Rhizobitoxine Biosynthetic Genes In Bradyrhizobium Elkanii

T. Yasuta S. Okazaki1, K. Yuhashi1, H. Ezura2, K. Minamisawa1

1 Institute of Genetic Ecology, Tohoku University, Sendai 980-8577, Japan

2Plant Biotechnology Institute, Nishi-Ibaraki, 319-0292, Japan

We cloned and sequenced a cluster of genes involved in the biosynthesis of rhizobitoxine, a nodulation enhancer produced by Bradyrhizobium elkanii. The nucleotide sequence of the cloned 28.4-kb DNA region encompassing rtxA showed that several open reading frames (ORFs) were located downstream of rtxA. A large deletion mutant of B. elkanii, USDA94Ar£x::Ql, which lacks rtxA, ORF1 (rtxC), ORF2 and ORF3, did not produce rhizobitoxine, dihydrorhizobitoxine, or serinol. The broad host-range cosmid pLAFRl, which contains rtxA and these ORFs, complemented rhizobitoxine production in USDA94Arfcc:Ql. Further complementation experiments involving cosmid derivatives randomly inserted by using a kanamycin cassette revealed that at least rtxA and rtxC are necessary for rhizobitoxine production. Insertional mutagenesis of the N-terminal and C-terminal regions of rtxA indicated that rtxA is responsible for two crucial steps, serinol formation and dihydrorhizobitoxine biosynthesis. An insertional mutant of rtxC produced serinol and dihydrorhizobitoxine, but no rhizobitoxine. Moreover, rtxC was highly homologous to the fatty acid desaturase of Pseudomonas syringae and included the copper-binding signature and eight histidine residues conserved in membrane-bound desaturase. This result suggested that rtxC encodes dihydrorhizobitoxine desaturase for the final step of rhizobitoxine production.


C3 CI D8 D2

D3 E26 E9 E10 E2 E8


C3 CI D8 D2

D3 E26 E9 E10 E2 E8

Serinol production

+++ +++ ++ ND




+++ +++

+++ +++

Dihydrorhizobitoxine production

+++ +++ ND ND



+ + +

+ + ++ +

+++ ++++

Rhizobitoxine production

T !



+ + + u—V—i

+ + ++ +

+++ ++++ >i—^-"N/





pRTFl derivatives




Figure 1. [ATI]Serinol, dihydrorhizobitoxine and rhizobitoxine production of Bradyrhizobium elkanii USDA94Artx::Ql as complemented with pRTFl cosmid derivatives created by insertion of a kanamycin cassette. The insertion point of the kanamycin cassette is indicated by the arrowhead [AT2] on pRTFl. N-domain and C-domain of the rtxA gene show the regions where their deduced amino acid sequences are homologous to aminotransferase and O-acetylhomoserine sulfhydrylase, respectively. P, putative promoter (according to the sequence).


Dept of Biology, Queen's University, Kingston, ON, K7L 3N6 Canada

1. Introduction

Legume nodules are involved in the simultaneous exchange of four gases: CO2, O2, H2 and N2. These exchanges are linked integrally through the metabolism of C, N and O within the nodule tissues. In theory, it should be possible to use a measure of nitrogenase activity and our understanding of nodule biochemistry to predict, with accuracy, the relative exchanges of C02 and

02. This report describes preliminary efforts to make these predictions and measurements.

2. Methods

A detailed mathematical model was constructed to describe the current understanding of the complex quantitative relationships that exist among bacteroid and plant metabolism in nodules. By extending an earlier study (Cen et al. 2001), the model calculated CO2 and 02 exchanges associated with N2 fixation, ATP, carbohydrate, polyhydroxybutyrate and glycogen synthesis in bacteroids, as well as the metabolism of sucrose and the synthesis of malate, ATP, carbohydrate, asparagine and ureides in the plant fraction.

Measurements of H2 production in excised soybean (Glycine max cv. Maple Glenn) nodules were combined with assumptions regarding nodule growth rate and the substrates and end products of metabolism to predict the C02 and 02 exchanges associated with metabolism. To test the model predictions, a gas exchange system was built to measure simultaneously the production of H2 and CO2 and the uptake of 02, the latter involving a novel differential oxygen analyzer (Willms et al. 1997). Respiratory quotient (RQ) was calculated as -CO2/O2 exchange.

3. Results and Discussion

The model predicted that the bacteroid fraction of nodules is highly reductive (RQ = 1.96) whereas the plant fraction is highly oxidative (RQ = 0.13). Whole nodules were predicted to have a respiratory quotient (RQ = -C02/02) of 1.08, a value significantly lower than that measured in excised soybean nodules (1.25).

To explain the major discrepancy between measured and theoretical RQ values, a sensitivity analysis of the model was conducted, based on the possibility that the model's assumptions may be incorrect. The only reasonable changes in model assumptions that were able to generate predicted RQ values similar to measured values involved the complete elimination of the ureide biosynthetic pathway. Therefore, the model predicted that short term (<7 min) exposure of a nodule to Ar:02 should result in an increase in RQ of about 0.18 units. If this does not occur, there are likely to be major problems with our current understanding of the pathway of ureide synthesis in nodules. Experiments are currently underway to test this hypothesis.

4. References

Cen Y-P, Turpin DH, Layzell DB (2001) Plant Physiol. 126

Willms J, Dowling N, Dong Z, Hunt S, Shelp B, Layzell D (1997) Anal. Biochem. 254, 272-282


A. Ureta, S. Nordlund

Dept of Biochemistry and Biophysics, Stockholm University, S-106 91 Stockholm, Sweden

1. Introduction

Acetobacter diazotrophicus is a gram-negative diazotroph, endophytic in sugarcane, where it has been shown to play an important role in the nitrogen supply required for growth without nitrogen fertilizer. To understand its diazotrophic physiology under endophytic conditions, we need to establish how it functions diazotrophically in the presence of O2 as well as the source of reductant and energy. To address this question we have studied nitrogenase activity in the presence of glucose or pyruvate as energy/reductant source under low (5%) or high (20%) O2 levels.

2. Materials and Methods

A. diazotrophicus, strain PAL5, was grown as described (Stephan et al. 1991) and diazotrophic conditions were established by growing cells with a starter of 1 mM NH4CI. Nitrogenase activity was determined as acetylene reduction. SDS-PAGE, Western blotting and protein concentration were determined by standard protocols.

3. Results and Discussion

Nitrogenase activity was supported by 0.1 M glucose under 20% and 5% O2 whereas, with pyruvate, nitrogenase activity occurred only under 5% O2. When cells were transferred to 20% O2, with pyruvate, nitrogenase activity was lost and could not be recovered at 5% 02 concentration. However, glucose addition restored nitrogenase activity. Together, these results indicate that the diversion of electrons to O2 cannot account for this inactivation and that reactivation requires a reductant that can supply electrons for nitrogenase and for respiration, generating ATP as well as lowering the intracellular 02 concentration.

Western blot analysis showed that the Fe protein is initially protected from degradation but after longer exposure to 20% 02 with pyruvate, it was degraded. This indicated that a putative conformational mechanism protecting nitrogenase is present in A. diazotrophicus. Western blot using antibodies raised against the FeSII (Shetna) protein of Azotobacter vinelandii showed cross reaction with a protein of approximately 15Kd present in both N-limited or N+ extracts of A. diazotrophicus grown either under 5% or 20% O2. The interaction of the A. diazotrophicus FeSII-like protein and both components of nitrogenase was further confirmed by co-immunoprecipitation of the Fe or MoFe protein under aerobic but not anaerobic/reducing conditions. These results show that a transient conformational protection mechanism of nitrogenase operates in A. diazotrophicus when nitrogen-fixing cells experience a sudden change in redox-state.

4. References

Moshiri F et al. (1994) Mol. Microbiol. 14, 101-114 Stephan MP et al. (1991) FEMS Microbiol. Lett. 77, 67-72

5. Acknowledgements

We are grateful to F. Moshiri for providing antibodies against the FeSII protein and an expression plasmid pAVFESII. The work was supported by grants to SN from SJFR and CTS.

Was this article helpful?

0 0

Post a comment