During Stationary Phase, Beijerinckia Derxii Shows Nitrogenase Activity Concomitant with the Release and Accumulation of Nitrogenated Substances

Microbiol. Res. (2003) 158, 309–315 http://www.urbanfischer.de/journals/microbiolres

During stationary phase, Beijerinckia derxii shows nitrogenase activity concomitant with the release and accumulation of nitrogenated substances

Natália Reiko Sato Miyasaka1, Daniela Strauss Thuler1, Eny Iochevet Segal Floh2, Walter Handro2, Mariana Braga Duarte Toledo1, Sônia Maria Gagioti3 and Heloiza Ramos Barbosa1*

1 Department of Microbiology, Institute of Biomedical Sciences, University of São Paulo. Av. Prof. Lineu Prestes, 1374, CEP- 05508-900, São Paulo, Brasil 2 Plant Cell Biology Laboratory, Department of Botany, Institute of Biosciences, University of São Paulo. Rua do Matão, 277- CEP. 05422-970, São Paulo, Brasil 3 Department of Histology and Embriology, Institute of Biomedical Sciences, University of São Paulo. Av. Prof. Lineu Prestes, 1524, 05508-900, São Paulo, Brasil

Accepted: July 22, 2003

Abstract

Beijerinckia derxii, a free-living nitrogen-fixing bacterium, the activity of their nitrogenase enzyme. Metabolic and maintained an increasing nitrogenase specific activity during genetic adaptations ensure that the nitrogen fixed by the stationary growth phase. To verify the destination of the the bacteria may be used by the microorganism itself, nitrogen fixed during this phase, intra and extracellular nitro- as free-living bacteria or by the host plant in symbio- genated contents were analyzed. Organic nitrogen and amino tic associations (Schubert, 1995). The very low NH + acids were detected in the supernatant of the cultures. An 4 increase in intracellular content of both nitrogen and protein assimilation by bacteroids (Espín et al., 1994) and the occurred. Cytoplasmic granules indicated the presence of inhibition of microbial cell division (Postgate, 1998) + arginine. The ability of a non-diazotrophic bacterium (E. coli) are features that promote NH4 excretion. Free-living to use B. derxii proteins as a source of nitrogen was observed bacteria, on the other hand, being fully able to engage in concomitantly with E. coli growth. There is a suggestion that cell division, need to incorporate fixed nitrogen into B. derxii contributes to the environment by both releasing their own structures and proteins. These microorga- nitrogenated substances and accumulating substances capable + nisms have an efficient system for NH4 assimilation. of being consumed after its death. Since low intracellular nitrogen levels are necessary for diazotrophic microorganisms to perform nitrogen fixa- Key words: Beijerinckia – N fixation – nitrogenated sub- 2 tion (Pati et al., 1994), nitrogenated substances not stances incorporated into the cell structures and enzymes must be excreted or stored as insoluble material. In general, the excretion of ammonium by free-living diazotrophs is observed in mutants that suffered physiological Introduction suppression or genetic manipulation of the enzymes involved in ammonium assimilation (Bali et al., 1992). However, Narula et al. (1981) showed that several The ability of N2-fixing bacteria to survive as free-living diazotrophs or as symbionts associated to plants, deter- strains of Azotobacter chroococcum and two strains of mines the destination of the nitrogen fixed as a result of A. vinelandii, obtained from laboratories or isolated from the soil, were able to release ammonium into the culture medium (in concentrations varying from traces –1 Corresponding author: Heloiza Ramos Barbosa to 46 µg.ml ). There is evidence that free-living N2- e-mail: [email protected] fixing bacteria release substances that may be utilized

0944-5013/03/158/04-309 $15.00/0 Microbiol. Res. 158 (2003) 4 309 by other organisms. Pati et al. (1994) showed that phyl- Cultures preparation. B. derxii inoculum was carried lospheric isolates identified as Beijerinckia indica, out in N-free LGa medium. Culture I preparation: A. chroococcum and Corynebacterium sp were able to an Erlenmeyer flask (500 mL) containing 130 mL of the release several amino acids. LGb medium was inoculated with 20 mL of a 48h One way of keeping the product of nitrogen fixation B. derxii culture (about 108 CFU.mL–1), incubated for inside the cell is to accumulate a nitrogen reserve. 74 h at 30°C in a rotary shaker (200 rpm), followed by Such nitrogen reserves were only observed in some a still incubation at 30°C, for a maximum period of cyanobacteria. During the stationary phase, and when in 550 h. Periodically, samples of the culture were taken to the presence of an excessive large source of nitrogen determine: a) Viable Cells Counts (Colony Forming (Suarez et al., 1999) and a shortage of other essential Units-CFU) using the drop method (Barbosa et al., nutrients (Newton and Tyler, 1987), these organisms 1995). Six replicates of the dilutions of the culture were produced cyanophicin, a peptide consisting of co-poly- plated on solid LGa medium. The B. derxii CFU number mers of aspartic acid and arginine. in stationary phase was confirmed by direct counting This paper deals with the destination of the products (Koch, 1994); b) Cell Protein Content, according to of N2 fixation by Beijerinckia derxii, not used for cell Lowry (1951) and Bradford (1976) methods and c) growth, for the purpose of gaining a better understand- Nitrogenase Activity, using the acetylene reduction ing of the ecological role of this free-living N2-fixing assay (Turner and Gibson, 1980). In a 384 h B. derxii bacterium. Using bacteria grown in a N-free medium, culture, arginine was tested in cell granules by the Saka- the authors seeked to establish the correlation between guchi reaction (Pearse, 1968). nitrogenase activity, i.e., the enzyme responsible for Extracellular determinations in B. derxii cultures providing fixed nitrogen to the cells, and the factors were performed in two supernatants called A and B. listed as follows: growth phase, liberation of nitro- Supernatant A, was obtained by centrifugation of the genated substances and accumulation of both N/protein culture at 12,100 x g for 30 min and filtered through inside the cells. Moreover, a model using a non-diazo- Millipore membranes (0.22 µm). Supernatant B was trophic microorganism (E. coli) was constructed in obtained by initial centrifugation of the culture at order to observe the possible consumption by this 12,100 x g for 30 min. After that, several re-suspensions bacterium of nitrogenated material, measured as pro- of cells in water and centrifugations followed, until the tein, proceeding from disrupted B. derxii cells. cells were free from a mucous layer which detached itself from the cells as a gel. The supernatant B, a pool of all centrifugations, was filtered through a 0.22 µm Millipore membrane to discard the remaining cells. Materials and methods Supernatant A was used to determine ammonium content, using the method of Chaney and Marbach (1962), glucose using the method of glucose oxidase (Henry Bacterial strains. Beijerinckia derxii, a free-living, N - 2 et al., 1974), extracellular protein using the method of fixing, bacterium, was isolated by our group from acid Bradford (1976) with bovine serum albumin (Fluka) as soil, which is also poor in nitrogen and organic matter, standard, and amino acids (only in the supernatant of a in Pirassununga, Brazil. The bacterium was identified 250 h grown culture) by the method described below. To and catalogued as ATCC 33962. Escherichia coli ICB19 determine total extracellular nitrogen (converted to was isolated from human faeces in our laboratory and NH +) in supernatant B by the method of micro-Kjeldahl identified by biochemical and morphological tests 4 (Daniels et al., 1994; Eaton et al., 1995), this prepara- (Farmer III, 1995). This bacterium was chosen to indi- tion was concentrated to a volume compatible with the cate if B. derxii nitrogenated substances might be method used (8–80 µg.ml–1). consumed as a N source. – Amino acid analysis – Fifteen ml samples from super- Culture media. LGa medium, containing (mM): natant A, previously filtered through Millipore mem- K HPO , 0,57; CaCl .2H O, 0,14; MgSO .7H O, 0,81; 2 4 2 2 4 2 branes (0.22 µm), were centrifuged (14,000 x g, for Na MoO .2H O, 0,008; KH PO , 2,2; FeCl . 6H O, 2 4 2 2 4 3 2 0.5 h) to detect amino acids (Astarita et al. 2003) by 0,037; CoCl .2H O, 0,0054 and glucose, 55,0 (pH 5.7). 2 2 High-Performance Liquid Chromatography (HPLC). For solid medium, 12 g.l–1 agar were added. An LGb medium, similar to LGa except for the phosphate Culture II preparation: another culture, called the concentration being raised to 9,6mM K2HPO4 and culture II was prepared, in which E. coli cells were cul- 40 mM KH2PO4 (pH 6.2), was also employed. Media tivated in a LGb medium to which disrupted components were supplied by Merck AG (Darmstadt) B. derxii cells were added. A sonicated diazotroph or by Difco: Nutrient broth (NB) and nutrient agar suspension (8 ml) and 125 µl of a diluted E. coli suspen- (NA). sion (about 4.107 CFU.ml–1) were added to 42 ml of an

310 Microbiol. Res. 158 (2003) 4 LGb medium. The sonicated suspension was obtained plained by the following. The idea that bacteria are as follows. A 504 h B. derxii culture was centrifuged at capable of regu-lating their metabolic reactions to 12,100 x g for 30 min at 4°C. The cells were washed achieve maximum economy and efficiency, in order to with sterile distilled water. The cells were re-suspended obtain yield of cells proportional to the amount of ATP in LGb medium and sonicated in a Branson Sonifier produced, is contradicted by the observation that 450, in an ice bath, for about 5 min. The confirmation “resting-cells suspensions” can utilize energy sources in of cell burst was done by obser-vation under an optical the complete absence of growth (Russell and Cook, microscope. A volume of 8 ml of sonicated suspen- 1995). Besides, although different genera may regulate sion was used to obtain a final protein concentration of their nitrogen fixation processes differently, there is 20 µg.ml–1. Two controls were used: an E. coli culture general agreement around the idea that the synthesis of in LGb deprived of the sonicated diazotroph suspen- nitrogenase occurs when the enzyme is operating under –1 sion and a LGb medium to which 20 µg.ml of B. der- limited amounts of ammonium and under O2 starvation xii disrupted cells protein was added. The culture II (Merrick and Edwards, 1995). Under those conditions, was incubated at 30°C in a rotary shaker (200 rpm). nitrogen fixation may occur when a source of electrons, Periodically CFU numbers (Barbosa et al., 1995) were ATP and enzyme cofactors are available to nitrogenase. determined in both culture II and E. coli control culture. It is likely that the combination of the above elements Super-natants obtained as described below were used allowed for maintenance of nitrogenase activity in the to determine protein content by the method of Bradford stationary growth phase of B. derxii. Fig. 1 shows that (1976). this bacterium, which has a mechanism to protect nitrogenase against O2 inactivation (Barbosa et al., 1992), had a sufficient source of energy (as glucose) available, and possessed a strategy to maintain intracellular Results and discussion ammonium at a low level. In nature, the genus Beije- rinckia is often found adhered to the roots of exudate- B. derxii population remained in a long stationary producing plants (Ruschel and Britto, 1966). The exu- phase, until the end of the assay, as determined by CFU dates may constitute an energy source for the countings (Fig. 1) and confirmed by total direct coun- microorganism which, in turn, might fix nitrogen and, in tings (data not shown). Glucose was only partially certain conditions, release nitrogenated substances into consumed (Fig. 1) and pH values remained stable (data the environment. The released substances can stimulate not shown). The observations indicate respectively, plant growth and consequently enhance the production that there was no limitation of carbon source and that the of plant metabolites and utilize them for their own medium remained buffered during the course of the growth (Gaudin et al., 1994). Available information evaluation. shows that the processes of N2 fixation by symbionts Figure 1 shows that nitrogenase activity was main- and by free-living bacteria differ in the number of tained throughout the duration of the assay. There was steps involved and in the specific genes and substan- continuous increase in activity and the increase ex- ces taking part in gene regulation mechanisms (Suga- tended into the stationary phase. The upkeep of nitro- numa et al., 2003). However, the upkeep of nitroge- genase activity into the stationary phase may be ex- nase activity by both B. derxii in stationary phase, and

Fig. 1. Bacterial number (CFU), specific nitrogenase activity (ARA) and glucose consumption by B. derxii, in a not shaken culture - O CFU; * Nitrogenase; Glu- cose.

Microbiol. Res. 158 (2003) 4 311 Fig. 2. Cell total nitrogen, extracellular total nitrogen (micro-Kjeldahl method) and cell protein (Bradford method) content in a not shaken B. derxii culture – extracellular total N; cell total N; cell protein

by symbiont bacteroids (Rhizobium non-multiplying, medium, completely deprived of amino acids, was used differentiated cells) may represent a common link in the as control. The excretion of amino acids by diazotrophic processes developed by those two types of bacteria. bacteria has been reported by Pati et al. (1994). The The fact that nitrogenase remained active for a long release of such compounds was considered to be an period of time after multiplication ceased raises the advantage to the nitrogenase activity, since it helps to issue of what happens to the product of N2 fixation. keep intracellular nitrogen at a low level, which is neces- Three hypotheses were tested: either the products were sary for the active N2 fixation. In addition, the release of released into the medium or they remained outside the amino acids could mean a contribution to the environ- cell and became part of the mucus layer composition, or, ment, considering that they can be easily consumed by still, they remained inside the cell as a component of a number of different organisms. new cellular structures, independent of the cell’s vital Considering that in the natural environment bacteria structures. seldom encounter conditions that permit continuous The first hypotesis was tested by analysing the extra- growth, these microorganisms can survive for extreme- cellular nitrogen concentrations, as ammonium levels ly long periods in absence of nutrients. The efficiency in (method of Chaney and Marbach, 1962). The deter- transforming nutrients into biomass indicates that bac- minations were performed in supernatant A obtained terial populations are starved most of time (Kolter et al. free of the mucus layer. The inability of B. derxii to 1993). To secure viability for longer periods, bacteria excrete ammonium (at least in concentrations above develop survival mechanisms that may involve the –1 0.5 µg.ml ) confirmed that wild-type free-living N2- accumulation of reserve materials. Storages of nitrogen, fixing bacteria normally do not excrete this substance, phosphate, carbon and energy, are examples of inclu- which is a nitrogenase inhibitor (Postgate, 1998). sions and granules that have been detected and studied The concentration of extracellular nitrogen present in in several bacteria (Perry, 1997). The intracellular supernatant B (which included the mucus layer) content in nitrogen and protein increase during the increased six times for 72 h, and remained stable until stationary phase (Fig. 2) suggest that the cells were the end of the assay (Fig. 2). Given that supernatant B probably building up a kind of nitrogenated reserve also contained the bacterial mucus layer, these results (third hypothesis). A limited coloured region, inside the suggest that B. derxii might release combined nitrogen cells, evidenced by optical microscopy, shows the posi- as a component of, or simply linked to, the extracellular tive Sakaguchi reaction (Fig. 3). The presence of such mucus. It is unlikely that this nitrogen might have been inadvertently extracted from the cell wall since the mucus was separated from the cells by a gentle method Table 1. Amino acids released by B. derxii in a 250 h N-free involving simple centrifugation. Studies on the chemical grown culture. composition of the extracellular polysaccharide of Bei- jerinckia indica (Lopez and Becking, 1968) and Amino Acids Content in µM B. mobilis (Cooke and Percival, 1975) have not revealed Glutamic acid 3.80 ± 0.598 the presence of nitrogen in its composition. Extracellu- Threonine 0.16 ± 0.215 lar proteins were not detected. Alanina 2.90 ± 0.012 B. derxii released the following amino acids to the Valine 1.13 ± 0.036 culture medium: glutamic acid, threonine, alanine, Isoleucine 0.55 ± 0.324 valine, isoleucine and leucine (Table 1). A fresh culture Leucine 0.43 ± 0.098

312 Microbiol. Res. 158 (2003) 4 Fig. 3. Photomicrography of a 384 h B. derxii culture after treatment with the Sakaguchi reaction. The arrows indicate possible arginine granules. X 2000.

Fig. 4. E. coli number (CFU) determined in N-free medium supplemented with disrupted B. derxii cells and consumption of protein in the supernatant (Brad- ford method). – E. coli number in N-free medium; E. coli number in N-free medium supplemented with sonicated cells; supernatant protein.

reserves may be a consequence of the maintenance of information was analysed using a culture II in which the nitrogenase activity during stationary phase. Simon available nitrogenated material for E. coli consisted of (1973) employed this reaction to reveal arginine in puri- ruptured B. derxii cells. Although E. coli is an unlikely fied cyanophicin granule polypeptide (CGP) from the natural partner of B. derxii, it was chosen because cyanobacterium Anabaena cylindrica. it is a non-fastidious bacterium, able to grow in simple Since preliminary results suggest the presence of an medium composed of glucose and mineral salts. Des- intracellular nitrogenous reserve, the present work pro- pite the fact that other nitrogenated compounds such as posed a model to evaluate the possibility that a non-dia- vitamins or cofactors from disrupted B. derxii could be zotrophic microorganism might utilize the component used for E. coli growth, only proteins were followed material of B. derxii’s cellular nitrogen sources. This because proteins appear in higher concentrations and are

Microbiol. Res. 158 (2003) 4 313 easier to be determined. The results showing the ability Barbosa, H. R., Alterthum, F. (1992): The role of extracellular of E. coli to consume B. derxii proteins are reported in polysaccharide in cell viability and nitrogenase activity Figure 4. E. coli presented a higher CFU number in a of Beijerinckia derxii Can J. Microbiol. 38, 986–988. culture II than in a medium not supplemented with Bradford, M. M. (1976): A rapid and sensitive method for the disrupted B. derxii cells. The material was partially quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, consumed. The consumption coincided with a rise in the 248–254. CFU number. After this increase, the population did not Chaney, A. L., Marbach, E. P. (1962): Modified reagents deter- change in number and protein concentration was stable. mination for urea and ammonia. Clin. Chem. 8, 130–132. The protein concentration of the control medium (not + Christiansen-Weniger, C., Van Veen, J. A. (1991): NH4 ex- inoculated with E. coli) remained at around 20 µg.ml–1 creting Azospirillum brasiliense mutants enhance the nitro- during the 400 hours the assay lasted. The present gen supply of a wheat host. Appl. Environm. Microbiol. 57, results indicate that E. coli produced some enzymes, 3006–3012. either sufficiently active, or at such concentrations, that Cooke, A. A., Percival, E. (1975): Structural investigations of the degradation of some proteins in the medium was the extracellular polysaccharides elaborated by Beijerinckia possible. mobilis. Carbohydr. Res. 43, 117–132. Considering the intracellular presence of this nitro- Daniels, L., Hanson, R. S., Philips, J. A. (1994): Chemical analysis. – In: Methods for General and Molecular Bac- genated material, including a possible protein reserve, teriology. (Eds: Gerhardt P., Murray R. G. E., Wood W. A. the present study supports the hypothesis proposed by and Krieg N. R.). American Society for Microbiology, Yahalom et al. (1984) and by Christiansen-Weniger and Washington, 512–554. Van Veen (1991). Accordingly, free-living, diazotrophic Eaton, A. D., Clesceri, L. S., Greenberg, A. E. (eds.) (1995): microorganisms may contribute nitrogen to the environ- Standard Methods for the Examination of Water and Waste- ment, in the form of decomposed bacterial structures. water, 19th ed. American Public Health Association, Wa- The present study also shows that some N2-fixing cell shington, D.C. proteins can be recycled by any organism that presents Espín, G., Moreno, S., Guzmán, J. (1994): Molecular genetics the specific enzymes. of the glutamine synthetases in Rhizobium species. Crit. The increase of both intra and extra-cellular organic Rev. Microbiol. 20, 117–123. nitrogen, amounting to about 66 µg.ml–1 after growing Farmer III, J. J. (1995): Enterobacteriaceae: Introduction and Identification. In: Manual of Clinical Microbiology (Eds: the bacteria for a period of 560 h (Fig. 2), suggests an Murray, P. R., Baron, E. J., Pfaller, M. A., Tenover, F. C. and important ecological role for B. derxii. It is, certainly, Yolken, R. H.) 6th ed. American Society for Microbiology, a way for the environment to increase its supply of fixed Washington, 438–449. nitrogen. At the same time, this ability gives B. derxii an Gaudin V. Vrain T., Jouanin L. (1994): Bacterial genes modi- advantage over other organisms. fying hormonal balances in plants. Plant Physiology and Biochemistry 32, 11– 29. Henry, R. J., Cannon, D. C., Winkelman, J. (1974): Clinical Acknowledgements Chemistry, Principles and Techniques, 2nd ed., Herper and Row Publishers Inc. N. Y. We thank CNPq (Brazilian National Research Counceil) and Koch, A. L. (1994): Growth Measurement. In: Methods for CAPES (Coordenação de Aperfeiçoamento de Pessoal de general and molecular bacteriology. (Eds: Gerhardt P., Ensino Superior) for the graduate and undergraduate fellow- Murray R. G. E., Wood W. A. and Krieg N. R.). American ships. Society for Microbiology, Washington, 248–277. Kolter, R., Siegele, D. A., Tormo, A. (1993): The stationary phase of the bacterial life cicle. Annu. Rev. Microbiol. 47, References 855–874. López, R. Becking, J. H. (1968): Polysaccharide production – Astarita, L., Floh, E. I. S., Handro, W. (2003): Free amino by Beijerinckia and Azotobacter. Microbiol. Espãn. 21, acid, protein, and water content changes associated with 53–75. seed development in Araucaria angustifolia. Biologia Lowry, O. H., Rosebrough, N. J., Farr, A., Randal, R. J. Plantarum (in press). (1951): Protein measurement with the folin phenol reagent. Bali, A., Blanco, G., Hill, S. and Kennedy, C. (1992): Excre- J. Biol. Chem. 193, 265–275. tion of ammonium by a nifL mutant of Azotobacter vine- Merrick, M. J., Edwards, R. A. (1995): Nitrogen control in landii fixing nitrogen. Appl. Environ. Microbiol. 58, bacteria. Microbiol. Rev. 59, 604–622. 1711–1718. Narula, N. K., Lakshminarayana, N., Tauro, P. (1981): Barbosa, H. R., Rodrigues, M. F. A., Campos, C. C., Chaves, Ammonia excretion by Azotobacter chroococcum. Bio- M. E., Nunes, I., Juliano, Y., Novo, N. F. (1995): Counting technol. Bioeng. 23, 767–770. of viable cluster forming and non-forming bacteria: a com- Newton, J. W., Tyler, D. D. (1987): Cyanophycin granule parison between the drop and the spread methods. J. Micro- polypeptide in a facultatively hetereotrophic cyanobacte- biol. Methods. 22, 39–50. rium. Curr. Microbiol. 15, 207–211.

314 Microbiol. Res. 158 (2003) 4 Pati, B. R., Sengupta, S., Chandra, A. K. (1994): Studies on polypeptide contained in blue-green alga Anabaena cylin- the amino acids released by phyllosphere diazotrophic drica. J. Bacteriol. 114, 1213–1216. bacteria. Microbiol. Res. 149, 287–290. Suarez, C., Kohler, S. J., Allen, M. M., Kolodny, N. H. (1999): Pearse, A. G. E. (1968): Histochemistry: theoretical and app- NMR study of the metabolic 15N isotopic enrichment of lied. J. & A. Churchill Ltd., London. cyanophycin synthesized by the cyanobacterium Synecho- Perry, J. J., Staley, J. T. (1997): Microbiology: Dynamics and cystis sp. Strain PCC 6308. Bioch. Biophys. Acta. 1426, Diversity. Saunders College Publishing. 429–438. Postgate, J. (1998): Nitrogen Fixation. University Press, Suganuma, N., Nakamura, Y., Yamamoto, M., Ohta, T., Koiwa, Cambridge. H., Akao, S. Kawaguchi, M. The Lotus japonicus Sen1 Ruschel, A. P., Britto, D. P. P. S. (1966): Fixação assimbiótica gene controls rhizobial differentiation into nitrogen- de nitrogênio atmosférico em algumas gramíneas e na tiri- fixing bacteroids in nodules. Mol. Gen. Genomics., DOI rica pelas bactérias do gênero Beijerinckia Derx. Pesq. 10.1007/s00438-003-0840-4. Published online: 28 March Agropec. Bras. 1, 65–69 2003. Russell, J. B., Cook, G. M. (1995): Energetics of bacterial Turner, G. L., Gibson, A. H. (1980): Measurement of nitrogen growth: balance of anabolic and catabolic reactions. Micro- fixation by indirect means. – In: Methods for evaluating biol. Rev. 59, 48–62. biological nitrogen fixation. (Ed. Bergersen, F. J.). John Schubert, K. R. (1995): Nitrogen assimilation by legumes – Wiley & Son Publ., London, 111–139. processes and ecological limitations. Fert. Res. 42, Yahalom, E., Kapulnik, Y., Okon, Y. (1984): Response of Seta- 99–107. ria italica to inoculation with Azospirillum brasilense as Simon, R. D. (1973): Measurement of cyanophycin granule compared to Azotobacter chroococcum. Pl. Soil 82, 77–85.

Microbiol. Res. 158 (2003) 4 315