Effects of jasmonic acid, ethylene, and salicylic acid signaling on the rhizosphere bacterial community of Arabidopsis thaliana.

Abstract

Systemically induced resistance is a promising strategy to control plant diseases, as it affects numerous pathogens. However, since induced resistance reduces one or both growth and activity of plant pathogens, the indigenous microflora may also be affected by an enhanced defensive state of the plant. The aim of this study was to elucidate how much the bacterial rhizosphere microflora of Arabidopsis is affected by induced systemic resistance (ISR) or systemic acquired resistance (SAR). Therefore, the bacterial microflora of wild-type plants and plants affected in their defense signaling was compared. Additionally, ISR was induced by application of methyl jasmonate and SAR by treatment with salicylic acid or benzothiadiazole. As a comparative model, we also used wild type and ethylene-insensitive tobacco. Some of the Arabidopsis genotypes affected in defense signaling showed altered numbers of culturable bacteria in their rhizospheres; however, effects were dependent on soil type. Effects of plant genotype on rhizosphere bacterial community structure could not be related to plant defense because chemical activation of ISR or SAR had no significant effects on density and structure of the rhizosphere bacterial community. These findings support the notion that control of plant diseases by elicitation of systemic resistance will not significantly affect the resident soil bacterial microflora.

Full text

Page 1

Vol. 24, No. 4, 2011 / 395
MPMI Vol. 24, No. 4, 2011, pp. 395–407. doi:10.1094/MPMI-05-10-0115. © 2011 The American Phytopathological Society
Effects of Jasmonic Acid, Ethylene,
and Salicylic Acid Signaling
on the Rhizosphere Bacterial Community
of Arabidopsis thaliana
Rogier F. Doornbos,1 Bart P. J. Geraats,2 Eiko E. Kuramae,3 L. C. Van Loon,1 and Peter A. H. M. Bakker1
1Plant-Microbe Interactions, Institute of Environmental Biology, Utrecht University, Padualaan 8, 3584 CH Utrecht, The
Netherlands; 2Seed Technology, Nunhems Netherlands B.V., Voort 6, 6083 AC, Nunhem, The Netherlands; 3Microbial Ecology,
Netherlands Institute of Ecology, Droevendaalsesteeg 10, 6708 PB Wageningen, The Netherlands
Submitted 19 May 2010. Accepted 9 December 2010.
Systemically induced resistance is a promising strategy to
control plant diseases, as it affects numerous pathogens.
However, since induced resistance reduces one or both
growth and activity of plant pathogens, the indigenous mi-
croflora may also be affected by an enhanced defensive
state of the plant. The aim of this study was to elucidate
how much the bacterial rhizosphere microflora of Arabi-
dopsis is affected by induced systemic resistance (ISR) or
systemic acquired resistance (SAR). Therefore, the bacterial
microflora of wild-type plants and plants affected in their
defense signaling was compared. Additionally, ISR was
induced by application of methyl jasmonate and SAR by
treatment with salicylic acid or benzothiadiazole. As a com-
parative model, we also used wild type and ethylene-insen-
sitive tobacco. Some of the Arabidopsis genotypes affected
in defense signaling showed altered numbers of culturable
bacteria in their rhizospheres; however, effects were depend-
ent on soil type. Effects of plant genotype on rhizosphere
bacterial community structure could not be related to
plant defense because chemical activation of ISR or SAR
had no significant effects on density and structure of the
rhizosphere bacterial community. These findings support
the notion that control of plant diseases by elicitation of
systemic resistance will not significantly affect the resident
soil bacterial microflora.
The bacterial rhizosphere microflora plays an important role
in plant health. A well-studied phenomenon is the suppression
of soilborne plant diseases by plant root–inhabiting bacteria,
including members of genera Bacillus, Erwinia, Pseudomonas,
Rhizobium, Serratia, and Xanthomonas (Weller 1988; Whipps
2001). Mechanisms of suppression of plant diseases by such
bacteria include competition for substrates, competition for
iron by siderophores, antibiosis, lytic activity, and induced sys-
temic resistance (ISR) (Van Loon and Bakker 2003). Induced
resistance is the state of enhanced defensive capacity devel-
oped by plants when appropriately stimulated (Van Loon et al.
1998; Van Wees et al. 2008; Zehnder et al. 2001). Rhizobacte-
ria-mediated ISR is effective against a wide range of patho-
gens on dicotyledonous plant species, including Arabidopsis,
bean, carnation, eucalyptus, radish, tobacco, and tomato (Bakker
et al. 2007), as well as the monocot rice (De Vleesschauwer et
al. 2008). ISR resembles pathogen-induced systemic acquired
resistance (SAR) in that i) upon challenge inoculation, induced
plants show an enhanced defensive capacity, enabling the plant
to respond faster, more effectively, or both to microbial attack-
ers (Conrath et al. 2002, 2006; Van Wees et al. 2008) and ii)
both are dependent on a functional NPR1 gene (Pieterse and
Van Loon 2004). However, whereas SAR is dependent upon
the plant hormone salicylic acid (SA) and is associated with
the expression of pathogenesis-related (PR) proteins, rhizobac-
teria-mediated ISR in Arabidopsis does not require SA signal-
ing, nor is it associated with the expression of known defense-
related genes (Pieterse et al. 1996, 1998; Van der Ent et al.
2008; Van Wees et al. 1999; Verhagen et al. 2004). Instead,
ISR requires responsiveness to jasmonic acid (JA) and ethyl-
ene (ET); yet, it is not associated with endogenous increases of
these hormones (Pieterse et al. 2000).
In general, pathogens with a necrotrophic lifestyle are re-
sisted by JA- and ET-dependent defenses, whereas SA-depend-
ent defenses are effective against pathogens with a biotrophic
lifestyle (Glazebrook 2005; Thomma et al. 1998). This differ-
ential effectiveness of plant defenses is also displayed by ISR
and SAR (Ton et al. 2002). For example, ISR is effective
against the necrotrophic fungus Alternaria brassicicola,
whereas SAR is not; SAR is effective against the biotrophic
Turnip crinkle virus, whereas ISR is not. However, almost
nothing is known about effects of the augmented defensive
state on the indigenous rhizosphere microflora. A recent study
by Micallef and associates (2009) assessed the rhizobacterial
community structure of eight Arabidopsis accessions. Of
these, the non-ISR–expressing accessions RLD and WS-0
showed a bacterial community structure that tended to differ
from that of the other six, which are ISR-inducible. Such dif-
ferences might be related to differences in plant defensive ca-
pacity.
The aim of this study was to investigate whether the bacte-
rial rhizosphere microflora is affected by one or both the JA/ET-
or SA-dependent defense-signaling pathways. Two experimental
approaches were used. First, we analyzed the bacterial rhizos-
phere microflora of Arabidopsis thaliana accession Col-0 and
derivatives affected in specific defense signal-transduction
pathways. Second, we activated the JA- and SA-signaling
pathways by exogenous application of the hormones, to study
the effect of activated defenses on the bacterial abundance and
community structure.
Corresponding author: P. A. H. M Bakker; E-mail: p.a.h.m.bakker@uu.nl;
Telephone: +31 (0)30-2536861; Fax +31 (0)30-2518366.

Page 2

396 / Molecular Plant-Microbe Interactions
In addition, we used tobacco and its ET-insensitive transfor-
mant, Tetr18 (Knoester et al. 1998; Wilkinson et al. 1997).
Tetr18 plants have reduced resistance against several soilborne
fungi and oomycetes, including species belonging to genera
Fusarium, Thielaviopsis, and Pythium (Geraats et al. 2003;
Knoester et al. 1998). Geraats (2003) suggested that Tetr18
plants differ from nontransformed plants in root characteristics
that influence bacterial community structure in the rhizos-
phere. The Arabidopsis JA-response mutant jar1 and ET-sig-
naling mutant ein2 also have an enhanced susceptibility to dif-
ferent soilborne Pythium isolates (Geraats et al. 2002). More-
over, Arabidopsis genotypes affected in SA signaling display
an enhanced sensitivity to the necrotrophic soilborne oomycete
Phytophthora parasitica, the leaf spot fungus Cercospora nico-
tianae, and the hemibiotrophic bacterial leaf pathogen Pseudo-
monas syringae pv. tomato (Cao et al. 1994; Delaney et al.
1994). Conversely, the constitutive expressor of SA-dependent
defenses cpr1 has increased resistance to the downey mildew
oomycete Hyaloperonospora arabidopsidis as well as to P.
syringae pv. maculicola (Bowling et al. 1994).
In order to obtain representative results, plants were grown
on different soil types: i) a mixture of commercially available
potting soil and sand and ii) a clay soil. Both soils were used
as such or were autoclaved to allow possible species-specific
recolonization from surviving bacteria or the ambient environ-
ment. The rhizosphere microflora, with focus on the total bac-
terial community and Pseudomonas spp., was explored using
two complementary techniques, i.e., culturable-dependent
semiselective plating was used for bacterial quantification and
denaturing gradient–gel electrophoresis (DGGE), a culturable
independent-fingerprinting technique, was used to monitor
possible shifts in microbial community structure.
RESULTS
Abundance of rhizosphere populations
of culturable bacteria and Pseudomonas spp.
The numbers of culturable bacteria and Pseudomonas spp.
in the rhizospheres of Arabidopsis and tobacco plants were
quantified by selective plating. To study a possible role of
defense signaling, mutants affected in either the JA/ET or SA
response were used. Power to detect differences between treat-
ments was 100% for the total bacterial populations and 97%
for numbers of pseudomonads in the nonautoclaved potting
soil, as determined by power analysis. Population densities of
total culturable bacteria in the rhizospheres of the different
Arabidopsis genotypes grown on distinct soils ranged from 2 ×
107 to 1 × 109 colony-forming units (CFU) per gram of rhizos-
phere soil (Fig. 1). Significant differences in bacterial numbers
were only observed for plants grown in untreated potting soil-
sand mixture (Fig. 1A). The JA-response mutant jar1, the ET-
response mutant ein2, and the constitutive SA-producing cpr1
showed significantly lower numbers of culturable bacteria com-
pared with the Col-0 wild type. These differences were not
observed when the potting soil and sand mixture was auto-
claved before use or when plants were grown on nonauto-
claved or autoclaved clay soil (Fig. 1B through D). Numbers
of CFU of Pseudomonas spp. in the rhizosphere were between
5 × 105 and 5 × 107 per gram of root and demonstrated
tendencies similar to total bacterial numbers, except for ein2
(Fig. 2). However, Pseudomonas populations seemed more
sensitive to SA-dependent defenses, as illustrated by a de-
creased abundance in cpr1 and a tendency of increased bac-
terial numbers in the NahG rhizospheres (Fig. 2A and B).
Although differences in bacterial abundance between the geno-
Fig. 1. Number of total culturable bacteria in the rhizosphere (log CFU per gram of root) of Arabidopsis Col-0 and Col-0 derivatives affected in jasmonic
acid- (jar1 and npr1), ethylene- (etr1, ein2 and npr1), or salicylic acid– (cpr1, npr1 and NahG) dependent defense responses. Rhizosphere samples of five
individual plants were dilution plated on 1/10 TSA+ (3 g of tryptic soy broth per liter, 13 g of agar technical per liter, and 100 mg of natamycin per liter). A,
Plants grown on potting soil-sand mixture. B, Plants grown on autoclaved potting soil-sand mixture. C, Plants grown on clay soil. D, Plants grown on
autoclaved clay soil. Different letters indicate significant differences (analysis of variance and Tukey post-hoc test, P < 0.05); error bars represent standard
errors.

Page 3

Vol. 24, No. 4, 2011 / 397
types were not observed when plants were grown on the clay
soil, Pseudomonas population densities were also consistently
lower in the cpr1 rhizosphere (Fig. 2C and D).
The transgenic Tetr18 tobacco contains the mutant etr1 ET
receptor of Arabidopsis, resulting in ET insensitivity. When
grown in either a potting soil-sand mixture or clay soil,
autoclaved or not, no differences were observed in rhizosphere
population densities of total culturable bacteria or Pseudomo-
nas spp. (data not shown). These results are largely compara-
ble to those of the Arabidopsis etr1 and ein2 genotypes.
Composition of total bacterial
and Pseudomonas populations in the rhizosphere
of Arabidopsis defense signaling mutants.
Possible differences in the bacterial rhizosphere community
of the different Arabidopsis genotypes were studied in a cul-
turable-independent manner, using polymerase chain reaction
(PCR)-DGGE with eubacterial- and Pseudomonas spp.–specific
primers. DGGE analysis revealed complex banding patterns
containing 22 to 30 and nine to 18 distinct bands for plants
grown on potting soil and clay soil, respectively. We used re-
dundancy analysis (RDA) to evaluate effects of plant genotype
on the rhizosphere microflora. Ordination plots of the eubacte-
rial data are shown in Figure 3, and those for the Pseudomonas
data are shown in Figure 4.
Analysis of similarity (ANOSIM) showed that the bacterial
rhizosphere microflora was not significantly affected by plant
genotype (Table 1). However, in all soils, the eubacterial
rhizosphere microflora from mutants jar1, etr1, and ein2 con-
sistently clustered away from the Col-0 wild type. Similar re-
sults were obtained for the Pseudomonas spp. community.
Independent of the soil used, the genotypes jar1, etr1, and ein2
clustered separately from the wild type, whereas the other
genotypes showed more variable patterns in the different soils.
Bacterial and Pseudomonas community structure
in the rhizosphere of ET-insensitive tobacco.
The eubacterial and Pseudomonas rhizosphere communities
of wild type and Tetr18 tobacco were not significantly differ-
ent, as assessed by ANOSIM analysis of PCR-DGGE finger-
prints (Table 2). Plants grown in nonautoclaved soil showed
distinct clusters of the eubacterial microflora between the wild
type and the Tetr18 transformant. This effect was less pro-
nounced for the rhizosphere community of plants grown in
autoclaved soil (Fig. 5). The rhizosphere Pseudomonas com-
munity only revealed separate clusters between wild type and
Tetr18 tobacco for plants grown on a nonautoclaved potting
soil and sand mixture (Fig. 6).
Effects of foliar application of methyl jasmonate (MeJA),
SA, and benzothiadiazole (BTH)
on local and systemic VSP2 and PR-1 expression.
In the experiments with the defense-signaling mutants, no
consistent differential effects of plant genotypes on rhizosphere
bacterial communities were observed. Given that defense signal-
ing was not elicited in these experiments, such an experimental
outcome may not be surprising. To investigate if activetion of
JA- and SA-signaling pathways affects the composition of the
rhizosphere microflora, MeJA, SA, or BTH were applied to the
leaves. Application of MeJA activated the JA-dependent defense
response both locally (leaves) and systemically (roots), indicated
by an about 7.5-fold increase in the expression of the JA-respon-
sive marker gene VSP2 (Fig. 7A). Although SA and BTH appli-
cation did not induce expression of VSP2 in the leaves, the ex-
Fig. 2. Number of total culturable Pseudomonas spp. in the rhizosphere (log CFU per gram of root) of Arabidopsis Col-0 and Col-0 derivatives affected in
jasmonic acid- (jar1 and npr1), ethylene- (etr1, ein2 and npr1), or salicylic acid–dependent (cpr1, npr1 and NahG) defense responses. Rhizosphere samples
of five individual plants were dilution plated on KB+. A, Plants grown on potting soi-sand mixture. B, Plants grown on autoclaved potting soil-sand mixture.
C, Plants grown on clay soil. D, Plants grown on autoclaved clay soil. Different letters indicate significant differences (analysis of variance and Tukey post-
hoc test, P < 0.05); error bars represent standard errors.

Page 4

398 / Molecular Plant-Microbe Interactions
pression of VSP2 in the roots was increased 2.8- and 1.6-fold.
Local expression of the SAR marker gene PR-1 was increased
49-fold in response to SA and 73-fold in response to BTH treat-
ment (Fig. 7B). Foliar application of SA or BTH also resulted in
activation of the corresponding defense-signaling pathway in the
roots, as indicated by a nine- and fourfold induction of PR-1
expression, respecttively. As expected, application of MeJA did
not affect PR-1 expression.
Effect of activated defense signal-transduction pathways
on bacterial abundance.
The abundance of the bacterial and Pseudomonas spp. micro-
flora in the rhizospheres of mock-, MeJA-, SA-, and BTH-
treated plants was quantified by selective plating two weeks
after treatment (Fig. 8). The numbers of CFU of total culturable
bacteria in the rhizosphere of Col-0 revealed no significant influ-
ences of the hormone treatments. Likewise, activation of plant
defenses had no effect on the numbers of total culturable bacte-
ria in the rhizospheres of the mutants npr1 and jar1 nor in trans-
genic NahG plants. Colony counts of Pseudomonas spp. showed
comparable results. However, compared with the mock-treated
control, significantly higher numbers of Pseudomonas spp. were
measured in NahG plants treated with SA or BTH.
Effect of activated plant defense signaling
on bacterial community structure.
Ordination plots derived from DGGE fingerprints of the
bacterial rhizosphere communities were generated by RDA
and revealed that all samples clustered predominantly by plant
genotype (Fig. 9). Induction of defense signaling in the differ-
ent genotypes did not affect rhizobacterial communities, as the
variables mock, MeJA, SA, and BTH are all in the center of
the ordination plot. Pseudomonas-specific DGGE analysis re-
vealed a higher variability among the samples, resulting in less
Fig. 3. Ordination biplots generated by redundancy analysis of eubacterial denaturing gradient–gel electrophoresis fingerprints of the rhizosphere of
Arabidopsis Col-0 and Col-0 derivatives affected in jasmonic acid- (jar1 and npr1), ethylene- (etr1, ein2 and npr1), or salicylic acid–defense (cpr1, npr1 and
NahG) signal-transduction pathways. A, Plants grown on potting soil-sand mixture. B, Plants grown on autoclaved potting soil-sand mixture. C, Plants
grown on clay soil. D, Plants grown on autoclaved clay soil. Crosses represent individual samples, triangles the centroid position of each genotype.

Page 5

Vol. 24, No. 4, 2011 / 399
distinct clusters compared with the eubacterial analysis. How-
ever, RDA analyses indicated that none of the hormone treat-
ments had a relevant influence on the Pseudomonas commu-
nity structure. Indeed, two-way ANOSIM analysis showed that
eubacterial communities were significantly different with
some overlapping (R = 0.41924) in the rhizospheres of differ-
ent Arabidopsis genotypes, while there were no significant
effects of the hormone treatments (Table 3). For the Pseudo-
monas community, ANOSIM revealed significant differences
in the rhizospheres of different Arabidopsis genotypes, but
also, activation of defense signaling by applying hormones had
no significant impact (Table 3) for this microbial group.
DISCUSSION
Rhizobacteria-mediated ISR and pathogen-induced SAR can
reduce disease severity and microbial proliferation of a wide
range of plant pathogens (Durrant and Dong 2004; Loake and
Grant 2007; Van Loon 2007; Van Loon et al. 1998; Van Wees et
al. 2008). However, it is not known if augmented plant-defense
responses affect indigenous populations of plant-associated
microorganisms. To elucidate if the bacterial rhizosphere
microflora of Arabidopsis is affected by one or more JA-, ET-,
or SA-dependent defenses, we investigated the abundance and
community structure of the total bacterial microflora as well as
the prevalent Pseudomonas spp. in the rhizosphere of different
Arabidopsis and tobacco genotypes with altered defense
signaling characteristics.
Effect of plant defense signaling on bacterial abundances.
Total bacterial and Pseudomonas spp. population densities
in the rhizospheres of Arabidopsis and tobacco and of deriva-
tives of these plants with altered defense signaling properties
were studied by dilution plating on selective media. Bacterial
Fig. 4. Ordination biplots generated by redundancy analysis of Pseudomonas spp.-specific denaturing gradient–gel electrophoresis fingerprints of the
rhizosphere Arabidopsis Col-0 and Col-0 derivatives affected in jasmonic acid- (jar1 and npr1), ethylene- (etr1, ein2 and npr1), or salicylic acid–defense
(cpr1, npr1 and NahG) signal-transduction pathways. A, Plants grown on potting soil-sand mixture. B, Plants grown on autoclaved potting soil-sand mixture.
C, Plants grown on clay soil. D, Plants grown on autoclaved clay soil. Crosses represent individual samples, triangles the centroid position of each genotype.