Extracellular Polysaccharide Composition of Azospirillum Brasilense and Its Relation with Cell Aggregation

FEMS Microbiology Letters 189 (2000) 259^264 www.fems-microbiology.org

Extracellular polysaccharide composition of Azospirillum brasilense and its relation with cell aggregation

S. Burdman a, E. Jurkevitch a, Mar|¨a E. Soria-D|¨az b, Antonio M. Gil Serrano b,

Y. Okon a;* Downloaded from https://academic.oup.com/femsle/article/189/2/259/524810 by guest on 27 September 2021

a Department of Plant Pathology and Microbiology, and The Otto Warburg Center for Agricultural Biotechnology, Faculty of Agricultural, Food and Environmental Quality Sciences, The Hebrew University of Jerusalem, P.O. Box 12, Rehovot 76100, Israel b Department of Organic Chemistry, Faculty of Chemistry, University of Sevilla, 41012 Sevilla, Spain

Received 13 December 1999; received in revised form 14 June 2000; accepted 19 June 2000

Abstract

The exopolysaccharide (EPS) and capsular polysaccharide (CPS) composition of four Azospirillum brasilense strains differing in their aggregation capacity was analyzed by high performance anion exchange chromatography. When growing the different strains in an aggregation inducing medium containing a high carbon:nitrogen (C:N) ratio, both EPS and CPS showed a positive correlation between aggregation and the relative amount of arabinose. Arabinose was not detected in polysaccharides from Sp72002, a pleiotrophic Tn5 mutant strain impaired in aggregation. Arabinose was also not detected in extracellular polysaccharides of bacteria grown in a low C:N ratio, non- inducing aggregation medium, with exception for a relatively small amount found in the CPS of FAJ0204, a super-aggregating mutant strain. The only monosaccharides able to significantly inhibit aggregation at low sugar concentration when tested in a bioassay were arabinose (at a higher extent) and galactose. The possibility that residues of arabinose present in the extracellular polysaccharides are involved in the aggregation of A. brasilense is discussed. ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.

Keywords: Exopolysaccharide; Capsular polysaccharide; Cell aggregation; Flocculation; Azospirillum brasilense

1. Introduction est in the production of bacterial inoculants for agricul- ture. Neyra et al. [9] proposed the generation of inoculants

The free-living N2-¢xing rhizobacteria of the genus Azo- containing £ocs of Azospirillum and Rhizobium for com- spirillum live in close association with plant roots, and mon bean plants. may exert bene¢cial e¡ects on plant growth and yield of Data suggesting the involvement of extracellular poly- many crops of agronomic importance [1,2]. Azospirillum saccharides (exopolysaccharide, EPS; capsular polysaccha- brasilense cells possess the capacity to aggregate and £oc- ride, CPS) in aggregation of Azospirillum have been pub- culate, and this property may positively a¡ect their disper- lished. The production of surface polysaccharide by sal and survival in soil [3,4]. Azospirilla are particularly Azospirillum has been demonstrated by growing colonies responsive to variations in the carbon:nitrogen (C:N) ra- on media containing the £uorescent dye calco£uor, which tio. Rich media with a low C:N ratio tend to promote a binds predominantly to L-1,4- and L-1,3-linked glucans dispersive growth mode, whereas in the presence of a me- [10]. Non-£uorescent mutants of A. brasilense and Azospi- dium with high C:N ratio the cells tend to aggregate and rillum lipoferum have lost the ability to form £ocs and to £occulate [4^8]. anchor to wheat roots, indicating that the calco£uor-bind- Along with the fact that £ocs can be produced on a ing polysaccharide may be necessary for cell aggregation large scale and separated easily from the growth medium, and root attachment [11,12]. Sadasivan and Neyra [4] re- the phenomenon of bacterial aggregation is of great inter- lated £oc formation in A. brasilense Sp7 and A. lipoferum to L-linked EPSs. It has recently been found that a spon- taneous mutant of A. brasilense strain Sp7 which forms * Corresponding author. Tel.: +972 (8) 9489216; colonies that stain weakly with Congo red does not £oc- Fax: +972 (8) 9466794; E-mail: [email protected] culate and lacks cell-surface materials found in the paren-

FEMSLE 9518 26-7-00 260 S. Burdman et al. / FEMS Microbiology Letters 189 (2000) 259^264 tal strain [13]. Genetic complementation of this mutant led pended in and dialyzed against distilled water for 72 h at to the identi¢cation of £cA, a gene involved in the regu- 4³C, using a dialysis membrane with a molecular mass lation of di¡erent processes in A. brasilense, including £oc- cuto¡ of 2 kDa. The bacterial pellet from the ¢rst centri- culation [14]. fugation was resuspended in 10 volumes of phosphate- In a previous work, a medium for consistent induction bu¡ered saline and left to stand for 7 days at 4³C. There- of aggregation of A. brasilense cells was developed and the after, cells were removed by centrifugation (4000Ug,20 e¡ects of chemical and physical factors on this phenom- min, 4³C), and the CPS containing supernatant was ¢l- enon were studied [8]. Growth of A. brasilense strain Cd in trated through a 0.22-Wm membrane and dialyzed against a medium with a high C:N ratio using fructose and am- distilled water as described for the EPS. The amounts of monium chloride as C and N sources, respectively, re- sugar in the polysaccharidic fractions were evaluated by sulted in £occulation visible to the naked eye after 24 h, the anthrone method, using glucose as a standard [17], and whereas no cell aggregates were formed after 72 h of were in accordance with previous results [8].

growth in a low C:N medium. The concentration of Downloaded from https://academic.oup.com/femsle/article/189/2/259/524810 by guest on 27 September 2021 EPS produced by di¡erent strains of A. brasilense varying 2.3. High performance anion exchange (HPAE) in their capacity to aggregate was shown to strongly cor- chromatography relate to the extent of aggregation [8]. Since the extent of aggregation may be determined not only by quantitative Following polysaccharide extraction, samples were but also by qualitative di¡erences in surface polysaccha- lyophilized and then hydrolyzed in 0.5 M H2SO4 at ride production, the present study was carried out with the 90³C for 18 h. After neutralization with Ba(OH)2 and objective of comparing these strains for the composition of centrifugation (7000Ug, 15 min, 4³C), the supernatants their extracellular polysaccharides. were ¢ltered through 0.45-Wm membranes and samples of 25 Wl were applied onto a Dionex HPAE chromatography coupled with pulsed amperometric detection using a 2. Materials and methods Carbo Pac1-PA1 column (250U4 mm; Dionex Co., Sun- nyvale, CA, USA) with 16 mM NaOH as eluant and a 2.1. Bacterial strains and growth conditions £ow rate of 1.0 ml min31. Glucose, galactose, fucose, rhamnose and arabinose, which are ¢ve from the main The A. brasilense strains used in this study were: wild- components of the extracellular polysaccharides of A. bra- type strains Cd [15] and Sp7 [16]; FAJ0204, a Tn5 mutant silense grown on fructose as carbon source [18,19], were of Sp7 defective in the production of both lateral and well resolved using this procedure. polar £agella, kindly supplied by J. Vanderleyden (Catho- lic University of Leuven, Belgium); Sp72002 [14], a pleio- 2.4. Aggregation bioassay trophic Tn5 mutant of Sp7 impaired in £cA, and a¡ected in its aggregation capacity, kindly supplied by C. Elmerich The bioassay is based on ¢ndings from previous work in (Institute Pasteur, Paris, France). Bacteria were main- which sonication of A. brasilense £ocs for a relatively short tained on nutrient agar (Difco, Detroit, MI, USA) slants. period (20 s) was shown to be e¤cient in temporarily dis- Kanamycin was added at 50 Wgml31 for the Tn5 mutant rupting the cell aggregates [8]. The assay is fully described strains. For aggregation assays and extracellular material in Burdman et al. [20]. Brie£y, it consists of adding the extraction, bacteria were grown in 250-ml Erlenmeyer supernatant of 20-s sonicates of high C:N-grown (aggre- £asks, with 100 ml liquid medium. High and low C:N gating) strain Cd cells to low C:N-grown (non-aggregat- media were as previously described [8]. Flasks were inoc- ing) cells. The latter are then able to aggregate and £oc- ulated with exponential phase cultures at an initial OD540 culate after an incubation period of 2^3 h in a conical tube of approximately 0.05 (about 107 colony forming units [20]. In the present study, the e¡ects of di¡erent monosac- ml31), and incubated on a rotary shaker (150 rpm) at charides on aggregation were tested by adding them to the 30³C. bioassay at various concentrations. The percentage of aggregation was measured as described previously [8]. The 2.2. Extraction of extracellular polysaccharides aggregate containing suspension was allowed to stand and after 20 min, aggregates had settled to the bottom of the Polysaccharides were extracted as described by Del Gal- tube, and the suspension was mostly composed of free lo and Haegi [6] with some modi¢cations. Following (dispersed) cells. Turbidity was measured from the suspen- growth for 24 h, bacterial cultures were centrifuged at sion using a Genesis 5 spectrophotometer (Spectronic Ins.)

4000Ug for 20 min at 4³C. The supernatant was collected, at 540 nm (ODs). The culture was then dispersed by a ¢ltered through a 0.22-Wm membrane and left to stand in tissue homogenizer (Heidolph RzR 50) for 1 min, and three volumes of cold ethanol for 24 h at 4³C, resulting in the total turbidity was measured (ODt). Percentage of ag- EPS precipitation. The precipitated polysaccharides were gregation was estimated as follows: % aggregation = isolated by centrifugation (30 000Ug, 20 min, 4³C), resus- (ODt3ODs)U100/ODt.

FEMSLE 9518 26-7-00 S. Burdman et al. / FEMS Microbiology Letters 189 (2000) 259^264 261 Downloaded from https://academic.oup.com/femsle/article/189/2/259/524810 by guest on 27 September 2021

Fig. 1. Relative concentration (%) among ¢ve of the main monosaccharides in the EPS of four A. brasilense strains grown for 24 h in high (H) and low (L) C:N medium. Columns represent averages (from three replicates) from one of two independent experiments with similar results. nd = not detect- able. Di¡erent letters indicate signi¢cant di¡erences (P = 0.05) among the various sugars in each strain, at high or low C:N ratio, as determined by one- way analysis of variance (ANOVA). Asterisks in (H) indicate signi¢cant di¡erences (P = 0.05) in the relative amount of the referred sugar in each strain between high and low C:N conditions, as determined by t-test analysis. Di¡erences in the relative amount of each sugar among the di¡erent strains as determined by one-way ANOVA are referred in the text.

3. Results of galacturonic acid, glucosamine and galactosamine that were also reported to be present in the extracellular poly- The relative concentrations (in %) of the tested monosaccharides of some A. brasilense strains [6,18,19], but at saccharides detected from the EPS and CPS of four A. lower concentrations than the sugars tested in the present brasilense strains, grown in aggregation and non-aggrega- study. These sugars are not well resolved by the used tion inducing conditions (high and low C:N, respectively), method. are presented in Figs. 1 and 2. These percentages do not In general, glucose was statistically signi¢cant present at take into account other sugar components such as residues the highest concentrations in all strains, both in the EPS

Fig. 2. Relative concentration (%) among ¢ve of the main monosaccharides in the CPS of four A. brasilense strains grown for 24 h in high (H) and low (L) C:N medium. Columns represent averages (from three replicates) from one of two independent experiments with similar results. nd = not detect- able. Statistical analysis was carried out and expressed as for the EPS (see Fig. 1 for legend).

FEMSLE 9518 26-7-00 262 S. Burdman et al. / FEMS Microbiology Letters 189 (2000) 259^264 and the CPS fractions. This was true for most of the strains grown under high as well as low C:N. The highest relative amounts of glucose were observed in the EPS of low C:N-grown strains (Fig. 1). In contrast, increased relative amounts of glucose in the CPS of high C:N-grown bacteria as compared to low C:N-grown bacteria were observed (Fig. 2). Rhamnose was also found to be gener- ally present in relatively high amounts, especially in the CPS fractions. In the tested strains, the relative amounts of rhamnose in the CPS fractions obtained from low C:N medium were signi¢cantly higher when compared to CPS fractions originating from high C:N medium (Fig. 2). This

trend was not observed in the di¡erent EPS fractions Downloaded from https://academic.oup.com/femsle/article/189/2/259/524810 by guest on 27 September 2021 (Fig. 1). Similarly to previous reported data [8], the extent of aggregation of the di¡erent strains after 24 h of growth in high C:N medium was determined to be approximately 15, 40 and 60% for strains Cd, Sp7 and FAJ0204, respectively. As previously indicated, strain Sp72002 is a pleio- Fig. 3. E¡ects of di¡erent monosaccharides at various concentrations on trophic Tn5 mutant impaired in aggregation. The relative aggregation of low C:N-grown A. brasilense strain Cd after 4 h of incu- amounts of arabinose present in the EPS of these four bation with 20-s sonicates of high C:N-grown cells, according to Burd- man et al. [20]. Bars represent averages from one representative experi- strains when grown in the high C:N medium (Fig. 1) ment (three replicates per treatment). Di¡erent letters indicate signi¢cant were signi¢cantly di¡erent between each other (P = 0.05), di¡erences (P = 0.05) among the various monosaccharides for each sugar and strongly correlated with the extent of aggregation they concentration, as determined by one-way ANOVA. Ar, arabinose; Ga, exhibited (r2 = 0.85). Moreover, in the super-aggregative galactose; Rh, rhamnose; Fu, fucose; Gl, glucose; Fr, fructose; Co, mutant strain FAJ0204, arabinose was signi¢cantly found control (without sugar addition). to be the predominant sugar (Fig. 1). In the CPS from high C:N-grown bacteria (Fig. 2), the relative amount of the EPS composition of the mutant Sp72002, which ex- arabinose also correlated with the aggregation values but hibited concentration ratios of 1.0, 0.6, 0.3, 0.2 and 0.1 for to a smaller extent (r2 = 0.75). No arabinose could be de- glucose, galactose, rhamnose, mannose and allose, respec- tected in the extracellular polysaccharides of strain tively. Sp72002. Moreover, no arabinose was found in the extra- An aggregation bioassay was utilized to evaluate the cellular polysaccharides of bacteria grown in low C:N in£uence of various monosaccharides (most of them (Figs. 1 and 2), except for a small amount present in the present in the extracellular polysaccharides of A. brasi- CPS of strain FAJ0204, which, under these growth con- lense) on aggregation. If a speci¢c component in the extra- ditions, exhibited an extent of aggregation of about 20% cellular polysaccharides plays a role in aggregation, it is after 24 h. No correlation was found between the relative expected that its addition will cause an inhibition of this amount of any of the other tested monosaccharides and process by obstructing the adhesion sites. Fig. 3 shows the extent of aggregation. results of one of three independent experiments, all exhib- The EPS fractions of the four tested strains grown iting statistically similar results. At a low sugar concentra- under high C:N conditions were also analyzed by gas liq- tion of 0.01 M, arabinose (at a higher extent) and galac- uid chromatography-mass spectrometry (GLC-MS). This tose signi¢cantly inhibited cell aggregation while the other analysis was carried out on the O-trimethylsilyl derivatives tested monosaccharides were ine¡ective, not being signi¢- of the methyl glycosides [21]. In addition to the previously cantly di¡erent from the untreated controls. The extent of HPAE-detected monosaccharides, residues of mannose, inhibition was shown to be concentration-dependent (Fig. xylose and allose were also found in the EPS of the di¡er- 3). Only at higher concentrations, relatively low inhibitory ent strains when analyzed by GLC-MS. In contrast, fucose e¡ects (in comparison to arabinose and galactose) were could not be detected using this technique. In general, observed with the other sugars with the exception of fruc- similar results to those obtained by HPAE chromatogra- tose, which among the tested sugars is the only monosac- phy were obtained by GLC-MS, con¢rming the relatively charide not present in the extracellular polysaccharides of high amount of arabinose present in the EPS of high C:N- A. brasilense. grown FAJ0204 cells. Indeed, this strain showed concentration ratios of 2.0, 1.0, 0.5, 0.5, 0.3, 0.2 and 0.1 for arabinose, glucose, galactose, rhamnose, xylose, allose 4. Discussion and mannose, respectively. In contrast, arabinose could not be detected also by this technique when analyzing Published data support the hypothesis that extracellular

FEMSLE 9518 26-7-00 S. Burdman et al. / FEMS Microbiology Letters 189 (2000) 259^264 263 polysaccharides are involved in aggregation of Azospiril- Paris, France) for supplying strains for this research, and lum (see Section 1). We previously reported on evidences M. Megias (University of Sevilla, Spain) for helpful dis- indicating that outer membrane proteins (OMPs) are in- cussions. This work was supported by a grant from `The volved in this phenomenon [20]. The strains used in the Israel Science Foundation' founded by `The Academy of present study di¡er in their aggregation capacity but ex- Sciences and Humanities'. hibit similar OMPs pro¢les. Moreover, no di¡erences are observed in OMPs pro¢les between cells grown under high and low C:N conditions [8]. However, these strains did References di¡er in the amounts of EPS produced, which were shown to correlate to their extents of aggregation [8]. As was [1] Okon, Y. and Itzigsohn, R. (1995) The development of Azospirillum shown in the present study, the composition of the EPS as a commercial inoculant for improving crop yields. Biotechnol. and CPS also di¡ered between the tested strains. There- Adv. 13, 415^424. [2] Okon, Y. and Vanderleyden, J. (1997) Root-associated Azospirillum fore, it is possible that the OMP(s) involved in aggregation Downloaded from https://academic.oup.com/femsle/article/189/2/259/524810 by guest on 27 September 2021 species can stimulate plants. ASM News 63, 366^370. is (are) constitutively present in A. brasilense, with di¡er- [3] Nur, I., Steinitz, Y.L., Okon, Y. and Henis, Y. (1981) Carotenoid ences in the extent of aggregation among the various composition and function in nitrogen ¢xing bacteria of the genus strains being related to the amount, composition and Azospirillum. J. Gen. Microbiol. 122, 27^32. structure of the extracellular polysaccharide. Interactions [4] Sadasivan, L. and Neyra, C.A. (1985) Flocculation in Azospirillum between proteins and polysaccharides leading to cell ag- brasilense and Azospirillum lipoferum: exopolysaccharides and cyst formation. J. Bacteriol. 163, 716^723. gregation are well established in the microbial world, with [5] Del Gallo, M., Negi, M. and Neyra, C.A. (1989) Calco£uor- and good examples being the well characterized systems of lectin-binding exocellular polysaccharides of Azospirillum brasilense Myxobacteria [22] and dental plaques [23]. and Azospirillum lipoferum. J. Bacteriol. 171, 3504^3510. Analysis of the relative concentration of ¢ve from the [6] Del Gallo, M. and Haegi, A. (1990) Characterization and quanti¢ca- main monosaccharides present in the extracellular polysac- tion of exocellular polysaccharides in Azospirillum brasilense and Azospirillum lipoferum. Symbiosis 9, 155^161. charides of the di¡erent strains suggests that arabinose [7] Arunakumari, A., Lamm, R.B. and Neyra-Estens, C.A. (1992) present in the EPS and in the CPS may play an important Changes in cell surface properties of azospirilla in relation to cell role in determining the aggregation capability of A. brasi- pleomorphism and aggregation. Symbiosis 13, 291^305. lense. This hypothesis was strengthened by the fact that L- [8] Burdman, S., Jurkevitch, E., Schwartsburd, B., Hampel, M. and arabinose is able to strongly inhibit aggregation when Okon, Y. (1998) Aggregation in Azospirillum brasilense: e¡ects of chemical and physical factors and involvement of extracellular com- added to the bacterial medium in an aggregation bioassay. ponents. Microbiology 144, 1989^1999. Recently, in a screening for cell-surface lectins of A. bra- [9] Neyra, C.A., Atkinson, A. and Olubayi, O. (1995) Coaggregation of silense, it was found that arabinose was the only sugar Azospirillum with other bacteria: basis for functional diversity. capable of completely inhibiting agglutination of A. brasi- NATO ASI Ser. G 37, 429^439. lense Sp245 cells using latex beads covered with ¢ve di¡er- [10] Wood, P.J. and Fulcher, R.G. (1978) Interaction of some dyes with cereal L-glucans. Cereal Chem. 55, 952^966. ent neoglycoproteins [24], thus providing other evidence [11] Michiels, K., Verreth, C. and Vanderleyden, J. (1990) Azospirillum on the possible involvement of this sugar in aggregation. lipoferum and Azospirillum brasilense surface polysaccharide mutants These ¢ndings do not exclude that residual sugars in the that are a¡ected in £occulation. J. Appl. Bacteriol. 69, 705^711. extracellular polysaccharides of the bacterium other than [12] Michiels, K.W., Croes, C.L. and Vanderleyden, J. (1991) Two di¡er- arabinose may also be involved in cell aggregation ent modes of attachment of Azospirillum brasilense Sp7 to wheat roots. J. Gen. Microbiol. 137, 2241^2246. although evidences for this possibility are lacking from [13] Katupitiya, S., Millet, J., Vesk, M., Viccars, L., Zeman, A., Lidong, this as well as from previous works. Z., Elmerich, C. and Kennedy, I.R. (1995) A mutant of Azospirillum We are currently investigating other features of the brasilense Sp7 impaired in £occulation with a modi¢ed colonization OMPs and of the extracellular polysaccharides of A. bra- pattern and superior nitrogen ¢xation in association with wheat. silense in order to elucidate the biological factors involved Appl. Environ. Microbiol. 61, 1987^1995. [14] Pereg-Gerk, L., Paquelin, A., Gounon, P., Kennedy, I.R. and Elmer- in cell to cell adhesion of this bacterium. Since it seems ich, C. (1998) A transcriptional regulator of the LuxR-UhpA family, that surface components that are associated with the ag- FlcA, controls £occulation and wheat root surface colonization by gregation process are also important in plant root coloni- Azospirillum brasilense Sp7. Mol. Plant-Microbe Interact. 11, 177^ zation [11,12,19], advances in the understanding of the 187. aggregation process may also broaden our knowledge on [15] Eskew, D.L., Focht, D.D. and Ting, I.P. (1977) Nitrogen ¢xation, denitri¢cation and pleomorphic growth in highly pigmented Spirillum the interactions that lead to the adhesion and colonization lipoferum. Can. J. Microbiol. 34, 582^585. of plant roots by the bacteria. [16] Tarrand, J.J., Krieg, N.R. and Do«bereiner, J. (1978) A taxonomic study of Spirillum lipoferum group, with description of a new genus, Azospirillum gen. nov., and two species Azospirillum lipoferum (Bei- Acknowledgements jerinck) comb. nov., and Azospirillum brasilense, sp. nov.. Can. J. Microbiol. 24, 967^980. [17] Dische, Z. (1962) General color reactions. Methods Charbohydr. We thank J. Vanderleyden (Catholic University of Chem. 1, 477^479. Leuven, Belgium) and C. Elmerich (Institute Pasteur, [18] Del Gallo, M. and Fendrik, I. (1994) The rhizosphere and Azospi-

FEMSLE 9518 26-7-00 264 S. Burdman et al. / FEMS Microbiology Letters 189 (2000) 259^264

rillum.In:Azospirillum/Plant Associations (Okon, Y., Ed.), pp. 57^ of carbohydrate components in glycoproteins using gas-liquid chro- 85. CRC Press, Boca Raton, FL. matography. Anal. Biochem. 123, 336^341. [19] De Troch, P. and Vanderleyden, J. (1996) Surface properties and [22] Dworkin, M. (1996) Recent advances in the social and developmental motility of Rhizobium and Azospirillum in relation to plant root at- biology of the Myxobacteria. Microbiol. Rev. 60, 70^102. tachment. Microb. Ecol. 32, 149^169. [23] Marsh, P.D. and Bradshaw, D.J. (1995) Dental plaque as a bio¢lm. [20] Burdman, S., Jurkevitch, E., Schwartsburd, B. and Okon, Y. (1999) J. Ind. Microbiol. 15, 169^175. Involvement of outer membrane proteins in the aggregation of Azo- [24] Castellanos, T., Ascencio, F. and Bashan, Y. (1998) Cell-surface lec- spirillum brasilense. Microbiology 145, 1145^1152. tins of Azospirillum spp.. Curr. Microbiol. 36, 241^244. [21] Chaplin, M.F. (1982) A rapid and sensitive method for the analysis Downloaded from https://academic.oup.com/femsle/article/189/2/259/524810 by guest on 27 September 2021

FEMSLE 9518 26-7-00