Evidence that a nucleotide sequence, "boxA," is involved in the action of the NusA protein

iTerm-PseKNC: a sequence-based tool for predicting bacterial transcriptional terminators

MOTIVATION

Transcription termination is an important regulatory step of gene expression. If there is no terminator in gene, transcription could not stop, which will result in abnormal gene expression. Detecting such terminators can determine the operon structure in bacterial organisms and improve genome annotation. Thus, accurate identification of transcriptional terminators is essential and extremely important in the research of transcription regulations.

RESULTS

In this study, we developed a new predictor called 'iTerm-PseKNC' based on support vector machine to identify transcription terminators. The binomial distribution approach was used to pick out the optimal feature subset derived from pseudo k-tuple nucleotide composition (PseKNC). The 5-fold cross-validation test results showed that our proposed method achieved an accuracy of 95%. To further evaluate the generalization ability of 'iTerm-PseKNC', the model was examined on independent datasets which are experimentally confirmed Rho-independent terminators in Escherichia coli and Bacillus subtilis genomes. As a result, all the terminators in E. coli and 87.5% of the terminators in B. subtilis were correctly identified, suggesting that the proposed model could become a powerful tool for bacterial terminator recognition.

AVAILABILITY AND IMPLEMENTATION

For the convenience of most of wet-experimental researchers, the web-server for 'iTerm-PseKNC' was established at http://lin-group.cn/server/iTerm-PseKNC/, by which users can easily obtain their desired result without the need to go through the detailed mathematical equations involved.

RNase E/G-dependent degradation of metE mRNA, encoding methionine synthase, in Corynebacterium glutamicum

Corynebacterium glutamicum is used for the industrial production of various metabolites, including L-glutamic acid and L-lysine. With the aim of understanding the post-transcriptional regulation of amino acid biosynthesis in this bacterium, we investigated the role of RNase E/G in the degradation of mRNAs encoding metabolic enzymes. In this study, we found that the cobalamin-independent methionine synthase MetE was overexpressed in ΔrneG mutant cells grown on various carbon sources. The level of metE mRNA was also approximately 6- to 10-fold higher in the ΔrneG mutant strain than in the wild-type strain. A rifampicin chase experiment showed that the half-life of metE mRNA was approximately 4.2 times longer in the ΔrneG mutant than in the wild-type strain. These results showed that RNase E/G is involved in the degradation of metE mRNA in C. glutamicum.

Nus Factors of Escherichia coli

The highly conserved Nus factors of bacteria were discovered as essential host proteins for the growth of temperate phage λ in Escherichia coli. Later, their essentiality and functions in transcription, translation, and, more recently, in DNA repair have been elucidated. Close involvement of these factors in various gene networks and circuits is also emerging from recent genomic studies. We have described a detailed overview of their biochemistry, structures, and various cellular functions, as well as their interactions with other macromolecules. Towards the end, we have envisaged different uncharted areas of studies with these factors, including their participation in pathogenicity.

Nus Factors of Escherichia coli

The Nus factors-NusA, NusB, NusE, and NusG-area set of well-conserved proteins in bacteria and are involved in transcription elongation, termination, antitermination, and translation processes. Originally, Escherichia coli host mutations defective for supporting bacteriophage λ N-mediated antitermination were mapped to the nusA (nusA1), nusB (nusB5, nusB101), and nusE (nusE71) genes, and hence, these genes were named nus for Nutilization substances (Nus). Subsequently,the Nus factors were purified and their roles in different host functions were elucidated. Except for NusB, deletion of which is conditionally lethal, all the other Nus factors are essential for E. coli. Among the Nus factors, NusA has the most varied functions. It specifically binds to RNA polymerase (RNAP), nascent RNA, and antiterminator proteins like N and Q and hence takes part in modulating transcription elongation, termination, and antitermination. It is also involved in DNA repair pathways. NusG interacts with RNAP and the transcription termination factor Rho and therefore is involved in both factor-dependent termination and transcription elongation processes. NusB and NusE are mostly important in antitermination at the ribosomal operon-transcription. NusE is a component of ribosome and may take part in facilitating the coupling between transcription and translation. This chapter emphasizes the structure-function relationship of these factors and their involvement in different fundamental cellular processes from a mechanistic angle.

Structural basis for RNA recognition by NusB and NusE in the initiation of transcription antitermination

Processive transcription antitermination requires the assembly of the complete antitermination complex, which is initiated by the formation of the ternary NusB-NusE-BoxA RNA complex. We have elucidated the crystal structure of this complex, demonstrating that the BoxA RNA is composed of 8 nt that are recognized by the NusB-NusE heterodimer. Functional biologic and biophysical data support the structural observations and establish the relative significance of key protein-protein and protein-RNA interactions. Further crystallographic investigation of a NusB-NusE-dsRNA complex reveals a heretofore unobserved dsRNA binding site contiguous with the BoxA binding site. We propose that the observed dsRNA represents BoxB RNA, as both single-stranded BoxA and double-stranded BoxB components are present in the classical lambda antitermination site. Combining these data with known interactions amongst antitermination factors suggests a specific model for the assembly of the complete antitermination complex.

RNA-binding specificity of E. coli NusA

The RNA sequences boxA, boxB and boxC constitute the nut regions of phage λ. They nucleate the formation of a termination-resistant RNA polymerase complex on the λ chromosome. The complex includes E. coli proteins NusA, NusB, NusG and NusE, and the λ N protein. A complex that includes the Nus proteins and other factors forms at the rrn leader. Whereas RNA-binding by NusB and NusE has been described in quantitative terms, the interaction of NusA with these RNA sequences is less defined. Isotropic as well as anisotropic fluorescence equilibrium titrations show that NusA binds only the nut spacer sequence between boxA and boxB. Thus, nutR boxA5-spacer, nutR boxA16-spacer and nutR boxA69-spacer retain NusA binding, whereas a spacer mutation eliminates complex formation. The affinity of NusA for nutL is 50% higher than for nutR. In contrast, rrn boxA, which includes an additional U residue, binds NusA in the absence of spacer. The K_d values obtained for rrn boxA and rrn boxA-spacer are 19-fold and 8-fold lower, respectively, than those for nutR boxA-spacer. These differences may explain why λ requires an additional protein, λ N, to suppress termination. Knowledge of the different affinities now describes the assembly of the anti-termination complex in quantitative terms.

Evolutionary comparison of ribosomal operon antitermination function

Transcription antitermination in the ribosomal operons of Escherichia coli results in the modification of RNA polymerase by specific proteins, altering its basic properties. For such alterations to occur, signal sequences in rrn operons are required as well as individual interacting proteins. In this study we tested putative rrn transcription antitermination-inducing sequences from five different bacteria for their abilities to function in E. coli. We further examined their response to the lack of one known rrn transcription antitermination protein from E. coli, NusB. We monitored antitermination activity by assessing the ability of RNA polymerase to read through a factor-dependent terminator. We found that, in general, the closer the regulatory sequence matched that of E. coli, the more likely there was to be a successful antitermination-proficient modification of the transcription complex. The rrn leader sequences from Pseudomonas aeruginosa, Bacillus subtilis, and Caulobacter crescentus all provided various levels of, but functionally significant antitermination properties to, RNA polymerase, while those of Mycobacterium tuberculosis and Thermotoga maritima did not. Possible RNA folding structures of presumed antitermination sequences and specific critical bases are discussed in light of our results. An unexpected finding was that when using the Caulobacter crescentus rrn leader sequence, there was little effect on terminator readthrough in the absence of NusB. All other hybrid antitermination system activities required this factor. Possible reasons for this finding are discussed.

Conservation and diversity in the immunity regions of wild phages with the immunity specificity of phage lambda

The gene regulatory circuitry of phage lambda is among the best-understood circuits. Much of the circuitry centres around the immunity region, which includes genes for two repressors, CI and Cro, and their cis-acting sites. Related phages, termed lambdoid phages, have different immunity regions, but similar regulatory circuitry and genome organization to that of lambda, and show a mosaic organization, arising by recombination between lambdoid phages. We sequenced the immunity regions of several wild phages with the immunity specificity of lambda, both to determine whether natural variation exists in regulation, and to analyse conservation and variability in a region rich in well-studied regulatory elements. CI, Cro and their cis-acting sites are almost identical to those in lambda, implying that regulatory mechanisms controlled by the immunity region are conserved. A segment adjacent to one of the operator regions is also conserved, and may be a novel regulatory element. In most isolates, different alleles of two regulatory proteins (N and CII) flank the immunity region; possibly the lysis-lysogeny decision is more variable among isolates. Extensive mosaicism was observed for several elements flanking the immunity region. Very short sequence elements or microhomologies were also identified. Our findings suggest mechanisms by which fine-scale mosaicism arises.

Rho-dependent terminators and transcription termination

Rho-dependent transcription terminators participate in sophisticated genetic regulatory mechanisms, in both bacteria and phages; they occur in regulatory regions preceding the coding sequences of genes and within coding sequences, as well as at the end of transcriptional units, to prevent readthrough transcription. Most Rho-dependent terminators have been found in enteric bacteria, but they also occur in Gram-positive bacteria and may be widespread among bacteria. Rho-dependent termination requires both cis-acting elements, on the mRNA, and trans-acting factors. The only cis-acting element common to Rho-dependent terminators is richness in rC residues. Additional sequence elements have been observed at different Rho termination sites. These 'auxiliary elements' may assist in the termination process; they differ among terminators, their occurrence possibly depending on the function and sequence context of the terminator. Specific nucleotides required for termination have also been identified at Rho sites. Rho is the main factor required for termination; it is a ring-shaped hexameric protein with ATPase and helicase activities. NusG, NusA and NusB are additional factors participating in the termination process. Rho-dependent termination occurs by binding of Rho to ribosome-free mRNA, C-rich sites being good candidates for binding. Rho's ATPase is activated by Rho-mRNA binding, and provides the energy for Rho translocation along the mRNA; translocation requires sliding of the message into the central hole of the hexamer. When a polymerase pause site is encountered, the actual termination occurs, and the transcript is released by Rho's helicase activity. Many aspects of this process are still being studied. The isolation of mutants suppressing termination, site-directed mutagenesis of cis-acting elements in Rho-dependent termination, and biochemistry, are and will be contributing to unravelling the still undefined aspects of the Rho termination machinery. Analysis of the more sophisticated regulatory mechanisms relying on Rho-dependent termination may be crucial in identifying new essential elements for termination.

Transcription termination and anti-termination in E. coli

Transcription termination in Escherichia coli is controlled by many factors. The sequence of the DNA template, the structure of the transcript, and the actions of auxiliary proteins all play a role in determining the efficiency of the process. Termination is regulated and can be enhanced or suppressed by host and phage proteins. This complex reaction is rapidly yielding to biochemical and structural analysis of the interacting factors. Below we review and attempt to unify into basic principles the remarkable recent progress in understanding transcription termination and anti-termination.

Analysis of tryptophanase operon expression in vitro: accumulation of TnaC-peptidyl-tRNA in a release factor 2-depleted S-30 extract prevents Rho factor action, simulating induction

Expression of the tryptophanase (tna) operon in Escherichia coli is regulated by catabolite repression and tryptophan-induced transcription antitermination. The key feature of this antitermination mechanism has been shown to be the retention of uncleaved TnaC-peptidyl-tRNA in the translating ribosome. This ribosome remains stalled at the tna stop codon and blocks the access of Rho factor to the tna transcript, thereby preventing transcription termination. In normal S-30 preparations, synthesis of a TnaC peptide containing arginine instead of tryptophan at position 12 (Arg(12)-TnaC) was shown to be insensitive to added tryptophan, i.e. Arg(12)-TnaC-peptidyl-tRNA was cleaved, and there was normal Rho-dependent transcription termination. When the S-30 extract used was depleted of release factor 2, Arg(12)-TnaC-tRNA(Pro) was accumulated in the absence or presence of added tryptophan. Under these conditions the accumulation of Arg(12)-TnaC-tRNA(Pro) prevented Rho-dependent transcription termination, mimicking normal induction. Using a minimal in vitro transcription system consisting of a tna template, RNA polymerase, and Rho, it was shown that RNA sequences immediately adjacent to the tnaC stop codon, the presumed boxA and rut sites, contributed most significantly to Rho-dependent termination. The tna boxA-like sequence appeared to serve as a segment of the Rho "entry" site, despite its likeness to the boxA element.

Control of intrinsic transcription termination by N and NusA: the basic mechanisms

Intrinsic transcription termination plays a crucial role in regulating gene expression in prokaryotes. After a short pause, the termination signal appears in RNA as a hairpin that destabilizes the elongation complex (EC). We demonstrate that negative and positive termination factors control the efficiency of termination primarily through a direct modulation of hairpin folding and, to a much lesser extent, by changing pausing at the point of termination. The mechanism controlling hairpin formation at the termination point relies on weak protein interactions with single-stranded RNA, which corresponds to the upstream portion of the hairpin. Escherichia coli NusA protein destabilizes these interactions and thus promotes hairpin folding and termination. Stabilization of these contacts by phage lambda N protein leads to antitermination.

Interactions of an Arg-rich region of transcription elongation protein NusA with NUT RNA: implications for the order of assembly of the lambda N antitermination complex in vivo

The E. coli NusA transcription elongation protein (NusA(Ec)), identified because of its requirement for transcription antitermination by the N protein, has an Arg-rich S1 RNA-binding domain. A complex of N and NusA with other host factors binding at NUT sites in the RNA renders RNA polymerase termination-resistant. An E. coli haploid for nusA944, having nine different codons replacing four normally found in the Arg-rich region, is defective in support of N action. Another variant, haploid for the nusAR199A allele, with a change in a highly conserved Arg codon in the S1 domain, effectively supports N-mediated antitermination. However, nusAR199A is recessive to nusA944, while nusA(Ec) is dominant to nusA944 for support of N-mediated antitermination, suggesting a competition between NusA944 and NusAR199A during complex formation. Complex formation with the variant NusA proteins was assessed by mobility gel shifts. NusAR199A, unlike NusA(Ec) and NusA944, fails to form a complex with N and NUT RNA. However, while NusAR199A, like wild-type NusA, forms an enlarged complex with NUT RNA, N, RNA polymerase, and other host proteins required for efficient N-mediated antitermination, NusA944 does not form this enlarged complex. Consistent with the in vivo results, NusA944 prevents NusAR199A but not NusA(Ec) from forming the enlarged complex. The simplest conclusion from these dominance studies is that in the formation of the complete active antitermination complex in vivo, NusA and N binding to the newly synthesized NUT RNA precedes addition of the other factors. Alternative less effective routes to the active complex that allows bypass of this preferred pathway may also exist.

An extended RNA binding surface through arrayed S1 and KH domains in transcription factor NusA

The crystal structure of Thermotoga maritima NusA, a transcription factor involved in pausing, termination, and antitermination processes, reveals a four-domain, rod-shaped molecule. An N-terminal alpha/beta portion, a five-stranded beta-barrel (S1 domain), and two K-homology (KH) modules create a continuous spine of positive electrostatic potential, suitable for nonspecific mRNA attraction. Homology models suggest how, in addition, specific mRNA regulatory sequences can be recognized by the S1 and KH motifs. An arrangement of multiple S1 and KH domains mediated by highly conserved residues is seen, creating an extended RNA binding surface, a paradigm for other proteins with similar domain arrays. Structural and mutational analyses indicate that the motifs cooperate, modulating strength and specificity of RNA binding.

The alpha subunit of E. coli RNA polymerase activates RNA binding by NusA

The Escherichia coli NusA protein modulates pausing, termination, and antitermination by associating with the transcribing RNA polymerase core enzyme. NusA can be covalently cross-linked to nascent RNA within a transcription complex, but does not bind RNA on its own. We have found that deletion of the 79 carboxy-terminal amino acids of the 495-amino-acid NusA protein allows NusA to bind RNA in gel mobility shift assays. The carboxy-terminal domain (CTD) of the alpha subunit of RNA polymerase, as well as the bacteriophage lambda N gene antiterminator protein, bind to carboxy-terminal regions of NusA and enable full-length NusA to bind RNA. Binding of NusA to RNA in the presence of alpha or N involves an amino-terminal S1 homology region that is otherwise inactive in full-length NusA. The interaction of the alpha-CTD with full-length NusA stimulates termination. N may prevent termination by inducing NusA to interact with N utilization (nut) site RNA rather than RNA near the 3' end of the nascent transcript. Sequence analysis showed that the alpha-CTD contains a modified helix-hairpin-helix motif (HhH), which is also conserved in the carboxy-terminal regions of some eubacterial NusA proteins. These HhH motifs may mediate protein-protein interactions in NusA and the alpha-CTD.

Rho-dependent transcription termination in the tna operon of Escherichia coli: roles of the boxA sequence and the rut site

Expression of the tryptophanase (tna) operon of Escherichia coli is regulated by catabolite repression and by tryptophan-induced transcription antitermination. Tryptophan induction prevents Rho-dependent transcription termination in the leader region of the operon. Induction requires translation of a 24-residue leader peptide-coding region, tnaC, containing a single, crucial Trp codon. Studies with a lacZ reporter construct lacking the tnaC-tnaA spacer region suggest that, in the presence of excess tryptophan, the TnaC leader peptide acts in cis on the ribosome translating tnaC to inhibit its release. The stalled ribosome is thought to block Rho's access to the transcript. In this paper we examine the roles of the boxA sequence and the rut site in Rho-dependent termination. Deleting six nucleotides (CGC CCT) of boxA or introducing specific point mutations in boxA results in high-level constitutive expression. Some constitutive changes introduced in boxA do not change the TnaC peptide sequence. We confirm that deletion of the rut site results in constitutive expression. We also demonstrate that, in each constitutive construct, replacement of the tnaC start codon by a UAG stop codon reduces expression significantly, suggesting that constitutive expression requires translation of the tnaC coding sequence. Addition of bicyclomycin, an inhibitor of Rho, to these UAG constructs increases expression, demonstrating that reduced expression is due to Rho action. Combining a boxA point mutation with rut site deletion results in constitutive expression comparable to that of a maximally induced operon. These results support the hypothesis that in the presence of tryptophan the ribosome translating tnaC blocks Rho's access to the boxA and rut sites, thereby preventing transcription termination.

Functional importance of regions in Escherichia coli elongation factor NusA that interact with RNA polymerase, the bacteriophage lambda N protein and RNA

The association of the essential Escherichia coli protein NusA with RNA polymerase increases pausing and the efficiency of termination at intrinsic terminators. NusA is also part of the phage lambda N protein-modified antitermination complex that functions to prevent transcriptional termination. We have investigated the structure of NusA using various deletion fragments of NusA in a variety of in vitro assays. Sequence and structural alignments have suggested that NusA has both S1 and KH homology regions that are thought to bind RNA. We show here that the portion of NusA containing the S1 and KH homology regions is important for NusA to enhance both termination and antitermination. There are two RNA polymerase-binding regions in NusA, one in the amino-terminal 137 amino acids and the other in the carboxy-terminal 264 amino acids; only the amino-terminal RNA polymerase-binding region provides a functional contact that enhances termination at an intrinsic terminator or antitermination by N. The carboxy-terminal region of NusA is also required for interaction with N and is important for the formation of an N-NusA-nut site or N-NusA-RNA polymerase-nut site complex; the instability of complexes lacking this carboxy-terminal region of NusA that binds N and RNA polymerase can be compensated for by the presence of the additional E. coli elongation factors, NusB, NusG and ribosomal protein S10.

rut Sites in the nascent transcript mediate Rho-dependent transcription termination in vivo

The in vitro function of the coliphage lambda tR1 Rho-dependent terminator is governed primarily by a tripartite upstream sequence element designated rut. To determine the contribution of the different components of the rut site to terminator function in the normal context of coupled translation of the nascent cro message, tR1 variants lacking different rut site sequences were tested for terminator function in vivo. Intact rutA and rutB sequences were both necessary for efficient termination. However, deletion of the upstream rutA was far more detrimental than deletion of rutB. The intervening boxB, which encodes a short RNA stem and loop, could be deleted without reducing termination or detectably altering Rho's interaction with the corresponding cro transcript. The relative importance of these sequence elements was also the same in a minimal in vitro termination assay system. Rut sequences are therefore essential for terminator function in vivo and rutA contributes substantially more to tR1 function than does rutB. The relative contribution of these elements can be ascribed to differences in Rho's binding affinity for the encoded transcripts. If other cellular factors also bind the rut element RNA, they do not alter the relative contribution of its two regions to Rho-dependent transcription termination in vivo.

RNA-mediated signaling in transcription

Structures of phage transcriptional antitermination complexes define novel motifs for recognition of RNA hairpins by arginine-rich peptides. A bent alpha-helix in each case follows the contour of an induced GNRA-like fold. A phage-specific pattern of base pairing, base stacking and base flipping underlies biological specificity and permits engagement with RNA polymerase. The structures suggest a mechanism of RNA-mediated signaling in transcriptional regulation.

Assembly of the N-dependent antitermination complex of phage lambda: NusA and RNA bind independently to different unfolded domains of the N protein

The N protein of bacteriophage lambda activates expression of the delayed early genes of this phage by modifying RNA polymerase (RNAP) into a form that is resistant to termination signals. N binds to the boxB hairpin that forms in the nascent RNA transcript upon transcription of the nut regulatory element, and then interacts with RNAP by RNA looping. The binding of the N-boxB subassembly to the transcription complex is further stabilized by interaction with the Escherichia coli NusA protein. N, free in solution, exists as an unfolded protein that becomes partially structured upon binding specifically to boxB RNA. Because NusA does not assist in antitermination unless N is specifically bound to boxB, we have asked whether the structural change induced by binding to boxB affects the interaction of N with NusA. Using fluorescence spectroscopy, we have measured the affinity of N for NusA in the presence and absence of boxB RNA. We find that NusA binds to the unfolded N protein with a dissociation constant (Kd) of approximately 70 nM, and although N undergoes a significant structural change upon binding to boxB, the binding affinity of NusA for a N protein complexed with boxB is not altered. We have also shown that the boxA element of nut does not affect NusA binding to N-boxB. These results demonstrate that the interaction of N with NusA is independent of RNA binding, arguing that NusA must interact with an unfolded region of the polypeptide that remains unstructured even when N binds to boxB RNA. To further establish this point we isolated a truncated peptide containing the amino-terminal 36 residues of the N protein. Binding of boxB RNA to this peptide showed that all of the structural change in N that occurs upon binding to boxB RNA is localized within the amino-terminal 36 residues of N, therefore the C terminus of N, including the regions necessary for NusA binding and RNAP activation, remains unfolded when the full length N binds to boxB RNA. Thus it appears that N can be described as an unfolded multi-domain protein that becomes ordered in a modular fashion as it encounters its various binding partners within the N-dependent antitermination complex.

An RNA enhancer in a phage transcriptional antitermination complex functions as a structural switch

Antitermination protein N regulates the transcriptional program of phage lambda through recognition of RNA enhancer elements. Binding of an arginine-rich peptide to one face of an RNA hairpin organizes the other, which in turn binds to the host antitermination complex. The induced RNA structure mimics a GNRA hairpin, an organizational element of rRNA and ribozymes. The two faces of the RNA, bridged by a sheared GA base pair, exhibit a specific pattern of base stacking and base flipping. This pattern is extended by stacking of an aromatic amino acid side chain with an unpaired adenine at the N-binding surface. Such extended stacking is coupled to induction of a specific internal RNA architecture and is blocked by RNA mutations associated in vivo with loss of transcriptional antitermination activity. Mimicry of a motif of RNA assembly by an RNA-protein complex permits its engagement within the antitermination machinery.

Localization of nusA-suppressing amino acid substitutions in the conserved regions of the beta' subunit of Escherichia coli RNA polymerase

Escherichia coli RNA polymerase is composed of four different subunits, alpha (present in two copies), beta, beta' and sigma. Among these, the beta' polypeptide shares nine conserved regions with the largest subunits of eukaryotic RNA polymerases, but its role is poorly understood. We isolated novel mutations in a plasmid-borne copy of rpoC, which encodes beta', as dominant suppressors of two temperature-sensitive nusA alleles. All 20 suppressors of nusA11 (single missense mutation) isolated had either of two specific substitutions: Lys for Glu-402 (rpoC10) and Thr for Ala-904 (rpoC111) in the beta' subunit. In vivo and in vitro transcription assays revealed that the rpoC10 allele of beta' participates in Rho-dependent transcription termination. On the other hand, of 20 suppressors of nusA134 (deletion of C-terminal one-third) scattered at 18 distinct sites, 16 were assigned to one of six conserved regions C-I. These results suggested that the conserved domains of the beta' subunit of E. coli RNA polymerase are involved in transcript termination or interaction with termination factor(s).

A NusG-like protein from Thermotoga maritima binds to DNA and RNA

The NusG-like protein from Thermotoga maritima was expressed in Escherichia coli and purified to homogeneity. Purified T. maritima NusG exhibited a generalized, non-sequence-specific and highly cooperative DNA and RNA binding activity. The complexes formed between nucleic acid and T. maritima NusG were unable to penetrate a polyacrylamide or agarose gel. The affinity of the protein for DNA was highest in buffers containing about 50 mM salt. The DNA-protein complexes could not be stained with ethidium bromide, were resistant to digestion by TaqI endonuclease, were able to be transcribed in vitro by T. maritima RNA polymerase, and contained a minimum of about 30 to 40 monomers of NusG per kb of duplex DNA. The protein had comparable affinities for duplex DNA and RNA but a lower affinity for single-stranded DNA. Electron microscopy showed that the DNA in the complex is condensed within a large structure that resembles the complex between DNA and histone-like protein Hcl from Chlamydia trachomatis. Neither the wild-type T. maritima nusG gene nor a deletion derivative more similar to the E. coli gene was able to substitute for the essential E. coli nusG. Two variants of the NusG protein were constructed, expressed, and purified: one contains only the entire 171-amino-acid insertion that is unique to T. maritima NusG, and the other has only the sequences present in NusG homologs from E. coli and other eubacteria. Both variants exhibited similar DNA and RNA binding behavior, although their apparent affinities were 5- to 10-fold lower than that of the wild-type T. maritima NusG.

A protein-RNA interaction network facilitates the template-independent cooperative assembly on RNA polymerase of a stable antitermination complex containing the lambda N protein

The stable association of the N gene transcriptional antiterminator protein of bacteriophage lambda with transcribing RNA polymerase requires a nut site (boxA+boxB) in the nascent transcript and the Escherichia coli factors NusA, NusB, NusG, and ribosomal protein S10. We have used electrophoretic mobility shift assays to analyze the assembly of N protein, the E. coli factors, and RNA polymerase onto the nut site RNA in the absence of a DNA template. We show that N binds boxB RNA and that subsequent association of NusA with the N-nut site complex is facilitated by both boxA and boxB. In the presence of N, NusA, and RNA polymerase the nut site assembles ribonucleoprotein complexes containing NusB, NusG, and S10. The effects on assembly of mutations in boxA, boxB, NusA, and RNA polymerase define multiple weak protein-protein and protein-RNA interactions (e.g., NusB with NusG; NusA with boxB; NusA, NusB, and NusG with boxA) that contribute to the overall stability of the complex. Interaction of each component of the complex with two or more other components can explain the many observed cooperative binding associations in the DNA-independent assembly of a stable antitermination complex on RNA polymerase.

Escherichia coli-Salmonella typhimurium hybrid nusA genes: identification of a short motif required for action of the lambda N transcription antitermination protein

The Escherichia coli nusA gene, nusAEc, encodes an essential protein that influences transcription elongation. Derivatives of E. coli in which the Salmonella typhimurium nusA gene, nusASt, has replaced nusAEc are viable. Thus, NusASt can substitute for NusAEc in supporting essential bacterial activities. However, hybrid E. coli strains with the nusASt substitution do not effectively support transcription antitermination mediated by the N gene product of phage lambda. We report the DNA sequence of nusASt, showing that the derived amino acid sequence is 95% identical to the derived amino acid sequence of nusAEc. The alignment of the amino acid sequences reveals scattered single amino acid differences and one region of significant heterogeneity. In this region, called 449, NusAEc has four amino acids and NusASt has nine amino acids. Functional studies of hybrid nusA genes, constructed from nusAEc and nusASt, show that the 449 region of the NusAEc protein is important for lambda N-mediated transcription antitermination. A hybrid that has a substitution of the four E. coli codons for the nine S. typhimurium codons, but is otherwise nusASt, supports the action of the N antitermination protein. The 449 region and, presumably, adjacent sequences appear to compose a functional domain of NusAEc important for the action of the N transcription antitermination protein of phage lambda.

Pleiotropic effects of the rpoC10 mutation affecting the RNA polymerase beta' subunit of Escherichia coli on factor-dependent transcription termination and antitermination

Escherichia coli RNA polymerase is composed of four different subunits, 2 alpha, beta, beta' and sigma. Among these subunits, the role of beta' is poorly understood. The rpoC10 mutation affecting beta' has been isolated as a suppressor mutation of the temperature-sensitive nusA11 mutant. DNA sequence analysis revealed that the rpoC10 mutant is a substitution of Lys for Glu-402. This increased positive charge appears to compensate for the increased negative charge present in the nusA11 protein (Asp for Gly-181). In vivo measurements of reporter gene expression have revealed that rpoC10 restores rho-dependent termination but fails to restore rho-independent termination in nusA11. Moreover, the rpoC10 mutation, in combination with any nusA mutation, inhibited lambda Q-mediated antitermination without affecting N antitermination and severely restricted lambda phage development. The inhibition of Q function and lambda growth could be compensated for by overproducing Q. These results suggest that the RNA polymerase beta' subunit plays a crucial role in factor-dependent transcription termination and antitermination.

Ribosomal RNA antitermination in vitro: requirement for Nus factors and one or more unidentified cellular components

Using an in vitro transcription assay, we have successfully demonstrated read through of a Rho-dependent terminator by the ribosomal RNA antitermination system. The assay used a DNA template containing a promoter-antiterminator-terminator arrangement, RNA polymerase, termination factor Rho, antitermination factors NusA, NusB, NusE, and NusG, and a cellular extract depleted of NusB. Terminator read-through was highly efficient only in the presence of the extract and Nus factors, suggesting that an as yet uncharacterized cellular component is required for ribosomal antitermination. The NusB-depleted extract had no activity in the absence of NusB, confirming an absolute requirement for this protein in ribosomal RNA antitermination. The DNA template requirements were the same as those previously established in vivo; transcription of a wild-type boxA sequence is both necessary and sufficient to promote RNA polymerase modification into a terminator-resistant form.

Recognition of boxA antiterminator RNA by the E. coli antitermination factors NusB and ribosomal protein S10

The boxA sequences of the E. coli ribosomal RNA (rrn) operons are sufficient to cause RNA polymerase to read through Rho-dependent transcriptional terminators. We show that a complex of the transcription antitermination factors NusB and ribosomal protein S10 interacts specifically with boxA RNA. Neither NusB nor S10 binds boxA RNA on its own, and neither NusA nor NusG affects the interaction of the NusB-S10 complex with boxA RNA. Mutations in boxA that impair its antitermination activity compromise its interaction with NusB and S10, suggesting that ribosomal protein S10 regulates the synthesis of ribosomal RNA in bacteria. RNA containing the closely related boxA sequence from the bacteriophage lambda nutR site is not stably bound by NusB and S10. This probably explains why antitermination in phage lambda depends on the phage lambda N protein and the boxB component of the nut site, in addition to boxA.

The kis and kid genes of the parD maintenance system of plasmid R1 form an operon that is autoregulated at the level of transcription by the co-ordinated action of the Kis and Kid proteins

Stability mediated by the parD system of plasmid R1 is modulated by a killer protein, Kid, and by an antagonist of this function, Kis. Determination of the 5' ends of ParD transcripts, revealed that the genes coding for these proteins are transcribed from a single promoter. Analysis of the 3' end of the ParD RNAs indicated the existence of two transcripts: one of them coding for the Kis and Kid proteins, and the other coding only for Kis. Analysis of the effects of parD+ and kis+ recombinants on the beta-galactosidase levels expressed by different transcriptional and translational parD-lacZ fusions, and on the ParD RNA levels determined by a derepressed parD mutant, indicated that the Kis and Kid proteins repress coordinately the parD system at the transcriptional level. We discuss the relevance of these results in terms of the activities of the Kis and Kid proteins and in the context of the stabilization mediated by parD.

RNA-protein interactions in a Nus A-containing Escherichia coli transcription complex paused at an RNA hairpin

We have isolated Escherichia coli transcription complexes, paused in the presence and absence of Nus A, which contain RNA substituted at every UMP residue with a photocrosslinking nucleotide analog. The pause site is immediately downstream from an RNA stem-loop structure, and although pausing occurs in the absence of Nus A, it is substantially enhanced in the presence of Nus A. We have analyzed the secondary structure of this RNA and show that the analog does not interfere with the formation of the normal stem-loop structures. Additionally, the analog substrate does not alter transcriptional pausing, in the presence or absence of Nus A, indicating that Nus A recognition of the transcription complex is not affected by the presence of the crosslinking groups in the RNA. Ribonuclease digestion of the RNA in paused complexes identifies two accessible regions, two nucleotides in the loop and one near the base of the upstream side of the stem-loop. Cleavage at one loop nucleotide is enhanced by Nus A, while the nucleotide near the base of the stem-loop is partially protected. Upon irradiation of the transcription complex, Nus A is not photoaffinity labeled by the RNA, even at a high molar ration to RNA polymerase (250:1). Both the beta and beta' subunits are labeled, however, indicating that the putative stem-loop binding domain on the core polymerase involves both subunits. Because the nucleotide protected from ribonuclease by Nus A is very near two analogs, yet Nus A is not crosslinked to the RNA, it is unlikely that Nus A could be protecting this position through direct contact. Furthermore, analog is substituted at positions in both the loop and at several positions in the stem, and again, no crosslinking to Nus A is observed. We conclude that enhancement of pausing by Nus A probably does not require direct interaction with the bases in the RNA stem-loop.

Genetic interaction between the beta' subunit of RNA polymerase and the arginine-rich domain of Escherichia coli nusA protein

The nusA11 mutation causes reduced transcription termination and temperature-sensitive growth of Escherichia coli. Suppressor mutations that restored growth of nusA11 mutant cells were isolated and named sna mutations. The intergenic suppressor mutation sna-10 was located in the rpoC gene at 90 min, which encodes the beta' subunit of RNA polymerase. sna-10 complemented the defect in tR1 termination caused by nusA11 and by itself stimulated termination of transcription at the lambda tR1 terminator. sna-10 is specific to the nusA11 allele and unable to suppress cold-sensitive growth of the nusA10 mutant. nusA10 carried two base substitutions at positions 311 and 634, causing two amino acid changes from the wild-type sequence. During these studies, we found three -1 frameshift errors in the wild-type nusA sequence; the correct sequence was confirmed by the peptide sequence and gene fusion analyses. The revised sequence revealed that nusA1 and nusA11 are located in an arginine-rich peptide region and substitute arginine and aspartate for leucine 183 and glycine 181, respectively. The intragenic suppressor study indicated that the nusA11 mutation can be suppressed by changing the mutated aspartate 181 to alanine or changing aspartate 84 to tyrosine.

Two-step cloning and expression in Escherichia coli of the DNA restriction-modification system StyLTI of Salmonella typhimurium

The StyLTI restriction-modification system is common to most strains of the genus Salmonella, including Salmonella typhimurium. We report here the two-step cloning of the genes controlling the StyLTI system. The StyLTI methylase gene (mod) was cloned first. Then, the companion endonuclease gene (res) was introduced on a compatible vector. A strain of S. typhimurium sensitive to the coliphage lambda was constructed and used to select self-modifying recombinant phages from a Res- Mod+ S. typhimurium genomic library in the lambda EMBL4 cloning vector. The methylase gene of one of these phages was then subcloned in pBR328 and transferred into Escherichia coli. In the second step, the closely linked endonuclease and methylase genes were cloned together on a single DNA fragment inserted in pACYC184 and introduced into the Mod+ E. coli strain obtained in the first step. Attempts to transform Mod- E. coli or S. typhimurium strains with this Res+ Mod+ plasmid were unsuccessful, whereas transformation of Mod+ strains occurred at a normal frequency. This can be understood if the introduction of the StyLTI genes into naive hosts is lethal because of degradation of host DNA by restriction activity; in contrast to most restriction-modification systems, StyLTI could not be transferred into naive hosts without killing them. In addition, it was found that strains containing only the res gene are viable and lack restriction activity in the absence of the companion mod gene. This suggests that expression of the StyLTI endonuclease activity requires at least one polypeptide involved in the methylation activity, as is the case for types I and III restriction-modification systems but not for type II systems.

Nucleotide sequence of the Escherichia coli recJ chromosomal region and construction of recJ-overexpression plasmids

The nucleotide sequence of the recJ gene of Escherichia coli K-12 and two upstream coding regions was determined. Three regions were identified within these two upstream genes that exhibited weak to moderate promoter activity in fusions to the galK gene and are candidates for the recJ promoter. recJ appeared to be poorly translated: the recJ nucleotide sequence revealed a suboptimal initiation codon GUG, no discernible ribosome-binding consensus sequence, and relatively nonbiased synonymous codon usage. Comparison of the sequence of this region of the chromosome with DNA data bases identified the gene immediately downstream of recJ as prfB, which encodes translational release factor 2 and has been mapped near recJ at 62 min. No significant homology between recJ and other previously sequenced regions of DNA was detected. However, protein sequence comparisons with a gene upstream of recJ, denoted xprB, revealed significant homology with several site-specific recombination proteins. Its genetic function is presently unknown. Knowledge of the nucleotide sequence of recJ allowed the construction of a plasmid from which overexpression of RecJ protein could be induced. Supporting the notion that translation of recJ is limiting, a strong T7 bacteriophage promoter upstream of recJ did not, by itself, allow high-level expression of RecJ protein. The addition of a ribosome-binding sequence fused to the initiator GTG of recJ in this construction was necessary to promote expression of high levels of RecJ protein.

Functional importance of the Escherichia coli ribosomal RNA leader box A sequence for post-transcriptional events

To shed more light on the controversial findings concerning the functional participation of the highly conserved nut-like leader box A sequence element in ribosomal RNA transcription antitermination we have carried out a mutational study. We have substituted the box A and combined this mutation with several deletions comprising the rRNA leader elements box B, box C and the tL region. The mutations are located within the genuine rrnB operon cloned on multicopy plasmids. We determined the effects of the mutations on cell growth, rRNA accumulation and ribosomal subunit stoichiometry. Cells transformed with the mutated plasmids were affected in their growth rate, and showed a surprising deficiency of the promoter-proximal 16S compared to the 23S RNA, indicative of a post-transcriptional degradation event. Accordingly, we could demonstrate a reduced amount of free 30S relative to 50S ribosomal subunits in exponentially growing cells. Similar stoichiometric aberrations in the ribosome pool were detected in conditionally Nus factor-defective strains. The results show that the leader box A sequence within rRNA operons has important post-transcriptional functions for 16S RNA stability and ribosomal subunit stoichiometry. A model is proposed, describing the biogenesis and quality control of ribosomes based on rRNA leader and Nus-factor interactions. It is compatible with the previously observed effects of box A in antitermination.

A site in the T4 bacteriophage major head protein gene that can promote the inhibition of all translation in Escherichia coli

The cryptic DNA element, e14, synthesizes a protein, Lit, which can inhibit gene expression late in T4 bacteriophage development. This inhibition is due to the interaction between the Lit protein and a short region, the gol region, within gene 23, the major head protein gene of phage T4. We have constructed plasmids in which the gol region is transcribed from the lac promoter and fused translationally and transcriptionally to lacZ and cat (chloramphenicol acetyltransferase). These fusion plasmids were used to demonstrate that, in the presence of Lit protein, the gol region inhibits the expression of genes downstream in the same transcription unit. This local inhibition does not require the gene 23 polypeptide from the gol region. In addition, inducing the transcription and translation of the gol region in the presence of Lit protein causes an immediate global inhibition of all translation in Escherichia coli. This global inhibition does require the gene 23 polypeptide. No more than 75 base-pairs of DNA from the gol region are required for both the local and global inhibitions. The gol region sequence contains a short dyad symmetry. However, it is the sequence of bases in the region of dyad symmetry and not the ability to form a hairpin in the RNA that is required for gol region activity.

An analysis of the role of host factors in transcription antitermination in vitro by the Q protein of coliphage lambda

We used two different approaches to study the requirement for Escherichia coli Nus factors for the activity of bacteriophage lambda late antiterminator Q. Using an in vitro coupled transcription-translation assay, based on Q-dependent synthesis of galactokinase from a pR'-tR'-galK template, we showed that mutations in the host nusB and nusE genes do not affect Q activity. A mutation in nusA (nusA1) only partially affects Q action at all temperatures tested. Defective Q function in the nusA1 mutant extract could be restored by the addition of pure NusA but not by excess Q. In a pure transcription system, measurement of the run-off transcript produced by Q-mediated suppression of tR' revealed that NusA is greatly stimulatory to Q activity, whereas NusB and S10, in the presence or absence of NusA, have no effect. Unidentified E. coli factor(s) present in an S30 extract efficiently suppress the natural pausing by RNA polymerase at +15, +16 of pR' without affecting Q activity. These results show that NusA is the only host protein that directly participates in Q function.

Inhibition of transcription of cytosine-containing DNA in vitro by the alc gene product of bacteriophage T4

The alc gene product (gpalc) of bacteriophage T4 inhibits the transcription of cytosine-containing DNA in vivo. We examined its effect on transcription in vitro by comparing RNA polymerase isolated from Escherichia coli infected with either wild-type T4D+ or alc mutants. A 50 to 60% decline in RNA polymerase activity, measured on phage T7 DNA, was observed by 1 min after infection with either T4D+ or alc mutants; this did not occur when the infecting phage lacked gpalt. In the case of the T4D+ strain but not alc mutants, this was followed by a further decrease. By 5 min after infection the activity of alc mutants was 1.5 to 2.5 times greater than that of the wild type on various cytosine-containing DNA templates, whereas there was little or no difference in activity on T4 HMdC-DNA, in agreement with the in vivo specificity. Effects on transcript initiation and elongation were distinguished by using a T7 phage DNA template. Rifampin challenge, end-labeling with [gamma-32P]ATP, and selective initiation with a dinucleotide all indicate that the decreased in vitro activity of the wild-type polymerase relative to that of the alc mutants was due to inhibition of elongation, not to any difference in initiation rates. Wild-type (but not mutated) gpalc copurified with RNA polymerase on heparin agarose but not in subsequent steps. Immunoprecipitation of modified RNA polymerase also indicated that gpalc was not tightly bound to RNA polymerase intracellularly. It thus appears likely that gpalc inhibits transcript elongation on cytosine-containing DNA by interacting with actively transcribing core polymerase as a complex with the enzyme and cytosine-rich stretches of the template.

Lambda nutR mutations convert HK022 Nun protein from a transcription termination factor to a suppressor of termination

The Nun protein of the lambdoid phage HK022 blocks lambda growth by terminating transcription at (or near) the lambda nut sites. An HK022 lysogen carrying a fusion of the lambda pR promoter and nutR site to a gal operon that lacks its own promoter is, therefore, Gal-. To characterize the target of Nun action, spontaneous Gal+ revertants of this strain were isolated and characterized. Two cis-acting mutations are located in the fusion and represent transversions of conserved nucleotides within the boxA sequence (CGCTCTTA) of nutR. One mutation, (CTCTCTTA), is identical with boxA5. The second, boxA16 (CGCTATTA), has not been reported previously. In the absence of Nun, both boxA mutants reduce gal expression. Analysis of in vivo fusion RNA indicates that the mutations increase termination at or near tR1, a rho-dependent lambda terminator located upstream from the fusion point. In contrast to the nutR+ fusion, Nun stimulates gal expression in the boxA mutants by suppressing transcription termination in the tR1 region. Nun antitermination, however, does not extend to distal terminators. The lambda N-function also suppresses termination at or near tR1 in the mutant fusions. N fails to suppress terminators distal to tR1 in the boxA5 fusion, but displays persistent antitermination activity in the boxA16 fusion. A similar reversal of Nun activity occurs when wild-type fusions are introduced into nusA1, nusB5 or nusE71 hosts. We therefore suggest that Nun and N can interact with RNA polymerase in the absence of wild-type boxA, nusA, nusB or nusE, but that the complex formed with mutant components differs functionally from wild-type.

Ribosomal protein L4 stimulates in vitro termination of transcription at a NusA-dependent terminator in the S10 operon leader

The 11-gene S10 ribosomal protein operon of Escherichia coli is under the autogenous control of L4, the product of the third gene of the operon. Ribosomal protein L4 inhibits both transcription and translation of the operon. Our in vivo studies indicated that L4 regulates transcription by causing premature termination within the untranslated S10 operon leader. We have now used an in vitro transcription system to study the effect of purified L4 on expression of the S10 operon. We find that the cell-free system reproduces the in vivo observations. Namely, in the absence of L4, most of the RNA polymerases read through the termination site in the S10 attenuator; the addition of L4 results in increased termination at this site. However, RNA polymerase does not terminate at the S10 attenuator, with or without L4, unless an additional factor, protein NusA, is added to the transcription reaction. These results suggest that the attenuator in the S10 operon is a NusA-dependent terminator whose efficiency is regulated by ribosomal protein L4.

Structural and functional properties of the segments of lambda cro mRNA that interact with transcription termination factor Rho

Termination of transcription at tR1, the Rho-dependent terminator between genes cro and cII of bacteriophage lambda, is dependent upon the structure of segments near the 3' end of the nascent cro gene transcript and on contacts between Rho protein and a 3' proximal segment called rut. The characteristics of the structure of cro RNA in the region from residue 220 to residue 355 in free, isolated RNA and in the presence of Rho or NusA proteins were analyzed by measuring relative rates of reactivity of individual nucleotides with chemicals and enzymes of defined specificities. The results indicate that the rut segments are single-stranded and become blocked to the action of the various probes in the presence of Rho factor. They also show that this region contains two stem-loop structures; one involves the boxB sequence of nutR, the other precedes the tR1 subsite II end points. The results provide direct evidence for a primary binding contact between Rho protein and the rut segment of cro RNA and demonstrate that this binding contact remains stable when the cro RNA is serving as a cofactor for ATP hydrolysis, an observation that is consistent with a mechanism in which Rho maintains contact with the rut region while it makes additional interactions with RNA that are coupled to ATP hydrolysis.

Sequence-specific recognition of RNA hairpins by bacteriophage antiterminators requires a conserved arginine-rich motif

We have dissected the protein and nucleic acid determinants that direct a group of transcriptional antiterminators to their specific target operons. These antiterminators, the N gene products of phages lambda, 21, and P22, function solely with their respective recognition sites, nut, to modify RNA polymerase to a termination-resistant form. We demonstrate that a unique hairpin sequence within each nut site, called boxB, confers genome specificity by interacting with a small amino-terminal domain of the cognate N protein. This interaction is dependent upon an arginine-rich subdomain, which is conserved not only among the N proteins but also in many RNA binding proteins from ribosomes and RNA virus capsids. Notably, this motif constitutes an essential domain of the HIV protein Tat whose function as a trans-activator requires a specific hairpin sequence.

Ribosomal RNA operon anti-termination. Function of leader and spacer region box B-box A sequences and their conservation in diverse micro-organisms

All Escherichia coli rrn operons show a common motif in which anti-terminator box B-box A sequences occur twice, first in the leader and again in the 16 S-23 S spacer. In this study we have analyzed several aspects of rrn anti-termination by leader and spacer anti-terminator sequences. Using DNA synthesis and a plasmid test system, we incorporated random changes into the leader anti-terminator region and examined these mutations for their ability to read through a strong terminator. We also examined anti-termination by synthetic box A and by rrn spacer region sequences. Information derived from these experiments was used to search the rrn sequences of other micro-organisms for possible anti-termination features. Our principal conclusions were that: (1) box A was sufficient for terminator readthrough; (2) we could show no positive requirement for box B in our test system; (3) many of the negative anti-terminator mutations caused a promoter up-effect in the absence of a terminator; (4) the search of rrn operons from other micro-organisms revealed that anti-terminator-like box B-box A sequences exist in leader and spacer regions of both eubacteria and archaebacteria. The frequent occurrence of this pattern suggested that the E. coli rrn anti-termination motif is widespread in nature and has been conserved in microbial evolution.

Features of the rho-dependent transcription termination polar element within the hisG cistron of Salmonella typhimurium

Previous genetic analysis showed that the polar effects of mutations in the hisG cistron of Salmonella typhimurium are dependent on the presence of a single putative transcription termination element within the hisG gene. In fact, all proximal mutations causing translation termination are strongly polar, whereas distal ones are not. The element was mapped by isolating mutations able to relieve the polar phenotype, and they were found to be small deletions in the region downstream of the translational stop codon (M. S. Ciampi and J. R. Roth, Genetics 118:193-202, 1988). In this study, we analyzed the his-specific RNAs synthesized in vivo in different strains harboring the polar frameshift hisG2148 mutation. The nature of the polarity effects is clearly transcriptional, since shorter RNA molecules were produced. When the hisG2148 mutation was transferred in a rho background or in strains harboring the small distal deletions, an increase in readthrough transcription was observed. The transcriptional termination element was characterized in more detail by performing high-resolution S1 nuclease mapping experiments. This analysis showed that (i) termination or exonucleolytic degradation following termination produced transcripts with heterogeneous 3' ends; (ii) this process is dependent on the transcription termination factor Rho, since relief of termination occurs in a rho background; and (iii) the element appears to function as a transcription terminator, at least to some extent, even in the course of active translation of the hisG cistron.