Levels of Periplasmic Nitrate Reductase during Denitrification are Lower in Bradyrhizobium japonicum than in Bradyrhizobium diazoefficiens

Abstract

Soybean plants host endosymbiotic dinitrogen (N₂)-fixing bacteria from the genus Bradyrhizobium. Under oxygen-limiting conditions, Bradyrhizobium diazoefficiens and Bradyrhizobium japonicum perform denitrification by sequentially reducing nitrate (NO₃^–) to nitrous oxide (N₂O) or N₂. The anaerobic reduction of NO₃^– to N₂O was previously shown to be lower in B. japonicum than in B. diazoefficiens due to impaired periplasmic nitrate reductase (Nap) activity in B. japonicum. We herein demonstrated that impaired Nap activity in B. japonicum was due to low Nap protein levels, which may be related to a decline in the production of FixP and FixO proteins by the cbb₃-type oxidase.

Soybean is a globally important leguminous crop that generally hosts endosymbiotic dinitrogen (N₂)-fixing bacteria from the genus Bradyrhizobium (Argaw, 2014). Under oxygen-limiting conditions, some species of soybean-associated Bradyrhizobium may perform denitrification as an alternative respiratory process and sequentially reduce nitrate (NO₃^–) or nitrite (NO₂^–) to nitric oxide (NO), nitrous oxide (N₂O), and N₂ gases. Denitrification is functional in free-living bradyrhizobia at the soybean rhizosphere and in symbiotic bradyrhizobia inside the root nodules (Mesa et al., 2004; Sameshima-Saito et al., 2006b; Sánchez et al., 2011; Inaba et al., 2012).

Bradyrhizobium diazoefficiens (reclassified from Bradyrhizobium japonicum [Delamuta et al., 2013]; type strain USDA 110) may completely reduce NO₃^– to N₂. These reactions depend on napEDABC, which encodes periplasmic NO₃^– reductase (Nap); nirK, which encodes copper-containing NO₂^– reductase (NirK); norCBQD, which encodes c-type NO reductase (Nor); and nosRZDYFLX, which encodes N₂O reductase (Nos) (Sánchez et al., 2011). While B. diazoefficiens is a complete denitrifier, B. japonicum (type strain USDA 6) lacks the nos gene cluster and is, thus, unable to reduce N₂O to N₂ (Sameshima-Saito et al., 2006b; Itakura et al., 2009; Kaneko et al., 2011; Siqueira et al., 2014).

We previously screened the growth of 11 strains of B. japonicum and 15 strains of B. diazoefficiens and found that anaerobic (≈0% O₂) growth by B. japonicum with NO₃^– as the electron acceptor was significantly lower than that by B. diazoefficiens; however, no significant differences were observed between the growth rates of B. japonicum and B. diazoefficiens strains under microaerobic (≈2% O₂) and aerobic (≈12% O₂) growth conditions in the presence of NO₃^– (Siqueira et al., 2017). The lower growth rate of B. japonicum in anaerobiosis was not explained by the absence of nos, but by markedly impaired Nap activity that had a negative impact on the reduction of NO₃^– to N₂O (Siqueira et al., 2017). B. japonicum conserved the FixLJ–FixK₂ regulatory cascade, which mediates the response of the nap operon to low oxygen and NO₃^–; the binding sites for FixK/FNR (fumarate and nitrate reductase) regulators upstream of napE were also conserved (Robles et al., 2006; Bueno et al., 2017; Siqueira et al., 2017). Additionally, napA transcript levels were similar in B. japonicum and B. diazoefficiens (Siqueira et al., 2017). Thus, we proposed that impaired Nap activity in B. japonicum may be due to posttranscriptional events (Siqueira et al., 2017).

In the present study, we used the type strains B. diazoefficiens USDA 110 (Bd-USDA 110) and B. japonicum USDA 6 (Bj-USDA 6) to expand on our earlier research on the mechanisms responsible for impaired Nap activity in B. japonicum under denitrifying conditions.

Bradyrhizobium cells were precultured at 30°C in HM salt medium (Cole and Elkan, 1973) supplemented with 0.1% L-(+)-arabinose and 0.25% (w/v) yeast extract. HM medium supplemented with trace metals (Sameshima-Saito et al., 2006a) and 10‍ ‍mM KNO₃ (HMMN medium) was used in all assays. Escherichia coli cells were grown at 37°C in Luria–Bertani medium (Miller, 1972). The following antibiotics were used for the Bradyrhizobium culture: tetracycline (Tc, 100‍ ‍μg mL^–1), spectinomycin (Sp, 100‍ ‍μg mL^–1), streptomycin (Sm, 100‍ ‍μg mL^–1), and polymyxin B (50‍ ‍μg mL^–1). Tc (50‍ ‍μg mL^–1), Sp (50‍ ‍μg mL^–1), and Sm (50‍ ‍μg mL^–1) were used for the E. coli culture.

In growth experiments, precultured cells were inoculated into 34-mL test tubes containing 5‍ ‍mL of HMMN medium. The initial optical density at 660 nm (OD₆₆₀) was adjusted to 0.01. In the competition experiment, initial cell numbers were adjusted to 3×10⁷‍ ‍cells‍ ‍mL^–1 and cells were used at a 1:1 ratio. Aerobic (≈12% O₂), microaerobic (≈2% O₂), and anaerobic (≈0% O₂) treatments were prepared and monitored as previously described (Siqueira et al., 2017). Cells were grown at 30°C with reciprocal shaking at 300 rpm. Growth was monitored daily by measuring the OD₆₆₀ of the cultures, and the number of cells was directly counted using a 20-μm-deep hemocytometer (Sunlead Glass) and BX51 Fluorescence Microscope (Olympus).

To construct Bd-USDA 110 tagged with GFP and Bj-USDA 6 tagged with DsRed, pRJPaph-gfp (Ledermann et al., 2015) and pBjGroEL4::DsRed2 (Hayashi et al., 2014) plasmids were transferred by triparental mating using pRK2013 as a helper plasmid (Figurski and Helinski, 1979). Single recombination events, indicating the integration of pRJPaph-gfp into the Bd-USDA 110 chromosome or pBjGroEL4::DsRed2 into the Bj-USDA 6 chromosome, were selected based on their acquisition of resistance to Tc or Sp/Sm, respectively. Single recombinant strains were further confirmed using the BX51 Fluorescence Microscope (Olympus).

In immunoblotting and heme-staining analyses, precultured Bradyrhizobium cells were inoculated into 500-mL flasks containing 100‍ ‍mL of HMNN (OD₆₆₀≈0.03) and incubated under microaerobic or anaerobic conditions at 30°C, with reciprocal shaking at 100 rpm. Soluble and membrane fractions were prepared as described previously (Fischer et al., 2001; Torres et al., 2014a). Soluble fractions were concentrated using Amicon Ultra-15 centrifugal filter units (Merck). In immunoblotting experiments, rabbit polyclonal antibodies were raised against the synthetic NapA peptide, CPLQKKATGAAKAND (GenScript Japan), which is common between the Nap of Bd-USDA 110 and Bj-USDA 6. Soluble fractions (10‍ ‍μg) were separated by SDS-PAGE (10% polyacrylamide) and transferred to P plus PVDF membranes (ATTO) using a HorizeBLOT 4M-R semi-dry transfer system (ATTO). Membranes were incubated with an anti-NapA antibody (1:10,000) at room temperature for 30‍ ‍min and then with HRP-conjugated anti-rabbit IgG (1:2,500; Promega) for 30‍ ‍min. Immunoreactive bands were detected using the chemiluminescent substrate, Promega^TM ECL Western Blotting Substrate. Regarding heme staining, membrane fractions (10‍ ‍μg) were separated by SDS-PAGE at 4°C (stacking phase at 5% acrylamide and resolving phase at 12% acrylamide), transferred to a 0.45 μm Hybond-P membrane (GE Healthcare), and stained with the Promega^TM ECL Western Blotting Substrate to detect heme-dependent activity by chemiluminescence (Vargas et al., 1993). Blots were visualized using a Molecular Imager Gel Doc XR+ System with Image Lab software (Bio-Rad). Relative band intensities were calculated using ImageJ software (Schneider et al., 2012). Experiments were performed with at least two independent biological replicates.

The alignment of the napEDABC genome regions of Bj-USDA 6 (accession number NC_017249) and Bd-USDA 110 (accession number NC_004463) was performed using Genome Matcher version 2.2 (Ohtsubo et al., 2008). Amino acid alignment was conducted using MEGA version 7.0 and BoxShade version 3.21 (https://www.ch.embnet.org/software/BOX_form.html). A protein structural similarity analysis was performed using TM-align version 20170708 (Zhang and Skolnick, 2005).

The impaired activity of Nap in B. japonicum during anaerobic growth in the presence of NO₃^– may rely on posttranscriptional effects that affect the structure, function, or level of Nap (Siqueira et al., 2017). The napEDABC gene cluster of B. diazoefficiens encodes NapE, a membrane protein of 6.6‍ ‍kDa with unknown function; NapD (11.8‍ ‍kDa), a chaperone involved in the maturation of NapA; NapA (94.5‍ ‍kDa), the catalytic subunit that contains a Mobis-MGD active site and [4Fe4S] cluster; and NapB (16.9‍ ‍kDa), the electron transfer subunit, which is a diheme cytochrome c₅₅₂. NapA and NapB form the heterodimeric subunit NapAB, which is present in the periplasm. NapC (25‍ ‍kDa) is a membrane-anchored, tetraheme, c-type cytochrome that is responsible for the transfer of electrons from the quinol pool to the NapAB complex (Zumft, 1997; Delgado et al., 2003). A sequence comparison of the napEDABC gene cluster of Bd-USDA 110 and Bj-USDA 6 showed high identity (>85%) at the nucleotide level (Fig. S1). The amino acid sequences of Bj-USDA 6 NapEDABC were conserved, exhibiting 89–97% identity with Bd-USDA 110 (Fig. S1). In NapA of Bj-USDA 6, the twin arginine motif, the residues involved in the binding of the Mobis-MGD cofactor, and the [4Fe4S] cluster were conserved, except in some cases in which there was a substitution with a similar residue. In Bj-USDA 6 NapB, both the cytochrome c-binding sites and motif for translocation via the general secretory pathway were conserved. The structural similarity of the Bj-USDA 6 and Bd-USDA 110 Nap proteins was high, with a TM-score of 0.81, which indicated the same type of folding (Fig. S1). These results suggested that the structure and function of Nap were conserved in Bj-USDA 6.

We compared the amount of Nap in Bj-USDA 6 and Bd-USDA 110 cells incubated under microaerobic and anaerobic conditions in the presence of NO₃^–. An immunoblotting analysis of soluble fractions revealed a bradyrhizobial NapA protein band at ≈100‍ ‍kDa (Fig. 1A), which was consistent with the expected molecular weight of 94.5‍ ‍kDa (Delgado et al., 2003). This was further confirmed by the absence of this band in soluble fractions of the B. diazoefficiens ΔnapA mutant (Fig. 1A). The intensity of the NapA band was weaker in Bj-USDA 6 than in Bd-USDA 110, in cells incubated microaerobically (≈0.57 times; Fig. 1A and Table S1), and particularly in cells incubated anaerobically (≈0.33 times; Fig. 1A and Table S1). This result suggested that Bj-USDA 6 produced less NapA than Bd-USDA 110 under low-oxygen conditions (i.e. ≤2% O₂) in the presence of NO₃^–. Additionally, we detected NapC in the membrane fractions of bradyrhizobial cells by heme staining; the staining of heme covalently bound to c-type cytochromes (Vargas et al., 1993). Bj-USDA 6 and Bd-USDA 110 showed the typical profile of five stained bands previously identified in membrane fractions of Bd-USDA 110 (Fig. 1B) (Preisig et al., 1993; Mesa et al., 2002; Delgado et al., 2003; Bueno et al., 2008; Torres et al., 2014a). The 32-kDa and 28-kDa bands, almost co-migrating, corresponded to the FixP and FixO proteins, respectively, of cbb₃-type high affinity cytochrome oxidase, encoded by the fixNOQP operon (Preisig et al., 1993, 1996); the 25-kDa band corresponded to NapC (Delgado et al., 2003); the 20-kDa band corresponded to cytochrome CycM (Bott et al., 1991); and the 16-kDa band corresponded to the NorC subunit of Nor (Mesa et al., 2002). We confirmed the identities of bands by analyzing soluble fractions of the B. diazoefficiens ΔnapA mutant under microaerobic conditions; bands correspondig to NapC and NorC were absent (Fig. 1B), as described previously (Delgado et al., 2003). Under anaerobic conditions, the intensity of the NapC band was ≈0.3-fold weaker in Bj-USDA 6 than in Bd-USDA 110 (Fig. 1B and Table S1). Thus, immunoblotting and heme-staining results indicated that Bj-USDA 6 produced lower levels of Nap under low-oxygen conditions in the presence of NO₃^–. Therefore, impaired Nap activity in B. japonicum during NO₃^–-dependent anaerobic growth, which was reported previously (Siqueira et al., 2017), may be caused by a low level of Nap.

Fig. 1.

Detection of Nap in Bradyrhizobium diazoefficiens USDA 110 (Bd) and Bradyrhizobium japonicum USDA 6 (Bj) in HMMN medium under microaerobic and anaerobic conditions. (A) Immunoblotting analysis of NapA in soluble fractions (10‍ ‍μg). (B) Heme staining of membrane fractions (10‍ ‍μg). In panels A and B, the B. diazoefficiens USDA 110 ΔnapA mutant grown in HMMN medium under microaerobic conditions is shown as the control.

Based on the results of heme staining, the amount of NorC was higher in Bj-USDA 6 than in Bd-USDA 110 under microaerobic and anaerobic conditions (≈1.3- and 1.6-fold, respectively) in the presence of NO₃^– (Fig. 1B and Table S1). This result was supported by a previous finding showing that norB transcript levels were higher in Bj-USDA 6 than in Bd-USDA 110 under anaerobic NO₃^–-dependent growth (Siqueira et al., 2017). Heme staining also revealed that besides NapC, FixP and FixO were the only other cytochromes affected in Bj-USDA 6 under low-oxygen conditions; the amounts of FixP and FixO were lower in Bj-USDA 6 than in Bd-USDA 110 by ≈0.45-fold under microaerobic conditions and ≈0.8-fold under anaerobic conditions (Fig. 1B and Table S1). This result suggested that cbb₃ oxidase is a critical enzyme for the adaptation process to NO₃^– respiration under low-oxygen conditions in Bradyrhizobium. Accordingly, the maximal expression of Nap does not appear to occur until the oxygen concentration becomes very low, and this is only observed after cbb₃ oxidase has consumed the oxygen present in the growth medium (Bueno et al., 2008). Since the transcription of nap genes appeared to be unaffected in B. japonicum cells grown anaerobically in the presence of NO₃^– (Siqueira et al., 2017), the remaining oxygen in the growth medium may affect the amount of the Nap protein at the posttranscriptional level by inhibiting the translation of nap messenger RNA or inducing the degradation of Nap proteins.

The mechanisms responsible for low levels of cbb₃ oxidase in B. japonicum under low-oxygen conditions in the presence of NO₃^– may be related to B. japonicum being unable to make an effective switch to denitrification in the absence of oxygen, similar to other bacteria (Aida et al., 1986; Bergaust et al., 2011; Torres et al., 2014b; Siqueira et al., 2017). We tested the NO₃^–-dependent anaerobic growth of bradyrhizobial cells preincubated microaerobically, instead of aerobically as previously reported (Siqueira et al., 2017). The microaerobic preincubation resulted in a reduced growth rate of Bj-USDA 6, to a similar extent as the aerobic preincubation (Fig. S2), indicating the absence of a significant difference between a rapid (aerobic preincubation) and gradual (microaerobic preincubation) transition from aerobiosis to anaerobiosis. However, further studies are needed to elucidate the mechanisms responsible for the changes observed in cbb₃ oxidase levels in B. japonicum under denitrifying conditions.

Our previous findings prompted the hypothesis that B. japonicum may be less competitive than B. diazoefficiens due to energy depletion under anaerobic denitrifying growth (Siqueira et al., 2017). We examined the competitive growth of Bd-USDA 110 and Bj-USDA 6 tagged with GFP and DsRed proteins, respectively. The tagged strains were tested to exclude the possible effects of tagging on growth rates (Fig. S3). Under microaerobic and anaerobic conditions, in the presence of NO₃^–, the growth of DsRed-tagged Bj-USDA 6 was lower than that of GFP-tagged Bd-USDA 110, whereas the aerobic growth rates of these two species were similar (Fig. 2). This result indicated that Bd-USDA 110 is more competitive than Bj-USDA 6 at low-oxygen levels in the presence of NO₃^–. In Bradyrhizobium, growth at low-oxygen levels in the presence of NO₃^– is supported by both cbb₃ oxidase, which is active during free-living microaerobic growth and in N₂-fixing bacteroids, and the denitrification pathway (Preisig et al., 1993, 1996; Bueno et al., 2008). The weaker competitiveness of B. japonicum USDA 6 under low-oxygen conditions in the presence of NO₃^– may correlate with low levels of Nap (Fig. 1 and Table S1).

Fig. 2.

Average number of cells in co-cultures (1:1) of Bradyrhizobium diazoefficiens USDA 110 tagged with GFP (white bars) and Bradyrhizobium japonicum USDA 6 tagged with DsRed (grey bars) under anaerobic, microaerobic, and aerobic conditions in HMMN medium. Error bars indicate SE. * Values significantly different between USDA 110 and USDA 6 (t-test, P<0.05; n=3).

In conclusion, the present study demonstrated that the level of Nap under low-oxygen conditions in the presence of NO₃^– was lower in B. japonicum USDA 6 than in B. diazoefficiens USDA 110 and suggests that the capacity to maintain a sufficient quantity of Nap is an advantage under denitrifying conditions. The lower levels of Nap in B. japonicum USDA 6 presumably resulted in a competitive disadvantage against B. diazoefficiens USDA 110 due to energy depletion during denitrifying growth. Ecologically, this advantage of B. diazoefficiens may be an important factor that influences the predominance of B. diazoefficiens over B. japonicum in soils exposed to low-oxygen conditions (Shiina et al., 2014).

Citation

Siqueira, A. F., Sugawara, M., Arashida, H., Minamisawa, K., and Sánchez, C. (2020) Levels of Periplasmic Nitrate Reductase during Denitrification are Lower in Bradyrhizobium japonicum than in Bradyrhizobium diazoefficiens. Microbes Environ 35: ME19129.

https://doi.org/10.1264/jsme2.ME19129

Acknowledgements

We thank Hans-Martin Fischer (Institute of Microbiology, ETH Zürich, Switzerland) and Hiroshi Kouchi (International Christian University, Tokyo, Japan) for kindly providing the plasmids, pRJPaph-gfp and pBjGroEL4::DsRed2, respectively. This work was supported by Grants-in-Aid for Scientific Research (A) 26252065 and (B) 18H02112 from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and by grants from the Project of the NARO Bio-oriented Technology Research Advancement Institution (Research Program on Development of Innovative Technology). A.F.S. was supported by fellowships from the National Council for Scientific and Technological Development-Brazil and the Japan Society for the Promotion of Science.

References

•  Aida,  T.,  Hata,  S., and  Kusunoki,  H. (1986) Temporary low oxygen conditions for the formation of nitrate reductase and nitrous oxide reductase by denitrifying Pseudomonas sp. G59. Can J Microbiol 32: 543–547.
•  Argaw,  A. (2014) Symbiotic effectiveness of inoculation with Bradyrhizobium isolates on soybean [Glycine max (L.) Merrill] genotypes with different maturities. SpringerPlus 3: 753.
•  Bergaust,  L.,  Bakken,  L.R., and  Frostegård,  Å. (2011) Denitrification regulatory phenotype, a new term for the characterization of denitrifying bacteria. Biochem Soc Trans 39: 207–212.
•  Bott,  M.,  Ritz,  D., and  Hennecke,  H. (1991) The Bradyrhizobium japonicum cycM gene encodes a membrane-anchored homolog of mitochondrial cytochrome c. J Bacteriol 173: 6766–6772.
•  Bueno,  B.,  Robles,  E.F.,  Torres,  M.J.,  Krell,  T.,  Bedmar,  E.J.,  Delgado,  M.J., and  Mesa,  S. (2017) Disparate response to microoxia and nitrogen oxides of the Bradyrhizobium japonicum napEDABC, nirK and norCBQD denitrification genes. Nitric Oxide 6: 1–13.
•  Bueno,  E.,  Bedmar,  E.J.,  Richardson,  D.J., and  Delgado,  M.J. (2008) Role of Bradyrhizobium japonicum cytochrome c550 in nitrite and nitrate respiration. FEMS Microbiol Lett 279: 188–194.
•  Cole,  M.A., and  Elkan,  G.H. (1973) Transmissible resistance to penicillin G, neomycin, and chloramphenicol in Rhizobium japonicum. Antimicrob Agents Chemother 4: 248–253.
•  Delamuta,  J.R.,  Ribeiro,  R.A.,  Ormeño-Orrillo,  E.,  Melo,  I.S.,  Martínez-Romero,  E., and  Hungria,  M. (2013) Polyphasic evidence supporting the reclassification of Bradyrhizobium japonicum group Ia strains as Bradyrhizobium diazoefficiens sp. nov. Int J Syst Evol Microbiol 63: 3342–3351.
•  Delgado,  M.J.,  Bonnard,  N.,  Tresierra-Ayala,  A.,  Bedmar,  E.J., and  Müller,  P. (2003) The Bradyrhizobium japonicum napEDABC genes encoding the periplasmic nitrate reductase are essential for nitrate respiration. Microbiology 149: 3395–3403.
•  Figurski,  D.H., and  Helinski,  D.R. (1979) Replication of an origin-containing derivative of plasmid Rk2 dependent on a plasmid function provided in trans. Proc Natl Acad Sci U S A 76: 1648–1652.
•  Fischer,  H.M.,  Velasco,  L.,  Delgado,  M.J.,  Bedmar,  E.J.,  Schären,  S.,  Zingg,  D., et al. (2001) One of two hemN genes in Bradyrhizobium japonicum is functional during anaerobic growth and in symbiosis. J Bacteriol 183: 1300–1311.
•  Hayashi,  M.,  Shiro,  S.,  Kanamori,  H.,  Mori-Hosokawa,  S.,  Sasaki-Yamagata,  H.,  Sayama,  T., et al. (2014) A thaumatin-like protein, Rj4, controls nodule symbiotic specificity in soybean. Plant Cell Physiol 55: 1679–1689.
•  Inaba,  S.,  Ikenishi,  F.,  Itakura,  M.,  Kikuchi,  M.,  Eda,  S.,  Chiba,  N., et al. (2012) N₂O emission from degraded soybean nodules depends on denitrification by Bradyrhizobium japonicum and other microbes in the rhizosphere. Microbes Environ 24: 470–476.
•  Itakura,  M.,  Saeki,  K.,  Omori,  H.,  Yokoyama,  T.,  Kaneko,  T.,  Tabata,  S., et al. (2009) Genomic comparison of Bradyrhizobium japonicum strains with different symbiotic nitrogen-fixing capabilities and other Bradyrhizobiaceae members. ISME J 3: 326–339.
•  Kaneko,  T.,  Maita,  H.,  Hirakawa,  H.,  Uchiike,  N.,  Minamisawa,  K.,  Watanabe,  A., and  Sato,  S. (2011) Complete genome sequence of the soybean symbiont Bradyrhizobium japonicum strain USDA6^T. Genes 2: 763–787.
•  Ledermann,  R.,  Bartsch,  I.,  Remus-Emsermann,  M.N.,  Vorholt,  J.A., and  Fischer,  H.M. (2015) Stable fluorescent and enzymatic tagging of Bradyrhizobium diazoefficiens to analyze host-plant infection and colonization. Mol Plant Microbe Interact 28: 959–967.
•  Mesa,  S.,  Velasco,  L.,  Manzanera,  M.E.,  Delgado,  M.J., and  Bedmar,  E.J. (2002) Characterization and regulation of the nitric oxide reductase-encoding region of Bradyrhizobium japonicum. Microbiology 148: 3553–3560.
•  Mesa,  S.,  Alché,  J.D.,  Bedmar,  E.J., and  Delgado,  M.J. (2004) Expression of nir, nor and nos denitrification genes from Bradyrhizobium japonicum in soybean root nodules. Physiol Plant 120: 205–212.
• Miller, J.H. (1972) Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
•  Ohtsubo,  Y.,  Ikeda-Ohtsubo,  W.,  Nagata,  Y., and  Tsuda,  M. (2008) Genome-Matcher: a graphical user interface for DNA sequence comparison. BMC Bioinformatics 16: 9: 376.
•  Preisig,  O.,  Anthamatten,  D., and Hennecke H. (1993) Genes for a microaerobically induced oxidase complex in Bradyrhizobium japonicum are essential for a nitrogen-fixing endosymbiosis. Proc Natl Acad Sci U S A 90: 3309–3313.
•  Preisig,  O.,  Zufferey,  R.,  Thöny-Meyer,  L.,  Appleby,  C.A., and  Hennecke,  H. (1996) A high-affinity cbb₃-type cytochrome oxidase terminates the symbiosis-specific respiratory chain of Bradyrhizobium japonicum. J Bacteriol 178: 1532–1538.
•  Robles,  E.F.,  Sánchez,  C.,  Bonnard,  N.,  Delgado,  M.J., and  Bedmar,  E.J. (2006) The Bradyrhizobium japonicum napEDABC genes are controlled by the FixLJ-FixK₂-NnrR regulatory cascade. Biochem Soc Trans 34: 108–110.
•  Sameshima-Saito,  R.,  Chiba,  K.,  Hirayama,  J.,  Itakura,  M.,  Mitsui,  H.,  Eda,  S., and  Minamisawa,  K. (2006a) Symbiotic Bradyrhizobium japonicum reduces N₂O surrounding the soybean root system via nitrous oxide reductase. Appl Environ Microbiol 72: 2526–2532.
•  Sameshima-Saito,  R.,  Chiba,  K., and  Minamisawa,  K. (2006b) Correlation of denitrifying capability with the existence of nap, nir, nor and nos genes in diverse strains of soybean bradyrhizobia. Microbes Environ 21: 174–184.
• Sánchez, C., Bedmar, E.J., and Delgado, M.J. (2011) Denitrification in legume-associated endosymbiotic bacteria. In Nitrogen Cycling in Bacteria: Molecular Analysis. Moir, J.W.B. (ed.) Norfolk: Caister Academic Press, pp. 197–210.
•  Schneider,  C.A.,  Rasband,  W.S., and  Eliceiri,  K.W. (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9: 671–675.
•  Shiina,  Y.,  Itakura,  M.,  Choi,  H.,  Saeki,  Y.,  Hayatsu,  M., and  Minamisawa,  K. (2014) Relationship between soil type and N₂O reductase genotype (nosZ) of indigenous soybean bradyrhizobia: nosZ-minus populations are dominant in Andosols. Microbes Environ 29: 420–426.
•  Siqueira,  A.F.,  Ormeño-Orrillo,  E.,  Souza,  R.C.,  Rodrigues,  E.P.,  Almeida,  L.G.,  Barcellos,  F.G., et al. (2014) Comparative genomics of Bradyrhizobium japonicum CPAC 15 and Bradyrhizobium diazoefficiens CPAC 7: elite model strains for understanding symbiotic performance with soybean. BMC Genomics 15: 420.
•  Siqueira,  A.F.,  Minamisawa,  K., and  Sánchez,  C. (2017) Anaerobic reduction of nitrate to nitrous oxide is lower in Bradyrhizobium japonicum than in Bradyrhizobium diazoefficiens. Microbes Environ 32: 398–401.
•  Torres,  M.J.,  Argandoña,  M.,  Vargas,  C.,  Bedmar,  E.J.,  Fischer,  H.M.,  Mesa,  S., and  Delgado,  M.J. (2014a) The global response regulator RegR controls expression of denitrification genes in Bradyrhizobium japonicum. PLoS One 9: e99011.
•  Torres,  M.J.,  Rubia,  M.I.,  Coba de la Peña,  T.,  Pueyo,  J.J.,  Bedmar,  E.J., and  Delgado,  M.J. (2014b) Genetic basis for denitrification in Ensifer meliloti. BMC Microbiol 14: 142.
•  Vargas,  C.,  McEwan,  A.G., and  Downie,  J.A. (1993) Detection of c-type cytochromes using enhanced chemiluminescence. Anal Biochem 209: 323–326.
•  Zhang,  Y., and  Skolnick,  J. (2005) TM-align: A protein structure alignment algorithm based on TM-score. Nucleic Acids Res 33: 2302–2309.
•  Zumft,  W.G. (1997) Cell biology and molecular basis of denitrification. Microbiol Mol Biol Rev 61: 533–616.