Microbiology Papers in Press. Published January 27, 2011 as doi:10.1099/mic.0.

043943-0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Role of the antimicrobial compound 2,4-diacetylphloroglucinol in the impact of biocontrol Pseudomonas fluorescens F113 on Azospirillum brasilense phytostimulators Olivier Couillerot, Emeline Combes-Meynet, Jol F. Pothiera, Floriant Bellvert, Elita Challita, Marie-Andre Poirier, Ren Rohr, Gilles Comte, Yvan Monne-Loccoz, Claire PrigentCombaret *

Universit de Lyon, F-69622, Lyon, France; Universit Lyon 1, Villeurbanne, France; CNRS, UMR5557, Ecologie Microbienne, Villeurbanne, France

Present address: Agroscope Changins-Wdenswil ACW, Swiss Federal Research Station,

Plant Protection Division, 8820 Wdenswil, Switzerland

Running title: Phl and impact of Pseudomonas on Azospirillum

Contents category: Environmental and Evolutionary Microbiology

Corresponding author. Mailing address: UMR CNRS 5557 Ecologie Microbienne, Universit

Lyon 1, 43 bd du 11 Novembre, 69622 Villeurbanne cedex, France. Phone: +33 4 72 44 58 89. Fax: +33 4 72 43 12 23. E-mail: claire.prigent@univ-lyon1.fr

Couillerot et al., 1 Microbiology

28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

SUMMARY

Pseudomonads producing the antimicrobial metabolite 2,4-diacetylphloroglucinol (Phl) can control soil-borne phytopathogens, but their impact on other plant-beneficial bacteria remains poorly documented. Here, the effects of synthetic Phl and Phl+ Pseudomonas fluorescens F113 on Azospirillum brasilense phytostimulators were investigated. Most A. brasilense strains were moderately sensitive to Phl. In vitro, Phl induced accumulation of carotenoids and poly--hydroxybutyrate-like granules, cytoplasmic membrane damage and growth inhibition in Azospirillum brasilense Cd. Experiments with P. fluorescens F113 and a Phlmutant indicated that Phl production ability contributed to in vitro growth inhibition of A. brasilense Cd and Sp245. Under gnotobiotic conditions, each of the three strains P. fluorescens F113, A. brasilense Cd and Sp245 stimulated wheat growth. Co-inoculation of A. brasilense Sp245 and Pseudomonas resulted in the same level of phytostimulation as in single inoculations, whereas it abolished phytostimulation when A. brasilense Cd was used. Pseudomonas Phl-production ability resulted in lower Azospirillum cell numbers per root system (based on colony counts) and restricted microscale root colonization of neighbouring Azospirillum cells (based on confocal microscopy), regardless of the A. brasilense strain used. Therefore, this work establishes that Phl+ pseudomonads have the potential to interfere with A. brasilense phytostimulators on roots and with their plant growth promotion capacity.

Couillerot et al., 2 Microbiology

47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71

INTRODUCTION

Microbial interactions in the rhizosphere are of paramount importance for plant growth and health (Barea et al., 2005; Raaijmakers et al., 2009). Many strains of fluorescent Pseudomonas are effective at colonizing plant roots and have been extensively studied for their plant-beneficial effects (Couillerot et al., 2009). Pseudomonas strains may benefit the plant directly, via induction of systemic resistance to pathogens (Bakker et al., 2007), deamination of the ethylene precursor 1-aminocyclopropane-1-carboxylate (Hontzeas et al., 2004), auxin production (Picard & Bosco, 2005) and/or associative nitrogen fixation (Mirza et al., 2006). Plant-beneficial effects by fluorescent pseudomonads may also entail the inhibition of soil-borne phytoparasitic microorganisms, which often involves production of siderophores (Lemanceau et al., 1992) and especially antimicrobial secondary metabolites (Raaijmakers et al., 2002; Haas & Dfago 2005). 2,4-Diacetylphloroglucinol (Phl), pyoluteorin, pyrrolnitrin, hydrogen cyanide and viscosinamide are some examples of well-studied antimicrobial metabolites produced by biocontrol strains of fluorescent pseudomonads (Haas & Keel 2003). The ability to produce Phl was shown, both in vitro and in vivo, as a particularly important biocontrol trait by comparing wild-types and non-producing mutants (Vincent et al., 1991; Fenton et al., 1992; Keel et al., 1992). In addition, Phl+ strains protected plants better than naturally nonproducing counterparts when assessing collections of wild-type biocontrol pseudomonads (Rezzonico et al., 2007). The polyketide Phl inhibits the growth of several phytopathogenic bacteria (Pectobacterium carotovorum; Cronin et al., 1997a), oomycetes (Pythium spp.) and fungi (e.g. Rhizoctonia solani, Thielaviopsis basicola, Gaeumannomyces graminis var. tritici) (Howell & Stipanovic 1979; Keel et al., 1992; Shanahan et al., 1992), and is also active against nematodes (Cronin et al., 1997b).

Couillerot et al., 3 Microbiology

72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96

The inhibitory properties of Phl are not restricted to phytopathogens, as nonpathogenic rhizosphere fungi (Girlanda et al., 2001) and bacteria (Natsch et al., 1998) might be inhibited as well. As far as saprophytic rhizobacteria are concerned, this possibility has been assessed in detail only for a very limited number of taxa, especially Bacillus (Natsch et al., 1998), Rhizobium leguminosarum (Walsh et al., 2003), and Cytophaga-like bacteria (Johansen et al., 2002). However, many other rhizobacterial taxa are also important to consider, because they can occur in the same rhizosphere as Phl+ pseudomonads (Barea et al., 2005; Kyselkov et al., 2009) and may have positive effects on the host plant. Whether these rhizobacteria can be inhibited by Phl (and Phl+ pseudomonads) on roots is usually unknown. In addition to microbial inhibition, Phl can also have an effect on root physiology (Brazzelton et al., 2008), which in turn may influence the conditions in which saprophytic rhizobacteria colonize the rhizosphere. Indeed, Phl (or the presence of Phl+ pseudomonads) can elicit an induced systemic resistance in the plant (Iavicoli et al., 2003), but the consequences are probably negligible when considering root colonization by other saprophytic bacteria. Much more significant, roots exposed to Phl display enhanced exudation of amino acids (Phillips et al., 2004), and this may enhance the ability of saprophytic bacteria to colonize roots. Therefore, it can be anticipated that Phl+ pseudomonads may have negative, neutral or positive effects on other root-colonizing saprophytic bacteria, depending on whether or not the latter (i) are sensitive to Phl and/or (ii) can benefit from Phl-driven amino acid root exudation. The aim of this study was to assess whether the ability of root-associated pseudomonads to produce Phl could have an impact on growth and root colonization by other saprophytic rhizobacteria. This possibility was investigated in the case of A. brasilense, one of the most important species of plant growth-promoting rhizobacteria. A. brasilense phytostimulatory effects on cereals are extensively documented (Dobbelaere et al., 2001;

Couillerot et al., 4 Microbiology

97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113

Fuentes-Ramirez & Caballero-Mellado 2006) and their plant-beneficial traits include associative nitrogen fixation (Bally & Elmerich 2007), production of nitric oxide (Creus et al., 2005; Pothier et al., 2007) and phytohormones especially auxins (Costacurta & Vanderleyden 1995; Dobbelaere et al., 1999). This may stimulate root growth, which in turn can lead to a better uptake of nutrients and water by the plant, and thus to better plant health and development (Reid & Renquist 1997; Dobbelaere et al., 2003; Richardson et al., 2009). Interestingly, Azospirillum and Pseudomonas indigenous populations have been found together in the rhizosphere of tobacco (Kyselkov et al., 2009) and wheat (Sanguin et al., 2009). In this work, A. brasilense strains from different geographic origins and host plants were exposed in vitro to synthetic Phl to determine whether this compound had any deleterious effect. A Phl+ Pseudomonas fluorescens strain and its Phl- mutant were then compared for their effect on cells of A. brasilense Cd and Sp245, both in vitro and in planta. The co-inoculation experiments were done under gnotobiotic conditions using autofluorescent bacterial derivatives to assess whether colonization patterns of particular root zones (and phytostimulatory properties) of A. brasilense can be affected in the presence of a Phl+ pseudomonad.

Couillerot et al., 5 Microbiology

114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137

METHODS

Strains and culture conditions All bacterial strains (Table 1) were routinely grown at 28 C with shaking in LBm, i.e. LuriaBertani medium (Sambrook et al., 1989) containing only 5 g NaCl L-1. The other media were the N-free medium NFb (Nelson & Knowles, 1978) supplemented with Congo red (0.25 % w/v) when in plates, AB medium (Chilton et al.,, 1974) containing 5 g malic acid L-1 as carbon source (i.e. ABmal), Kings B (King et al., 1954), and SA-Fe (Cronin et al., 1997a), i.e. Sucrose Asparagine (Scher & Baker 1982) supplemented with 100 M FeCl3. Antibiotics were used at the following concentrations: ampicillin, 40 g mL-1 (Amp40); chloramphenicol, 15 and 30 g mL-1 (Cm15 and Cm30, respectively); gentamycin, 25 g mL-1 (Gm25).

Effect of synthetic Phl on growth and cell morphology of Azospirillum brasilense Cd The effect of synthetic Phl (Toronto Research Chemicals Inc., North York, Ontario) on A. brasilense Cd was investigated in Petri dishes, where the compound was spotted onto water agar containing Cd cells, as follows. Strain Cd was grown for 48 h in liquid NFb supplemented with 1/40 (v/v) LBm. The cells were washed twice in a 10 mM MgSO4 solution and adjusted to 4 107 cells mL-1 (based on optical density). Five mL of cell suspension were then mixed with 5 mL of molten water agar (1.5 % w/v), and the mixture was immediately poured onto SA-Fe or LBm agar. Phl was dissolved in methanol (to reach 0.1 to 100 mM Phl) and 15 L were spotted onto the Cd agar layer. Methanol was used as Phl-negative control. For SA-Fe and LBm sublayers, the effect of each of the ten Phl concentrations studied and the Phl-negative control were analysed on two occasions, using two independent Cd cell cultures at each time. The plates were incubated 72 h at 28 C.

Couillerot et al., 6 Microbiology

138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162

To assess the impact of Phl on growth, the diameter of the inhibition zone was measured (using a ruler). To assess the impact of Phl on cell morphology, small pieces of water agar were cut off at different locations in the plate. Samples were then fixed with osmium tetroxide 2 %, contrasted with uranyl acetate 1 %, dehydrated in a graded ethanol series and embedded in Epon. Ultra-thin sections (0.1 m) were cut using a Reichert ultramicrotone (Vienna, Austria), contrasted with lead citrate and observed using a CM 120 transmission electron microscope (Philips, Eindhoven, The Netherlands) at 100 kV. Several dozen cells were examined in each of the 2 replicates.

Effect of synthetic Phl on carotenoid production by A. brasilense Cd The effect of synthetic Phl on the production of carotenoids by A. brasilense Cd was studied by Ultra-High-Pressure Liquid Chromatography (UHPLC), after growing cells in 24-well microtiter plates (Nalgene NUNC, Roskilde, Denmark). Phl was dissolved in methanol (from 0.0001 to 1 mM Phl) and 15 L was added to 1.5 mL of ABmal broth in each well. Methanol (1% v/v) was used as Phl-negative control. Cd was inoculated at 1.5 106 cells mL-1. At least two replicates were prepared per treatment. After 7 d of growth at 28C, Cd cells were recovered from each well, and cell density was estimated by measuring the optical density of cultures at 600 nm (OD600). Cells were pelleted, lyophilised, and extracted with 1 mL of pure methanol by ultrasonication for 15 min. After centrifugation (10 min at 16,000 g), the supernatant was collected. Another extraction with 0.75 mL of pure methanol was performed on the pellet. Both supernatants obtained were pooled, filtrated and concentrated to 1 mg mL-1. UHPLC analysis of the extracts was performed with Agilent 1290 series coupled with a G4212A DAD (Agilent technologies, Santa Carla, CA). The system was managed by the Mass Hunter (Agilent technologies). Separation was carried out using a Hypersil Gold-C18

Couillerot et al., 7 Microbiology

163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186

column (100 x 1 mm, 1.9 m particle size; Thermo scientific, San Jose, CA), thermostated at 40C. The mobile phase was a linear gradient of formic acid 1% in water (solvent A) and acetonitrile (solvent B). The linear gradient at the flow rate 0.15 mL min-1 was 0 to 0.5 min, 50% solvent B; 0.5 to 7 min, 50% to 100% solvent B; 7 to 8 min, 100% solvent B. Ten microliters of each sample were injected. Spectra were recorded between 200 and 600 nm and the response at 450 nm was used for quantification of the Area Under the Curve (AUC) of all characteristic peaks. The latter were defined on the basis of the typical shape of their UVvisible spectra (Nur et al., 1981). The total content of carotenoids was expressed as arbitrary AUC units s-1 OD600-1. The number of distinct carotenoid metabolites was also recorded for each sample.

Sensitivity of A. brasilense strains to synthetic Phl Overnight liquid NFb cultures of 11 A. brasilense strains (Table 1) were used to inoculate NFb liquid medium in 96-well microtiter plates (50 L inoculum into 150 L of medium). Each strain was inoculated in at least six wells (i.e. six replicates). After a 24 h incubation at 28 C (without agitation), each liquid culture was drop-inoculated onto solid LBm or SA-Fe medium containing Phl at final concentrations of 5, 10, 20, 50, 100, 200, 500 or 1000 M (previously dissolved in methanol and resulting in 1 % v/v methanol in media), and methanol (1 % v/v) was used in the Phl-negative control. The plates were incubated for 72 h at 28 C. Growth of colonies was scored visually, and the Phl concentration needed to abolish growth in at least 50 % of the case (LC50) was determined. Each of the eight Phl concentrations studied and the Phl-negative control were investigated using 6 to 22 replicates.

A. brasilense in vitro inhibition experiments using Phl+ pseudomonads and mutants

Couillerot et al., 8 Microbiology

187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211

The two A. brasilense strains Sp245 and Cd were exposed in vitro to the Phl+ sugarbeet isolate P. fluorescens F113 (Shanahan et al.,, 1992), its Tn5::lacZY-induced Phl- biosynthetic derivative F113G22 (Shanahan et al.,, 1992), and the Phl+ complemented mutant F113G22(pCU203) (Fenton et al., 1992). Strains F113 and F113G22 were grown in LBm and F113G22(pCU203) in LBm Cm30. The cells were then washed twice in a 10 mM MgSO4 solution and adjusted to 109 cells mL-1 (based on optical density). Each A. brasilense strain was prepared in water agar, which was poured on LBm or SA-Fe medium, as described above, and each pseudomonad was spotted as 15 L of cell suspension. For each A. brasilense strain, growth inhibition tests were performed in duplicate and the whole experiment was done twice. After 72 h of incubation at 28 C, the width (radius) of the inhibition zone surrounding the Pseudomonas colonies was measured using a ruler. Conversely, the potential inhibitory activity of A. brasilense Cd and Sp245 strains against P. fluorescens F113 was checked, using a similar protocol.

Growth chamber experiments The effect of P. fluorescens F113, F113G22 and F113G22(pCU203) on growth and root colonization of the wheat isolate A. brasilense Sp245 and the wheat-adapted strain A. brasilense Cd strains was assessed on wheat (Triticum aestivum) under gnotobiotic conditions. The Plac-egfp plasmid pMP2444 was introduced (as described in Pothier et al., 2007) into the two A. brasilense strains retained and the Plac-rfp plasmid pMP4661 into P. fluorescens F113 and F113G22, so as to monitor cells by fluorescence microscopy (Bloemberg et al., 2000). Both plasmids derive from the broad host-range vector pBBR1MCS-5 and confer gentamycin resistance. The inoculants were obtained after overnight growth in liquid LBm Cm30 for strain F113G22(pCU203) or LBm Gm25 for all other strains. The cells were washed twice in a 10 mM MgSO4 solution and adjusted to 2

Couillerot et al., 9 Microbiology

212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236

107 CFU mL-1 for co-inoculation and 107 CFU mL-1 for single inoculation (based on optical density). Seeds of spring wheat (cv. Fiorina), obtained from Florimond-Desprez (Cappelle en Pvle, France), were surface-disinfected, as described by Pothier et al., (2007). They were placed on water agar (15 % w/v) and incubated 48 h in the dark at 28 C to enable germination. The plants were then co-inoculated using one Pseudomonas strain and one Azospirillum strain, and in the controls they were inoculated using only one strain or were not inoculated. This was done by treating pre-germinated seeds with 50 l containing 106 CFU of Azospirillum and 50 l containing 106 CFU of Pseudomonas or with 100 l containing 106 CFU of each strain. Plants were placed near one edge of square plates (12 cm 12 cm) containing water agar (15 % w/v). Each A. brasilense strain was studied in a distinct experiment carried out independently. Four plates (containing four plants each) were used per treatment. The plates were placed standing in a growth chamber at 75 % relative humidity, with 16 h of light (63 Em-2 s-1) at 26 C and 8 h of dark at 18 C, and they were sampled seven days later. Root development at 7 d after inoculation was assessed by measuring root biomass and characterizing root system architecture using WinRHIZO (Rgent Instruments Inc., Qubec, Canada). Four plants, each from a distinct plate, were studied per treatment.

Analysis of bacterial root colonization in the growth chamber experiments Confocal Laser Scanning Microscopy (CLSM) observations were done at 7 d using two plants, each from a distinct plate, per treatment. Samples 1-2 cm in length were cut from different root zones (root apex, hair root zone, and older root parts) and mounted in AquaPoly/Mount (Polysciences, Eppelheim, Germany). A 510 Meta microscope (Carl Zeiss, Le Pecq, France) equipped with argon-krypton and He-Ne lasers was used for analysis of green

Couillerot et al., 10 Microbiology

237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255

fluorescence (excitation at 488 nm and detection at 510-531 nm) and red fluorescence (excitation at 543 nm and detection at 563-628 nm). After acquisition of transmitted (in bright field mode) and/or reflected lights (in blue, detection at 456-499 nm), the single-colour images were overlaid into a single image using LSM software release 3.5 (Carl Zeiss). For quantification of inoculant bacteria, four plants were sampled per treatment at 7 d. Bacteria were extracted by vortexing each root system 5 min in a 15-mL Falcon tube containing 5 mL of 10-mM MgSO4 solution. A serial dilution was prepared in the same solution, and six 10-l drops from each dilution were spotted onto Kings B Amp40 Cm15 Gm25 to quantify F113(pMP4661) or F113G22(pMP4661), Kings B Amp40 Cm30 to quantify F113G22(pCU203), and NFb Gm25 to quantify Cd(pMP2444) or Sp245(pMP2444). Colonies were not found on plates in the absence of inoculation of seedlings with the corresponding strain(s).

Statistical analysis CFU of root-colonizing inoculants were expressed per g of dry root and were log-transformed before analysis. All results were processed by analysis of variance (ANOVA), followed when appropriate with Fishers least significant difference (LSD) tests. All analyses were conducted at P < 0.05, using S-plus software 6.1 (Hearne Scientific Software, Kilkenny, Ireland).

Couillerot et al., 11 Microbiology

256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280

RESULTS

Effect of synthetic Phl on A. brasilense Cd on plates Synthetic Phl (amount ranging from 0 to 1500 nmol) was used in a simple plate assay to determine its effect on the model strain A. brasilense Cd. Growth of A. brasilense Cd in complex LBm medium and minimal defined SA-Fe medium was inhibited when at least 15 nmol of synthetic Phl was added (as 15 L spot). With both media, the diameter of the inhibition zone increased with increasing Phl quantities (Fig. 1a). Only few cells were found by photonic microscopy in this inhibition zone. Surprisingly, a dark pink halo (10-20 mm wide) of Azospirillum growth was observed around the inhibition zone when Cd was grown on SA-Fe (but not on LBm), provided that added Phl exceeded 150 nmol (Fig. 1b). Photonic microscopy indicated that cell population biomass was higher in the dark pink halo compared with normal growth areas of the plate located further away (Fig. 1c). In addition to growth inhibition, electronic microscopy indicated that cell morphology of A. brasilense Cd was also affected upon exposure to Phl. In comparison with normal growth areas of the plate (Fig. 1d), cells in the dark pink halo displayed accumulation of carbon storage material (Fig. 1e), presumably poly--hydroxybutyrate. In the inhibition zone, the cytoplasmic membrane of the few Cd cells present was physically damaged (Fig. 1f). Inhibition zones and microscopic observations in the two independent repetitions were similar, and none of the effects of Phl was observed when the Phl solvent (i.e. methanol) was added alone. This indicates that these effects were solely due to Phl.

Couillerot et al., 12 Microbiology

281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305

Effect of synthetic Phl on carotenoid production by A. brasilense Cd in liquid broth The pink pigmentation of A. brasilense Cd is linked to the synthesis of bacterioruberin-type carotenoids (Nur et al., 1981; Thirunavukkarasu et al., 2008). UHPLC analysis showed that A. brasilense Cd produced four distinct carotenoids in liquid minimal broth (Fig. S1). When 1 M to 200 M Phl was added, the same carotenoids were produced as well as one (at 1 and 10 M Phl) to five other carotenoids (between 50 M and 200 M Phl), all displaying typical UV-visible spectrum (i.e. 3 max between 400 and 550 nm, Fig. S1). At 500 and 1000 M Phl, the growth of Cd was strongly affected and no carotenoids could be detected. The total content of carotenoids was higher at 100 and especially 50 M Phl in comparison with the control (Fig. 2).

LC50 of A. brasilense strains to synthetic Phl in complex and defined media There was some fluctuation in Phl sensitivity in vitro from one A. brasilense strain to the next (Table 1). For the three main model strains of this species (i.e. Cd, Sp245 and Sp7), lethal concentration LC50 (i.e. concentration necessary for at least 50 % of growth inhibition) was (i) between 200 and 500 M Phl on LBm, and (ii) 500 M Phl (for Cd and Sp245) or between 500 and 1000 M Phl (for Sp7) on SA-Fe agar, on which Azospirillum growth is slower (data not shown). For A. brasilense Cd and Sp245 (which were used hereafter), Phl sensitivity was also observed in liquid cultures. Strain Cd reached 9.2 1.6 107 CFU mL-1 in SA-Fe and 4.7 1.0 108 CFU mL-1 in LBm, but numbers decreased to 0.9 0.3 107 CFU mL-1 in SA-Fe and 0.4 0.1 108 CFU mL-1 in LBm one day after exposure to 1000 M Phl. Similar results were obtained with A. brasilense Sp245 (not shown). Phl sensitivity levels in A. amazonense, A. doebereinerae and A. halopraeferens (LC50 on LBm between 200 and 500 M Phl depending on the strain) were similar to those in A.

Couillerot et al., 13 Microbiology

306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330

brasilense (not shown). A. irakense still grew at 1000 M Phl on both media, whereas a higher proportion of A. lipoferum strains than A. brasilense strains had a LC50 200 M Phl on LBm and LC50 < 500 M Phl on SA-Fe (not shown).

Effect of Phl+ P. fluorescens F113 and mutants on growth of A. brasilense Cd and Sp245 in vitro Whether Phl production by Pseudomonas could have an effect on A. brasilense growth was determined by comparing the in vitro effects of the Phl+ strain P. fluorescens F113, its Phlmutant F113G22 and the complemented derivative F113G22(pCU203) on growth of A. brasilense Cd and Sp245, using the two previously used solid media, which are conducive to Phl production by strain F113. Strain F113 inhibited the growth of the two A. brasilense strains within 2 d, on both medium (Table 2), and inhibition was greater on complex medium LBm compared with defined medium SA-Fe. Similar strain differences were observed at 7 d, with wider inhibition zones especially on SA-Fe agar (data not shown). Conversely, the Azospirillum strains did not inhibit P. fluorescens F113 strain even at 7 d. Compared with the wild-type F113, the Phl- biosynthetic mutant F113G22 clearly failed to inhibit the two Azospirillum strains. Mutant complementation partially restored the Azospirillum inhibition phenotype of F113 (Table 2).

Effect of Phl+ P. fluorescens F113 and mutants on Azospirillum phytostimulation To establish whether Phl+ Pseudomonas could interfere with Azospirillum phytostimulation ability, wheat seedlings were co-inoculated with both Azospirillum and Pseudomonas strains. When inoculated alone on wheat, P. fluorescens F113 improved at 7 d (i) total root length and total root surface in both wheat experiments, and (ii) root dry weight and the number of roots (only in the Cd experiment), or total root volume (in the Sp245 experiment) (Table 3). With

Couillerot et al., 14 Microbiology

331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355

the Phl- biosynthetic mutant F113G22, these phytostimulatory effects were either not found or of less magnitude. The effects of the complemented derivative F113G22(pCU203) were intermediate between those of strains F113 and F113G22. Indeed, strain F113G22(pCU203) fared like F113 for total root length and total root surface (in both experiments), but rather like F113G22 for other root parameters especially in the Cd experiment. Single inoculation of wheat with A. brasilense Cd resulted in higher root dry weight and total root length (Table 3). However, in comparison with the non-inoculated control, plants co-inoculated with strain Cd and P. fluorescens F113 (i) did not benefit from any increase in root dry weight, total root length and the number of roots, and (ii) even displayed lower total root surface. Plants co-inoculated with A. brasilense Cd and either P. fluorescens F113G22 or F113G22(pCU203) were comparable to non-inoculated plants, except that they displayed enhanced root dry weight in the case of F113G22(pCU203). Single inoculation of wheat with A. brasilense Sp245 resulted in higher total root length, total root volume and total root surface (Table 3). In comparison with the noninoculated control, plants co-inoculated with strain Sp245 and P. fluorescens F113 displayed enhanced total root length, total root volume and total root surface, similarly to plants inoculated singly. A similar trend was also observed when A. brasilense Sp245 was coinoculated with F113G22 or F113G22(pCU203).

Effect of Phl+ P. fluorescens F113 and mutants on Azospirillum cell numbers in planta When co-inoculated with P. fluorescens F113, A. brasilense Sp245 reached lower cell numbers on wheat roots at 7 d compared to those attained when inoculated singly, but the decrease was about 1 log CFU g-1 only (Fig. 3a). Unlike F113, the Phl- biosynthetic mutant F113G22 had no negative effect on strain Sp245 at 7 d, whereas the complemented derivative F113G22(pCU203) decreased cell numbers of strain Sp245 as the wild-type did.

Couillerot et al., 15 Microbiology

356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380

Unexpectedly, the Phl- biosynthetic mutant F113G22 facilitated root colonization by A. brasilense Sp245, which at 7 d reached higher levels in comparison with those attained after single inoculation. Similar findings were made with A. brasilense Cd, except that several of these differences were not significant at P < 0.05 because of higher data fluctuation levels (Fig. 3b).

Effect of Phl+ P. fluorescens F113 and mutants on root colonization patterns of A. brasilense strains To establish whether Phl+ Pseudomonas could also change microscale root colonization patterns of Azospirillum strains, distinct molecular tags were used for both types of bacteria. CLSM observations at 7 d of wheat roots colonized by P. fluorescens F113 or its Phl- mutant F113G22 (both marked with a plasmid expressing constitutively the red fluorescent protein DsRed) evidenced that each pseudomonad formed large patches of cells, which were of moderate thickness (maximum 5 m thick) (Fig. 4a-b). These cell patches were found throughout the root system (i.e. at the apex, in the root hair zones, and on older parts of the roots), but were mainly located in the grooves between epidermal cells (Fig. 4b). This colonization all along the root system of P. fluorescens F113 and F113G22 was not significantly altered when these pseudomonads were co-inoculated with an Azospirillum strain. A. brasilense Cd and Sp245 (both marked with a plasmid expressing constitutively the enhanced green fluorescent protein EGFP) were found as single cells as well as large, thick (more than 20 m high) clumps of cells forming biofilm-like structures on wheat roots (Fig. 4c and 4f). Both strains extensively colonized the root hair zone (root surface and root hairs), but much less the older parts of the wheat root system. In the presence of F113 or its Phlmutant F113G22, A. brasilense Cd and Sp245 still formed cell clumps on wheat roots, and

Couillerot et al., 16 Microbiology

381 382 383 384 385 386 387 388

some of them were very close to the cell clumps produced by F113 or F113G22 (Fig. 4d-e; gh). However, in the case of the Sp245 strain, a distinct spatialization of Azospirillum and Pseudomonas strains was observed when comparing co-inoculation with F113 and F113G22 (Fig. 4g-h). Indeed, Pseudomonas cells were found around Sp245 cell clumps in the case of F113G22, whereas Sp245 and F113 were often observed on distinct areas within a same root zone. In addition, colonization by Sp245 cells within a given root zone was more extensive when colonization by F113 cells was of lesser extent, a trend not found in the case of F113G22.

Couillerot et al., 17 Microbiology

389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413

DISCUSSION

Phl+ pseudomonads have been extensively studied for their ability to influence plant development and physiology (Iavicoli et al., 2003; Phillips et al., 2004; Brazelton et al., 2008) and to inhibit a large variety of soil-borne phytopathogenic fungi and bacteria (Weller et al., 2002; Haas & Dfago 2005; Couillerot et al., 2009). In contrast, their capacity to affect other saprophytic microbial inhabitants in the rhizosphere has received much less research attention (Natsch et al., 1998; Girlanda et al., 2001; Johansen et al., 2002), and in particular their impact on most genera of plant-beneficial microorganisms is largely unknown except for rhizobia (Walsh et al., 2003). In this work, the effect of Phl+ fluorescent Pseudomonas strains on root-colonizing A. brasilense phytostimulators was investigated, and Pseudomonas mutants were used to assess the role of Phl production ability. A. brasilense strains differed in Phl sensitivity, but most were moderately sensitive to Phl. They were more sensitive to Phl than many other saprophytic rhizobacteria (Keel et al., 1992; Monne-Loccoz et al., 2001; Walsh et al., 2003) or phytopathogenic root-associated microorganisms (Keel et al., 1992; Cronin et al., 1997a; Schouten et al., 2004), and the strain to strain fluctuation within A. brasilense was comparable to the fluctuation documented among isolates of a same species or genus in other taxa. The mode of action of Phl is poorly understood. This phenolic metabolite, which inhibits both prokaryotic and eukaryotic organisms, can probably diffuse through the cytoplasmic membrane. On one hand, Phl is involved in cross-communications between Phlproducing Pseudomonas strains (Maurhofer et al., 2004), and can enhance at very low concentrations the expression of a wide range of A. brasilense Sp245 genes, including genes involved in phytostimulation (Combes-Meynet et al., 2010). On the other hand, Phl induces the synthesis of ABC-transporter efflux pumps in Botrytis cinerea (Schouten et al., 2008) and

Couillerot et al., 18 Microbiology

414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438

impairs mitochondrial functioning in yeast by depolarisation of the mitochondrial membrane (Gleeson et al., 2010). In Pythium ultimum, it affects the plasma membrane (de Souza et al., 2003). Here, a similar effect of Phl on the cytoplasmic membrane of A. brasilense Cd was observed in areas of the plates where Phl had been deposited and where Azospirillum cells were sparse (Fig. 1f). At further distance from the Phl deposit, Cd cells became more numerous, resulting in even higher cell population biomass (pink concentration halo) compared with plates without Phl. This was also found when Phl+ strains F113 or Q2-87 were used instead of synthetic Phl (not shown). In the pink concentration halo, where the higher bacterial biomass might have resulted in carbon source depletion, Cd cells displayed cytoplasmic accumulation of poly-hydroxybutyrate-like granules. These granules favor the survival of A. brasilense under stress conditions such as carbon starvation (Tal & Okon 1985; Kadouri et al., 2002; 2003). In another experiment, visual observations of liquid cultures of A. brasilense Cd revealed an enhanced pink pigmentation when added Phl was increased from 50 to 200 M (Fig. 2). This is consistent with the increase in carotenoid content found in presence of 50 or 100 M Phl (Fig. 2). In A. brasilense Cd, carotenoids are involved in protection against oxidative damage (Hartmann & Hurek, 1988; Nur et al., 1981; Thirunavukkarasu et al., 2008), and future work with mutants impaired in carotenoid synthesis will be needed to decipher the role of carotenoids in the resistance of A. brasilense Cd to Phl. However, no correlation was found between colony pigmentation and resistance to Phl when assessing a collection of A. brasilense strains (Table 1). The Phl+ strain P. fluorescens F113 inhibited the growth on plates of the two moderately-sensitive A. brasilense strains Cd and Sp245. For A. brasilense Cd, the extent of inhibition implemented by F113 was equivalent to that caused by 13 nmol Phl on LBm and 8 nmol Phl on SA-Fe. These amounts of Phl require in the order of 5 108 CFU of F113 on rich

Couillerot et al., 19 Microbiology

439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463

medium (Duffy & Dfago, 1999), which is comparable to the inoculum level in the inhibition bioassay. Indeed, experiments performed with the Phl- biosynthetic mutant F113G22 and its Phl+ complemented derivative F113G22(pCU203) indicated that the ability of F113 to produce Phl was involved in its growth inhibition effect on Azospirillum strains. Inhibition was only partially restored with F113G22(pCU203), in accordance with the fact that it produces less Phl than F113 under laboratory conditions (Fenton et al., 1992; Cronin et al., 1997a). The effect of Phl production ability in vitro was confirmed with another Phl+ pseudomonad, i.e. strain Q2-87 (Vincent et al., 1991), which like strain F113 inhibited growth of the two A. brasilense strains within 2 d on both media tested (data not shown). In comparison with strain Q2-87, the Phl- biosynthetic mutant Q2-87::Tn5-1 (Vincent et al., 1991) displayed a lower ability to inhibit growth of the two A. brasilense strains (not shown), confirming the importance of Phl production in Azospirillum inhibition by P. fluorescens. Regarding results from wheat growth test, plants co-inoculated with A. brasilense Cd and Phl+ P. fluorescens F113 did not display phytostimulation. The significance of Phl production ability was not clear cut, because an unexpected phytostimulation effect (not found in previous work done on other crop plants) was evidenced with strains F113, but results were essentially similar with F113G22 or F113G22(pCU203). This phytostimulatory potential might be linked to ACC deaminase activity (Blaha et al., 2006) and/or production of acid indole-3-acetic evidenced in this strain (personal communication of J. CaballeroMellado, UNAM Cuernavaca, Mexico). This potential could have masked some of the inhibition effects on co-inoculated Azospirillum, as co-inoculation of wheat phytostimulator A. brasilense Sp245 with Phl+ P. fluorescens F113 lead to similar stimulation of wheat as single inoculations did. Most Rhizobium leguminosarum strains are more Phl sensitive than A. brasilense Sp245, but field inoculation with Phl+ P. fluorescens F113 had no effect on functioning of the Rhizobium-clover symbiosis (Walsh et al., 2003). Plant-growth promotion

Couillerot et al., 20 Microbiology

464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488

was not observed in co-inoculated plants with A. brasilense Cd, suggesting that strain Cd affected Pseudomonas strains, but P. fluorescens F113 was not inhibited by A. brasilense Cd in vitro. Future work will be needed to better understand the interaction mechanisms between Azospirillum and Pseudomonas strains. Root colonization is a property that may vary according to the plant species (Michiels et al., 1989; Bashan et al., 1991) as well as the plant-beneficial strains (Amus et al., 1997; Schloter & Hartmann, 1998). Here, CLSM observations evidenced that A. brasilense Cd and Sp245 did not differ from each other in their patterns of wheat root colonization, whereas they displayed a spatially-distinct distribution in comparison with Pseudomonas cells. Coinoculation of Azospirillum and Pseudomonas strains did not change overall their respective cell distribution patterns on root zones of preferential colonization, except at a microscale level where the local distribution of Azospirillum cells in the vicinity of Pseudomonas cell clusters seems to depend on Phl production ability. Thus, especially in the case of Sp245, Azospirillum and Phl+ Pseudomonas cells were often observed on distinct areas within a same root zone. Plating of wheat root samples indicated that A. brasilense Cd and Sp245 were approximately 10 times less abundant in presence of Phl+ strain P. fluorescens F113, and comparison with F113 mutants showed that this effect was due to Phl production ability. The inhibition of Azospirillum growth on roots could be explained (i) by a direct impact of Phl on Azospirillum growth in the context of a competitive interaction, and/or (ii) perhaps also by an indirect impact as Phl can change root exudation patterns (Phillips et al., 2004). In conclusion, results suggest that Phl production ability could interfere with microscale root colonization and phytostimulation potential of A. brasilense strains. However, even under gnotobiotic conditions maximizing microbial interactions on roots, phytostimulation still took place in co-inoculation experiment when A. brasilense Sp245 was used.

Couillerot et al., 21 Microbiology

489 490

ACKNOWLEDGEMENTS This study was supported in part by the Ministre Franais de la Recherche. We are grateful to E. Chapelle, V. Walker, and G. Comte (UMR CNRS 5557 Ecologie Microbienne, Universit Lyon 1) for technical assistance and/or helpful discussion. We thank G. Dfago (ETH Zrich, Switzerland), F. OGara (UCC, Cork, Ireland) and L. Thomashow (USDAARS, Pullman, Washington State) for providing Pseudomonas strains and mutant derivatives and G.V. Bloemberg (University of Zurich, Switzerland) for plasmid pMP2444. This work made use of the technical platforms CESN (at UMR 5557 Ecologie Microbienne), DTAMB, Serre and Centre Technologique des Microstructures (at IFR 41, Universit Lyon 1).

491 492 493 494 495 496 497 498

Couillerot et al., 22 Microbiology

499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523

REFERENCES Amus, B., Schloter, M., Kirchhof, G., Hutzler, P. & Hartmann, A. (1997). Improved in situ tracking of rhizosphere bacteria using dual staining with fluorescence-labeled antibodies and rRNA targeted oligonucleotides. Microb Ecol 33, 32-40. Bakker, P. A. H. M., Pieterse, C. M. J. & van Loon, L. C. (2007). Induced systemic resistance by fluorescent Pseudomonas spp. Phytopathology 97, 239-243. Bally, R. & Elmerich, C. (2007). Biocontrol of plant diseases by associative and endophytic nitrogen-fixing bacteria. In Associative and Endophytic Nitrogen-fixing Bacteria and Cyanobacterial Associations. pp. 171-190. Kluwer Academic Publishers, Dordrecht, The Netherlands. Barea, J. M., Pozo, M. J., Azcon, R. & Azcon-Aguilar, C. (2005). Microbial co-operation in the rhizosphere. J Exp Bot 56, 1761-1778. Bashan, Y., Levanony, H. & Whitmoyer, R. E. (1991). Root surface colonization of noncereal crop plants by pleomorphic Azospirillum brasilense Cd. J Gen Microbiol 137, 187196. Blaha, D., Prigent-Combaret, C., Mirza, M. S. & Monne-Loccoz, Y. (2006). Comparison of the 1-aminocyclopropane-1-carboxylic acid deaminase-encoding phytobeneficial gene acdS in phytobeneficial and pathogenic Proteobacteria. FEMS Microbiol Ecol 56, 455-470. Bloemberg, G. V., Wijfjes, A. H., Lamers, G. E., Stuurman, N. & Lugtenberg, B. J. (2000). Simultaneous imaging of Pseudomonas fluorescens WCS365 populations expressing three different autofluorescent proteins in the rhizosphere: new perspectives for studying microbial communities. Mol Plant-Microbe Interact 13, 1170-1176. Brazelton, J. N., Pfeufer, E. E., Sweat, T. A., McSpadden Gardener, B. B. & Coenen, C. (2008). 2,4-Diacetylphloroglucinol alters plant root development. Mol Plant-Microbe Interact 121, 1349-1358.

Couillerot et al., 23 Microbiology

524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548

Chilton, M. D., Currier, T. C., Farrand, S. K., Bendich, A. J., Gordon, M. P. & Nester, E. W. (1974). Agrobacterium tumefaciens DNA and PS8 bacteriophage DNA not detected in crown gall tumors. Proc. Natl. Acad. Sci. USA 71, 3672-3676. Combes-Meynet, E., Pothier, J.F., Monne-Loccoz, Y. & Prigent-Combaret C. (2010). The Pseudomonas secondary metabolite 2,4-diacetylphloroglucinol is a signal inducing rhizoplane expression of Azospirillum genes involved in plant-growth promotion. Mol Plant-Microbe Interact (In press). Costacurta, A. & Vanderleyden, J. (1995). Synthesis of phytohormones by plant-associated bacteria. Crit Rev Microbiol 21, 1-18. Couillerot, O., Prigent-Combaret, C., Caballero-Mellado, J. & Monne-Loccoz, Y. (2009). Pseudomonas fluorescens and closely-related fluorescent pseudomonads as biocontrol agents of soil-borne phytopathogens. Lett Appl Microbiol 48, 505-512. Creus, C. M., Graziano, M., Casanovas, E.M., Pereyra, M. A., Simontacchi, M., Puntarulo, S., Barassi, C. A. & Lamattina, L. (2005). Nitric oxide is involved in the Azospirillum brasilense-induced lateral root formation in tomato. Planta 221, 297-303. Cronin, D., Monne-Loccoz, Y., Fenton, A., Dunne, C., Dowling, D. N. & OGara, F. (1997a). Ecological interaction of a biocontrol Pseudomonas fluorescens strain producing 2,4-diacetylphloroglucinol with the soft rot potato pathogen Erwinia carotovora subsp. atroseptica. FEMS Microbiol Ecol 23, 95-106. Cronin, D., Monne-Loccoz, Y., Fenton, A., Dunne, C., Dowling, D. N. & OGara, F. (1997b). Role of 2,4-diacetylphloroglucinol in the interactions of the biocontrol pseudomonad strain F113 with the potato cyst nematode Globodera rostochiensis. Appl Environ Microbiol 63, 1357-1361. de Souza, J.T., Arnould, C., Deulvot, C., Lemanceau, P., Gianinazzi-Pearson, V. & Raaijmakers, J. M. (2003). Effect of 2,4-diacetylphloroglucinol on Pythium: cellular

Couillerot et al., 24 Microbiology

549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572

responses and variation in sensitivity among propagules and species. Phytopathology 93, 966-975. Dobbelaere, S., Croonenborghs, A., Thys, A., Vande Broek, A. & Vanderleyden, J. (1999). Phytostimulatory effect of Azospirillum brasilense wild type and mutant strains altered in IAA production on wheat. Plant Soil 212, 155-164. Dobbelaere, S., Croonenborghs, A., Amber, T., Ptacek, D., Vanderleyden, J., Dutto, P., Labandera-Gonzalez, C., Caballero-Mellado, J., Aguirre, J. F., Kapulnik, Y., Shimon, B., Burdman, S., Kadouri, D., Sarig, S. & Okon, Y. (2001). Responses of agronomically important crops to inoculation with Azospirillum. Aust J Plant Physiol 28, 1-9. Dobbelaere, S., Vanderleyden, J. & Okon, Y. (2003). Plant growth-promoting effects of diazotrophs in the rhizosphere. Crit Rev Plant Sci 22, 107-149. Duffy, B. K. & Dfago, G. (1999). Environmental factors modulating antibiotic and siderophore biosynthesis by Pseudomonas fluorescens biocontrol strains. Appl Environ Microbiol 65, 2429-2438. Elbeltagy, A., Nishioka, K., Sato, T., Suzuki, H., Ye, B., Hamada, T., Isawa, T., Mitsui, H. & Minamisawa, K. (2001). Endophytic colonization and in planta nitrogen fixation by a Herbaspirillum sp. isolated from wild rice species. Appl Environ Microbiol 67, 52855293. Eskew, D. L., Focht, D. D. & Ting, P. (1977). Nitrogen fixation, denitrification and pleomorphic growth in a highly pigmented Spirillum lipoferum. Appl Environ Microbiol 34, 582-585. Fenton, A. M., Stephens, P. M., Crowley, J., O'Callaghan, M. & O'Gara, F. (1992). Exploitation of gene(s) involved in 2,4-diacetylphloroglucinol biosynthesis to confer a new biocontrol capability to a Pseudomonas strain. Appl Environ Microbiol 58, 3873-3878.

Couillerot et al., 25 Microbiology

573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595

Fuentes-Ramirez, L. & Caballero-Mellado, J. (2006). Bacterial biofertilizers. In PGPR: Biocontrol and Biofertilization. pp. 143-172. Springer-Verlag, Heidelberg, Germany. Girlanda, M., Perotto, S., Monne-Loccoz, Y., Bergero, R., Lazzari, A., Dfago, G., Bonfante, P. & Luppi, A. M. (2001). Impact of biocontrol Pseudomonas fluorescens CHA0 and a genetically modified derivative on the diversity of culturable fungi in the cucumber rhizosphere. Appl Environ Microbiol 67, 1851-1864. Gleeson, O., O'Gara, F. & Morrissey, J. P. (2010). The Pseudomonas fluorescens secondary metabolite 2,4 diacetylphloroglucinol impairs mitochondrial function in Saccharomyces cerevisiae. Anton Leeuw 97, 261-73. Haas, D. & Dfago, G. (2005). Biological control of soil-borne pathogens by fluorescent pseudomonads. Nat Rev Microbiol 3, 307-319. Haas, D. & Keel, C. (2003). Regulation of antibiotic production in root-colonizing Peudomonas spp. and relevance for biological control of plant disease. Annu Rev Phytopathol 41, 117-153. Hartmann, A. & Hurek, T. (1988). Effect of carotenoid overproduction on oxygen tolerance of nitrogen fixation in Azospirillum brasilense Sp7. J Gen Microbiol 134, 2449-2455. Hontzeas, N., Zoidakis, J., Glick, B. R. & Abu-Omar, M. M. (2004). Expression and characterization of 1-aminocyclopropane-1-carboxylate deaminase from the

rhizobacterium Pseudomonas putida UW4: a key enzyme in bacterial plant growth promotion. Biochim Biophys Acta 1703, 11-19. Howell, C. R. & Stipanovic, R. D. (1979). Control of Rhizoctonia solani on cotton seedlings with Pseudomonas fluorescens and with an antibiotic produced by the bacterium. Phytopathology 69, 480-482.

Couillerot et al., 26 Microbiology

596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619

Iavicoli, A., Boutet, E., Buchala, A. & Mtraux, J. -P. (2003). Induced systemic resistance in Arabidopsis thaliana in response to root inoculation with Pseudomonas fluorescens CHA0. Mol Plant-Microbe Interact 16, 851-858. Johansen, J. E., Binnerup, S. J., Bang Lejblle, K., Mascher, F., Srensen, J. & Keel, C. (2002). Impact of biocontrol strain Pseudomonas fluorescens CHA0 on rhizosphere bacteria isolated from barley (Hordeum vulgare L.) with special reference to Cytophagalike bacteria. J Appl Microbiol 93, 1065-1074. Kabir, M., Faure, D., Heulin, T., Achouak, W. & Bally, R. (1996). Azospirillum populations in soils infested by a parasitic weed (Striga) under Sorghum cultivation in Mali, West Africa. Eur J Soil Biol 32, 157-163. Kadouri, D., Burdman, S., Jurkevitch, E. & Okon, Y. (2002). Identification and isolation of genes involved in poly(beta-hydroxybutyrate) biosynthesis in Azospirillum brasilense and characterization of a phbC mutant. Appl Environ Microbiol 68, 2943-2949. Kadouri, D., Jurkevitch, E. & Okon, Y. (2003). Involvement of the reserve material poly-hydroxybutyrate in Azospirillum brasilense stress endurance and root colonization. Appl Environ Microbiol 69, 3244-3250. Keel, C., Schnider, U., Maurhofer, M., Voisard, C., Laville, J., Burger, U., Wirthner, P., Haas, D. & Dfago, G. (1992). Suppression of root diseases by Pseudomonas fluorescens CHA0: Importance of the bacterial secondary metabolite 2,4-diacetylphloroglucinol. Mol Plant-Microbe Interact 5, 4-13. King, E. O, Ward, M. K. & Raney, D. E. (1954). Two simple media for the demonstration of pyocyanin and fluorescin. J Lab Clin Med 44, 301-307. Kyselkov, M., Kopecky, J., Frapolli, M., Dfago, G., Sagova-Mareckova, M., Grundmann, G. L. & Monne-Loccoz, Y. (2009). Comparison of rhizobacterial

Couillerot et al., 27 Microbiology

620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644

community composition in soil suppressive or conducive to tobacco black root rot disease. ISME J 3, 1127-1138. Lemanceau, P., Bakker, P. A. H. M., de Kogel, W. J., Alabouvette, C. & Schippers, B. (1992). Effect of pseudobactin 358 production by Pseudomonas putida WCS358 on suppression of fusarium wilt of carnation by nonpathogenic Fusarium oxysporum Fo47. Appl Environ Microbiol 58, 2978-2982. Maurhofer, M., Baehler, E., Notz, R., Martinez, V. & Keel, C. (2004). Cross talk between 2,4-diacetylphloroglucinol-producing biocontrol pseudomonads on wheat roots. Appl Environ Microbiol 70, 1990-1998. Michiels, K., Vanderleyden, J. & Van Gool, A. (1989). Azospirillum-plant root associations: a review. Biol Fertil Soils 8, 356-368. Mirza, M. S., Mehnaz, S., Normand, P., Prigent-Combaret, C., Monne-Loccoz, Y., Bally, R. & Malik, K. A. (2006). Molecular characterization and PCR detection of a nitrogen-fixing Pseudomonas strain promoting rice growth. Biol Fertil Soils 43, 163-170. Monne-Loccoz, Y., Tichy, H. V., O'Donnell, A., Simon, R. & O'Gara, F. (2001). Impact of 2,4-diacetylphloroglucinol-producing biocontrol strain Pseudomonas fluorescens F113 on intraspecific diversity of resident culturable fluorescent pseudomonads associated with the roots of field-grown sugar beet seedlings. Appl Environ Microbiol 67, 3418-3425. Natsch, A., Keel, C., Hebecker, N., Laasik, E. & Dfago, G. (1998). Impact of Pseudomonas fluorescens strain CHA0 and a derivative with improved biocontrol activity on the culturable resident bacterial community on cucumber roots. FEMS Microbiol Ecol 27, 365-380. Nelson, L. M. & Knowles, R. (1978). Effect of oxygen and nitrate on nitrogen fixation and denitrification by Azospirillum brasilense grown in continuous culture. Can J Microbiol 24, 1395-1403.

Couillerot et al., 28 Microbiology

645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668

Nur, I., Yuval, L. S., Okon, Y. & Henis, Y. (1981). Carotenoid composition and function in nitrogen-fixing bacteria of the genus Azospirillum. J Gen Microbiol 122, 27-32. Penot, I., Bergs, N., Guinguen, C. & Fages, J. (1992). Characterization of Azospirillum associated with maize (Zea mays) in France using biochemical tests and plasmid profiles. Can J Microbiol 38, 798-803. Phillips, D. A., Fox, T. C., King, M. D., Bhuvaneswari, T. V. & Teuber, L. R. (2004). Microbial products trigger amino acid exudation from plant roots. Plant Physiol 136, 28872894. Picard, C. & Bosco, M. (2005). Maize heterosis affects the structure and dynamics of indigenous rhizospheric auxins-producing Pseudomonas populations. FEMS Microbiol Ecol 53, 349-357. Pothier, J. F., Wisniewski-Dy, F., Weiss-Gayet, M., Monne-Loccoz, Y. & PrigentCombaret, C. (2007). Promoter-trap identification of wheat seed extract-induced genes in the plant-growth-promoting rhizobacterium Azospirillum brasilense Sp245. Microbiology 153, 3608-3622. Raaijmakers, J. M., Vlami, M. & de Souza, J. T. (2002). Antibiotic production by bacterial biocontrol agents. Antonie van Leeuwenhoek 81, 537-547. Raaijmakers, J. M., Paulitz, T. C., Steinberg, C., Alabouvette, C. & Monne-Loccoz, Y. (2009). The rhizosphere: a playground and battlefield for soilborne pathogens and beneficial microorganisms. Plant Soil 321, 341-361. Reid, J. B. & Renquist, A. R. (1997). Enhanced root production as a feed-forward response to soil water deficit in field-grown tomatoes. Aust J Plant Physiol 24, 685-692. Rezzonico, F., Zala, M., Keel, C., Duffy, B., Monne-Loccoz, Y. & Dfago, G. (2007). Is the ability of biocontrol fluorescent pseudomonads to produce the antifungal metabolite

Couillerot et al., 29 Microbiology

669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692

2,4-diacetylphloroglucinol really synonymous with higher plant protection? New Phytol 173, 861-872. Richardson, A. E., Barea, J. -M., McNeill, A. M. & Prigent-Combaret, C. (2009). Acquisition of phosphorus and nitrogen in the rhizosphere and plant growth promotion by microorganisms. Plant Soil 321, 305-339. Rinaudo, G. (1982). Fixation htrotrophe de l'azote dans la rhizosphre du riz. Thse de Doctorat d'Etat. Universit Paris-Sud, Orsay, France. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual. 2nd edn. Cold Spring Harbor Laboratory Press, New York, NY, USA. Sanguin, H., Sarniguet, A., Gazengel, K., Monne-Loccoz, Y. & Grundmann, G.L. (2009). Rhizosphere bacterial communities associated with disease suppressiveness stages of take-decline in wheat monoculture. New phytol 184, 694-707. Scher, F. M. & Baker, R. (1982). Effect of Pseudomonas putida and a synthetic iron chelator on induction of soil suppressiveness to Fusarium wilt pathogens. Phytopathology 72, 1567-1573. Schloter, M. & Hartmann, A. (1998). Endophytic and surface colonization of wheat roots (Triticum aestivum) by different Azospirillum brasilense strains studied with strain-specific monoclonal antibodies. Symbiosis 25, 159-179. Schouten, A., Van den Berg, G., Edel-Hermann, V., Steinberg, C., Gautheron, N., Alabouvette, C., De Vos, C. H., Lemanceau, P. & Raaijmakers, J. M. (2004). Defense responses of Fusarium oxysporum to 2,4-DAPG, a broad spectrum antibiotic produced by Pseudomonas fluorescens. Mol Plant-Microbe Interact 17, 1201-1211. Schouten, A., Maksimova, O., Cuesta-Arenas, Y., Van Den Berg, G. & Raaijmakers, J. M. (2008). Involvement of the ABC transporter BcAtrB and the laccase BcLCC2 in

Couillerot et al., 30 Microbiology

693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717

defence

of

Botrytis

cinerea

against

the

broad-spectrum

antibiotic

2,4-

diacetylphloroglucinol. Environ Microbiol 10, 11451157. Shanahan, P., OSullivan, D. J., Simpson, P., Glennon, J. D. & OGara, F. (1992). Isolation of 2,4-diacetylphloroglucinol from a fluorescent pseudomonad and investigation of physiological parameters influencing its production. Appl Environ Microbiol 58, 353358. Tal, S. & Okon, Y. (1985). Production of the reserve material poly-beta-hydroxybutyrate and its function in Azospirillum brasilense Cd. Can J Microbiol 31, 608-613. Tarrand, J. J., Krieg, N. R. & Dbereiner, J. (1978). A taxonomic study of the Spirillum lipoferum group, with descriptions of a new genus, Azospirillum gen. nov. and two species, Azospirillum lipoferum (Beijerinck) comb. nov. and Azospirillum brasilense sp. nov. Can J Microbiol 24, 967-980. Thirunavukkarasu, N., Mishra, M. N., Spaepen, S., Vanderleyden, J., Gross, C. A. & Tripathi, A. K. (2008). An extra-cytoplasmic function sigma factor and anti-sigma factor control carotenoid biosynthesis in Azospirillum brasilense. Microbiology 154, 2096-2105. Vincent, M. N., Harrison, L. A., Brackin, J. M., Kovacevich, P. A., Mukerji, P., Weller, D. M. & Pierson, E. A. (1991). Genetic analysis of the antifungal activity of a soilborne Pseudomonas aureofaciens strain. Appl Environ Microbiol 57, 2928-2934. Walsh, U. F., Monne-Loccoz, Y., Tichy, H. -V., Gardner, A., Corkery, D. M., Lohrke, S. & O'Gara, F. (2003). Residual impact of the biocontrol inoculant Pseudomonas fluorescens F113 on the resident population of rhizobia nodulating a red clover rotation crop. Microbial Ecol 45, 145-155. Weller, D. M, Raaijmakers, J. M, Gardener, B. B. M. & Thomashow, L. S. (2002). Microbial populations responsible for specific soil suppressiveness to plant pathogens. Annu Rev Phytopathol 40, 309-348.

Couillerot et al., 31 Microbiology

Table 1. Phl sensitivity of 11 A. brasilense strains

Strain L4 PH1 Wb1 Wb3 WN1 Sp245 Cd WS1 B506 Sp7 ZN1 Host plant Sorghum Rice Wheat Wheat Wheat Wheat Cynodon dactilon Wheat Rice Digitaria Maize Geographic Origin France France Pakistan Pakistan Pakistan Brazil USA Pakistan Japan Brazil Pakistan Reference Kabir et al., 1996 Rinaudo 1982 Blaha et al., 2006 Blaha et al., 2006 Blaha et al., 2006 Penot et al., 1992 Eskew et al., 1977 Blaha et al., 2006 Elbeltagy et al., 2001 Tarrand et al., 1978 Blaha et al., 2006 Phl LC50* LBm 50-100 200 200 200 200-500 200-500|| 200-500 200-500|| 200-500|| 200-500|| 200-500|| SA-Fe 200 500 200-500# 500-1000# 200-500# 500 500|| 500# 500 500-1000|| 500-1000#||

*Minimal concentration of synthetic Phl to inhibit growth in at least 50 % of the replicates. Concentrations tested were 5, 10, 20, 50, 100, 200, 500 and 1000 M Phl. Growth inhibition test performed using 6 replicates. LC50 was between the two Phl concentrations indicated. Growth inhibition test performed on 22 replicates. #Growth inhibition test performed on 14 replicates. || Colony pigmentation enhanced with increasing Phl quantities.

Couillerot et al., 32 Microbiology

Table 2. Growth inhibition* of A. brasilense Cd and Sp245 by Phl+ P. fluorescens F113, its Phl- mutant F113G22 and the complemented derivative F113G22(pCU203) at 2 days on LBm and SA-Fe agar, as indicated by the width (mm) of the inhibition zone around the Pseudomonas colony (mean standard error)
LBm agar F113 A. brasilense Cd A. brasilense Sp245 2.8 0.9 a 3.0 0.0 a F113G22 0b 0.8 0.5 b F113G22 (pCU203) 1.0 0.6 b 1.8 1.2 a F113 0.8 0.3 a 2.8 0.5 a F113G22 0c 0b SA-Fe agar F113G22 (pCU203) 1.0 0.6 b 00b

*In absence of Pseudomonas, there was no growth inhibition of Azospirillum strains observed. For each medium and each Azospirillum strain, letters a-b are used to indicate statistical relations between

treatments based on ANOVA and Fishers LSD tests (P < 0.05).

Couillerot et al., 33 Microbiology

Table 3. Effect of single and co-inoculation on root system properties of wheat (mean SE; n = 4 plants)
Plant parameters* Units Control Azospirillum Single inoculation F113 F113G22 F113G22 (pCU203) Pseudomonas co-inoculated with Azospirillum F113G22 F113 F113G22 (pCU203)

A. brasilense Sp245 experiment 7.2 0.9 7.3 0.4 6.2 0.3 5.7 0.7 6.9 0.8 6.3 0.3 6.9 0.2 7.7 0.4 Root dry weight mg plant-1 10 1 c 19 2 ab 19 2 ab 16 1 b 24 1 a 20 2 ab 18 2 b 15 3 b Total root length cm plant-1 13 2 b 83 18 a 73 3 a 63 13 a 78 9 a 66 8 a 70 16 a 71 9 a Total root volume mm3 plant-1 122 14 c 440 65 ab 417 21 ab 347 41 b 484 39 a 406 39 ab 393 69 ab 368 58 ab Total root surface mm2 plant-1 Number of roots 20 6 17 2 17 2 16 4 22 5 16 2 16 3 10 2 A. brasilense Cd experiment 5.5 0.5 c 6.8 0.3 b 8.0 0.4 a 6.8 0.4 b 6.7 0.2 b 5.7 0.4 bc 6.0 0.3 bc 7.8 0.3 a Root dry weight mg plant-1 19 2 c 33 3 b 46 3 a 31 3 b 43 7 a 12 2 c 14 1 c 18 2 c Total root length cm plant-1 92 9 76 3 102 8 83 10 92 7 54 14 77 14 103 17 Total root volume mm3 plant-1 565 48 cd 557 33 bc 763 55 a 564 31 bc 700 82 ab 284 48 e 359 38 de 473 37 cd Total root surface mm2 plant-1 Number of roots 25 5 bc 39 3 b 69 11 a 35 8 b 39 11 b 11 2 c 12 2 c 16 3 c *For each root system parameter studied, statistical differences between treatments are indicated with letters a-e (ANOVA and Fisher LSD tests; P < 0.05). For the A. brasilense Sp245 experiment, root dry weight and number of roots, and for the A. brasilense Cd experiment, total root volume were not significantly different between treatments.

Couillerot et al., 34 Microbiology

Diameter zone of inhibition (mm)

(a)

25 20 15 10 5 0 1 10 Phl (nmol) 100 1000

(b)
(c)
EG IG

NG

IG

EG

NG

(d)

(e)

(f)

500 nm

500 nm

500 nm

Fig. 1. Effect of synthetic Phl on A. brasilense Cd in complex medium LBm and defined minimal medium SA-Fe, after 72 h incubation of plates at 28 C: (a) diameter of growth inhibition on SA-Fe (open squares) and LBm (filled squares) according to Phl amount in the 15-L drops; (b) a darkpink

C
Couillerot et al., 35 Microbiology

halo (10-20 mm wide) of enhanced growth (EG) was observed around the zone of inhibited growth (IG) when added Phl (arrow) exceeded 150 nmol, in comparison with normal growth conditions further away in the plate (NG). White triangles indicate locations in the plate where small pieces of water agar were cut off for scanning microscopy; (c) photonic microscopy (using a binocular loop) of a vertical section performed in the A. brasilense Cd cell layer across the zones of inhibited growth (IG), enhanced growth (EG) and normal growth (NG); (d-e) transmission electronic microscopy of A. brasilense Cd cells upon exposure to Phl in the zones of (d) normal growth, (e) enhanced growth and (f) inhibited growth. Poly--hydroxybutyrate-like granules (arrows) and damage to the cytoplasmic membrane (dashed arrow) of A. brasilense Cd are shown.

Couillerot et al., 36 Microbiology

70

12
a

Carotenoid content (arbitrary AUC unit s-1OD600-1)

60 50 40 30 20 10 0 0 0.01 0.1 1 10
d d d d cd

8
b

6
d

4 2

50

100

200

Phl (M)

Fig.2. Effect of synthetic Phl (from 0 to 200 M) on the total carotenoid content (means SE), and the number of different carotenoid compounds produced by A. brasilense Cd, in ABmal, after 7d of growth. Letters a-d are used to indicate statistical differences between treatments based on ANOVA and Fishers LSD tests (P < 0.05).

Couillerot et al., 37 Microbiology

Number of different carotenoid compounds

10

(a)
12

(b)
a
11

13

ab ab a b ab

ab ab

b bc b

12

Cell number (log CFU / g)

10

Cell number (log CFU / g)

c d

11

10

c c

F113G22 (pCU203)

F113G22 (pCU203)

F113G22 (pCU203)

F113G22

F113G22

Single inoculation

Co-inoculation

Co-inoculation

Co-inoculation

Single inoculation

Co-inoculation

Co-inoculation

Co-inoculation

Fig. 3. Effect of Phl+ P. fluorescens F113, its Phl- mutant F113G22 and the complemented derivative F113G22(pCU203) on cell numbers (means SE) of (a) A. brasilense Sp245 and (b) Cd. In each panel, letters a-d are used to indicate statistical differences between treatments based on ANOVA and Fishers LSD tests (P < 0.05).

Couillerot et al., 38 Microbiology

F113G22 (pCU203)

F113G22

F113G22

7
F113 F113 Cd Cd

Sp245

Sp245

Sp245

Sp245

F113

F113

Cd

Cd

P. fluorescens F113 (Phl+)

P. fluorescens F113G22 (Phl-)

(a)
rh

(b)
F113G22 rs F113 cc rh cc rs

(c)

rs rh cc

(d)
rh rs Cd mcc Cd rs F113

(e)

rh

Cd

F113G22

A. brasilense Cd

rh

(f)

rs

(g)
Sp245 cc

(h)
Sp245 cc

A. brasilense Sp245
Sp245 rs F113G22 F113 apex

cc
20 m

Couillerot et al., 39 Microbiology

Fig. 4. CLSM observations at 7 d of wheat roots colonized by A. brasilense and/or P. fluorescens strains. Single inoculations were performed with (a) P. fluorescens F113 or (b) F113G22, or (c) A. brasilense Cd or (f) Sp245. A. brasilense strains were coinoculated with the wild-type strain F113 (Cd in (d), Sp245 in (g)) or the Phl- mutant F113G22 (Cd in (e), Sp245 in (h)). Azospirillum cells constitutively expressing EGFP are green, Pseudomonas cells constitutively expressing DsRed are red, whereas yellow indicates mixed cell clumps (mcc), and white/grey and blue backgrounds correspond to the root image formed by the transmitted light and the reflected light, respectively. rs root surface, rh root hair, cc cell clump.

Couillerot et al., 40 Microbiology