2. Akiyama, M., E. Crooke, and A. Kornberg. 1992. The polyphosphate kinase gene of Escherichia coli. Isolation and sequence of the ppk gene and membrane location of the protein. J. Biol. Chem. 267:22556–22561.
3. Akiyama, M., E. Crooke, and A. Kornberg. 1993. An exopolyphosphatase of Escherichia coli. The enzyme and its ppx gene in a polyphosphate operon. J. Biol. Chem. 268:633–639.
4. Aldea, M., T. Garrido, C. Hernandez-Chico, M. Vicente, and S. R. Kushner. 1989. Induction of a growth-phase-dependent promoter triggers transcription of bolA, an Escherichia coli morphogene. EMBO J. 8:3923–3931.
5. Aldea, M., T. Garrido, J. Pla, and M. Vicente. 1990. Division genes in Escherichia coli are expressed coordinately to cell septum requirements by gearbox promoters. EMBO J. 9:3787–3794.
6. Alfoldi, L., G. S. Stent, and R. C. Clowes. 1962. The chromosomal site for the RNA control (R.C.) locus in Escherichia coli. J. Mol. Biol. 5:348–355.
7. Alfoldi, L., G. S. Stent, M. Hoogs, and R. Hill. 1964. Physiological effects of the RNA control (RC) in E. coli. Z. Vererbungsl. 94:285–302.
8. Almiron, M., A. J. Link, D. Furlong, and R. Kolter. 1992. A novel DNA-binding protein with regulatory and protective roles in starved Escherichia coli. Genes Dev. 6:2646–2654.
9. Altuvia, S., M. Almiron, G. Huisman, R. Kolter, and G. Storz. 1994. The dps promoter is activated by OxyR during growth and by IHF and σS in stationary phase. Mol. Microbiol. 13:265–272.
10. An, G., J. Justesen, R. J. Watson, and J. D. Friesen. 1979. Cloning the spoT gene project. J. Bacteriol. 137:1100–1110.
11. Aono, R., M. Yamasaki, and G. Tamura. 1979. High and selective resistance to mecillinam in adenylate cyclase-deficient or cyclic adenosine 3',5'-monophosphate receptor protein-deficient mutants of Escherichia coli. J. Bacteriol. 137:839–845.
12. Artz, S. W., and J. R. Broach. 1975. Histidine regulation in Salmonella typhimurium: an activator-attenuator model of gene regulation. Proc. Natl. Acad. Sci. USA 72:3453–3457.
13. Artz, S. W., and D. Holzschu. 1983. Histidine biosynthesis and its regulation, p. 379–404. In K. M. Herrmann and R. L. Sommerville (ed.), Amino Acids: Biosynthesis and Genetic Regulation. Addison-Wesley Publishing Co., New York.
14. Atherly, A. G. 1979. Escherichia coli mutant containing a large deletion from relA to argA. J. Bacteriol. 138:530–534.
15. Atlung, T., A. Lobner-Olesen, and F. G. Hansen. 1987. Overproduction of dnaA protein stimulates initiation of chromosome and minichromosome replication in E. coli. Mol. Gen. Genet. 206:51–59.
16. Aviv, M., H. Giladi, G. Schreiber, A. B. Oppenheim, and G. Glaser. 1994. Expression of the genes coding for the E. coli integration host factor are controlled by growth phase, rpoS, ppGpp, and by autoregulation. Mol. Microbiol. 14:1021–1031.
17. Baer, M., K. B. Low, and D. Soll. 1979. Regulation of the biosynthesis of aminoacyl-transfer ribonucleic acid synthetases and of transfer ribonucleic acid in Escherichia coli. V. Mutants with increased levels of valyl-transfer ribonucleic acid synthetase. J. Bacteriol. 139:165–175.
18. Baracchini, E., and H. Bremer. 1988. Stringent and growth control of rRNA synthesis in Escherichia coli are both mediated by ppGpp. J. Biol. Chem. 263:2597–2602.
19. Baracchini, E., and H. Bremer. 1991. Control of rRNA synthesis in Escherichia coli at increased rrn dosage. Role of guanosine tetraphosphate and ribosome feedback. J. Biol. Chem. 266:11753–11760.
20. Baracchini, E., R. Glass, and H. Bremer. 1988. Studies in vivo on Escherichia coli RNA polymerase mutants altered in the stringent response. Mol. Gen. Genet. 213:379–387.
21. Bartlett, M. S., and R. L. Gourse. 1994. Growth rate-dependent control of the rrnB P1 core promoter in Escherichia coli. J. Bacteriol. 176:5560–5564.
22. Bauer, B. F., R. M. Elford, and W. M. Holmes. 1993. Mutagenesis and functional analysis of the Escherichia coli tRNALeu1 promoter. Mol. Microbiol. 7:265–273.
23. Bech, F. W., S. T. Jorgenson, B. Diderichsen, and O. H. Karlstrom. 1985. Sequence of the relB transcription unit from Escherichia coli and identification of the relB gene. EMBO J. 4:1059–1066.
24. Belitskii, B. R., S. T. Kulakauskas, V. V. Sukhodolets, and R. S. Shakulov. 1986. Precise mapping of the gpp gene involved in guanosine tetraphosphate synthesis and ilvC-gpp deletion in the region of the Escherichia coli chromosome. Genetika 22:2775–2783.
25. Belitskii, B. R., and R. S. Shakulov. 1982. Role of spoT gene product in the degradation of pppGpp in bacteria. Mol. Biol. (Moscow) 16:857–864.
26. Belitskii, B. R., and R. S. Shakulov. 1987. Cloning of Escherichia coli gpp gene and insertion of its mutant allele into chromosome of recBC, sbcB cells. Genetika 24:1333–1342.
27. Berg, K. L., C. Squires, and C. L. Squires. 1989. Ribosomal RNA operon anti-termination. Function of leader and spacer region boxB-boxA sequences and their conservation in diverse micro-organisms. J. Mol. Biol. 209:345–358.
28. Block, R. 1976. Synthesis of ribosomal RNA in a partially purified extract from Escherichia coli, p. 226–238. In N. O. Kjeldgaard and O. Maaloe (ed.), Control of Ribosome Synthesis. Alfred Benzon Symposium IX. Munksgaard, Copenhagen.
29. Block, R., and W. A. Haseltine. 1974. In vitro synthesis of ppGpp and pppGpp, p. 747–761. In M. Nomura, A. Tissieres, and P. Lengyel (ed.), Ribosomes. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
30. Blumenthal, R. M., P. G. Lemaux, F. C. Neidhardt, and P. P. Dennis. 1976. The effects of the relA gene on the synthesis of aminoacyl-tRNA synthetases and other transcription and translation proteins in Escherichia coli B. Mol. Gen. Genet. 149:291–296.
31. Bochkanov, S. S., E. V. Kliachko, and R. S. Shakulov. 1983. Stringent control of relA gene transcription in cells of Escherichia coli. Mol. Biol. (Moscow) 17:958–964.
32. Bochner, B. R., and B. N. Ames. 1982. ZTP (5-amino, 4-imidazole carboxamide riboside 5'-triphosphate): a proposed alarmone for 10-formyl-tetrahydrofolate deficiency. Cell 29:929–937.
33. Bochner, B. R., and B. N. Ames. 1982. Complete analysis of cellular nucleotides by two dimensional thin layer chromatography. J. Biol. Chem. 247:9759–9769.
34. Bock, A., and F. C. Neidhardt. 1966. Isolation of a mutant of Escherichia coli with a temperature-sensitive fructose-1,6-diphosphate aldolase activity. J. Bacteriol. 92:464–469.
35. Bock, A., and F. C. Neidhardt. 1966. Properties of a mutant of Escherichia coli with a temperature-sensitive fructose-1,6-diphosphate aldolase activity. J. Bacteriol. 92:470–476.
36. Boquet, P. L., M. A. Devynck, C. Monnier, P. Fromageot, and R. Roschenthaler. 1973. Inhibition of stable RNA synthesis by levallorphan in Escherichia coli. Implication of compounds MSI and MSII. Eur. J. Biochem. 40:31–42.
37. Borek, E., J. Rockenbach, and A. Ryan. 1956. Studies on a mutant of Escherichia coli with unbalanced ribonucleic acid synthesis. J. Bacteriol. 71:318–323.
38. Borukhov, S., V. Sagitov, C. A. Josaitis, R. L. Gourse, and A. Goldfarb. 1993. Two modes of transcription initiation in vitro at the rrnB P1 promoter of Escherichia coli. J. Biol. Chem. 268:23477–23482.
39. Bosch, L., L. Nilsson, E. Vijgenboom, and H. Verbeek. 1993. Fis-dependent transactivation of stable RNA operons and bacterial growth, p. 139–150. In K. H. Nierhaus, F. Franceski, A. R. Subramanian, V. A. Erdmann, and B. Wittmann-Liebold (ed.), The Translational Apparatus. Plenum Press, New York.
40. Bossi, L., and R. Cortese. 1977. Biosynthesis of tRNA in histidine regulatory mutants of Salmonella typhimurium. Nucleic Acids Res. 4:1945–1956.
41. Bossi, L., and D. M. Smith. 1984. Conformational change in the DNA associated with an unusual promoter mutation in a tRNA operon of Salmonella. Cell 39:643–652.
42. Bouloc, P., A. Jaffe, and R. D’Ari. 1988. Preliminary physiologic characterization and genetic analysis of a new Escherichia coli mutant, lov, resistant to mecillinam. Rev. Infect. Dis. 19:905–910.
43. Bouloc, P., A. Jaffe, and R. D’Ari. 1989. The Escherichia coli lov gene product connects peptidoglycan synthesis, ribosomes and growth rate. EMBO J. 8:317–323.
44. Bouloc, P., D. Vinella, and R. D’Ari. 1992. Leucine and serine induce mecillinam resistance in Escherichia coli. Mol. Gen. Genet. 235:242–246.
45. Boyle, S. M., and K. Adachi. 1982. Biosynthetic ornithine and arginine decarboxylases: correlation of rates of synthesis and activities in Escherichia coli during exponential growth and following nutritional shift-up. Can. J. Microbiol. 28:945–950.
46. Bramhill, D., and A. Kornberg. 1988. Duplex opening by DnaA protein at novel sequences in initiation of replication at the origin of the E. coli chromosome. Cell 52:743–755.
47. Breeden, L., and M. Yarus. 1982. Amber suppression relaxed stringent control by elongating stringent factor. Mol. Gen. Genet. 187:254–264.
48. Breeden, L., and M. Yarus. 1982. Mutations that overcome plasmid-mediated relaxation affect (p)ppgpp. Mol. Gen. Genet. 179:119–124.
49. Brenner, M., and B. N. Ames. 1971. The histidine operon and its regulation, p. 349–387. In D. M. Greenberg and H. J. Vogel (ed.), Metabolic Pathways, vol. 5. Academic Press, Inc., New York.
50. Brikun, I. A., and V. V. Sukhodolets. 1982. Effects of mutations in the relC and spoT genes of the stringent control regulatory system on the gene expression of Escherichia coli K-12 nucleoside metabolism. Genetika 18:929–938.
51. Broda, P. 1968. Ribonucleic acid synthesis and glutamate excretion in Escherichia coli. J. Bacteriol. 96:1528–1534.
52. Brunschede, H., T. Dove, and H. Bremer. 1977. Establishment of exponential growth after a nutritional shift-up in Escherichia coli B/r: accumulation of deoxyribonucleic acid, ribonucleic acid, and protein. J. Bacteriol. 129:1020–1033.
53. Carter-Muenchau, P., and R. E. Wolf, Jr. 1989. Growth-rate-dependent regulation of 6-phosphogluconate dehydrogenase level mediated by an anti-Shine-Dalgarno sequence located within the Escherichia coli gnd structural gene. Proc. Natl. Acad. Sci. USA 86:1138–1142.
54. Cashel, M. 1969. The control of ribonucleic acid synthesis in Escherichia coli. IV. Relevance of unusual phosphorylated compounds from amino acid starved stringent strains. J. Biol. Chem. 244:3133–3141.
55. Cashel, M. 1974. Preparation of guanosine tetraphosphate (ppGpp) and guanosine pentaphosphate (pppGpp) from Escherichia coli ribosomes. Anal. Biochem. 57:100–107.
56. Cashel, M. 1975. Regulation of bacterial ppGpp and pppGpp. Annu. Rev. Microbiol. 29:301–318.
57. Cashel, M. 1994. Detection of (p)ppGpp accumulation patterns in Escherichia coli mutants, p. 341–356. In K. W. Adolph (ed.), Methods in Molecular Genetics, vol. 3. Molecular Microbiology Techniques, part A. Academic Press, New York.
58. Cashel, M., and J. Gallant. 1969. Two compounds implicated in the function of the RC gene of Escherichia coli. Nature (London) 221:838–841.
59. Cashel, M., and J. Gallant. 1974. Cellular regulation of guanosine tetraphosphate and guanosine pentaphosphate. p. 733–745. In M. Nomura, A. Tissieres, and P. Lengyel (ed.), Ribosomes. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
60. Cashel, M., and B. Kalbacher. 1970. The control of ribonucleic acid synthesis in Escherichia coli. V. Characterization of a nucleotide associated with the stringent response. J. Biol. Chem. 245:2309–2318.
61. Cenatiempo, Y., N. Robakis, L. Meza-Basso, N. Brot, H. Weissbach, and B. R. Reid. 1982. Use of different tRNA-ser isoacceptor species in vitro to discriminate between expression of plasmid genes. Proc. Natl. Acad. Sci. USA 79:1466–1468.
62. Chaloner-Larsson, G., and H. Yamazaki. 1976. Synthesis of guanosine 5'-triphosphate, 3'-diphosphate in a spoT strain of Escherichia coli. Can. J. Biochem. 54:935–940.
63. Chaloner-Larsson, G., and H. Yamazaki. 1977. Adjustment of RNA content during temperature upshift in Escherichia coli. Biochem. Biophys. Res. Commun. 77:503–508.
64. Chan, C. L., and R. Landick. 1994. New perspectives on RNA chain elongation and termination by E. coli RNA polymerase, p. 297–321. In R. C. Conaway and J. W. Conaway (ed.), Transcription: Mechanisms and Regulation. Raven Press, Ltd., New York.
65. Chiaramello, A. E., and J. W. Zyskind. 1989. Expression of Escherichia coli dnaA and mioC genes as a function of growth rate. J. Bacteriol. 171:4272–4280.
66. Chiaramello, A. E., and J. W. Zyskind. 1990. Coupling of DNA replication to growth rate in Escherichia coli: a possible role for guanosine tetraphosphate. J. Bacteriol. 172:2013–2019.
67. Chinali, G., J. Horowitz, and J. Ofengang. 1978. Replacement of pseudouridine in transfer RNA by 5-fluorouridine does not affect the ability to stimulate the synthesis of guanosine 5'-triphosphate 3'-diphosphate. Biochemistry 17:2755–2760.
68. Chung, C. H., and A. L. Goldberg. 1981. The product of the lon (capR) gene in Escherichia coli is the ATP-dependent protease, protease La. Proc. Natl. Acad. Sci. USA 78:4931–4935.
69. Churchward, G., H. Bremer, and R. Young. 1982. Transcription in bacteria at different DNA concentrations. J. Bacteriol. 150:572–581.
70. Cochran, J. W., and R. W. Byrne. 1974. Isolation and properties of a ribosome-bound factor required for ppGpp and pppGpp synthesis in Escherichia coli. J. Biol. Chem. 249:353–360.
71. Cole, J. R., C. L. Olsson, J. W. B. Hershey, M. Grunberg-Manago, and M. Nomura. 1987. Feedback regulation of rRNA synthesis in Escherichia coli: requirement for initiation factor IF2. J. Mol. Biol. 198:383–392.
72. Collins, J., N. P. Fiil, P. Jorgensen, and J. D. Friesen. 1976. Gene cloning of Escherichia coli chromosomal genes important in the regulation of ribosomal RNA synthesis, p. 356–367. In N. O. Kjeldgaard and O. Maaloe (ed.), Control of Ribosome Synthesis. Alfred Benzon Symposium IX. Munksgaard, Copenhagen.
73. Condon, C., S. French, C. Squires, and C. L. Squires. 1993. Deletion of functional ribosomal RNA operons in Escherichia coli causes increased expression of the remaining intact copies. EMBO J. 12:4305–4315.
74. Condon, C., J. Phillips, Z.-Y. Fu, C. Squires, and C. Squires. 1992. Comparison of the expression of the seven ribosomal RNA operons in Escherichia coli. EMBO J. 11:4175–4185.
75. Cortay, J. C., and A. J. Cozzone. 1983. Accumulation of guanosine tetraphosphate induced by polymixin and gramicidin in Escherichia coli. Biochim. Biophys. Acta 755:467–473.
76. Cortese, R., H. O. Kammen, S. J. Spengler, and B. N. Ames. 1974. Biosynthesis of pseudouridine in transfer ribonucleic acid. J. Biol. Chem. 249:1103–1108.
77. Cozzone, A. J. 1980. Stringent control and protein synthesis in bacteria. Biochimie (Paris) 62:647–664.
78. Crooke, E., M. Akiyama, N. N. Rao, and A. Kornberg. 1994. Genetically altered levels of inorganic polyphosphate in Escherichia coli. J. Biol. Chem. 269:6290–6294.
79. Danchin, A. 1977. A new technique for selection of sensitive and auxotrophic mutants of E. coli: isolaton of a strain sensitive to an excess of one-carbon metabolites. Mol. Gen. Genet. 150:293–299.
80. Danchin, A., and L. Dondon. 1979. Regulatory features of tRNA(leu1) expression in Escherichia coli K-12. Biochem. Biophys. Res. Commun. 90:1280–1286.
81. Danchin, A., and L. Dondon. 1980. Serine sensitivity of Escherichia coli K 12: partial characterization of a serine resistant mutant that is extremely sensitive to 2-ketobutyrate. Mol. Gen. Genet. 178:155–164.
82. Daniel, J., L. Dondon, and A. Danchin. 1983. 2-Ketobutyrate: a putative alarmone of Escherichia coli. Mol. Gen. Genet. 190:452–458.
83. Dassa, E., M. Cahu, B. Desjoyaux-Cherel, and P. L. Bouquet. 1982. The acid phosphatase with optimum pH of 2.5 of Escherichia coli; physiological and biochemical study. J. Biol. Chem. 257:6669–6676.
84. Davidson, J. P., D. J. Wilson, and L. S. Williams. 1982. Role of hisU gene in the control of stable RNA synthesis in Salmonella typhimurium. J. Mol. Biol. 157:237–264.
85. Davis, L., and L. S. Williams. 1982. Characterization of a cold-sensitive hisW mutant of Salmonella typhimurium. J. Bacteriol. 151:867–878.
86. Davis, L., and L. S. Williams. 1982. Altered regulation of isoleucine-valine biosynthesis in a hisW mutant of Salmonella typhimurium. J. Bacteriol. 151:860–866.
87. Debenham, P. G., R. Buckland, and A. A. Travers. 1979. The regulation of RNA polymerase by guanoside 5'-diphosphate 3'-diphosphate, p. 77–84. In G. Koch and D. Richter (ed.), Regulation of Macromolecular Synthesis by Low Molecular Weight Mediators. Academic Press, Inc., New York.
88. de Boer, H., and M. Nomura. 1979. In vivo transcription of rRNA operons in Escherichia coli initiates with purine nucleoside triphosphates at the first promoter and with CTP at the second promoter. J. Biol. Chem. 254:5609–5612.
89. de Boer, H. A., W. J. Weyer, J. G. deBoer, S. van Der Heide, and M. Gruber. 1977. Synthesis of guanosine 5'-diphosphate, 3'-diphosphate in spoT mutants of Escherichia coli. Biochim. Biophys. Acta 474:165–172.
90. de Massy, B., O. Fayet, and T. Kogoma. 19843. Multiple origin usage for DNA replication in sdrA (rnh) mutant of Escherichia coli K12: initiation in the absence of oriC. J. Mol. Biol. 178:227–236.
91. Dennis, D. P. 1977. Influence of the stringent control system on the transcription of ribosomal ribonucleic acid and ribosomal protein genes in Escherichia coli. J. Bacteriol. 129:580–588.
92. Dennis, P. P., and M. Nomura. 1974. Stringent control of ribosomal protein gene expression in Escherichia coli. Proc. Natl. Acad. Sci. USA 71:3819–3823.
93. Deuster, G., R. M. Elford, and W. M. Holmes. 1982. Fusion of the Escherichia coli tRNA leu1 promoter to the galK gene: analysis of sequences necessary for growth-rate-dependent regulation. Cell 30:855–864.
94. Dickson, R. R., T. Gaal, H. A. deBoer, P. L. deHaseth, and R. L. Gourse. 1989. Identification of promoter mutants defective in growth-rate-dependent regulation of rRNA transcription in Escherichia coli. J. Bacteriol. 171:4862–4870.
95. Diderichsen, B., and L. Desmarez. 1980. Variations in phenotype of relB mutants of E. coli and the effect of pus and sup mutations. Mol. Gen. Genet. 180:429–437.
96. Diderichsen, B., N. P. Fiil, and R. Lavalle. 1977. Genetics of the relB locus in Escherichia coli. J. Bacteriol. 131:30–33.
97. Dietzler, D. N., and M. P. Leckie. 1977. Regulation of ADP-glucose synthetase, the rate limiting enzyme of bacterial glycogen synthesis, by the pleiotropic nucleotides ppGpp and pppGpp. Biochem. Biophys. Res. Commun. 77:1459–1467.
98. Dohi, M., A. Kikuchi, and L. Gorini. 1978. Some regulation profiles of ornithine transcarbamylase synthesis in vivo. J. Biochem. (Tokyo) 84:1401–1409.
99. Donachie, W. D. 1968. Relationship between cell size and time of initiation of DNA replication. Nature (London) 219:1077–1079.
100. Donini, P. 1972. Turnover of ribosomal RNA during the stringent response in Escherichia coli. J. Mol. Biol. 72:553–569.
101. Donini, P., V. Santonastaso, J. Roche, and A. J. Cozzone. 1978. The relationship between guanosine tetraphosphate, polysomes and RNA synthesis in amino acid starved Escherichia coli. Mol. Biol. Rep. 4:15–19.
102. Drew, H. R., J. R. Weeks, and A. A. Travers. 1984. Negative supercoiling induces spontaneous unwinding of a bacterial promoter. EMBO J. 4:1025–1032.
103. Edlin, G., and P. Broda, 1968. Physiology and genetics of the ribonucleic acid control locus in Escherichia coli. Bacteriol. Rev. 32:206–226.
104. Elhardt, D., R. Wirth, and A. Bock. 1982. Regulation of formation of threonyl-tRNA synthetase and protein synthesis initiation factor 3 from Escherichia coli in vivo and in vitro. Eur. J. Biochem. 123:477–482.
105. Emilsson, V., and C. G. Kurland. 1990. Growth rate dependence of transfer RNA abundance in Escherichia coli. EMBO J. 13:4359–4366.
106. Erdmann, V. A., S. Lorenz, M. Sprinzl, and R. T. Wagner. 1976. T and W of loop IV in tRNAs are essential for stringent factor dependent synthesis of pppGpp and ppGpp, p. 427–436. In N. O. Kjeldgaard and O. Maaloe (ed.), Control of Ribosome Synthesis. Alfred Benzon Symposium IX. Munksgaard, Copenhagen.
107. Erlich, H., T. Laffler, and J. Gallant. 1971. ppGpp formation in Escherichia coli treated with rifampicin. J. Biol. Chem. 246:6121–6125.
108. Fallon, A. M., C. S. Jinks, G. D. Strycharz, and M. Nomura. 1979. Regulation of ribosomal protein synthesis by selective mRNA inactivation. Proc. Natl. Acad. Sci. USA 76:3411–3415.
109. Fast, R., and O. Skold. 1977. Biochemical mechanism of uracil uptake regulation in Escherichia coli B. Allosteric effects on uracil phosphoribosyltransferase under stringent conditions. J. Biol. Chem. 252:7620–7624.
110. Fehr, S., and D. Richter. 1981. Stringent response of Bacillus stearothermophilus: evidence for the existence of two distinct guanosine 3',5'-polyphosphate synthetases. J. Bacteriol. 145:68–73.
111. Figueroa, N., N. Wills, and L. Bossi. 1991. Common sequence determinants of the response of a prokaryotic promoter to DNA bending and supercoiling. EMBO J. 10:941–949.
112. Fiil, N. 1969. A functional analysis of the rel gene in Escherichia coli. J. Mol. Biol. 45:195–203.
113. Fiil, N., and J. D. Friesen. 1968. Isolation of relaxed mutants of Escherichia coli. J. Bacteriol. 95:729–731.
114. Fiil, N. P., K. von Meyenberg, and J. D. Friesen. 1972. Accumulation and turnover of guanosine tetraphosphate in Escherichia coli. J. Mol. Biol. 71:769–783.
115. Fiil, N. P., B. M. Willumsen, J. D. Friesen, and K. von Meyenberg. 1977. Interaction of alleles of the relA, relC and spoT genes in Escherichia coli: analysis of the interconversion of GTP, ppGpp, and pppGpp. Mol. Gen. Genet. 150:87–101.
116. Fournier, M. J., and H. Ozeki. 1985. Structure and organization of the transfer ribonucleic acid genes of Escherichia coli K-12. Microbiol. Rev. 49:379–397.
117. Frank, S., W. Eck, W. Kersten, and A. Ogilvie. 1981. On the cytostatic mechanism of cyclophosphamide. Inhibition of aminoacylation of transfer RNA and induction of stringent control in Escherichia coli by 4-hydroperoxycyclophosphamide. Hoppe-Seyler’s Z. Physiol. Chem. 363:1059–1068.
118. Freundlich, M. 1977. Cyclic AMP can replace the RelA-dependent requirement for derepression of acetohydroxy acid synthase in E. coli K-12. Cell 12:1121–1126.
119. Friesen, J. D., G. An, and N. P. Fiil. 1978. Nonsense and insertion mutants in the rel gene of E. coli: cloning relA. Cell 15:1187–1197.
120. Friesen, J. D., and N. Fiil. 1973. Accumulation of guanosine tetraphosphate in T7 bacteriophage-infected Escherichia coli. J. Bacteriol. 113:697–703.
121. Friesen, J. D., N. P. Fiil, J. M. Parker, and W. Haseltine. 1974. A new relaxed mutant of Escherichia coli with an altered 50S ribosomal subunit. Proc. Natl. Acad. Sci. USA 71:3465–3469.
122. Friesen, J. D., N. P. Fiil, and K. von Meyenberg. 1975. Synthesis and turnover of basal level guanosine tetraphosphate in Escherichia coli. J. Biol. Chem. 250:304–309.
123. Friesen, J. D., J. Parker, R. J. Watson, N. P. Fiil, S. Pedersen, and F. S. Pedersen. 1976. Isolation of a lambda transducing bacteriophage carrying the relA gene of Escherichia coli. J. Bacteriol. 127:917–922.
124. Fukuda, R., A. Nishimura, and H. Serizawa. 1988. Genetic mapping of the Escherichia coli gene for the stringent starvation protein and its dispensibility for normal cell growth. Mol. Gen. Genet. 211:511–519.
125. Furano, A. V. 1975. Content of elongation factor Tu in Escherichia coli. Proc. Natl. Acad. Sci. USA 72:4780–4784.
126. Furano, A. V., and F. P. Witel. 1976. Effect of the relA gene on the synthesis of individual proteins in vivo. Cell 8:115–122.
127. Gaal, T., J. Barkei, R. R. Dickson, H. A. deBoer, P. L. deHaseth, H. Alavi, and R. L. Gourse. 1989. Saturation mutagenesis of an Escherichia coli rRNA promoter and initial characterization of promoter variants. J. Bacteriol. 171:4852–4861.
128. Gaal, T., and R. L. Gourse. 1990. Guanosine 3'-diphosphate 5'-diphosphate is not required for growth rate-independent control of rRNA synthesis in Escherichia coli. Proc. Natl. Acad. Sci. USA 87:5533–5537.
129. Gaal, T., L. Rao, S. T. Estrem, J. Yang, R. M. Wartell, and R. L. Gourse. 1994. Localization of the intrinsically bent DNA region upstream of the E. coli rrnB P1 promoter. Nucleic Acids Res. 22:2344–2350.
130. Gafny, R., S. Cohen, N. Nachaliel, and G. Glaser. 1994. Isolated P2 promoters of Escherichia coli are strong promoters that are subject to stringent control. J. Mol. Biol. 243:152–156.
131. Gallant, J., and B. Harada. 1969. The control of RNA synthesis in Escherichia coli. III. The functional relationship between purine ribonucleoside triphosphate pool sizes and the rate of ribonucleic acid accumulation. J. Biol. Chem. 244:3125–3152.
132. Gallant, J., J. Irr, and M. Cashel. 1971. The mechanism of amino acid control of guanylate and adenylate synthesis. J. Biol. Chem. 246:5812–5816.
133. Gallant, J., and R. A. Lazzarini. 1976. The regulation of ribosomal RNA synthesis and degradation in bacteria, p. 310–349. In E. McConkey (ed.), Protein Synthesis, a Series of Advances, vol. 2. Marcel Dekker, Inc., New York.
134. Gallant, J., G. Margason, and B. Finch. 1972. On the turnover of ppGpp in Escherichia coli. J. Biol. Chem. 247:6055–6058.
135. Gallant, J., L. Palmer, and C. C. Pao. 1977. Anomalous synthesis of ppGpp in growing cells. Cell 11:181–185.
136. Gallant, J., L. Shell, and R. Bittner. 1976. A novel nucleotide implicated in the response of E. coli to energy source downshift. Cell 7:75–84.
137. Gallant, J. A. 1979. Stringent control in E. coli. Annu. Rev. Genet. 13:393–415.
138. Garrido, T., M. Sanchez, P. Placios, M. Aldea, and M. Vicente. 1993. Transcription of ftsZ oscillates during the cell cycle of Escherichia coli. EMBO J. 12:3957–3965.
139. Gausing, K. 1976. Synthesis of rRNA and r-protein mRNA in E. coli at different growth rates, p. 292–303. In N. O. Kjeldgaard and O. Maaloe (ed.), Control of Ribosome Synthesis. Alfred Benzon Symposium IX. Munksgaard, Copenhagen.
140. 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.
141. Gentry, D. R., C. Bengra, K. Ikehara, and M. Cashel. 1993. Guanylate kinase of Escherichia coli K-12. J. Biol. Chem. 268:14316–14321.
142. Gentry, D. R., and R. R. Burgess. 1986. The cloning and sequence of the gene encoding the omega subunit of Escherichia coli RNA polymerase. Gene 48:33–40.
143. Gentry, D. R., V. J. Hernandez, D. E. Nguyen, D. B. Jensen, and M. Cashel. 1993. Synthesis of the stationary phase specific sigma factor, σS, is positively regulated by ppGpp. J. Bacteriol. 175:7982–7989.
144. Gentry, D. R., H. Xiao, R. Burgess, and M. Cashel. 1991. The omega subunit of Escherichia coli K-12 RNA polymerase is not required for stringent RNA control in vivo. J. Bacteriol. 173:3901–3903.
145. Gerdes, K., L. K. Poulsen, T. Thisted, A. K. Nielsen, J. Martinussen, P. H. Andresasen. 1990. The hok killer gene family in gram-negative bacteria. New Biol. 2:946–956.
146. Gerdes, K., F. W. Bech, S. T. Jorgensen, A. Lobner-Olesen, P. B. Rasmussen, T. Atlung, L. Boe, O. Karlstrom, S. Molin, and K. von Meyenburg. 1986. Mechanism of postsegregational killing by hok gene product of the parB system of plasmid R1 and its homology with the relF gene product of the E. coli relB operon. EMBO J. 5:2023–2029.
147. Gervais, F. G., P. Phoenix, and G. R. Drapeau. 1992. The rcsB gene, a positive regulator of colanic acid biosynthesis in Escherichia coli, is also an activator of ftsZ expression. J. Bacteriol. 174:3964–3971.
148. Gilbert, S. F., H. A. deBoer, and M. Nomura. 1979. Identification of initiation sites for the in vitro transcription of rRNA operons rrnE and rrnA in E. coli. Cell 17:211–224.
149. Ginther, C. L., and J. L. Ingraham. 1974. Cold-sensitive mutant of Salmonella typhimurium defective in nucleoside diphosphate kinase. J. Bacteriol. 118:1020–1026.
150. Glaser, G., S. Kobi, and A. B. Oppenheim. 1980. Fusion of the promoter region of rRNA operon rrnB to lacZ gene. Nucleic Acids Res. 8:4327–4335.
151. Glaser, G., P. Sarmientos, and M. Cashel. 1983. Functional inter-relationship between two tandem E. coli ribosomal RNA promoters. Nature (London) 302:74–76.
152. Glass, R. E., S. T. Jones, and A. Ishihama. 1986. Genetic studies on the β subunit of Escherichia coli RNA polymerase. VII. RNA polymerase is a target for ppGpp. Mol. Gen. Genet. 203:265–268.
153. Goff, S. A., L. P. Casson, and A. L. Goldberg. 1984. Heat shock regulatory gene htpR influences rates of protein degradation and expression of the lon gene in Escherichia coli. Proc. Natl. Acad. Sci. USA 81:6647–6651.
154. Golden, N. G., and G. L. Powell. 1972. Stringent and relaxed control of phospholipid metabolism in Escherichia coli. J. Biol. Chem. 247:6651–6658.
155. Goldman, E., and H. Jakubowski. 1990. Uncharged tRNA, protein synthesis, and the bacterial stringent response. Mol. Microbiol. 4:2035–2040.
156. Goldstein, J., N. S. Pollitt, and M. Inouye. 1990. Major cold shock protein of Escherichia coli. Proc. Natl. Acad. Sci. USA 87:283–287.
157. Goodell, W., and A. Tomasz. 1980. Alteration of Escherichia coli murein during amino acid starvation. J. Bacteriol. 144:1009–1016.
158. Gordeev, V. K., and M. I. Turkov. 1983. Expression of amino acid operons in Escherichia coli strains with altered transcription and translation machinery. IV. The influence of mutations altering coupling of transcription and translation systems on expression of the ilv operon in cells carrying the spoT mutation. Genetika 9:1433–1438.
159. Gosink, K. K., W. Ross, S. Liermo, R. Osuna, S. E. Finkel, R. C. Johnson, and R. L. Gourse. 1993. DNA binding and bending are necessary but not sufficient for Fis-dependent activation of rrnB P1. J. Bacteriol. 175:1580–1589.
160. Gourse, R. L. 1988. Visualization and quantitative analysis of complex formation between E. coli RNA polymerase and rRNA promoter in vitro. Nucleic Acids Res. 16:9789–9799.
161. Gourse, R. L., H. A. deBoer, and M. Nomura. 1986. DNA determinants of rRNA synthesis in E. coli: growth rate dependent regulation, feedback inhibition, upstream activation, antitermination. Cell 44:197–205.
162. Gourse, R. L., and M. Nomura. 1984. Level of rRNA, not tRNA, synthesis controls transcription of rRNA and tRNA operons in Escherichia coli. J. Bacteriol. 160:1022–1026.
163. Gourse, R. L., M. J. R. Stark, and A. E. Dahlberg. 1983. Regions of DNA involved in the stringent control of plasmid-encoded rRNA in vivo. Cell 32:1347–1357.
164. Gourse, R. L., Y. Takebe, R. A. Sharrock, and M. Nomura. 1985. Feedback regulation of rRNA and tRNA synthesis and accumulation of free ribosomes after conditional expression of rRNA genes. Proc. Natl. Acad. Sci. USA 82:1069–1073.
165. Gropp, M., E. Eizenman, G. Glaser, W. Samarrai, and R. Rudner. 1994. A relA (S) suppressor mutant allele of Bacillus subtilis which maps to relA and responds only to carbon limitation. Gene 140:91–96.
166. Grossman, A. D., W. E. Taylor, Z., F. Burton, R. R. Burgess, and C. A. Gross. 1985. Stringent response in E. coli induces expression of heat shock proteins. J. Mol. Biol. 186:357–365.
167. Grundy, F. J., S. M. Rollins, and T. M. Henkin. 1994. Interaction between the acceptor end of tRNA and the T box stimulates antitermination in the Bacillus subtilis tyrS gene: a new role for the discriminator base. J. Bacteriol. 176:4518–4526.
168. Gustafson, C. E., S. Kaul, and E. E. Ishiguro. 1993. Identification of the Escherichia coli lytB gene, which is involved in penicillin tolerance and control of the stringent response. J. Bacteriol. 175:1203–1205.
169. Gustafsson, C., P. H. R. Lindstrom, T. G. Hagervall, K. B. Esberg, and G. R. Bjork. 1991. The trmA promoter has regulatory features and sequence elements in common with rRNA P1 promoter family of Escherichia coli. J. Bacteriol. 173:1757–1764.
170. Guzman, E. C., F. J. Carrillo, and A. Jimenez-Sanchez. 1988. Differential inhibition of the initiation of the DNA replication in stringent and relaxed strains of Escherichia coli. Genet. Res. 51:173–177.
171. Guzman, E. C., R. H. Pritchard, and A. Jimenez-Sanchez. 1991. A calcium-binding protein that may be required for the initiation of chromosome replication in Escherichia coli. Res. Microbiol. 142:137–140.
172. Hagervall, T. T., B. Esberg, J.-N. Li, T. M. F. Tuohy, J. F. Atkins, J. F. Curran, and G. R. Bjork. 1993. Functional aspects of three modified nucleosides, Y, ms2io6A and m1G, present in the anticodon loop of tRNA, p. 67–78. In K. H. Nierhaus, F. Franceshi, A. R. Subramanian, V. A. Erdmann, and B. Wittmann-Liebold (ed.), The Translational Apparatus. Plenum Press, New York.
173. Hamel, E., and M. Cashel. 1973. Role of guanine nucleotides in protein synthesis: elongation factor G and guanosine 5'-triphosphate, 3'-diphosphate. Proc. Natl. Acad. Sci. USA 70:3250–3254.
174. Hamming, J., M. Gruber, and G. Ab. 1979. Interaction between RNA polymerase and a ribosomal promoter of E. coli. Nucleic Acids Res. 7:1019–1033.
175. Hansen, F. G., T. Atlung, R. E. Braun, A. Wright, P. Hughes, and M. Kohiyama. 1991. Initiator (DnaA) protein concentration as a function of growth rate in Escherichia coli and Salmonella typhimurium. J. Bacteriol. 173:5194–5199.
176. Hansen, F. G., B. B. Christensen, and T. Atlung. 1991. The initiator titration model: computer simulation of chromosome and minichromosome control. Res. Microbiol. 142:161–167.
177. Hansen, M. J., L.-H. Chen, M. L. S. Fejzo, and J. G. Belasco. 1994. The ompA 5' untranslated region impedes a major pathway for mRNA degradation in Escherichia coli. Mol. Microbiol. 12:707–716.
178. Hansen, M. T., M. L. Pato, S. Molin, N. P. Fiil, and K. von Meyenburg. 1975. Simple downshift and resulting lack of correlation between ppGpp pool size and ribonucleic acid accumulation. J. Bacteriol. 122:585–591.
179. Hara, A., and J. Sy. 1983. Guanosine 5'-triphosphate, 3'-diphosphate 5'-phosphohydrolase. Purification and substrate specificity. J. Biol. Chem. 258:1678–1693.
180. Harkness, R. E., and E. E. Ishiguro. 1983. Temperature-sensitive autolysis-defective mutants of Escherichia coli. J. Bacteriol. 155:15–21.
181. Harshman, R. B., and H. Yamazaki. 1971. Formation of ppGpp in a relaxed and stringent strain of Escherichia coli during diauxic lag. Biochemistry 10:3980–3982.
182. Harshman, R. B., and H. Yamazaki. 1972. MS I accumulation induced by sodium chloride. Biochemistry 11:615–618.
183. Haseltine, W. A. 1972. In vitro transcription of E. coli ribosomal RNA genes. Nature (London) 235:329–333.
184. Haseltine, W. A., and R. Block. 1973. Synthesis of guanosine tetra- and pentaphosphate requires the presence of a codon-specific, uncharged transfer ribonucleic acid in the acceptor site of ribosomes. Proc. Natl. Acad. Sci. USA 70:1564–1568.
185. Haseltine, W. A., R. Block, W. Gilbert, and K. Weber. 1972. MSI and MSII are made on ribosomes in an idling step of protein synthesis. Nature (London) 238:381–384.
186. Heath, R. J., S. Jackowski, and C. O. Rock. 1994. Guanosine tetraphosphate inhibition of fatty acid and phospholipid synthesis in Escherichia coli is relieved by overexpression of glycerol-3-phosphate acyltransferase (plsB). J. Biol. Chem. 269:26584–26590.
187. Hecker, M., and A. Schroeter. 1985. Synthesis of alkaline phosphatase in a stringent and relaxed strain of Escherichia coli under amino acid and phosphate limitation. J. Basic Microbiol. 25:341–347.
188. Hecker, M., A. Schroeter, and F. Mach. 1983. Replication of pBR322 DNA in stringent and relaxed strains of Escherichia coli. Mol. Gen. Genet. 190:355–357.
189. Heinemeyer, E. A., M. Gels, and D. Richter. 1978. Degradation of guanosine 3'-diphosphate 5'-diphosphate in vitro by the spoT gene product of Escherichia coli. Eur. J. Biochem. 89:125–131.
190. Heinemeyer, E. A., and D. Richter. 1977. In vitro degradation of guanosine tetraphosphate (ppGpp) by an enzyme associated with the ribosomal fraction from Escherichia coli. FEBS Lett. 84:357–361.
191. Heinemeyer, E. A., and D. Richter. 1978. Mechanism of the in vitro breakdown of guanosine 5'-diphosphate 3'-diphosphate in Escherichia coli. Proc. Natl. Acad. Sci. USA 75:4180–4183.
192. Heinemeyer, E. A., and D. Richter. 1978. Characterization of the guanosine 5'-triphosphate 3'-diphosphate and guanosine 5'-diphosphate 3'-diphosphate degradation reaction catalyzed by a specific pyrophosphorylase from Escherichia coli. Biochemistry 12:5368–5372.
193. Herman, A., A. Wegrzyn, and G. Wegrzyn. 1994. Differential replication of plasmids during stringent and relaxed response of Escherichia coli. Plasmid 32:89–94.
194. Herman, A., A. Wegrzyn, and G. Wegrzyn. 1994. Regulation of replication of plasmid pBR322 in amino acid-starved Escherichia coli strains. Mol. Gen. Genet. 243:374–378.
195. Hernandez, V. J., and H. Bremer. 1990. Guanosine tetraphosphate (ppGpp) dependence of the growth rate control of rrnB P1 promoter activity in Escherichia coli. J. Biol. Chem. 265:11605–11614.
196. Hernandez, V. J., and H. Bremer. 1991. Escherichia coli ppGpp synthetase II activity requires spoT. J. Biol. Chem. 266:5991–5999.
197. Hernandez, V. J., and H. Bremer. 1993. Characterization of RNA and DNA synthesis in Escherichia coli strain devoid of ppGpp. J. Biol. Chem. 268:10851–10862.
198. Hochstadt-Ozer, J., and M. Cashel. 1972. The regulation of purine utilization in bacteria. V. Inhibition of purine phosphoribosyltransferase activities and purine uptake in isolated membrane vesicles by guanosine tetraphosphate. J. Biol. Chem. 247:7067–7072.
199. Holtje, J.-V. 1979. Regulation of polyamine and streptomycin transport during the stringent control in Escherichia coli. J. Bacteriol. 137:661–663.
200. Holtta, E., J. Janne, and J. Pispa. 1974. The regulation of polyamine synthesis during the stringent control in Escherichia coli. Biochem. Biophys. Res. Commun. 57:1104–1111.
201. Hsu, D., L.-M. Shih, and Y. C. Zee. 1994. Degradation of rRNA in Salmonella strains: a novel mechanism to regulate the concentrations of rRNA and ribosomes. J. Bacteriol. 176:4761–4765.
202. Hughes, J., and G. Mellows. 1978. Inhibition of isoleucyl-transfer ribonucleic acid synthetase in Escherichia coli by pseudomonic acid. Biochem. J. 176:305–318.
203. Ichikawa, J. K., C. Li, J. Fu, and S. Clark. 1994. A gene at 59 minutes on the Escherichia coli chromosome encodes a lipoprotein with unusual amino acid repeat sequences. J. Bacteriol. 176:1630–1638.
204. Igarashi, K., N. Fujita, and A. Ishihama. 1989. Promoter selectivity of Escherichia coli RNA polymerase: omega factor is responsible for ppGpp sensitivity. Nucleic Acids Res. 17:8755–8765.
205. Ikemura, T., and J. E. Dahlberg. 1973. Small ribonucleic acids of Escherichia coli. II. Noncoordinate accumulation during stringent control. J. Biol. Chem. 248:5033–5041.
206. Ingraham, J. L., O. Maaloe, and F. C. Neidhardt. 1983. Growth of the Bacterial Cell. Sinauer Associates, Inc., Sunderland, Mass.
207. Irr, J., and J. Gallant. 1969. The control of RNA synthesis in Escherichia coli. II. Stringent control of energy metabolism. J. Biol. Chem. 244:2233–2239.
208. Ishiguro, E. E., and W. D. Ramey. 1976. Stringent control of peptidoglycan biosynthesis in Escherichia coli K-12. J. Bacteriol. 127:1119–1126.
209. Ishiguro, E. E., and W. D. Ramey. 1978. Involvement of the relA gene product and feedback inhibition in the regulation of UDP-N-acetylmuramyl peptide synthesis in Escherichia coli. J. Bacteriol. 135:766–774.
210. Ishiguro, E. E., and W. D. Ramey. 1980. Inhibition of in vitro biosynthesis in Escherichia coli by guanosine 5'-diphosphate 3'-diphosphate. Can. J. Microbiol. 26:1514–1518.
211. Ishihama, A., and T. Saitoh. 1979. Subunits of RNA polymerase in function and structure. IX. Regulation of RNA polymerase activity by stringent starvation protein (SSP). J. Mol. Biol. 129:517–530.
212. James, R., J. Y. Haga, and A. B. Pardee. 1975. Inhibition of an early event in the cell division cycle of Escherichia coli by FL 1060, an amidinopenicillinac acid. J. Bacteriol. 122:1283–1292.
213. Jensen, K. F., and S. Pedersen. 1990. Metabolic growth rate control in Escherichia coli may be a consequence of subsaturation of the macromolecular biosynthetic apparatus with substrates and catalytic components. Microbiol. Rev. 54:89–100.
214. Jin, D. J., R. R. Burgess, J. P. Richardson, and C. A. Gross. 1992. Termination efficiency at rho-dependent terminators depends on kinetic coupling between RNA polymerase and rho. Proc. Natl. Acad. Sci. USA 89:1453–1457.
215. Jin, D. J., M. Cashel, D. J. Friedman, Y. Nakamura, W. A. Walter, and C. A. Gross. 1988. Effects of rifampicin resistant rpoB mutations on antitermination and interaction with nusA in Escherichia coli. J. Mol. Biol. 204:247–261.
216. Jin, D. J., and C. A. Gross. 1988. Mapping and sequencing of mutations in Escherichia coli rpoB gene that lead to rifampicin resistance. J. Mol. Biol. 202:45–58.
217. Jin, D. J., W. A. Walter, and C. A. Gross. 1988. Characterization of the termination phenotypes of rifampicin-resistant mutants. J. Mol. Biol. 202:245–253.
218. Jinks-Robertson, S., G. Baughman, and M. Nomura. 1984. Regulation of ribosome biosynthesis in E. coli, p. 395–409. In B. V. C. Clark and H. U. Petersen (ed.), Gene Expression. Alfred Benzon Symposium XIX. Munksgaard, Copenhagen.
219. Jinks-Robertson, S., R. L. Gourse, and M. Nomura. 1983. Expression of rRNA and tRNA genes in Escherichia coli: evidence for feedback regulation by products of rRNA operons. Cell 33:865–876.
220. Johnson, G. S., C. R. Adler, J. J. Collins, and D. Court. 1979. Role of the spoT gene product and manganese ion in the metabolism of guanosine 5'-diphosphate 3'-diphosphate in Escherichia coli. J. Biol. Chem. 254:5483–5487.
221. Jones, P. G., M. Cashel, G. Glaser, and F. C. Neidhardt. 1992. Function of a relaxed-like state following temperature downshifts in Escherichia coli. J. Bacteriol. 174:3903–3914.
222. Jones, P. G., and M. Inouye. 1994. The cold-shock response—a hot topic. Mol. Microbiol. 11:811–818.
223. Jones, P. G., R. Krah, S. R. Tafuri, and A. P. Wolffe. 1992. DNA gyrase, CS7.4, and the cold shock response in Escherichia coli. J. Bacteriol. 174:5798–5802.
224. Jones, P. G., R. A. VanBogelen, and F. C. Neidhardt. 1987. Induction of proteins in response to low temperaure in Escherichia coli. J. Bacteriol. 169:2092–2095.
225. Josaitis, C. A., T. Gaal, and R. L. Gourse. 1995. Stringent control and growth-rate-dependent control have non-identical promoter sequence requirements. Proc. Natl. Acad. Sci. USA 92:1117–1121.
226. Joseleau-Petit, D., D. Thevenet, and R. D’Ari. 1994. ppGpp concentration, growth without PBP2 activity and growth rate control in Escherichia coli. Mol. Microbiol. 13:911–917.
227. Kajitani, M., and A. Ishihama. 1983. Determination of promoter strength in the mixed transcription system: promoters of lactose, tryptophan, and ribosomal L10 operons from Escherichia coli. Nucleic Acids Res. 11:671–686.
228. Kajitani, M., and A. Ishihama. 1983. Determination of promoter strength in the mixed transcription system. II. Promoters of ribosomal RNA, ribosomal protein S1 and recA protein operons from Escherichia coli. Nucleic Acids Res. 11:3873–3888.
229. Kajitani, M., and A. Ishihama. 1984. Promoter selectivity of Escherichia coli RNA polymerase. Differential stringent control of the multiple promoters from ribosomal RNA and protein operons. J. Biol. Chem. 259:1951–1957.
230. Kalman, M., H. Murphy, and M. Cashel. 1991. rhlB, a new Escherichia coli K-12 gene with an RNA helicase-like protein sequence motif, one of at least five such possible genes in a prokaryote. New Biol. 3:886–895.
231. Kalman, M., H. Murphy, and M. Cashel. 1992. The nucleotide sequence of recG, the distal spo operon gene in Escherichia coli K-12. Gene 110:95–99.
232. Kaplan, S., A. G. Atherly, and A. Barrett. 1973. Synthesis of stable RNA in stringent Escherichia coli cells in the absence of charged transfer RNA. Proc. Natl. Acad. Sci. USA 70:689–692.
233. Kari, C., I. Torok, and A. Travers. 1977. ppGpp cycle in Escherichia coli. Mol. Gen. Genet. 150:249–255.
234. Keasling, J. D., L. Bertsch, and A. Kornberg. 1993. Guanosine pentaphosphate phosphohydrolase of Escherichia coli is a long-chain exopolyphosphatase. Proc. Natl. Acad. Sci. USA 90:7029–7033.
235. Kingston, R. E. 1983. Effects of deletions near Escherichia coli rrnB P2 on inhibition of in vitro transcription by guanosine tetraphosphate. Biochemistry 22:5249–5254.
236. Kingston, R. E., and M. J. Chamberlin. 1981. Pausing and attenuation of in vitro transcription in the rrnB operon of E. coli. Cell 27:523–531.
237. Kingston, R. R., R. R. Gutell, A. R. Taylor, and M. J. Chamberlin. 1981. Transcriptional mapping of plasmid pKK3535. Quantitation of the effect of guanosine tetraphosphate on binding of the rrnB promoters and a λ promoter with sequence homologies to the cII binding region. J. Mol. Biol. 146:433–449.
238. Kingston, R. E., W. C. Nierman, and M. J. Chamberlin. 1981. A direct effect of guanosine tetraphosphate on pausing of Escherichia coli RNA polymerase during RNA chain elongation. J. Biol. Chem. 256:2787–2797.
239. Kjeldgaard, N. O., O. Maaloe, and M. Schacter. 1958. The transition between different physiological states during balanced growth of Salmonella typhimurium. J. Gen. Microbiol. 19:607–616.
240. Kliachko, E. V., S. S. Bochanov, and R. S. Shakulov. 1983. Metabolic regulation of threonine operon transcription in E. coli cells. Biokhimiya 48:1095–1102.
241. Kogoma, T., and K. von Meyenburg. 1983. The origin of replication, oriC, and the DnaA protein are dispensable in stable DNA replication (sdrA) mutants of Escherichia coli. EMBO J. 2:463–468.
242. Kohiyama, M., D. Cousin, A. Ryter, and F. Jacob. 1966. Mutants thermosensibles d’Escherichia coli K12. I. Isolement et caracterization rapide. Ann. Inst. Pasteur (Paris) 110:465–486.
243. Komine, Y., T. Adachi, H. Inokuchi, and H. Ozeki. 1990. Genomic organization and physical mapping of the transfer RNA genes in Escherichia coli K12. J. Mol. Biol. 212:579–598.
244. Koonin, E. V., and K. E. Rudd. 1993. SpoU protein of Escherichia coli belongs to a new family of putative rRNA methylases. Nucleic Acids Res. 21:5519.
245. Kramer, G. F., J. C. Baker, and B. N. Ames. 1988. Near-UV stress in Salmonella typhimurium: 4-thiouridine in tRNA, ppGpp, and ApppGpp as components of an adaptive response. J. Bacteriol. 170:2344–2351.
246. Kurland, C. G., and O. Maaloe. 1962. Regulation of ribosomal and transfer RNA synthesis. J. Mol. Biol. 4:193–210.
247. Kusser, W., and E. E. Ishiguro. 1985. Involvement of the relA gene in the autolysis of Escherichia coli induced by inhibitors of peptidoglycan biosynthesis. J. Bacteriol. 164:861–865.
248. Kusser, W., and E. E. Ishiguro. 1986. Lysis of nongrowing Escherichia coli by combination of β-lactam antibiotics and inhibitors of ribosome function. Antimicrob. Agents Chemother. 29:451–455.
249. Kusser, W., and E. E. Ishiguro. 1987. Suppression of mutations conferring penicillin tolerance by interference with the stringent control mechanism of Escherichia coli. J. Bacteriol. 169:4396–4398.
250. Laffler, T., and J. Gallant. 1974. spoT, a new genetic locus involved in the stringent response of E. coli. Cell 1:27–30.
251. Laffler, T., and J. Gallant. 1974. Stringent control of protein synthesis in E. coli. Cell 3:47–49.
252. Lagosky, P. A., and F. N. Chang. 1978. The extraction of guanosine 5'-diphosphate, 3'-diphosphate (ppGpp) from Escherichia coli using low pH reagents: a re-evaluation. Biochem. Biophys. Res. Commun. 84:1016–1024.
253. Lagosky, P. A., and F. N. Chang. 1980. Influence of amino acid starvation of guanosine 5'-diphosphate, 3'-diphosphate basal level synthesis in Escherichia coli. J. Bacteriol. 144:499–508.
254. Lagosky, P. A., and F. N. Chang. 1981. Correlation between RNA synthesis and basal level guanosine 5'-diphosphate, 3'-diphosphate in relaxed mutants of Escherichia coli. J. Biol. Chem. 256:11651–11656.
255. Lamond, A. I. 1985. Supercoiling response of a bacterial tRNA gene. EMBO J. 4:501–507.
256. Lamond, A. I., and A. A. Travers. 1983. Requirement for an upstream element for optimal transcription of a bacterial tRNA gene. Nature (London) 305:248–250.
257. Lamond, A. I., and A. A. Travers. 1985. Genetically separable functional elements mediate the optimal expression and stringent regulation of a bacterial tRNA gene. Cell 40:319–326.
258. Landick, R., J. Stewart, and D. N. Lee. 1990. Amino acid changes in conserved regions of the β-subunit of Escherichia coli RNA polymerase alter transcription pausing and termination. Genes Dev. 4:1623–1636.
259. Landick, R. C. 1983. Ph.D. thesis. University of Michigan, Ann Arbor.
260. Lange, R., and R. Hengge-Aronis. 1991. Growth-phase regulated expression of bolA and of stationary-phase Escherichia coli cells are controlled by the novel sigma factor σS. J. Bacteriol. 173:4474–4481.
261. Lange, R., and R. Hengge-Aronis. 1994. The cellular concentraton of the σS subunit of RNA polymerase in Escherichia coli is controlled at the levels of transcription, translation, and protein stability. Genes Dev. 8:1600–1612.
262. LaTeana, A., A. Brandi, M. Falconi, R. Spurio, C. C. Pon, and C. O. Gualerzi. 1991. Identification of a cold shock transcriptional enhancer of the Escherichia coli gene encoding nucleoid protein H-NS. Proc. Natl. Acad. Sci. USA 88:10907–10911.
263. Lavalle, R. 1965. Nouveaux mutants de regulation de la synthese de l’ARN. Bull. Soc. Chim. Biol. 47:1567–1570.
264. Lavalle, R., L. Desmarez, and G. deHauwer. 1976. Natural message translation impairment in an E. coli mutant, p. 408–416. In N. O. Kjeldgaard and O. Maaloe (ed.), Control of Ribosome Synthesis. Alfred Benzon Symposium IX. Munksgaard, Copenhagen.
265. Lazarus, L. R., and A. A. Travers. 1993. The Escherichia coli Fis protein is not required for the activation of tyrT transcription on entry into exponential growth. EMBO J. 12:2483–2494.
266. Lazzarini, R. A., M. Cashel, and J. Gallant. 1971. On the regulation of guanosine tetraphosphate levels in stringent and relaxed strains of Escherichia coli. J. Biol. Chem. 246:4381–4385.
267. Lazzarini, R. A., and A. E. Dahlberg. 1971. The control of ribonucleic acid synthesis during amino acid deprivation in Escherichia coli. J. Biol. Chem. 246:420–429.
268. Lehner, A. F., S. Harvey, and C. W. Hill. 1984. Mapping and spacer identification of rRNA operons of Salmonella typhimurium. J. Bacteriol. 160:682–686.
269. Leung, K. L., and H. Yamazaki. 1977. Synthesis of ppGpp by ribosomes from an Escherichia coli spoT mutant and the metabolic relationship between pppGpp and ppGpp. Can. J. Biochem. 55:1207–1212.
270. Lewis, J. A., and B. N. Ames. 1972. Histidine regulation in Salmonella typhimurium. XI. The percentage of transfer RNAHis charged in vivo and its relation to the repression of the histidine operon. J. Mol. Biol. 66:131–142.
271. Li, S., C. L. Squires, and C. Squires. 1984. Antitermination of E. coli rRNA transcription is caused by a control region segment containing a nut-like sequence. Cell 38:851–860.
272. Lideman, L. F., O. I. Ponomarenko, and R. S. Shakulov. 1979. Synthesis of beta, beta'-subunits of RNA-polymerase in E. coli cells starved for an essential amino acid. Biokhimiya 44:1859–1863.
273. Liebke, H. H., and J. F. Speyer. 1983. A new gene in E. coli RNA synthesis. Mol. Gen. Genet. 189:314–320.
274. Liermo, S., and R. L. Gourse. 1991. Factor-independent activation of Escherichia coli rRNA transcription. I. Kinetic analysis of the roles of the upstream activator region and supercoiling on transcription of the rrnB P1 promoter in vitro. J. Mol. Biol. 220:555–568.
275. Lindahl, L., and J. M. Zengel. 1982. Expression of ribosomal genes in bacteria. Adv. Genet. 21:53–121.
276. Lipmann, F., and J. Sy. 1976. The enzymatic mechanism of guanosine 5'-3'-polyphosphate synthesis. Prog. Nucleic Acid Res. Mol. Biol. 17:1–14.
277. Little, R., and H. Bremer. 1982. Quantitation of guanosine 5',3'-bis-diphosphate in extracts from bacterial cells by ion-pair reverse-phase high-performance liquid chromatography. Anal. Biochem. 126:381–388.
278. Little, R., and H. Bremer. 1984. Transcription of ribosomal component genes and lac in a relA +/relA pair of Escherichia coli strains. J. Bacteriol. 159:863–869.
279. Little, R., S. W. Mou, and H. Bremer. 1984. Isolation of rel mutants of Escherichia coli B/r. J. Bacteriol. 155:1426–1428.
280. Little, R., J. Ryals, and H. Bremer. 1983. rpoB mutation in Escherichia coli alters control of ribosome synthesis by guanosine tetraphosphate. J. Bacteriol. 154:787–792.
281. Little, R., J. Ryals, and H. Bremer. 1983. Physiological characterization of Escherichia coli rpoB mutants with abnormal control of ribosome synthesis. J. Bacteriol. 155:1162–1170.
282. Lloyd, R. G., and G. J. Sharples. 1991. Molecular organization and nucleotide sequence of the recG locus of Escherichia coli K-12. J. Bacteriol. 173:6837–6843.
283. Lobner-Oleson, A., K. Skarstad, F. G. Hansen, K. von Meyenburg, and E. Boye. 1989. The DnaA protein determines the initiation mass of Escherichia coli K-12. Cell 57:881–889.
284. Loewen, P. C., I. von Ossowski, J. Switala, and M. Mulvey. 1993. KatF (σS) synthesis in Escherichia coli is subject to postranscriptional regulation. J. Bacteriol. 175:2150–2153.
285. Lonetto, M., M. Gribskov, and C. A. Gross. 1992. The σ70 family: sequence conservation and evolutionary relationships. J. Bacteriol. 174:3843–3849.
286. Lotz, W. 1978. Effect of guanosine tetraphosphate on in vitro protein synthesis directed by E1 and E3 colicinogenic factors. J. Bacteriol. 135:707–712.
287. Lueking, D. R., and H. Goldfine. 1975. The involvement of ppGpp in the regulation of phospholipid biosynthesis in Escherichia coli: lack of ppGpp inhibition of acyl transfer from acyl-ACP to sn-glycerol-3-phosphate. J. Biol. Chem. 250:4911–4917.
288. Lund, E., and J. E. Dahlberg. 1979. Initiation of Escherichia coli ribosomal RNA synthesis in vivo. Proc. Natl. Acad. Sci. USA 76:5480–5484.
289. Lund, E., and N. O. Kjeldgaard. 1972. Metabolism of guanosine tetraphosphate in Escherichia coli. Eur. J. Biochem. 28:316–326.
290. Lund, E., and N. O. Kjeldgaard. 1972. Protein synthesis and formation of guanosine tetraphosphate. FEBS Lett. 26:306–310.
291. Lund, F., and G. Tybring. 1972. 6β-Amidinopenicillanic acid - a new group of antibiotics. Nature (London) New Biol. 236:135–137.
292. Maaloe, O. 1979. Regulation of the protein synthesizing machinery—ribosomes, tRNA, factors, and so on, p. 487–542. In R. Goldberger (ed.), Biological Regulation and Development, vol. I. Gene Expression. Plenum Publishing Corp., New York.
293. Maaloe, O., and N. O. Kjeldgaard. 1966. Control of Macromolecular Synthesis. W. A. Benjamin, New York.
294. Marchal, J., J. C. Cortay, and A. J. Cozzone. 1983. Functional aspects of bacterial polysomes during limited protein synthesis. Biochim. Biophys. Acta 739:326–333.
295. Masuda, Y., K. Miyakawa, Y. Nishimura, and E. Ohtsubo. 1993. chpA and chpB, Escherichia coli chromosomal homologs of the pem locus responsible for stable maintenance of plasmid R100. J. Bacteriol. 175:6850–6856.
296. Matsuhashi, S., T. Kamiryo, P. M. Blumberg, P. Linnett, E. Willoughby, and J. L. Strominger. 1974. Mechanism of action and development of resistance to a new amidino penicillin. J. Bacteriol. 117:578–587.
297. Matzura, H., and P. Broda. 1968. Sensitization of Escherichia coli to actinomycin D by the arrest of protein synthesis. J. Bacteriol. 96:1877–1879.
298. McCann, M. P., C. D. Fraley, and A. Matin. 1993. The putative σ factor KatF is regulated posttranscriptionally during carbon starvation. J. Bacteriol. 175:2143–2149.
299. 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. 226:581–596.
300. Merlie, J. P., and L. I. Pizer. 1973. Regulation of phospholipid synthesis in Escherichia coli by guanosine tetraphosphate. J. Bacteriol. 116:355–366.
301. Metzger, S., I. Ben-Dror, E. Aizenman, G. Schreiber, M. Toone, J. D. Friesen, M. Cashel, and G. Glaser. 1988. The nucleotide sequence and characterization of the relA gene of Escherichia coli. J. Biol. Chem. 264:15699–15704.
302. Metzger, S., E. Sarubbi, G. Glaser, and M. Cashel. 1989. Protein sequences encoded by the relA and spoT genes of Escherichia coli are interrelated. J. Biol. Chem. 264:9122–9125.
303. Metzger, S., G. Schreiber, E. Aizenman, M. Cashel, and G. Glaser. 1989. Characterization of the relA1 mutation and a comparison of relA1 with new relA null alleles in Escherichia coli. J. Biol. Chem. 264:21146–21152.
304. Meza-Basso, L., N. Robakis, Y. Cenatiempo, H. Weissbach, and N. Brot. 1981. Guanosine-5'-diphosphate-3'-diphosphate inhibits the in vitro synthesis of β-lactamase from pBR322 DNA. Biochem. Biophys. Res. Commun. 101:459–463.
305. Mitchell, J. J., and J. M. Lucas-Lenard. 1980. The effect of alcohols on guanosine 5'-diphosphate-3'-diphosphate metabolism in stringent and relaxed Escherichia coli. J. Biol. Chem. 255:6307–6313.
306. Miura, A., J. H. Kreuger, S. Itoh, H. A. deBoer, and M. Nomura. 1981. Growth-rate-dependent regulation of ribosome synthesis in E. coli: expression of the lacZ and galK genes fused to ribosomal promoters. Cell 25:773–782.
307. Miyajima, A., and Y. Kaziro. 1978. Coordination of levels of elongation factors Tu, Ts and G, and ribosomal protein S1 in Escherichia coli. J. Biochem. (Tokyo) 83:453–462.
308. Miyajima, A., M. Shibuya, Y. Kuchino, and Y. Kaziro. 1981. Transcription of the E. coli tufB gene: cotranscription with four tRNA genes and inhibition by guanosine-5'-diphosphate-3'-diphosphate. Mol. Gen. Genet. 183:13–19.
309. Mizushima-Sugano, J., and Y. Kaziro. 1985. Regulation of the expression of the tufB operon: DNA sequences directly involved in the stringent control. EMBO J. 4:1053–1058.
310. Mizushima-Sugano, J., J. A. Miyajima, and Y. Kaziro. 1983. Selective inhibition of transcription of the E. coli operon by guanosine-5'-diphosphate-3'-diphosphate. Mol. Gen. Genet. 189:185–192.
311. Moazed, D., and H. F. Noller. 1989. Interaction of tRNA with 23S rRNA in the ribosomal A, P, and E sites. Cell 57:585–597.
312. Molin, S., K. von Meyenburg, O. Maaloe, M. T. Hansen, and M. L. Pato. 1977. Control of ribosome synthesis in Escherichia coli: analysis of an energy source shift-down. J. Bacteriol. 131:7–17.
313. Morgan, S. D., and D. Soll. 1978. Regulation of the biosynthesis of aminoacid:tRNA ligases and of tRNA. Prog. Nucleic Acid Res. Mol. Biol. 21:181–207.
314. Morton, D. P., and S. M. Parsons. 1977. Synergistic inhibition of ATP phosphoribosyltransferase by guanosine tetraphosphate and histidine. Biochem. Biophys. Res. Commun. 74:172–177.
315. Mosteller, R. D. 1978. Evidence that glucose starvation-sensitive mutants are altered in the relB locus. J. Bacteriol. 133:1034–1037.
316. Mosteller, R. D., and S. F. Kwan. 1976. Isolation of relaxed-control mutants of Escherichia coli K-12 which are sensitive to glucose starvation. Biochem. Biophys. Res. Commun. 69:325–332.
317. Muto. A. 1977. Control of ribosomal RNA synthesis in Escherichia coli. III. Cytoplasmic factors for ribosomal RNA synthesis. Mol. Gen. Genet. 152:161–165.
318. Muto, A. 1981. Control of ribosomal RNA synthesis in Escherichia coli. V. Stimulation of rrnC gene transcription in vitro by a protein factor. Mol. Gen. Genet. 181:69–73.
319. Nachaliel, N., J. Melnick, R. Gafny, and G. Glaser. 1989. Ribosome associated protein(s) specifically bind(s) to the upstream activator sequence of the E. coli rrnA P1 promoter. Nucleic Acids Res. 17:9811–9822.
320. Nagase, T., S. Ishii, and F. Imamoto. 1988. Differential transcriptional control of the two tRNA(fmet) genes of Escherichia coli K-12. Gene 67:49–57.
321. Neidhardt, F. C. 1963. Properties of a bacterial mutant lacking amino acid control of RNA synthesis. Biochim. Biophys. Acta 68:365–379.
322. Neidhardt, F. C. 1966. Role of amino acid activating enzymes in cellular physiology. Bacteriol. Rev. 30:701–719.
323. Neidhardt, F. C., J. L. Ingraham, and M. Schaechter. 1990. Physiology of the Bacterial Cell. A Molecular Approach. Sinauer Associates, Sunderland, Mass.
324. Neidhardt, F. C., V. Vaughn, T. A. Phillips, and P. L. Bloch. 1983. Gene-protein index of Escherichia coli K-12. Microbiol. Rev. 47:231–284.
325. Nene, V., and R. E. Glass. 1982. Genetic studies on the β subunit of Escherichia coli RNA polymerase. I. The effect of known, single amino acid substitutions in an essential protein. Mol. Gen. Genet. 188:399–404.
326. Nene, V., and R. E. Glass. 1983. Relaxed mutants of Escherichia coli RNA polymerase. FEBS Lett. 153:307–310.
327. Newlands, J. T., T. Gaal, J. Mecsas, and R. L. Gourse. 1993. Transcription of the Escherichia coli rrnB P1 promoter by the heat shock RNA polymerase (Eσ32) in vitro. J. Bacteriol. 175:661–668.
328. Newlands, J. T., C. A. Josaitis, W. Ross, and R. L. Gourse. 1992. Both fis-dependent and factor-independent upstream activation of the rrnB P1 promoter are face of the helix dependent. Nucleic Acids Res. 20:719–726.
329. Newlands, J. T., W. Ross, K. K. Gosink, and R. L. Gourse. 1991. Factor-independent activation of Escherichia coli rRNA transcription. II. Characterization of complexes of rrnB P1 promoters containing or lacking the upstream activator region with Escherichia coli RNA polymerase. J. Mol. Biol. 220:569–583.
330. Nierlich, D. P. 1968. Amino acid control over RNA synthesis, a re-evaluation. Proc. Natl. Acad. Sci. USA 60:1345–1352.
331. Nierlich, D. P. 1978. Regulation of bacterial growth, RNA, and protein synthesis. Annu. Rev. Microbiol. 32:393–432.
332. Nilsson, G., J. G. Belasco, S. N. Cohen, and A. von Gabain. 1984. Growth-rate-dependent regulation of mRNA stability in E. coli. Nature (London) 312:75–77.
333. Nilsson, L., A. Vanet, E. Vigenboom, and L. Bosch. 1990. The role of Fis in trans activation of stable RNA operons of E. coli. EMBO J. 9:727–734.
334. Nilsson, L., H. Verbeek, E. Vijgenboom, C. van Drunen, A. Vanet, and L. Bosch. 1992. Fis-dependent trans activation of stable RNA operons of Escherichia coli under various growth conditions. J. Bacteriol. 174:921–929.
335. Ninnemann, O., C. Koch, and R. Kahmann. 1992. The Escherichia coli fis promoter is subject to stringent control and autoregulation. EMBO J. 11:1075–1083.
336. Nishizuka, Y., and F. Lipmann. 1966. The interrelationships between guanosine triphosphatase and amino acid polymerization. Arch. Biochem. Biophys. 116:344–351.
337. Nomura, M., R. Gourse, and G. Baughman. 1984. Regulation of the synthesis of ribosomes and ribosomal components. Annu. Rev. Biochem. 53:75–117.
338. Nomura, M., S. Jinks-Robertson, and A. Miura. 1982. Regulation of ribosome biosynthesis in Escherichia coli, p. 91–104. In M. Grunberg-Manago and B. Saver (ed.), Interaction of Translational and Transcriptional Controls in the Regulation of Gene Expression. Elsevier Science Publishing Co., New York.
339. Norris, T. E., and A. Koch. 1972. Effect of growth rate on the relative rates of messenger, ribosomal, and transfer RNA in Escherichia coli. J. Mol. Biol. 64:633–649.
340. Nunn, W., and J. Cronan. 1974. rel gene control of lipid synthesis in Escherichia coli. Evidence for eliminating fatty acid synthesis as the sole regulatory site. J. Biol. Chem. 249:3994–3996.
341. Ny, T., and G. R. Bjork. 1977. Stringent regulation of the synthesis of a transfer ribonucleic acid biosynthetic enzyme: transfer ribonucleic acid(m5U)methyltransferase from Escherichia coli. J. Bacteriol. 130:635–641.
342. Ochi, K., J. Kandala, and E. Freese. 1982. Evidence that Bacillus subtilis sporulation induced by the stringent response is caused by the decrease in GTP or GDP. J. Bacteriol. 151:1062–1065.
343. O’Farrell, P. H. 1978. The suppression of defective translation by ppGpp and its role in the stringent response. Cell 14:545–557.
344. Ogawa, T., and T. Okazaki. 1991. Concurrent transcription from the gid and mioC promoters activates replication of an Escherichia coli minichromosome. Mol. Gen. Genet. 230:193–200.
345. Ogawa, Y., and J. Sy. 1977. Synthesis of guanosine polyphosphates (pppGpp and ppGpp) and its regulation by aminoacyl-tRNA. J. Biochem. 82:947–953.
346. Ogura, T., P. Bouloc, H. Niki, R. D’Ari, S. Hiraga, and A. Jaffe. 1989. Penicillin-binding protein 2 is essential in wild-type Escherichia coli but not in lov or cya mutants. J. Bacteriol. 171:3025–3030.
347. Ohlsen, K. L., and J. D. Gralla. 1992. Interrelated effects of DNA supercoiling, ppGpp, and low salt on melting within the Escherichia coli ribosomal RNA rrnB P1 promoter. Mol. Microbiol. 6:2243–2251.
348. Ohlsen, K. L., and J. D. Gralla. 1992. DNA melting within stable closed complexes at Escherichia coli rrnB P1 promoter. J. Biol. Chem. 267:19813–19818.
349. Ohlsen, K. L., and J. D. Gralla. 1992. Melting during steady-state transcription of the rrnB P1 promoter in vivo and in vitro. J. Bacteriol. 174:6071–6075.
350. O’Neal, C. R., W. M. Gabriel, A. K. Turk, S. J. Libby, F. C. Fang, and M. P. Spector. 1994. RpoS is necessary for both positive and negative regulation of starvation genes during phosphate, carbon, and nitrogen starvation in Salmonella typhimurium. J. Bacteriol. 176:4610–4616.
351. Oostra, B. A., G. Ab, and M. Gruber. 1980. Specific stimulation of ribosomal RNA synthesis in E. coli by a protein factor. Mol. Gen. Genet. 177:291–295.
352. Oostra, B. A., A. J. J. van Ooyen, and M. Gruber. 1977. In vitro transcription of three different ribosomal RNA cistrons of E. coli: heterogeneity of control regions. Mol. Gen. Genet. 152:106.
353. Ota, Y., A. Kikuchi, and M. Cashel. 1979. Gene expression of an Escherichia coli ribosomal RNA promoter fused to structural genes of the galactose operon. Proc. Natl. Acad. Sci. USA 76:5799–5803.
354. Pao, C. C., and B. T. Dyess. 1981. Stringent control of RNA synthesis in the absence of guanosine 5'-diphosphate-3'-diphosphate. J. Biol. Chem. 256:2252–2257.
355. Pao, C. C., and B. T. Dyess. 1981. Effect of unusual guanosine nucleotides on the activities of some Escherichia coli cellular enzymes. Biochim. Biophys. Acta 677:358–362.
356. Pao, C. C., and J. Gallant. 1978. A gene involved in the metabolic control of ppGpp synthesis. Mol. Gen. Genet. 158:271–277.
357. Pao, C. C., and J. Gallant. 1979. A new nucleotide involved in the stringent response in Escherichia coli. J. Biol. Chem. 254:688–692.
358. Parker, J., and J. D. Friesen. 1980. "Two out of three" codon reading leading to mistranslation in vivo. Mol. Gen. Genet. 177:439–445.
359. Parker, J., J. W. Pollard, J. D. Friesen, and C. P. Stanners. 1978. Stuttering: high-level mistranslation in animal and bacterial cells. Proc. Natl. Acad. Sci. USA 75:1091–1095.
360. Patte, J., P. Morand, E. Boy, C. Richaud, and F. Borne. 1980. The relA locus and the regulation of lysine biosynthesis in Escherichia coli. Mol. Gen. Genet. 179:319–325.
361. Pedersen, F. S. 1976. Stringent factor, p. 419–425. In N. O. Kjeldgaard and O. Maaloe (ed.), Control of Ribosome Synthesis. Alfred Benzon Symposium IX. Munksgaard, Copenhagen.
362. Pedersen, F. S., and N. O. Kjeldgaard. 1977. Analysis of the relA gene product of Escherichia coli. Eur. J. Biochem. 76:91–97.
363. Pedersen, F. S., E. Lund, and N. O. Kjeldgaard. 1973. Codon specific, tRNA-dependent in vitro synthesis of ppGpp and pppGpp. Nature (London) 243:13–15.
364. Perelman, B. V., and R. S. Shakulov. 1981. Dependence of threonine operon expression on the relA gene allelic state and the level of guanosine tetraphosphate in E. coli. Biokhimiya 46:1267–1276.
365. Perelman, B. V., R. S. Shakulov, I. V. Smirnov, A. F. Lisenkov, and G. V. Astvatsaturiants. 1988. Metabolic regulation of the histidine operon in Escherichia coli and Salmonella typhimurium. Mol. Gen. Microbiol. Virusol. 33:36–40.
366. Piepersberg, W., D. Geyl, P. Buckel, and A. Bock. 1979. Studies on the coordination of tRNA-charging and polypeptide synthesis activity in Escherichia coli, p. 39–52. In G. Koch and D. Richter (ed.), Regulation of Macromolecular Synthesis by Low Molecular Weight Mediators. Academic Press, Inc., New York.
367. Pisbarro, A. G., M. A. De Pedro, and E. E. Ishiguro. 1990. Dissociation of the ampicillin-induced lysis of amino acid-deprived Escherichia coli into two stages. J. Bacteriol. 172:2187–2190.
368. Polakis, S. E., R. B. Guchhait, and M. D. Lane. 1973. Stringent control of fatty acid synthesis in Escherichia coli: possible regulation of acetyl CoA carboxylase by ppGpp. J. Biol. Chem. 248:7957–7966.
369. Primakoff, P. 1981. In vivo role of the relA + gene in the regulation of the lac operon. J. Bacteriol. 145:410–416.
370. Primakoff, P., and S. W. Artz. 1979. Positive control of lac operon expression in vitro by guanosine 5'-diphosphate, 3'-diphosphate. Proc. Natl. Acad. Sci. USA 76:1726–1730.
371. Pritchard, R. H., P. T. Barth, and J. Collins. 1969. Control of DNA synthesis in bacteria. Microbial growth. Symp. Soc. Gen. Microbiol. 19:263–297.
372. Quay, S. C., and D. L. Oxender. 1979. The relA locus specifies a positive effector in branched-chain amino acid transport regulation. J. Bacteriol. 137:1059–1062.
373. Que, L., G. R. Willie, M. Cashel, J. W. Bodley, and G. R. Gray. 1973. Guanosine 5'-diphosphate, 3'-diphosphate: assignment of structure by 13C nuclear magnetic resonance spectroscopy. Proc. Natl. Acad. Sci. USA 70:2563–2566.
374. Ramabhadrian, T. V., and J. Jagger. 1976. Mechanism of growth delay induced in Escherichia coli by near ultraviolet radiation. Proc. Natl. Acad. Sci. USA 73:59–63.
375. Ramagopal, S., and B. D. Davis. 1974. Localization of the stringent protein of Escherichia coli on the 50S ribosomal subunit. Proc. Natl. Acad. Sci. USA 71:820–824.
376. Rao, L., W. Ross, J. A. Appleman, T. Gaal, S. Liermo, P. J. Schlax, M. T. Record, Jr., and R. L. Gourse. 1994. Factor independent activation of rrnB P1. An extended promoter with an upstream element that dramatically increases promoter strength. J. Mol. Biol. 235:1421–1435.
377. Raue, H. A., and M. Cashel. 1974. Regulation of RNA synthesis in Escherichia coli. III. Degradation of guanosine 5'-diphosphate 3'-diphosphate in cold-shocked cells. Biochim. Biophys. Acta 383:290–289.
378. Reddy, P. S., A. Raghavan, and D. Chatterji. 1995. Evidence for a ppGpp-binding site on Escherichia coli RNA polymerase: proximity relationship with the rifampicin-binding domain. Mol. Microbiol. 15:225–265.
379. Reeh, S., S. Pedersen, and J. D. Friesen. 1976. Biosynthetic regulation of individual proteins in relA+ and relA strains of Escherichia coli during amino acid starvation. Mol. Gen. Genet. 149:279–289.
380. Reiness, G., H. Yang, G. Zubay, and M. Cashel. 1975. Effects of guanosine tetraphosphate on cell-free synthesis of Escherichia coli ribosomal RNA and other gene products. Proc. Natl. Acad. Sci. USA 72:2881–2885.
381. Reizer, J., A. Reizer, and M. H. Saier, Jr., P. Bork, and C. Sander. 1993. Exopolyphosphate phosphatase and guanosine pentaphosphate phosphatase belong to the sugar kinase/actin/hsp70 superfamily. Trends Biochem. Sci. 18:247–248.
382. Richter, D. 1976. Stringent factor from Escherichia coli directs ribosomal binding and release of uncharged tRNA. Proc. Natl. Acad. Sci. USA 73:707–711.
383. Richter, D. 1979. Synthesis and degradation of the pleiotropic effector, guanosine 3',5-bis(diphosphate) in bacteria, p. 85–94. In G. Koch and D. Richter (ed.), Regulation of Macromolecular Synthesis by Low Molecular Weight Mediators. Academic Press, Inc., New York.
384. Richter, D. 1980. In vitro degradation of guanosine 3',5'-bis(diphosphate) [ppGpp] by the spoT gene product [ppGppase] from auxotrophic strains of Escherichia coli: effects of various antibiotics and drugs. Arch. Microbiol. 124:229–232.
385. Richter, D. 1980. Uncharged tRNA inhibits guanosine 3',5'-bis(diphosphate) 3'-pyrophosphohydrolase [ppGppase], the spoT gene product, from Escherichia coli. Mol. Gen. Genet. 178:325–327.
386. Richter, D., V. A. Erdmann, and M. Sprinzl. 1974. A new transfer RNA fragment reaction: TpYpCpGp bound to a ribosome-messenger RNA complex induces the synthesis of guanosine tetra- and pentaphosphates. Proc. Natl. Acad. Sci. USA 71:3226–3229.
387. Richter, D., S. Fehr, and R. Harder. 1979. The guanosine 3',5'-bis(diphosphate) (ppGpp) cycle. comparison of synthesis and degradation of guanosine 3',5'-bis(diphosphate) in various bacterial systems. Eur. J. Biochem. 99:57–64.
388. Riggs, D. L., R. D. Mueller, H. Kwan, and S. W. Artz. 1986. Promoter domain mediates guanosine tetraphosphate activation of the histidine operon. Proc. Natl. Acad. Sci. USA 83:9333–9337.
389. Robin, A., D. Joseleau-Petit, and R. D’Ari. 1990. Transcription of the ftsZ gene and cell division in Escherichia coli. J. Bacteriol. 172:1392–1399.
390. Robinson, A. C., D. J. Kenan, J. Sweeney, and W. D. Donachie. 1986. Further evidence for overlapping transcriptional units in an Escherichia coli cell envelope-cell division gene cluster: DNA sequence and transcriptional organization of the ddl ftsQ region. J. Bacteriol. 167:809–817.
391. Rodriguez, S. B., and J. L. Ingraham. 1983. Location on the Salmonella typhimurium chromosome of the gene encoding nucleoside diphosphokinase (ndk). J. Bacteriol. 153:1101–1103.
392. Rohlman, C. E., and R. G. Matthews. 1990. Role of purine biosynthetic intermediates in response to folate stress in Escherichia coli. J. Bacteriol. 172:7200–7210.
393. Rojiani, M. V., H. Jakubowski, and M. Goldman. 1989. Effect of variation of charged and uncharged tRNA(trp) levels on ppGpp synthesis in Escherichia coli. J. Bacteriol. 171:6493–6502.
394. Rojiani, M. V., H. Jakubowski, and M. Goldman. 1990. Relationship between protein synthesis and concentrations of charged and uncharged tRNATrp in Escherichia coli. Proc. Natl. Acad. Sci. USA 87:1511–1515.
395. Rokeach, L. A., G. A. Kassavetis, and J. W. Zyskind. 1987. RNA polymerase pauses in vitro within the Escherichia coli origin of replication at the same sites where termination occurs in vivo. J. Biol. Chem. 262:7264–7272.
396. Rokeach, L. A., and J. W. Zyskind. 1986. RNA terminating within the E. coli origin of replication: stringent regulation and control by DnaA protein. Cell 46:763–771.
397. Romeo, T., M. Gong, M. Y. Liu, and A. M. Brun-Zinkernagel. 1993. Identification and molecular characterization of csrA, a pleiotropic gene from Escherichia coli that affects glycogen biosynthesis, gluconeogenesis, cell size, and surface properties. J. Bacteriol. 175:4744–4755.
398. Ross, W., K. K. Gosink, J. Salomon, K. Igarashi, C. Zou, A. Ishihama, K. Severinov, and R. L. Gourse. 1993. A third recognition element in bacterial promoters: DNA binding by the alpha subunit of RNA polymerase. Science 262:1407–1413.
399. Ross, W., J. F. Thompson, J. T. Newlands, and R. L. Gourse. 1990. E. coli Fis protein activates ribosomal RNA transcription in vitro and in vivo. EMBO J. 9:3733–3742.
400. Rowley, K. B., R. M. Elford, I. Roberts, and W. M. Holmes. 1993. In vivo regulatory responses of four Escherichia coli operons which encode leucyl-tRNAs. J. Bacteriol. 175:1309–1315.
401. Rudd, K. E., B. R. Bochner, M. Cashel, and J. R. Roth. 1985. Mutations in the spoT gene of Salmonella typhimurium: effects on his operon expression. J. Bacteriol. 163:534–542.
402. Rudd, K. E., and R. Menzel. 1987. his operons of Escherichia coli and Salmonella typhimurium are regulated by DNA supercoiling. Proc. Natl. Acad. Sci. USA 84:517–521.
403. Ryals, J., and H. Bremer. 1982. relA-dependent RNA polymerase activity in Escherichia coli. J. Bacteriol. 150:168–179.
404. Ryals, J., R.-Y. Hsu, M. N. Lipsett, and H. Bremer. 1982. Isolation of single-site Escherichia coli mutants deficient in thiamine and 4-thiouridine syntheses: identification of a NUV mutant. J. Bacteriol. 151:899–904.
405. Ryals, J., R. Little, and H. Bremer. 1982. Control of RNA synthesis in Escherichia coli after a shift to higher temperature. J. Bacteriol. 151:1425–1432.
406. Ryals, J., R. Little, and H. Bremer. 1982. Temperature dependence of RNA synthesis parameters in Escherichia coli. J. Bacteriol. 151:879–887.
407. Ryals, J., R. Little, and H. Bremer. 1982. Control of rRNA and tRNA synthesis in Escherichia coli by guanosine tetraphosphate. J. Bacteriol. 151:1261–1268.
408. Sands, M. K., and R. B. Roberts. 1952. The effects of a tryptophan-histidine deficiency in a mutant of Escherichia coli. J. Bacteriol. 63:505–511.
409. Sarmientos, P., and M. Cashel. 1983. Carbon starvation and growth rate-dependent regulation of the Escherichia coli ribosomal RNA promoters: differential control of dual promoters. Proc. Natl. Acad. Sci. USA 80:7010–7013.
410. Sarmientos, P., J. E. Sylvester, S. Contente, and M. Cashel. 1983. Differential stringent control of the tandem Escherichia coli ribosomal RNA promoters from the rrnA operon expresssed in vivo in multicopy plasmids. Cell 32:1337–1346.
411. Sarubbi, E., K. E. Rudd, and M. Cashel. 1988. Basal ppGpp level adjustment shown by new spoT mutants affect steady state growth rates and rrnA ribosomal promoter regulation in Escherichia coli. Mol. Gen. Genet. 213:214–222.
412. Sarubbi, E., K. E. Rudd, H. Xiao, K. Ikehara, M. Kalman, and M. Cashel. 1989. Characterization of the spoT gene of Escherichia coli. J. Biol. Chem. 264:15074–15082.
413. Schaechter, M., O. Maaloe, and N. O. Kjeldgaard. 1958. Dependency on medium and temperature of cell size and chemical composition during balanced growth of Salmonella typhimurium. J. Gen. Microbiol. 19:592–606.
414. Schreiber, G. S., Metzger, E. Aizenman, S. Roza, M. Cashel, and G. Glaser. 1991. Overexpression of the relA gene in Escherichia coli. J. Biol. Chem. 266:3760–3767.
415. Schreiber, G. S., E. Z. Ron, and G. Glaser. 1995. ppGpp mediated regulation of DNA replication and cell division in Escherichia coli. Curr. Microbiol. 30:27–32.
416. Serizawa, H., and R. Fukuda. 1987. Structure of the gene for the stringent starvation protein of Escherichia coli. Nucleic Acids Res. 15:1153–1163.
417. Seyfzadeh, M., J. Keener, and M. Nomura. 1993. spoT-dependent accumulation of guanosine tetraphosphate in response to fatty acid starvation in Escherichia coli. Proc. Natl. Acad. Sci. USA 90:11004–11008.
418. Shand, R. F., P. H. Blum, R. D. Mueller, D. L. Riggs, and S. W. Artz. 1989. Correlation between histidine operon expression and guanosine 5'-diphosphate-3'-diphosphate levels during amino acid downshift in stringent and relaxed strains of Salmonella typhimurium. J. Bacteriol. 171:737–743.
419. Sheperd, N., G. Churchward, and H. Bremer. 1980. Synthesis and activity of ribonucleic acid polymerase in Escherichia coli. J. Bacteriol. 141:1089–1108.
420. Shimm, L. C., D. Vanderwel, R. E. Harkness, B. R. Currie, A. Galloway, and E. E. Ishiguro. 1984. Temperature-sensitive β-lactam-tolerant mutants of Escherichia coli. J. Gen. Microbiol. 130:1315–1323.
421. Silverman, R. H., and A. G. Atherly. 1978. Unusual effects of 5α,6-anhydrotetracycline and other tetracyclines. Inhibition of guanosine 5'-diphosphate 3'-diphosphate metabolism, RNA accumulation and other growth-related processes in Escherichia coli. Biochim. Biophys. Acta 518:267–276.
422. Singer, C. E., G. R. Smith, R. Cortese, and B. N. Ames. 1972. Mutant of tRNAHis ineffective in repression and lacking two pseudouridine modifications. Nature (London) New Biol. 238:72–74.
423. Singer, M., W. A. Walter, B. M. Cali, P. Rouviere, H. H. Liebke, R. L. Gourse, and C. A. Gross. 1991. Physiological effects of the fructose-1,6-diphosphate aldolase ts8 mutation on stable RNA synthesis in Escherichia coli. J. Bacteriol. 173:6249–6257.
424. Singer, P., P. Rossmeisse, B. M. Cali, H. Liebke, and C. A. Gross. 1991. The Escherichia coli ts8 mutation is an allele of fda, the gene encoding fructose-1,6-diphosphate aldolase. J. Bacteriol. 173:6242–6248.
425. Smith, R. J., and J. E. M. Midgley. 1973. The effect of trimethoprim on macromolecular synthesis in Escherichia coli: general effects on RNA and protein synthesis. Biochem. J. 136:225–234.
426. Smith, R. J., and J. E. M. Midgley. 1973. The effect of trimethoprim on macromolecular synthesis in Escherichia coli: regulation of RNA synthesis by "magic spot" nucleotides. Biochem. J. 136:249–257.
427. Smolin, D. E., and H. E. Umbarger. 1975. Specificity of the stimulation of in vitro ribonucleic acid synthesis by guanosine 5'-diphosphate, 3'-diphosphate. Mol. Gen. Genet. 141:277–284.
428. Sokawa, J., and Y. Kaziro. 1969. Amino acid-dependent control of the transport of α-methyl glucoside in E. coli. Biochem. Biophys. Res. Commun. 34:99–103.
429. Sokawa, J., Y. Sokawa, and Y. Kaziro. 1971. Stringent control in Escherichia coli. Nature (London) New Biol. 240:242–245.
430. Somerville, C. R., and A. Ahmed. 1977. rel-dependent methionine requirement in revertants of a methionyl-transfer RNA synthetase mutant of Escherichia coli. J. Mol. Biol. 111:77–81.
431. Somerville, C. R., and A. Ahmed. 1979. Mutants of Escherichia coli defective in the degradation of guanosine 5'-triphosphate, 3'-diphosphate (pppGpp). Mol. Gen. Genet. 169:315–323.
432. Sorensen, M. A., K. F. Jensen, and S. Pedersen. 1994. High concentrations of ppGpp decrease the RNA chain growth rate. Implications for protein synthesis and translational fidelity during amino acid starvation in Escherichia coli. J. Mol. Biol. 236:441–454.
433. Sorensen, M. A., U. Vogel, K. F. Jensen, and S. Pedersen. 1993. The rates of macromolecular chain elongation modulate the initiation frequencies for transcription and translation in Escherichia coli. Antonie van Leeuwenhoek 63:323–336.
434. Spadaro, A., A. Spena, V. Santonastaso, and P. Donini. 1981. Stringency without ppGpp accumulation. Nature (London) 291:256–258.
435. Spassky, A., S. Rimsky, H. Garreau, and H. Buc. 1984. Hla, an E. coli DNA-binding protein which accumulates in stationary phase, strongly compacts DNA in vitro. Nucleic Acids Res. 12:5321–5340.
436. Spector, M. P., and C. L. Cubitt. 1992. Starvation-inducible loci of Salmonella typhimurium: regulation and roles in starvation-survival. Mol. Microbiol. 6:1467–1476.
437. Spratt, B. G. 1977. The mechanism of action of mecillinam. J. Antimicrob. Chemother. 3(Suppl. B):13–19.
438. Spratt, B. G., and A. B. Pardee. 1975. Penicillin-binding protein and cell shape in E. coli. Nature (London) 254:515–517.
439. Squires, C. L., J. Greenblatt, J. Li, C. Condon, and C. L. Squires. 1993. Ribosomal RNA antitermination in vitro: requirement for Nus factors and one or more unidentified cellular components. Proc. Natl. Acad. Sci. USA 90:970–974.
440. Stamminger, G., and R. A. Lazzarini. 1974. Altered metabolism of the guanosine tetraphosphate, ppGpp, in mutants of E. coli. Cell 1:85–90.
441. Stayton, M. M., and H. J. Fromm. 1979. ppGpp inhibition of adenylosuccinate synthetase. J. Biol. Chem. 254:2579–2581.
442. Stein, J. P., and K. E. Bloch. 1976. Inhibition of E. coli β-hydroxydecanoyl thioester dehydrase by ppGpp. Biochem. Biophys. Res. Commun. 73:881–884.
443. Stent, G. S., and S. Brenner. 1961. A genetic locus for the regulation of ribonucleic acid synthesis. Proc. Natl. Acad. Sci. USA 47:2005–2014.
444. Stephens, J. C., S. W. Artz, and B. N. Ames. 1975. Guanosine 5'-diphosphate 3'-diphosphate (ppGpp): positive effector for histidine operon transcription and general signal for amino-acid deficiency. Proc. Natl. Acad. Sci. USA 72:4389–4393.
445. Sumegi, J., A. Udvardy, and P. Venetianer. 1977. In vitro transcription of the ribosomal RNA genes of E. coli DNA. Mol. Gen. Genet. 151:305–312.
446. Svitil, A. L., M. Cashel, and J. W. Zyskind. 1993. Guanosine tetraphosphate inhibits protein synthesis in vivo. A possible protective mechanism for starvation stress in Escherichia coli. J. Biol. Chem. 268:2307–2311.
447. Sy, J. 1974. Reversibility of the pyrophosphoryl transfer from ATP to GTP by Escherichia coli stringent factor. Proc. Natl. Acad. Sci. USA 71:3470–3473.
448. Sy, J. 1977. In vitro degradation of guanosine 5'-diphosphate, 3'-diphosphate. Proc. Natl. Acad. Sci. USA 74:5529–5533.
449. Sy, J. 1980. Activation of ppGpp-3'pyrophosphohydrolase by a supernatant factor and ATP. J. Biol. Chem. 255:10056–10059.
450. Sy, J., and F. Lipmann. 1973. Identification of the synthesis of guanosine tetraphosphate (MSI) as insertion of a pyrophosphoryl group into the 3'-position in guanosine 5'-diphosphate. Proc. Natl. Acad. Sci. USA 70:306–309.
451. Sy, J., Y. Ogawa, and F. Lipmann. 1973. Nonribosomal synthesis of guanosine 5',3'-polyphosphates by the ribosomal wash of stringent Escherichia coli. Proc. Natl. Acad. Sci. USA 70:2145–2148.
452. Taguchi, M., K. Izui, and H. Katsuki. 1978. Stringent control of glycolysis in Escherichia coli. Biochem. Biophys. Res. Commun. 84:195–201.
453. Taguchi, M., K. Izui, and H. Katsuki. 1979. Activation of E. coli phosphoenolpyruvate carboxylase by ppGpp. FEBS Lett. 77:270–272.
454. Takayanagi, Y., K. Tanaka, and H. Takahashi. 1994. Structure of the 5' upstream region and the regulation of the rpoS gene of Escherichia coli. Mol. Gen. Genet. 243:525–531.
455. Tedin, K., and H. Bremer. 1992. Toxic effects of high levels of ppGpp in Escherichia coli are relieved by rpoB mutations. J. Biol. Chem. 267:2337–2344.
456. Tetu, C., E. Dassa, and P. L. Boquet. 1980. The energy dependent degradation of guanosine 5'-diphosphate 3'diphosphate in Escherichia coli. Lack of correlation with ATP levels in vivo and the role of the transmembrane proton gradient. Eur. J. Biochem. 103:117–124.
457. Thiam, K., and A. Favre. 1984. Role of the stringent response in the expression and mechanism of near-ultraviolet induced growth delay. Eur. J. Biochem. 145:137–142.
458. Thomas, G., and A. Favre. 1975. 4-Thiouridine as the target for near-ultraviolet light induced growth delay in Escherichia coli. Biochem. Biophys. Res. Commun. 66:1454–1461.
459. Toone, W. M., K. E. Rudd, and J. D. Friesen. 1991. deaD, a new Escherichia coli gene encoding a presumed ATP-dependent RNA helicase, can suppress a mutation in rpsB, the gene encoding ribosomal protein S2. J. Bacteriol. 173:3291–3302.
460. Toone, W. M., K. E. Rudd, and J. D. Friesen. 1992. Mutations causing aminotriazole resistance and temperature sensitivity reside in gyrB, which encodes the β subunit of DNA gyrase. J. Bacteriol. 174:5479–5481.
461. Torok, I., and C. Kari. 1980. Accumulation of ppGpp in a relA mutant of Escherichia coli during amino acid starvation. J. Biol. Chem. 255:3838–3840.
462. Tosa, T., and L. I. Pizer. 1971. Biochemical basis for the antimetabolite action of l-serine hydroxamate. J. Bacteriol. 106:972–982.
463. Travers, A. 1973. Control of ribosomal RNA synthesis in vitro. Nature (London) 244:15–18.
464. Travers, A. 1976. Modulation of RNA polymerase specificity by ppGpp. Mol. Gen. Genet. 147:225–232.
465. Travers, A. 1980. Promoter sequence for stringent control of bacterial ribonucleic acid synthesis. J. Bacteriol. 41:973–976.
466. Travers, A. 1984. Conserved features of coordinately regulated E. coli promoters. Nucleic Acids Res. 12:2605–2618.
467. Travers, A., and F. E. Baralle. 1976. In vitro transcription of E. coli ribosomal RNA, p. 241–251. In N. O. Kjeldgaard and O. Maaloe (ed.), Control of Ribosome Synthesis. Alfred Benzon Symposium IX. Munksgaard, Copenhagen.
468. Travers, A. A., P. G. Debenham, and O. Pongs. 1980. Translation initiation factor 2 alters E. coli ribonucleic acid polymerase holoenzyme. Biochemistry 19:1651–1656.
469. Travers, A. A., R. Kamen, and M. Cashel. 1970. The in vitro synthesis of RNA. Cold Spring Harbor Symp. Quant. Biol. 35:415–418.
470. Tuomanen, E. 1986. Phenotpic tolerance: the search for beta-lactam antibiotics that kill nongrowing bacteria. Rev. Infect. Dis. 8(Suppl. 3):S279–S291.
471. Tuomanen, E., and R. Cozens. 1987. Changes in peptidoglycan composition and penicillin-binding proteins in slowly growing Escherichia coli. J. Bacteriol. 169:5308–5310.
472. Tuomanen, E., Z. Markiewicz, and A. Tomasz. 1988. Autolysis-resistant peptidoglycan of anomalous composition in amino-acid-starved Escherichia coli. J. Bacteriol. 170:1373–1376.
473. Tuomanen, E., and A. Tomasz. 1986. Induction of autolysis in nongrowing Escherichia coli. J. Bacteriol. 167:1077–1080.
474. Turnbough, C. L., R. J. Neill, R. Landsberg, and B. N. Ames. 1979. Pseudouridylation of tRNAs and its role in regulation in Salmonella typhimurium. J. Biol. Chem. 254:5111–5119.
475. Uzan, M., and A. Danchin. 1976. A rapid test for the relA mutation in E. coli. Biochem. Biophys. Res. Commun. 69:751–758.
476. Uzan, M., and A. Danchin. 1978. Correlation between the serine sensitivity and the derepressibility of the ilv genes in Escherichia coli relA mutants. Mol. Gen. Genet. 165:21–30.
477. VanBogelen, R. A., P. M. Kelley, and F. C. Neidhardt. 1987. Differential induction of heat shock, SOS, and oxidation-stress regulons and accumulation of nucleotides in Escherichia coli. J. Bacteriol. 313:22–72.
478. VanBogelen, R. A., and F. C. Neidhardt. 1990. Ribosomes as sensors of heat and cold shock in Escherichia coli. Proc. Natl. Acad. Sci. USA 87:5589–5593.
479. Vanderwel, D., and E. E. Ishiguro. 1984. Properties of cell wall peptidoglycan synthesized by amino acid deprived relA mutants of Escherichia coli. Can. J. Microbiol. 30:1239–1246.
480. Van Ooyen, A. J. J., H. A. deBoer, G. Ab, and M. Gruber. 1975. Specific inhibition of ribosomal RNA synthesis in vitro by guanosine 3'-diphosphate 5'-diphosphate. Nature (London) 254:530–531.
481. Van Ooyen, A. J. J., M. Gruber, and P. Jorgenson. 1976. The mechanism of action of ppGpp on rRNA synthesis in vitro. Cell 8:123–128.
482. Van Veen, H. W., T. Abee, G. J. J. Korstee, W. N. Konings, and A. J. B. Zehnder. 1994. Translocation of metal phosphate via the phosphate inorganic transport system of Escherichia coli. Biochemistry 33:1766–1770.
483. Venetianer, P. 1968. Preferential synthesis of the messenger RNA of the histidine operon during histidine starvation. Biochem. Biophys. Res. Commun. 33:959–963.
484. Venetianer, P. 1969. Level of messenger RNA transcribed from the histidine operon in repressed, derepressed, and histidine-starved Salmonella typhimurium. J. Mol. Biol. 45:375–384.
485. Vinella, D., and R. D’Ari. 1994. Thermoinducible filamentation in Escherichia coli due to an altered RNA polymerase beta subunit is suppressed by high levels of ppGpp. J. Bacteriol. 176:966–972.
486. Vinella, D., R. D’Ari, A. Jaffe, and P. Bouloc. 1992. Penicillin-binding protein 2 is dispensable in Escherichia coli when ppGpp synthesis is induced. EMBO J. 11:1493–1501.
487. Vinella, D., D. Joseleau-Petit, D.Thevenet, P. Bouloc, and R. D’Ari. 1993. Penicillin-binding protein 2 inactivation in Escherichia coli results in cell division inhibition, which is relieved by FtsZ overexpression. J. Bacteriol. 175:6704–6710.
488. Vogel, U., and K. F. Jensen. 1994. The RNA chain elongation rate in Escherichia coli depends on the growth rate. J. Bacteriol. 176:2807–2813.
489. Vogel, U., and K. F. Jensen. 1994. Effects of guanosine 3',5'-diphosphate (ppGpp) on rate of transcription elongation in isoleucine-starved Escherichia coli. J. Biol. Chem. 269:16236–16241.
490. Vogel, U., S. Pedersen, and J. F. Jensen. 1991. An unusual correlation between ppGpp pool size and rate of ribosome synthesis during partial pyrimidine starvation of Escherichia coli. J. Bacteriol. 173:1168–1174.
491. Vogel, U., M. Sorensen, S. Pedersen, K. F. Jensen, and M. Kilstrup. 1992. Decreasing transcription elongation rates in Escherichia coli exposed to amino acid starvation. Mol. Microbiol. 6:2191–2200.
492. von Meyenburg, K., E. Boye, K. Skarstad, L. Koppes, and T. Kogoma. 1987. Mode of initiation of constitutive stable DNA replication in RNase H-defective mutants of Escherichia coli K-12. J. Bacteriol. 169:2650–1658.
493. Wachi, M., M. Doi, S. Tamaki, W. Park, S. Nakajima-Iijima, and M. Matsuhashi. 1987. Mutant isolation and molecular cloning of mre genes, which determine cell shape, sensitivity to mecillinam, and amount of penicillin-binding proteins in Escherichia coli. J. Bacteriol. 169:4935–4940.
494. Wagner, R., G. Theissen, and M. Zacharias. 1993. Regulation of ribosomal RNA synthesis and control of ribosome formation in E. coli, p. 119–130. In K. H. Nierhaus, F. Franceshi, A. R. Subramanian, V. A. Erdmann, and B. Wittmann-Liebold (ed.), The Translational Apparatus. Plenum Press, New York.
495. Wahab, S. Z., K. O. Rowley, and W. M. Holmes. 1993. Effects of tRNA(1Leu) overproduction in Escherichia coli. Mol. Microbiol. 7:253–263.
496. Watson, R., and H. Yamazaki. 1972. Expression of the rel gene during R17 phage infection. Biochemistry 11:611–614.
497. Wegrzyn, G., F. Neubauer, S. Kreuger, M. Hecker, and K. Taylor. 1991. Stringent control of replication of plasmids derived from coliphage lambda. Mol. Gen. Genet. 225:94–98.
498. Weyer, W. J., H. A. de Boer, J. G. de Boer, and M. Gruber. 1976. The sequence of ppGpp and pppGpp in a reaction scheme for magic spot synthesis. Biochim. Biophys. Acta 442:123–127.
499. Whitby, M. C., L. Ryder, and R.G. Lloyd. 1993. Reverse branch migration of Holliday junctions by RecG protein: a new mechanism for resolution of intermediates in recombination and DNA repair. Cell 75:341–350.
500. Williams, M. D., T. X. Ouyang, and M. C. Flickinger. 1994. Starvation-induced expression of SspA and SspB: the effects of a null mutation in sspA on Escherichia coli protein synthesis and survival during growth and prolonged starvation. Mol. Microbiol. 11:1029–1043.
501. Williams, M. G., and P. Rogers. 1987. Expression of arg genes in Escherichia coli during arginine limitation dependent upon stringent control of translation. J. Bacteriol. 169:1644–1650.
502. Winkler, M. E., D. J. Roth, and P. E. Hartman. 1978. Promoter- and attenuator-related metabolic regulation of the Salmonella typhimurium histidine operon. J. Bacteriol. 133:830–843.
503. Winkler, M. E., R. V. Zawodny, and P. E. Hartman. 1979. Mutation spoT of Escherichia coli increases expression of the histidine operon deleted for the attenuator. J. Bacteriol. 139:993–1000.
504. Winslow, R. M. 1971. A consequence of the rel gene during a glucose to lactate downshift in Escherichia coli. J. Biol. Chem. 246:4872–4877.
505. Winslow, R. M., and R. A. Lazzarini. 1969. The rates of synthesis and chain elongation of ribonucleic acid in Escherichia coli. J. Biol. Chem. 244:1128–1137.
506. Wirth, R., V. Kohles, and A. Bock. 1981. Factors modulating transcription and translation in vitro of ribosomal protein S20 and isoleucyl-tRNA synthetase from Escherichia coli. Eur. J. Biochem. 114:429–437.
507. Wold, S., K. Skarstad, H. B. Steen, T. Stokke, and E. Boye. 1994. The initiation mass for DNA replication in Escherichia coli K-12 is dependent on growth rate. EMBO J. 13:2097–2102.
508. Wong, J. T.-F., and R. N. Nazar. 1970. Relationship of the MS nucleotides to the regulation of ribonucleic acid synthesis in Escherichia coli. J. Biol. Chem. 245:4591–4593.
509. Woody, A. Y., R. W. Woody, and A. D. Malcolm. 1987. Effects of ppGpp on transcription by DNA-dependent RNA polymerase from Escherichia coli: circular dichroism, absorption and specific transcription studies. Biochim. Biophys. Acta 909:115–125.
510. Xiao, H., M. Kalman, K. Ikehara, S. Zemel, G. Glaser, and M. Cashel. 1991. Residual guanosine 3',5'-bispyrophosphate synthetic activity of relA null mutants can be eliminated by spoT null mutations. J. Biol. Chem. 266:5980–5990.
511. Yamagishi, M., J. R. Cole, M. Nomura, F. W. Studier, and J. J. Dunn. 1987. Stringent control in Escherichia coli applies also to transcription by T7 RNA polymerase. J. Biol. Chem. 262:3940–3943.
512. Yamagishi, M., H. Matsushima, A. Wada, M. Sakagami, N. Fujita, and A. Ishihama. 1993. Regulation of the Escherichia coli rmf gene encoding the ribosome modulation factor: growth phase- and growth rate-independent control. EMBO J. 12:625–630.
513. Yang, H., K. Heller, M. Gellert, and G. Zubay. 1979. Differential sensitivity of gene expression in vitro to inhibition of DNA gyrase. Proc. Natl. Acad. Sci. USA 76:3304–3308.
514. Yaniv, M., A. Chestier, F. Gros, and A. Favre. 1971. Biological activity of irradiated tRNAVal containing a 4-thiouridine-cytosine dimer. J. Mol. Biol. 58:381–388.
515. 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.
516. Yarus, M. 1979. Relaxation of stable RNA synthesis by a plasmid borne locus. Mol. Gen. Genet. 170:309–317.
517. Yegian, C. D., and G. S. Stent. 1969. An unusual condition of leucine transfer RNA appearing during leucine starvation of Escherichia coli. J. Mol. Biol. 39:45–58.
518. Yegian, C. D., and G. S. Stent. 1969. Differential aminoacylation of three species of isoleucine transfer RNA from Escherichia coli. J. Mol. Biol. 39:59–71.
519. Yi, Q.-M., S. Rockenbach, J. E. J. Ward, and J. Lutkenhaus. 1985. Structure and expression of the cell division genes ftsQ, ftsA and ftsZ. J. Mol. Biol. 184:399–412.
520. Zacharias, M., H. U. Goringer, and R. Wagner. 1989. Influence of the GCGC discriminator motif introduced into the ribosomal RNA p2- and tac promoter on growth-rate control and stringent sensitivity. EMBO J. 8:3357–3413.
521. Zacharias, M., H. U. Goringer, and R. Wagner. 1992. Analysis of the Fis-dependent and Fis-independent transcription activation mechanisms of the Escherichia coli ribosomal RNA P1 promoter. Biochemistry 31:2621–2628.
522. Zacharias, M., G. Thiessen, C. Bradaczek, and R. Wagner. 1991. Analysis of sequence elements important for the synthesis and control of ribosomal RNA in E. coli. Biochimie 73:699–712.
523. Zhou, Y. N., W. A. Walter, and C. A. Gross. 1992. A mutant of σ32 with a small deletion in conserved region 3 of σ has reduced affinity for core RNA polymerase. J. Bacteriol. 174:5005–5012.
524. Zidwick, M. J., J. Korshus, and P. Rogers. 1984. Positive control of expression of the argECBH gene cluster in vitro by guanosine 5'-diphosphate 3'-diphosphate. J. Bacteriol. 159:647–651.
525. Zyskind, J. W., and D. W. Smith. 1992. DNA replication, the bacterial cell cycle, and cell growth. Cell 69:5–8.
Chapter92
