Daidzein and coumestrol concentrations in the root exudates of the common bean were found to be increased in the presence of Rhizobium spp. (2). No effects ...
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 1996, p. 3030–3033 0099-2240/96/$04.0010 Copyright q 1996, American Society for Microbiology
Vol. 62, No. 8
Promotion of nod Gene Inducers and Nodulation in Common Bean (Phaseolus vulgaris) Roots Inoculated with Azospirillum brasilense Cd SAUL BURDMAN,1 HANNE VOLPIN,1 JAIME KIGEL,1 YORAM KAPULNIK,2
AND
YAACOV OKON1*
Faculty of Agriculture, The Hebrew University of Jerusalem, Rehovot 76100,1 and Institute of Field and Garden Crops, Agricultural Research Organization, The Volcani Center, Bet Dagan 50250,2 Israel Received 6 November 1995/Accepted 28 May 1996
Inoculation of Phaseolus vulgaris with Azospirillum brasilense Cd promoted root hair formation in seedling roots and significantly increased total and upper nodule numbers at different concentrations of Rhizobium inoculum. In experiments carried out in a hydroponic system, A. brasilense caused an increase in the secretion of nod gene-inducing flavonoids, as was observed by nod gene induction assays of root exudates fractionated by high-performance liquid chromatography. Possible mechanisms involved in the influence of A. brasilense on this symbiotic system are discussed. Roux flasks. Cells were extracted by adding rolling sterile glass beads to the agar and then suspending the bacteria in 0.1 M phosphate buffer. Azospirillum brasilense Cd (ATCC 29729) was grown on liquid malate minimal medium (13) for 24 h at 308C. Cells were washed, harvested by centrifugation (1,000 3 g, twice for 10 min each time), and resuspended for inoculation in 0.1 M phosphate buffer (pH 6.8) for pot experiments or in nutrient solution for hydroponic systems. R. etli UBP102 was obtained from Carmen Quinto (Universidad Nacional Autonoma de Mexico) and used for nod gene induction assays. This bacterium contains a nodA-lacZ fusion on plasmid p42d (22). Cultures were maintained on YMA slants with 12 mg of chloramphenicol per ml and were transferred to liquid YMA at least 16 h prior to assay. Pot experiments. Seeds of P. vulgaris cv. Bulgarian (Gedera Seeds Co., Gedera, Israel) were surface sterilized for 2 min in 95% ethyl alcohol and then for 1 min in 1% sodium hypochlorite and were washed five times with sterile water. Nodulation experiments were performed under gnotobiotic conditions. Pots with a capacity of 1.5 liters were lined with polyester bags, and a 2- to 3-cm layer of tuff (volcanic gravel) was added. A glass tube (17-cm length) extending 7 cm above the rim of the pot was placed above the tuff layer, and 1.4 kg of sand containing 1 g (each) of Ca(PO4)2 and K2SO4 per kg of sand was poured around the glass tube, filling the pot. Jensen’s nutrient solution (24) was added (at 200 ml per pot), and the pots were then autoclaved for 1 h. Two surface-sterilized seeds were sown in each pot, and inoculation was performed with a mixture of R. tropici CIAT899 and R. etli TAL182 at 0, 101, 102, 103, 104, 105, and 106 CFU/ml, with or without A. brasilense at 108 CFU/ml. One milliliter of inoculum per seed was applied. The sand surface then was covered with a layer of approximately 3 cm of dry sterilized perlite. The plants were grown in a greenhouse on a cycle of 16 h of light (at 228C) and 8 h of dark (at 178C). Irrigation with sterile water or Jensen’s solution was done via the glass tubes that penetrated to the bottom of the pot, thus keeping the perlite layer dry and consequently preventing the transfer of contaminating bacteria. Seedlings were thinned out to one per pot at 1 week after emergence, and the plants (eight replicates per treatment) were harvested 3 weeks later. Both the total nodules per plant and the upper nodules (those present in the upper 2 cm of the crown roots) were counted. Four pot experiments, with some modifications, were performed. The results obtained were very similar. One
Bacteria of the genus Rhizobium interact with the roots of leguminous plants, leading to the development of symbiotic nitrogen-fixing nodules (9). The roots of the common bean (Phaseolus vulgaris) are nodulated by Rhizobium tropici (10) and Rhizobium etli (20). However, in contrast to other legumes, the common bean is an inefficient nitrogen fixer. Several factors, both genetic and environmental, contribute to this poor performance. The nodulation process in the common beanRhizobium symbiosis is relatively slow (1, 19). Consequently, in bean cultivars with a short growth cycle, biological nitrogen fixation does not contribute significantly to the N nutrition of the plant (1, 7, 19). An initial stage in nodule formation is the transcription of the common nodulation genes, nodABC, in Rhizobium spp. This process is triggered by the interaction between flavonoid compounds from root exudates and the protein product of the regulatory nodD gene of the bacterium (23). Several nod gene inducers from common bean root exudates have been identified as isoflavone and flavanone molecules (2, 5, 22). Free-living nitrogen-fixing bacteria of the genus Azospirillum live in close association with plants in the rhizosphere and are beneficial to plant growth and yield (16). These effects are attributed primarily to improved root development and enhanced water and mineral uptake (15). There is some evidence that secretion of plant-growth-promoting substances by the bacteria is at least partially responsible for these effects (3). The positive effects of combined inoculation with Azospirillum and Rhizobium spp. have been reported for several legumes (6, 8, 17, 18). The positive response of dually inoculated plants and the enhanced growth of such plants may be attributed to early nodulation, an increase in the number of nodules, and a general improvement in root development (26). The objective of this research was to study the effects of Azospirillum application on nodulation of the common bean and to investigate possible mechanisms involved in this interaction, such as the influence of Azospirillum spp. on the secretion of nod gene induction flavonoids by the roots. Bacteria and growth conditions. R. tropici CIAT899 (UMR1899) and R. etli TAL182 (SEMIA4021) were obtained from Eduardo Schroder (University of Puerto Rico). For inoculation treatments in pot experiments, cultures were grown for 72 h on yeast mannitol agar (YMA; pH 6.8) (24) at 308C in
* Corresponding author. 3030
VOL. 62, 1996
NOTES
3031
TABLE 1. Effects of A. brasilense (108 CFU/ml) on total and upper nodule numbers on bean roots at different concentrations of Rhizobium spp.a Rhizobium concn (CFU/ml)
101 102 103 104 105 106 Average
Total no. of nodules per plant Without A. brasilense
With A. brasilense
227 6 11 216 6 17 233 6 10 238 6 11 233 6 17 191 6 17
260 6 23 262 6 10 261 6 22 259 6 15 252 6 12 266 6 25
223
260
Change (%)b
No. of upper nodules per plantc
Change (%)b
Without A. brasilense
With A. brasilense
14.5 21.3 12.0 8.8 8.1 39.2
63 6 4 61 6 9 80 6 10 80 6 9 79 6 5 59 6 6
63 6 7 79 6 11 86 6 8 92 6 5 83 6 8 75 6 5
29.5 7.5 15.0 5.1 27.1
17.3
70
80
14.0
a
Plants were grown in 1.5-liter pots, and nodules were counted 4 weeks after emergence. No nodules were formed in controls without Rhizobium spp. Each value represents the average number 6 standard error of nodules per plant (eight replicates per treatment). b Percent change from without to with A. brasilense. Two-way analysis of variance indicates significant differences at P 5 0.05 for treatments with and without A. brasilense for both variables. No significant differences were found for different concentrations of Rhizobium inoculum. c Upper nodules are those formed in the upper 2 cm of the crown root.
representative experiment is reported. Results for each experiment were subjected to a two-way analysis of variance. Plants that were not inoculated with Rhizobium spp. did not develop nodules. A. brasilense enhanced nodule formation at all concentrations of Rhizobium inoculum (Table 1). Coinoculation with Rhizobium spp. and A. brasilense caused a significant increase (17.3%) in the total number of nodules per plant, and there was a 14.2% increase in the number of upper nodules (an average of different Rhizobium levels) by comparison with the number for plants treated with Rhizobium spp. alone (Table 1). No significant differences in the numbers of total and upper nodules were found for the different concentrations of Rhizobium inoculum. This result may indicate that a low concentration of Rhizobium inoculum is enough for optimal nodulation of this bean cultivar in pots under gnotobiotic conditions. In the field, where the common bean is regarded as a poor nodulating and N-fixing legume (19), the positive effects of A. brasilense on nodulation may be more advantageous. Earlier nodulation and increases in total nodule number have been reported for legumes inoculated with A. brasilense (6, 17, 18, 21). An increase in the number of upper nodules (those present in the upper 2 cm of the roots) could indicate early nodulation (upper nodules form earlier than others) and/or a greater susceptibility of the roots to nodulation by Rhizobium spp. (14). The effect of A. brasilense inoculation on root hair formation in seedling roots was studied in 0.6-liter pots. Surface-sterilized seeds were sown (9 seeds per pot) in fine sand, and 15 ml of A. brasilense inoculum (107 CFU/ml) was added. An equal volume of 0.1 M phosphate buffer (pH 6.8) was added to uninoculated controls. The pots were kept on a cycle of 16 h of light (at 228C) and 8 h of dark (at 178C). Seedlings were harvested 96 h after the seeds were sown. The roots (3 to 3.5 cm in length) were washed, and root hairs were counted with a light microscope (3100). A. brasilense significantly increased the root hair density of the roots of bean seedlings. This increase was on the order of 66% (an average of 25 root hairs per mm, compared with 15 root hairs per mm for uninoculated controls). These results are consistent with previous work with alfalfa (8) and tomatoes (4) as well as with several cereals and grasses (15). Rhizobium infection takes place by the formation of infection threads in root hairs (9). The enhanced susceptibility of bean plants to Rhizobium infection following Azospirillum inoculation could be caused by the differentiation of a greater number of epidermal cells into infectible root hairs (14, 27). Hydroponic experiments. Twenty-five seeds of P. vulgaris
were surface sterilized and placed on a steel screen covered with sterile cheesecloth which was placed over 350 ml of sterile, aerated, 10-fold-diluted Jensen’s solution in a 400-ml plastic container. A sterile clear plastic cover was positioned over the container, and roots developed into the nutrient solution (11). The containers were maintained under an illumination of 40 microeinsteins z m22 z s21 and on a cycle of 16 h of light (at 258C) and 8 h of dark (at 208C). Nutrient solutions were changed 96 h after the sowing, and treatments were then initiated. A. brasilense was added to the nutrient solution to a final concentration of 107 CFU/ml in the inoculated containers. Untreated controls received an equal amount of nutrient solution. Liquid media containing root exudates were collected and freeze-dried after 48 h. For the purification of nod gene inducers, freeze-dried root exudates were dissolved in 15 ml of high-performance liquid chromatography (HPLC)-grade water (Merck), centrifuged at 6,000 3 g for 20 min, and passed through 0.2-mm-pore-size polycarbonate filters (Schleicher & Schuell). The supernatant was adsorbed into C18 Maxi-Clean Cartridges (Alltech Associates Inc.). Flavonoids were eluted with 5 ml of acetone, concentrated to a final volume of 3 ml with a N2 gas stream at 258C (11), and then divided into 1-ml samples which were dried under vacuum. For HPLC separation, the samples were dissolved in 120 ml of 50% methanol and sonicated for 30 s. Aliquots (100 ml each) were loaded onto a Waters HPLC system (Millipore Corp.) fitted with a 250-by-4-mm LiChrospher 100 RP-18 column and were eluted at 0.5 ml/min from 0 to 10 min with a linear gradient of water-methanol-acetic acid (from 69:30:1 to 49:50:1, vol/vol/vol) and from 10 to 60 min with a linear gradient of methanol-acetic acid (99:1, vol/vol). The analysis continued isocratically at this solvent composition for another 20 min. Eluting compounds were monitored every 2.01 s with a Waters 990 photodiode array detector with an absorption spectrum of 240 to 400 nm and a 1.4-nm resolution. Eluant fractions were collected every 2 min and vacuum dried. The nod gene induction capacity of the root exudates and HPLC fractions was assayed as b-galactosidase activity with R. etli UBP102. Assays were carried out by methods previously described (12). Genistein at 0.2 mM was used as a positive control, and pure nutrient solutions with and without A. brasilense were used as negative controls. Total purified nutrient solution at 0.1% from every container was added to the assay. The background observed in the negative controls was subtracted from the treatments. The hydroponic experiments were carried out twice, each time with three replicates per
3032
NOTES
FIG. 1. HPLC characteristics (A240–400) (top panels) and nod gene induction assays (bottom panels) of fractions from root exudates of 6-day-old bean seedlings inoculated or not inoculated with A. brasilense at 107 CFU/ml. Root exudates were collected at 48 h after inoculation. For assays, fractions were collected every 2 min. The means of three replicates for each treatment of a representative experiment are shown.
treatment (each replicate containing roots from 25 seedlings). The results were very similar; therefore, we report on one of them. Inoculation with A. brasilense caused a threefold increase in the nod gene induction activity of crude root exudates of 6-dayold seedlings (an average of 133 b-galactosidase units compared with 33 units for untreated controls). Exudates of roots inoculated with A. brasilense showed higher levels of flavonoid compounds than did those of controls (Fig. 1). Nine peaks were located in fractions that showed nod induction activity. Some of the peaks (e.g., 1, 3, and 9 [Fig. 1]) were observed only in the root exudates of inoculated seedlings. Peaks 4 to 8 were clearly increased by factors of 14.5, 3.7, 4.5, 1.7, and 3.8, respectively, by Azospirillum inoculation. Thus, A. brasilense increased the number of active fractions, and fractions from Azospirillum-treated roots generally showed higher nod induction activity than comparable fractions from controls (Fig. 1). The more highly active fractions from both treatments were located between 25 and 35 min (retention time), corresponding to peaks 5 to 7. On the basis of UVvisible spectra, retention time, and nod gene induction activities, it was determined that these peaks cochromatographed with daidzein, naringenin, and genistein, respectively. Peak 8 was located in a slightly active fraction, and it cochromatographed with the phytoalexin coumestrol. More rigorous identification by gas chromatography-mass spectrometry procedures is needed. These compounds were previously reported to be found in bean root exudates (2, 5). Daidzein and coumestrol concentrations in the root exudates of the common bean were found to be increased in the presence of Rhizobium spp. (2). No effects on nod gene induction activity (above background values) were observed in plant-free pure nutrient solution with or without A. brasilense (results not shown). This result indicates that the increase in nod gene induction in inoculated seedlings was derived from the enhanced secretion of flavonoids by the roots. The results of this study show that A. brasilense enhances nodulation in common bean roots. We propose that this increase could be explained, at least in part, by the promotive effects of the bacterium on root hair formation and by an increase in the secretion of nod gene inducer signals by the
APPL. ENVIRON. MICROBIOL.
roots. The first effect is a well-demonstrated effect of A. brasilense on plants. To the best of our knowledge, however, there are no previous reports of A. brasilense affecting the secretion of nod gene inducers. In a parallel study with alfalfa, different flavonoid profiles were observed after the root exudates of A. brasilense- and yeast extract-treated plants were compared, indicating that the changes in root metabolism caused by A. brasilense are not a defense-like response (26). The activities of phenylalanine ammonia-lyase and chalcone isomerase and the mRNA levels of these enzymes were not induced by A. brasilense (25). We cannot conclude from this work that the observed increase in nod gene signal secretion is a specific effect of A. brasilense. In one of these experiments, we added a Pseudomonas fluorescens inoculation treatment (strain 94.61, obtained from Alicia Arias, Instituto de Investigacion Biologica Clemente Estable). This bacterium caused a slight increase in flavonoid secretion by roots, but not to the same degree as that caused by A. brasilense (results not shown). However, other plant-growth-promoting rhizobacteria may cause a similar effect. This work was supported by The Wolfson Foundation for Scientific Research. REFERENCES 1. Chaverra, M. H., and P. H. Graham. 1992. Cultivar variation in traits affecting early nodulation of common bean. Crop Sci. 32:1432–1436. 2. Dakora, F. D., C. M. Joseph, and D. A. Phillips. 1993. Common bean root exudates contain elevated levels of daidzein and coumestrol in response to Rhizobium inoculation. Mol. Plant-Microbe Interact. 6:665–668. 3. Fallik, E., S. Sarig, and Y. Okon. 1994. Morphology and physiology of plant roots associated with Azospirillum, p. 87–110. In Y. Okon (ed.), Azospirillum/ plant associations. CRC Press, Boca Raton, Fla. 4. Hadas, R., and Y. Okon. 1987. Effect of Azospirillum brasilense inoculation on root morphology and respiration in tomato seedlings. Biol. Fertil. Soils 5:241–247. 5. Hungria, M., C. M. Joseph, and D. A. Phillips. 1991. Rhizobium nod gene inducers exuded naturally from roots of common bean (Phaseolus vulgaris L.). Plant Physiol. (Bethesda) 97:759–764. 6. Iruthayathas, E. E., S. Gunasekaran, and K. Vlassak. 1983. Effect of combined inoculation of Azospirillum and Rhizobium on nodulation and N2 fixation of winged bean and soybean. Sci. Hortic. (Canterbury) 20:231–240. 7. Isoi, T., and S. Yoshida. 1991. Low nitrogen fixation of common bean (Phaseolus vulgaris). Soil Sci. Plant Nutr. 37:559–563. 8. Itzigsohn, R., Y. Kapulnik, Y. Okon, and A. Dovrat. 1993. Physiological and morphological aspects of interactions between Rhizobium meliloti and alfalfa (Medicago sativa) in association with Azospirillum brasilense. Can. J. Microbiol. 39:610–615. 9. Long, S. R. 1989. Rhizobium-legume nodulation: life together in the underground. Cell 56:203–214. 10. Martinez-Romero, E., L. Segovia, F. M. Mercante, A. A. Franco, P. Graham, and M. A. Pardo. 1991. Rhizobium tropici, a novel species nodulating Phaseolus vulgaris L. beans and Leucaena sp. trees. Int. J. Syst. Bacteriol. 41:417– 426. 11. Maxwell, C. A., U. E. Hartwig, C. M. Joseph, and D. A. Phillips. 1989. A chalcone and two related flavonoids from alfalfa roots induce nod genes of Rhizobium meliloti. Plant Physiol. (Bethesda) 91:842–847. 12. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 13. Okon, Y., S. L. Albrecht, and R. H. Burris. 1977. Methods for growing Spirillum lipoferum and for counting it in pure culture and in association with plants. Appl. Environ. Microbiol. 33:85–88. 14. Okon, Y., R. Itzigsohn, S. Burdman, and M. Hampel. 1995. Advances in agronomy and ecology of the Azospirillum/plant association, p. 635–640. In I. A. Tikhanovich, N. A. Provorov, V. I. Romanov, and W. E. Newton (ed.), Nitrogen fixation: fundamentals and applications. Kluwer Academic Publishers, Dordrecht, The Netherlands. 15. Okon, Y., and Y. Kapulnik. 1986. Development and function of Azospirilluminoculated roots. Plant Soil 90:3–16. 16. Okon, Y., and C. A. Labandera-Gonzales. 1994. Agronomic applications of Azospirillum: an evaluation of 20 years worldwide field inoculation. Soil Biol. Biochem. 26:1591–1601. 17. Plazinski, J., and B. G. Rolfe. 1985. Influence of Azospirillum strains on the nodulation of clovers by Rhizobium strains. Appl. Environ. Microbiol. 49: 984–989.
VOL. 62, 1996 18. Sarig, S., Y. Kapulnik, and Y. Okon. 1986. Effect of Azospirillum inoculation on nitrogen fixation and growth of several winter legumes. Plant Soil 90: 335–342. 19. Schroder, E. C. 1992. Improvement of the Phaseolus/Rhizobium symbiosis, with particular reference to the Caribbean region, p. 79–95. In K. Mulongoy, M. Gueye, and D. S. C. Spencer (ed.), Biological nitrogen fixation and sustainability of tropical agriculture. J. Wiley & Sons, Chichester, United Kingdom. 20. Segovia, L., J. P. W. Young, and E. Martinez-Romero. 1993. Reclassification of American Rhizobium leguminosarum biovar phaseoli type I strains as Rhizobium etli sp. nov. Int. J. Syst. Bacteriol. 43:374–377. 21. Singh, C. S., and N. S. Subba Rao. 1979. Associative effects of Azospirillum brasilense with Rhizobium japonicum on nodulation and yield of soybean (Glycine max). Plant Soil 90:335–342. 22. Va ´zquez, M., A. Da ´valos, A. de las Pen ˜ as, F. Sa ´nchez, and C. Quinto. 1991. Novel organization of the common nodulation genes in Rhizobium legumino-
NOTES
3033
sarum bv. phaseoli strains. J. Bacteriol. 173:1250–1258. 23. Verma, D. P. S., and K. Nadler. 1984. Legume-Rhizobium symbiosis: host’s point of view, p. 58–93. In D. P. S. Verma and T. Hohn (ed.), Genes involved in microbe-plant interactions. Springer-Verlag, New York. 24. Vincent, J. M. 1970. A manual for the practical study of root-nodule bacteria. International Biological Programme Handbook, vol. 15. Blackwell Scientific Publishers, Oxford. 25. Volpin, H., S. Burdman, S. Castro-Sowinski, Y. Kapulnik, and Y. Okon. Inoculation with Azospirillum increased exudation of rhizobial nod gene inducers by alfalfa roots. Mol. Plant-Microbe Interact., in press. 26. Volpin, H., and Y. Kapulnik. 1994. Interaction of Azospirillum with beneficial soil microorganisms, p. 111–118. In Y. Okon (ed.), Azospirillum/plant associations. CRC Press, Boca Raton, Fla. 27. Yahalom, E., Y. Okon, and A. Dovrat. 1989. Possible mode of action of Azospirillum brasilense strain Cd on the root morphology and nodule formation in burr medic (Medicago polymorpha). Can. J. Microbiol. 36:10–14.