role of the antimicrobial compound 2,4-diacetylphloroglucinol in the impact of

8/3/2019 Role of the Antimicrobial Compound 2,4-Diacetylphloroglucinol in the Impact Of

1/40

Role of the antimicrobial compound 2,4-diacetylphloroglucinol in the impact of1

biocontrolPseudomonas fluorescensF113 onAzospirillum brasilense phytostimulators2

3

Olivier Couillerot, Emeline Combes-Meynet, Jol F. Pothiera, Floriant Bellvert, Elita Challita,4

Marie-Andre Poirier, Ren Rohr, Gilles Comte, Yvan Monne-Loccoz, Claire Prigent-5

Combaret *6

7

8

Universit de Lyon, F-69622, Lyon, France; Universit Lyon 1, Villeurbanne, France; CNRS,9

UMR5557, Ecologie Microbienne, Villeurbanne, France10

11

12

aPresent address: Agroscope Changins-Wdenswil ACW, Swiss Federal Research Station,13

Plant Protection Division, 8820 Wdenswil, Switzerland14

15

16

17

Running title: Phl and impact ofPseudomonas onAzospirillum18

19

20

Contents category: Environmental and Evolutionary Microbiology21

22

23

24

*Corresponding author Mailing address: UMR CNRS 5557 Ecologie Microbienne Universit25

Microbiology Papers in Press. Published January 27, 2011 as doi:10.1099/mic.0.043943-0


2/40

SUMMARY28

29

Pseudomonads producing the antimicrobial metabolite 2,4-diacetylphloroglucinol (Phl) can30

control soil-borne phytopathogens, but their impact on other plant-beneficial bacteria remains31

poorly documented. Here, the effects of synthetic Phl and Phl+ Pseudomonas fluorescens32

F113 on Azospirillum brasilense phytostimulators were investigated. Most A. brasilense33

strains were moderately sensitive to Phl. In vitro, Phl induced accumulation of carotenoids34

and poly--hydroxybutyrate-like granules, cytoplasmic membrane damage and growth35

inhibition in Azospirillum brasilense Cd. Experiments with P. fluorescens F113 and a Phl-36

mutant indicated that Phl production ability contributed to in vitro growth inhibition ofA.37

brasilense Cd and Sp245. Under gnotobiotic conditions, each of the three strains P.38

fluorescens F113,A. brasilense Cd and Sp245 stimulated wheat growth. Co-inoculation ofA.39

brasilense Sp245and Pseudomonas resulted in the same level of phytostimulation as in single40

inoculations, whereas it abolished phytostimulation when A. brasilense Cd was used.41

Pseudomonas Phl-production ability resulted in lowerAzospirillum cell numbers per root42

system (based on colony counts) and restricted microscale root colonization of neighbouring43

Azospirillum cells (based on confocal microscopy), regardless of theA. brasilense strain used.44

Therefore, this work establishes that Phl+ pseudomonads have the potential to interfere with45

A. brasilensephytostimulators on roots and with their plant growth promotion capacity.46


3/40

INTRODUCTION47

48

Microbial interactions in the rhizosphere are of paramount importance for plant growth and49

health (Barea et al., 2005; Raaijmakers et al., 2009). Many strains of fluorescent50

Pseudomonas are effective at colonizing plant roots and have been extensively studied for51

their plant-beneficial effects (Couillerot et al., 2009). Pseudomonas strains may benefit the52

plant directly, via induction of systemic resistance to pathogens (Bakkeret al., 2007),53

deamination of the ethylene precursor 1-aminocyclopropane-1-carboxylate (Hontzeas et al.,54

2004), auxin production (Picard & Bosco, 2005) and/or associative nitrogen fixation (Mirza et55

al., 2006). Plant-beneficial effects by fluorescent pseudomonads may also entail the inhibition56

of soil-borne phytoparasitic microorganisms, which often involves production of siderophores57

(Lemanceau et al., 1992) and especiallyantimicrobial secondary metabolites (Raaijmakers et58

al., 2002; Haas & Dfago 2005).59

2,4-Diacetylphloroglucinol (Phl), pyoluteorin, pyrrolnitrin, hydrogen cyanide and60

viscosinamide are some examples of well-studied antimicrobial metabolites produced by61

biocontrol strains of fluorescent pseudomonads (Haas & Keel 2003). The ability to produce62

Phl was shown, both in vitro and in vivo, as a particularly important biocontrol trait by63

comparing wild-types and non-producing mutants (Vincent et al., 1991; Fenton et al., 1992;64

Keel et al., 1992). In addition, Phl+ strains protected plants better than naturally non-65


4/40

The inhibitory properties of Phl are not restricted to phytopathogens, as non-72

pathogenic rhizosphere fungi (Girlanda et al., 2001) and bacteria (Natsch et al., 1998) might73

be inhibited as well. As far as saprophytic rhizobacteria are concerned, this possibility has74

been assessed in detail only for a very limited number of taxa, especially Bacillus (Natsch et75

al., 1998), Rhizobium leguminosarum (Walsh et al., 2003), and Cytophaga-like bacteria76

(Johansen et al., 2002). However, many other rhizobacterial taxa are also important to77

consider, because they can occur in the same rhizosphere as Phl+ pseudomonads (Barea et al.,78

2005; Kyselkov et al., 2009) and may have positive effects on the host plant. Whether these79

rhizobacteria can be inhibited by Phl (and Phl+ pseudomonads) on roots is usually unknown.80

In addition to microbial inhibition, Phl can also have an effect on root physiology81

(Brazzelton et al., 2008), which in turn may influence the conditions in which saprophytic82

rhizobacteria colonize the rhizosphere. Indeed, Phl (or the presence of Phl+ pseudomonads)83

can elicit an induced systemic resistance in the plant (Iavicoli et al., 2003), but the84

consequences are probably negligible when considering root colonization by other85

saprophytic bacteria. Much more significant, roots exposed to Phl display enhanced exudation86

of amino acids (Phillips et al., 2004), and this may enhance the ability of saprophytic bacteria87

to colonize roots. Therefore, it can be anticipated that Phl+

pseudomonads may have negative,88

neutral or positive effects on other root-colonizing saprophytic bacteria, depending on89

whether or not the latter (i) are sensitive to Phl and/or (ii) can benefit from Phl-driven amino90


5/40

Fuentes-Ramirez & Caballero-Mellado 2006) and their plant-beneficial traits include97

associative nitrogen fixation (Bally & Elmerich 2007), production of nitric oxide (Creus et al.,98

2005; Pothieret al., 2007) and phytohormones especially auxins (Costacurta& Vanderleyden99

1995; Dobbelaere et al., 1999). This may stimulate root growth, which in turn can lead to a100

better uptake of nutrients and water by the plant, and thus to better plant health and101

development (Reid & Renquist 1997; Dobbelaere et al., 2003; Richardson et al., 2009).102

Interestingly, Azospirillum and Pseudomonas indigenous populations have been found103

together in the rhizosphere of tobacco (Kyselkov et al., 2009) and wheat (Sanguin et al.,104

2009).105

In this work, A. brasilense strains from different geographic origins and host plants106

were exposed in vitro to synthetic Phl to determine whether this compound had any107

deleterious effect. A Phl+ Pseudomonas fluorescens strain and its Phl- mutant were then108

compared for their effect on cellsofA. brasilense Cd and Sp245, both in vitro and in planta.109

The co-inoculation experiments were done under gnotobiotic conditions using autofluorescent110

bacterial derivatives to assess whether colonization patterns of particular root zones (and111

phytostimulatory properties) of A. brasilense can be affected in the presence of a Phl+112

pseudomonad.113


6/40

METHODS114

115

Strains and culture conditions116

All bacterial strains (Table 1) were routinely grown at 28 C with shaking in LBm, i.e. Luria-117

Bertani medium (Sambrooket al., 1989) containing only 5 g NaCl L-1. The other media were118

the N-free medium NFb (Nelson & Knowles, 1978) supplemented with Congo red (0.25 %119

w/v) when in plates, AB medium (Chilton et al.,, 1974) containing 5 g malic acid L-1 as120

carbon source (i.e. ABmal), Kings B (King et al., 1954), and SA-Fe (Cronin et al., 1997a),121

i.e. Sucrose Asparagine (Scher & Baker 1982) supplemented with 100 M FeCl3. Antibiotics122

were used at the following concentrations: ampicillin, 40 g mL-1 (Amp40); chloramphenicol,123

15 and 30 g mL-1 (Cm15 and Cm30, respectively); gentamycin, 25 g mL-1 (Gm25).124

125

Effect of synthetic Phl on growth and cell morphology ofAzospirillum brasilense Cd126

The effect of synthetic Phl (Toronto Research Chemicals Inc., North York, Ontario) on A.127

brasilense Cd was investigated in Petri dishes, where the compound was spotted onto water128

agar containing Cd cells, as follows. Strain Cd was grown for 48 h in liquid NFb129

supplemented with 1/40 (v/v) LBm. The cells were washed twice in a 10 mM MgSO4 solution130

and adjusted to 4 107 cells mL-1 (based on optical density). Five mL of cell suspension were131

then mixed with 5 mL of molten water agar (1.5 % w/v), and the mixture was immediately132


7/40

To assess the impact of Phl on growth, the diameter of the inhibition zone was138

measured (using a ruler). To assess the impact of Phl on cell morphology, small pieces of139

water agar were cut off at different locations in the plate. Samples were then fixed with140

osmium tetroxide 2 %, contrasted with uranyl acetate 1 %, dehydrated in a graded ethanol141

series and embedded in Epon. Ultra-thin sections (0.1 m) were cut using a Reichert142

ultramicrotone (Vienna, Austria), contrasted with lead citrate and observed using a CM 120143

transmission electron microscope (Philips, Eindhoven, The Netherlands) at 100 kV. Several144

dozen cells were examined in each of the 2 replicates.145

146

Effect of synthetic Phl on carotenoid production byA. brasilense Cd147

The effect of synthetic Phl on the production of carotenoids by A. brasilense Cd was studied148

by Ultra-High-Pressure Liquid Chromatography (UHPLC), after growing cells in 24-well149

microtiter plates (Nalgene NUNC, Roskilde, Denmark). Phl was dissolved in methanol (from150

0.0001 to 1 mM Phl) and 15 L was added to 1.5 mL of ABmal broth in each well. Methanol151

(1% v/v) was used as Phl-negative control. Cd was inoculated at 1.5 106 cells mL-1. At152

least two replicates were prepared per treatment. After 7 d of growth at 28C, Cd cells were153

recovered from each well, and cell density was estimated by measuring the optical density of154

cultures at 600 nm (OD600). Cells were pelleted, lyophilised, and extracted with 1 mL of pure155

methanol by ultrasonication for 15 min. After centrifugation (10 min at 16,000 g), the156


8/40

column (100 x 1 mm, 1.9 m particle size; Thermo scientific, San Jose, CA), thermostated at163

40C. The mobile phase was a linear gradient of formic acid 1% in water (solvent A) and164

acetonitrile (solvent B). The linear gradient at the flow rate 0.15 mL min-1 was 0 to 0.5 min,165

50% solvent B; 0.5 to 7 min, 50% to 100% solvent B; 7 to 8 min, 100% solvent B. Ten166

microliters of each sample were injected. Spectra were recorded between 200 and 600 nm and167

the response at 450 nm was used for quantification of the Area Under the Curve (AUC) of all168

characteristic peaks. The latter were defined on the basis of the typical shape of their UV-169

visible spectra (Nuret al., 1981). The total content of carotenoids was expressed as arbitrary170

AUC units s-1 OD600-1. The number of distinct carotenoid metabolites was also recorded for171

each sample.172

173

Sensitivity ofA. brasilense strains to synthetic Phl174

Overnight liquid NFb cultures of 11 A. brasilense strains (Table 1) were used to inoculate175

NFb liquid medium in 96-well microtiter plates (50 L inoculum into 150 L of medium).176

Each strain was inoculated in at least six wells (i.e. six replicates). After a 24 h incubation at177

28 C (without agitation), each liquid culture was drop-inoculated onto solid LBm or SA-Fe178

medium containing Phl at final concentrations of 5, 10, 20, 50, 100, 200, 500 or 1000 M179

(previously dissolved in methanol and resulting in 1 % v/v methanol in media), and methanol180

(1 % v/v) was used in the Phl-negative control. The plates were incubated for 72 h at 28 C.181


9/40

The twoA. brasilense strains Sp245 and Cd were exposed in vitro to the Phl+ sugarbeet187

isolate P. fluorescens F113 (Shanahan et al.,, 1992), its Tn5::lacZY-induced Phl-

biosynthetic188

derivative F113G22 (Shanahan et al.,, 1992), and the Phl+ complemented mutant189

F113G22(pCU203) (Fenton et al., 1992). Strains F113 and F113G22 were grown in LBm and190

F113G22(pCU203) in LBm Cm30. The cells were then washed twice in a 10 mM MgSO4191

solution and adjusted to 109 cells mL-1 (based on optical density). Each A. brasilense strain192

was prepared in water agar, which was poured on LBm or SA-Fe medium, as described193

above, and each pseudomonad was spotted as 15 L of cell suspension. For eachA. brasilense 194

strain, growth inhibition tests were performed in duplicate and the whole experiment was195

done twice. After 72 h of incubation at 28 C, the width (radius) of the inhibition zone196

surrounding the Pseudomonas colonies was measured using a ruler. Conversely, the potential197

inhibitory activity ofA. brasilense Cd and Sp245 strains against P. fluorescens F113 was198

checked, using a similar protocol.199

200

Growth chamber experiments201

The effect of P. fluorescens F113, F113G22 and F113G22(pCU203) on growth and root202

colonization of the wheat isolate A. brasilense Sp245 and the wheat-adapted strain A.203

brasilense Cd strains was assessed on wheat (Triticum aestivum) under gnotobiotic204

conditions. The Plac-egfp plasmid pMP2444 was introduced (as described in Pothier et al.,205


10/40

107 CFU mL-1 for co-inoculation and 107 CFU mL-1 for single inoculation (based on optical212

density).213

Seeds of spring wheat (cv. Fiorina), obtained from Florimond-Desprez (Cappelle en214

Pvle, France), were surface-disinfected, as described by Pothier et al., (2007). They were215

placed on water agar (15 % w/v) and incubated 48 h in the dark at 28 C to enable216

germination. The plants were then co-inoculated using one Pseudomonas strain and one217

Azospirillum strain, and in the controls they were inoculated using only one strain or were not218

inoculated. This was done by treating pre-germinated seeds with 50 l containing 106 CFU of219

Azospirillum and 50 l containing 106 CFU ofPseudomonas or with 100 l containing 10

6220

CFU of each strain. Plants were placed near one edge of square plates (12 cm 12 cm)221

containing water agar (15 % w/v). Each A. brasilense strain was studied in a distinct222

experiment carried out independently. Four plates (containing four plants each) were used per223

treatment. The plates were placed standing in a growth chamber at 75 % relative humidity,224

with 16 h of light (63 Em-2 s-1) at 26 C and 8 h of dark at 18 C, and they were sampled225

seven days later.226

Root development at 7 d after inoculation was assessed by measuring root biomass227

and characterizing root system architecture using WinRHIZO (Rgent Instruments Inc.,228

Qubec, Canada). Four plants, each from a distinct plate, were studied per treatment.229

230


11/40

fluorescence (excitation at 488 nm and detection at 510-531 nm) and red fluorescence237

(excitation at 543 nm and detection at 563-628 nm). After acquisition of transmitted (in bright238

field mode) and/or reflected lights (in blue, detection at 456-499 nm), the single-colour239

images were overlaid into a single image using LSM software release 3.5 (Carl Zeiss).240

For quantification of inoculant bacteria, four plants were sampled per treatment at 7 d.241

Bacteria were extracted by vortexing each root system 5 min in a 15-mL Falcon tube242

containing 5 mL of 10-mM MgSO4 solution. A serial dilution was prepared in the same243

solution, and six 10-l drops from each dilution were spotted onto Kings B Amp40 Cm15244

Gm25 to quantify F113(pMP4661) or F113G22(pMP4661), Kings B Amp40 Cm30 to245

quantify F113G22(pCU203), and NFb Gm25 to quantify Cd(pMP2444) or Sp245(pMP2444).246

Colonies were not found on plates in the absence of inoculation of seedlings with the247

corresponding strain(s).248

249

Statistical analysis250

CFU of root-colonizing inoculants were expressed per g of dry root and were log-transformed251

before analysis. All results were processed by analysis of variance (ANOVA), followed when252

appropriate with Fishers least significant difference (LSD) tests. All analyses were conducted253

at P < 0.05, using S-plus software 6.1 (Hearne Scientific Software, Kilkenny, Ireland).254

255


12/40

RESULTS256

257

Effect of synthetic Phl onA. brasilense Cd on plates258

Synthetic Phl (amount ranging from 0 to 1500 nmol) was used in a simple plate assay to259

determine its effect on the model strain A. brasilense Cd. Growth of A. brasilense Cd in260

complex LBm medium and minimal defined SA-Fe medium was inhibited when at least 15261

nmol of synthetic Phl was added (as 15 L spot). With both media, the diameter of the262

inhibition zone increased with increasing Phl quantities (Fig. 1a). Only few cells were found263

by photonic microscopy in this inhibition zone.264

Surprisingly, a dark pink halo (10-20 mm wide) ofAzospirillum growth was observed265

around the inhibition zone when Cd was grown on SA-Fe (but not on LBm), provided that266

added Phl exceeded 150 nmol (Fig. 1b). Photonic microscopy indicated that cell population267

biomass was higher in the dark pink halo compared with normal growth areas of the plate268

located further away (Fig. 1c).269

In addition to growth inhibition, electronic microscopy indicated that cell morphology270

of A. brasilense Cd was also affected upon exposure to Phl. In comparison with normal271

growth areas of the plate (Fig. 1d), cells in the dark pink halo displayed accumulation of272

carbon storage material (Fig. 1e), presumably poly--hydroxybutyrate. In the inhibition zone,273

the cytoplasmic membrane of the few Cd cells present was physically damaged (Fig. 1f).274


13/40

Effect of synthetic Phl on carotenoid production byA. brasilense Cd in liquid broth281

The pink pigmentation ofA. brasilense Cd is linked to the synthesis of bacterioruberin-type282

carotenoids (Nuret al., 1981; Thirunavukkarasu et al., 2008). UHPLC analysis showed that283

A. brasilense Cd produced four distinct carotenoids in liquid minimal broth (Fig. S1). When 1284

M to 200 M Phl was added, the same carotenoids were produced as well as one (at 1 and285

10 M Phl) to five other carotenoids (between 50 M and 200 M Phl), all displaying typical286

UV-visible spectrum(i.e. 3 max between 400 and 550 nm, Fig. S1). At 500 and 1000 M287

Phl, the growth of Cd was strongly affected and no carotenoids could be detected. The total288

content of carotenoids was higher at 100 and especially 50 M Phl in comparison with the289

control (Fig. 2).290

291

LC50 ofA. brasilense strains to synthetic Phl in complex and defined media292

There was some fluctuation in Phl sensitivity in vitro from oneA. brasilense strain to the next293

(Table 1). For the three main model strains of this species (i.e. Cd, Sp245 and Sp7), lethal294

concentration LC50 (i.e. concentration necessary for at least 50 % of growth inhibition) was (i)295

between 200 and 500 M Phl on LBm, and (ii) 500 M Phl (for Cd and Sp245) or between296

500 and 1000 M Phl (for Sp7) on SA-Fe agar, on whichAzospirillum growth is slower (data297

not shown).298

ForA. brasilense Cd and Sp245 (which were used hereafter), Phl sensitivity was also299


14/40

brasilense (not shown). A. irakense still grew at 1000 M Phl on both media, whereas a306

higher proportion ofA. lipoferum strains thanA. brasilense strains had a LC50 200 M Phl307

on LBm and LC50 < 500 M Phl on SA-Fe (not shown).308

309

Effect of Phl+P. fluorescens F113 and mutants on growth ofA. brasilense Cd and Sp245310

in vitro311

Whether Phl production by Pseudomonas could have an effect on A. brasilense growth was312

determined by comparing the in vitro effects of the Phl+ strain P. fluorescens F113, its Phl-313

mutant F113G22 and the complemented derivative F113G22(pCU203) on growth of A.314

brasilense Cd and Sp245, using the two previously used solid media, which are conducive to315

Phl production by strain F113. Strain F113 inhibited the growth of the two A. brasilense316

strains within 2 d, on both medium (Table 2), and inhibition was greater on complex medium317

LBm compared with defined medium SA-Fe. Similar strain differences were observed at 7 d,318

with wider inhibition zones especially on SA-Fe agar (data not shown). Conversely, the319

Azospirillum strains did not inhibit P. fluorescens F113 strain even at 7 d.320

Compared with the wild-type F113, the Phl- biosynthetic mutant F113G22 clearly321

failed to inhibit the two Azospirillum strains. Mutant complementation partially restored the322

Azospirillum inhibition phenotype of F113 (Table 2).323

324


15/40

the Phl- biosynthetic mutant F113G22, these phytostimulatory effects were either not found or331

of less magnitude. The effects of the complemented derivative F113G22(pCU203) were332

intermediate between those of strains F113 and F113G22. Indeed, strain F113G22(pCU203)333

fared like F113 for total root length and total root surface (in both experiments), but rather334

like F113G22 for other root parameters especially in the Cd experiment.335

Single inoculation of wheat with A. brasilense Cd resulted in higher root dry weight336

and total root length (Table 3). However, in comparison with the non-inoculated control,337

plants co-inoculated with strain Cd and P. fluorescens F113 (i) did not benefit from any338

increase in root dry weight, total root length and the number of roots, and (ii) even displayed339

lower total root surface. Plants co-inoculated withA. brasilense Cd and eitherP. fluorescens340

F113G22 or F113G22(pCU203) were comparable to non-inoculated plants, except that they341

displayed enhanced root dry weight in the case of F113G22(pCU203).342

Single inoculation of wheat with A. brasilense Sp245 resulted in higher total root343

length, total root volume and total root surface (Table 3). In comparison with the non-344

inoculated control, plants co-inoculated with strain Sp245 and P. fluorescens F113 displayed345

enhanced total root length, total root volume and total root surface, similarly to plants346

inoculated singly. A similar trend was also observed when A. brasilense Sp245 was co-347

inoculated with F113G22 or F113G22(pCU203).348

349


16/40

Unexpectedly, the Phl- biosynthetic mutant F113G22 facilitated root colonization by A.356

brasilense Sp245, which at 7 d reached higher levels in comparison with those attained after357

single inoculation. Similar findings were made with A. brasilense Cd, except that several of358

these differences were not significant at P < 0.05 because of higher data fluctuation levels359

(Fig. 3b).360

361

Effect of Phl+ P. fluorescens F113 and mutants on root colonization patterns ofA.362

brasilense strains363

To establish whether Phl+ Pseudomonas could also change microscale root colonization364

patterns ofAzospirillum strains, distinct molecular tags were used for both types of bacteria.365

CLSM observations at 7 d of wheat roots colonized by P. fluorescens F113 or its Phl- mutant366

F113G22 (both marked with a plasmid expressing constitutively the red fluorescent protein367

DsRed) evidenced that each pseudomonad formed large patches of cells, which were of368

moderate thickness (maximum 5 m thick) (Fig. 4a-b). These cell patches were found369

throughout the root system (i.e. at the apex, in the root hair zones, and on older parts of the370

roots), but were mainly located in the grooves between epidermal cells (Fig. 4b). This371

colonization all along the root system of P. fluorescens F113 and F113G22 was not372

significantly altered when these pseudomonads were co-inoculated with an Azospirillum373

strain.374


17/40

some of them were very close to the cell clumps produced by F113 or F113G22 (Fig. 4d-e; g-381

h). However, in the case of the Sp245 strain, a distinct spatialization ofAzospirillum and382

Pseudomonas strains was observed when comparing co-inoculation with F113 and F113G22383

(Fig. 4g-h). Indeed, Pseudomonas cells were found around Sp245 cell clumps in the case of384

F113G22, whereas Sp245 and F113 were often observed on distinct areas within a same root385

zone. In addition, colonization by Sp245 cells within a given root zone was more extensive386

when colonization by F113 cells was of lesser extent, a trend not found in the case of387

F113G22.388


18/40

DISCUSSION389

390

Phl+ pseudomonads have been extensively studied for their ability to influence plant391

development and physiology (Iavicoli et al., 2003; Phillips et al., 2004; Brazelton et al., 2008)392

and to inhibit a large variety of soil-borne phytopathogenic fungi and bacteria (Welleret al.,393

2002; Haas & Dfago 2005; Couillerot et al., 2009). In contrast, their capacity to affect other394

saprophytic microbial inhabitants in the rhizosphere has received much less research attention395

(Natsch et al., 1998; Girlanda et al., 2001; Johansen et al., 2002), and in particular their396

impact on most genera of plant-beneficial microorganisms is largely unknown except for397

rhizobia (Walsh et al., 2003). In this work, the effect of Phl+ fluorescent Pseudomonas strains398

on root-colonizing A. brasilense phytostimulators was investigated, and Pseudomonas399

mutants were used to assess the role of Phl production ability.400

A. brasilense strains differed in Phl sensitivity, but most were moderately sensitive to401

Phl. They were more sensitive to Phl than many other saprophytic rhizobacteria (Keel et al.,402

1992; Monne-Loccoz et al., 2001; Walsh et al., 2003) or phytopathogenic root-associated403

microorganisms (Keel et al., 1992; Cronin et al., 1997a; Schouten et al., 2004), and the strain404

to strain fluctuation within A. brasilense was comparable to the fluctuation documented405

among isolates of a same species or genus in other taxa.406

The mode of action of Phl is poorly understood. This phenolic metabolite, which407


19/40

impairs mitochondrial functioning in yeast by depolarisation of the mitochondrial membrane414

(Gleeson et al., 2010). In Pythiumultimum, it affects the plasma membrane (de Souza et al.,415

2003). Here, a similar effect of Phl on the cytoplasmic membrane ofA. brasilense Cd was416

observed in areas of the plates where Phl had been deposited and where Azospirillum cells417

were sparse (Fig. 1f).418

At further distance from the Phl deposit, Cd cells became more numerous, resulting in419

even higher cell population biomass (pink concentration halo) compared with plates without420

Phl. This was also found when Phl+ strains F113 or Q2-87 were used instead of synthetic Phl421

(not shown). In the pink concentration halo, where the higher bacterial biomass might have422

resulted in carbon source depletion, Cd cells displayed cytoplasmic accumulation of poly--423

hydroxybutyrate-like granules. These granules favor the survival ofA. brasilense under stress424

conditions such as carbon starvation (Tal & Okon 1985; Kadouri et al., 2002; 2003). In425

another experiment, visual observations of liquid cultures ofA. brasilense Cd revealed an426

enhanced pink pigmentation when added Phl was increased from 50 to 200 M (Fig. 2). This427

is consistent with the increase in carotenoid content found in presence of 50 or 100 M Phl428

(Fig. 2).InA. brasilense Cd, carotenoids are involved in protection against oxidative damage429

(Hartmann & Hurek, 1988; Nuret al., 1981; Thirunavukkarasu et al., 2008), and future work430

with mutants impaired in carotenoid synthesis will be needed to decipher the role of431

carotenoids in the resistance ofA. brasilense Cd to Phl. However, no correlation was found432


20/40

medium (Duffy & Dfago, 1999), which is comparable to the inoculum level in the inhibition439

bioassay. Indeed, experiments performed with the Phl-

biosynthetic mutant F113G22 and its440

Phl+ complemented derivative F113G22(pCU203) indicated that the ability of F113 to441

produce Phl was involved in its growth inhibition effect on Azospirillum strains. Inhibition442

was only partially restored with F113G22(pCU203), in accordance with the fact that it443

produces less Phl than F113 under laboratory conditions (Fenton et al., 1992; Cronin et al.,444

1997a). The effect of Phl production ability in vitro was confirmed with another Phl+445

pseudomonad, i.e. strain Q2-87(Vincent et al., 1991), which like strain F113 inhibited growth446

of the two A. brasilense strains within 2 d on both media tested (data not shown). In447

comparison with strain Q2-87, the Phl- biosynthetic mutant Q2-87::Tn5-1 (Vincent et al.,448

1991) displayed a lower ability to inhibit growth of the two A. brasilense strains (not shown),449

confirming the importance of Phl production inAzospirillum inhibition by P. fluorescens.450

Regarding results from wheat growth test, plants co-inoculated with A. brasilense Cd451

and Phl+ P. fluorescens F113 did not display phytostimulation. The significance of Phl452

production ability was not clear cut, because an unexpected phytostimulation effect (not453

found in previous work done on other crop plants) was evidenced with strains F113, but454

results were essentially similar with F113G22 or F113G22(pCU203). This phytostimulatory455

potential might be linked to ACC deaminase activity (Blaha et al., 2006) and/or production of456

acid indole-3-acetic evidenced in this strain (personal communication of J. Caballero-457


21/40

was not observed in co-inoculated plants with A. brasilense Cd, suggesting that strain Cd464

affected Pseudomonas strains, but P. fluorescens F113 was not inhibited byA. brasilense Cd465

in vitro. Future work will be needed to better understand the interaction mechanisms between466

Azospirillum and Pseudomonas strains.467

Root colonization is a property that may vary according to the plant species (Michiels et468

al., 1989; Bashan et al., 1991) as well as the plant-beneficial strains (Amus et al., 1997;469

Schloter & Hartmann, 1998). Here, CLSM observations evidenced that A. brasilense Cd and470

Sp245 did not differ from each other in their patterns of wheat root colonization, whereas they471

displayed a spatially-distinct distribution in comparison with Pseudomonas cells. Co-472

inoculation ofAzospirillum and Pseudomonas strains did not change overall their respective473

cell distribution patterns on root zones of preferential colonization, except at a microscale474

level where the local distribution ofAzospirillum cells in the vicinity ofPseudomonas cell475

clusters seems to depend on Phl production ability. Thus, especially in the case of Sp245,476

Azospirillum and Phl+Pseudomonas cells were often observed on distinct areas within a same477

root zone. Plating of wheat root samples indicated that A. brasilense Cd and Sp245 were478

approximately 10 times less abundant in presence of Phl+ strain P. fluorescens F113, and479

comparison with F113 mutants showed that this effect was due to Phl production ability. The480

inhibition ofAzospirillum growth on roots could be explained (i) by a direct impact of Phl on481

Azospirillum growth in the context of a competitive interaction, and/or (ii) perhaps also by an482


22/40

489

ACKNOWLEDGEMENTS490

This study was supported in part by the Ministre Franais de la Recherche. We are grateful491

to E. Chapelle, V. Walker, and G. Comte (UMR CNRS 5557 Ecologie Microbienne,492

Universit Lyon 1) for technical assistance and/or helpful discussion. We thank G. Dfago493

(ETH Zrich, Switzerland), F. OGara (UCC, Cork, Ireland) and L. Thomashow (USDA-494

ARS, Pullman, Washington State) for providing Pseudomonas strains and mutant derivatives495

and G.V. Bloemberg (University of Zurich, Switzerland) for plasmid pMP2444. This work496

made use of the technical platforms CESN (at UMR 5557 Ecologie Microbienne), DTAMB,497

Serre and Centre Technologique des Microstructures (at IFR 41, Universit Lyon 1).498


23/40

REFERENCES499

Amus, B., Schloter, M., Kirchhof, G., Hutzler, P. & Hartmann, A. (1997). Improved in500

situ tracking of rhizosphere bacteria using dual staining with fluorescence-labeled501

antibodies and rRNA targeted oligonucleotides.Microb Ecol33, 32-40.502

Bakker, P. A. H. M., Pieterse, C. M. J. & van Loon, L. C. (2007). Induced systemic503

resistance by fluorescent Pseudomonas spp. Phytopathology97, 239-243.504

Bally, R. & Elmerich, C. (2007). Biocontrol of plant diseases by associative and endophytic505

nitrogen-fixing bacteria. In Associative and Endophytic Nitrogen-fixing Bacteria and506

Cyanobacterial Associations. pp. 171-190.Kluwer Academic Publishers, Dordrecht, The507

Netherlands.508

Barea, J. M., Pozo, M. J., Azcon, R. & Azcon-Aguilar, C. (2005). Microbial co-operation509

in the rhizosphere.J Exp Bot56, 1761-1778.510

Bashan, Y., Levanony, H. & Whitmoyer, R. E. (1991). Root surface colonization of non-511

cereal crop plants by pleomorphic Azospirillum brasilense Cd.J Gen Microbiol137, 187-512

196.513

Blaha, D., Prigent-Combaret, C., Mirza, M. S. & Monne-Loccoz, Y. (2006). Comparison514

of the 1-aminocyclopropane-1-carboxylic acid deaminase-encoding phytobeneficial gene515

acdS in phytobeneficial and pathogenic Proteobacteria. FEMS Microbiol Ecol56, 455-470.516

Bloemberg, G. V., Wijfjes, A. H., Lamers, G. E., Stuurman, N. & Lugtenberg, B. J.517


24/40

Chilton, M. D., Currier, T. C., Farrand, S. K., Bendich, A. J., Gordon, M. P. & Nester,524

E. W.(1974).Agrobacterium tumefaciens DNA and PS8 bacteriophage DNA not detected525

in crown gall tumors. Proc. Natl. Acad. Sci. USA71, 3672-3676.526

Combes-Meynet, E., Pothier, J.F., Monne-Loccoz, Y. & Prigent-Combaret C. (2010).527

The Pseudomonas secondary metabolite 2,4-diacetylphloroglucinol is a signal inducing528

rhizoplane expression ofAzospirillum genes involved in plant-growth promotion. Mol529

Plant-Microbe Interact(In press).530

Costacurta, A. & Vanderleyden, J. (1995). Synthesis of phytohormones by plant-associated531

bacteria. Crit Rev Microbiol21, 1-18.532

Couillerot, O., Prigent-Combaret, C., Caballero-Mellado, J. & Monne-Loccoz, Y.533

(2009). Pseudomonas fluorescens and closely-related fluorescent pseudomonads as534

biocontrol agents of soil-borne phytopathogens.Lett Appl Microbiol48, 505-512.535

Creus, C. M., Graziano, M., Casanovas, E.M., Pereyra, M. A., Simontacchi, M.,536

Puntarulo, S., Barassi, C. A. & Lamattina, L. (2005). Nitric oxide is involved in the537

Azospirillum brasilense-induced lateral root formation in tomato. Planta221, 297-303.538

Cronin, D., Monne-Loccoz, Y., Fenton, A., Dunne, C., Dowling, D. N. & OGara, F.539

(1997a). Ecological interaction of a biocontrol Pseudomonas fluorescens strain producing540

2,4-diacetylphloroglucinol with the soft rot potato pathogen Erwinia carotovora subsp.541

atroseptica. FEMS Microbiol Ecol 23, 95-106.542


25/40

responses and variation in sensitivity among propagules and species. Phytopathology93,549

966-975.550

Dobbelaere, S., Croonenborghs, A., Thys, A., Vande Broek, A. & Vanderleyden, J.551

(1999). Phytostimulatory effect of Azospirillum brasilense wild type and mutant strains552

altered in IAA production on wheat. Plant Soil212, 155-164.553

Dobbelaere, S., Croonenborghs, A., Amber, T., Ptacek, D., Vanderleyden, J., Dutto, P.,554

Labandera-Gonzalez, C., Caballero-Mellado, J., Aguirre, J. F., Kapulnik, Y., Shimon,555

B., Burdman, S., Kadouri, D., Sarig, S. & Okon, Y. (2001). Responses of agronomically556

important crops to inoculation withAzospirillum.Aust J Plant Physiol28, 1-9.557

Dobbelaere, S., Vanderleyden, J. & Okon, Y. (2003). Plant growth-promoting effects of558

diazotrophs in the rhizosphere. Crit Rev Plant Sci22, 107-149.559

Duffy, B. K. & Dfago, G. (1999). Environmental factors modulating antibiotic and560

siderophore biosynthesis by Pseudomonas fluorescens biocontrol strains. Appl Environ561

Microbiol65, 2429-2438.562

Elbeltagy, A., Nishioka, K., Sato, T., Suzuki, H., Ye, B., Hamada, T., Isawa, T., Mitsui,563

H. & Minamisawa, K. (2001). Endophytic colonization and in planta nitrogen fixation by564

a Herbaspirillum sp. isolated from wild rice species. Appl Environ Microbiol 67, 5285-565

5293.566

Eskew, D. L., Focht, D. D. & Ting, P. (1977). Nitrogen fixation, denitrification and567


26/40

Fuentes-Ramirez, L. & Caballero-Mellado, J. (2006). Bacterial biofertilizers. In PGPR:573

Biocontrol and Biofertilization. pp.143-172. Springer-Verlag, Heidelberg, Germany.574

Girlanda, M., Perotto, S., Monne-Loccoz, Y., Bergero, R., Lazzari, A., Dfago, G.,575

Bonfante, P. & Luppi, A. M. (2001). Impact of biocontrol Pseudomonas fluorescens576

CHA0 and a genetically modified derivative on the diversity of culturable fungi in the577

cucumber rhizosphere.Appl Environ Microbiol67, 1851-1864.578

Gleeson, O., O'Gara, F. & Morrissey, J. P. (2010). The Pseudomonas fluorescens579

secondary metabolite 2,4 diacetylphloroglucinol impairs mitochondrial function in580

Saccharomyces cerevisiae. Anton Leeuw 97, 261-73.581

Haas, D. & Dfago, G. (2005). Biological control of soil-borne pathogens by fluorescent582

pseudomonads.Nat Rev Microbiol3, 307-319.583

Haas, D. & Keel, C. (2003). Regulation of antibiotic production in root-colonizing584

Peudomonas spp. and relevance for biological control of plant disease. Annu Rev585

Phytopathol41, 117-153.586

Hartmann, A. & Hurek, T. (1988). Effect of carotenoid overproduction on oxygen tolerance587

of nitrogen fixation inAzospirillum brasilense Sp7.J Gen Microbiol134, 2449-2455.588

Hontzeas, N., Zoidakis, J., Glick, B. R. & Abu-Omar, M. M. (2004). Expression and589

characterization of 1-aminocyclopropane-1-carboxylate deaminase from the590

rhizobacterium Pseudomonas putida UW4: a key enzyme in bacterial plant growth591


27/40

Iavicoli, A., Boutet, E., Buchala, A. & Mtraux, J. -P. (2003). Induced systemic resistance596

in Arabidopsis thaliana in response to root inoculation with Pseudomonas fluorescens597

CHA0.Mol Plant-Microbe Interact16, 851-858.598

Johansen, J. E., Binnerup, S. J., Bang Lejblle, K., Mascher, F., Srensen, J. & Keel, C.599

(2002). Impact of biocontrol strain Pseudomonas fluorescens CHA0 on rhizosphere600

bacteria isolated from barley (Hordeum vulgare L.) with special reference to Cytophaga-601

like bacteria.J Appl Microbiol93, 1065-1074.602

Kabir, M., Faure, D., Heulin, T., Achouak, W. & Bally, R. (1996). Azospirillum603

populations in soils infested by a parasitic weed (Striga) under Sorghum cultivation in604

Mali, West Africa.Eur J Soil Biol32, 157-163.605

Kadouri, D., Burdman, S., Jurkevitch, E. & Okon, Y. (2002). Identification and isolation606

of genes involved in poly(beta-hydroxybutyrate) biosynthesis in Azospirillum brasilense607

and characterization of aphbCmutant.Appl Environ Microbiol68, 2943-2949.608

Kadouri, D., Jurkevitch, E. & Okon, Y. (2003). Involvement of the reserve material poly--609

hydroxybutyrate in Azospirillum brasilense stress endurance and root colonization. Appl610

Environ Microbiol69, 3244-3250.611

Keel, C., Schnider, U., Maurhofer, M., Voisard, C., Laville, J., Burger, U., Wirthner, P.,612

Haas, D. & Dfago, G. (1992). Suppression of root diseases by Pseudomonas fluorescens613

CHA0: Importance of the bacterial secondary metabolite 2,4-diacetylphloroglucinol. Mol614


28/40

community composition in soil suppressive or conducive to tobacco black root rot disease.620

ISME J3, 1127-1138.621

Lemanceau, P., Bakker, P. A. H. M., de Kogel, W. J., Alabouvette, C. & Schippers, B.622

(1992). Effect of pseudobactin 358 production by Pseudomonas putida WCS358 on623

suppression of fusarium wilt of carnation by nonpathogenic Fusarium oxysporum Fo47.624

Appl Environ Microbiol58, 2978-2982.625

Maurhofer, M., Baehler, E., Notz, R., Martinez, V. & Keel, C. (2004). Cross talk between626

2,4-diacetylphloroglucinol-producing biocontrol pseudomonads on wheat roots. Appl627

Environ Microbiol70,1990-1998.628

Michiels, K., Vanderleyden, J. & Van Gool, A. (1989). Azospirillum-plant root629

associations: a review.Biol Fertil Soils8, 356-368.630

Mirza, M. S., Mehnaz, S., Normand, P., Prigent-Combaret, C., Monne-Loccoz, Y.,631

Bally, R. & Malik, K. A. (2006). Molecular characterization and PCR detection of a632

nitrogen-fixing Pseudomonas strain promoting rice growth.Biol Fertil Soils43, 163-170.633

Monne-Loccoz, Y., Tichy, H. V., O'Donnell, A., Simon, R. & O'Gara, F. (2001). Impact634

of 2,4-diacetylphloroglucinol-producing biocontrol strain Pseudomonas fluorescens F113635

on intraspecific diversity of resident culturable fluorescentpseudomonads associated with636

the roots of field-grown sugar beet seedlings. Appl Environ Microbiol67, 3418-3425.637

Natsch, A., Keel, C., Hebecker, N., Laasik, E. & Dfago, G. (1998). Impact of638


29/40

Nur, I., Yuval, L. S., Okon, Y. & Henis, Y. (1981). Carotenoid composition and function in645

nitrogen-fixing bacteria of the genusAzospirillum.J Gen Microbiol122, 27-32.646

Penot, I., Bergs, N., Guinguen, C. & Fages, J. (1992). Characterization ofAzospirillum647

associated with maize (Zea mays) in France using biochemical tests and plasmid profiles.648

Can J Microbiol38, 798-803.649

Phillips, D. A., Fox, T. C., King, M. D., Bhuvaneswari, T. V. & Teuber, L. R. (2004).650

Microbial products trigger amino acid exudation from plant roots. Plant Physiol136, 2887-651

2894.652

Picard, C. & Bosco, M. (2005). Maize heterosis affects the structure and dynamics of653

indigenous rhizospheric auxins-producing Pseudomonas populations. FEMS Microbiol654

Ecol53, 349-357.655

Pothier, J. F., Wisniewski-Dy, F., Weiss-Gayet, M., Monne-Loccoz, Y. & Prigent-656

Combaret, C. (2007). Promoter-trap identification of wheat seed extract-induced genes in657

the plant-growth-promoting rhizobacterium Azospirillum brasilense Sp245. Microbiology658

153, 3608-3622.659

Raaijmakers, J. M., Vlami, M. & de Souza, J. T. (2002). Antibiotic production by bacterial660

biocontrol agents.Antonie van Leeuwenhoek81, 537-547.661

Raaijmakers, J. M., Paulitz, T. C., Steinberg, C., Alabouvette, C. & Monne-Loccoz, Y.662

(2009). The rhizosphere: a playground and battlefield for soilborne pathogens and663


30/40

2,4-diacetylphloroglucinol really synonymous with higher plant protection? New Phytol669

173, 861-872.670

Richardson, A. E., Barea, J. -M., McNeill, A. M. & Prigent-Combaret, C. (2009). 671

Acquisition of phosphorus and nitrogen in the rhizosphere and plant growth promotion by672

microorganisms. Plant Soil321, 305-339.673

Rinaudo, G. (1982). Fixation htrotrophe de l'azote dans la rhizosphre du riz. Thse de674

Doctorat d'Etat. Universit Paris-Sud, Orsay, France.675

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: A Laboratory676

Manual. 2nd edn. Cold Spring Harbor Laboratory Press, New York, NY, USA.677

Sanguin, H., Sarniguet, A., Gazengel, K., Monne-Loccoz, Y. & Grundmann, G.L.678

(2009). Rhizosphere bacterial communities associated with disease suppressiveness stages679

of take-decline in wheat monoculture.New phytol184, 694-707.680

Scher, F. M. & Baker, R. (1982). Effect of Pseudomonas putida and a synthetic iron681

chelator on induction of soil suppressiveness to Fusariumwilt pathogens. Phytopathology682

72, 1567-1573.683

Schloter, M. & Hartmann, A. (1998). Endophytic and surface colonization of wheat roots684

(Triticum aestivum) by differentAzospirillum brasilense strains studied with strain-specific685

monoclonal antibodies. Symbiosis25, 159-179.686

Schouten, A., Van den Berg, G., Edel-Hermann, V., Steinberg, C., Gautheron, N.,687


31/40

defence of Botrytis cinerea against the broad-spectrum antibiotic 2,4-693

diacetylphloroglucinol.Environ Microbiol10, 11451157.694

Shanahan, P., OSullivan, D. J., Simpson, P., Glennon, J. D. & OGara, F. (1992). 695

Isolation of 2,4-diacetylphloroglucinol from a fluorescent pseudomonad and investigation696

of physiological parameters influencing its production. Appl Environ Microbiol58, 353-697

358.698

Tal, S. & Okon, Y. (1985). Production of the reserve material poly-beta-hydroxybutyrate and699

its function inAzospirillum brasilenseCd. Can J Microbiol31, 608-613.700

Tarrand, J. J., Krieg, N. R. & Dbereiner, J. (1978). A taxonomic study of the Spirillum701

lipoferum group, with descriptions of a new genus,Azospirillum gen. nov. and two species,702

Azospirillum lipoferum (Beijerinck) comb. nov. andAzospirillum brasilense sp. nov. Can J703

Microbiol 24, 967-980.704

Thirunavukkarasu, N., Mishra, M. N., Spaepen, S., Vanderleyden, J., Gross, C. A. &705

Tripathi, A. K. (2008). An extra-cytoplasmic function sigma factor and anti-sigma factor706

control carotenoid biosynthesis inAzospirillum brasilense.Microbiology154, 2096-2105.707

Vincent, M. N., Harrison, L. A., Brackin, J. M., Kovacevich, P. A., Mukerji, P., Weller,708

D. M. & Pierson, E. A. (1991). Genetic analysis of the antifungal activity of a soilborne709

Pseudomonasaureofaciens strain.Appl Environ Microbiol57, 2928-2934.710

Walsh, U. F., Monne-Loccoz, Y., Tichy, H. -V., Gardner, A., Corkery, D. M., Lohrke, S.711


32/40

Table 1. Phl sensitivity of 11A. brasilense strains

Phl LC50*Strain Host plant

Geographic

OriginReference

LBm SA-Fe

L4 Sorghum France Kabiret al., 1996 50-100 200

PH1 Rice France Rinaudo 1982 200 500

Wb1 Wheat Pakistan Blaha et al., 2006 200 200-500#

Wb3 Wheat Pakistan Blaha et al., 2006 200 500-1000#

WN1 Wheat Pakistan Blaha et al., 2006 200-500 200-500#

Sp245 Wheat Brazil Penot et al., 1992 200-500|| 500

Cd Cynodon dactilon USA Eskew et al., 1977 200-500 500||

WS1 Wheat Pakistan Blaha et al., 2006 200-500|| 500#

B506 Rice Japan Elbeltagy et al., 2001 200-500|| 500

Sp7 Digitaria Brazil Tarrand et al., 1978 200-500|| 500-1000||

ZN1 Maize Pakistan Blaha et al., 2006 200-500|| 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.


33/40

Couillerot et al., 33Microbiology

Table 2. Growth inhibition* ofA. 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 SA-Fe agar

F113 F113G22F113G22

(pCU203)F113 F113G22

F113G22

(pCU203)

A. brasilense Cd 2.8 0.9 a 0 b 1.0 0.6 b 0.8 0.3 a 0 c 1.0 0.6 b

A. brasilense Sp245 3.0 0.0 a 0.8 0.5 b 1.8 1.2 a 2.8 0.5 a 0 b 0 0 b

*In absence ofPseudomonas, there was no growth inhibition ofAzospirillum 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).


34/40


Table 3. Effect of single and co-inoculation on root system properties of wheat (mean SE; n = 4 plants)Plant Single inoculation Pseudomonas co-inoculated withAzospirillum

parameters* Units Control Azospirillum F113 F113G22F113G22(pCU203)

F113 F113G22F113G22(pCU203)

A. brasilense Sp245 experimentRoot dry weight mg plant-1 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 0.4Total root length cm 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 volume mm3 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 aTotal root surface mm2 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 abNumber of roots 20 6 17 2 17 2 16 4 22 5 16 2 16 3 10 2A. brasilense Cd experimentRoot dry weight mg plant-1 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 aTotal root length cm 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 cTotal root volume mm3 plant-1 92 9 76 3 102 8 83 10 92 7 54 14 77 14 103 17Total root surface mm2 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 cdNumber 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 Sp245experiment, root dry weight and number of roots, and for theA. brasilense Cd experiment, total root volume were not significantly different between treatments.


35/40

0

5

10

15

20

25

1 10 100 1000

Phl (nmol)

Diameterzoneofinhibition

(mm)

(b)

EG IG

NG

(c)

EGIG NG

(d) (e) (f)

(a)


36/40

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 ofA. 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) ofA. brasilense Cd are shown.


37/40

d d d d

d

cd

b

a

0

10

20

30

40

50

60

70

0 0.01 0.1 1 10 50 100 200

2

4

6

8

10

12

Carotenoidcontent

(arb

itraryAUCunits-1OD

600

-1)

Phl (M)

Numberofdifferent

carotenoidcompounds

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).


38/40


6

7

8

9

10

11

12

Cellnumber(logCFU/g)

c

b

a

b

d

bcb

d

bb

Sp245

Sp245

Sp245

Sp245

F113

F113

F113G22

(pCU203)

F113G22

F113G22

(pCU203)

F113G22

Single inoculation Co-inoculation Co-inoculation Co-inoculation

7

8

9

10

11

12

13

Cellnumber(logCFU/g)

Cd

Cd

Cd

Cd

F113

F113

F113G22

(pCU203)

F113G22

F113G22

(pCU203)

F113G22

Single inoculation Co-inoculation Co-inoculation Co-inoculation

c

ab a

ab

c

bab

c

ab

ab

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).

(a) (b)

(a) b

P. fluorescens F113 (Phl+) P. fluorescens F113G22 (Phl-)


39/40


cc

F113

rhrs

cc

F113G22

rh

rs

d

(g)

Cd

rh

F113

rs

rs

mcc

rh

F113G22

Cd

rh

Cd

cc

rs

(f)

F113G22

Sp245

apex

cc

rs

Sp245

F113

cc

rs

Sp245

cc

(a) b

c e

(h)

A. brasilense Sp245

20 m

rh

A. brasilense Cd

Fig. 4. CLSM observations at 7 d of wheat roots colonized by A. brasilense and/orP. fluorescens strains. Single inoculations were performed with (a) P.


40/40


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.

role of the antimicrobial compound 2,4-diacetylphloroglucinol in the impact of

Documents