Role of DNA methylation at GATC sites in the dnaA promoter, dnaAp2.

DnaA protein is required for the initiation of DNA replication at the bacterial chromosomal origin, oriC, and at the origins of many plasmids. The concentration of DnaA protein is an important factor in determining when initiation occurs during the cell cycle. Methylation of GATC sites in the dnaAp2 promoter, two of which are in the -35 and -10 sequences, has been predicted to play an important role in regulating dnaA gene expression during the cell cycle because the promoter is sequestered from methylation immediately following replication. Mutations that eliminate these two GATC sites but do not substantially change the activity of the promoter were introduced into a reporter gene fusion and into the chromosome. The chromosomal mutants are able to initiate DNA replication synchronously at both moderately slow and fast growth rates, demonstrating that GATC methylation at these two sites is not directly involved in providing the necessary amount of DnaA for precise timing of initiation during the cell cycle. Either sequestration does not involve these GATC sites, or cell cycle control of DnaA expression is not required to supply the concentration necessary for correct timing of initiation.

Chromosomal insertions localized around oriC affect the cell cycle in Escherichia coli.

The present work reports the effects of localized insertions around the origin of Escherichia coli chromosome, oriC, on cell cycle parameters. These insertions cause an increase of the C period with an inverse correlation to the distance from oriC. In addition, Omega insertion near oriC causes an increase in the number of replication forks per chromosome, n, and Tn10 insertion causes a decrease in growth rate. We found that the same insertion positioned in another region of the chromosome, outside of oriC, has a negligible effect on the C period. Marker frequency analysis suggests a slower replication velocity along the whole chromosome. We propose that the insertions positioned at less than 2 kbp from oriC could create a structural alteration in the origin of replication that would result in a longer C period. Flow cytometry reveals that asynchrony is not associated with these alterations.

Chitobiase, a new reporter enzyme.

N,N'-diacetylchitobiase (chitobiase) from the marine organism Vibrio harveyi is a highly stable reporter enzyme for gene fusions. This enzyme hydrolyzes the disaccharide chitobiose to N-acetyl glucosamine. The advantages of the reporter gene encoding chitobiase (chb) are: (i) that chitobiase and N-acetyl-beta-D-glucosaminidase activities are missing in E. coli strains, (ii) chitobiase can be monitored using blue/white colony indicator plates and (iii) convenient substrates for this enzyme are commercially available. The use of chitobiase as a reporter enzyme is generally applicable to the study of gene expression in those bacteria that do not contain N-acetyl-beta-D-glucosaminidases. We constructed plasmid vectors containing a multiple cloning site for producing in-frame fusions to chitobiase, the attP of lambda phase for movement into the bacterial chromosome for single-copy analysis, the gene encoding chloramphenicol acetyltransferase (cat), the pACYC184 origin of replication and the rrnBt1t2 terminator region upstream of the chb gene to prevent read-through from other promoters. In-frame fusions between the dnaA gene and chb were moved to the chromosome by site-specific recombination with the chromosomal attB site. These single-copy fusions were assayed for chitobiase to examine the effects of a deletion in the dnaA regulatory region.

Fis binding in the dnaA operon promoter region.

The region between the rpmH and dnaA genes contains five promoters that divergently express the ribosomal protein L34 and the proteins of the dnaA operon, including DnaA, the beta clamp of DNA polymerase III holoenzyme, and RecF. The DNA-binding protein Fis was shown by the band shift assay to bind near the rpmHp2 and dnaAp2 promoters and by DNase I footprinting to bind to a single site in the dnaAp2 promoter overlapping the -35 and spacer sequences. There were no observable differences in Fis affinity or the angle of bending induced by Fis between methylated and unmethylated DNA fragments containing the Fis binding site in the dnaAp2 promoter. Fis directly or indirectly represses the expression of DnaA protein and the beta clamp of DNA polymerase III. A fis null mutant containing a dnaA-lacZ in-frame fusion had twofold greater beta-galactosidase activity than a fis wild-type strain, and induced expression of Fis eliminated the increase in activity of the fusion protein. A two- to threefold increase in the levels of DnaA and beta clamp proteins was found in a fis null mutant by immunoblot gel analysis.

Cloning of nod gene regions from mesquite rhizobia and bradyrhizobia and nucleotide sequence of the nodD gene from mesquite rhizobia.

Nitrogen-fixing symbiosis between bacteria and the tree legume mesquite (Prosopis glandulosa) is important for the maintenance of many desert ecosystems. Genes essential for nodulation and for extending the host range to mesquite were isolated from cosmid libraries of Rhizobium (mesquite) sp. strain HW17b and Bradyrhizobium (mesquite) sp. strain HW10h and were shown to be closely linked. All of the cosmid clones of rhizobia that extended the host range of Rhizobium (Parasponia) sp. strain NGR234CS to mesquite also supported nodulation of a Sym- mesquite strain. The cosmid clones of bradyrhizobia that extended the host range of Rhizobium (Parasponia) sp. strain NGR234CS to mesquite were only able to confer nodulation ability in the Sym- mesquite strain if they also contained a nodD-hybridizing region. Subclones containing just the nodD genes of either genus did not extend the host range of Rhizobium (Parasponia) sp. to mesquite, indicating that the nodD gene is insufficient for mesquite nodulation. The nodD gene region is conserved among mesquite-nodulating rhizobia regardless of the soil depth from which they were collected, indicating descent from a common ancestor. In a tree of distance relationships, the NodD amino acid sequence from mesquite rhizobia clusters with homologs from symbionts that can infect both herbaceous and tree legumes, including Rhizobium tropici, Rhizobium leguminosarum bv; phaseoli, Rhizobium loti, and Bradyrhizobium japonicum.

Variation of clonal, mesquite-associated rhizobial and bradyrhizobial populations from surface and deep soils by symbiotic gene region restriction fragment length polymorphism and plasmid profile analysis.

Genetic characteristics of 14 Rhizobium and 9 Bradyrhizobium mesquite (Prosopis glandulosa)-nodulating strains isolated from surface (0- to 0.5-m) and deep (4- to 6-m) rooting zones were determined in order to examine the hypothesis that surface- and deep-soil symbiont populations were related but had become genetically distinct during adaptation to contrasting soil conditions. To examine genetic diversity, Southern blots of PstI-digested genomic DNA were sequentially hybridized with the nodDABC region of Rhizobium meliloti, the Klebsiella pneumoniae nifHDK region encoding nitrogenase structural genes, and the chromosome-localized ndvB region of R. meliloti. Plasmid profile and host plant nodulation assays were also made. Isolates from mesquite nodulated beans and cowpeas but not alfalfa, clover, or soybeans. Mesquite was nodulated by diverse species of symbionts (R. meliloti, Rhizobium leguminosarum bv. phaseoli, and Parasponia bradyrhizobia). There were no differences within the groups of mesquite-associated rhizobia or bradyrhizobia in cross-inoculation response. The ndvB hybridization results showed the greatest genetic diversity among rhizobial strains. The pattern of ndvB-hybridizing fragments suggested that surface and deep strains were clonally related, but groups of related strains from each soil depth could be distinguished. Less variation was found with nifHDK and nodDABC probes. Large plasmids (>1,500 kb) were observed in all rhizobia and some bradyrhizobia. Profiles of plasmids of less than 1,000 kb were related to the soil depth and the genus of the symbiont. We suggest that interacting selection pressures for symbiotic competence and free-living survival, coupled with soil conditions that restrict genetic exchange between surface and deep-soil populations, led to the observed patterns of genetic diversity.

Guanosine tetraphosphate inhibits protein synthesis in vivo. A possible protective mechanism for starvation stress in Escherichia coli.

Guanosine 3',5'-bispyrophosphate (ppGpp) accumulates in bacteria in response to either amino acid or energy source starvation. Here we demonstrate that levels of ppGpp similar to those induced by amino acid starvation inhibit the rate of protein synthesis by 84-91%. The intracellular concentration of ppGpp is manipulated in our studies by increasing the expression of a truncated relA gene encoding a smaller but catalytically active peptide with ppGpp synthetase activity. We find that the intracellular activity of the truncated RelA peptide is insensitive to chloramphenicol, unlike the product of the wild-type relA gene, ppGpp synthetase I. Previously, this same ppGpp expression system was used (Schreiber, G., Metzger, S., Aizenman, E., Roza, S., Cashel, M., and Glaser, G. (1991) J. Biol. Chem. 226, 3760-3767) to demonstrate that increasing the ppGpp concentration inhibits growth and ribosomal RNA transcription, and they found suggestive evidence for ppGpp inhibition of protein synthesis. We further investigated the effect of ppGpp on protein synthesis and find that ppGpp is a potent inhibitor of protein synthesis as well as glycerol accumulation but has no effect on transport of methionine, the amino acid used in measuring protein synthesis rates, or on uptake of alpha-methylglucoside, a non-metabolizable analogue of glucose.

RecA protein of Escherichia coli and chromosome partitioning.

Escherichia coli cells deficient in RecA protein frequently contain an abnormal number of chromosomes after completion of ongoing rounds of DNA replication. This suggests that RecA protein may be required for correct timing of initiation of DNA replication; however, we show here that initiation of DNA replication is properly timed in recA mutants. We also find that more than 10% of recA mutant cells contain no DNA. These anucleate cells appear to arise from partitioning of all the DNA into one daughter cell and no DNA into the other daughter cell. Based on these and previously published results, we propose that RecA protein is required for equal partitioning of chromosomes into the two daughter cells.

Escherichia coli cells lacking methylation-blocking factor (leucine-responsive regulatory protein) have precise timing of initiation of DNA replication in the cell cycle.

A protein that is required for specific methylation inhibition of two GATC sites in the papBA pilin promoter region, known as methylation-blocking factor (Mbf) and recently shown to be identical to the leucine-responsive regulatory protein (Lrp), is not responsible for the delayed methylation at oriC implicated in an eclipse period following initiation of DNA replication. Cells containing a transposon mutation within the mbf (lrp) gene initiate DNA replication at the correct time during the cell cycle, whereas cells with increased amounts of the Dam methyltransferase initiate DNA replication randomly throughout the cell cycle.

DNA replication, the bacterial cell cycle, and cell growth.

The coupling of replication to the cell cycle and cell growth involves events that occur at oriC. Immediately after initiation, there is an eclipse phase during which reinitiation from the newly synthesized origins is prevented. GATC sites in oriC remain in a hemimethylated state longer than other sites because of their association with the outer membrane, which prevents DnaA from binding and activating additional rounds of initiation. After the origins are methylated and released from the outer membrane, the concentration of newly synthesized DnaA and the activation of oriC by transcription from the nearby mioC and gid promoters determine when the next rounds of replication initiate. If growth rate is reduced, the synthesis of (p)ppGpp will increase, and this will lead to a decrease in dnaA, mioC, and gid transcription. On the other hand, if growth rate is increased by access to a tasty meal, synthesis of (p)ppGpp will decrease, expression of dnaA, mioC, and gid genes will increase, and a shortening of the interinitiation time will result. The participation of all these control features ensures rapid and precise coordination of DNA replication with cell growth.

Coupling of DNA replication to growth rate in Escherichia coli: a possible role for guanosine tetraphosphate.

Two promoters for the Escherichia coli operon that contains the four genes dnaA, dnaN, recF, and gyrB were found to be growth rate regulated and under stringent control. Transcript abundance relative to total RNA increased with the growth rate. Changes in transcription from the dnaAp1 and dnaAp2 promoters that were induced by amino acid starvation and chloramphenicol and were relA dependent were correlated with the stringent response. The abundance of these transcripts per total RNA also decreased in spoT mutants as the severity of the mutation increased (guanosine 5'-diphosphate 3'-diphosphate [ppGpp] basal levels increased). Because expression of these promoters appears to be inhibited by ppGpp, it is proposed that one mechanism for coupling DNA replication to the growth rate of bacteria is through ppGpp synthesis at the ribosome.

A model for the initiation of replication in Escherichia coli.

The role of the protein DnaA as the principal control of replication initiation is investigated by a mathematical model. Data showing that DnaA is growth rate regulated suggest that its concentration alone is not the only factor determining the timing of initiation. A mathematical model with stochastic and deterministic components is constructed from known experimental evidence and subdivides the total pool of DnaA protein into four forms. The active form, DnaA.ATP, can be bound to the origin of replication, oriC, where it is assumed that a critical level of these bound molecules is needed to initiate replication. The active form can also exist in a reserve pool bound to the chromosome or a free pool in the cytoplasm. Finally, a large inactive pool of DnaA protein completes the state variables and provides an explanation for how the DnaA.ATP form could be the principal controlling element in the timing of initiation. The fact that DnaA protein is an autorepressor is used to derive its synthesis rate. The model studies a single exponentially growing cell through a series of cell divisions. Computer simulations are performed, and the results compare favorably to data for different cell cycle times. The model shows synchrony of initiation events in agreement with experimental results.

N,N'-diacetylchitobiase of Vibrio harveyi. Primary structure, processing, and evolutionary relationships.

The nucleotide sequence of the gene, chb, encoding the outer membrane protein, N,N'-diacetylchitobiase (chitobiase), of the marine bacterium, Vibrio harveyi, has been determined. The amino acid sequence of prechitobiase was derived from the nucleotide sequence. Prechitobiase has a molecular mass of 97,771 Da and consists of 883 amino acid residues. A characteristic signal peptide is present at the amino terminus whose removal is inhibited by the antibiotic, globomycin, suggesting that mature chitobiase is a lipoprotein with a maturation pathway similar to that of the Escherichia coli major outer membrane lipoprotein. A perfect homology to six amino acids at the processing and modification region of the outer membrane lipoprotein of E. coli was found with amino acids 15-19 of the deduced prechitobiase protein sequence. Chitobiase shares similarities and possibly common ancestry with the alpha-chain of the human beta-hexosaminidase. A comparison of the amino acid sequences of chitobiase and the alpha-chain of beta-hexosaminidase gave a highly significant alignment score of 19.1 standard deviation units above a mean randomized alignment score. Primer extension analysis of the promoter region revealed three transcription initiation sites used by E. coli cells harboring the chb gene, two of which were also evident in V. harveyi cells.

Expression of Escherichia coli dnaA and mioC genes as a function of growth rate.

The synthesis of specific cellular components related to the initiation process of DNA replication was correlated with changes in growth rate. The concentrations of DnaA protein and mioC mRNA were determined for cells grown at six different growth rates; both increased relative to either total protein or total RNA, respectively, as the growth rate increased. Expression from the chromosomal mioC promoter, which contains a DnaA protein-binding site, was not repressed when the DnaA protein concentration was increased and was not derepressed in a dnaA46 mutant at 42 degrees C. The mioC transcript had a characteristic mRNA-type half-life of 1.51 min.

Translocation of Vibrio harveyi N,N'-diacetylchitobiase to the outer membrane of Escherichia coli.

The gene encoding N,N'-diacetylchitobiase (chitobiase) of the chitinolytic marine bacterium Vibrio harveyi has been isolated. While expression of the chitobiase gene (chb) was inducible by N,N'-diacetylchitobiose in V. harveyi, it was expressed constitutively when cloned in Escherichia coli, suggesting that controlling elements are not closely linked to chb. Chitobiase was found in the membrane fraction of E. coli cells containing plasmids with the cloned V. harveyi chb gene. When membranes of such cells were separated on Osborn gradients, chitobiase activity was found mainly in the outer membrane band. Translocation of the enzyme to the outer membrane was accompanied by cleavage of a signal peptide. A fusion protein, in which 22 amino acids from the amino terminus of prechitobiase were replaced with 21 amino acids from the pUC19 lacZ amino terminus, was not processed, and 99% of the activity was located in the cytoplasmic fraction. A homology to six amino acids surrounding the lipoprotein processing and modification site was found near the amino terminus of prechitobiase.

RNA polymerase pauses in vitro within the Escherichia coli origin of replication at the same sites where termination occurs in vivo.

An in vitro transcription system able to distinguish initiation at the 16-kDa promoter from elongation events was used to identify factors that might participate in transcription termination within oriC. Pausing in the oriC region occurs at the same sites where termination occurs in vivo. Ten of these sites overlap RNA:DNA junctions in oriC. The pausing that occurs in vitro was not converted to efficient termination by guanosine 5'-diphosphate 3'-diphosphate, NusA, and Rho alone or in combination, or by DnaA suggesting that in vivo other or additional factors contribute to termination at oriC. Transcription from the 16-kDa promoter was 90% inhibited by the nucleotides guanosine 5'-diphosphate 3'-diphosphate and guanosine 5'-triphosphate 3'-diphosphate in agreement with previous observations that this promoter is stringently regulated.

RNA terminating within the E. coli origin of replication: stringent regulation and control by DnaA protein.

RNA entering the E. coli replication origin, oriC, in the counterclockwise direction terminates at several sites throughout the origin sequence. The significant finding was that nine clusters of these termination sites are found at the nine clusters of RNA to DNA transitions in oriC. The majority of these transcripts terminates with cytosine. Termination sites are associated with 9 of the 11 GATC sites and all DnaA protein-binding sites. Chloramphenicol-treated cells contain an increased amount of this RNA species, while cells starved for isoleucine have greatly reduced levels, indicating that synthesis of these transcripts is stringently regulated. Both decreased and increased intracellular levels of DnaA protein decrease the fraction of transcription that enters oriC.

Transcription termination within the Escherichia coli origin of DNA replication, oriC.

Initiation of DNA replication from the Escherichia coli origin, oriC, is dependent on an RNA polymerase-mediated transcription event. The function of this RNA synthetic event in initiation, however, remains obscure. Since control of the synthesis of this RNA could serve a key role in the overall initiation process, transcription regulatory sites within and near oriC were identified using the galK fusion vector system. Our results confirm the existence of a transcription termination signal within oriC, first identified by Hansen et al. (1981), for the 16 kd transcript that is transcribed counterclockwise towards oriC. Termination is shown to be 92% efficient. A similar approach led to the detection of transcription termination within the chromosomal replication origin of Klebsiella pneumoniae. Approximately 50% of the E. coli 16 kd transcripts appear to terminate before reaching oriC between the XhoI (+416 bp) and the HindIII (+243 bp) sites. The predominant 3' ends of RNA that enter oriC, as determined by SI nuclease mapping, were located at positions +20 +/- 2, +23 +/- 2, +37, +39, +52, +66, +92, and +107. These termination sites, which map cl to RNA . DNA junctions identified by Kohara et al. (1985), appear as triplets and quadruplets. The E. coli oriC Pori-L promoter described in in vitro transcription studies by Lother and Messer (1981) was not detected in this study in either wildtype cells or isogenic dnaA mutants at the nonpermissive temperature. A new promoter activity, Pori-R1, was identified within the E. coli origin in the clockwise direction.

Importance of state of methylation of oriC GATC sites in initiation of DNA replication in Escherichia coli.

In vivo and in vitro evidence is presented implicating a function of GATC methylation in the Escherichia coli replication origin, oriC, during initiation of DNA synthesis. Transformation frequencies of oriC plasmids into E. coli dam mutants, deficient in the GATC-specific DNA methylase, are greatly reduced compared with parental dam+ cells, particularly for plasmids that must use oriC for initiation. Mutations that suppress the mismatch repair deficiency of dam mutants do not increase these low transformation frequencies, implicating a new function for the Dam methylase. oriC DNA isolated from dam- cells functions 2- to 4-fold less well in the oriC-specific in vitro initiation system when compared with oriC DNA from dam+ cells. This decreased template activity is restored 2- to 3-fold if the DNA from dam- cells is first methylated with purified Dam methylase. Bacterial origin plasmids or M13-oriC chimeric phage DNA, isolated from either base substitution or insertion dam mutants of E. coli, exhibit some sensitivity to digestion by DpnI, a restriction endonuclease specific for methylated GATC sites, showing that these dam mutants retain some Dam methylation activity. Sites of preferred cleavage are found within the oriC region, as well as in the ColE1-type origin.