Resolution of Holliday junctions in Escherichia coli: identification of the ruvC gene product as a 19-kilodalton protein.

The ruvC gene of Escherichia coli specifies a nuclease that resolves Holliday junction intermediates in genetic recombination (B. Connolly, C.A. Parsons, F.E. Benson, H.J. Dunderdale, G.J. Sharples, R.G. Lloyd, and S.C. West, Proc. Natl. Acad, Sci. USA 88:6063-6067, 1991). The gene was located between aspS and the ruvAB operon by DNA sequencing and deletion analysis of ruvC plasmids and was shown to encode a protein of 18,747 Da. Analysis of the DNA flanking ruvC indicated that the gene is transcribed independently of the LexA-regulated ruvAB operon and is not under direct SOS control. ruvC lies downstream of an open reading frame, orf-33, for a protein which migrates during sodium dodecyl sulfate-polyacrylamide gel electrophoresis as a 33-kDa polypeptide. These two genes probably form an operon. However, expression of ruvC was found to be very poor relative to that of orf-33. A double ribosomal frameshift between these genes is proposed as a possible reason for the low level of RuvC. Two further open reading frames of unknown function were identified, one on either side of the orf-33-ruvC operon.     

Escherichia coli RuvC protein is an endonuclease that resolves the Holliday structure.

Genetic evidence suggests that the Escherichia coli ruvC gene is involved in DNA repair and in the late step of RecE and RecF pathway recombination. To study the biochemical properties of RuvC protein, we overproduced and highly purified the protein. By employing model substrates, we examined the possibility that RuvC protein is an endonuclease that resolves the Holliday structure, an intermediate in genetic recombination in which two double-stranded DNA molecules are linked by single-stranded crossover. RuvC protein cleaves cruciform junctions, which are formed by the extrusion of inverted repeat sequences from a supercoiled plasmid and which are structurally analogous to Holliday junctions, by introducing nicks into strands with the same polarity. The nicked ends are ligated by E.coli or T4 DNA ligases. Analysis of the cleavage sites suggests that DNA topology rather than a particular sequence determines the cleavage site. RuvC protein also cleaves Holliday junctions which are formed between gapped circular and linear duplex DNA by the function of RecA protein. However, it does not cleave a synthetic four-way junction that does not possess homology between arms. The active form of RuvC protein, as studied by gel filtration, is a dimer. This is mechanistically suited for an endonuclease involved in swapping DNA strands at the crossover junctions. From these properties of RuvC protein and the phenotypes of the ruvC mutants, we infer that RuvC protein is an endonuclease that resolves Holliday structures in vivo.     

Escherichia coli gene ydeA encodes a major facilitator pump which exports L-arabinose and isopropyl-beta-D-thiogalactopyranoside.

Inactivation of the Escherichia coli gene ydeA, which encodes a member of the major facilitator superfamily, decreased the efflux of L-arabinose, thereby affecting the expression of AraC-regulated genes. In addition, overexpression of ydeA decreased the expression of genes regulated by isopropyl-beta-D-thiogalactopyranoside.     

Molecular analysis of the Escherichia coli recO gene.

The plasmid pLC7-47, which contains lep, rnc, and era, was found to complement the UV-sensitive and recombination-deficient phenotypes caused by the recO1504::Tn5 mutation. Southern blotting analysis demonstrated that pLC7-47 contained a segment of Escherichia coli DNA that covered the region of the E. coli chromosome containing the recO1504::Tn5 mutation. A combination of deletion mapping and insertional mutagenesis localized the recO-complementing region to an approximately 1-kilobase region of a 1.6-kilobase BamHI fragment. The DNA sequence of the 1.6-kilobase BamHI fragment was determined and contained part of era and a 726-base-pair recO open reading frame. The recO open reading frame contained three possible translation start codons and could potentially encode a polypeptide of Mr 26,000. Computer analysis indicated that the putative RecO protein had suboptimal codon usage and did not show significant homology with previously identified proteins whose sequences were present in protein data bases. A combination of primary sequence analysis and secondary structure predictions suggested that recO contains a mononucleotide-binding fold.     

Molecular analysis of the Salmonella typhimurium phoN gene, which encodes nonspecific acid phosphatase.

The phoN gene of Salmonella typhimurium encodes nonspecific acid phosphatase (EC 3.1.3.2), which is regulated by a two-component regulatory system consisting of the phoP and phoQ genes. We cloned the phoN region into a plasmid vector by complementation of a phoN mutant strain and determined the nucleotide sequence of the phoN gene and its flanking regions. The phoN gene could encode a 26-kDa protein, which was identified by the maxicell method as the product of phoN. Results of the enzyme assay and Southern hybridization with chromosomal DNA of Escherichia coli K-12 suggests that there is no phoN gene in E. coli. The regulatory pattern of phoN in E. coli and Southern hybridization analysis of the E. coli chromosome with the S. typhimurium phoP gene suggest that E. coli K-12 also harbors the phoP and phoQ genes.     

Molecular analysis of the Escherichia coli hns gene encoding a DNA-binding protein, which preferentially recognizes curved DNA sequences.

Summary We previously demonstrated that the E. coli protein, H-NS (or Hla), encoded by the gene hns (or osmZ or bglY preferentially recognizes curved DNA sequences in vitro. In order to gain further insight into the complex function of H-NS and the significance of DNA curvature, we constructed a structurally defined hns deletion mutant on the E. coli chromosome. The hns deletion mutant thus obtained showed a variety of phenotypes previously for other lesions in hns. It was further demonstrated that, in this hns deletion background, numerous E. coli cellular proteins were either strongly expressed or remarkably repressed, as compared to their expression levels in wild-type cells.     

Deletion analysis of gene minE which encodes the topological specificity factor of cell division in Escherichia coli

Division inhibition caused by the minCD gene products of Escherichia coli is suppressed specifically at mid-cell by MinE protein expressed at physiological levels. Excess MinE allows division to take place also at the poles, leading to a minicell-forming (Min⁻) phenotype. In order to investigate the basis of this topological specificity, we have analysed the ability of truncated derivatives of MinE to suppress either minCD -dependent division inhibition in a chromosomal Δ( minB ) background, or the division inhibition exerted by MinCD at the cell poles in a minB,+ strain. Our results indicate that these two effects are not mediated by identical interactions of MinE protein. In addition, gel filtration and the yeast two-hybrid system indicated that MinE interacts with itself by means of its central segment. Taken together, our results favour a model in which wild-type MinE dimer molecules direct the division inhibitor molecules to the cell poles, thus preventing polar divisions and allowing non-polar sites to divide. This model explains how excess MinE, or an excess of certain MinE derivatives which prevent the accumulation of the division inhibitor at the poles, can confer a Min⁻ phenotype in a minB ⁺ strain.     

The Escherichia coli polB gene, which encodes DNA polymerase II, is regulated by the SOS system.

The dinA (damage inducible) gene was previously identified as one of the SOS genes with no known function; it was mapped near the leuB gene, where the polB gene encoding DNA polymerase II was also mapped. We cloned the chromosomal fragment carrying the dinA region from the ordered Escherichia coli genomic library and mapped the dinA promoter precisely on the physical map of the chromosome. The cells that harbored multicopy plasmids with the dinA region expressed very high levels of DNA polymerase activity, which was sensitive to N-ethylmaleimide, an inhibitor of DNA polymerase II. Expression of the polymerase activity encoded by the dinA locus was regulated by SOS system, and the dinA promoter was the promoter of the gene encoding the DNA polymerase. From these data we conclude that the polB gene is identical to the dinA gene and is regulated by the SOS system. The product of the polB (dinA) gene was identified as an 80-kDa protein by the maxicell method.     

The structure of Escherichia coli RusA endonuclease reveals a new Holliday junction DNA binding fold

Holliday junction resolution performed by a variety of structure-specific endonucleases is a key step in DNA recombination and repair. It is believed that all resolvases carry out their reaction chemistries in a similar fashion, utilizing a divalent cation to facilitate the hydrolysis of the phosphodiester backbone of the DNA, but their architecture varies. To date, with the exception of bacteriophage T4 endonuclease VII, each of the known resolvase enzyme structures has been categorized into one of two families: the integrases and the nucleases. We have now determined the structure of the Escherichia coli RusA Holliday junction resolvase, which reveals a fourth structural class for these enzymes. The structure suggests that dimer formation is essential for Mg(2+) cation binding and hence catalysis and that like the other resolvases, RusA distorts its Holliday junction target upon binding. Key residues identified by mutagenesis experiments are well positioned to interact with the DNA.     

Revised nucleotide sequence of the gltP gene, which encodes the proton-glutamate-aspartate transport protein of Escherichia coli K-12.

The gene encoding the proton-glutamate carrier (GltP) of Escherichia coli K-12 was sequenced, and the primary structure of the protein was analyzed. The nucleotide sequence was found to differ in several aspects from the previously published sequence (B. Wallace, Y. Yang, J. Hong, and D. Lum, J. Bacteriol. 172:3214-3220, 1990). The corrected open reading frame encodes a protein of 437 (instead of 395) amino acids. Hydropathy analysis predicts 12 membrane-spanning alpha-helical regions. The complementary strand does contain an open reading frame possibly encoding a highly hydrophilic polypeptide of 272 amino acids.