Werner Klipp

New derivatives of transposon tn5 suitable for mobilization of replicons, generation of operon fusions and induction of genes in Gram-negative bacteria

Three types of new variants of the broad-host-range transposon Tn5 are described. (i) Tn5-mob derivatives with the new selective resistance (R) markers GmR, SpR and TcR facilitate the efficient mobilization of replicons within a wide range of Gram-negative bacteria. (ii) Promoter probe transposons carry the promoterless reporter genes lacZ, nptII, or luc, and NmR, GmR or TcR as selective markers. These transposons can be used to generate transcriptional fusions upon insertion, thus facilitating accurate determinations of gene expression. (iii) Tn5-P-out derivatives carry the npt- or tac-promoter reading out from the transposon, and TcR, NmR or GmR genes. These variants allow the constitutive expression of downstream genes. The new Tn5 variants are available on mobilizable Escherichia coli vectors suitable as suicidal carriers for transposon mutagenesis of non-E. coli recipients and some on a phage lambda mutant to be used for transposon mutagenesis in E. coli.

Nucleotide sequence of a 24,206-base-pair DNA fragment carrying the entire nitrogen fixation gene cluster of Klebsiella pneumoniae☆

The complete nucleotide sequence (24,206 base-pairs) of the Klebsiella pneumoniae gene region for nitrogen fixation (nif) is presented. Coding regions corresponding to the 19 known nif genes (including nifW and nifZ) could be identified. An additional open reading frame of 216 base-pairs, called nifT, was detected between nifK and nifY. Search for transcriptional signal structures revealed some unusual features: (1) several possible NifA-binding motifs are present in the intergenic regions between nifJ and nifH as well as between nifX and nifU; (2) a perfect NifA-binding motif, preceding the nifENX promoter, is located within an inverted repeat structure; (3) structures resembling the consensus nif promoter are found within the coding regions of nifW and nifZ and, together with a NifA-binding motif, in nifN. Typical rho-independent termination structures were detected only downstream from the nifHDKTY and the nifBQ operons. Analysis of the deduced amino acid sequences revealed the presence of two Cys-X2-Cys-X2-Cys-X3-Cys-Pro clusters in the pyruvate-flavodoxin oxidoreductase NifJ. This arrangement of cysteine residues is normally present only in ferredoxins. A high degree of homology between the two gene products (NifE and NifN) involved in iron-molybdenum cofactor biosynthesis and the two nitrogenase component I structural proteins (NifD and NifK) was found. All four proteins are characterized by the conserved motif His-Gly-X2-Gly-Cys, which may play a role in binding the iron-molybdenum cofactor.

Identification of a new class of nitrogen fixation genes in Rhodobacter capsalatus: a putative membrane complex involved in electron transport to nitrogenase

DNA sequence analysis of a 12236 by fragment, which is located upstream of nifE in Rhodobacter capsulatus nif region A, revealed the presence of ten open reading frames. With the exception of fdxC and fdxN, which encode a plant-type and a bacterial-type ferredoxin, the deduced products of these coding regions exhibited no significant homology to known proteins. Analysis of defined insertion and deletion mutants demonstrated that six of these genes were required for nitrogen fixation. Therefore, we propose to call these genes rnfA, rnfB, rnfC, rnfD, rnfE and rnfF (for Rhodobacter nitrogen fixation). Secondary structure predictions suggested that the rnf genes encode four potential membrane proteins and two putative iron-sulphur proteins, which contain cysteine motifs (C-X2-C-X2-C-X3-C-P) typical for [4Fe-4S] proteins. Comparison of the in vivo and in vitro nitrogenase activities of fdxN and rnf mutants suggested that the products encoded by these genes are involved in electron transport to nitrogenase. In addition, these mutants were shown to contain significantly reduced amounts of nitrogenase. The hypothesis that this new class of nitrogen fixation genes encodes components of an electron transfer system to nitrogenase was corroborated by analysing the effect of metronidazole. Both the fdxN and rnf mutants had higher growth yields in the presence of metronidazole than the wild type, suggesting that these mutants contained lower amounts of reduced ferredoxins.

Genetic characterization and sequence analysis of the duplicated nifA/nifB gene region of Rhodobacter capsulatus

SummaryA DNA region showing homology to Klebsiella pneumoniae nifA and nifB is duplicated in Rhodobacter capsulatus. The two copies of this region are called nifA/nifB copy I and nifA/nifB copy II. Deletion mutagenesis demonstrated that either of the two copies is sufficient for growth in nitrogen-free medium. In contrast, a double deletion mutant turned out to be deficient in nitrogen fixation. The complete nucleotide sequence of a 4838 bp fragment containing nifA/nifB copy I was determined. Two open reading frames coding for a 59653 (NifA) and a 49453 (NifB) dalton protein could be detected. Comparison of the amino acid sequences revealed that the R. capsulatus nifA and nifB gene products are more closely related to the NifA and NifB proteins of Rhizobium meliloti and Rhizobium leguminosarum than to those of K. pneumoniae. A rho-independent termination signal and a typical nif promoter region containing a putative NifA binding site and a consensus nif promoter are located within the region between the R. capsulatus nifA and nifB genes. The nifB sequence is followed by an open reading frame (ORF1) coding for a 27721 dalton protein in nifA/nifB copy I. DNA sequence analysis of nifA/nifB copy II showed that both copies differ in the DNA region downstream of nifB and in the noncoding sequence in front of nifA. All other regions compared, i.e. the 5′ part of nifA, the intergenic region and the 3′ part of nifB, are identical in both copies.

nif gene expression studies in Rhodobacter capsulatus: nfrC‐independent repression by high ammonium concentrations

The expression of nif genes in Rhodobacter capsulatus depends on the two regulatory genes, rpoN and nifA, encoding a nif‐specific alternative sigma factor of RNA polymerase and a nif‐specific transcriptional activator, respectively. The expression of the rpoN gene itself is also RPON/NIFA dependent. In order to better characterize the regulation of nif gene induction, chromosomal nifH‐, rpoN‐, nifAl‐ and nifA2‐ lacZ fusions were constructed and the expression of these different nif‐lacZ fusions was determined under photoheterotrophic conditions at different starting ammonium concentrations. The two nifA genes were found to be induced first, followed by nifH and finally by rpoN upon weak, medium and strong nitrogen starvation, respectively. This induction profile and the correlation between the expression of the different nif genes suggested that nifAl expression is the limiting factor for nif gene induction. This hypothesis was tested by construction of different nifA1 overexpressing mutants. Contrary to the current model of n/Y gene expression in R. capsulatus, which predicted constitutive nif gene expression in such mutants, a strong repression of nifH and rpoW was found at high ammonium concentration. The low nifH expression under these conditions is unaffected by nifA2 and is not increased in a ntrC mutant, ruling out any role of NTRC as a mediator of this repression. This finding implies an additional, so far unidentified, regulation by fixed nitrogen in R capsulatus. Changing the expression level of rpoN indicated that low levels of RPON are already sufficient for full nifH induction. The nifA1 and rpoN expression mutants were also tested for diazotrophic growth. Similar generation times were determined for the mutants and for the wild type, but diazotrophic growth of the nifA1 overexpressing ntrC mutant RCM14 did not start until after a prolonged lag phase of two to three days.

Xanthine dehydrogenase from the phototrophic purple bacterium Rhodobacter capsulatus is more similar to its eukaryotic counterparts than to prokaryotic molybdenum enzymes

Fourteen Rhodobacter capsulatus mutants unable to grow with xanthine as sole nitrogen source were isolated by random Tn5 mutagenesis. Five of these Tn5 insertions were mapped within two adjacent chromosomal EcoRI fragments hybridizing to oligonucleotides synthesized according to conserved amino acid sequences of eukaryotic xanthine dehydrogenases. DNA sequence analysis of this region revealed two open reading frames, designated xdhA and xdhB, encoding xanthine dehydrogenase. The deduced amino acid sequence of XDHA contains binding sites for two [2Fe–2S] clusters and FAD, whereas XDHB is predicted to contain the molybdopterin cofactor. In contrast to R. capsulatus, these three cofactor binding sites reside within a single polypeptide chain in eukaryotic xanthine dehydrogenases. The amino acid sequence of xanthine dehydrogenase from R. capsulatus showed a higher degree of similarity to eukaryotic xanthine dehydrogenases than to the xanthine dehydrogenase‐related aldehyde oxidoreductase from Desulphovibrio gigas. The expression of an xdhA–lacZ fusion was induced when hypoxanthine or xanthine was added as sole nitrogen source. Mutations in nifR1 (ntrC) and nifR4 (rpoN, encoding σ54) had no influence on xdh gene expression. A putative activator sensing the availability of substrate seems to respond to xanthine but not to hypoxanthine. The transcriptional start site of xdhA was mapped by primer extension analysis. Comparison with known promoter elements revealed no significant homology. Xanthine dehydrogenase from R. capsulatus was purified to homogeneity. The enzyme consists of two subunits with molecular masses of 85 kDa and 50 kDa respectively. N‐terminal amino acid sequencing of both subunits confirmed the predicted start codons. The molecular mass of the native enzyme was determined to be 275 kDa, indicating an α2β2‐subunit structure. Analysis of the molybdenum cofactor of xanthine dehydrogenase from R. capsulatus revealed that it contains the molybdopterin cofactor and not a molybdopterin dinucleotide derivative.

Isolation of periplasmic nitrate reductase genes from Rhodobacter sphaeroides DSM 158: structural and functional differences among prokaryotic nitrate reductases

The prototrophic bacterium Rhodobacter sphaeroides DSM 158 has a periplasmic nitrate reductase which is induced by nitrate and it is not repressed by ammonium or oxygen. In a Tn5 mutant lacking nitrate reductase activity, transposon insertion is localized in a 1.2 kb EcoRI fragment. A 0.6 kb BamHI‐EcoRI segment of this region was used as a probe to isolate, from the wild‐type strain, a 6.8 kb Pstl fragment carrying the putative genes coding for the periplasmic nitrate reductase. In vivo protein expression and DNA sequence analysis reveal the presence in this region of three genes, napABC, probably organized in an operon. These genes are required for nitrate reduction, as deduced by mutational and complementation studies. The napA gene codes for a protein with a high homology to the periplasmic nitrate reductase from Alcali‐genes eutrophus and, to a lesser extent, to other prokaryotic nitrate reductases and molybdenum‐containing enzymes. The napB gene product has two haem c‐binding sites and shows a high homology with the cytochrome c‐type subunit of the periplasmic nitrate reductase from A. eutrophus. NAPA and NAPB proteins appear to be translated with signal peptides of 29 and 24 residues, respectively, indicating that mature proteins are located in the periplasm. The napC gene codes for a 25 kDa protein with a transmembrane sequence of 17 hydrophobic residues. NAPC has four haem c‐binding sites and is homologous to the membrane‐bound c‐type cytochromes encoded by Pseudomonas stutzeri nirT and Escherichia coli torC genes. The phenotypes of defined insertion mutants constructed for each gene also indicate that periplasmic nitrate reductase from R. sphaeroides DSM 158 is a dimeric complex of a 90kDa catalytic subunit (NAPA) and a 15kDa cytochrome c (NAPB), which receives electrons from a membrane‐anchored tetrahaem protein (NAPC), thus allowing electron flow between membrane and periplasm. This nitrate‐reducing system differs from the assimilatory and respiratory bacterial nitrate reductases at the level of cellular localization, regulatory properties, biochemical characteristics and gene organization.

Organization and regulation of genes encoding the molybdenum nitrogenase and the alternative nitrogenase in Rhodobacter capsulatus

Abstract The phototrophic non-sulfur purple bacterium Rhodobacter capsulatus is able to fix atmospheric dinitrogen either via a conventional molybdenum nitrogenase or via an alternative iron-only nitrogenase. At least 53 genes are involved in the synthesis and regulation of these two nitrogenase systems, most of which are clustered in four regions widely spread in the genome. Expression of both nitrogenase systems is regulated at the transcriptional level by NifR1 and NifR2, homologues of NtrC and NtrB, respectively. However, this ntr system is only involved in the regulation of the two nitrogenase systems and the high-affinity molybdenum transport system and is not required for utilization of other N sources such as proline and arginine. In contrast to enteric bacteria, the R. capsulatus NtrC homologue does not act in concert with the alternative sigma factor RpoN (σ54). Nitrogen fixation in R. capsulatus is regulated at the transcriptional level and also at the post-translational level. The draTG gene products are responsible for covalent modification of the dinitrogenase reductases of both nitrogenase systems. In addition, mutations in hvrA, a gene previously described as being responsible for low-light activation of the photosynthetic apparatus, also affect regulation of nitrogen fixation.

Yeast two-hybrid studies on interaction of proteins involved in regulation of nitrogen fixation in the phototrophic bacterium Rhodobacter capsulatus.

Rhodobacter capsulatus contains two PII-like proteins, GlnB and GlnK, which play central roles in controlling the synthesis and activity of nitrogenase in response to ammonium availability. Here we used the yeast two-hybrid system to probe interactions between these PII-like proteins and proteins known to be involved in regulating nitrogen fixation. Analysis of defined protein pairs demonstrated the following interactions: GlnB-NtrB, GlnB-NifA1, GlnB-NifA2, GlnB-DraT, GlnK-NifA1, GlnK-NifA2, and GlnK-DraT. These results corroborate earlier genetic data and in addition show that PII-dependent ammonium regulation of nitrogen fixation in R. capsulatus does not require additional proteins, like NifL in Klebsiella pneumoniae. In addition, we found interactions for the protein pairs GlnB-GlnB, GlnB-GlnK, NifA1-NifA1, NifA2-NifA2, and NifA1-NifA2, suggesting that fine tuning of the nitrogen fixation process in R. capsulatus may involve the formation of GlnB-GlnK heterotrimers as well as NifA1-NifA2 heterodimers. In order to identify new proteins that interact with GlnB and GlnK, we constructed an R. capsulatus genomic library for use in yeast two-hybrid studies. Screening of this library identified the ATP-dependent helicase PcrA as a new putative protein that interacts with GlnB and the Ras-like protein Era as a new protein that interacts with GlnK.

Characterization of anf genes specific for the alternative nitrogenase and identification of nif genes required for both nitrogenases in Rhodobacter capsulatus

To identify Rhodobacter capsulatus nif genes necessary for the alternative nitrogenase, strains carrying defined mutations in 32 genes and open reading frames of nif region A, B or C were constructed. The ability of these mutants to grow on nitrogen‐free medium with molybdenum (Nif phenotype) or in a nifHDK deletion background on medium without molybdenum (Anf phenotype) was tested. Nine nif genes and nif‐associated coding regions are absolutely essential for the alternative nitrogenase. These genes comprise nifV and nifB, the nif‐specific ntr system (nifR1, R2, R4) and four open reading frames, which exhibit no homology to known genes. In addition, a significantly reduced activity of both the alternative nitrogenase and the molybdenum‐dependent nitrogenase was found for fdxN mutants. By random Tn5 mutagenesis of a nifHDK deletion strain 42 Anf− mutants were isolated. Southern hybridization experiments demonstrated that 17 of these Tn5 mutants were localized in at least 13 different restriction fragments outside of known nif regions. Ten different Anf− Tn5 mutations are clustered on a 6 kb DNA fragment of the chromosome designated anf region A. DNA sequence analysis revealed that this region contained the structural genes of the alternative nitrogenase (anfHDGK). The identification of several Tn5 insertions mapping outside of anf region A indicated that at least 10 genes specific for the alternative nitrogenase are present in R. capsulatus.