Symbiotic Bradyrhizobium japonicum Reduces N₂O Surrounding the Soybean Root System via Nitrous Oxide Reductase

Abstract

N₂O reductase activity in soybean nodules formed with Bradyrhizobium japonicum was evaluated from N₂O uptake and conversion of ¹⁵N-N₂O into ¹⁵N-N₂. Free-living cells of USDA110 showed N₂O reductase activity, whereas a nosZ mutant did not. Complementation of the nosZ mutant with two cosmids containing the nosRZDFYLX genes of B. japonicum USDA110 restored the N₂O reductase activity. When detached soybean nodules formed with USDA110 were fed with ¹⁵N-N₂O, they rapidly emitted ¹⁵N-N₂ outside the nodules at a ratio of 98.5% of ¹⁵N-N₂O uptake, but nodules inoculated with the nosZ mutant did not. Surprisingly, N₂O uptake by soybean roots nodulated with USDA110 was observed even in ambient air containing a low concentration of N₂O (0.34 ppm). These results indicate that the conversion of N₂O to N₂ depends exclusively on the respiratory N₂O reductase and that soybean roots nodulated with B. japonicum carrying the nos genes are able to remove very low concentrations of N₂O.

Nitrous oxide (N₂O) is a key atmospheric greenhouse gas that contributes to global climate change through radiative warming and depletion of stratospheric ozone (1, 24). Agricultural land is a major source through denitrification and nitrification (14, 35) and contributes significantly to the net increase in atmospheric N₂O (1, 34). Several attempts have been made to reduce the emission of N₂O from agricultural systems (34, 35).

The complete denitrification of nitrate by bacteria to dinitrogen (N₂) is generally an anaerobic respiratory process, where the last step is mediated by N₂O reductase (54). The corresponding structural gene is nosZ and is assembled in the nosRZDFYL gene operon (54). Several species capable of denitrification are also nitrogen-fixing bacteria, including rhizobia such as Bradyrhizobium japonicum (4, 6, 44, 49) and Sinorhizobium meliloti (7, 20). Indeed, genes responsible for denitrification have been found in rhizobial genomes (15, 23).

Earlier studies (19, 37) reported the evolution of ¹⁵N-N₂ from ¹⁵N-N₂O from sliced or detached soybean nodules. Recently, Velasco et al. (50) reported that nosZ and nosR insertion mutants of B. japonicum USDA110 accumulate N₂O when cultured microaerobically in the presence of nitrate. The nosZ gene was also expressed in soybean nodules (29). However, it has not yet been fully proved that N₂ evolution from N₂O by soybean nodules is mediated by N₂O reductase encoded by the nosZ gene in B. japonicum (4, 9, 38). The aims of this work were to confirm whether the nos gene cluster of B. japonicum is responsible for respiratory N₂O reduction to N₂ in nodules and to evaluate the capability of the nodulated roots to transform N₂O into N₂.

MATERIALS AND METHODS

Bacterial strains, plasmids, and media.

The bacterial strains and plasmids used are listed in Table 1. Bradyrhizobium cells were grown at 30°C in HM salt medium (8) supplemented with 0.1% arabinose and 0.025% (wt/vol) yeast extract (Difco, Detroit, MI), which is termed HM medium here. HM medium was further supplemented with 0.55 μM Na₂MoO₄ · 2H₂O, 1 μM FeCl₃, and 1 μM CuSO₄ · 5H₂O (HMM medium) for the denitrification assay (44). Escherichia coli cells were grown at 37°C in Luria-Bertani medium (42). Antibiotics were added to the media at the following concentrations: for B. japonicum, 100 μg of tetracycline (Tc)/ml, 100 μg of spectinomycin (Sp)/ml, 100 μg of streptomycin (Sm)/ml, 100 μg of kanamycin (Km)/ml, and 50 μg of polymyxin B/ml; for E. coli, 15 μg of Tc/ml, 50 μg of Sp/ml, 50 μg of Sm/ml, 50 μg of Km/ml, and 100 μg of ampicillin/ml.

TABLE 1.

Bacterial strains and cosmids used in this study

Strain or plasmid	Characteristicsa	Source or reference
Strains
Bradyrhizobium japonicum
USDA110	Wild type; nosZ+	23
T7	Field isolate in Tokachi, Hokkaido, Japan; nosZ natural mutant	32, 43
USDA110ΔnosZ	USDA110 nosZ::del/ins Ω cassette; Smr Spr	This study
Escherichia coli
DH5α	recA; cloning strain	Toyobo Inc.
JM109	recA; cloning strain	Toyobo Inc.
HB101	recA hsdR hsdM pro Smr	5
S17-1	hsdR RP4-2(kan::Tn7)(tet::Mu); Smr Spr	47
Plasmids
pRK2013	ColE1 replicon carrying RK2 transfer genes; Kmr	13
pTZ18R	Cloning vector; pMB1ori; Apr	46
pTZ18RnosZ	pTZ18R carrying 4.0-kb nosZ fragment; Apr	This study
pTZ18rRΔSacI-nosZ	SacI-deleted pTZ18R carrying 4.0-kb nosZ fragment; Apr	This study
pHP45Ω	Plasmid carrying 2.1-kb Ω cassette; Spr Smr Apr	40
pK18mob	Cloning vector; pMB1ori oriT; Kmr	46
pK18mob-nosZ::Ω	pK18mob carrying 4.0-kb nosZ fragment; Kmr Smr Spr	This study
Cosmids
pKS800	Derivative of broad-host-range cosmid pLAFR1; IncP Tcr	17, 23
brc02856	pKS800 carrying nos gene clusterb; Tcr	23
brc01733	pKS800 carrying nos gene clusterb; Tcr	23

^aAp^r, ampicillin resistant; Tc^r, tetracycline resistant; Km^r, kanamycin resistant; Sm^r, streptomycin resistant; Sp^r, spectinomycin resistant.

^bSee Fig. 1.

DNA manipulations.

Isolation of plasmids, DNA ligation, and transformation of E. coli were performed as described by Sambrook et al. (42). DNA preparation and Southern hybridization were carried out as described previously (22, 30, 43).

Construction of a B. japonicum USDA110 nosZ mutant.

A 4-kb BamHI DNA fragment identified from the genome sequence of B. japonicum USDA110 (23) was excised from BamHI-digested total DNA and inserted into the BamHI site of pTZ18R (Fig. 1A). The plasmid containing the fragment was selected by PCR amplification specific for the nosZ gene of B. japonicum USDA110 with primer 3 (5′-GAGGATCGTTTCGCATGAGCGACAGCGACAACAT-3′) and primer 4 (5′-CGCTCCCGATCAGACCGATTT-3′), in which the underlined nucleotides form the USDA110 nosZ sequence. From the cloned 4-kb DNA fragment, pK18mob-nosZ::Ω was constructed (Fig. 1A). Biparental mating was conducted on HM agar plates using E. coli S17-1 (47). Double crossover was verified by Southern hybridization with the 4-kb nosRZD fragment as a probe (Fig. 1A).

FIG. 1.

Construction of a nosZ mutant (A) and complementation cosmid inserts (B) of Bradyrhizobium japonicum USDA110. (A) Cloned fragments in pTZ18RΔSacI-nosZ and pK18mob-ΔnosZ::Ω are shown alongside the physical map of the nos gene cluster of B. japonicum USDA110. B, BamHI; S, SacI; X, XhoI; Sb, SacI blunt end; Bb, BamHI blunt end; Xb, XhoI blunt end. During blunt-end ligation, the Sb/Bb site produced a BamHI site in pK18mob-ΔnosZ::Ω. Upstream of nosR lies a consensus sequence of an FNR box (TTGAT-N₄-ATCAA) (50). The cloned 4-kb DNA fragment containing the nosZ gene in pTZ18R-nosZ was reconnected to pTZ18RΔSacI, which was produced by blunting of the SacI site of pTZ18R, and pTZ18RΔSacI-nosZ was generated. The Ω cassette, which was excised from pHP45Ω (40), digested with BamHI, and blunt ended, was inserted into the blunt-ended SacI and XhoI sites of pTZ18RΔSacI-nosZ to generate pTZ18RΔSacI-nosZ::Ω. Finally, the fragment containing the nosZ and Ω cassette fusion was inserted between the SmaI and XbaI sites of pK18mob to generate pK18mob-nosZ::Ω. (B) Positions covered by two cosmid inserts, brc02856 and brc01733 (Table 1), in the genome of B. japonicum USDA110 (http://www.kazusa.or.jp/rhizobase/) (23).

Genetic complementation by nos genes.

Two cosmids, brc02856 and brc01733, carrying the nos gene cluster (nosRZDFYLX) were selected from the pKS800 cosmid library, which was constructed for the sequencing of B. japonicum USDA110 (23). The cosmids were introduced into B. japonicum strains by triparental mating via pRK2013 (39, 47, 52).

Gas chromatography.

To determine N₂O concentrations, we used two ⁶³Ni electron capture detector gas chromatographs. In the quick N₂O analysis, 0.2 ml of sample gas was injected into a Shimadzu (Kyoto, Japan) GC-17A gas chromatograph with a CP-PoraBOND Q capillary column (internal diameter, 0.32 mm, length, 25 m; Varian, Palo Alto, CA). The temperatures of injection, the column, and the detector were 250, 65, and 300°C, respectively. In the sensitive N₂O analysis, 0.8 ml of sample gas was injected into a Shimadzu GC-14BPsE gas chromatograph equipped with tandem packed columns of a Porapak N (80/100 mesh; diameter, 0.3 mm; length, 1 m) and a Porapak Q (80/100 mesh; diameter, 0.3 mm; length, 3 m). The temperatures of injection, the column, and the detector were 80, 80, and 340°C, respectively. The carrier gas used for both N₂O analyses was composed of 5% (vol/vol) CH₄ in Ar. ¹⁵N-N₂ was determined with a thermal conductivity detector gas chromatograph as described previously (44). For the acetylene reduction assay, ethylene was measured by a flame ionization detection gas chromatograph, as described previously (12).

N₂O uptake by free-living cells.

B. japonicum cells were washed with HMM medium by centrifugation (5,000 × g, 15 min, 4°C), and the cell density was adjusted to 6 × 10⁹ cells/ml (33). The cell suspension (4.6 ml, 6 × 10⁹ cells/ml) was introduced into a 123-ml airtight vial (V-100; Nichiden-Rika Glass, Kobe, Japan). After the gas phase was replaced with N₂, N₂O gas was introduced into the vial at a final concentration of approximately 0.1% (vol/vol).

¹⁵N-N₂ emission from ¹⁵N-labeled N₂O by free-living cells.

B. japonicum cells (7 ml; 6 × 10⁹ cells/ml) were introduced into a 14-ml airtight specimen vial. After the N₂ gas replacement, ¹⁵N-N₂O (¹⁵N, >98 atom%; Cambridge Isotope Laboratories Inc., Andover, MA) was introduced. After incubation at 30°C for 67 h, the concentration of ¹⁵N-N₂ in the gas phase was determined by TCD gas chromatography (44).

Real-time PCR.

Beacon Designer software (Premier Biosoft International, Palo Alto, CA) was used to design the nosZ primer set (nosZ2F, 5′-GTCGTCATCTTCAACCTCAAG; nosZ2R, 5′-TATTCATGCCATGCGGATTG). Total RNA was prepared as described previously (53). Quantitative reverse transcription-PCR analysis was carried out by the i-Cycler optical system (Bio-Rad Laboratories, Inc., Tokyo, Japan) as described previously (53).

Plant cultivation and inoculation.

The surface-sterilized soybean seeds (Glycine max cv. Enrei) were germinated and transplanted to a Leonard jar (25, 48), which contained sterile vermiculite and nitrogen-free sterilized nutrient solution (31, 39). The cell suspension of B. japonicum (1 ml) was used for inoculation at 1 × 10⁷ cells per seed. For N₂O-feeding experiments, plants were grown in a growth chamber (LH200; Nippon Medical & Chemical Industries, Tokyo, Japan). For the nitrogen fixation experiment, soybeans were grown in pots (1 liter) filled with vermiculite in a phytotron (Koitotron type KC; Koito Industries, Tokyo, Japan). A nitrogen-free sterilized nutrient solution was periodically supplied to the pots (31, 39).

N₂O uptake, ¹⁵N-N₂ emission, and ¹⁵N incorporation by nodules.

After the nodulated roots of 45-day-old soybean plants were washed with water, the nodules were detached from the roots and weighed. The nodules were introduced into a 15-ml airtight vial. After the vials were sealed, ¹⁵N-N₂O was injected. After the gas measurements, the nodules were dried at 80°C for 3 days and then powdered with a mortar and pestle. The nodule samples were subjected to tracer ¹⁵N and total-N analyses by an automatic gas chromatograph-mass spectrometer (EA1110-DELTA^plus Advantage ConFloIII system, Thermo Electron Co., Bremen, Germany).

The whole root system of 32-day-old soybean plants was inserted into a 51-ml test tube (19.4 mm in diameter by 176 mm in height) containing 20 ml nitrogen-free nutrient solution (31, 39). The test tube was sealed with a silicon rubber stopper with a gas sampling port and a hole in the center. This system enclosed the whole root system of intact soybeans in the test tube, leaving the aerial parts in air. Then N₂O gas was injected into the closed test tube.

Nitrogen fixation and plant growth.

The nitrogen-fixing activity of nodules was examined by acetylene reduction assays in an incubation jar (300-ml capacity) containing the excised root system. Acetylene gas was injected into the jar at 15% (vol/vol), and the roots were incubated for 60 min at 25°C. Separated shoot, roots, and nodules were weighed.

Bacteroid antibiotic test.

Nodules were removed from soybeans 18 days after inoculation with T7 or T7 with brc01733. Homogenates of surface-sterilized nodules were plated on HM agar medium with and without tetracycline.

RESULTS

Construction of the nosZ deletion mutant.

To examine whether the nos gene cluster of B. japonicum was responsible for N₂O reductase activity, we constructed a nosZ deletion mutant of B. japonicum USDA110 (Table 1). After deletion of most of the nosZ gene from the cloned DNA fragment containing nosRZD, the Ω cassette was inserted into the position (Fig. 1). When free-living cells of USDA110 and USDA110ΔnosZ were anaerobically incubated in gas containing 0.1% (vol/vol) N₂O, the parent strain consumed N₂O, whereas the mutant did not show N₂O uptake (Fig. 2A). To further examine the N₂O reductase reaction, ¹⁵N-N₂O was supplied to cultures of B. japonicum USDA110 and USDA110ΔnosZ. Wild-type USDA110 stoichiometrically produced ¹⁵N-N₂ from ¹⁵N-N₂O, whereas USDA110ΔnosZ did not (data not shown). These results suggest that nosZ of B. japonicum USDA110 is responsible for the N₂O reductase activity that converts N₂O to N₂.

FIG. 2.

N₂O uptake by free-living cells of the nosZ mutant (A) and nos complementation of USDA110ΔnosZ (B) and T7 (C). pKS800 is the cosmid vector used for complementation (17). Two cosmid clones, brc02856 and brc01733, have inserts containing nos genes of USDA110 (Table 1; Fig. 1). N₂O reductase activities (means ± SDs) estimated from maximum rates of N₂O uptake were 4.6 ± 0.2 [USDA110(pKS800)], 22.9 ± 0.7 [USDA110ΔnosZ(brc02856)], and 23.6 ± 1.9 N₂O [USDA110ΔnosZ(brc01733)] μmol h⁻¹ 10⁷ cells⁻¹.

Complementation of USDA110ΔnosZ with nos genes.

To further examine the validity of nosZ-dependent N₂O reductase activity, two cosmids containing the nos gene cluster were introduced into USDA110ΔnosZ (Fig. 1B). The transconjugants were anaerobically incubated in the presence of 0.1% (vol/vol) N₂O and tetracycline, and USDA110 and USDA110ΔnosZ carrying the pKS800 vector were used as controls (Fig. 2B). USDA110 and USDA110ΔnosZ carrying pKS800 (Fig. 2B) showed N₂O uptake profiles similar to those of wild-type USDA110 and USDA110ΔnosZ, respectively (Fig. 2A). However, the two cosmids in Fig. 1B restored strong N₂O uptake (Fig. 2B). The amino acid sequences of B. japonicum nosRZDFYLX (23, 50) have a high degree of similarity to those of other N₂O reductases from Pseudomonas stutzeri (21). Both B. japonicum cosmids contain nosRZDFYLX and adjacent genes (Fig. 1B). Their addition to USDA110ΔnosZ produced similar activities of N₂O reductase, as indicated by N₂O uptake (Fig. 2B). Moreover, the introduction of a cosmid (brc02856 or brc01733) into USDA110ΔnosZ restored the strain's ability to convert ¹⁵N-N₂O into ¹⁵N-N₂ (data not shown). Thus, the complementation experiments confirmed that the nos gene cluster per se is responsible for N₂O reductase activity in B. japonicum USDA110.

Expression of nosZ.

Strain USDA110ΔnosZ with the two cosmids containing the nos cluster rapidly and strongly absorbed N₂O, much more than the parent strain USDA110 containing the pKS800 vector (Fig. 2B). When the N₂O reductase activity was estimated from the maximum rate of N₂O uptake in Fig. 2B, the activities of USDA110ΔnosZ containing brc02856 and brc01733 (22.9 ± 0.7 and 23.6 ± 1.9 N₂O μmol h⁻¹ 10⁷ cells⁻¹) were five times that of wild-type USDA110 (4.6 ± 0.2 N₂O μmol h⁻¹ 10⁷ cells⁻¹). In addition, the lag time for the induction of N₂O reductase activity in the wild-type USDA110(pKS800) disappeared in the cosmid-complemented USDA110ΔnosZ.

Because we used pKS800, a derivative of cosmid pLAFR1 (RK2), as a vector for the nos complementation (17), it is possible that the cosmid introduction increases the copy number and thus results in higher expression of nos genes. USDA110ΔnosZ carrying brc01733 expressed the nosZ gene at 5-times-higher levels on the basis of total RNA than USDA110ΔnosZ carrying the pKS800 vector 1 h after anaerobic incubation in the presence of 0.1% (vol/vol) N₂O. This result suggests that the higher expression of the nos genes enhanced N₂O reductase activity and shortened the lag time for induction in USDA110ΔnosZ with the nos cosmids.

N₂O uptake and N₂ emission by nodules.

Denitrification genes of B. japonicum are expressed in a microaerobic state in both legume nodules and free-living bacteria via the fixLJ and fixK₂ regulatory cascades (3, 28). Mesa et al. (29) reported that the nir, nor, and nos denitrification genes in B. japonicum were expressed in soybean root nodules. We thus examined the nodulation phenotypes of the nosZ mutant. The nodule weight and nitrogen-fixing activity of USDAΔnosZ showed no significant differences from those of USDA110 (Table 2), indicating that nosZ mutation does not alter symbiotic nitrogen fixation. This agrees with the results of a report on the nodulation kinetics (29) of a nosZ insertion mutant of B. japonicum (50).

TABLE 2.

Growth and N₂-fixing activity of soybean plants inoculated with Bradyrhizobium japonicum USDA110 or the nosZ mutant^a

Strain	Growth (g [fresh wt] plant−1)			N2-fixing activity (C2H2-reducing activity) (μmol h−1 g of nodule [fresh wt]−1)
	Shoot	Root	Nodule
USDA110	10.8 ± 0.7	4.8 ± 0.4	1.0 ± 0.1	10.3 ± 1.3
USDA110ΔnosZ	10.4 ± 0.9	4.5 ± 0.4	0.9 ± 0.1	9.0 ± 2.0
Uninoculated	6.9 ± 0.2	5.5 ± 0.2	0	0

^aValues for growth and N₂-fixing activity are shown as means ± SDs, with six replications.

To evaluate the extent of N₂O uptake and N₂ emission by N₂O reductase in soybean nodules, ¹⁵N-N₂O was supplied to detached nodules formed with the nosZ mutant or the parent strain (Fig. 3). Soybean nodules formed with USDA110 rapidly absorbed ¹⁵N-N₂O and emitted ¹⁵N-N₂ outside the nodules, whereas nodules formed with the nosZ mutant did not mediate the conversion of ¹⁵N-N₂O to ¹⁵N-N₂ at all (Fig. 3).

FIG. 3.

¹⁵N-N₂O uptake and ¹⁵N-N₂ emission by detached soybean nodules formed with Bradyrhizobium japonicum USDA110 (A) and the nosZ mutant USDA110ΔnosZ (B). ¹⁵N-labeled N₂O was supplied to vials (15 ml) containing soybean nodules (1.6 to 1.8 g). Five independent experiments gave similar results.

Because the transformation of ¹⁵N-N₂O into ¹⁵N-N₂ was likely stoichiometric (Fig. 3), we determined the amount of fixed ¹⁵N in the nodules and made a ¹⁵N balance sheet for ¹⁵N-N₂O exposure. ¹⁵N atoms of ¹⁵N-N₂O were distributed 98.5% ± 0.1% (mean ± standard deviation [SD]) outside the nodules as ¹⁵N-N₂ and 1.5% ± 0.1% (mean ± SD) fixed in the nodules. Thus, most N₂ molecules produced by N₂O reductase in bacteroids were released outside the nodules, and nitrogenase was not able to utilize most of them as a substrate for nitrogen fixation. This is probably due to the diffusion of N₂ into various compartments for N₂O reductase and nitrogenase. The apparent rate of N₂O uptake by USA110 nodules under about 3% N₂O (Fig. 3) was estimated to be 7.8 μmol h⁻¹ per g (fresh weight) of nodule; this was comparable with the acetylene-reducing activity (Table 2). These results indicate that the conversion of N₂O into N₂ in soybean nodules depends on the function of the nos gene cluster in B. japonicum.

N₂O uptake by intact soybeans nodulated with USDA110.

To examine the extent of N₂O concentrations that the nodulated roots can take up, the whole root systems of intact plants nodulated by USDA110 or the nosZ mutant were enclosed. After the introduction of N₂O gas, N₂O concentrations around the whole root system were monitored (Fig. 4A and B). At an initial N₂O concentration of 800 ppm (Fig. 4A), the root systems nodulated by USDA110 absorbed N₂O, whereas the nosZ mutant-inoculated roots did not take up N₂O at all. After a 21-h incubation, the N₂O concentration reached 4 to 8 ppm around the roots nodulated with USDA110 (Fig. 4A). Thus, the initial concentrations of N₂O were gradually decreased to 50, 5, 1, and 0.34 ppm. Surprisingly, the root systems nodulated by USDA110 were able to take up 0.34 ppm N₂O, reducing the concentration to 0.11 ± 0.02 ppm (mean ± SD) 22 h after the start of incubation (Fig. 4B). When a shoot-decapitated root system was completely enclosed in the test tube with a rubber stopper to avoid gas leakage, the N₂O concentration around the root system nodulated by USDA110 decreased to below the detection limit (<0.02 ppm) by 17 h (Fig. 4C).

FIG. 4.

N₂O uptake from root systems of intact plants (A and B) and decapitated root systems (C) inoculated with the wild-type strain or the nosZ mutant of Bradyrhizobium japonicum USDA110. Initial gas phases were adjusted to approximately 800 ppm (A) (intact plant) and 0.34 ppm (B [intact plant] and C [decapitated root system]) of N₂O in air. Soybean plants that harbored 0.26 to 0.30 g (A), 0.15 to 0.22 g (B), or 0.18 to 0.36 g (C) of nodules (fresh weight per plant) were examined by a test tube system (see the text). Triplicate determinations were carried out, except that data for USDA110ΔnosZ inoculation in panel A show means of duplicate determinations. Error bars, SDs.

Generally, the N₂O concentration in soils (0.36 ppm to 8,300 ppm) is higher than that in ambient air (0.31 ppm) (1, 26, 34, 45). Therefore, these results strongly suggest that soybean nodules formed with B. japonicum carrying nos genes potentially scavenge N₂O gas in soil.

Introduction of nos genes into a naturally occurring nosZ-deficient strain.

Strains of B. japonicum can apparently be classified into three denitrification types: full denitrifiers (up to N₂), truncated denitrifiers (up to N₂O), and nondenitrifiers (44, 49). Their incidence depends on the field site and the phylogeny (44). We tested whether introduction of the nos gene cluster into the naturally occurring nosRZD-deficient strain T7 would induce N₂O reductase activity. Introduction of brc02856 or brc01733 into strain T7 induced N₂O uptake, but T7 carrying the cosmid vector did not show such activity (Fig. 2C).

To test whether strain T7 carrying brc01733 exhibited N₂O reductase activity in nodules, this strain was inoculated into soybean plants. However, the instability of cosmid brc01733 interfered with this experiment: Bacteroid cells from the nodules lost cosmid brc01733 during nodule development, as suggested by plate counts of tetracycline-resistant colonies (<10⁻⁶). Because similar instability of pLAFR1 derivatives has been reported (10), other strategies are needed for the symbiotic testing of strain T7 carrying nos genes.

DISCUSSION

The nosZ mutant of USDA110 was not able to convert ¹⁵N-N₂O to ¹⁵N-N₂ in nodules (Fig. 3) or as free-living cells (Fig. 2) and continued to produce nitrogen-fixing nodules (Table 2). Thus, the evolution of ¹⁵N-N₂ from ¹⁵N-N₂O in soybean nodules is due exclusively to N₂O reductase, which is encoded by nosZ.

An important finding of this work is that N₂O reductase expressed in B. japonicum can take up very low concentrations of N₂O from outside of the nodules, equivalent to the natural concentration of N₂O in air (approximately 0.31 ppm). Probably this indicates that N₂O reductase has a high affinity for N₂O as a substrate (54), and the gas movement through the tissues of the nodules does not severely interfere with the enzyme property.

Agricultural soil is a major source of the N₂O emissions that contribute to global warming (1, 34). The N₂O concentration in the soil gas phase fluctuates with water, organic matter, and nitrogen fertilizer but is generally higher than that of ambient air (1, 14, 18, 24, 26, 45). The results of N₂O uptake by soybean nodules (Fig. 3 and 4) suggest that soybean nodules formed with B. japonicum carrying nos genes scavenge N₂O in soil, thus lessening N₂O emission to the atmosphere from soybean fields.

However, several groups have reported that cultivation of legume crops often enhances N₂O emission from fields of alfalfa (11), soybean (51), white clover (36), and Bengal gram (16). Thus, we want to discuss the N₂O paradox of legume root nodules for those results and ours. One possible explanation for the N₂O paradox is that the nitrate absorbed via legume roots and nodule surfaces from the soil solution might be transported in nodules (27, 41), which would emit N₂O from nitrate in bacteroids that lack nos genes. Another explanation is that N₂O is produced by rhizosphere organisms from degrading nodules (51), where fixed nitrogen would become a source of N₂O emission irrespective of the nos genes in the bacteroids. If either assumption is correct, the presence of nos genes in rhizobia ought still to decrease N₂O levels in the rhizosphere and N₂O emission from field soils.

Rhizobial inoculation technology focuses on increasing legume yields. The nos gene-dependent transformation of N₂O into N₂ in soybean nodules prompted us to enhance this activity in bradyrhizobial inoculants—a technology that could be environmentally friendly. The introduction of cosmids carrying nos genes enhanced nosZ transcription and N₂O reductase activity (Fig. 2B), suggesting that up-regulation of nos gene expression would result in the enhancement of N₂O reductase activity in B. japonicum.

Natural populations of B. japonicum that lacks N₂O reductase activity were often dominant in soybean fields (44). To generate bradyrhizobial inoculants with high N₂O reductase activity in soybean nodules, one can adopt several strategies: (i) introduction of nos genes by using stable vectors or chromosome integration with strong promoters, (ii) selection of natural isolates with high N₂O reductase activity, and (iii) simple use of USDA110 and other wild-type strains carrying nos genes as inoculants. Denitrification steps, including N₂O reduction, are anaerobic respiration processes. It is thus important to examine the allocations of electrons to O₂ and N₂O in soybean nodules, because energy metabolism in bacteroids is supported by cbb₃ terminal oxidase, which has a high affinity for O₂ and adapts to a microoxic environment in root nodules (2).