Cold-temperature induction of Escherichia coli polynucleotide phosphorylase occurs by reversal of its autoregulation

When Escherichia coli cells are shifted to low temperatures (e.g. 15 degrees C), growth halts while the 'cold shock response' (CSR) genes are induced, after which growth resumes. One CSR gene, pnp, encodes polynucleotide phosphorylase (PNPase), a 3'-exoribonuclease and component of the RNA degradosome. At 37 degrees C, ribonuclease III (RNase III, encoded by rnc) cleaves the pnp untranslated leader, whereupon PNPase represses its own translation by an unknown mechanism. Here, we show that PNPase cold-temperature induction involves several post-transcriptional events, all of which require the intact pnp mRNA leader. The bulk of induction results from reversal of autoregulation at a step subsequent to RNase III cleavage of the pnp leader. We also found that pnp translation occurs throughout cold-temperature adaptation, whereas lacZ(+) translation was delayed. This difference is striking, as both mRNAs are greatly stabilized upon the shift to 15 degrees C. However, unlike the lacZ(+) mRNA, which remains stable during adaptation, pnp mRNA decay accelerates. Together with other evidence, these results suggest that mRNA is generally stabilized upon a shift to cold temperatures, but that a CSR mRNA-specific decay process is initiated during adaptation.

The widely conserved Era G-protein contains an RNA-binding domain required for Era function in vivo

Era is a small G-protein widely conserved in eubacteria and eukaryotes. Although essential for bacterial growth and implicated in diverse cellular processes, its actual function remains unclear. Several lines of evidence suggest that Era may be involved in some aspect of RNA biology. The GTPase domain contains features in common with all G-proteins and is required for Era function in vivo. The C-terminal domain (EraCTD) bears scant similarity to proteins outside the Era subfamily. On the basis of sequence comparisons, we argue that the EraCTD is similar to, but distinct from, the KH RNA-binding domain. Although both contain the consensus VIGxxGxxI RNA-binding motif, the protein folds are probably different. We show that bacterial Era binds RNA in vitro and can form higher-order RNA-protein complexes. Mutations in the VIGxxGxxI motif and other conserved residues of the Escherichia coli EraCTD decrease RNA binding in vitro and have corresponding effects on Era function in vivo, including previously described effects on cell division and chromosome partitioning. Importantly, mutations in L-66, located in the predicted switch II region of the E. coli Era GTPase domain, also perturb binding, leading us to propose that the GTPase domain regulates RNA binding in response to unknown cellular cues. The possible biological significance of Era RNA binding is discussed.

Escherichia coli RNase III (rnc) autoregulation occurs independently of rnc gene translation

Control of mRNA stability is an established means of regulating gene expression. However, the detailed mechanisms by which such control is achieved are only now emerging. In particular, there remains a question about the involvement of translation. Escherichia coli ribonuclease III (RNase III) negatively autoregulates expression of its own gene (rnc) approximately 10-fold, by cleaving the untranslated leader and initiating approximately 10-fold more rapid decay of the rnc mRNA, after which RNase III plays no further role. Here, we define the mechanism of this control further. Mutations that increase rnc gene translation abolish autoregulation by increasing the stability of the RNase III-cleaved transcript RNA approximately 10-fold, with no effect on the uncleaved species. Mutations that decrease translation destabilize the rnc mRNA in the presence or absence of RNase III. In so doing, they reveal a pathway of rnc transcript decay distinct from the RNase III-dependent pathway. Stability of a 'mini-rnc' transcript containing the rnc leader and only the first two codons of the rnc gene is unaffected by decreased translation, presumably because sequences required for this pathway were removed. Importantly, this mini-rnc transcript is regulated normally by RNase III. Moreover, rnc transcripts synthesized in vitro do not decay in cell-free extracts lacking ribosomes, unless they are first cleaved by RNase III. These two results show that RNase III cleavage can initiate rnc transcript decay independently of rnc gene translation, unambiguously establishing that control of mRNA decay need not involve changes in translation. How rnc gene translation is optimized for efficient autoregulation will also be discussed.

Molecular characterisation of the pifC gene encoding translation initiation factor 3, which is required for normal photosynthetic complex formation in Rhodobacter sphaeroides NCIB 8253

In order to determine whether translation initiation events play a selective role in regulating the expression of photosynthetic complexes in the photosynthetic bacterium Rhodobacter sphaeroides, we have undertaken an initial study to investigate the potential role of translation initiation factor IF3, which also behaves as a pleiotropic regulatory factor in some bacteria. Following the isolation and purification of a 24-kDa IF3-like protein (PifC) from R. sphaeroides, we used nested PCR to clone and characterise the encoding gene, pifC (photosynthesis-affecting initiation factor). The 545-bp pifC encodes a protein exhibiting 60% identity (78.6% similarity) with the Escherichia coli IF3 (InfC) protein and, in common with all other IF3 genes identified to date, pifC possesses a rare initiation codon (AUA). Furthermore, in common with IF3, PifC was shown here to perform a discriminatory function towards CUG start codons, confirming its role and function as an IF3 in R. sphaeroides. Insertion of a kanamycin resistance cassette into the 5' end of pifC resulted in a viable phenotype which exhibits growth rates similar to wild type but which possesses reduced bacteriochlorophyll and photosynthetic complexes in semi-aerobic cultures. It is shown here that the mutant is still able to produce a PifC protein but that it possesses reduced IF3 activity. This may account for the viable nature of the mutant strain, and may indicate that the effect of the mutation on photosynthesis can be more severe than shown in the present study. The mechanisms by which PifC may exert its selective regulatory effect on photosynthesis expression are discussed.

Regulation of the "tetCD" genes of transposon Tn10

In addition to the genes involved in tetracycline resistance, the loop region of the composite transposon Tn10 contains two other known genes, tetC and tetD, whose functions are unclear. Using primarily a genetic approach, we examined tetCD gene expression and regulation. The tetC gene product, TetC, is a diffusible repressor of both tetC and tetD transcription. Despite an earlier claim by others, we do not detect induction of either tetC or tetD by tetracycline (Tc) or several of its analogs. Although the 5' ends of the tetC and tetD messages overlap due to transcription from convergent promoters, we find no evidence for anti-sense RNA control. The operator for the TetC repressor has been localized. We also demonstrate that transcription from the tetD promoter probably terminates within IS10-Right and does not apparently interfere with Tn10 or IS10-Right transposition or its regulation.

Expression and regulation of the rnc and pdxJ operons of Escherichia coli

Escherichia coli rnc-era-recO operon (rnc operon) expression is negatively autoregulated at the level of message stability by ribonuclease III (RNase III), which is encoded by the rnc gene. RNase III, a double-stranded RNA-specific endoribonuclease involved in rRNA and mRNA processing and degradation, cleaves a stemloop structure in the 5' untranslated leader, initiating rapid decay of the rnc operon mRNA. Here, we examine rnc operon expression and regulation in greater detail. Northern, primer extension, and lacZ fusion analyses show that a single promoter (rncP) specifies two principal mRNAs: the 1.9 kb rnc-era transcript and the less-abundant 3.7 kb RNA encoding rnc-era-recO and the downstream pdxJ and acpS genes. A 1.3 kb pdxJ-acpS RNA is transcribed from a promoter (pdxP) located within recO. About 70% of pdxJ transcription depends on transcription from rncP. Both promoters were characterized genetically. RNase III reduces 1.9 kb and 3.7 kb transcript levels and stability, and corresponding effects are seen with genetic fusions. These detailed studies enabled us to show that the first 378 nucleotides of the rnc transcript comprise a portable RNA stability element (rncO) that contains all of the cis-acting elements required for RNase III-initiated decay of the rnc mRNA as well as the heterologous lacZ transcript. Moreover, mutations in rncO that block RNase III cleavage also block control, showing that RNase III initiates mRNA decay by cleaving at a single site.

RNase III autoregulation: structure and function of rncO, the posttranscriptional "operator"

Expression of the Escherichia coli rnc-era-recO operon is regulated posttranscriptionally by ribonuclease III (RNase III), encoded in the rnc gene. RNase III initiates rapid decay of the rnc operon mRNA by cleaving a double-stranded region of the rnc leader. This region, termed rncO, is portable, conferring stability and RNase III regulation to heterologous RNAs. Here, we report the detailed analysis of rncO structure and function. The first 215 nt of the rnc leader are sufficient for its function. Dimethylsulfate (DMS) modification in vivo revealed distinct structural elements in this region: a 13-nt single-stranded 5' leader, followed by a 6-bp stem-loop structure (I), a larger stem-loop structure (II) containing the RNase III site, a single-stranded region containing the rnc translation initiation site, and a small stem-loop structure (III) at the 3' terminus of rncO, wholly within the rnc coding region. Genetic analysis revealed the function of these structural elements. The single-stranded leader is not required for stability or RNase III control, stem-loop II is required only for RNase III control, and both stem-loops I and III are required for stability. Stem-loop II effectively serves only as the site at which RNase III cleaves to remove stem-loop I and thereby initiates decay, after which RNase III plays no role. Mutations at the cleavage site underscore the importance of base pairing for efficient RNase III attack. When stem-loops I and II were replaced with an artificial hairpin structure, stability was restored only partially, but was restored almost fully when a single-stranded leader was also added.

Escherichia coli translation initiation factor 3 discriminates the initiation codon in vivo

In a genetic selection designed to isolate Escherichia coli mutations that increase expression of the IS 10 transposase gene (tnp), we unexpectedly obtained viable mutants defective in translation initiation factor 3 (IF3). Several lines of evidence led us to conclude that transposase expression, per se, was not increased. Rather, these mutations appear to increase expression of the tnp'-'lacZ gene fusions used in this screen, by increasing translation initiation at downstream, atypical initiation codons. To test this hypothesis we undertook a systematic analysis of start codon requirements and measured the effects of IF3 mutations on initiation from various start codons. Beginning with an efficient translation initiation site, we varied the AUG start codon to all possible codons that differed from AUG by one nucleotide. These potential start codons fall into distinct classes with regard to translation efficiency in vivo: Class I codons (AUG, GUG, and UUG) support efficient translation; Class IIA codons (CUG, AUU, AUC, AUA, and ACG) support translation at levels only 1-3% that of AUG; and Class IIB codons (AGG and AAG) permit levels of translation too low for reliable quantification, importantly, the IF3 mutations had no effect on translation from Class I codons, but they increased translation from Class II codons 3-5-fold, and this same effect was seen in other gene contexts. Therefore, IF3 is generally able to discriminate between efficient and inefficient codons in vivo, consistent with earlier in vitro observations. We discuss these observations as they relate to IF3 autoregulation and the mechanism of IF3 function.

Structure and regulation of the Salmonella typhimurium rnc-era-recO operon

The Escherichia coli rnc-era-recO operon encodes ribonuclease III (RNase III; a dsRNA endonuclease involved in rRNA and mRNA processing and decay), Era (an essential G-protein of unknown functions and RecO (involved in the RecF homologous recombination pathway). Expression of the rnc and era genes is negatively autoregulated: RNase III cleaves the rncO 'operator' in the untranslated leader, destabilizing the operon mRNA. As part of a larger effort to understand RNase III and Era structure and function, we characterized rnc operon structure, function and regulation in the closely related bacterium Salmonella typhimurium. Construction of a S typhimurium strain conditionally defective for RNase III and Era expression showed that Era is essential for cell growth. This mutant strain also enabled selection of recombinant clones containing the intact S typhimurium rnc-era-recO operon, whose nucleotide sequence, predicted protein sequence, and predicted rncO RNA secondary structure were all highly conserved with those of E coli. Furthermore, genetic and biochemical analysis revealed that S typhimurium rnc gene expression is negatively autoregulated by a mechanism very similar or identical to that in E coli, and that the cleavage specificities of RNase IIIs.t. and RNase IIIE.c. are indistinguishable with regard to rncO cleavage and S typhimurium 23S rRNA fragmentation in vivo.

Control of translation by mRNA secondary structure: the importance of the kinetics of structure formation

RNA secondary structure is important in a wide variety of biological processes, but relatively little is known about the pathways and kinetics of RNA folding. When the IS10 transposase (tnp) gene is transcribed from a promoter outside the element, little increase in tnp expression is observed. This protection from outside transcription (pot) occurs at the translational level, presumably resulting from mRNA secondary structure proposed to sequester the tnp ribosome-binding site. Here, we confirm the pot RNA structure and show that it blocks 30S ribosomal subunit binding in vitro. Point mutations that abolish protection in vivo map to the pot structure. Surprisingly, these pot mutations do not severely alter the pot secondary structure or increase 30S subunit binding in vitro, except in one case. Using an oligonucleotide hybridization assay, we show that most of the pot mutations slow the kinetics of pot structure formation, with little or no effect on the inhibitory function of the final structure. Moreover, a suppressor mutation reverses this effect. We propose a pathway for pot mRNA folding that is consistent with the mutations and implicates the formation of important kinetic intermediates. The significance of these observations for the RNA folding problem in general is discussed.

Antisense RNA control in bacteria, phages, and plasmids

Antisense RNA control is now recognized as an efficient and specific means of regulating gene expression at the posttranscriptional level. Almost all naturally occurring cases have been found in prokaryotes, often in their accessory genetic elements. Several antisense RNA systems are now well-understood, and these display a spectrum of mechanisms of action, binding pathways, and kinetics. This review summarizes antisense RNA control in prokaryotes, emphasizing the biology of the systems involved.

Decay of the IS10 antisense RNA by 3' exoribonucleases: evidence that RNase II stabilizes RNA-OUT against PNPase attack

RNA-OUT, the 69-nucleotide antisense RNA that regulates Tn10/IS10 transposition folds into a simple stem-loop structure. The unusually high metabolic stability of RNA-OUT is dependent, in part, on the integrity of its stem-domain: mutations that disrupt stem-domain structure (Class II mutations) render RNA-OUT unstable, and restoration of structure restores stability. Indeed, there is a strong correlation between the thermodynamic and metabolic stabilities of RNA-OUT. We show here that stem-domain integrity determines RNA-OUT's resistance to 3' exoribonucleolytic attack: Class II mutations are almost completely suppressed in Escherichia coli cells lacking its principal 3' exoribonucleases, ribonuclease II (RNase II) and polynucleotide phosphorylase (PNPase). RNase II and PNPase are individually able to degrade various RNA-OUT species, albeit with different efficiencies: RNA-OUT secondary structure provides greater resistance to RNase II than to PNPase. Surprisingly, RNA-OUT is threefold more stable in wild-type cells than in cells deficient for RNase II activity, suggesting that RNase II somehow lessens PNPase attack on RNA-OUT. We discuss how this might occur. We also show that wild-type RNA-OUT stability changes only two-fold across the normal range of physiological growth temperatures (30-44 degrees C) in wild-type cells, which has important implications for IS10 biology.

Chromosomal supercoiling in Escherichia coli

The Escherichia coli chromosome is compacted into 40-50 negatively supercoiled domains. It has been proposed that these domains differ in superhelical density. Here, we present evidence that this is probably not the case. A modified Tn10 transposable element was inserted at a number of locations around the E. coli chromosome. This element, mTn10-plac-lacZ+, contains the lac operon promoter, plac, whose activity increases with increasing superhelical density, fused to a lacZ+ reporter gene. Although mTn10-plac-lacZ+ fusion expression varies as much as approximately threefold at different insertion sites, the relative levels of expression from these elements are unaffected by replacing plac with the gyrA promoter, pgyrA, which has a reciprocal response to changes in superhelical density. Importantly, topoisomerase mutations and coumermycin, which inhibits DNA gyrase activity, alter mTn10-plac-lacZ+ and mTn10-pgyrA-lacZ+ fusion expression in expected ways, showing that the elements remain responsive to supercoiling and that topoisomerase activity is required for maintaining superhelical density. Fusion expression is not affected by anaerobic growth or osmotic shock, two physiological conditions thought to alter supercoiling. The approximately threefold difference in mTn10-plac-lacZ+ and mTn10-pgyrA-lacZ+ fusion expression observed at different sites may be explained by regional differences in chromosomal copy number that arise from bidirectional replication. Together, these results strongly suggest that the E. coli chromosomal domains do not differ in functional superhelical density.

DNA from diverse sources manifests cryptic low-level transcription in Escherichia coli

We present evidence that DNA from diverse prokaryotic and eukaryotic sources gives rise to low-level fusion expression in Escherichia coli promoter-probe vectors. This expression may be as high as approximately 10% of the E. coli lacUV5 promoter. Although expression does not correlate with the presence of obvious E. coli promoter-like sequences, it is blocked by transcriptional terminators. Furthermore, transcription across the fusion junction is detected at levels that correlate with fusion expression. We suggest that this 'low-level transcription' (LLT) results from infrequent initiation by RNA polymerase at random sites and/or weak promoters. We propose that LLT has biological significance. In some instances, it may provide an advantageous basal level of gene expression, and we suggest that this may be true for the E. coli lacY gene. In other instances, LLT may be detrimental, in which case it may be blocked by mechanisms such as RNA secondary structure or transcriptional polarity. We present evidence to show that activation of the IS10 transposase gene by LLT is blocked at the translational level.

Vectors for constructing kan gene fusions: direct selection of mutations affecting IS10 gene expression

We describe several vectors for constructing translational fusions to the kan gene of Tn5. Fusions are constructed in vitro using multi-copy vectors containing unique cloning sites situated between upstream transcriptional terminators and a downstream kan gene lacking transcriptional and translational start signals. Multi-copy fusions can be converted to single-copy chromosomal fusions by in vivo recombination with specific phage lambda vectors and vice versa. We find that kan fusions are often more suitable than lacZ fusions for the direct selection of mutations that increase fusion expression. These vectors were developed for isolating mutations that increase IS10 transposase expression; we describe strategies used to isolate such mutations, which map to IS10 or the Escherichia coli himA, himD(hip), dam or infC genes.

The IS10 antisense RNA blocks ribosome binding at the transposase translation initiation site

Transposase (tnp) expression from insertion sequence IS10 is controlled, in part, by an antisense RNA, RNA-OUT, which pairs to the translation initiation region of the tnp mRNA, RNA-IN. Genetic experiments suggest that control occurs post-transcriptionally. Here, we present evidence that bears on the control mechanism. Specific ribosome binding at the tnp translation initiation site is demonstrated in vitro. Two mutations that alter tnp translation in vivo are shown to have corresponding effects in vitro. Most importantly, RNA-OUT/RNA-IN pairing is shown to block ribosome binding. In conjunction with the work described in the accompanying paper, we propose that inhibition of ribosome binding also occurs in vivo, and that it is sufficient to account for control. Implications for translational control in analogous systems are discussed.

The IS10 transposase mRNA is destabilized during antisense RNA control

RNA stability is an important component of gene expression, and antisense RNAs have been proposed to alter target RNA stability. We show here that the IS10 transposase mRNA, RNA-IN, is rendered unstable during control by the IS10 antisense RNA, RNA-OUT. Destabilization requires RNA-OUT/RNA-IN pairing and ribonuclease III cleavage. Independent of such cleavage, RNA-OUT is rendered unstable through disruption of its secondary structure. Pairing has no other obvious effects on RNA-IN transcription or stability. Nevertheless, RNA-IN destabilization is not required for antisense control in vivo. In the accompanying paper [Ma,C. and Simons, R.W. (1990) EMBO J., 9, 1267-1274 we show that pairing blocks ribosome binding to RNA-IN. Were it not for control at this level, destabilization would play a more prominent role.

The unusual stability of the IS10 anti-sense RNA is critical for its function and is determined by the structure of its stem-domain

IS10 transposition is regulated by an approximately 70 nt anti-sense RNA, RNA-OUT. RNA-OUT folds into a duplex 'stem-domain' topped by a loosely paired 'loop-domain'. The loop-domain is critical for RNA-RNA pairing per se; pairing initiates by interaction of the RNA-OUT loop with the 5' end of the target mRNA. We show here that RNA-OUT is unusually stable in vivo (half-life 60 min) and that this stability is conferred by specific features of the RNA-OUT stem-domain. One critical feature is stable base-pairing: mutations that disrupt stem pairing destabilize RNA-OUT in vivo and abolish anti-sense control; combinations of mutations that restore pairing also restore both stability and control. We propose that the stem renders RNA-OUT resistant to 3' exoribonucleases. Other features of the stem-domain prevent this essential duplex from being an effective substrate for double-strand nucleases: two single base mutations disrupt antisense control by making RNA-OUT susceptible to RNase III. Mutations in the loop region have little effect on RNA-OUT stability. Implications for IS10 biology and the design of efficient anti-sense RNAs are discussed.

Insertion sequence IS10 anti-sense pairing initiates by an interaction between the 5' end of the target RNA and a loop in the anti-sense RNA

Transposition of insertion sequence IS10 is regulated by an anti-sense RNA which inhibits transposase expression when IS10 is present in multiple copies per cell. The anti-sense RNA (RNA-OUT) consists of a stem domain topped by a flexibly paired loop; the 5' end of the target molecule, RNA-IN, is complementary to the top of the loop, and complementarity extends for 35 base-pairs down one side of RNA-OUT. We present here genetic evidence that anti-sense pairing, both in vitro and in vivo, initiates by interaction of the 5' end of RNA-IN and the loop domain of RNA-OUT; other features of the reaction are discussed. In the context of this model, we discuss features of this anti-sense system which are important for its biological effectiveness, and suggest that IS10 provides a convenient model for design of efficient artificial anti-sense RNA molecules.

Analysis of the promoters and transcripts involved in IS10 anti-sense RNA control

Genetic analysis of eleven mutations affecting the IS10 promoters, pIN and pOUT, involved in anti-sense RNA control of transposase gene expression, and characterization of the transcripts, reveal that: (i) The transposase message (RNA-IN) and the anti-sense RNA (RNA-OUT) have been unambiguously identified in vivo. (ii) Five mutations affect pIN activity, and establish that pIN is the only IS10 promoter transcribing the tnp gene, and the only such IS10 promoter that responds to DNA-adenine methylation. (iii) Six mutations alter pOUT activity, and establish that pOUT is the only IS10 promoter specifying the anti-sense RNA-OUT. (iv) The latter, however, need not be so: heterologous promoters, if properly positioned, can also specify active anti-sense RNAs. (v) These heterologously promoted anti-sense RNAs are processed to species closely resembling native RNA-OUT.

Naturally occurring antisense RNA control--a brief review

Biological control by naturally occurring anti-sense RNAs has been documented in a number of prokaryotic cases, and strongly suggested in several eukaryotic systems. The biological activities controlled are diverse, including transposition, phage development, chromosomal gene expression, and plasmid replication, compatibility and conjugation. Control is exerted at many different levels, by both direct and long-range effects. The stem/loop structures common to all anti-sense RNAs are important functional domains: loops are the sites of critical interactions in the initiation of pairing to the target RNA; stems determine anti-sense RNA stability in vivo. These features need to be considered in the design of artificial anti-sense RNA control. Details of RNA/RNA pairing have emerged; pairing initiates at single-stranded regions in anti-sense RNA loops, and stable complex formation involves the nearby end of one or both molecules.

Regulation of fatty acid degradation in Escherichia coli: fadR superrepressor mutants are unable to utilize fatty acids as the sole carbon source

Localized mutagenesis of the fadR region of the Escherichia coli chromosome resulted in the isolation of two classes of fadR regulatory mutants. The first class was constitutive for the fatty acid degradative enzymes and presumably defective for fadR function. The second class was rarer and resulted in the inability to utilize fatty acids as a sole carbon source (Fad-). These fadR superrepressor mutants [fadR(S)] had greatly reduced levels of the beta-oxidative enzymes required for growth on fatty acids. The fadR(S) mutants reverted to Fad+ at a high frequency (10(-5], and the resulting Fad+ revertants were constitutive for expression of the fad enzymes (fadR). Merodiploid analysis showed the fadR(S) allele to be dominant to both fadR+ and fadR alleles.

Improved single and multicopy lac-based cloning vectors for protein and operon fusions

We describe several new vectors for the construction of operon and protein fusions to the Escherichia coli lacZ gene. In vitro constructions utilize multicopy plasmids containing suitable cloning sites located between upstream transcription terminators and downstream lac operon segments whose lacZ genes retain or lack translational start signals. Single-copy lambda prophage versions of multicopy constructs can be made genetically, without in vitro manipulation. The new vectors, both single and multicopy, are improved in that they have very low levels of background lac gene expression, which makes possible the easy detection and accurate quantitation of very weak transcriptional and translational signals. These vectors were developed for analysis of the expression of IS10's transposase gene, which is transcribed less than, once per generation, and whose transcripts are translated on average less than once each. Both single and multicopy constructs can also be used to select mutations affecting fusion expression, and mutations isolated in single-copy constructs can be crossed genetically back onto multicopy plasmids for further analysis.

Tn10 protects itself at two levels from fortuitous activation by external promoters

Tn10 rarely transposes, primarily because its IS10-encoded transposase protein is synthesized infrequently. Since the 5' end of the transposase gene is immediately adjacent to flanking host sequences, insertion of Tn10 into an actively transcribed operon could conceivably result in dramatically increased transposition. We show here that Tn10 is protected from such fortuitous activation; high levels of transcription from an upstream promoter actually decrease its rate of transposition. Protection operates at two levels. First, externally-initiated transcripts yield only a small amount of additional transposase protein, primarily because of inhibition at a posttranscriptional level. We suggest that the transposase gene start codon is sequestered in an mRNA secondary structure not present in transcripts initiated at the normal promoter. Second, transcription per se across an IS10 terminus inhibits its activity, thus negating any small transposase increase. These observations provide additional evidence that Tn10 has evolved specific mechanisms for keeping its transposition activity low.

Translational control of IS10 transposition

We present genetic evidence that insertion sequence IS10, the active element in transposon Tn10, can negatively control expression of its own transposase protein at the translational level. This control process is manifested in trans in a phenomenon called "multicopy inhibition": the presence of a multicopy plasmid containing IS10 inhibits transposition of a single copy chromosomal Tn10 element by reducing its ability to express transposition functions. Fusion analysis suggests that expression is reduced at the translational and not the transcriptional level. Only the outer 180 bp of IS10-Right are required on the plasmid for full inhibition. Plasmid-encoded transposase protein is not involved. The genetic structure of the essential plasmid region and the effects of point and deletion mutations on multicopy inhibition lead us to propose that inhibition of transposase translation occurs by direct pairing between the transposase messenger RNA and a small, complementary, regulatory RNA specified by the IS10-encoded pOUT promoter.

Three promoters near the termini of IS10: pIN, pOUT, and pIII

We have identified three IS10-encoded promoters, pIN, the promoter for IS10's transposase gene, is intrinsically weak, contributing to the low frequency of IS10 transposition in vivo. Its transcripts begin near the "outside" end of IS10 and extend inward across the element. pOUT, a strong promoter just internal to and opposing pIN, directs transcription outward. Its transcripts are proposed to inhibit translation of the transposase gene in trans (accompanying paper). pOUT may also inhibit transcription from pIN in cis. pIII, a weak promoter near the "inside" end of IS10, is of unknown genetic importance. Many transposable elements activate, by adjacent insertion, silent genes lacking normal promoters. Such IS10-promoted turn-on is mediated by pOUT and results from continuation of pOUT-initiated transcripts past the IS10 terminus, into adjoining chromosomal material. Wild-type and mutant IS10 promoters have been analyzed in vitro. pIN is weaker than pOUT because of inefficient isomerization from closed to open complexes. Despite their proximity, pIN and pOUT do not interact before or during open complex formation.

DNA sequence organization of IS10-right of Tn10 and comparison with IS10-left

Tn10 is 9,300 base pairs long and has inverted repeats of an insertion sequence (IS)-like sequence (IS10) at its ends. IS10-right provides all of the Tn10-encoded functions used for normal Tn10 transposition. IS10-left can also provide these functions but at a much reduced level. We report here the complete nucleotide sequence of IS10-right and a partial sequence of IS10-left. From our analysis of this information, we draw the following conclusions. (i) IS10-right is 1,329 base pairs long. Like most IS elements, it has short (23-base pair) nearly perfect inverted repeats at its termini. We can divide these 23-base pair segments into at least two functionally distinct parts. IS10-right also shares with other elements the presence of a single long coding region that extends the entire length of the element. Genetic evidence suggests that this coding region specifies an essential IS10 transposition function. A second, overlapping, coding region may or may not be important. (ii) The "outside" end of IS10-right contains three suggestively positioned internal symmetries. Two of these (A1 and A2) are nearly identical in sequence. Symmetry A1 overlaps the terminal inverted repeat; symmetry A2 overlaps the promoter shown elsewhere to be responsible for expression of IS10 functions and lies very near a second characterized promoter that directs transcription outward across the end of IS10. Symmetries A1 and A2 may play a role in modulation of Tn10 activity and are likely to function at least in part as protein recognition sites. We propose that the third symmetry (B) acts to prevent fortuitous expression of IS10 functions from external promoters. The transcripts from such promoters can assume a stable secondary structure in which the AUG start codon of the long coding region is sequestered in a region of double-stranded mRNA formed by pairing between the two halves of symmetry B. (iii) IS10-left differs from IS10-right at many nucleotide positions in both the presumptive regulatory region and the long coding region. The available evidence suggests that Tn10 may be older than other analyzed drug-resistance transposons and thus have had more time to accumulate mutational changes.

Transport of long and medium chain fatty acids by Escherichia coli K12

Kinetic, metabolic, and physical parameters of long and medium chain fatty acid transport by Escherichia coli K12 were determined. Uptake of long chain fatty acids (C11-C18:1) mediated by the fadL gene involves concentrative transport. Evidence for this is as follows: (i) characteristic Ki and Vmax values were obtained for long chain fatty acids, (ii) long chain fatty acid transport was inhibited by energy inhibitors, (iii) long chain fatty acids were concentrated 10-fold inside the cell against a concentration gradient, (iv) efflux of transported long chain fatty acids did not occur, and (v) an energy of activation of 11.72 kcal mol-1 and Q10 of 2.3 were obtained for long chain fatty acid transport. The fadL gene product shows some activity with medium chain fatty acids (C7-C10) as well. Medium chain fatty acids also appear to enter the cell by simple diffusion since: (i) medium chain fatty acid transport by fadL strains is not saturable under our assay conditions, (ii) fadL strains do not concentrate medium chain fatty acids against a concentration gradient, and (iii) medium chain fatty acids are available for efflux in fadL strains. Physical parameters of long and medium chain fatty acid transport are also reported. These results present evidence for separate mechanisms of long and medium chain fatty acid transport in E. coli.

Regulation of fatty acid degradation in Escherichia coli: dominance studies with strains merodiploid in gene fadR

Strains stably merodiploid in the 25-min region of the chromosome of Escherichia coli were constructed and used in dominance tests between various wild-type and mutant alleles of the fadR gene. Whereas the monoploid fadR+ and fadR strains were inducible and constitutive, respectively, for the enzymes involved in fatty acid degradation (fad), merodiploids with at least one fadR+ allele were inducible. This observation was true whether the fadR+ allele resided on the main chromosome or on the episome. These results show that fadR+ is trans dominant to fadR, and they are consistent with the proposal that the fadR gene product is a repressor protein. Complementation tests were also performed by constructing 24 merodiploids harboring fadR alleles on both the main chromosome and the episome. All of these fadR/fadR diploids were able to utilize the noninducing substrate decanoate as sole carbon source, suggesting that only one polypeptide is encoded by the fadR gene.

Regulation of fatty acid degradation in Escherichia coli: isolation and characterization of strains bearing insertion and temperature-sensitive mutations in gene fadR

Transposon Tn10 was used to mutagenize the fadR gene in Escherichia coli. Mutants bearing fadR:Tn10 insertion mutations were found to (i) utilize the noninducing fatty acid decanoate as sole carbon source, (ii) beta-oxidize fatty acids at constitutive rates, and (iii) contain constitutive levels of the five key beta-oxidative enzymes. These characteristics were identical to those observed in spontaneous fadR mutants. The constitutive phenotype presented by the fadR:Tn10 mutants was shown to be genetically linked to the associated transposon-encoded drug resistance. These results suggest that the fadR gene product exerts negative control over the fatty acid degradative regulon. The fadR gene of E. coli has been mapped through the use of transposon-mediated fadR insertion mutations. The fadR locus is at 25.5 min on the revised map and cotransduces with purB, hemA, and trp. Three-factor conjugational and transductional crosses indicate that the order of loci in this region of the chromosome is purB-fadR-hemA-trp. Spontaneous fadR mutants were found to map at the same location. Strains that exhibit alterations in the control of the fad regulon in response to changes in temperature were also isolated and characterized. These fadR(Ts) mutants were constitutive for the fad enzymes at elevated temperatures and inducible for these activities at low temperatures. The fadR(Ts) mutations also map at the fadR locus. These results strongly suggest that the fadR gene product is a repressor protein.

Kinetics of the utilization of medium and long chain fatty acids by mutant of Escherichia coli defective in the fadL gene

Experiments were performed to assess the role of the fadL gene in Escherichia coli. These studies have revealed that this organism requires a functional fadL gene in order to (i) transport optimally the fatty acids C10 to C18:1 into the cell, (ii) optimally grow on and oxidize C10 to C18:1 fatty acids, and (iii) incorporate efficiently C12 to C18:1 fatty acids into its membrane phospholipids. A defect in the fadL gene does not prevent E. coli from optimally utilizing fatty acids with chain lengths less than 10 carbon atoms. These results suggest that the fadL gene governs a transport component(s) which is required for the optimal transport of fatty acids with chain lengths greater than 9 carbon atoms.

Transport of long-chain fatty acids by Escherichia coli: mapping and characterization of mutants in the fadL gene

A new locus (fadL) that is required for the utilization of long-chain fatty acids has been mapped and partially characterized in an Escherichia coli mutant. The fadL locus has been mapped at 50 min on the chromosome. A mutant bearing a defect in this locus cannot utilize long-chain fatty acids as a sole carbon source. Derivatives of this mutant that can grow on decanoate (termed fadR) are capable of growth on medium-chain but not long-chain fatty acids. It is believed that the fadL mutants is defective in the transport of long-chain fatty acids into the cell for the following reasons: (i) fadR fadL strains can oxidize in vivo decanoate but not oleate; (ii) neither fadL nor fadR fadL strains can incorporate oleate into their membrane lipids; (iii) the activity of the acyl-CoA synthetase (EC 6.2.1.3) in fadR fadL strains is comparable to the acyl-CoA synthetase activity in fadR fadL+ strains; and (iv) in vitro extracts from fadR fadL+ strains. If the above hypothesis is correct, the uptake of long-chain fatty acids by E. coli requires at least two gene products.