Identification of Genes Regulated by the Antitermination Factor NasT during Denitrification in Bradyrhizobium diazoefficiens.

The soybean symbiont Bradyrhizobium diazoefficiens grows anaerobically in the presence of nitrate using the denitrification pathway, which involves the nap, nir, nor, and nos genes. We previously showed that NasT acts as a transcription antitermination regulator for nap and nos gene expression. In the present study, we investigated the targets of NasT in B. diazoefficiens during denitrifying growth by performing transcription profiling with RNA-seq and quantitative reverse-transcription PCR. Most of the genes with altered expression in the absence of NasT were related to nitrogen metabolism, specifically several systems for branched-chain amino acid transport. The present results suggest that the reduced expression of genes involved in nitrogen acquisition leads to the induction of alternative sets of genes with similar functions. The ΔnasT mutant of B. diazoefficiens grew better than the wild type under denitrifying conditions. However, this enhanced growth was completely abolished by an additional loss of the narK or bjgb genes, which encode cytoplasmic systems for nitrite and nitric oxide detoxification, respectively. Since the expression of narK and bjgb was increased in the ΔnasT mutant, the growth of the ΔnasT mutant may be promoted by increased detoxification activity.

Soybean hosts the N2-fixing bacterium Bradyrhizobium diazoefficiens (reclassified from B. japonicum [3]), which has the ability to grow under low oxygen conditions by sequentially reducing nitrate (NO3−) to N2 via the denitrification pathway. Denitrification is a dissimilatory pathway that requires four enzymes in B. diazoefficiens: periplasmic nitrate reductase (Nap), nitrite (NO2−) reductase (NirK), nitric oxide (NO) reductase (cNor), and nitrous oxide (N2O) reductase (Nos). These enzymes and their accessory functions are encoded by the napEDABC, nirKV, norCBQD, and nosRZDYFLX gene clusters, respectively (28). Denitrification may be advantageous for bradyrhizobial cell survival in the soybean rhizosphere and for root colonization after oxygen depletion (18). Under symbiotic conditions, denitrification contributes to the production and detoxification of NO, an important signaling molecule for the establishment and functioning of symbiosis (27). Additionally, bradyrhizobial denitrification is involved in both the production and mitigation of the greenhouse gas N2O in soybean fields (9, 11).

The nasST operon encodes a NO3− and NO2− sensor/transcriptional antitermination regulatory system. This system was initially considered to be involved in the NO3−/NO2−-responsive regulation of nas genes for the NO3− assimilation pathway in bacteria, including B. diazoefficiens (1, 15, 16, 25, 38). Since nasST genes are located separately from the NO3−/NO2− assimilation gene cluster in some bacteria, a possible role for the regulation of other metabolic pathways was suggested (15). We showed that the expression of nap and nos genes was weaker under anaerobic NO3− respiration conditions (hereafter ‘denitrifying conditions’) in the B. diazoefficiens ΔnasT mutant than in the wild type (30). Other targets of this regulatory system are currently unknown in this bacterium.

NasS and NasT form a complex that dissociates when NasS senses NO3− at a micromolar concentration (8, 16, 30). NasT is an ANTAR (AmiR and NasR transcription antitermination regulator) family protein (34). When it is released from NasS, this protein interacts directly with the 5′-leader region of nosR mRNA and interferes with the formation of a terminator structure, allowing for the read-through transcription of nos genes (31). A similar antitermination mechanism is expected for other targets of NasT, as suggested for the regulation of nas genes in B. diazoefficiens and other bacteria (1, 16, 25, 38). The ΔnasT mutant has been shown to grow better than the wild type under denitrifying conditions (30); this was an unexpected observation because the growth of B. diazoefficiens is completely dependent on nap genes under denitrifying conditions (4).

The main objective of the present study was to investigate the targets of NasT in B. diazoefficiens under denitrifying conditions. We herein showed that NasT regulated a number of genes involved in nitrogen metabolism. We also characterized B. diazoefficiens mutants in different genes of the nas operon that are relevant to NO2− and NO detoxification in the cytoplasm (narK and bjgb) and demonstrated the involvement of these genes in the enhanced growth of the ΔnasT mutant under denitrifying conditions.

Materials and Methods

Bacterial strains and growth conditions

The strains used in the present study are listed in Table 1. Cells of B. diazoefficiens were pre-cultured aerobically with reciprocal shaking (300 rpm, 30°C) in HM salt medium (HEPES; 1.3 g L−1; MES, 1.1 g L−1; Na2HPO4, 0.125 g L−1; Na2SO4, 0.25 g L−1; NH4Cl, 0.32 g L−1; MgSO4 7H2O, 0.18 g L−1; FeCl3, 0.004 g L−1; CaCl2 2H2O, 0.013 g L−1; pH 6.8) supplemented with 0.1% l-(+)-arabinose and 0.25% (w/v) yeast extract (2, 26). Escherichia coli cells were grown at 37°C in Luria–Bertani medium (20). The following antibiotics were used for B. diazoefficiens: kanamycin (Km; 100 μg mL−1), spectinomycin (Sp; 100 μg mL−1), streptomycin (100 μg mL−1), and polymyxin B (50 μg mL−1); and for E. coli: Km (50 μg mL−1) and Sp (50 μg mL−1).

Table 1

Bacterial strains and plasmids used in the present study.

Strain/plasmid	Relevant characteristics	Source/reference
Strains
Bradyrhizobium diazoefficiens
USDA 110	Wild type	13
ΔnasT	nasT−; nasT::del	30
ΔnapA	napA−; napA::Ωcassette; Spr Smr	9
ΔnasC	nasC−; nasC::del	This study
ΔnarK	narK−; narK::del	This study
Δbjgb	bjgb−; bjgb::del	This study
ΔnapA-ΔnasT	napA− nasT−; napA::Ωcassette, nasT::del; Spr Smr	This study
ΔnasC-ΔnasT	nasC− nasT−; nasC::del, nasT::del	This study
ΔnarK-ΔnasT	narK− nasT−; narK::del, nasT::del	This study
Δbjgb-ΔnasT	bjgb− nasT−; bjgb::del, nasT::del	This study
Escherichia coli
DH5α	recA; cloning strain	Toyobo
Plasmids
pRK2013	ColE1 replicon carrying RK2 transfer genes; Kmr	5
pK18mobsacB	Suicide vector; Kmr	32
pΔnasT	pK18mobsacB::ΔnasT; Kmr	30
pΔnasC	pK18mobsacB::ΔnasC; Kmr	This study
pΔnarK	pK18mobsacB::ΔnarK; Kmr	This study
pΔbjgb	pK18mobsacB::Δbjgb; Kmr	This study

In growth experiments under denitrifying conditions, cells were inoculated into 5 mL (optical density ~0.01 at 660 nm) of HM medium supplemented with trace metals (26) and 10 mM KNO3 (HMMN) in 35-mL tubes. The tubes were sealed with rubber stoppers and the gas phase was replaced with 100% N2 in a vacuum line (26, 35). Cells were grown at 30°C with reciprocal shaking at 300 rpm. Growth was measured daily by recording optical density at 660 nm. Extracellular NO3− concentrations were measured as described previously (35).

RNA isolation and sequencing

Cells of B. diazoefficiens were inoculated into 20 mL (optical density ~0.01 at 660 nm) of HMMN medium in 100-mL bottles and reciprocally shaken (100 rpm, 30°C) for 24 h under denitrifying conditions. The isolation of total RNA, the DNaseI treatment, and cDNA synthesis were performed as described previously (31 and references therein). Two biological replications were processed for each strain (wild-type USDA 110 and ΔnasT mutant). In each cDNA sample (four in total), 5 μg was used for the RNA sequencing analysis. The removal of ribosomal RNAs with the Ribo-Zero Magnetic Kit for Gram-negative Bacteria (Epicentre, Madison, WI, USA), cDNA library preparation with the Illumina TruSeq Stranded mRNA LT Sample Prep Kit (Illumina, San Diego, CA, USA), and sequencing on the Illumina HiSeq 2000 Sequencing System in the paired-end mode running 100×2 cycles were performed by Hokkaido System Science (http://www.hssnet.co.jp/). In each of the four samples, ~45 million reads were generated; ~83% of the reads had Q (Phred quality score) ≥30 for each of the four samples (Table S1).

Bioinformatic analysis

Read trimming, the mapping of reads to the reference genome, read counting, normalization to RPKM (reads kilobase−1 million mapped reads−1), and calculations of expression values were performed with the CLC Genomics Workbench software 9.5.3. (https://www.qiagenbioinformatics.com/). Based on the total reads generated, approximately ~85% were trimmed, and from those, ~42% were mapped onto the B. diazoefficiens genome (GenBank accession number: NC_004463) (Table S1). Genes with fewer than 10 reads per 1 million mRNA reads were omitted from subsequent analyses. A gene with a fold change ≥2 or ≤−2 and a q value (estimate of the false discovery rate) ≤0.05 was considered to be up- or down-regulated, respectively. The Rhizobase (http://genome.annotation.jp/rhizobase) and KEGG (http://www.genome.jp/kegg/) databases were used for pathway analyses.

RNA sequencing data accession number

RNA sequencing data have been deposited in the NCBI Gene Expression Omnibus and are accessible through GEO Series accession number GSE130301 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE130301).

Validation of differential expression

To validate RNA sequencing results, quantitative reverse transcription-PCR (qRT-PCR) was performed on selected genes using a LightCycler Nano Instrument (Roche, Basel, Switzerland) with the FastStart Essential DNA Green Master (Roche) and specific primers for sigA (sigAf/sigAr), nosR (nosRf/nosRr), nosZ (nosZf/nosZr), napA (napAf/napAr), nirK (nirKf/nirKr), norB (norBf/norBr), narK (narKf/narKr), nasC (nasCf/nasCr), and bll3385 (bll3385f/bll3385r) (Table S2). The PCR program was set according to the manufacturer’s instructions. The specificity of PCR amplification was confirmed by a melting-curve analysis. This analysis was performed in duplicate for each of the two independent RNA samples. Expression levels calculated by the 2−ΔΔCt method (33) were normalized to the sigA level and expressed relative to the values for the wild type.

Construction of mutant strains

B. diazoefficiens ΔnarK, Δbjgb, and ΔnasC mutants were constructed by overlap extensions (10). In separate PCRs, two fragments (600–700 nucleotides each) of the target sequence were amplified using PrimeSTAR Max DNA Polymerase (TaKaRa Bio, Kusatsu, Japan) and the following primer sets: narK_01/narK_02 (to generate narK-A), narK_03/narK_04 (narK-B), bjgb_01/bjgb_02 (bjgb-A), bjgb_03/bjgb_04 (bjgb-B), nasC_01/nasC_02 (nasC-A), and nasC_03/nasC_04 (nasC-B). Primer sequences are listed in Table S2. narK-A and narK-B, bjgb-A and bjgb-B, and nasC-A and nasC-B fragments were then fused in second PCR with the same polymerase and the primer sets narK_01/narK_04, bjgb_01/bgjb_04, and nasC_01/nasC_04 (Table S2), respectively. PCR products were cloned as ~1.3-kb BamHI–PstI fragments for narK-AB and bjgb-AB and as ~1.3-kb EcoRI–BamHI fragments for nasC-AB into the pK18mobsacB vector (32). The resulting plasmids (pΔnarK, pΔbjgb, and pΔnasC; Table 1) were transferred by conjugation from E. coli DH5α to B. diazoefficiens using pRK2013 as a helper plasmid (5) to generate markerless deletions, as described previously (32). Kanamycin-resistant transconjugants were selected and grown in the presence of 10% sucrose to force the loss of the vector-encoded sacB gene. The resulting colonies were checked for Km sensitivity. The desired deletions were confirmed by PCR. To obtain ΔnapA-ΔnasT, ΔnarK-ΔnasT, Δbjgb-ΔnasT, and ΔnasC-ΔnasT double mutants, the plasmid pΔnasT (Table 1) was transferred by conjugation from E. coli DH5α to B. diazoefficiens Δ napA (Table 1), ΔnarK, Δbjgb, and ΔnasC, respectively, as described above. The desired deletions were confirmed by PCR using the primer set nasT_01/nasT_02 (Table S2).

Results and Discussion

The NasT regulon is mainly composed of genes involved in nitrogen metabolism

We used RNA-seq for the transcription profiling of the wild type and ΔnasT mutant, which were grown under anaerobic conditions in the presence of NO3− at 10 mM as the electron acceptor with both ammonia and organic nitrogen as nitrogen sources (2, 26). Under these conditions, NasT proteins are expected to be fully active (8, 30). We found that 77 genes were differentially expressed in the ΔnasT mutant from those in the wild-type strain, with 40 genes being down-regulated and 37 genes being up-regulated in the ΔnasT mutant (Table 2 and 3). Consistent with previous studies (30, 31), the complete napEDABC (for dissimilatory NO3− reductase) and nosRZDFYLX (for N2O reductase) gene clusters were decreased in the ΔnasT mutant (Table 2), whereas the expression of nirKV (for dissimilatory NO2− reductase) and norEBCQD (for NO reductase) gene clusters remained unchanged (Fig. 1). This result was further validated using qRT-PCR for eight genes, including the denitrification genes nosR, nosZ, napA, nirK, and norC (Fig. 1).

Table 2

List of down-regulated genes in the B. diazoefficiens ΔnasT mutant under denitrifying conditions.

Gene IDa	Gene product description	Fold changeb
bll2540	nadC; nicotinate-mononucleotide pyrophosphorylase	−3.08
bll2541	nadB; L-aspartate oxidase	−3.25
bll2542	nadA; quinolinate synthetase A	−2.47
blr7036	napE; periplasmic nitrate reductase protein	−6.8
blr7037	napD; periplasmic nitrate reductase chaperone	−5.01
blr7038	napA; periplasmic nitrate reductase large subunit	−6.95
blr7039	napB; periplasmic nitrate reductase small subunit	−7.73
blr7040	napC; membrane-anchored cytochrome c	−7.81
blr0314	nosR; transmembrane expression regulator/flavoprotein	−3.8
blr0315	nosZ; nitrous oxide reductase	−8.99
blr0316	nosD; periplasmic protein	−9.29
blr0317	nosF; cytoplasmic ABC transporter	−8.53
blr0318	nosY; transmembrane permease	−8.78
blr0319	nosL; periplasmic copper-binding lipoprotein	−9.12
blr0320	nosX; periplasmic flavoprotein	−7.43
blr0321	No similarity	−3.72
blr2896	paaI; phenylacetic acid degradation protein	−2.01
blr2897	paaK; phenylacetate-coenzyme A ligase	−2.15
bll3150	Putative oxalate:formate antiporter	−2.02
blr6246	ABC transporter substrate-binding protein; putative NitT/TauT family transport system	−2.02
blr6443	ABC transporter permease protein; putative branched-chain amino acid (LIV) transport	−2.46
blr6445	ABC transporter ATP-binding protein; putative LIV transport	−2.45
blr6446	ABC transporter substrate-binding protein; putative LIV transport	−2.64
blr6447	ABC transporter ATP-binding protein; putative LIV transport	−2.03
bll0913	ABC transporter substrate-binding protein; putative LIV transport	−2.07
blr7064	Putative ABC transporter substrate-binding protein	−2.07
blr0335	Putative carbon monoxide dehydrogenase small chain (coxS)	−2.02
blr0336	Putative carbon monoxide dehydrogenase large chain (coxL)	−2.02
blr3166	gcl; glyoxylate carboligase	−5.72
blr3167	hyi; hydroxypyruvate isomerase	−6.31
blr3168	glxR; oxidoreductase; putative tartronate semialdehyde reductase	−5.09
bll0332	Cytochrome-c like protein	−2.94
bll0333	Putative alcohol dehydrogenase	−3.07
bll7610	Conserved hypothetical protein	−2.31
blr2827	Conserved hypothetical protein	−2.29
blr3159	Conserved hypothetical protein	−2.02
blr6840	Conserved hypothetical protein	−2.52
bsr2315	Conserved hypothetical protein	−2.31
bll4571	nirA; assimilatory nitrite reductase	−3.14c
blr0612	glnK2; nitrogen regulatory protein PII	−2.05c

An underlined gene identifier indicates the presence of putative NasT-interaction hairpins in the leader region.

Fold change ≤2; q value ≤0.05, unless marked with “c”

P value ≤0.05.

Table 3

List of up-regulated genes in the B. diazoefficiens ΔnasT mutant under denitrifying conditions.

Gene IDa	Gene product description	Fold changeb
blr1311	Outer membrane protein	+2.31
blr5221	hspF; small heat shock protein	+4.43
blr2806	narK; nitrite extrusion protein	+4.86
blr2807	bjgb; single domain hemoglobin	+3.22
blr2808	flp; FAD-binding protein	+2.31
bll3383	ABC transporter permease protein; putative branched-chain amino acid (LIV) transport	+4.50
bll3384	ABC transporter ATP-binding protein; putative LIV transport	+4.89
bll3385	ABC transporter ATP-binding protein; putative LIV transport	+5.42
bll3386	AraC family transcriptional regulator	+2.30
blr2921	Conserved hypothetical protein	+31.63
blr2922	ABC transporter amino acid-binding protein; putative LIV transport	+18.76
blr2923	Amino acid ABC transporter permease protein; putative LIV transport	+25.39
blr2924	Amino acid ABC transporter permease protein; putative LIV transport	+24.71
blr2925	Amino acid ABC transporter ATP-binding protein; putative LIV transport	+22.85
blr2926	Amino acid ABC transporter ATP-binding protein; putative LIV transport	+24.49
blr6921	Putative multidrug resistance protein	+9.35
bll3369	Putative gluconolactonase	+3.95
bll3376	Oxidoreductase; putative aerobic carbon monoxide dehydrogenase small subunit (coxS)	+5.04
bll3377	Oxidoreductase; putative aerobic carbon monoxide dehydrogenase medium subunit (coxM)	+4.74
bll6500	Putative SAM (S-adenosyl-L-methionine)-dependent methyltransferase	+2.26
bll6502	Putative threonine dehydratase (ilvA)	+2.03
blr3831	mvrA; ferredoxin NADP+ reductase	+4.24
bll2855	rocD; ornithine aminotransferase	+2.18
bll3993	Conserved hypothetical protein	+2.09
bll3994	Conserved hypothetical protein	+2.48
bll6920	Conserved hypothetical protein	+5.42
blr3995	Conserved hypothetical protein	+2.50
blr4566	Conserved hypothetical protein	+6.29
blr4567	Conserved hypothetical protein	+6.19
blr4568	Conserved hypothetical protein	+2.67
bll4091	No similarity	+5.41
bll6133	No similarity	+2.13
bll6134	No similarity	+3.33
blr4022	No similarity	+5.35
blr4764	No similarity	+6.04
blr6135	Putative repressor LexA	+3.24
blr6136	No similarity	+2.96

An underlined gene identifier indicates the presence of putative NasT-interaction hairpins in the leader region.

Fold change ≥2; q value ≤0.05.

Fig. 1

Comparison of logarithm-transformed expression data generated by RNA-seq (white bars) and qRT-PCR (black bars). Fold-change values refer to differences in expression when the B. diazoefficiens ΔnasT mutant was compared with wild-type USDA 110. Data are means of two independent RNA samples.

In addition to denitrification genes, most of the genes with altered expression in the ΔnasT mutant were also related to nitrogen metabolism; a number of genes are putatively involved in the transport of amino acids, specifically, branched-chain amino acids (LIV, for L-leucine, L-isoleucine, and L-valine) (Table 2 and 3). This is in agreement with the observation that the majority of ANTAR-associated genes are related to nitrogen metabolism (24).

We found that some genes dedicated to similar processes varied in opposing directions in the ΔnasT mutant. In putative systems encoding LIV transport, blr6443/45-6447 and bll0913 were down-regulated, while the blr2922-2926 cluster and bll3383-3386 were up-regulated (Table 2 and 3). In addition, for aerobic carbon monoxide dehydrogenase, blr0335 (coxS) and blr0336 (coxL) were down-regulated, and bll3376 (coxS) and bll3377 (coxM) were up-regulated (Table 2 and 3).

The presence of alternative sets of LIV transport systems in B. diazoefficiens may be explained by LIV being important nutrients in bacterial physiology, with roles that range from supporting protein synthesis to signaling and fine-tuning the adaptation to amino acid starvation (12). Furthermore, LIV transport is essential for N2 fixation because symbiotic rhizobial cells become auxotrophs for LIV and depend on the plant for their supply (22, 23). The expression of bll3386, a transcriptional regulator that belongs to the bll3383-3386 gene cluster (Table 3), was induced in symbiotic cells of B. diazoefficiens (22).

We surveyed all genes that were down-regulated in the ΔnasT mutant (Table 2) for possible NasT-interaction hairpin formation within their mRNA leader regions (39). Among the genes examined, only two exhibited a clear NasT-interaction hairpin within the leader region: one was a putative coxS gene (blr0335; Fig. S1) and the other was nirA (bll4571; Fig. S1), encoding the assimilatory NO2− reductase. Thus, it is reasonable to expect a direct interaction between NasT and the hairpin within their leader regions (Fig. S1). However, the effects of NasT on the expression of the other genes may have been indirect; alternatively, NasT-binding motifs in these genes may not be well conserved and NasT may regulate these genes by an alternative mechanism. Nevertheless, we cannot rule out the possibility that some of these genes are regulated by unknown mechanisms that differ to NasT in response to the phenotype promoted by the nasT deletion.

Reduced expression of genes involved in nitrogen acquisition in the ΔnasT mutant is counteracted by the induction of genes with similar functions

The expression of a number of genes related to nitrogen acquisition, such as glc-hyi-glxR (allantoin degradation), paaIK (phenylacetate degradation), blr6443/45-6447 and bll0913 (LIV transport), nirA (assimilatory NO2− reductase), and glnK2 (nitrogen regulatory protein PII), was down-regulated (Table 2 and Fig. 2) (1, 6, 13, 21, 36). The PII protein encoded by glnB was also down-regulated in the ΔnasT mutant of Paracoccus denitrificans (17).

Fig. 2

Summary of transcription analysis results in the ΔnasT mutant under denitrifying conditions. See the text for details.

In contrast, other genes related to the transport, synthesis, and catabolism of amino acids were up-regulated in the absence of NasT (Table 3 and Fig. 2). Among them, we found two putative systems for LIV transport (blr2922-2926 and bll3383-3386); genes putatively involved in LIV synthesis: bll6500, bll6502 (threonine dehydratase ilvA), and mvrA (ferredoxin NADP+ reductase that may provide low-potential electrons for amino acid synthesis); and rocD encoding ornithine aminotransferase for arginine catabolism (7, 12, 13, 23). This result suggests the induction of alternative mechanisms to obtain nitrogen, counteracting the loss of other genes involved in nitrogen acquisition in the ΔnasT mutant.

narK-bjgb-flp genes are up-regulated in the ΔnasT mutant

Although NasT-recognizable hairpin formation was predicted in the leader regions of narK and bjgb mRNAs (Fig. S1), the expression of narK-bjgb-flp genes was stronger in the ΔnasT mutant than in the wild type (Fig. 1 and Table 3). Furthermore, the expression of the nasC gene, which is located downstream of flp, as part of the same operon (1, 29), remained unchanged in the ΔnasT mutant (Fig. 1). The narK-bjgb-flp-nasC operon encodes an integrated cytoplasmic system for NO3− assimilation and NO2−/NO detoxification in B. diazoefficiens (1, 29). NarK is an MFS (major facilitator superfamily)-type NO3−/NO2− transporter that lowers cytoplasmic NO2− levels by exporting NO2− to the periplasm, Bjgb is a single-domain hemoglobin that detoxifies NO in the cytoplasm, and Flp is a flavoprotein that functions as an electron donor to the hemoglobin Bjgb and assimilatory NO3− reductase NasC (1, 29).

In contrast, Cabrera et al. reported that the expression of narK was down-regulated in the absence of NasT. Therefore, they suggested the down-regulation of the narK-bjgb-flp-nasC operon in the ΔnasT mutant (1). A possible explanation for this discrepancy is that the researchers employed aerobic NO3−-assimilation conditions (i.e., NO3− as the sole nitrogen source). However, we herein employed denitrifying non-assimilation conditions; under anaerobiosis with NO3− as the electron acceptor and ammonia and organic nitrogen as nitrogen sources (2, 26). The present results suggest that the function of narK-bjgb-flp-nasC genes is important beyond NO3−-assimilation conditions and may be subjected to a complex regulatory system. In support of this hypothesis, RegR, the response regulator of the RegSR two-component regulatory system, has been shown to activate the transcription of the narK-bjgb-flp-nasC operon under denitrifying conditions, with a putative RegR box located upstream of narK (37). Additionally, we found a putative FixK box upstream of bjgb (data not shown), suggesting the control of a FixK-like transcriptional regulator in response to low oxygen conditions (14, 19).

Notably, the expression of nasC (encoding the assimilatory NO3− reductase) and nirA (encoding the assimilatory NO2− reductase) differed with respect to their dependence on NasT (Table 2 and 3). Under our experimental conditions, the nirA expression pattern was similar to that reported by Cabrera et al., in which nirA was down-regulated in the ΔnasT mutant (1). In contrast to the majority of bacteria in which the genes encoding an assimilatory NO3− reductase or NO2− reductase are arranged in the same operon (15), nirA in B. diazoefficiens is located downstream of NasST, while nasC is located at a separate locus (30). The separate location of nasC supports narK-bjgb-flp-nasC genes being subjected to NasT-independent regulation, as observed under the experimental conditions employed in the present study.

Periplasmic nitrate reductase is responsible for anaerobic nitrate reduction in the ΔnasT mutant

An interesting finding from a previous study showed that although nap expression and Nap activity both markedly decreased in the ΔnasT mutant under denitrifying conditions, growth was more rapid than that of the wild type (30). The enhanced growth of the ΔnasT mutant under denitrifying conditions was confirmed in the present study (Fig. 3A). Consistent with previous findings by Delgado et al. (4), the growth of the ΔnapA single mutant was completely abolished (Fig. 3A). The growth of the ΔnapA-ΔnasT double mutant was also abolished (Fig. 3A), indicating that Nap is the sole enzyme responsible for reducing NO3− to NO2− under anaerobic NO3−-respiring conditions. The ΔnasT mutant and wild-type strain consumed NO3− at a similar rate from the growth medium, but the consumption was completely abolished in the ΔnapA-ΔnasT double mutant (Fig. 3B). These results indicate that a reduced level of Nap is still sufficient to sustain the anaerobic NO3− reduction in the ΔnasT mutant.

Fig. 3

Growth of B. diazoefficiens under denitrifying conditions in HMMN medium. (A) Growth of wild-type USDA 110 and the indicated mutant strains. Growth was measured by recording optical density at 660 nm on a daily basis. (B) Extracellular concentrations of nitrate (NO3−) are indicated for the cultures shown in (A). The results presented are the means of at least three biological replicates±standard deviations (n=3–5).

Involvement of nas genes in the growth enhancement of the ΔnasT mutant

We investigated whether the narK-bjgb-flp-nasC operon is involved in the growth enhancement of the ΔnasT mutant under denitrifying conditions. The additional loss of the narK or bjgb genes in the ΔnasT mutant background suppressed the growth enhancement of the ΔnasT single mutant (Fig. 4A and B). This result indicates that the narK and bjgb genes are necessary for the enhanced growth of the ΔnasT mutant. The mechanisms responsible for this enhancement may be related to the increased capacity of the ΔnasT mutant to detoxify NO2− and NO in the cytoplasm (1) (Fig. 5). NarK may act to reduce cytoplasmic NO2− levels, which are presumably the result of the decreased expression of nirA (Table 2 and Fig. 5); Bjgb-Flp may reduce NO, which is produced by NasC during anaerobic nitrate-dependent growth, as reported previously (1).

Fig. 4

Involvement of narK, bjgb, and nasC genes in the growth of B. diazoefficiens under denitrifying conditions in HMMN medium. (A) Growth of ΔnarK and ΔnasT-ΔnarK mutants. (B) Growth of Δbjgb and ΔnasT-Δbjgb mutants. (C) Growth of ΔnasC and ΔnasT-ΔnasC mutants. Wild-type USDA 110 and the ΔnasT mutant are shown as a reference in all charts. Growth was measured by recording optical density at 660 nm on a daily basis. Results presented are the mean of at least three biological replicates±standard deviations (n=3–5).

Fig. 5

Model of the function of denitrification genes, the nas gene cluster, and nirA products in the ΔnasT mutant. Proteins in white, black, or gray indicate the down-regulation, up-regulation, or unchanged regulation of the respective genes in the ΔnasT mutant. See the text for details.

The additional loss of nasC in the ΔnasT background resulted in an intermediate growth phenotype between the ΔnasT single mutant and wild type (Fig. 4C). This result suggests that the assimilatory NO3− reductase NasC is active in the ΔnasT mutant and is involved in the growth enhancement of the ΔnasT mutant. The contribution of NasC to energy production may be related to its capacity to reduce NO2− to NO (1) (Fig. 5).

Conclusions

Our results suggest the following: (i) NasT is a key regulator for genes associated with nitrogen metabolism under denitrifying conditions, particularly for branched-chain amino acid transport; (ii) the direct NasT regulatory mechanism that was described for nos genes (31) may not be common for other targets because most of them did not exhibit a NasT-interaction hairpin; and (iii) the transcription of some NasT targets may be enhanced in a NasT-independent manner under non-assimilation denitrifying conditions, as observed for the narK-bjgb-flp-nasC operon.

According to the model proposed by Cabrera et al. (1), an explanation for the events that occur in the B. diazoefficiens ΔnasT mutant under denitrifying conditions is as follows (Fig. 5). The loss of genes in the ΔnasT mutant may induce genes responsible for alternative nitrogen acquisition, including narK-bjgb-flp-nasC (Fig. 2). As a consequence of the induction of the narK-bjgb-flp-nasC operon, the growth of B. diazoefficiens under denitrifying conditions is induced, which may be explained by the enhancement of NO2− and NO detoxification systems in the cytoplasm (1) (Fig. 5).

These results may provide a novel approach for enhancing the denitrifying growth of B. diazoefficiens and other bradyrhizobial strains by optimizing the NO2− and NO detoxification systems. This may have important implications for improving the survival of bradyrhizobial cells in the soybean rhizosphere and for root colonization, as well as for the modulation of NO levels in soybean nodules and N2O levels in soybean fields (9, 11, 18, 27).