INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

Rhodobacter capsulatus is a nonsulfur phototrophic purple bacterium which can grow with a variety of different nitrogen sources, such as ammonium, almost all amino acids, purines (xanthine and hypoxanthine), urea, polyamines (putrescine and spermidine), and molecular nitrogen. Like in many other bacteria, ammonium is a preferred N source, and consequently, the highly energy-demanding nitrogen fixation process is repressed by ammonium.

R. capsulatus measures the cellular nitrogen status (e.g., availability of ammonium) by an Ntr system similar to that of enterobacteria involving the two-component regulatory system NtrB-NtrC (10, 15, 18). NtrC is the transcriptional activator of a number of genes directly or indirectly involved in nitrogen fixation in R. capsulatus (glnB-glnA, nifA1, nifA2, anfA, and mopA-modABCD [3]). These genes code for the signal transduction protein PII (GlnB) and glutamine synthetase (GlnA), the transcriptional activators of the two nitrogenase systems (nif- and anf-encoded), a molybdenum repressor of the Anf system, and a high-affinity molybdate uptake system. Despite the similarities of the Ntr systems of R. capsulatus and of enterobacteria, mutations in R. capsulatus ntrC do not produce the "classical" Ntr phenotype of enterobacteria, since the NtrC protein in R. capsulatus is not required for the utilization of amino acids as an N source (11, 13, 25).

Under conditions of nitrogen limitation, the NtrB sensor kinase autophosphorylates and transfers the phosphate to the NtrC response regulator. NtrC~P in turn activates transcription of its target genes. Transcriptional activation of these genes requires binding of NtrC~P to enhancer-binding sites distant from the promoters that are activated. Members of the enhancer-binding protein family characteristically activate transcription from promoters recognized by the alternative sigma factor ⁵⁴ (NtrA). The NtrC protein from R. capsulatus (RcNtrC) is a unique enhancer-binding protein that does not require ⁵⁴ but instead activates transcription of the genes mentioned above together with RNA polymerase containing the ⁷⁰-like housekeeping sigma factor (RNAP-⁷⁰ [3, 8]).

Since an R. capsulatus ntrC mutant does not grow with urea as the sole N source (17), genes required for utilization of urea (ure genes) were likely targets for NtrC-mediated activation. Degradation of urea is catalyzed by the enzyme urease, which is an inducible enzyme in R. capsulatus E1F1 (7). Urease is a nickel-containing enzyme that catalyzes the hydrolysis of urea to form ammonia and carbamate, which spontaneously decomposes to produce carbonic acid and additional ammonia. In Klebsiella aerogenes and many other bacteria, the ureA, ureB, and ureC genes encode the catalytically inactive apoenzyme, and the ureD, ureE, ureF, and ureG gene products are required for assembly of the nickel metallocenter (5, 22).

This work describes genetic analyses of the R. capsulatus ure gene region showing that expression of the ure genes is activated under nitrogen-limiting conditions in an NtrC-dependent manner. DNA footprinting studies demonstrate direct binding of NtrC to the ureD promoter. This is the first example of an NtrC-activated target gene in R. capsulatus which is not somehow involved in the process of nitrogen fixation.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Bacterial strains and plasmids. The bacterial strains and plasmids used in this study are listed in Table 1. Methods for conjugational plasmid transfer between Escherichia coli and R. capsulatus and the procedures for selection of mutants, anaerobic growth conditions, and antibiotic concentrations have been previously described (14, 19, 23, 24).

DNA techniques. DNA isolation, restriction enzyme analysis, and cloning procedures were performed according to standard methods (27). Restriction endonucleases, T4 DNA ligase, and Superscript I reverse transcriptase were purchased from MBI Fermentas (St. Leon-Rot, Germany) or Life Technologies (Karlsruhe, Germany) and were used as recommended by the supplier. DNA sequence analysis was done by the chain termination method with fluorescein-labeled primers for use in the A. L. F. DNA sequencer (Amersham Pharmacia Biotech, Freiburg, Germany) according to the manufacturer's instructions.

Construction of an ureDA-lacZ fusion plasmid. A 3.9-kb BamHI-HindIII fragment encompassing the ureD promoter region (Fig. 1A) was inserted into the mobilizable broad-host-range vector pPHU235, resulting in hybrid plasmid pNIRUB35 (Fig. 1C) carrying an in-frame ureDA-lacZ fusion.

View larger version (15K): [in this window] [in a new window]   FIG. 1.   Physical and genetic map of the R. capsulatus ure gene region. Abbreviations: B, BamHI; E, EcoRI, H, HindIII; M, SmaI; S, SalI; X, XhoI. (A) The localizations of genes and open reading frames are given by arrows carrying their respective gene designations. Black arrows emphasize the highly conserved structural and accessory ure genes found in many bacteria (22). Vertical arrows above the genetic map indicate the locations of Tn5 insertions in mutant strains Xan-9, Xan-10, and Xan-19 (17) resulting in an Ure phenotype. Below the map, the locations of interposon insertions are shown. The interposon cassettes (Gm, gentamicin resistance; Km, kanamycin resistance) are not drawn to scale. The directions of transcription of interposon resistance genes are symbolized by arrowheads, indicating polar and nonpolar insertions. (B) The ability of the corresponding R. capsulatus mutant strains to grow with urea as the sole nitrogen source is indicated by + or . (C) Hybrid plasmid pNIRUB35 carrying an in-frame ureDA-lacZ fusion is based on the mobilizable broad-host-range plasmid pPHU235.

-Galactosidase and in vivo nitrogenase assays. To determine the -galactosidase activities of R. capsulatus strains carrying pNIRUB35 (ureDA-lacZ), cultures were grown in RCV minimal medium supplemented with tetracycline (23). For growth under nitrogen-limiting conditions, either serine or urea was added to final concentrations of 9.5 or 4.0 mM, respectively. Nitrogen-sufficient conditions were achieved by addition of 15 mM NH₄Cl to the medium. Following growth in the media to late exponential phase, -galactosidase activities of R. capsulatus strains were determined by the sodium dodecyl sulfate-chloroform method (10, 20).

RNA isolation and primer extension analysis. RNA was prepared from R. capsulatus nifHDK deletion strain KS36 harboring plasmid pNIRUB35 (ureDA-lacZ). For this purpose, cells were grown in RCV minimal medium containing 4 mM urea as the sole nitrogen source under photoheterotrophic conditions until the early exponential growth phase. Cells were harvested and RNA was isolated according to the method described by Chomczynski and Sacchi (4). The primer extension procedure was performed with Superscript I reverse transcriptase (Life Technologies), using a 5' fluorescein-labeled oligonucleotide (5'-TGCGTCCGAATCAAGCCCATAGC-3', corresponding to codons 35 to 42 of the ureD gene). Conditions for primer extension were as described by Myöhänen and Wahlfors (26). Analysis of primer extension products next to a sequencing ladder generated with the same oligonucleotide was performed using the A. L. F. DNA sequencer (Amersham Pharmacia Biotech).

DNase I footprinting. PCR was used to generate a 286-bp fragment spanning the ureD promoter region with oligonucleotide primers ureD.F (5'-CGGGGTCGGGCAGATCGAAG-3') and ureD.R (5'-CACAAGACCCTTCAGACGCG-3') using pNIRUB42-I as a template. In each reaction, one of the primers ureD.F and ureD.R was end labeled with T4 polynucleotide kinase (Epicentre Technologies, Madison, Wis.) and [-³²P]dATP. DNA binding reactions with the unphosphorylated RcNtrC were carried out by the addition of various concentrations of RcNtrC to binding buffer (50 mM Tris-HCl [pH 7.5], 0.1 mM EDTA, 10 mM MgCl₂, 50 mM KCl, 10 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride) which contained either upper- or lower-strand end-labeled probe (approximately 20,000 cpm), either 0 or 1 mM ATP, and 100 ng of poly(dI-dC) as a nonspecific competitor. Complexes were allowed to form for 10 min at 23°C in a total volume of 50 µl, after which DNase I digestion and DNA purification were performed as described previously (8). DNA binding reactions with phosphorylated RcNtrC were performed by incubation of maltose-binding protein-RcNtrB (MBP-RcNtrB; 1 µM) in binding buffer that contained 1 mM ATP as described previously (6). DNase I digestion reactions were analyzed on an 8% sequencing gel.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

Genetic organization of the R. capsulatus ure gene region. Random transposon Tn5 mutagenesis of an R. capsulatus nifHDK mutant strain (deleted for the structural genes of the molybdenum nitrogenase) led to the identification of five mutants (Xan-9, Xan-10, Xan-19, Xan-20, and Xan-22) unable to grow with urea as the sole nitrogen source (17). Tn5-containing DNA fragments from these mutants were cloned, and the sites of transposon insertions were determined by nucleotide sequence analyses. Comparison with the R. capsulatus genome sequence (http://wit.mcs.anl.gov/WIT2/CGI/org.cgi) showed that the Tn5-induced mutations in Xan-9, Xan-10, and Xan-19 mapped within the ureF and ureC genes (Fig. 1A). Due to the close proximity of the ureDABC-orf136-ureEFG genes, it seems likely that the ure genes are cotranscribed in R. capsulatus from a promoter upstream of ureD (see below). The organization of ure genes in R. capsulatus is similar to that of K. aerogenes and many other bacteria (22), except for the presence of an additional open reading frame (orf136) located between ureC and ureE. This orf did not show homology to any known ure gene or any other gene in the databases, indicating that orf136 might be specific for Rhodobacter. The remaining mutants Xan-20 and Xan-22 carried the transposon within the ntrC and ntrB genes, respectively (data not shown).

Mutational analysis of the R. capsulatus urease gene region. As mentioned above, urease catalyzes the hydrolysis of urea to form ammonia and carbamate, which spontaneously decomposes to produce carbonic acid and additional ammonia. Although this process might produce free ammonium within the cell, urea did not prevent nitrogenase synthesis or activity in R. capsulatus at least under the conditions used in this study (Table 2). This is similar to the situation in Azotobacter vinelandii, where urea has been used as a nonrepressing nitrogen source (e.g., references 1 and 12). Since urease contains a nickel cofactor, different amounts of NiCl₂ (Table 2) were added to the R. capsulatus medium to rule out that the nickel concentration was limiting urease activity. In other studies nickel has been added in the micromolar range (e.g., 200 µM for E. coli carrying the K. aerogenes ure gene cluster [5]).

Transcriptional analysis of the R. capsulatus ure operon. Since expression of the ure operon appears to depend on a promoter located directly upstream of ureD, pNIRUB35 carrying an in-frame ureDA-lacZ fusion was constructed (Fig. 1C; Materials and Methods). This broad-host-range reporter plasmid was introduced into the R. capsulatus wild type and selected mutants, and -galactosidase activity was determined in cultures grown with different nitrogen sources. The results shown in Table 3 can be summarized as follows. (i) Maximum expression of the ure genes occurred under nitrogen-limiting conditions (serine or urea as the sole nitrogen source). (ii) Transcription of urease genes was not substrate (urea) inducible in R. capsulatus B10S. As mentioned above, urease is an inducible enzyme in R. capsulatus E1F1 (7). However, the two R. capsulatus strains differ in several aspects (e.g., their susceptibilities to several bacteriophages) from each other, including those concerning the general nitrogen metabolism. R. capsulatus E1F1 is able to grow with nitrate as the sole source of nitrogen, whereas R. capsulatus B10S is devoid of any nitrate reductase activity. Therefore, it seems not unlikely that regulatory circuits also differ from each other in these two R. capsulatus strains. (iii) Expression of the ureDA-lacZ fusion was down-regulated (3.5-fold) under nitrogen-sufficient conditions, but significant expression remains in the presence of ammonium. (iv) NtrC is essential for expression of the ure genes under both nitrogen-limiting and nitrogen-sufficient conditions, suggesting that expression of the ure genes in the presence of ammonium somehow requires NtrC. At present it remains a matter of speculation whether ammonium-grown R. capsulatus cells contain low amounts of phosphorylated NtrC sufficient for partial activation of promoters with high affinity to NtrC~P. As shown by footprint analysis (see below), NtrC~P indeed efficiently binds to the ureD promoter. (v) ⁵⁴ (NtrA or RpoN) is not essential for urease activity or expression of ureD. Expression of ureDA-lacZ was somehow enhanced in an ntrA mutant background, but this was typically less than twofold.

DNase I footprint analysis of the ureD promoter region. Since transcription of ureDA-lacZ was dependent on the ntrC gene, we predicted that RcNtrC might directly contact the ureD promoter upstream of the transcriptional start site. To verify this assumption, DNase I footprinting was performed with unphosphorylated RcNtrC to identify DNA sequences involved in contacting RcNtrC. Labeled DNA fragments containing the ureD upstream region from 227 to +59 were incubated with increasing concentrations of purified RcNtrC and subjected to DNase I digestion. The DNase I protection patterns were localized by comparison with dideoxy DNA sequence ladders generated with the ureD.F primer for the upper strand (Fig. 2A) and the ureD.R primer for the lower strand (Fig. 2B). A DNase I-hypersensitive site could be seen at 148 on the upper strand at an RcNtrC concentration of 10 nM and at 118 and 139 on the lower strand at an RcNtrC concentration of 20 nM. At a higher RcNtrC concentration (160 nM), several minor hypersensitive sites were observed on the upper strand. Full protection of the ureD upstream region was observed between 157 and 115 (upper strand) and between 155 and 110 (lower strand) at concentrations of RcNtrC of 80 nM and higher (Fig. 2A and B). This RcNtrC concentration (80 nM) is similar to that reported for complete protection of the glnB (160 nM), nifA1 (80 nM to 160 nM), and nifA2 (80 nM to 160 nM) promoter regions (6, 8). These regions of DNase I protection by RcNtrC encompass the consensus tandem RcNtrC binding sites described previously (Fig. 2A and B, regions I and II [3]).

View larger version (80K): [in this window] [in a new window]   FIG. 2.   Protection of the ureD promoter region by RcNtrC from DNase I digestion. (A) Upper-strand protection by RcNtrC at the indicated concentrations (nM); (B) lower-strand protection by RcNtrC at the indicated concentrations; (C) upper-strand protection by the indicated concentrations of RcNtrC with 1 µM MBP-RcNtrB and 1 mM ATP; (D) lower-strand protection by the indicated level of RcNtrC with 1 µM MBP-RcNtrB and 1 mM ATP. The concentration of DNA probe was approximately 0.1 nM for each reaction. Shaded bars indicate potential RcNtrC binding sites (see text); numbers refer to the distance from the transcription start site; arrowheads mark areas of increased DNase I sensitivity. Brackets indicate regions of protection at 160 nM RcNtrC (A and B) or 10 nM RcNtrC with 1 µM MBP-RcNtrB and 1 mM ATP (C and D).

View larger version (12K): [in this window] [in a new window]   FIG. 3.   Summary of the R. capsulatus ureD promoter region analysis. Boldfaced letters indicate oligonucleotide primers used to generate the indicated probe for DNase I protection experiments. Arrowheads represent major and minor DNase I-hypersensitive sites, and potential RcNtrC binding sites are indicated by horizontal shaded bars. The horizontal arrows show the direction of transcription, with +1 indicating the ureD transcriptional start site. Putative "10" and "35" regions are boxed. The asterisk marks the translational stop codon for the lrp2 gene.