Professional Documents
Culture Documents
c 2009 Federation of European Microbiological Societies FEMS Microbiol Ecol 69 (2009) 62–74
Published by Blackwell Publishing Ltd. All rights reserved
Biocontrol of Rhizoctonia solani and rhizosphere competence 63
understanding of the ecology of BCA in the rhizosphere (75 mg mL 1). The 15 rifampicin-resistant mutant strains were
where the BCA interact not only with the plant and the tested for inhibitory activity against two highly pathogenic
pathogen but also with the indigenous microbial commu- isolates belonging to the R. solani anastomosis groups AG1-IB
nity (Götz et al., 2006; Scherwinski et al., 2008). and AG2 (Faltin et al., 2004; Grosch et al., 2004). In vitro
In this study, 15 strains were selected from a collection of inhibitory activity against R. solani AG1-IB and AG2 was
bacterial isolates with in vitro antagonistic activity towards determined essentially as described by Adesina et al. (2007).
R. solani AG3 (the causal agent of black scurf on potato) and/ The mutant strains were stored at 80 1C in Luria–
or Fusarium oxysporum (which causes wilting disease in Bertani broth (ROTH, Germany) containing 20% glycerol
diverse crops), which were previously isolated and character- supplemented with rifampicin (75 mg mL 1).
ized (Adesina et al., 2007). The in vitro antagonists that
originated from four disease-suppressive soils located in Lettuce rhizosphere colonization assays
France, the Netherlands, Sweden and the United Kingdom
To determine the rhizosphere competence of the 15 rifam-
were first screened for their ability to colonize the rhizosphere
picin-resistant strains, rhizosphere colonization assays were
of lettuce in nonsterile soil. Based on their colonization ability,
performed in nonsterile soil under greenhouse conditions.
eight strains were selected for growth chamber experiments.
Lettuce seeds (cultivar ‘Tizian’, Syngenta, Bad Salzuflen,
The work presented here aimed to investigate the potential of
Germany) were surface sterilized in 2% sodium hypochlor-
the in vitro antagonists to suppress bottom rot disease caused
ite (NaOCl) for 5 min, followed by three subsequent wash-
by R. solani AG1-IB on lettuce plants in vivo (growth chamber
ing steps with sterile distilled water. The seeds were
experiments) and to study the survival and root colonization
pregerminated on sterile moistened filter papers in sterile
efficiency of the antagonists by means of selective plating. The
Petri dishes for 2 days at room temperature. Pregerminated
abundance of the most promising antagonist and the response
lettuce seeds were soaked in a cell suspension (cell concen-
of the indigenous fungal and bacterial communities in the
tration adjusted to c. 109 cells mL 1) of each bacterial
rhizosphere to plant inoculation were evaluated by cultiva-
antagonist for 1 h. Four pregerminated and inoculated seeds
tion-independent molecular fingerprints. This is one of the
per pot were sown into nonsterile potting soil [turf sub-
few studies assessing not only biocontrol efficiency but also
strate/clay granulate No. 4230, Klasmann-Deilmann GmbH,
the survival of the bacterial inoculants and effects on the
Geeste, Germany, sieved (2-mm mesh width) and mixed
indigenous microbial communities.
with 80% (w/w) sand]. The pots were kept in the greenhouse
at 20 1C and 30% humidity. Each treatment was replicated
Materials and methods three times. Four weeks after sowing, the plants were care-
fully removed from the soil and vigorously shaken. The
Bacterial and fungal strains roots with adhering soil were placed in a Stomacher plastic
Fifteen out of 248 antagonists isolated from four suppressive bag and processed after adding sterile saline mechanically by
soils (Adesina et al., 2007) were selected on the basis of Stomacher blending for 3 min at high speed. CFU of the
either their (1) strong in vitro activity against R. solani AG3 inoculant per gram of root fresh weight (rfw) were deter-
(isolate Ben3; host plant, potato) or (2) activity towards mined by plating serial dilutions of the rhizosphere suspen-
both R. solani AG3 and F. oxysporum (isolate Foln3; host sions on R2A supplemented with rifampicin (75 mg mL 1)
plant, flax). All bacterial antagonists used in this study were and cycloheximide (100 mg mL 1). The CFU were counted
maintained on R2A agar (Difco, Detroit, MI). Rhizoctonia after 2 days of incubation at 28 1C.
solani AG1-IB (isolate 7/3; host plant, lettuce) and AG2
(isolate W4; host plant cabbage) were obtained from the Growth chamber experiments
strain collection of the Institute for Vegetables and Orna-
Based on the results of the rhizosphere colonization assays,
mental Crops, Großbeeren, Germany) and maintained on
eight antagonistic isolates with the ability to colonize lettuce
Waksman agar containing 5 g of proteose-peptone (Merck,
roots at cell densities of approximately log CFU 4 4 g 1 of
Darmstadt, Germany), 10 g of glucose (Merck), 3 g of meat
rfw or more were selected for the growth chamber experi-
extract (Chemex, München, Germany), 5 g of NaCl (Merck),
ments. The efficacy of eight antagonistic isolates in control-
20 g of agar (Difco), and distilled water (to 1 L) (pH 6.8).
ling bottom rot disease caused by R. solani AG1-IB on
lettuce plants (cultivar ‘Tizian’) was assessed under favor-
Generation of antibiotic-resistant mutants
able conditions for the pathogen in the growth chamber
To facilitate reisolation of the bacterial inoculants from the (York, Mannheim, Germany; 16 h/8 h day/night cycle,
rhizosphere, spontaneous rifampicin-resistant mutants of the 500 mmol m 2 s 1, 20/15 1C and 60%/80% relative humid-
bacterial antagonists were generated by plating overnight ity). In experiment 1, all eight in vitro antagonists were
cultures onto R2A agar supplemented with rifampicin evaluated. Four isolates with the best disease suppression
were included in experiments 2 and 3, while only two For each treatment and sampling time three symptomless
isolates with the best results in experiments 1 and 2 were plants were used. Loosely adhering soil was removed from
assessed in experiment 4. In experiment 3, the effect on plant the roots by shaking the plants. Five grams of roots with
growth was determined based on lettuce shoot and root dry adhering soil were suspended in 20 mL of sterile saline and
weight, in treatments without pathogen inoculation. shaken vigorously in sterile flask containing six glass beads
Seeds were treated with bacterial antagonists directly before (0.6 mm in diameter) on a rotary shaker for 1 h at 307 r.p.m.
sowing. Each antagonist was grown overnight on nutrient agar Aliquots of the rhizosphere suspension were immediately
supplemented with rifampicin at 75 mg mL 1. Overnight- processed for enumeration of the inoculated strains. Serially
grown colonies were resuspended in a sterile 0.3% NaCl diluted rhizosphere suspensions were plated on R2A med-
solution and shaken to obtain a homogeneous cell suspension. ium (Difco) supplemented with rifampicin (75 mg mL 1)
The concentration was adjusted in a spectrophotometer to a and cycloheximide (100 mg mL 1). The remaining rhizo-
density corresponding to c. 109 CFU mL 1. Lettuce seeds were sphere suspension was centrifuged at 13 000 g for 5 min.
soaked in the bacterial suspension for 1 h at room temperature. After discarding the supernatants, the cell pellets were stored
Seeds that were incubated in sterile saline for the same time frozen at 20 1C.
served as a control. Seeds were germinated in a seedling tray The ability of the most promising antagonist RU47 to
containing 92 holes filled with a nonsterile mixture of quartz colonize the vascular tissue of lettuce roots and leaves was
sand and substrate [Fruhsdorfer Einheitserde Typ P, Germany; determined at 7 weeks after sowing in experiment 3. Roots
chemical analysis (mg per 100 g): N = 75, P = 75, K = 125; from the inoculated plants were shaken to remove loosely
pH 5.9] at a 1 : 1 ratio (v/v). Lettuce plantlets were transferred adhering soils and washed in running water. One gram of
at the two- to three-leaf stage to pots filled with the same soil the washed roots or leaves were soaked in 1% sodium
mixture with one plant per pot. Drenching of lettuce plants hypochlorite for 1 min. Surface-sterilized root or leaf sam-
with bacterial antagonists was carried out 24 h after transplant- ples were rinsed four times with sterile distilled water and
ing. Each plant of the treatment inoculated with RU47 and macerated using a sterile mortal and pestle. The suspensions
R. solani (RU471Rs) was drenched with 20 mL of the bacterial of the macerated roots or leaves were serially diluted and
suspension (107 CFU mL 1) and the noninoculated healthy plated as mentioned above.
control (Ctrl) and inoculated with R. solani (Ctrl1Rs) treat-
ments received 20 mL of sterile saline. The pathogen was added DNA extraction from rhizosphere samples
on barley kernels infested with the R. solani AG1-IB (isolate 7/3;
Total community (TC)-DNA was obtained from the rhizo-
Schneider et al., 1997) 1 day after drenching the plants with
sphere pellets by means of the Bio101 extraction kit
bacterial antagonists. Five kernels per lettuce seedling were
(Q.BIOgene, Carlsbad, CA) after a harsh lysis step. The TC-
placed 1 cm deep at a distance of 2 cm from the lettuce plant.
DNA was purified using the GENECLEAN Spin kit (Q.BIO-
Noninfested barley kernels were added to the control treat-
gene). DNA yields were estimated after electrophoresis in
ments accordingly. The pots were watered daily to maintain the
1% agarose gel stained with ethidium bromide under UV
substrate moisture. Each treatment consisted of six replicates
light in comparison with the 1-kb gene-rulerTM DNA
(four pots per replicate) arranged in a randomized design. Nine
ladder (Fermentas, St Leon-Rot, Germany) run on the same
additional pots (three plants per time point and treatment)
agarose gel. Depending on the yield of the extracted TC-
were included for microbial analysis.
DNA, dilutions were made with sterile milliQ water between
Suppression of bottom rot disease by bacterial 1 : 50 and 1 : 100 (c. 1–5 ng) and were used as a template for
inoculants PCR amplification of bacterial 16S rRNA gene fragments.
Undiluted TC-DNA, or 1 : 10 dilutions (c. 20 ng DNA)
Seven weeks after sowing (4 weeks after pathogen inocula-
served as a template for PCR of fungal 18S rRNA gene
tion), the experiments were terminated and a total of 24
fragments.
plants were evaluated for suppression of bottom rot disease
of R. solani AG1-IB based on (1) the disease incidence and
PCR amplification of 16S and 18S rRNA and gacA
on (2) the effect on the shoot dry weight (SDW). The disease
gene fragments for denaturing gradient gel
incidence was assessed by recording the number of plants
electrophoresis (DGGE) analysis
with visual symptoms or the number of dead plants. The
shoot dry weight of each lettuce plant was measured. To amplify the 16S rRNA gene fragments of Pseudomonas
from TC-DNA the nested PCR system described by Milling
Root colonization by bacterial inoculants
et al. (2004) was used. Diluted amplicons obtained from the
To determine the survival and root colonization efficiency first PCR served as templates for a second PCR using the
of the inoculated antagonists, rhizosphere samples were bacterial primers F984-GC/R1378 and PCR conditions
collected at three time points: 3, 5 and 7 weeks after sowing. described by Heuer et al. (1997). To generate bacterial
c 2009 Federation of European Microbiological Societies FEMS Microbiol Ecol 69 (2009) 62–74
Published by Blackwell Publishing Ltd. All rights reserved
Biocontrol of Rhizoctonia solani and rhizosphere competence 65
R. solani AG1-IB R. solani AG2 R. solani AG3 F. oxysporum Protease Glucanase Cellulase Chitinase Siderophore 2,4-DAPG fresh root weight
Strain code, KS and KF strains isolated on Kings’B from Swedish and French soil; AFNS strain isolated on AGS from French soil; RN and RU strains isolated on R2A from the Netherlands and the United
Almost all plants died due to high disease severity by
Kingdom, respectively (Adesina et al., 2007). Inhibition zones: 1, mycelia die back and inhibition zone of 1.0–2.9 mm; 11, inhibition zone of 3.0–5.9 mm; 111, inhibition zone of 6.0 mm and more;
Log10CFU g 1 of
R. solani infection and thus the colonization density could
4.94 0.36
4.29 0.30
3.65 1.03
4.14 0.05
4.63 0.42
4.17 0.75
3.43 0.71
4.07 0.25
–, tests conducted in this study (otherwise: Adesina et al., 2007); DAPG, detection of the phlD gene by PCR-Southern blot hybridization (M.F. Adesina et al. in preparation); ND, not determined.
not be followed until the end of experiment 1 for these
isolates.
In experiment 3, the ability of isolate RU47 to colonize
the vascular tissue of lettuce leaves and roots was deter-
mined 7 weeks after sowing by dilution plating of the
ND
suspension of surface-sterilized macerated roots and leaves
1
1
1
1
from lettuce plants inoculated with RU47. We observed no
Prescreening
Table 1. In vitro characterization of eight selected antagonists and rhizosphere competence (CFU per gram fresh root weight) in prescreening greenhouse experiment
1
1
1
1
1
111
111
1
1
1
11
100
100
98
99
99
99
%
P. fluorescens
P. fluorescens
P. cannabina
P. jessenii
P. fulgida
c 2009 Federation of European Microbiological Societies FEMS Microbiol Ecol 69 (2009) 62–74
Published by Blackwell Publishing Ltd. All rights reserved
Biocontrol of Rhizoctonia solani and rhizosphere competence 67
Effects of bacterial antagonist on the growth of the pathogen. The shoot and root dry weight obtained for
lettuce plants plants inoculated was similar to the healthy Ctrl plants.
The effects of RU47, KF36, KS16 and KS74 (which showed
significant disease suppression in experiment 1) on lettuce Treatment effects on indigenous microbial
growth were evaluated in experiment 3. The effects of the communities
inoculants on plant growth were determined in treatments
inoculated with each bacterial antagonist without adding The effect of the pathogen R. solani1-IB and of the inoculant
strain P. jessenii RU47 on the relative abundance of domi-
nant bacterial, fungal and Pseudomonas populations in the
rhizosphere of lettuce plants was investigated using DGGE
Table 2. Colonization density of bacterial antagonists on lettuce ‘Tizian’ analysis of 16S rRNA, 18S rRNA or gacA gene fragments
(log10 CFU g 1 of fresh root weight) cultivated in growth chamber at amplified from TC-DNA. DGGE profiles were compared
20/15 1C 3, 5 and 7 weeks after sowing (WAS) in three independent
between three replicate rhizosphere samples from the non-
experiments (1, 2 and 3)
inoculated healthy control plants (Ctrl), plants grown with
Experiments 3 WAS 5 WAS 7 WAS R. solani (Ctrl1Rs) and plants with a combined inoculation
Experiment 1 of RU47 and R. solani (RU471Rs) at three sampling times
RU471Rs 7.17 0.04 cb 6.72 0.15 a 5.92 0.12 cb (3, 5 and 7 weeks after sowing). Although DGGE analyses
KF361Rs 7.57 0.27 c 6.55 0.22 a 6.63 0.12 c were performed for experiments 1, 2 and 3, only the data for
KS161Rs 6.39 0.37 ab 6.40 0.55 a 6.58 0.19 c
experiment 3 are presented here.
KS741Rs 6.57 0.26 ab 6.46 0.13 a 5.82 0.23 cb
KS901Rs 6.52 0.08 ab 6.81 0.66 a 5.95 0.39 cb
The bacterial DGGE profiles showed complex patterns
KS701Rs 6.20 0.53 a 6.21 0.58 a 4.75 0.39 a indicating many equally abundant bacterial populations.
AFNS311Rs 6.85 0.38 ab 6.15 0.33 a 5.53 0.42 ab Detection of strain RU47 in the bacterial rhizosphere
RN861Rs 6.68 0.02 ND ND patterns of inoculated plants was impossible due to the
Experiment 2 presence of a band with identical electrophoretic mobility as
RU471Rs 6.19 0.38 a 5.00 0.33 a 4.88 0.49 a the 16S rRNA gene fragment amplified from RU47 in the
KF361Rs 6.55 0.30 a 5.93 0.21 b 5.54 0.44 a
replicates of Ctrl and Ctrl1Rs treatments (Fig. 1a and b).
KS161Rs 6.53 0.07 a 5.48 0.35 ab 5.45 0.24 a
KS741Rs 5.86 0.47 a 4.96 0.38 a 4.92 0.33 a
Seven weeks after sowing, this band was much less intense in
Experiment 3 the replicates of RU471Rs (Fig. 1b). Three weeks after
RU471Rs 5.82 0.06 a 5.24 0.27 5.62 0.22 sowing, a clear effect of the seed inoculation could be
KF361Rs 5.92 0.06 a ND ND detected (data not shown). Cluster analysis (UPGMA of
KS161Rs 5.22 0.22 a ND ND Pearson’s correlation indices) of the bacterial DGGE profiles
KS741Rs 6.09 0.28 a ND ND showed that the bacterial patterns of rhizosphere samples
CFU of the same experiment followed by the same letter are not taken 5 weeks after sowing (2 weeks after transplanting)
significantly different according to Tukey’s test (P = 0.05). displayed a higher degree of variability between replicates,
ND, parameter not determined due to the death of nearly all plants, thus and the clustering of the treatments was less pronounced
no surviving plants for sampling. (Fig. 1a). Seven weeks after sowing, the separation of the
Table 3. Biological control effect of bacterial antagonists toward Rhizoctonia solani after seed and plant inoculation of lettuce ‘Tizian’ cultivated in
growth chamber at 20/15 1C for 7 weeks on number of dead plants (DP, from 24 plants) and shoot dry weight (SDW)
Experiment 1 Experiment 2 Experiment 3 Experiment 4
Treatments DP SDW (g per plant) SD DP SDW (g per plant) SD DP SDW (g per plant) SD DP SDW (g per plant) SD
Ctrl 0 5.2 a 0.4 0 3.9 a 0.2 0 2.8 a 0.1 0 1.6 a 0.05
Ctrl1Rs 23 0.9 bc 0.5 13 2.2 b 0.9 24 0.6 b 0.1 24 0.2 b 0.03
RU471Rs 0 4.7 a 0.2 0 3.5 ac 0.1 7 1.9 c 0.5 3 1.3 c 0.14
KF361Rs 0 4.9 a 0.6 0 3.4 ac 0.2 23 0.6 b 0.1 24 0.2 b 0.03
KS161Rs 1 4.1 a 0.5 5 2.8 bc 0.6 22 0.6 b 0.3
KS741Rs 4 3.9 a 0.7 3 3.0 ab 0.5 23 0.7 b 0.3
KS901Rs 11 2.2 c 1.2
KS701Rs 21 1.4 bc 1.0
AFNS311Rs 21 1.6 bc 0.8
RN861Rs 24 0.6 b 0.1
Dry weight followed by the same letter is not significantly different according to Tukey’s test (P = 0.05). DP, dead plant; SD, standard deviation.
Fig. 1. Comparison among DGGE fingerprints of bacterial 16S rRNA gene fragments amplified from community DNA extracts obtained from the
rhizosphere of lettuce plants without inoculation (Ctrl), with Rhizoctonia solani inoculation alone (Ctrl1Rs) and with combined inoculation of RU47
(seed and young plant inoculation) and R. solani (RU471Rs) at (a) 5 weeks after sowing and (b) 7 weeks after sowing. Beside each gel, the
corresponding UPGMA dendrogram based on Pearson’s correlation indices is given. A, B and C are the three independent replicates per treatment.
cluster formed by the fingerprints of the Ctrl1Rs replicates the antagonist 3 and 5 weeks after sowing (data not shown),
was more pronounced [ o 30% similarity to the Ctrl and whereas 7 weeks after sowing, the detection of RU47 was
RU471Rs treatments (Fig. 1b)]. The replicates of the not possible because a band with similar electrophoretic
RU471Rs and Ctrl treatments formed a joint cluster. mobility as the 16S rRNA gene fragment amplified from
Analysis of the Pseudomonas community pattern in the RU47 was detected in all treatments (data not shown). In
rhizosphere of lettuce plants showed a band with the same contrast to the bacterial community profiles, Pseudomonas
mobility as RU47 only in treatments inoculated with rhizosphere patterns of lettuce plants were less complex,
c 2009 Federation of European Microbiological Societies FEMS Microbiol Ecol 69 (2009) 62–74
Published by Blackwell Publishing Ltd. All rights reserved
Biocontrol of Rhizoctonia solani and rhizosphere competence 69
with fewer numbers of bands. Aside from a band that was among all treatments. However, the fingerprints of the
pronounced in the fingerprints of the Ctrl and RU471Rs RU471Rs treatment clearly clustered separately from the
samples taken 7 weeks after sowing, the banding patterns replicates of the Ctrl1Rs and Ctrl treatments most likely
did not markedly differ at each sampling time for all due to the presence of a strong band with the same
treatments. UPGMA analysis of the Pseudomonas DGGE electrophoretic mobility as RU47. All gacA fingerprints of
banding patterns confirmed that replicates of all treatments the samples taken 7 weeks after sowing displayed a high
shared a high similarity (4 80%) at all sampling times. similarity, although treatment-dependent clustering was
Overall, a rather low degree of variability was observed observed.
among treatments, and the diversity of this group was The fungal fingerprints for rhizosphere samples taken 5
almost not affected by inoculation with RU47 (data not weeks after sowing (Fig. 3a) revealed a high similarity of the
shown). Ctrl and RU471Rs treatments with treatment-dependent
The DGGE profiles of the Pseudomonas-specific gacA separate clusters. In contrast, the samples of the Ctrl1Rs
gene fragments displayed a higher number of bands, indi- treatments were variable and two of the three replicates of
cating a better separation of Pseudomonas populations Ctrl1Rs shared o 25% with the fingerprints of all other
(Fig. 2a and b) as compared with Pseudomonas 16S rRNA samples. This trend became even more pronounced 7 weeks
DGGE profiles. A band with the mobility of RU47 was after sowing. All replicates of Ctrl1Rs treatment formed a
detected only in the RU471Rs treatments, and this band separate cluster with o 20% similarity to the fingerprints of
was clearly separated from a dominant band that was the RU471Rs and Ctrl treatment. Several bands (Fig. 3b,
observed in the fingerprints of all treatments (Fig. 2a and bands a, b, c d) were detected in the fingerprints of these
b). Although this dominant band was less well separated treatments that were not found in the patterns of the
from the RU47 band in the gacA fingerprints of rhizosphere Ctrl1Rs treatments. Obviously, the inoculation with
communities sampled 7 weeks after sowing, the inoculant R. solani AG1-IB had a major impact on the fungal
strain was still detectable at that time point for all replicates fingerprints. In contrast, the fingerprints of the RU471Rs
of RU471Rs (Fig. 2a). Also, the gacA fingerprints of samples treatment displayed a high similarity to the noninoculated
taken 5 weeks after sowing shared a high similarity (4 60%) control, indicating no major effect in this treatment on the
Fig. 3. Comparison among DGGE fingerprints of fungal 18S rRNA gene fragments amplified from community DNA extracts obtained from the
rhizosphere of lettuce plants without inoculation (Ctrl), with Rhizoctonia solani inoculation alone (Ctrl1Rs) and with combined inoculation of RU47
(seed and young plant inoculation) and R. solani (RU471Rs) at (a) 5 weeks after sowing and (b) 7 weeks after sowing. Beside each gel, the
corresponding UPGMA dendrogram based on Pearson’s correlation indices is given. A, B and C are the three independent replicates per treatment. The
bands indicated as circled a, b, c and d are bands that are common to the replicates of the Ctrl and the RU47+Rs plants.
c 2009 Federation of European Microbiological Societies FEMS Microbiol Ecol 69 (2009) 62–74
Published by Blackwell Publishing Ltd. All rights reserved
Biocontrol of Rhizoctonia solani and rhizosphere competence 71
relative abundance of dominant fungal ribotypes. The displayed strong in vitro antagonistic activity against
DGGE profiles of fungal communities revealed that a band R. solani AG1-IB.
with electrophoretic mobility corresponding to that of the In the growth chamber experiments, the bacterial antago-
18S rRNA gene fragment amplified from R. solani was much nists were applied by seed inoculation and plant drenching.
stronger in the Ctrl1Rs treatment than in the RU471Rs Drenching of young lettuce plants with a bacterial cell
treatment (Fig. 3a and b), indicating an increased relative suspension can easily be carried out before transplanting
abundance of R. solani in the lettuce rhizosphere of the lettuce seedlings into the field. The CFU counts of the
Ctrl1Rs treatment. inoculant strains per gram of rfw determined 3 weeks after
sowing in three independent growth chamber experiments
were much higher than the counts obtained in the pre-
Discussion screening greenhouse experiment. The reasons for this
In this study, in vitro antagonists were tested in growth difference might be that the soil composition and availabil-
chamber experiments with lettuce artificially inoculated ity of nutrients were different in the soils used. The ratio of
with R. solani AG1-IB in order to assess their potential to sand to substrate of the soil used for the greenhouse
colonize lettuce roots and to suppress R. solani AG1-IB- prescreening experiment was 80 : 20 in contrast to the soil
based disease symptoms and effect on plant growth. Inter- used for the growth chamber experiments, where the ratio
estingly, only one strain (P. jessenii RU47) with weak in vitro was 50 : 50. This observation might point to an influence of
antagonistic activity against R. solani AG1-1B (Table 1) the soil on the ability of the inoculant strains to colonize the
displayed efficient and consistent disease suppression rhizosphere of lettuce. Maloney et al. (1997) found that the
throughout the four independent growth chamber experi- availability of carbon and nitrogen affected the rhizobacter-
ments. These findings confirm previous observations that in ial population. The ability of inoculant strains to colonize
vitro inhibition does not necessarily correlate with in vivo the root system after seed inoculation or root inoculation
performance (Milus & Rothrock, 1997; Schottel et al., 2001; has been pinpointed as a key factor for successful biocontrol
Faltin et al., 2004; Gravel et al., 2005). by many authors (Bloemberg & Lugtenberg, 2001; Haas &
Seven of the eight strains selected were previously as- Défago, 2005). All strains showed rather good seed and root
signed to Pseudomonas by 16S rRNA gene sequencing colonization as most inoculants established at cell densities
(Adesina et al., 2007), and for all of them proteolytic activity 4 106 g 1 rfw 3 weeks after sowing. In the case of P. fulgida
and siderophore production was detected on plates. Four KS70, which displayed strong in vitro antagonistic activity
Pseudomonas fluorescens strains (KF36, KS16, KS74 and towards a range of fungi and bacteria, less efficient coloniza-
KS90), which displayed strong in vitro antagonistic activity tion might have contributed to the failure to control the
towards R. solani AG1-IB, carried the phlD gene involved in pathogen AG1-IB in the growth chamber experiment. How-
2,4-diacetylphloroglucinol (2,4-DAPG) biosynthesis (M.F. ever, P. fluorescens KS90 and S. marcescens AFNS31 also
Adesina et al., unpublished data), suggesting that the strains failed to suppress bottom rot of R. solani, but nevertheless
might produce 2,4-DAPG. The important role of 2,4-DAPG they colonized lettuce roots at the same or similar popula-
production in efficient biocontrol was demonstrated re- tion densities as the biocontrol-efficient strain RU47. This
cently (Haas & Défago, 2005; Rezzonico et al., 2007). In implies that the inability of these isolates to suppress disease
growth chamber experiment 1 all potentially 2,4-DAPG- was not due to insufficient root colonization, but may have
producing P. fluorescens strains, except for KS90, showed resulted from insufficient production of antifungal metabo-
good disease suppression. Although the strong in vitro lites, different colonization patterns, for example the colo-
antagonist P. fluorescens KS90 carried the phlD gene and its nization of the strains at locations different from the
CFU counts per gram of rfw were not significantly different infectious site of the pathogen or insufficient cell density of
from the P. fluorescens strains KS16 and KS74, the disease the antagonists at the infectious site or differences in the
control by this strain was much less efficient. However, all plant response. However, because the colonization patterns
three other phlD-carrying P. fluorescens strains inoculated in were not investigated in this study, it cannot be excluded
experiment 3 failed to control the disease. The reasons for that differences in colonization patterns between the strains
the variability observed are not clear. Differences in the existed. Using gfp-tagged inoculant strains, several studies
expression of genes involved in antibiotic production due to reported heterogeneous colonization patterns and the ques-
lower cell densities in experiment 3 in comparison with tion of what triggers the colonization of some sites and not
experiment 1 (Table 2) in combination with high pathogen of others was raised (Gamalero et al., 2003; Götz et al.,
pressure might be discussed. Four out of the eight in vitro 2006). In contrast to studies performed under field condi-
antagonists of R. solani AG1-IB selected in our study failed tions on the control of R. solani by means of bacterial
to efficiently control the pathogen in the growth chamber antagonists (Grosch et al., 2005; Scherwinski et al., 2008),
experiment 1 (Table 3), although two of these four isolates the growth chamber experiments were performed under
well-controlled conditions and with an extremely high presence of R. solani AG1-IB and the inoculation with RU47
pathogen pressure. A variation still existed between the had almost no effect on the Pseudomonas group and that the
experiments (Tables 2 and 3). The major variable factor gacA DGGE fingerprints of all treatments, despite a treat-
was the growth of lettuce (Ctrl). In each of the independent ment-dependent clustering, had high similarity among the
growth chamber experiments, the Ctrl1Rs and RU471Rs treatments. This observation contradicts the hypothesis that
treatments were compared with the control. The variability organisms that are closely related to the BCA itself are most
in the lettuce growth might be due to different factors. The likely to be affected by the BCA due to competition for the
experiments were performed in the time period of c. 1 year same niches and resources (Winding et al., 2004; Götz et al.,
in the same chamber and thus it cannot be excluded that 2006). Moreover, the band patterns of the gacA-based
aging of the lamps resulted in less plant growth. The Pseudomonas community revealed remarkable diversity of
cultivation in the growth chamber started only after trans- the gacA gene in the rhizosphere of lettuce plant originating
planting at the two- to three-leaf stage, and slightly varying from all treatments. As found in this study, a better resolu-
conditions in the greenhouse might have also contributed to tion of gacA-based Pseudomonas community fingerprints
the decreased lettuce growth in experiment 2 and even more than of 16S-based Pseudomonas community fingerprints
pronounced in experiment 3. The reduced SDW determined was of reported by Costa et al. (2007).
paralleled a tendency toward decreased CFU per rfw 3 weeks In conclusion, out of the eight in vitro antagonists
after seed inoculation. Despite the variability in the growth investigated in this study, we found only one promising
of lettuce plants and CFU per rfw determined 3 weeks after strain: P. jessenii RU47. For this strain, in vitro proteolytic
seed inoculation, the biocontrol efficiency of strain RU47 activity and siderophore production were observed, while
was consistent. However, these factors might have contrib- chitinolytic, glucanolytic and cellulolytic activities, and
uted to the variable control by the phlD carrying P. fluor- genes encoding antibiotics such as 2,4-diacetylphloroglucinol
escens strains KF36, KS16 and KS74. (2,4-DAPG), phenazine, pyrrolnitrin and pyoluteorin were
Plant inoculation with bacterial antagonists has been not detected. In vitro production of salicylic acid as side-
reported to influence microbial community structure as rophores by several resistance-inducing bacteria under low-
determined by culture-independent methods that rely on iron conditions has been reported, and its role in the
amplification of rRNA gene fragments from rhizosphere induced-systemic resistance (ISR) elicitation process was
DNA extracts such as PCR-based DGGE (Götz et al., 2006). demonstrated in the case of Pseudomonas aeruginosa
In other studies, no or only transient changes were observed KMPCH (de Meyer et al., 1999). Hence, siderophore pro-
(Lottmann et al., 2000; Scherwinski et al., 2008). In the duction has been suggested to trigger the ISR signal pathway
present study, PCR-DGGE analysis enabled us to compare in plants (Audenaert et al., 2002; van Loon & Bakker, 2005).
the bacterial and fungal community profiles of the three Considering the strong and consistent suppression of the
different treatments and thus to evaluate the effect of the pathogen at very high disease severity by strain RU47 in vivo
pathogen R. solani AG1-IB in the presence or absence of on lettuce, which is contrary to the weak and rather
P. jessenii RU47 on the rhizosphere microbiota. We could insignificant direct inhibition of R. solani AG1-IB by this
demonstrate that the pathogen R. solani AG1-IB influenced strain in vitro and the in vitro production of siderophores by
not only the fungal but also the bacterial community strain RU47, it is assumed that induction of systemic
patterns. The effects of R. solani AG1-IB became more resistance in lettuce is the possible mechanism by which
pronounced during the experiment for both fungi and strain RU47 could protect lettuce plants against R. solani
bacteria. In the presence of RU47, this effect was less strong AG1-IB. Nonetheless, its exact mechanism still remains
and the fingerprints of the RU471Rs and Ctrl treatment unclear and its explanation needs future investigation.
formed a joint cluster for bacterial and fungal fingerprints. Strain RU47 did not promote shoot or root biomass of
However, the detection of RU47 was hampered in the lettuce compared with the noninoculated healthy plant in
bacterial community patterns due to the complexity of these the absence of the pathogen. On the basis of this observa-
patterns and the presence of ribotypes, which have similar tion, we suggest that for biological application under the
electrophoretic mobility as RU47. The complexity of the conditions where the strain was tested, it can only be used to
DGGE patterns was reduced with the use of group-specific control bottom rot disease in a soil infested with the
primers, thus enhancing the resolution for the Pseudomonas pathogen and not for a growth-promoting purpose under
group analyzed by DGGE. The gacA-based Pseudomonas noninfested conditions. However, future experiments under
DGGE profiles revealed that RU47 successfully established, field conditions and in different soil types have to show
as a band corresponding to the electrophoretic mobility to whether the RU47 conferred control effect is specific for R.
RU47 was found only in the rhizosphere patterns of lettuce solani AG1-IB and specific for lettuce plants or whether
of the RU471Rs treatment in the gacA-based Pseudomonas strain RU47 can also provide protection of lettuce plants
community patterns. Furthermore, we observed that the against other pathogens.
c 2009 Federation of European Microbiological Societies FEMS Microbiol Ecol 69 (2009) 62–74
Published by Blackwell Publishing Ltd. All rights reserved
Biocontrol of Rhizoctonia solani and rhizosphere competence 73
on the diversity of soil and rhizosphere bacteria and fungi. Schottel JL, Shimizu K & Kinkel LL (2001) Relationship of in vitro
Plant Soil 266: 23–29. pathogen inhibition and soil colonization to potato scab
Milus EA & Rothrock CS (1997) Efficacy of bacterial seed biocontrol by antagonistic Streptomyces spp. Biol Control 20:
treatments for controlling Pythium root rot of winter wheat. 102–112.
Plant Dis 81: 180–184. Smit E, Leeflang P, Glandorf B, van Elsas JD & Wernars K (1999)
Ogoshi A (1987) Ecology and pathogenicity of anastomosis and Analysis of fungal diversity in the wheat rhizosphere by
intraspecific groups of Rhizoctonia solani Kühn. Annu Rev sequencing of cloned PCR-amplified genes encoding 18S
Phytopathol 25: 125–143. rRNA and temperature gradient gel electrophoresis. Appl
Ogoshi A (1996) The genus Rhizoctonia. Rhizoctonia Species: Environ Microb 65: 2614–2621.
Taxonomy, Molecular Biology, Ecology, Pathology and Disease Sweetingham MW (1996) Integrated control of Rhizoctonia
Control (Sneh B, Jabaji-Hare S, Neate S & Dijst G, eds), species. Rhizoctonia Species: Taxonomy, Molecular Biology,
pp. 1–9. Kluwer Academic Publishers, Dordrecht, the Ecology, Pathology and Disease Control (Sneh B, Jabaji-Hare S,
Netherlands. Neate S & Dijst G, eds), pp. 549–558. Kluwer Academic
Raaijmakers JM & Weller DM (2001) Exploiting genotypic Publishers, Dordrecht, the Netherlands.
diversity of 2,4-diacetylphloroglucinol-producing Pseudomonas Vainio EJ & Hantula J (2000) Direct analysis of wood-inhabiting
spp.: characterization of superior root-colonizing P. fluorescens fungi using denaturing gradient gel electrophoresis of
strain Q8r1-96. Appl Environ Microb 67: 2545–2554. amplified ribosomal DNA. Mycol Res 104: 927–936.
Raaijmakers JM, Paulitz CT, Steinberg C, Alabouvette C & van Loon LC & Bakker PAHM (2005) Induced systemic resistance
Moenne-Loccoz Y (2008) The rhizosphere: a playground and as a mechanism of disease suppression by rhizobacteria. PGPR:
battlefield for soilborne pathogens and beneficial Biocontrol and Biofertilization (Siddiqui ZA, ed), pp. 39–66.
microorganisms. Plant Soil DOI: 10.1007/s11104-008-9568-6. Springer, Dordrecht, the Netherlands.
Rezzonico F, Zala M, Keel C, Duffy B, Moënne-Loccoz Y & Weller DM, Raaijmakers JM, McSpadden Gardener BB &
Défago G (2007) Is the ability of biocontrol fluorescent Thomashow LS (2002) Microbial population responsible for
pseudomonads to produce the antifungal metabolite specific soil suppressiveness to plant pathogens. Annu Rev
2,4-diacetylphloroglucinol really synonymous with higher Phytopathol 40: 309–348.
plant protection? New Phytol 173: 861–872. White TJ, Bruns TD, Lee S & Taylor J (1990) Analysis of
Scherwinski K, Grosch R & Berg G (2008) Effect of bacterial phylogenetic relationships by amplification and direct
antagonists on lettuce: active biocontrol of Rhizoctonia sequencing of ribosomal RNA genes. PCR Protocols: A Guide to
solani and negligible, short-term effects on nontarget Methods and Applications (Innis MA, Gelfand DH, Sninsky JJ
microorganism. FEMS Microbiol Ecol 64: 106–116. & White TJ, eds), pp. 315–322. New York Academic Press,
Schneider JHM, Schilder MT & Dijst G (1997) Characterization New York.
of Rhizoctonia solani AG-2 isolates causing bare patch in field- Winding A, Binnerup SJ & Pritchard H (2004) Non-target effects
grown tulips in the Netherlands. Eur J Plant Pathol 103: of bacterial biological control agents suppressing root
265–279. pathogenic fungi. FEMS Microbiol Ecol 47: 129–141.
c 2009 Federation of European Microbiological Societies FEMS Microbiol Ecol 69 (2009) 62–74
Published by Blackwell Publishing Ltd. All rights reserved