Logo of jbacterPermissionsJournals.ASM.orgJournalJB ArticleJournal InfoAuthorsReviewers
J Bacteriol. 2005 Nov; 187(22): 7784–7794.
PMCID: PMC1280311

Functional Genomic Analysis of Three Nitrogenase Isozymes in the Photosynthetic Bacterium Rhodopseudomonas palustris


The photosynthetic bacterium Rhodopseudomonas palustris is one of just a few prokaryotes described so far that has vnf and anf genes for alternative vanadium cofactor (V) and iron cofactor (Fe) nitrogenases in addition to nif genes for a molybdenum cofactor (Mo) nitrogenase. Transcriptome data indicated that the 32 genes in the nif gene cluster, but not the anf or vnf genes, were induced in wild-type and Mo nitrogenase-expressing strains grown under nitrogen-fixing conditions in Mo-containing medium. Strains that were unable to express a functional Mo nitrogenase due to mutations in Mo nitrogenase structural genes synthesized functional V and Fe nitrogenases and expressed vnf and anf genes in nitrogen-fixing growth media that contained Mo and V at concentrations far in excess of those that repress alternative nitrogenase gene expression in other bacteria. Thus, not only does R. palustris have multiple enzymatic options for nitrogen fixation, but in contrast to reports on other nitrogen-fixing bacteria, the expression of its alternative nitrogenases is not repressed by transition metals. Between 95 and 295 genes that are not directly associated with nitrogenase synthesis and assembly were induced under nitrogen-fixing conditions, depending on which nitrogenase was being used by R. palustris. Genes for nitrogen acquisition were expressed at particularly high levels during alternative nitrogenase-dependent growth. This suggests that alternative nitrogenase-expressing cells are relatively starved for nitrogen and raises the possibility that fixed nitrogen availability may be the primary signal that controls the synthesis of the V and Fe nitrogenases.

Nitrogenases convert nitrogen gas to ammonia with the concomitant obligate production of hydrogen (8). This difficult reaction requires large amounts of ATP and reductant. Rhodopseudomonas palustris is a purple facultatively photosynthetic bacterium that is an attractive organism to develop as a biocatalyst for hydrogen production by means of nitrogen fixation because it can generate ATP from light and reductant from acetate and green plant-derived aromatic compounds to drive this process (1, 13). It should be possible to configure bioreactors wherein R. palustris cultures illuminated by sunlight degrade agricultural waste and generate hydrogen as a product of nitrogen fixation.

A striking feature of the genome of R. palustris CGA009 is that it has genes predicted to encode three different nitrogenases, each with a different transition metal-containing cofactor at its active site (19) (Fig. (Fig.1).1). Most nitrogen-fixing bacteria encode only a molybdenum nitrogenase (Mo nitrogenase). Some bacteria, including the purple nonsulfur phototrophs Rhodobacter capsulatus and Rhodospirillum rubrum as well as various cyanobacteria and Clostridium pasteurianum, encode, in addition, either an iron nitrogenase (Fe nitrogenase) or a vanadium nitrogenase (V nitrogenase) (5, 23). Alternative nitrogenases have been proposed to serve as a route for nitrogen fixation in situations where molybdenum is limited in the environment. Only the obligately aerobic heterotroph Azotobacter vinelandii has been shown to have three different functional nitrogenases, although genes for all three nitrogenases are also present in Methanosarcina acetovorans (9).

FIG. 1.
Organization of Mo, V, and Fe nitrogenase gene clusters in R. palustris. Gene functions are annotated according to Rubio and Ludden (30).

Mo nitrogenases consist of two components, designated the dinitrogenase reductase (encoded by nifH) and the dinitrogenase or MoFe protein (encoded by nifDK) (8). The MoFe protein has an iron molybdenum cofactor at its active site. Reduced NifH serves as the electron donor to the MoFe protein, which contains the site of nitrogen reduction. The electron transfer reaction from NifH to NifDK is accompanied by the hydrolysis of ATP. Nitrogen is progressively reduced at the active site to produce partially reduced intermediates until, finally, ammonium is formed and released along with hydrogen. Alternative V and Fe nitrogenases are comprised of homologous VnfHDK and AnfHDK subunits. These enzymes also include VnfG and AnfG subunits as additional structural components of the dinitrogenase (23). Many accessory proteins participate in the synthesis of the transition metal cofactors and in the assembly of nitrogenases (Fig. (Fig.1).1). Some of the cofactor synthesis and assembly proteins encoded by the nif gene cluster participate in alternative nitrogenase cofactor synthesis and assembly (30).

Here we present work demonstrating that R. palustris can express three functional nitrogenase isozymes. Wild-type cells expressed nitrogenase isozymes in a hierarchy according to metal availability, with Mo nitrogenase expressed in preference to V and Fe nitrogenases when Mo was present and V nitrogenase expressed in preference to Fe nitrogenase when Mo was absent and V was present. Specific metal-responsive regulators have been shown to repress alternative nitrogenase gene expression in some bacteria (18), and this is the generally accepted mechanism for the differential regulation of nitrogenase isozymes (18, 28). There is one report that is not consistent with this mechanism in which convincing evidence is presented that Mo does not directly repress Fe nitrogenase synthesis in R. rubrum (20). This observation was never further investigated, though. We present evidence here that R. palustris resembles R. rubrum in that it synthesizes its alternative nitrogenases in the presence of Mo in situations where it is unable to express a functional Mo nitrogenase. We hypothesize based on transcriptome data that alternative nitrogenase gene expression occurs in R. palustris in response to increasing levels of fixed nitrogen starvation rather than in direct response to transition metals.


Bacterial strains and growth conditions.

The bacterial strains and plasmids used for this study are listed in Table Table1.1. R. palustris strains were grown anaerobically in light in defined mineral medium containing ammonium sulfate and 10 mM succinate at 30°C (non-nitrogen-fixing medium) (17). Nitrogen-fixing medium was the same medium without added ammonium sulfate. Cells were grown in sealed tubes with a nitrogen gas headspace. Mo-depleted growth medium was prepared by treating 10-fold-concentrated nitrogen-fixing medium with 8-hydroxyquinoline (8-HQ) at pH 3.5, followed by extraction with dichloromethane (33). The pH was then raised to 7.2 with a solution of ultrapure NaOH (Fluka MicroSelect). The trace element solution used in the growth medium was prepared without Mo salt but was not treated with 8-HQ. All glassware was soaked in a solution of 1% Count-OFF (New England Nuclear) and 10 mM EDTA for 24 h and then washed with deionized H2O (Millipore system) (34). The concentrations of Mo and V in growth media were determined by inductively coupled plasma mass spectrometry (ICP-MS; Varian UltraMass 700) at the University of Iowa Hygienic Laboratory. Mo was added to 8-HQ-treated medium as Na2MoO4, and V was added as VCl3.

Bacterial strains, plasmids, and primers used for this study

Nitrogenase activity and hydrogen production measurement.

Nitrogenase activity was measured by the acetylene reduction assay (2). Assays were carried out in sealed tubes that had been flushed with argon gas and autoclaved. Cultures in the mid-logarithmic phase of growth and 2% acetylene (final concentration) were injected into the tubes and incubated in light at 30°C. Gas-phase samples (100 μl) were withdrawn with a Hamilton sample lock syringe at intervals, and ethylene and ethane were measured with a Hewlett Packard model 5890 series II gas chromatograph fitted with a flame ionization detector and a Porapak N column (80/100 mesh, 1/8 in. by 6 ft). The temperatures of the injector, detector, and oven were 50°C, 150°C, and 40°C, respectively. Nitrogen gas was the carrier, at a flow rate of 40 ml/min. Hydrogen was measured with a thermal conductivity detector and a Molecular Sieve-13X column (80/100 mesh, 1/4 in. by 8 ft). The temperatures of the injector, detector, and oven were 100°C, 100°C, and 50°C, respectively. Nitrogen gas was the carrier, at a flow rate of 40 ml/min.

Construction of R. palustris mutant strains and transcriptional fusions.

In-frame deletions of the nifH, vnfH, anfH, and anfA genes were generated by overlap extension PCR (14) as described previously (12). PCR primers and recombinant plasmids are described in Table Table1.1. Broad-host-range plasmids were mobilized from Escherichia coli S17-1 into R. palustris CGA009 by conjugation, and double recombinant strains were selected as previously described (6). Sucrose-resistant and gentamicin-sensitive colonies were screened by colony PCR and sequencing to validate the expected chromosomal in-frame deletion mutations. A nifD::Tn5 mutant was identified from an R. palustris mini-Tn5-lacZ1 mutant library (11) that had been screened for strains defective in nitrogen fixation. This mutant was used as a starting strain for the construction of a ΔnifH nifD::Tn5 ΔanfA strain. Plasmids containing promoter-lacZ fusions were constructed as previously described (7). The primers used to amplify promoter fragments are given in Table Table1.1. β-Galactosidase activities were measured as described previously (7).

Whole-genome microarray construction.

An R. palustris whole-genome microarray was constructed by using the same procedures as those for a Shewanella oneidensis whole-genome microarray (10). PCR primers for 4,508 of 4,836 predicted open reading frames (ORFs) of the R. palustris chromosome were designed using PRIMEGENS software (36). Each gene was amplified 16 times in parallel in 100-μl reaction mixtures in 96-well plates. All of the amplified products (16 by 100 μl) were pooled and purified using a Biomek F/X automated workstation (Beckman). All of the amplified products were analyzed in 1.5% agarose gels and were considered correct if the PCRs amplified single products of the expected sizes. Of the 4,508 total predicted genes, 4,350 ORFs were correctly amplified, which represents approximately 90% of the genome. Specific 50-mer oligonucleotide probes (385 ORFs) were synthesized for the genes that were not amplified. In total, the PCR amplicons and oligonucleotide probes represented 98% of the total predicted gene content of R. palustris. The purified PCR products were diluted in 50% dimethyl sulfoxide to a minimum concentration of 50 ng/μl. Probes were then printed onto Corning UltraGAPS coated slides (Corning, NY) with a BioRobotics Microgrid II printer (Genomic Solutions, Ann Arbor, MI) as recommended by the manufacturer and were UV cross-linked to the surfaces of slides (Stratagene UV cross-linker; Stratagene, La Jolla, CA). Each slide was printed with probes representing two copies of the genome.

RNA isolation.

For RNA isolation, R. palustris strains were subcultured at least twice after initial inoculation from a plate. Cells were grown to the mid-logarithmic phase of growth in mineral medium containing ammonium sulfate or in nitrogen-fixing medium, chilled in an ice-water bath, and harvested by centrifugation. The cells were then frozen in liquid nitrogen and stored at −80°C for RNA isolation at a later time. Thawed cells were disrupted by bead beating, and RNAs were purified with RNeasy mini kits (QIAGEN), including DNase treatment on the columns, as described previously (12). The quality of RNA (integrity and DNA contamination) was determined with an Agilent 2100 bioanalyzer (Agilent, Palo Alto, CA) and by reverse transcription (RT)-PCR using the 16S rRNA-targeted primer set of 27F (Escherichia coli positions 8 to 27) and 519R (positions 536 to 519) (25).

Microarray experiments and data analysis.

Fluorescently labeled cDNAs were prepared by the direct incorporation of either Cy3-dCTP or Cy5-dCTP (Amersham Biosciences) during first-strand RT reactions. The labeling efficiency was calculated by measuring the absorbance at 260 nm, 550 nm (Cy3 incorporation), and 650 nm (Cy5 incorporation). Prior to performing hybridizations, the array slides were incubated at 65°C for 45 min in prehybridization buffer (3× SSC [1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate], 0.3% sodium dodecyl sulfate [SDS], and 1% bovine serum albumin). The two labeled cDNA samples to be compared were mixed, and 20× SSC (final concentration, 3×), 10 mg/ml salmon sperm DNA (Invitrogen; final concentration, 0.8 mg/ml), and 5% SDS (final concentration, 0.3%) were added to the sample. This hybridization mixture was divided into two tubes and incubated at 95°C for 3 min, and the mixture in each tube was applied to prehybridized slides that had been covered with Lifterslips (Erie Scientific Company, Portsmouth, NH). The slides were assembled with hybridization chambers (Corning, Corning, NY) and submerged in a 65°C water bath. After 14 to 16 h of hybridization, the slides were washed with 1× SSC and 0.2% SDS for 5 min, 0.1× SSC and 0.2% SDS for 5 min, and 0.1× SSC for 5 min. The slides were then dried by centrifugation (200 × g for 5 min) and scanned with a ScanArray 4000XL scanner (Perkin-Elmer, Boston, MA). Images (Cy3 and Cy5) were captured as TIFF files and were analyzed with the image processing software ImaGene, version 5.6 (BioDiscovery, Inc., El Segundo, CA). The median intensity of each spot was used for further analysis. Intensity values for portions of array slides that had not been printed with DNA were used as negative controls. The average plus 2 standard deviations of the intensity values of the negative control was calculated for each slide and used as the threshold for data filtering (26).

The software package lcDNA was used for data normalization and assessments of the statistical confidence intervals of gene expression (15, 35). Duplicate calibration experiments and three comparative experiments using RNAs from three separately grown cultures (three biological replicates), with duplicate slides for each (10 slides in total), were used to generate each data set. Genes whose ratios (nitrogen-fixing conditions/non-nitrogen-fixing conditions) were ≥2 and whose scores were <0.025 (26) were considered expressed at higher levels under nitrogen-fixing conditions. Genes whose ratios were ≤0.5 and whose scores were >0.975 (26) were considered expressed at lower levels under nitrogen-fixing conditions. The microarray data have been deposited at http://www.ncbi.nlm.nih.gov/geo under accession number GSE3030.


R. palustris encodes three functional nitrogenase isozymes.

Nitrogen fixation was tested by measuring growth with nitrogen gas as a sole nitrogen source and by assaying acetylene reduction. A diagnostic feature of V and Fe nitrogenases is that they catalyze the production of ethane (C2H6) along with ethylene (C2H4) as a product of acetylene reduction, whereas Mo nitrogenases reduce acetylene exclusively to ethylene (23). R. palustris wild-type cells grow well with nitrogen gas as a sole nitrogen source, and we concluded that this is mediated by a Mo nitrogenase because ethylene was generated as the sole product in acetylene reduction assays. To determine whether the gene clusters annotated as encoding V and Fe nitrogenases encode functional nitrogenase isozymes and to characterize the Mo nitrogenase, we constructed strains mutated in various combinations of the nif, vnf, and anf genes. ΔvnfH ΔanfH, ΔnifH nifD::Tn5 ΔanfA, and ΔnifH ΔvnfH mutants grew with nitrogen gas as a sole source of nitrogen, and each reduced acetylene to ethylene. AnfA is predicted to encode a regulator necessary for the expression of anf genes. A triple ΔnifH ΔvnfH ΔanfA mutant did not grow under nitrogen-fixing conditions (Table (Table22).

Growth rate, nitrogenase activity, and hydrogen production from various R. palustris strains in the presence or absence of vanadium in nitrogen-fixing mediumd

The ΔvnfH ΔanfH strain had the characteristics of a Mo nitrogenase-expressing strain. It had relatively high rates of acetylene reduction and failed to reduce acetylene to ethane. Its growth rate matched that of wild-type cells (Table (Table2).2). The standard growth medium that we used to cultivate R. palustris included about 15 nM Mo, as determined by ICP-MS. The ΔvnfH ΔanfH strain failed to grow in medium that had been extracted with 8-HQ to remove Mo to undetectable levels (<0.1 ppb). The addition of Mo to the 8-HQ-extracted medium to a final concentration of 15 nM restored growth and nitrogenase activity. The ΔnifH nifD::Tn5 ΔanfA strain had the expected characteristics of a V nitrogenase-expressing strain. Our standard nitrogen-fixing growth medium, which lacks added V and has an undetectable level of V (<0.1 ppb), as measured by ICP-MS, failed to support the growth of this strain. However, the addition of exogenous V to a final concentration of 10 μM allowed the ΔnifH nifD::Tn5 ΔanfA strain to grow and to reduce acetylene to ethane and ethylene at ratios of about 0.015 ethane to 1 ethylene. A ΔnifH ΔvnfH strain had the phenotype of an Fe nitrogenase-expressing strain. It had low rates of acetylene reduction and produced ethane and ethylene at a ratio of about 0.055 (Table (Table22).

We characterized a ΔnifH nifD::Tn5 strain and concluded that it could express both alternative nitrogenase isozymes depending on V availability. It had the characteristics of an Fe nitrogenase-expressing strain when grown in medium lacking V and of a V nitrogenase-expressing strain when grown in medium supplemented with V (Table (Table2).2). Similarly, we found that wild-type cells grown in metal-depleted (8-HQ-extracted) medium expressed Mo nitrogenase when the medium was supplemented with Mo, V nitrogenase when the medium was supplemented with V, and Fe nitrogenase when the medium remained unsupplemented with metals (data not shown).

V and Fe nitrogenases from other bacteria produce relatively more hydrogen and less ammonia than the traditional Mo nitrogenases that are synthesized by all nitrogen-fixing bacteria (5, 23). R. palustris strains expressing each of the nitrogenase isozymes produced hydrogen. The V and Fe nitrogenases catalyzed the production of twofold and fourfold more hydrogen per unit of biomass formed during growth, respectively, than the Mo nitrogenase (Table (Table2).2). R. palustris strain CGA009 has a frameshift in the hupV gene, encoding a regulatory protein necessary for the expression of a functional uptake hydrogenase (19). This renders strain CGA009 unable to recycle hydrogen produced by means of its nitrogenases.

Transcriptome profiles of strains expressing a single nitrogenase isozyme.

Transcriptome profiles of R. palustris strains grown in nitrogen-fixing medium were compared with those of strains grown in the same medium supplemented with ammonium sulfate, a condition that blocks nitrogenase activity in R. palustris, as evidenced by the observation that cells do not produce hydrogen or reduce acetylene when grown in the presence of ammonium. As expected based on our physiological studies, wild-type cells expressed genes in the nif cluster (RPA4602 to RPA4633) at as much as 300-fold higher levels under nitrogen-fixing conditions. Genes in the vnf and anf clusters were expressed at only very low levels or not at all (Table (Table3).3). This confirms that the wild-type strain synthesizes only its Mo nitrogenase when grown in our standard nitrogen-fixing growth medium. The gene expression profile of wild-type cells grown in nitrogen-fixing medium supplemented with 10 μM V was similar to that of wild-type cells grown in the same medium without added V (Table (Table3).3). The ΔvnfH ΔanfH (Mo nitrogenase-expressing) strain had a gene expression profile similar to that of wild-type cells (Table (Table33).

Expression of the three nitrogenase gene clusters in different nitrogenase-expressing strains

Most of the nif genes and all of the vnf and anf genes were expressed at a high level in the ΔnifH ΔvnfH (Fe nitrogenase-expressing) strain grown under nitrogen-fixing conditions, even though this strain was unable to synthesize functional Mo or V nitrogenase due to nifH and vnfH structural gene mutations. Also, most nif genes were expressed in the ΔnifH nifD::Tn5 ΔanfA strain, in spite of the fact that this strain synthesized only an active V nitrogenase. Several genes (RPA4611 to RPA4618) situated downstream of nifD were not expressed by the ΔnifH nifD::Tn5 ΔanfA strain. This most likely reflects polar effects of the mini-Tn5 insertion in nifD on other genes in its transcriptional unit. The observation that the ΔnifH nifD::Tn5 ΔanfA strain expressed the anf genes at only very low levels is consistent with the prediction that AnfA controls anf gene expression. Plasmids carrying transcriptional fusions of lacZ to the promoter regions of two different nif genes (fer-1 and fixA) and one vnf gene (RPA1377) were constructed and moved into several R. palustris strains. The relative levels of expression of these genes, as determined by their β-galactosidase activities, in cells that were grown with nitrogen gas or with ammonium as a nitrogen source were in general agreement with our transcriptome data (Table (Table44).

Expression of promoter-lacZ fusion plasmids under nitrogen-fixing or non-nitrogen-fixing conditionsa

The transition metals Mo and V do not repress the synthesis of alternative nitrogenases in R. palustris.

In the experiments described above, R. palustris expressed active Fe and V nitrogenases when grown in nitrogen-fixing medium that included about 15 nM molybdate salt. This was surprising because Mo has been reported to repress the synthesis of alternative nitrogenases. Mo at concentrations of >10 nM, for example, prevented the diazotrophic growth of an R. capsulatus nifHDK mutant and the expression of its Fe nitrogenase activity (31). We thought it possible that the relatively small amount of Mo present in our nitrogen-fixing growth medium might only partially block the expression and activity of the Fe and V nitrogenases. However, the addition of Mo to concentrations as high as 100 μM did not substantially affect the growth rates of or rates of acetylene reduction by the Fe and V nitrogenase-expressing strains (Table (Table55 and data not shown). Also, the addition of V at a concentration of 100 μM did not block Fe nitrogenase expression, in contrast to what has been observed in A. vinelandii (16; data not shown).

Effect of metal ion addition on vanadium nitrogenase synthesis and activitya

We grew R. palustris in medium that had been extracted with 8-HQ to check whether the removal of Mo might derepress Fe nitrogenase or V nitrogenase activity further from what was detected in nitrogen-fixing medium. The 8-HQ-treated medium contained undetectable amounts of Mo and V (<0.1 ppb for both Mo and V), and the Fe nitrogenase-expressing strain reduced acetylene at about the same rate as that when grown in untreated medium (data not shown). Further addition of Mo or V to the metal-depleted medium did not result in reduced rates of acetylene reduction by this strain. In all these experiments, the cells produced relatively high levels of ethane, indicative of an active Fe nitrogenase. Similar results were obtained with the V nitrogenase-expressing strain. The removal of all Mo did not substantially affect the rates of acetylene reduction, and the addition of Mo to concentrations as high as 100 μM did not depress the rates of acetylene reduction (Table (Table55).

Many nitrogen acquisition genes are coordinately regulated with the anf and vnf gene sets.

Our findings show that metal ion availability per se does not control the expression of vnf and anf genes in R. palustris. Instead, this species synthesizes V and Fe nitrogenases in situations where it is unable to synthesize a functional Mo nitrogenase regardless of the amount of Mo that is present. The three nitrogenases might be regulated such that they are each synthesized at different thresholds of fixed nitrogen or redox availability rather than by direct metal ion repression. We examined our transcriptome data to further investigate this possibility. About twice as many genes were expressed at twofold or higher levels in the V nitrogenase (324 genes)- and Fe nitrogenase (369 genes)-expressing strains than in the wild-type (164 genes) and Mo nitrogenase-expressing (185 genes) strains under nitrogen-fixing conditions (see Table S1 in the supplemental material). Among these were about 20 genes that are located immediately adjacent to the vnf genes on the R. palustris chromosome and 4 genes that are located immediately adjacent to the anf genes (Table (Table6).6). Genes or groups of genes unlinked to nitrogenase gene clusters that were expressed at fivefold or higher levels in one or more strains are listed in Table Table7.7. The gene or operon functions can be loosely grouped into the following four categories related to nitrogen fixation: oxygen stress, reductant supply, iron acquisition, and nitrogen acquisition. Hypothetical and conserved hypothetical genes are also listed. Of these categories, a significant proportion of genes involved in nitrogen acquisition were expressed at higher levels in V or Fe nitrogenase-expressing strains than in Mo nitrogenase-expressing strains (Table (Table7).7). One explanation for this is that cells express genes involved in fixed nitrogen acquisition at higher levels to try to compensate for the fixed nitrogen limitation that occurs during V or Fe nitrogenase-dependent growth.

Expression of genes adjacent to vanadium and iron nitrogenase gene clusters
Genes or groups of genes unlinked to nitrogenase gene clusters that were expressed at fivefold or higher levels in one or more strains under nitrogen-fixing conditionsb


The R. palustris genome sequence has enabled a comprehensive analysis of gene expression during nitrogen fixation by a single bacterial species expressing three nitrogenase isozymes. In this study, transcriptome data were valuable for defining the full extent of the nif, vnf, and anf gene clusters and for identifying genes that are coordinately expressed with a particular nitrogenase. As has been observed in other bacteria, metal ion availability dictated the type of nitrogenase isozyme expressed by wild-type R. palustris. In contrast to what has been concluded for many other bacteria, however, our results indicate that the repression of gene expression by Mo or V is not the mechanism that underlies the hierarchical expression of nitrogenases in R. palustris. We reasoned that transcriptome data might point to physiological signals governing the differential expression of the three nitrogenase isozymes. A surprisingly large number of genes that are not obviously involved in nitrogenase synthesis were expressed under nitrogen-fixing conditions. Some of these genes are located directly adjacent to the anf or vnf gene cluster (Table (Table6).6). The anf-associated genes RPA1431, RPA1433, and RPA1434 are homologous to genes located adjacent to the R. rubrum anf gene cluster (https://maple.lsd.ornl.gov/microbial/rrub/). RPA1431 encodes a dinitrogen reductase ADP-ribosyltransferase (DRAT) predicted to be involved in inactivating dinitrogenase reductase by ADP-ribosylation in response to darkness or ammonia (21). R. palustris and R. rubrum each have a second DRAT gene that is paired with a dinitrogenase reductase-activating glycohydrolase gene elsewhere on their chromosomes. The vnf-associated genes RPA1381 to -1386 are homologous to genes that are located adjacent to the A. vinelandii vnf gene cluster (https://maple.lsd.ornl.gov/microbial/avin/). Among these is a set of ABC transport genes that likely encode a vanadate permease. The other highly expressed vnf-associated genes have no obvious functions. Two exceptions are genes for a glutamine amidotransferase (RPA1400) and a glutamine synthetase (glnAIII; RPA1401) that were expressed at about 10-fold higher levels under nitrogen-fixing conditions in the V and Fe nitrogenase-expressing strains but were not differentially expressed in Mo nitrogenase-expressing strains.

A large number of genes or groups of genes unlinked to nitrogenase gene clusters were also expressed at elevated levels under nitrogen-fixing conditions, and we grouped these into four physiological categories (Table (Table7).7). One of the categories is oxygen stress. Nitrogenases have metal ion clusters that render these enzymes extremely sensitive to oxygen and reactive oxygen species. This may explain why we saw elevated expression in all strains of genes encoding cytochrome aa3 oxidase and catalase, proteins that consume oxygen or detoxify toxic oxygen species. We also cannot exclude the possibility that cells were exposed to small amounts of oxygen during harvesting and breakage and that as a consequence of their high metal content, nitrogenase-containing cell extracts generated reactive oxygen species that in turn led to the elevated expression of oxygen stress-related genes. All nitrogenases require Fe, and several Fe acquisition genes were expressed at equivalent elevated levels in strains expressing each type of nitrogenase isozyme. Since nitrogenase activity demands large amounts of reductant, it is also logical that ferredoxin and flavodoxin genes (RPA1927 and -1928 and RPA2116 and -2117) were expressed at high levels. The Fe nitrogenase is the most reductant demanding of the nitrogenases. The observation that the NAD-dependent formate dehydrogenase genes (RPA0732 to RPA0736) were expressed at relatively high levels in the Fe nitrogenase-expressing strain may reflect the fact that this strain is seeking more reducing power from various sources. Many genes involved in nitrogen acquisition were expressed at higher levels in cells growing under nitrogen-fixing conditions. These included a glutamine synthetase encoded by RPA0984 and genes for a number of transport systems for nitrogenous compounds, including transporters for dipeptides (RPA1471 to -1473), oligopeptides (RPA0758 to -0762), nitrate (RPA3201), urea (RPA2409 to -2410), and amides (RPA2497 to -2500). Genes for enzymes that function to access fixed nitrogen, including amidase and cyanate lyase, were also expressed at elevated levels. Of the gene categories depicted in Table Table7,7, only those involved in fixed nitrogen acquisition were more highly expressed in both V nitrogenase- and Fe nitrogenase-expressing strains than in Mo nitrogenase-expressing strains. These data and growth rate data (Table (Table2)2) suggest that V and Fe nitrogenase-expressing cells are starved for nitrogen compared to Mo nitrogenase-expressing cells, and this suggests that fixed nitrogen availability may be a key physiological signal that mediates the differential expression of the nitrogenase isozymes.

R. palustris encodes a set of nitrogen signal transduction proteins that overlap with those known to control nitrogen fixation in other purple nonsulfur bacteria, nitrogen-fixing symbionts, and A. vinelandii in response to the intracellular nitrogen status (4, 22, 24, 38). It is possible that these proteins control transcription of the regulatory genes nifA, vnfA, and anfA in a hierarchy in response to various levels of fixed nitrogen starvation. Transcription of the anfA and vnfA genes does seem to be an important point of control for V and Fe nitrogenase expression considering that when cells were shifted to nitrogen-fixing conditions, vnfA was expressed at 78-fold higher levels in the V nitrogenase-expressing strain and anfA was expressed at 69-fold-higher levels in the Fe nitrogenase-expressing strain. R. palustris regulator and signal transduction proteins that could be encoded by one or more of the hypothetical genes that were induced under nitrogen-fixing conditions (Table (Table7)7) may also control anf and vnf gene expression. The intracellular redox status, oxygen availability, and intracellular energy status are other signals that might influence alternative nitrogenase gene expression.

A thorough understanding of the regulation of nitrogenase expression and activity will be helpful in the design of strategies to metabolically engineer R. palustris to produce hydrogen. It has been speculated that nitrogenases are essentially hydrogenase enzymes that have been modified by evolutionary pressures to reduce the triple bond of nitrogen as well as to reduce protons to hydrogen gas. The alternative V and Fe nitrogenases have good potential as catalysts for hydrogen production because they produce relatively more hydrogen and less ammonia than the traditional Mo nitrogenases that are synthesized by all nitrogen-fixing bacteria. Although the R. palustris alternative nitrogenases were less active than Mo nitrogenase, and accordingly produced hydrogen at a lower rate, these enzymes may nevertheless be superior hydrogen-producing catalysts in some situations.

Supplementary Material

[Supplemental material]


This research was supported by the U.S. Army Research Office (grant W911NF-05-1-0176) and the U.S. Department of Energy under the Microbial Genome Program (grant no. DE-FG02-05ER64063) of the Office of Biological and Environmental Research, Office of Science. The Oak Ridge National Laboratory is managed by University of Tennessee-Battelle LLC for the Department of Energy under contract DE-AC05-00OR22725.

We thank Jim Liao for his help with microarray data analysis.


Supplemental material for this article may be found at http://jb.asm.org/.


1. Barbosa, M. J., J. M. Rocha, J. Tramper, and R. H. Wijffels. 2001. Acetate as a carbon source for hydrogen production by photosynthetic bacteria. J. Biotechnol. 85:25-33. [PubMed]
2. Capone, D. G. 1993. Determination of nitrogenase activity in aquatic sediments using the acetylene reduction procedure, p. 621-631. In P. F. Kemp, B. F. Sherr, E. B. Sherr, and J. J. Cole (ed.), Handbook of methods in aquatic microbial ecology. Lewis Publishers, Boca Raton, Fla.
3. de Lorenzo, V., M. Herrero, U. Jakubzik, and K. N. Timmis. 1990. Mini-Tn5 transposon derivatives for insertion mutagenesis, promoter probing, and chromosomal insertion of cloned DNA in gram-negative eubacteria. J. Bacteriol. 172:6568-6572. [PMC free article] [PubMed]
4. Dixon, R., and D. Kahn. 2004. Genetic regulation of biological nitrogen fixation. Nat. Rev. Microbiol. 2:621-631. [PubMed]
5. Eady, R. R. 1996. Structure-function relationships of alternative nitrogenases. Chem. Rev. 96:3013-3030. [PubMed]
6. Egland, P. G., J. Gibson, and C. S. Harwood. 1995. Benzoate-coenzyme A ligase, encoded by badA, is one of three ligases able to catalyze benzoyl-coenzyme A formation during anaerobic growth of Rhodopseudomonas palustris on benzoate. J. Bacteriol. 177:6545-6551. [PMC free article] [PubMed]
7. Egland, P. G., and C. S. Harwood. 2000. HbaR, a 4-hydroxybenzoate sensor and FNR-CRP superfamily member, regulates anaerobic 4-hydroxybenzoate degradation by Rhodopseudomonas palustris. J. Bacteriol. 182:100-106. [PMC free article] [PubMed]
8. Fisher, K., and W. E. Newton. 2002. Nitrogen fixation—a general overview, p. 1-34. In G. J. Leigh (ed.), Nitrogen fixation in the millennium. Elsevier, Amsterdam, The Netherlands.
9. Galagan, J. E., C. Nusbaum, A. Roy, M. G. Endrizzi, P. Macdonald, W. FitzHugh, S. Calvo, R. Engels, S. Smirnov, D. Atnoor, A. Brown, N. Allen, J. Naylor, N. Stange-Thomann, K. DeArellano, R. Johnson, L. Linton, P. McEwan, K. McKernan, J. Talamas, A. Tirrell, W. Ye, A. Zimmer, R. D. Barber, I. Cann, D. E. Graham, D. A. Grahame, A. M. Guss, R. Hedderich, C. Ingram-Smith, H. C. Kuettner, J. A. Krzycki, J. A. Leigh, W. Li, J. Liu, B. Mukhopadhyay, J. N. Reeve, K. Smith, T. A. Springer, L. A. Umayam, O. White, R. H. White, E. Conway de Macario, J. G. Ferry, K. F. Jarrell, H. Jing, A. J. Macario, I. Paulsen, M. Pritchett, K. R. Sowers, R. V. Swanson, S. H. Zinder, E. Lander, W. W. Metcalf, and B. Birren. 2002. The genome of M. acetivorans reveals extensive metabolic and physiological diversity. Genome Res. 12:532-542. [PMC free article] [PubMed]
10. Gao, H., Y. Wang, X. Liu, T. Yan, L. Wu, E. Alm, A. Arkin, D. K. Thompson, and J. Zhou. 2004. Global transcriptome analysis of the heat shock response of Shewanella oneidensis. J. Bacteriol. 186:7796-7803. [PMC free article] [PubMed]
11. Harrison, F. H. 2005. Peripheral pathways of anaerobic benzoate degradation in Rhodopseudomonas palustris. Ph.D. thesis. University of Iowa, Iowa City.
12. Harrison, F. H., and C. S. Harwood. 2005. The pimFABCDE operon from Rhodopseudomonas palustris mediates dicarboxylic acid degradation and participates in anaerobic benzoate degradation. Microbiology 151:727-736. [PubMed]
13. Harwood, C. S., and J. Gibson. 1988. Anaerobic and aerobic metabolism of diverse aromatic compounds by the photosynthetic bacterium Rhodopseudomonas palustris. Appl. Environ. Microbiol. 54:712-717. [PMC free article] [PubMed]
14. Horton, R. M., S. N. Ho, J. K. Pullen, H. D. Hunt, Z. Cai, and L. R. Pease. 1993. Gene splicing by overlap extension. Methods Enzymol. 217:270-279. [PubMed]
15. Hyduke, D. R., L. Rohlin, K. C. Kao, and J. C. Liao. 2003. A software package for cDNA microarray data normalization and assessing confidence intervals. Omics 7:227-234. [PubMed]
16. Joerger, R. D., E. D. Wolfinger, and P. E. Bishop. 1991. The gene encoding dinitrogenase reductase 2 is required for expression of the second alternative nitrogenase from Azotobacter vinelandii. J. Bacteriol. 173:4440-4446. [PMC free article] [PubMed]
17. Kim, M.-K., and C. S. Harwood. 1991. Regulation of benzoate-CoA ligase in Rhodopseudomonas palustris. FEMS Microbiol. Lett. 83:199-204.
18. Kutsche, M., S. Leimkuhler, S. Angermuller, and W. Klipp. 1996. Promoters controlling expression of the alternative nitrogenase and the molybdenum uptake system in Rhodobacter capsulatus are activated by NtrC, independent of sigma54, and repressed by molybdenum. J. Bacteriol. 178:2010-2017. [PMC free article] [PubMed]
19. Larimer, F. W., P. Chain, L. Hauser, J. Lamerdin, S. Malfatti, L. Do, M. L. Land, D. A. Pelletier, J. T. Beatty, A. S. Lang, F. R. Tabita, J. L. Gibson, T. E. Hanson, C. Bobst, J. L. Torres, C. Peres, F. H. Harrison, J. Gibson, and C. S. Harwood. 2004. Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nat. Biotechnol. 22:55-61. [PubMed]
20. Lehman, L. J., and G. P. Roberts. 1991. Identification of an alternative nitrogenase system in Rhodospirillum rubrum. J. Bacteriol. 173:5705-5711. [PMC free article] [PubMed]
21. Ludden, P. W., and G. P. Roberts. 1995. The biochemistry and genetics of nitrogen fixation by photosynthetic bacteria, p. 929-947. In M. T. Madigan, C. E. Bauer, and R. E. Blankenship (ed.), Anoxygenic photosynthetic bacteria. Kluwer Academic Publishers, Dordrecht, The Netherlands.
22. Masepohl, B., T. Drepper, A. Paschen, S. Gross, A. Pawlowski, K. Raabe, K. U. Riedel, and W. Klipp. 2002. Regulation of nitrogen fixation in the phototrophic purple bacterium Rhodobacter capsulatus. J. Mol. Microbiol. Biotechnol. 4:243-248. [PubMed]
23. Masepohl, B., K. Schneider, T. Drepper, A. Müller, and W. Klipp. 2002. Alternative nitrogenases, p. 191-222. In G. J. Leigh (ed.), Nitrogen fixation at the millennium. Elsevier, Amsterdam, The Netherlands.
24. Ninfa, A. J., and P. Jiang. 2005. PII signal transduction proteins: sensors of alpha-ketoglutarate that regulate nitrogen metabolism. Curr. Opin. Microbiol. 8:168-173. [PubMed]
25. Oda, Y., W. G. Meijer, J. L. Gibson, J. C. Gottschal, and L. J. Forney. 2004. Analysis of diversity among 3-chlorobenzoate-degrading strains of Rhodopseudomonas palustris. Microb. Ecol. 47:68-79. [PubMed]
26. Oh, M. K., L. Rohlin, K. C. Kao, and J. C. Liao. 2002. Global expression profiling of acetate-grown Escherichia coli. J. Biol. Chem. 277:13175-13183. [PubMed]
27. Parales, R. E., and C. S. Harwood. 1993. Construction and use of a new broad-host-range lacZ transcriptional fusion vector, pHRP309, for gram-negative bacteria. Gene 133:23-30. [PubMed]
28. Premakumar, R., R. N. Pau, L. A. Mitchenall, M. Easo, and P. E. Bishop. 1998. Regulation of the transcriptional activators AnfA and VnfA by metals and ammonium in Azotobacter vinelandii. FEMS Microbiol. Lett. 164:63-68. [PubMed]
29. Quandt, J., and M. F. Hynes. 1993. Versatile suicide vectors which allow direct selection for gene replacement in gram-negative bacteria. Gene 127:15-21. [PubMed]
30. Rubio, L. M., and P. W. Ludden. 2005. Maturation of nitrogenase: a biochemical puzzle. J. Bacteriol. 187:405-414. [PMC free article] [PubMed]
31. Schneider, K., A. Muller, U. Schramm, and W. Klipp. 1991. Demonstration of a molybdenum- and vanadium-independent nitrogenase in a nifHDK-deletion mutant of Rhodobacter capsulatus. Eur. J. Biochem. 195:653-661. [PubMed]
32. Simon, R., U. Priefer, and A. Pühler. 1983. A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in gram-negative bacteria. Bio/Technology 1:784-791.
33. Strandberg, G. W., and P. W. Wilson. 1968. Formation of the nitrogen-fixing enzyme system in Azotobacter vinelandii. Can. J. Microbiol. 14:25-31. [PubMed]
34. Thiel, T., B. Pratte, and M. Zahalak. 2002. Transport of molybdate in the cyanobacterium Anabaena variabilis ATCC 29413. Arch. Microbiol. 179:50-56. [PubMed]
35. Tseng, G. C., M. K. Oh, L. Rohlin, J. C. Liao, and W. H. Wong. 2001. Issues in cDNA microarray analysis: quality filtering, channel normalization, models of variations and assessment of gene effects. Nucleic Acids Res. 29:2549-2557. [PMC free article] [PubMed]
36. Xu, D., G. Li, L. Wu, J. Zhou, and Y. Xu. 2002. PRIMEGENS: robust and efficient design of gene-specific probes for microarray analysis. Bioinformatics 18:1432-1437. [PubMed]
37. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phage cloning vectors and host strains; nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33:103-119. [PubMed]
38. Zhang, Y., E. L. Pohlmann, P. W. Ludden, and G. P. Roberts. 2001. Functional characterization of three GlnB homologs in the photosynthetic bacterium Rhodospirillum rubrum: roles in sensing ammonium and energy status. J. Bacteriol. 183:6159-6168. [PMC free article] [PubMed]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • BioProject
    BioProject links
  • Compound
    PubChem chemical compound records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records. Multiple substance records may contribute to the PubChem compound record.
  • GEO DataSets
    GEO DataSets
    Gene expression and molecular abundance data reported in the current articles that are also included in the curated Gene Expression Omnibus (GEO) DataSets.
  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...