- Volume 145, Issue 11, 1999

Growth of Alcaligenes sp. strain O-1 with 2-aminobenzenesulfonate (ABS; orthanilate) as sole source of carbon and energy requires expression of the soluble, multicomponent 2-aminobenzenesulfonate 2,3-dioxygenase system (deaminating) (ABSDOS) which is plasmid-encoded. ABSDOS was separated by anion-exchange chromatography to yield a flavin-dependent reductase component and an iron-dependent oxygenase component. The oxygenase component was purified to about 98% homogeneity and an α2β2 subunit structure was deduced from the molecular masses of 134, 45 and 16 kDa for the native complex, and the α and β subunits, respectively. Analysis of the amount of acid labile sulfur and total iron, and the UV spectrum of the purified oxygenase component indicated one [2Fe–2S] Rieske centre per α subunit. The inhibition of activity by the iron-specific chelator o-phenanthroline indicated the presence of an additional iron-binding site. Recovery of active protein required strictly anoxic conditions during all purification steps. The FAD-containing reductase could not be purified. ABSDOS oxygenated nine sulfonated compounds; no oxygen uptake was detected with carboxylated aromatic compounds or with aliphatic sulfonated compounds. K m values of 29, 18 and 108 μM and V max values of 140, 110 and 72 pkat for ABS, benzenesulfonate and 4-toluenesulfonate, respectively, were observed. The N-terminal amino acid sequences of the α- and β-subunits of the oxygenase component allowed PCR primers to be deduced and the DNA sequence of the α-subunit was thereafter determined. Both redox centres were detected in the deduced amino acid sequence. Sequence data and biochemical properties of the enzyme system indicate a novel member of the class IB ring-hydroxylating dioxygenases.

Anaerobic catabolism of toluene is initiated by addition of the methyl group of toluene to the double bond of a fumarate cosubstrate to yield the first intermediate, benzylsuccinate. This reaction is catalysed by the glycyl-radical enzyme benzylsuccinate synthase, as shown for the denitrifying bacterium Thauera aromatica. Benzylsuccinate is further oxidized to benzoyl-CoA, the central intermediate of anaerobic degradation of aromatic compounds. The authors show here by experiments with cell extracts of toluene-grown T. aromatica that the pathway of benzylsuccinate oxidation requires activation of the free acid to a CoA-thioester, catalysed by a toluene-induced, reversible succinyl-CoA-dependent CoA-transferase. The product of the CoA-transferase reaction, benzylsuccinyl-CoA, is oxidized to benzoyl-CoA and succinyl-CoA in extracts of toluene-grown cells, adding proof to the proposed anaerobic toluene-catabolic pathway. The stereochemical preferences of the enzymes catalysing formation and activation of benzylsuccinate have been analysed. Benzylsuccinate synthase was found to produce exclusively (R)-(+)-benzylsuccinate, although the proposed reaction mechanism of this enzyme proceeds via radical intermediates. In accordance, the reaction of succinyl-CoA:benzylsuccinate CoA-transferase is also specific for (R)-(+)-benzylsuccinate and does not proceed with the (S)-(−)-enantiomer.

The 5′-cap structure of eukaryotic mRNA is methylated at the terminal guanosine by RNA (guanine-N 7 -)-methyltransferase (cap MTase). Saccharomyces cerevisiae ABD1 (ScABD1) and human hMet (also called CMT1) genes are responsible for this enzyme. The ABD1 homologue was cloned from the pathogenic fungus Candida albicans and named C. albicans ABD1 (CaABD1). When expressed as a fusion with glutathione S-transferase (GST), CaAbd1p displayed cap MTase activity in vitro and rescued S. cerevisiae abd1Δ null mutants, indicating that CaABD1 specifies an active cap MTase. Although the human cap MTase binds to the human capping enzyme (Hce1p), which possesses both mRNA guanylyltransferase (mRNA GTase) and mRNA 5′-triphosphatase (mRNA TPase) activities, yeast two-hybrid analysis demonstrated that in yeast neither mRNA GTase nor mRNA TPase physically interacted with the Abd1 protein. Comparison of the amino acid sequences of known and putative cap MTases revealed a highly conserved amino acid sequence motif, Phe/Val-Leu-Asp/Glu-Leu/Met-Xaa-Cys-Gly-Lys-Gly-Gly-Asp-Leu-Xaa-Lys, which encompasses the sequence motif characteristic of divergent methyltransferases. Mutations in CaAbd1p of leucine at the second and the twelfth positions (so far uncharacterized) to alanine severely diminished the enzyme activity and the functionality in vivo, whereas those of leucine at the fourth, cysteine at the sixth, lysine at the eighth, and glycine at the tenth positions did not. Furthermore, valine substitution for the twelfth, but not for the second, leucine in that motif abolished the activity and functionality of CaAbd1p. Thus, it appears that leucine at the second and the twelfth positions in the motif, together with a previously identified acidic residue in the third, glycine at the sixth and glutamic acid at the eleventh positions, play important roles in the catalysis, and that side chain length is crucial for the activity at the twelfth position in the motif.

The starvation-stress response (SSR) of Salmonella typhimurium includes gene products necessary for starvation avoidance, starvation survival and virulence for this bacterium. Numerous genetic loci induced during carbon-source starvation and required for the long-term-starvation survival of this bacterium have been identified. The SSR not only protects the cell against the adverse effects of long-term starvation but also provides cross-resistance to other environmental stresses, e.g. thermal challenge (55 °C) or acid-pH challenge (pH 2·8). One carbon-starvation-inducible lac fusion, designated stiA was previously reported to be a σS-dependent SSR locus that is phosphate-starvation, nitrogen-starvation and H2O2 inducible, positively regulated by (p)ppGpp in a relA-dependent manner, and negatively regulated by cAMP:cAMP receptor protein complex and OxyR. We have discovered through sequence analysis and subsequent biochemical analysis that the stiA::lac fusion, and a similarly regulated lac fusion designated sti-99, lie at separate sites within the first gene (narZ) of an operon encoding a cryptic nitrate reductase (narZYWV) of unknown physiological function. In this study, it was demonstrated that narZ was negatively regulated by the global regulator Fnr during anaerobiosis. Interestingly, narZ(YWV) was required for carbon-starvation-inducible thermotolerance and acid tolerance. In addition, narZ expression was induced ∼20-fold intracellularly in Madin-Darby canine kidney epithelial cells and ∼16-fold in intracellular salts medium, which is believed to mimic the intracellular milieu. Also, a narZ1 knock-out mutation increased the LD50 ∼10-fold for S. typhimurium SL1344 delivered orally in the mouse virulence model. Thus, the previously believed cryptic and constitutive narZYWV operon is in fact highly regulated by a complex network of environmental-stress signals and global regulatory functions, indicating a central role in the physiology of starved and stressed cells.

Each of the Paracoccus denitrificans genes in the c-type cytochrome biogenesis gene cluster ccmABCDG, plus the two flanking genes ORF117 and hisH, were individually disrupted by Ω insertion. Resultant phenotypes were restored to the wild-type by complementation from a set of plasmids. All of the ccm genes, but neither ORF117 nor hisH, were required for c-type cytochrome biogenesis; only ccmG was also implicated in the biosynthesis of cytochrome aa 3. Disruption of ccmC or ccmG resulted in failure to grow on rich media, but disruption of ccmA, ccmB or ccmD did not. The ccmC mutant, but not the ccmA, ccmB or ccmD mutants, also exhibited the increased sensitivity to growth inhibition by oxidized thiol compounds previously observed for the ccmG mutant. In contrast to the ccmG mutant, however, growth of the ccmC mutant on rich media could not be restored by DTT. Siderophore biosynthesis and/or secretion by P. denitrificans was also attenuated by disruption of ccmC and ccmG but not of ccmA, ccmB or ccmD. These results indicate that CcmC can function independently of CcmA, CcmB and CcmD despite other evidence that these gene products form an ATP-binding cassette (ABC)-type-transporter with the subunit composition (CcmA)2-CcmB-CcmC or (CcmA)2-CcmB-CcmC-CcmD, and also suggest a possible link between the functions of CcmC and CcmG.

The antibiotic TA of Myxococcus xanthus is a complex macrocyclic polyketide, produced through successive condensations of acetate by a type I PKS (polyketide synthase) mechanism. The genes encoding TA biosynthesis are clustered on a 36 kb DNA fragment, which has been cloned and analysed. The chemical structure of TA and the mechanism by which it is synthesized indicate the need for several post-modification steps, which are introduced into the carbon chain of the polyketide to form the final bioactive molecule. These include the addition of several carbon atoms originating from acetate carbonyl, three C-methylations, O-methylation and a specific hydroxylation. This paper reports the analysis of five genes which are involved in the post-modification of TA. Their functional analysis, by specific gene disruption, suggests that they may be essential for the production of the active antibiotic. The characteristics and organization of the genes suggest that they may be involved in the addition of the carbon atoms which arise from acetate.

Escherichia coli contains a stationary-phase aconitase (AcnA) that is induced by iron and oxidative stress, and a major but less stable aconitase (AcnB) synthesized during exponential growth. These enzymes were shown to resemble the bifunctional iron-regulatory proteins (IRP1)/cytoplasmic aconitases of vertebrates in having alternative mRNA-binding and catalytic activities. Affinity chromatography and gel retardation analysis showed that the AcnA and AcnB apo-proteins each interact with the 3′ untranslated regions (3′UTRs) of acnA and acnB mRNA at physiologically significant protein concentrations. AcnA and AcnB synthesis was enhanced in vitro by the apo-aconitases and this enhancement was abolished by 3′UTR deletion from the DNA templates, presumably by loss of acn-mRNA stabilization by bound apo-aconitase. In vivo studies showed that although total aconitase activity is lowered during oxidative stress, synthesis of the AcnA and AcnB proteins and the stabilities of acnA and acnB mRNAs both increase, suggesting that inactive aconitase mediates a post-transcriptional positive autoregulatory switch. Evidence for an iron–sulphur-cluster-dependent switch was inferred from the more than threefold higher mRNA-binding affinities of the apo-aconitases relative to the holo-enzymes. Thus by modulating translation via site-specific interactions between apo-enzyme and relevant transcripts, the aconitases provide a new and rapidly reacting component of the bacterial oxidative stress response.

Escherichia coli transcription factor sigma 54 contains motifs that resemble closely those used for RNA polymerase II in mammalian cells, including two hydrophobic heptad repeats, a very acidic region and a glutamine-rich region. Triple changes in hydrophobic or multiple changes in acidic residues in Region III are known to severely impair core-binding ability. To investigate whether all the changes in triple mutants are necessary for core binding, site-directed mutagenesis was performed to create single and double mutants in the leucine or isoleucine residues in the heptad repeat in Region III. Single mutants showed no discernible loss of function. Double mutants showed partial protection of the −12 promoter element of the glnAp2 promoter due to the partial loss of their ability to bind core RNA polymerase. These mutations were deleterious to the function of sigma 54, which retained only 30–40% of wild-type mRNA levels. However, double mutants retained nearly normal ability to form open complexes. Two triple mutants created during previous work lost most, if not all, of their ability to bind core RNA polymerase, to protect the −12 promoter element of the glnAp2 promoter and to open the transcription start site. The two triple mutants produced about 20% or less than 10% of the wild-type transcripts from the glnAp2 promoter. These results demonstrate that the hydrophobic heptad repeat in Region III is essential for core RNA polymerase binding. Progressive loss of hydrophobicity of the hydrophobic heptad repeat in Region III of sigma 54 resulted in a progressive loss of core-binding ability, leading to the loss of −12 promoter element recognition and mRNA production.

RNAI is a short RNA, 108 nt in length, which regulates the replication of the plasmid ColE1. RNAI turns over rapidly, enabling plasmid replication rate to respond quickly to changes in plasmid copy number. Because RNAI is produced in abundance, is easily extracted and turns over quickly, it has been used as a model for mRNA in studying RNA decay pathways. The enzymes polynucleotide phosphorylase, poly(A) polymerase and RNase E have been demonstrated to have roles in both messenger and RNAI decay; it is reported here that these enzymes can work independently of one another to facilitate RNAI decay. The roles in RNAI decay of two further enzymes which facilitate mRNA decay, the exonuclease RNase II and the endonuclease RNase III, are also examined. RNase II does not appear to accelerate RNAI decay but it is found that, in the absence of RNase III, polyadenylated RNAI, unprocessed by RNase E, accumulates. It is also shown that RNase III can cut RNAI near nt 82 or 98 in vitro. An RNAI fragment corresponding to the longer of these can be found in extracts of an rnc + pcnB strain (which produces RNase III) but not of an rnc pcnB strain, suggesting that RNAI may be a substrate for RNase III in vivo. A possible pathway for the early steps in RNAI decay which incorporates this information is suggested.

Cellulosomes prepared by the cellulose affinity digestion method from Clostridium thermocellum culture supernatant hydrolysed carob galactomannan during incubation at 60 °C and pH 6·5. A recombinant phage expressing mannanase activity was isolated from a library of C. thermocellum genomic DNA constructed in λZAPII. The cloned fragment of DNA containing a putative mannanase gene (manA) was sequenced, revealing an ORF of 1767 nt, encoding a protein (mannanase A; Man26A) of 589 aa with a molecular mass of 66816 Da. The putative catalytic domain (CD) of Man26A, identified by gene sectioning and sequence comparisons, displayed up to 32% identity with other mannanases belonging to family 26. Immediately downstream of the CD and separated from it by a short proline/threonine linker was a duplicated 24-residue dockerin motif, which is conserved in all C. thermocellum cellulosomal enzymes described thus far and mediates their attachment to the cellulosome-integrating protein (CipA). Man26A consisting of the CD alone (Man26A′) was hyperexpressed in Escherichia coli BL21(DE3) and purified. The truncated enzyme hydrolysed soluble and insoluble mannan, displaying a temperature optimum of 65 °C and a pH optimum of 6·5, but exhibited no activity against other plant cell wall polysaccharides. Antiserum raised against Man26A′ cross-reacted with a polypeptide with a molecular mass of 70000 Da that is part of the C. thermocellum cellulosome. A second variant of Man26A containing the N-terminal segment of 130 residues and the CD (Man26A′′) bound to ivory-nut mannan and weakly to soluble Carob galactomannan and insoluble cellulose. Man26A′ consisting of the CD alone did not bind to these polysaccharides. These results indicate that the N-terminal 130 residues of mature Man26A may constitute a weak mannan-binding domain. Sequence comparisons revealed a lack of identity between this region of Man26A and other polysaccharide-binding domains, but significant identity with a region conserved in the three family 26 mannanases from the anaerobic fungus Piromyces equi.

Two genes, accA1 and accA2, with nearly identical nucleotide sequences were cloned from Streptomyces coelicolor A3(2). The deduced amino acid sequences of the product of these two genes showed high similarity to BcpA2 of Saccharopolyspora erythraea and other biotin-containing proteins from different organisms assumed to be the α subunit of a propionyl-CoA carboxylase. A gene, pccB, encoding the carboxyl transferase subunit of this enzyme complex was also characterized. Strains disrupted in accA1 did not show any change in acetyl- or propionyl-CoA carboxylase activity, whilst cell-free extracts of a pccB mutant strain contained a reduced level of propionyl-CoA carboxylase. No mutants in accA2 could be isolated, suggesting that the gene may be essential. Heterologous expression of accA1, accA2 and pccB in Escherichia coli and in vitro reconstitution of enzyme activity confirmed that PccB is the β subunit of a propionyl-CoA carboxylase and that either AccA1 or AccA2 could act as the α component of this enzyme complex. The fact that accA2 mutants appear to be inviable suggests that this gene encodes a biotinylated protein that might be shared with other carboxyl transferases essential for the growth of S. coelicolor.

aprX is a 1326 bp gene of Bacillus subtilis strain 168 that encodes a serine protease, probably intracellular, characterized by significant similarity with subtilisins, thermitases and pyrolysins. Transcription analysis, performed by RT-PCR and primer extension, allowed the localization of the active promoter and showed that aprX is expressed in stationary phase. The pattern of expression of aprX and its dependence on various transition state regulatory genes (degU, degQ, hpr, abrB, sinR), monitored by lacZ transcriptional fusions, are distinctive from those of subtilisin and other degradative enzymes. aprX is not essential for either growth or sporulation.

PlcR is a pleiotropic regulator of extracellular virulence factors in the opportunistic human pathogen Bacillus cereus and the entomopathogenic Bacillus thuringiensis, and is induced in cells entering stationary phase. Among the genes regulated by PlcR are: plcA, encoding phosphatidylinositol-specific phospholipase C (PI-PLC); plc, encoding phosphatidylcholine-preferring phospholipase C (PC-PLC); nhe, encoding the non-haemolytic enterotoxin; hbl, encoding haemolytic enterotoxin BL (HBL); and genes specifying a putativeS-layer like surface protein and a putative extracellular RNase. By analysing 37·1 kb of DNA sequence surrounding hbl, plcA and plcR, 28 ORFs were predicted. Three novel genes putatively regulated by PlcR and encoding a neutral protease (NprB), a subtilase family serine protease (Sfp) and a putative cell-wall hydrolase (Cwh) were identified. The corresponding sfp and cwh genes were located in the immediate upstream region of plcA and could both be regulated by a putative PlcR-binding site positioned between the inversely transcribed genes. Similarly, nprB was positioned directly upstream and transcribed in the opposite orientation to plcR. Genes surrounding plcA, plcR and hblCDAB that were lacking an upstream PlcR regulatory sequence did not appear to serve functions apparently related to PlcR and did not exhibit a conserved organization in Bacillus subtilis.