The ISME Journal (2014) 8, 790–803

& 2014 International Society for Microbial Ecology All rights reserved 1751-7362/14
www.nature.com/ismej

ORIGINAL ARTICLE
Rhizosphere microbiome assemblage is affected by
plant development
Jacqueline M Chaparro, Dayakar V Badri and Jorge M Vivanco
Center for Rhizosphere Biology, Department of Horticulture and Landscape Architecture, Colorado State
University, Fort Collins, CO, USA

There is a concerted understanding of the ability of root exudates to influence the structure of
rhizosphere microbial communities. However, our knowledge of the connection between plant
development, root exudation and microbiome assemblage is limited. Here, we analyzed the structure
of the rhizospheric bacterial community associated with Arabidopsis at four time points
corresponding to distinct stages of plant development: seedling, vegetative, bolting and flowering.
Overall, there were no significant differences in bacterial community structure, but we observed that
the microbial community at the seedling stage was distinct from the other developmental time
points. At a closer level, phylum such as Acidobacteria, Actinobacteria, Bacteroidetes, Cyanobac-
teria and specific genera within those phyla followed distinct patterns associated with plant
development and root exudation. These results suggested that the plant can select a subset of
microbes at different stages of development, presumably for specific functions. Accordingly,
metatranscriptomics analysis of the rhizosphere microbiome revealed that 81 unique transcripts
were significantly (Po0.05) expressed at different stages of plant development. For instance, genes
involved in streptomycin synthesis were significantly induced at bolting and flowering stages,
presumably for disease suppression. We surmise that plants secrete blends of compounds and
specific phytochemicals in the root exudates that are differentially produced at distinct stages of
development to help orchestrate rhizosphere microbiome assemblage.
The ISME Journal (2014) 8, 790–803; doi:10.1038/ismej.2013.196; published online 7 November 2013
Subject Category: Microbial population and community ecology
Keywords: Arabidopsis thaliana; development; microbiome; rhizosphere; root exudation

Introduction root secretion of defense-related proteins is enhanced

during flowering time (De-la-Pena et al., 2010),
Plants such as bean, maize, soybean, cowpea, suggesting an involvement of plant roots in ARR.
cabbage, cotton and Arabidopsis exhibit age-related Similarly, Chaparro et al. (2013) have shown that
resistance (ARR) (Develey-Riviere and Galiana, Arabidopsis roots release more phenolic-related com-
2007). For example, susceptibility to Puccinia sorghi pounds at later stages of life which might be correlated
(common rust) in maize is manifested in younger to defense strategies against pathogens as secondary
plants but as the plants mature, the level of disease
metabolites are involved in plant immunity against
resistance augments (Abedon and Tracy, 1996). In
bacterial and fungal pathogens (Rogers et al., 1996;
Arabidopsis, transitions from the vegetative to the
Clay et al., 2009; Millet et al., 2010; An and Mou,
floral phase correlates with resistance to Pseudomonas
2011; Bednarek, 2012). Thus, there is a need to
syringae (Kus et al., 2002; Rusterucci et al., 2005).
understand the influence of plant development on
While ARR at the molecular level has been studied
microbial associations that might occur naturally in
with respect to leaf pathogens, little discussion has
focused on root defense strategies and their role in the rhizosphere related to defense but also to other
ARR. However, there are some indirect correlations; vital plant necessities such as nutrient acquisition.
for example, cotton (Zaki et al., 1998) or bean (Nicoli Under natural conditions, plants tend to require
et al., 2012) plants are more susceptible to root disease higher quantities of N at later stages of development
(Rizoctonia solani or Fusarium root rot, respectively) (Kelly et al., 1995; Rossato et al., 2001; Nazoa et al.,
at the seedling stage. Recent studies have shown that 2003; Malagoli et al., 2004) but exactly how this N is
obtained under natural conditions is unknown. The
classically studied symbiotic relationship between
Correspondence: JM Vivanco, Center for Rhizosphere Biology, rhizobia and legumes has shown that symbiosis
Department of Horticulture and Landscape Architecture, Colorado occurs only when the plant is under N-limiting
State University, Fort Collins, CO, USA.
E-mail: [email protected]
conditions (Davidson and Robson, 1986; Eaglesham,
Received 6 March 2013; revised 22 August 2013; accepted 29 1989; Zahran, 1999). Similarly, the secretion of
September 2013; published online 7 November 2013 flavones and flavonols that initiates rhizobia-legume
 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
791
symbiosis is enhanced under N-limiting conditions decreased through plant growth. In contrast, the
(Coronado et al., 1995; Zhang et al., 2009). This cumulative secretion levels of amino acids and
combined information suggests that the plant may phenolics increased over time. Accordingly, it was
have some control over the identity and function- hypothesized that seedlings of roots release sugars
ality of the rhizosphere microbiome. as substrates for a wide diversity of microbes at early
Studies have shown that rhizospheric fungal and stages of development but as the plant ages it
bacterial communities of a wide range of plants (i.e., releases specific substrates and potentially antimi-
Arabidopsis, Medicago, maize, pea, wheat and sugar crobial compounds in an effort to select for
beet) change according to a plant developmental particular microbial inhabitants of the rhizosphere
gradient (Baudoin et al., 2002; Mougel et al., 2006; (Badri et al., 2013a; Chaparro et al., 2013). This
Houlden et al., 2008; Micallef et al., 2009a). In these potential selection of microbes in the rhizosphere as
studies the microbial communities were assessed the plant ages might be associated with the ability of
through automated ribosomal intergenic spacer beneficial microbes to suppress pathogenic ones
analysis or denaturing gradient gel electrophoresis (Mendes et al., 2011), trigger induced systemic
techniques that produce a fingerprint of the com- tolerance to overcome abiotic stress (Selvakumar
munity structure but not of its members’ identity. et al., 2012), increase the plant’s innate immunity
While these studies demonstrated that plant micro- (Zamioudis and Pieterse, 2012), help in mineral
bial communities change in response to plant nutrition (Bolan, 1991; van der Heijden et al., 2008)
development they were not able to distinguish and in overall plant health (Berendsen et al., 2012;
how or which microbes contribute to the changes Chaparro et al., 2012). Here, we tested this hypothesis
observed. For example, Micallef et al. (2009a) by analyzing the rhizosphere microbial composition
through denaturing gradient gel electrophoresis of Arabidopsis by 454 pyrosequencing at four
analysis observed that Arabidopsis rhizosphere distinct physiological stages of development: seed-
microbial communities varied with plant develop- ling (four–five leaf stage), vegetative (rosette), bolt-
ment and that microbial communities in early plant ing and flowering. We did not include samples past
development were more distinct to the bulk soil and the flowering stage because previous studies have
that this difference decreased with plant age. determined that rhizosphere microbial communities
Similarly, an assessment of the potato rhizosphere converge past the flowering stage (Micallef et al.,
demonstrated that young potato plants showed 2009a; Lundberg et al., 2012). Further, a metatran-
cultivar-dependent rhizosphere microbial commu- scriptomics analysis of the rhizomicrobiome was
nities but these differences in the microbiomes also conducted to ascertain a relationship between
disappeared as the plants aged (Inceoglu et al., plant growth and microbiome functioning.
2011). Soybean rhizosphere microbial communities
were also influenced by plant development; early
reproductive growth stages of the soybean plant Experimental procedures
produced more complex microbial communities
than late stage soybean plants (Xu et al., 2009). An Soil experiment
assessment of the microbial community structure Soil where Arabidopsis thaliana genotype (Pna-10)
through plant development focusing on the mem- (Li et al., 2010) has naturally grown for more than 8
bers that make up the community is warranted. years was collected in July 2011 from the Michigan
Incidentally, the recent characterization of the Extension Station, Benton Harbor, MI (421 050 3400 N,
Arabidopsis thaliana core microbiome provides a 861 210 1900 W, elevation 630 feet). The soil is
tool to decipher the influence of the plant on the described in detail in Broeckling et al. (2008); soil
rhizosphere microbiome at different stages of devel- from the same site was used in previous studies
opment (Bulgarelli et al., 2012; Lundberg et al., 2012). (Broeckling et al., 2008; Badri et al., 2009, 2013a).
Evidence demonstrating the close ties root exu- Arabidopsis thaliana seeds were surface-sterilized
dates have on the microbial composition of the with Clorox for one minute and subsequently rinsed
rhizosphere is mounting (Broeckling et al., 2008; four times with sterile distilled water. Sterile seeds
Badri et al., 2009, 2013a; Micallef et al., 2009b; were placed on Murashige and Skoog (MS) media
Chaparro et al., 2012, 2013), whereby many chemi- (Murashige and Skoog, 1962) supplemented with
cals present in root exudates act as substrates, 0.9% bactoagar and 3% sucrose Petri plates. Seeds
chemotactic or signaling molecules to orchestrate were incubated in a growth chamber with photo-
changes in microbial composition (Shaw, 1991; period 16 h light/8 h night at 25 1C for seven days.
de Weert et al., 2002; Jain and Nainawatee, 2002; The Arabidopsis seven-day-old seedlings did not
Horiuchi et al., 2005; Bais et al., 2006; Badri and introduce any microbes to the system as they were
Vivanco, 2009; Neal et al., 2012; Badri et al., 2013a). surface sterilized with bleach and no bacterial
Recently, it was reported that the composition of growth was observed on the MS agar plates even
Arabidopsis root exudates change following a plant after 7 days of growth. Six replicate pots were
developmental gradient (Chaparro et al., 2013). maintained for each of the four developmental
Cumulative secretion levels of sugars and sugar time points, and one 7-day-old seedling was
alcohols were higher in early time points and transplanted to each pot. Individual plants were

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
792
grown until the following stages were reached: products (3 replicates per time point) were pooled
seedling (17 days), vegetative (24 days), bolting and subjected to unidirectional pyrosequencing in
(31 days) or flowering (38 days); see Chaparro 1/8 of a pico titer-plate at the W.M. Keck Center for
et al. (2013). Comparative and Functional Genomics, Roy J.
Carver Biotechnology Center, University of Illinois
at Urbana-Champaign on a Roche/454 Genome
Extraction of microbial DNA from soil Sequencer GS-FLX þ instrument (Roche, Branford,
In our study we used the classical definition of the CT, USA). Similarly, isolated total RNA (15 mg per
rhizosphere which consist of three zones: the sample) was sent to the same facility. Briefly, mRNA
endorhizosphere (root tissue area), the rhizoplane isolation using the Ribozero rRNA removal Meta-
(root surface with epidermis) and the ectorhizo- bacterial kit (Epicentre Biotechnologies, Madison,
sphere (soil directly surrounding the root), and we WI, USA) was performed; using individually bar-
did not distinguish between these zones (Lynch, coded random hexamer primers the isolated mRNA
1990; Morgan et al., 2005; Brimecombe et al., 2007; was converted to cDNA. cDNA library normalization
Badri and Vivanco, 2009). Rhizosphere soils (or ‘soil’ was performed using the Trimmer Direct kit
thereafter) for each of the time point’s six biological (Evrogen, Moscow, Russia), the samples were
replicates (24 samples; 4 time points) were collected pooled in equimolar concentrations and pyrose-
by gently removing the plants from the pots quencing was performed on 1/8 of a pico titer-plate
and obtaining the soil attached to the roots (see Chaparro et al., 2013).
(Supplementary Figure S1) and stored at  80 1C
for future use. It is worth noting that the removal of
the rhizosphere soil was done in such a manner to 16S rRNA sequencing analysis
prevent mechanical root shearing. However, our Sequence reads were processed using Mothur v.
rhizosphere soil (as per the classical definition) 1.25.1 (Schloss et al., 2009) as previously described
consists of the rhizosphere and the roots present in (Badri et al., 2013a). Sampling effort was equalized
that soil (Supplementary Figure S1). Once the total to the depth of the smallest sample (2769 reads) and
RNA was extracted from the soil using the PowerSoil operational taxonomic units (OTUs) were defined at
total RNA isolation kit (Mo Bio, Carlsbad, CA, USA) X97% sequence identity, using the average neigh-
(see Chaparro et al., 2013), total DNA was subse- bor algorithm. Reads were classified within Mothur
quently extracted using the RNA PowerSoil DNA using the naı̈ve Baysian classifier (Wang et al., 2007).
Elution Accessory Kit (Mo Bio) according to the Final taxonomic assignment was based on the
manufacturer’s instructions. The isolated DNA was consensus identification for each OTU (see
quantified using a Nanodrop ND-1000 spectrophot- Supplementary Table S1). Sequences were also
ometer (Thermo, Waltham, MA, USA). The bacterial assigned to phylotypes using the phylotype com-
hypervariable regions V1-V3 of the 16S rRNA gene mand in Mothur. A multivariate data analysis of the
were PCR-amplified using individually bar-coded OTUs was performed using METAGENassist (Arndt
forward primers 27F, 50 -AGAGTTTGATYMTGGCT et al., 2012), followed by normalization based on
CAG-30 and reverse primer 533R, 50 - TTACCGCGGCT interquantile range (IQR) (Hackstadt and Hess, 2009)
GCTGGC -30 . PCR was performed using Taq DNA and log2-transformation. IQR normalization allows
polymerase (Takara, Mountain View, CA, USA) as one to increase statistical power by removing
previously described (Badri et al., 2013a). Briefly, sequences that do not fall within the middle 50%
DNA samples were diluted to a concentration of 5 ng of observations and thus reducing the number of
ml  1 and one microliter was used per PCR reaction. statistical tests one has to perform. Principal
The reaction mix (20 ml) contained 0.4 mmoles of component analysis (PCA) and significant features
each gene-specific primer, 200 mmoles of dNTPs, were identified for all treatments using METAGEN-
1  reaction buffer and one unit of Taq DNA assist (Arndt et al., 2012). The Vegan package
polymerase (Takara). PCR included 39 cycles of (Oksanen et al., 2012) for R was used for community
94 1C for 30 s, 56 1C for 30 s and 72 1C for 1 minute in dissimilarity calculations (Bray–Curtis index) and
a thermal cycler (GeneAmp PCR system 2700; principal coordinate analysis (PCoA).
Applied Biosystems, Grand Island, NY, USA). After
PCR amplification of the 24 soil DNA samples
(6 reps per time point), repetitions were pooled in Metatranscriptomics analysis
groups leaving three biological replicates per time Sequence reads were processed using the Metage-
point (12 samples). Amplicon products were gel nomics-RAST (MG-RAST) server (Meyer et al., 2008)
purified using Wizard SV gel (Promega, Madison, and Mothur (Schloss et al., 2009) (see Chaparro et al.,
WI, USA) and PCR clean-up system followed by 2013). Briefly, 454 pyrosequencing of the isolated
Agencourt AMPure XP purification kit (Beckman mRNA produced 166 250 sequence reads for seed-
Coulter, Brea, CA, USA). The concentration of DNA ling, vegetative, bolting, flowering and bulk soil
in each sample was determined using the Nanodrop samples (see Chaparro et al., 2013). Host-specific
ND-1000 spectrophotometer (Thermo). Approxi- species sequences (Arabidopsis thaliana) were
mately equal amounts of the 12 purified amplicon removed using the DNA level matching Bowtie

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
793
algorithm (Langmead et al., 2009) within MG-RAST. subsequently standardized early (seedling and vege-
Artificially replicated sequences were removed tative) and late (bolting and flowering) metatran-
(Gomez-Alvarez et al., 2009). Sequences were also scriptomics data (see Chaparro et al., 2013) were
removed if their length differed by more than two done using a two-sample t-test. Additionally, Pear-
standard deviations from the mean length. Addi- son correlation analysis was performed with the log
tionally, sequences identified via MG-RAST as transformed root exudate compounds and the
ribosomal RNAs were removed using Mothur. Once transformed and standardized functional genes that
the ribosomal RNAs were removed (6473 rRNA significantly changed (81 transcripts; 413 reads)
sequences total) from each of the samples they were with plant growth to determine if developmen-
normalized to 14 740 high-quality sequence reads tally-dependent root exudation mediates the func-
per time point (Chaparro et al., 2013). Sequence tions carried out by the rhizosphere microbiome.
reads were assigned to the Kyoto Encyclopedia of The interactive pathways explorer 2 (iPATH2)
Genes and Genomes (KEGG) protein database (Yamada et al., 2011) was used to map the functional
(Kanehisa et al., 2004, 2008). Further quality control transcripts that were differentially expressed
consisted of selecting sequences with a minimum between early and late plant development
percent identity cutoff of 70% and an E-value cutoff (Supplementary Figure S2).
of 10  5.

Root exudation
Results
Root exudate data was obtained from Chaparro et al. Plant development influences the soil microbial
(2013) from in vitro grown Arabidopsis plants that community
were grown to the designated plant developmental We analyzed the influence of plant development
time points (seedling, vegetative, bolting and on the rhizosphere microbial community by 454
flowering). Briefly, 7-day old seedlings (see above pyrosequencing and obtained 55 921 high-quality
for seedling growth) were transferred to Magenta 16S rRNA sequence reads. After equalizing sampling
boxes containing 10 ml of MS media supplemented effort, 33 228 reads were retained for analysis. These
with 1% sucrose. Plants were grown until they were 7, reads clustered into 7452 OTUs at 3% distance
14, 21 or 28 days and were subsequently transferred sequence dissimilarity (Supplementary Table S1).
to new Magenta boxes containing 10 ml of sterilized We visualized the Bray–Curtis distances between
Millipore water. Root exudation was collected after samples using PCoA to determine how dissimilar
3 days of constant secretion (10, 17, 24 or 31 days) the soil microbial communities were at each plant
where solutions were filtered using nylon filters developmental time point. The rhizospheric micro-
(0.45 mm; Millipore, MA, USA). Root exudates were bial community of seedling was statistically
freeze-dried, dissolved in 5 ml of 80% methanol, (Po0.05) distinct from that of all other time points
dried under N gas and sent to the Genome Center (vegetative, bolting and flowering; Figure 1a). On the
Core Services at the University of California, Davis other hand, the rhizosphere microbial community
for GC–MS analysis on an Agilent 6890 gas established at vegetative, bolting and flowering were
chromatograph (see Chaparro et al., 2013). similar to each other (Figure 1a). We used PCA to
identify the factors that influence the soil microbial
community at each time point (Figure 1b). The first
Statistical analyses two principle components explain 90.6% of the
All statistical analyses were done using SAS (ver. variability in the data. Principle component 1
9.3; SAS Institute). The PROC MIXED function was explains 86.4% of the data while principle compo-
used to implement a two-way ANOVA analysis with nent 2 explains 4.2% of the data. These data clearly
a Tukey post-hoc adjustment to determine pairwise show that the soil microbial community found at
differences between the microbial communities at seedling was distinct from bolting and flowering.
each plant developmental time point. To ensure that However, we observed that the soil microbial
the data followed the assumptions of normality, community corresponding to the vegetative stage
sequences were log2 transformed. To identify if overlapped with the microbial communities estab-
developmentally dependent root exudation influ- lished at seedling and with those at bolting and
enced the soil microbial community, Pearson corre- flowering. This suggests that there is a transition
lation analysis was performed with the compounds state of the microbial community between seedling
released as root exudates (identified in Chaparro and bolting/flowering (Figure 1b).
et al. (2013)) and the corresponding phylogenetic We also determined the total OTU richness,
data. To ensure that the data abided by the evenness and diversity of the sequencing data for
assumptions of normality, the sequencing data was each time point (seedling, vegetative, bolting and
log2 transformed and the metabolomics data was log flowering; Table 1). Although there were no statis-
transformed. To determine how the functional tically significant differences between the time
microbiome changes with plant growth pairwise points with respect to overall community character-
comparisons between log2 transformed and istics, we observed that vegetative had the largest

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
794

Figure 1 Multivariate analyses of the rhizosphere microbial community through plant development analyzed by 454 pyrosequencing.
(a) Principal Coordinate Analysis (PCoA) for the visualization of pairwise community dissimilarity (Bray–Curtis index) of the
rhizosphere microbial community at each plant developmental stage (seedling, vegetative, bolting and flowering). 95% confidence
ellipses are shown around each developmental stage. (b) Principal Component Analysis (PCA) of the rhizosphere microbial community
at each plant developmental stage. 95% confidence ellipses are shown around each developmental stage.

Table 1 Observed species richness (Sobs), Shannon diversity

and evenness of the OTU soil microbial community for each plant
developmental time point

Richness Sobs Shannon

Evenness Diversity

Seedling 1107.33 0.8009 5.6136

Vegetative 1262.67 0.8638 6.1675
Bolting 1114.00 0.8192 5.7484
Flowering 1040.33 0.8245 5.7278

community richness, diversity and evenness when

compared to the other developmental stages,
whereas seedling had the lowest diversity and
evenness (Table 1). This suggests that while the Figure 2 Relative abundance (%) of the major bacterial phyla
structure of the overall soil microbial community present in the rhizosphere microbial community at each plant
developmental stage.
does not change, specific microbes may be changing
through development.
After aligning the OTUs with the SILVA database (Figure 3c). Likewise, we observed an increase in the
we classified the soil microbial community into abundance of Cyanobacteria from early time points
phylotypes consisting of 21 phyla and unclassified (seedling and vegetative) to the later time points
(Figure 2). ANOVA analysis with a Tukey post-hoc (bolting and flowering; Figure 3d). These data
test identified significant differences between the indicate that while the soil microbial community
developmental time points and four phyla (Acid- as a whole did not dramatically change, specific-soil
obacteria, Actinobacteria, Bacteroidetes and Cyano- microbial phyla were influenced by plant
bacteria). All other phyla did not significantly development.
change with development. For example, the abun- We further analyzed the soil microbial community
dance of Acidobacteria significantly increased at the genus level to determine which genera were
(Po0.05) from seedling to vegetative where it influencing the changes observed through develop-
reached its peak and then significantly decreased ment within the phyla Acidobacteria, Actinobac-
at bolting and flowering to levels similar to those at teria, Bacteroidetes and Cyanobacteria. We observed
seedling (Figure 3a). Similarly, the abundance of that four genera belonging to Acidobacteria
Actinobacteria was at its highest at the early time increased in abundance from seedling to vegetative
points (seedling and vegetative) but then signifi- when they reached their peak abundance and then
cantly decreased at the later time points (bolting and decreased at bolting reaching levels similar to that
flowering; Figure 3b). On the other hand, the of seedling during flowering (Supplementary
abundance of Bacteroidetes increased with plant Figure S3A-D). Eight genera belonging to Actino-
growth reaching its highest abundance at flowering bacteria significantly changed according to plant

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
795

Figure 3 Bacterial phyla that significantly (Po0.05) change with plant development. (a) Acidobacteria, (b) Actinobacteria,
(c) Bacteroidetes and (d) Cyanobacteria. The bars with different letters are significantly different (ANOVA Tukey post-hoc Po0.05)
from one another. Each point represents one repetition and graphs show mean±SE.

development (Supplementary Figure S4A-H). For Table 2 Pearson correlation analysis of the OTUs classified as
example, Streptomyces increased in abundance from Acidobacteria, Actinobacteria, Bacteroidetes or Cyanobacteria
seedling to vegetative when the highest abundance with the compounds released as root exudates
was reached and then significantly decreased at
Phylum Number of Number of Total number
bolting stage to levels below those observed at positive negative of correlations
seedling and remaining at these levels throughout correlations correlations
flowering (Supplementary Figure S4E). On the other
hand, six genera belonging to the Bacteroidetes Acidobacteria 156 75 231
phylum significantly increased with plant aging Actinobacteria 103 46 149
(Supplementary Figures S5A-F). In general, bacteria Bacteroidetes 8 16 24
Cyanobacteria 130 243 373
classified as Cyanobacteria were significantly more
abundant at bolting than at seedling and vegetative The values indicate the number of significant (Po0.05) Pearson
and then decreased in abundance at flowering correlations for each phyla.
(Supplementary Figures S6A-C).
categorized as amino acids, phenolics, sugars or
Plant root exudation correlates with rhizosphere sugar alcohols. Phenolic compounds significantly
microbes through plant development correlated with the soil microbial community of
To determine how developmentally dependent root Acidobacteria, Actinobacteria, Bacteroidetes and
exudate changes may influence the soil microbiome, Cyanobacteria (412 significant correlations; Table 3),
Pearson correlation analysis was performed between followed by amino acids (151), sugars (137) and sugar
the compounds released as root exudates (data from alcohols (77).
Chaparro et al. (2013)) and the rhizosphere bacteria
that significantly changed through the life span of
the plant (Supplementary Table S2). We observed Plant development influences the functional
that Cyanobacteria significantly (Po0.05) correlated microbiome
with the most root exudate compounds (373 correla- To determine the functional genes that were differ-
tions; Table 2) while Bacteroidetes correlated with entially expressed at early (seedling and vegetative)
the least (24 correlations). To determine whether a compared to late plant development (bolting and
specific class of root exudate compound was flowering) two-sample t-tests on log2 transformed
involved in soil microbial community, dynamics and standardized data were performed on the
root exudates from Chaparro et al. (2013) were metatranscriptomics data from Chaparro et al.

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
796
Table 3 Pearson correlations analysis of the OTUs classified as Acidobacteria, Actinobacteria, Bacteroidetes or Cyanobacteria with the
group of compounds released as root exudates

Phylum Amino Acids Phenolics Sugars Sugar Alcohols

Positively Negatively Positively Negatively Positively Negatively Positively Negatively

Acidobacteria 35 12 60 38 24 18 37 7
Actinobacteria 3 21 58 25 24 0 18 0
Bacteroidetes 2 0 4 10 1 5 1 1
Cyanobacteria 67 11 60 157 0 65 3 10
Total Correlations 151 412 137 77

The values indicate the number of significant (Po0.05) Pearson correlations.

(2013). A total of 81 unique transcripts out of 1240 Table 4 Taxonomic assignments of the differentially expressed
assigned to the KEGG database were significantly (81) transcripts were categorized based on their activity and
(Po0.05) differentially expressed with plant age whether the corresponding transcript were significantly (t-test
Po0.05) expressed early or late in plant development
(Supplementary Table S3). iPATH2 (Yamada et al.,
2011) was used to map at what time point (early or Early Late Total
late) these 81 functional transcripts were more
abundant (Supplementary Figure S2). Of those Symbiotic N fixers 5 9 14
transcripts, 32 were more abundant during seed- Free N fixers 6 11 17
ling and vegetative stages, while 49 were more Antagonistic 5 6 11
Plant growth promoting 0 3 3
numerous during plant bolting and flowering. The Pathogen 8 4 12
majority of the transcripts that were differently Xenobiotic/metal detoxification 4 5 9
expressed between early and late development are Unclassified function 4 11 15
involved in metabolism and genetic information
processing. Two transcripts are involved in
toluene degradation (succinate dehydrogenase
iron-sulfur protein) and the transcript involved Pseudomonas syringae. Accordingly, we categorized
in nitrotoluene degradation (hydrogenase large these microbes into seven groups depending on
subunit) were significantly (Po0.05) more abun- their potential activity towards the plant (i.e.,
dant early in plant development. On the other symbiotic N fixation, free N fixation, antagonists,
hand, a transcript involved in bacterial chemo- PGPRs, pathogens, xenobiotic/metal detoxifiers or
taxis (two-component system, chemotaxis family, unclassified; Table 4). We observed that many of
response regulator) and one involved in streptomy- these transcripts aligned to symbiotic N fixers, free
cin biosynthesis (dTDP-glucose 4,6-dehydratase) N fixers or pathogens (14, 17 or 12 transcripts
were significantly more abundant in late develop- respectively; Table 4), such as Nitrobacter,
mental stages. Among five transcripts involved in Rhodospirillum, Nitrosospira, Mesorhizobium,
nitrogen metabolism, three and two were signifi- Cyanobacteria, Azorhizobium or Dickeya dadantii.
cantly expressed at early or late plant developmen- Interestingly, most of the transcripts that aligned to
tal stages, respectively. For example, the transcript pathogens were significantly expressed early during
formamidase, carbonic anhydrase and nitric-oxide development (8 genes) while transcripts that aligned
reductase NorQ protein were highly expressed to microbes involved in plant growth promotion or
during plant seedling and vegetative stages. Con- N fixation were more abundant late (23 genes;
versely, the transcripts nitrite reductase (NAD(P)H) Table 4). For example, the transcripts involved
large subunit and periplasmic nitrate reductase in bacterial chemotaxis (two-component system,
NapA were more abundant during plant bolting chemotaxis family, response regulator) were signifi-
and flowering. cantly more abundant at late time points and aligned
to the nitrogen fixing PGPR Methylobacterium
extorquens (Ivanova et al., 2001; Sy et al., 2001;
Beneficial microbes are more active during late plant Lidstrom and Chistoserdova, 2002). On the other
development hand, the transcript involved in amoebiasis
The 81 transcripts that were significantly expressed (Ras-related protein Rab-7A) was significantly more
through development were aligned to the BLAT abundant at early time points, and it aligned with
database (Kent, 2002) within MG-RAST (Meyer et al., the pathogen Moniliophthora perniciosa. Addition-
2008) to determine which microbes were carrying ally, microbes that are antagonistic and produce
out the expressed functions (Supplementary Table S3). fungicides or bactericides such as Streptosporan-
These transcripts were attributed to bacteria gium, Streptomyces avermitilis and Sorangium
such as Bradyrhizobia, Streptomyces, Azoarcus or cellulosum were transcriptionally active late

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
797
(Table 4 and Supplementary Table S3) (Burg et al., Acidobacteria (Figure 3a) and Cyanobacteria
1979; Pradella et al., 2002). (Figure 3d) also comprised the core rhizosphere
microbiome but these phyla along with Bacteroi-
detes (Figure 3c) and Actinobacteria (Figure 3b)
Plant root exudation correlates with the functional significantly changed with plant development
microbiome through plant development suggesting that the plant can select a subset of microbes
To determine whether root exudation potentially at different stages of growth. Acidobacteria is one of the
mediates the functions carried out by the rhizo- most abundant bacterial phyla found in terrestrial
sphere microbiome, we correlated the 81 unique and ecosystems (Barns et al., 1999) and they have an
significant transcripts with the compounds released important role in the carbon cycle due to their
as root exudates (from Chaparro et al. (2013); see ability to degrade complex plant derived polysac-
Supplementary Table S4). Similar to the 16S rRNA charides such as cellulose and lignin (Ward et al.,
analysis, phenolic compounds appear to mediate 2009). Unfortunately, the specific role they play in
the expression of the transcripts (449 significant the soil ecosystem and their role in the rhizosphere
correlations; Table 5). Additionally, dTDP-glucose (apart from being metabolically active (Lee et al.,
4,6-dehydratase positively correlated with the sugar 2008)) is relatively unknown. Actinobacteria, on the
alcohol myo-Inositol. The transcript two-component other hand, has recently been associated with
system, chemotaxis family, response regulator posi- disease suppressive soils (Mendes et al., 2011).
tively correlated with glycine and xylitol. Streptomyces species, which were significantly
more abundant in the vegetative stage of Arabidopsis
(Supplementary Figure S4E), are able to increase
Discussion root nodulation efficiency and promote plant growth
Plant developmental changes affect the rhizosphere of the legume Pisum sativum (Tokala et al., 2002)
microbial community while simultaneously triggering plant defense in
Analysis of the overall bacterial rhizosphere com- Arabidopsis or apple trees (Cohen et al., 2005; Conn
munity through plant development revealed that the et al., 2008; Lin et al., 2012). On the other hand,
community did not significantly change with Bacteroidetes’ role in the rhizosphere has not yet
respect to richness, diversity and evenness been elucidated but it has been reported that they
(Table 1) but Bray–Curtis community dissimilarity are important contributors to nutrient turnover in
analysis revealed that the microbial community at the soil (Yousuf et al., 2012). Bacterial species
seedling was significantly different from the other belonging to Bacteroidetes contain genes involved
developmental stages (vegetative, bolting and flow- in denitrification indicating a possible involvement
ering; Figure 1a). These results are in agreement in N cycling (Van Spanning et al., 2005). Cyanobac-
with previous reports as the Arabidopsis rhizo- teria have been shown to colonize plant roots
sphere microbial communities after the bolting stage (Gantar et al., 1991; Lundberg et al., 2012), promote
were not distinct (Micallef et al., 2009a; Lundberg plant growth (Prasanna et al., 2009) and are an
et al., 2012). A more detailed look at the assembled important plant source for inorganic N due to their
rhizosphere microbial communities through plant ability to fix N (Franche et al., 2009).
development revealed that a core microbiome was
established and this constituted bacteria comprising
Actinobacteria, Bacteroidetes and Proteobacteria, as Plant development influences the functional capacity
was previously observed (Bulgarelli et al., 2012; of the rhizomicrobiome
Lundberg et al., 2012). In addition, the present study Metatranscriptomics analyses of mRNA have only
demonstrated that Chloroflexi, Firmicutes, Gemma- recently been used as a means to study microbial
timonadetes, Nitrospirae, Planctomycetes and communities at a functional level in distinct
Verrucomicrobia were also consistently present environments such as the human gut (Gosalbes
throughout plant development (Figure 2). Additionally, et al., 2011), the mouse gut (Xiong et al., 2012) and

Table 5 Pearson correlation analysis of the significantly expressed transcripts at early or late plant development correlated with the
group of compounds released as root exudates

Early Late Overall Total

Positive Negative Total Positive Negative Total

Amino Acids 7 160 167 198 1 199 366

Phenolics 31 139 170 222 57 279 449
Sugars 51 0 51 0 82 82 133
Sugar Alcohols 0 47 47 85 6 91 138

The values indicate the number of significant (Po0.05) correlations.

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
798
oceans (Poretsky et al., 2009; Baker et al., 2013). out by bacteria involved in N assimilation were
In soils, functional microarrays or GeoChip that target prevalent (Table 4 and Supplementary Table S3).
specific microbial functions have been successfully This may be indicative of modulation of the
used to determine the metabolic potential of the microbiome to express specific functions through-
microbial communities (He et al., 2007, 2011; Bai out plant growth, that is, differentially express
et al., 2013; Zhang et al., 2013). A drawback of the transcripts from beneficial bacteria. For example,
GeoChip is the fact that novel functions and transcripts more abundant in early time points align
transcripts cannot be identified, and one is biased to to bacteria involved in providing the plant with N
the probes present in the chip (Dugat-Bony et al., 2012). such as Nitrobacter, Rhodospirillum, Nitrosospira,
Recently, metatranscriptomics in soil has been Mesorhizobium or Azorhizobium. Similarly, during
attempted on the rRNA to identify the active bolting and flowering, functional genes expressed in
members of the microbial community in the rhizo- the rhizosphere align to PGPRs such as Bacillus
sphere of various crop species (Turner et al., 2013); licheniformis (Gutiérrez-Mañero et al., 2001) or
Urich et al. (2008) described the isolation and Burkholderia ambifaria (Chiarini et al., 2006), free
sequencing of mRNA from a sandy lawn as a means N-fixers such as Cyanothece sp. (Junier et al., 2009),
of describing the microbial transcriptome in the soil. as well as symbiotic N-fixing bacteria (Bradyrhizo-
The rhizosphere microbiome plays an essential role bium) (Stacey et al., 1995) or Herbaspirillum which is
in plant health and productivity and it is often involved in endophytic N fixation (Elbeltagy et al.,
referred as the plant’s second genome (Berendsen 2001; Franche et al., 2009). It should be noted that in
et al., 2012; Chaparro et al., 2012). Accordingly, our addition to fixing N through legume symbiosis,
metatranscriptomics data permitted a glimpse at the Bradyrhizobium promotes plant growth of non-
genes that the microbiome was expressing as a leguminous plants (Antoun et al., 1998).
whole at each stage of plant development. More The functional microbiome also expressed genes
transcripts were significantly expressed at late plant that aligned to the free N-fixing Cyanobacteria
developmental time points (Supplementary Table S3) (Supplementary Table S3). These wide changes in
and this may be indicative of the soil microbial the soil bacterial community hint at the soil
community selecting specific functions throughout microbiome shifting and changing according to the
plant development. For example, dTDP-glucose 4,6- needs of the plant and that the rhizosphere
dehydratase involved in streptomycin biosynthesis functional microbiome can express specific func-
is more abundant during plant bolting and flowering. tions at precise stages of plant growth.
Streptomycin is an antibiotic mainly produced Interestingly, we determined that the expression
by Streptomyces and it is antagonistic against gram- of a functional gene is altered through plant
positive and -negative bacteria and has been shown development even when the bacterium performing
to enhance plant defenses and trigger systemic the function does not change in abundance. For
resistance (Schatz et al., 2005; Conn et al., 2008; example, Bradyrhizobia classified as Proteobacteria
Lin et al., 2012). Additionally, we observe that dTDP- does not significantly change with plant growth
glucose-4,6-dehydratase positively correlates with (Figure 2) yet transcripts expressed by Bradyrhizobia
the root exudate compound myo-Inositol which such as imidazolonepropionase, phenol 2-monoox-
increases streptomycin biosynthesis (Majumdar and ygenase or glutaminase are more abundant in later
Kutzner, 1962; Heding, 1964). Interestingly, we did stages of plant development (Supplementary Table S3).
not find many transcripts that are attributed to plant Alternatively, while Actinobacteria, more specifi-
pathogens at late stages of plant development which cally Streptomyces, were more abundant during
is consistent with the increased production of the vegetative stage (Figure 3b, Supplementary
streptomycin and in accordance with ARR. Further- Figure S4E) we observed that genes involved in
more, we determined that Sorangium cellulosum synthesis of the antibiotic streptomycin were higher
which produce bacteriocides and fungicides during plant bolting and flowering (Supplementary
(Pradella et al., 2002) was also present in late stages Table S3). This indicates that root exudates are
of plant development. presumably able to modulate the expression of
The rhizosphere microbiome can also supply the specific functional genes without altering the
plant with essential nutrients such as nitrogen. bacterial taxonomic composition of the rhizosphere.
Nitrogen is essential for plant growth (Burns 1996;
Rossato et al., 2001) and it is usually deficient in
soils (Novoa and Loomis, 1981). Thus, in natural Root exudates act as potential stimulants for
environments (as compared to agricultural systems) rhizomicrobiome functions
the plant depends on N fixing and nitrifying bacteria While microbial colonization of the rhizosphere has
for their N supply. It should be noted that in our shown an array of benefits to the plant, the exact
study the plants grew in natural soil without mechanism by which the plant is able to attract
external fertilization. In accordance, we observed these microbes varies. One such mechanism used is
that throughout plant development functional genes bacterial chemotaxis toward root exudate compounds.
involved in nitrogen metabolism were transcription- Interestingly, we see that the two-component
ally active (Table 4). Additionally, functions carried system, chemotaxis family, response regulator

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
799
involved in bacterial chemotaxis is significantly observed (Chaparro et al., 2013). Therefore, it
expressed late in plant development and its expres- appears that sugars present in the root exudates do
sion positively correlates with the root exudates not necessarily function as general substrates for soil
glycine and xylitol (Supplementary Tables S3 and microbes as is widely reported in the literature
S4). Interestingly, PGPRs and endophytic bacteria (Behera and Wagner, 1974; Jaeger et al., 1999; Fierer
show chemotaxis towards glycine (Gaworzewska et al., 2007; Eilers et al., 2010), whereby compounds
and Carlile, 1982; de Weert et al., 2002; Gupta Sood, such as phenolics or amino acids more readily
2003), while the non-symbiotic nitrogen-fixing influence soil rhizosphere microbial communities as
bacteria Azotobacter vinelandii shows chemotactic well as modulate their transcription (Tables 3 and 5).
behavior towards xylitol (Haneline et al., 1991). Recently, Badri et al. (2013a) reported that fractions
This provides an additional mechanism for the plants of root exudates containing phenolic and phenolic-
ability to manipulate and orchestrate the rhizo- related compounds when applied to the soil (in the
sphere microbiome. While the results presented are absence of the plant) significantly modulated the
only correlative they do highlight the importance of soil microbiome. In addition, analysis of the rhizo-
root exudation in rhizosphere plant-microbiome sphere microbial community of an Arabidopsis
interactions and provide strong evidence to warrant mutant that secretes more phenolic compounds
further investigation that would conclusively than sugars showed that it cultured more beneficial
determine how specific components of the root microbes such as PGPRs and those involved in
exudates in the absence of the plant could influence N-fixation such as Bradyrhizobia and Cyanobacteria
the rhizosphere metatranscriptome. Incidentally, a when compared to wild type (Badri et al., 2009).
recent study depicting the microbial communities Interestingly, the reduction of phenolic exudates by
established from the addition of root exudate blends inhibition of phenylalanine ammonia-lyase gene
to the soil demonstrated that different blends expression in transgenic rice influenced the rhizo-
produced changes in the microbial community. For spheric microbial community, with eight phyla
example, soil supplemented with ethyl acetate- or decreasing in abundance in transgenic rice when
water-extracted root exudates generated microbial compared to wild-type plants (Fang et al., 2013).
communities that presumably had the ability to Further studies pinpointing which phenolic
metabolize the pesticide atrazine (Badri et al., compounds are involved in coordinating these
2013a). Similar experiments that add specific com- microbial interactions are needed.
pounds (sugars or oxalic acids) to the soil have also The data presented here implies and alludes to
shown dramatic shifts in the composition of the the fact that the plant through root exudation may be
microbial community (Eilers et al., 2010; Shi et al., selecting microbes for different functions; however,
2011). we would be remiss not to point out other potential
The general root exudation patterns reported in mechanisms that may be playing a role. For
Chaparro et al. (2013) do not seem to correlate with example, the changes observed in the rhizosphere
general microbiome characteristics as initially microbial community could be due to microbial
hypothesized (i.e., increased root exudation of community succession with respect to microbial
sugars at early plant development would result in competition. Unfortunately, presently there is little
enhanced richness of the rhizosphere microbial research regarding microbial succession and stabi-
community at these early time points). This could lity in soils (Fierer et al., 2010). Recently, it was
be due to the fact that root exudates were collected shown that microbial succession is similar to that of
in vitro while rhizosphere microbial communities previously described plant and animal succession
were analyzed in vivo; in other words, the root with respect to species-time relationships (Shade
secretion patterns could be different under soil et al., 2013) but what this could mean in the
conditions. Accordingly, it has been reported that rhizosphere is unclear. There is still much to be
under in vitro conditions plants grown alone or co- explored with respect to plant-microbiome interac-
cultured with a microorganism show different tions to better understand and decipher the complex
patterns of proteins present in the root exudates patterns and associations that arise in this unique
(De-la-Pena et al., 2008). However, Chaparro et al. ecological niche.
(2013) showed evidence that root exudation through Much like the plant can influence the rhizo-
development is genetically programmed since there sphere microbiome the rhizosphere microbiome
were strong correlations between in vitro root can also influence the plant. For example, fungal
exudation patterns and the ability of the soil communities have been shown to influence root
microbiome to utilize these compounds in vivo. exudation rates which can in turn influence the
During the plant vegetative stage (transition state; rhizosphere microbiome (Meier et al., 2013). Simi-
Figure 1b) we identified the largest OTU richness, larly, Lau and Lennon (2011) demonstrated that
diversity and evenness in the rhizosphere (Table 1). microbial community structure affects natural
This rhizospheric microbial transition state corre- plant trait selections. Additionally, it has been
sponds with a root exudation transition state where established that distinct microbial communities
the highest diversity of phytochemicals (sugars, influence plants’ ability to tolerate abiotic stress
sugar alcohols, phenolics and amino acids) was such as drought (Zolla et al., 2013) and even affect

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
800
leaf metabolome and subsequent insect feeding Arndt D, Xia J, Liu Y, Zhou Y, Guo AC, Cruz JA et al.
(Badri et al., 2013b). These examples highlight the (2012). METAGENassist: a comprehensive web server
multifaceted nature of the interactions in the for comparative metagenomics. Nucleic Acids Res 40:
rhizosphere microbiome. W88–W95.
Badri DV, Quintana N, El Kassis EG, Kim HK, Choi YH,
Sugiyama A et al. (2009). An ABC transporter
mutation alters root exudation of phytochemicals that
Conclusions provoke an overhaul of natural soil microbiota. Plant
The conclusions of this study could be summarized Physiol 151: 2006–2017.
as follows: (1) the plant maintains a core rhizo- Badri DV, Vivanco JM. (2009). Regulation and function of
root exudates. Plant Cell Environ 32: 666–681.
sphere microbiome; (2) this core microbiome is Badri DV, Chaparro JM, Zhang R, Shen Q, Vivanco JM.
likely to express different functions at different (2013a). Application of natural blends of phytochem-
stages of plant development; (3) the plant can icals derived from the root exudates of Arabidopsis to
enhance the expression of a subset of microbial the soil reveal that phenolic related compounds
functions at specific times that the core microbiome predominantly modulate the soil microbiome. J Biol
is not currently expressing; (4) the plant can select a Chem 288: 4502–4512.
subset of microbes at different stages of develop- Badri DV, Zolla G, Bakker MG, Manter DK, Vivanco JM.
ment presumably for specific functions that the core (2013b). Potential impact of soil microbiomes on the
microbiome can’t express (i.e., N fixation, antibiosis leaf metabolome and on herbivore feeding behavior.
against pathogens, etc.); and (5) the plant secretes New Phytol 198: 264–273.
Bai S, Li J, He Z, Van Nostrand JD, Tian Y, Lin G et al.
blends of compounds and specific phytochemicals (2013). GeoChip-based analysis of the functional gene
in the root exudates that are differentially produced diversity and metabolic potential of soil microbial
at distinct stages of plant growth to help orchestrate communities of mangroves. Appl Microbiol Biotechnol
the activities described in 1, 2, 3 and 4. Overall, 97: 7035–7048.
these concepts suggest that plants and the rhizomi- Bais HP, Weir TL, Perry LG, Gilroy S, Vivanco JM. (2006).
crobiome are in constant communication through The role of root exudates in rhizosphere interactions
the exchange of signals. Experiments targeting some with plants and other organisms. Annu Rev Plant Biol
of these ideas are essential to conclusively deter- 57: 233–266.
mine the interactive functionalities that occur in the Baker BJ, Sheik CS, Taylor CA, Jain S, Bhasi A, Cavalcoli JD
rhizosphere between crops and their microbiome. et al. (2013). Community transcriptomic assembly
reveals microbes that contribute to deep-sea carbon
and nitrogen cycling. Isme J 7: 1962–1973.
Barns SM, Takala SL, Kuske CR. (1999). Wide distribution
Conflict of Interest and diversity of members of the bacterial kingdom
The authors declare no conflict of interest. Acidobacterium in the environment. Appl Environ
Microbiol 65: 1731–1737.
Baudoin E, Benizri E, Guckert A. (2002). Impact of growth
Acknowledgements stage on the bacterial community structure along
maize roots, as determined by metabolic and genetic
Work was supported by a grant from the National Science fingerprinting. Appl Soil Ecol 19: 135–145.
Foundation to JMV (MCB-0950857). JMC participated in Bednarek P. (2012). Chemical warfare or modulators of
the experimental design, performed the experiments, defence responses - the function of secondary meta-
analyzed the data and wrote the manuscript. DVB bolites in plant immunity. Curr Opin Plant Biol 15:
participated in the experimental design, helped in data 407–414.
analyses and critical reading of the manuscript. JMV Behera B, Wagner GH. (1974). Microbial Growth Rate
participated in the experimental design, coordinated the In Glucose-amended Soil. Soil Sci Soc Am J 38:
studies and wrote the manuscript. 591–594.
Berendsen RL, Pieterse CM, Bakker PA. (2012). The
rhizosphere microbiome and plant health. Trends
Plant Sci 17: 478–486.
Bolan NS. (1991). A critical review on the role of
References mycorrhizal fungi in the uptake of phosphorus by
plants. Plant and Soil 134: 189–207.
Abedon BG, Tracy WF. (1996). Corngrass 1 of Maize (Zea Brimecombe MJ, De Leij FA, Lynch JM. (2007). Rhizode-
mays L.) Delays Development of Adult Plant Resis- position and Microbial Populations. In: Pinton R,
tance to Common Rust (Puccinia sorghi Schw.) and Varanini Z, Nannapaneni P (eds) The Rhizosphere
European Corn Borer (Ostrinia nubilalis Hubner). Biochemistry and Organic Substances at the Soil-
J Hered 87: 219–223. Plant Interface, Second Edition CRC Press: Boca
An C, Mou Z. (2011). Salicylic acid and its function in Raton, Fl, pp 73–109.
plant immunity. J Integr Plant Biol 53: 412–428. Broeckling CD, Broz AK, Bergelson J, Manter DK, Vivanco JM.
Antoun H, Beauchamp C, Goussard N, Chabot R, Lalande R. (2008). Root exudates regulate soil fungal community
(1998). Potential of Rhizobium and Bradyrhizobium composition and diversity. Appl Environ Microbiol 74:
species as plant growth promoting rhizobacteria on 738–744.
non-legumes: effect on radishes (Raphanus sativus L.). Bulgarelli D, Rott M, Schlaeppi K, Ver Loren van Themaat E,
Plant Soil 204: 57–67. Ahmadinejad N, Assenza F et al. (2012). Revealing

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
801
structure and assembly cues for Arabidopsis root- Eilers KG, Lauber CL, Knight R, Fierer N. (2010). Shifts in
inhabiting bacterial microbiota. Nature 488: 91–95. bacterial community structure associated with inputs
Burg RW, Miller BM, Baker EE, Birnbaum J, Currie SA, of low molecular weight carbon compounds to soil.
Hartman R et al. (1979). Avermectins, new family of Soil Biol and Biochem 42: 896.
potent anthelmintic agents: producing organism and Elbeltagy A, Nishioka K, Sato T, Suzuki H, Ye B, Hamada T
fermentation. Antimicrob Agents Chemother 15: et al. (2001). Endophytic colonization and in planta
361–367. nitrogen fixation by a Herbaspirillum sp. isolated from
Burns IG. (1996). Nitrogen supply, growth and develop- wild rice species. Appl Environ Microbiol 67:
ment. Acta Hort 428: 21–30. 5285–5293.
Chaparro JM, Sheflin AM, Manter DK, Vivanco JM. (2012). Fang C, Zhuang Y, Xu T, Li Y, Li Y, Lin W. (2013). Changes
Manipulating the soil microbiome to increase soil in rice allelopathy and rhizosphere microflora by
health and plant fertility. Biol Fertil Soils 48: inhibiting rice phenylalanine ammonia-lyase gene
489–499. expression. J Chem Ecol 39: 204–212.
Chaparro JM, Badri DV, Bakker MG, Sugiyama A, Manter DK, Fierer N, Bradford MA, Jackson RB. (2007). Toward an
Vivanco JM. (2013). Root exudation of phytochemicals ecological classification of soil bacteria. Ecology 88:
in Arabidopsis follows specific patterns that are 1354.
developmentally programmed and correlate with soil Fierer N, Nemergut D, Knight R, Craine JM. (2010).
microbial functions. PLoS One 8: e55731. Changes through time: integrating microorganisms
Chiarini L, Bevivino A, Dalmastri C, Tabacchioni S, into the study of succession. Res Microbiol 161:
Visca P. (2006). Burkholderia cepacia complex species: 635–642.
health hazards and biotechnological potential. Trends Franche C, Lindstrom K, Elmerich C. (2009). Nitrogen-
Microbiol 14: 277–286. fixing bacteria associated with leguminous and
Clay NK, Adio AM, Denoux C, Jander G, Ausubel FM. non-leguminous plants. Plant and Soil 321: 35–59.
(2009). Glucosinolate metabolites required for an Gantar M, Kerby NW, Rowell P, Obreht Z. (1991).
Arabidopsis innate immune response. Science 323: Colonization of wheat Triticum vulgare L.) by
95–101. N2-fixing cyanobacteria: I. A survey of soil cyanobac-
Cohen MF, Yamasaki H, Mazzola M. (2005). Brassica terial isolates forming associations with roots. New
napus seed meal soil amendment modifies microbial Phytologist 118: 477.
community structure, nitric oxide production and Gaworzewska ET, Carlile MJ. (1982). Positive Chemotaxis
incidence of Rhizoctonia root rot. Soil Biol and Of Rhizobium-Leguminosarum And Other Bacteria
Biochem 37: 1215.
Towards Root Exudates From Legumes And Other
Conn VM, Walker AR, Franco CM. (2008). Endophytic
Plants. J Gen Microbiol 128: 1179–1188.
actinobacteria induce defense pathways in Arabidopsis
Gomez-Alvarez V, Teal TK, Schmidt TM. (2009). Systematic
thaliana. Mol Plant Microbe Interact 21: 208–218.
artifacts in metagenomes from complex microbial
Coronado C, Zuanazzi J, Sallaud C, Quirion JC, Esnault R,
communities. Isme J 3: 1314–1317.
Husson HP et al. (1995). Alfalfa root flavonoid
Gosalbes MJ, Durban A, Pignatelli M, Abellan JJ,
production is nitrogen regulated. Plant Physiol 108:
533–542. Jimenez-Hernandez N, Perez-Cobas AE et al. (2011).
Davidson IA, Robson MJ. (1986). Effect of contrasting Metatranscriptomic approach to analyze the functional
patterns of nitrate application on the nitrate uptake, human gut microbiota. PLoS One 6: e17447.
N2-fixation, nodulation and growth of white clover. Gupta Sood S. (2003). Chemotactic response of plant-
Annals of Botany 57: 331–338. growth-promoting bacteria towards roots of vesicular-
De-la-Pena C, Badri DV, Lei Z, Watson BS, Brandao MM, arbuscular mycorrhizal tomato plants. FEMS Microbiol
Silva-Filho MC et al. (2010). Root secretion of Ecol 45: 219–227.
defense-related proteins is development-dependent Gutiérrez-Mañero FJ, Ramos-Solano B, An Probanza,
and correlated with flowering time. J Biol Chem 285: Mehouachi J, Tadeo R, Talon FM. (2001). The plant-
30654–30665. growth-promoting rhizobacteria Bacillus pumilus and
De-la-Pena C, Lei Z, Watson BS, Sumner LW, Vivanco JM. Bacillus licheniformis produce high amounts of
(2008). Root-microbe communication through protein physiologically active gibberellins. Physiologia Plan-
secretion. J Biol Chem 283: 25247–25255. tarum 111: 206.
de Weert S, Vermeiren H, Mulders IHM, Kuiper I, Hackstadt AJ, Hess AM. (2009). Filtering for increased
Hendrickx N, Bloemberg GV et al. (2002). Flagella- power for microarray data analysis. BMC Bioinformatics
driven chemotaxis towards exudate components is an 10: 11.
important trait for tomato root colonization by pseu- Haneline S, Connelly CJ, Melton T. (1991). Chemotactic
domonas fluorescens. Mol Plant Microbe Interact 15: behavior of azotobacter vinelandii. Appl Environ
1173–1180. Microbiol 57: 825–829.
Develey-Riviere MP, Galiana E. (2007). Resistance to He Z, Gentry TJ, Schadt CW, Wu L, Liebich J, Chong SC et al.
pathogens and host developmental stage: a multi- (2007). GeoChip: a comprehensive microarray for
faceted relationship within the plant kingdom. New investigating biogeochemical, ecological and environ-
Phytol 175: 405–416. mental processes. Isme J 1: 67–77.
Dugat-Bony E, Peyretaillade E, Parisot N, Biderre-Petit C, He ZL, Van Nostrand JD, Deng Y, Zhou JZ. (2011).
Jaziri F, Hill D et al. (2012). Detecting unknown Development and applications of functional gene
sequences with DNA microarrays: explorative probe microarrays in the analysis of the functional diversity,
design strategies. Environ Microbiol 14: 356–371. composition, and structure of microbial communities.
Eaglesham ARJ. (1989). Nitrate inhibition of root-nodule Front Environ Sci Engin China 5: 1–20.
symbiosis in doubly rooted soybean plants. Crop Heding H. (1964). Radioactive myoinositol: incorporation
science 29: 115–119. into streptomycin. Science 143: 953–954.

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
802
Horiuchi J, Prithiviraj B, Bais HP, Kimball BA, Vivanco JM. Lynch JM (ed) (1990). The Rhizosphere. Wiley Inter-
(2005). Soil nematodes mediate positive interactions science: Chinchester, UK, p 458.
between legume plants and rhizobium bacteria. Planta Majumdar SK, Kutzner HJ. (1962). Myo-inositol in the
222: 848–857. biosynthesis of streptomycin by Streptomyces griseus.
Houlden A, Timms-Wilson TM, Day MJ, Bailey MJ. (2008). Science 135: 734.
Influence of plant developmental stage on microbial Malagoli P, Laine P, Le Deunff E, Rossato L, Ney B, Ourry A.
community structure and activity in the rhizosphere (2004). Modeling nitrogen uptake in oilseed rape cv
of three field crops. FEMS Microbiol Ecol 65: 193–201. Capitol during a growth cycle using influx kinetics of
Inceoglu O, Al-Soud WA, Salles JF, Semenov AV, van Elsas JD. root nitrate transport systems and field experimental
(2011). Comparative analysis of bacterial communities data. Plant Physiol 134: 388–400.
in a potato field as determined by pyrosequencing. Meier IC, Avis PG, Phillips RP. (2013). Fungal commu-
PLoS One 6: e23321. nities influence root exudation rates in pine seedlings.
Ivanova EG, Doronina NV, Trotsenko IuA. (2001). Aerobic FEMS Microbiol Ecol 83: 585–595.
methylobacteria are capable of synthesizing auxins. Mendes R, Kruijt M, De Bruijn I, Dekkers E, van der Voort M,
Mikrobiologiia 70: 452–458. JHM Schneider et al. (2011). Deciphering the rhizo-
Jaeger CH 3rd, Lindow SE, Miller W, Clark E, Firestone sphere microbiome for disease-suppressive bacteria.
MK. (1999). Mapping of sugar and amino acid Science 332: 1097–1100.
availability in soil around roots with bacterial sensors Meyer F, Paarmann D, D’Souza M, Olson R, Glass EM,
of sucrose and tryptophan. Appl Environ Microbiol 65: Kubal M et al. (2008). The metagenomics RAST
2685–2690. server—a public resource for the automatic phyloge-
Jain V, Nainawatee HS. (2002). Plant flavonoids: signals to netic and functional analysis of metagenomes. Bmc
legume nodulation and soil microorganisms. J Plant Bioinformatics 9: 386.
Biochem Biot 11: 1–10. Micallef SA, Channer S, Shiaris MP, Colon-Carmona A.
Junier P, Junier T, Witzel KP, Caru M. (2009). Composition (2009a). Plant age and genotype impact the progres-
of diazotrophic bacterial assemblages in bean-planted sion of bacterial community succession in the
soil compared to unplanted soil. Eur J Soil Biol 45: Arabidopsis rhizosphere. Plant Signal Behav 4:
153–162. 777–780.
Kanehisa M, Goto S, Kawashima S, Okuno Y, Hattori M. Micallef SA, Shiaris MP, Colon-Carmona A. (2009b).
(2004). The KEGG resource for deciphering the Influence of Arabidopsis thaliana accessions on
genome. Nucleic Acids Res 32: D277–D280.
rhizobacterial communities and natural variation in
Kanehisa M, Araki M, Goto S, Hattori M, Hirakawa M,
root exudates. J Exp Bot 60: 1729–1742.
Itoh M et al. (2008). KEGG for linking genomes to life
Millet YA, Danna CH, Clay NK, Songnuan W, Simon MD,
and the environment. Nucleic Acids Res 36: D480–D484.
Werck-Reichhart D et al. (2010). Innate immune
Kelly JT, Bacon RK, Wells BR. (1995). Genetic variability
responses activated in Arabidopsis roots by microbe-
in nitrogen utilization at four growth stages in soft red
associated molecular patterns. Plant Cell 22: 973–990.
winter wheat. J Plant Nutr 18: 969.
Morgan JA, Bending GD, White PJ. (2005). Biological costs
Kent WJ. (2002). BLAT–the BLAST-like alignment tool.
Genome Res 12: 656–664. and benefits to plant-microbe interactions in the
Kus JV, Zaton K, Sarkar R, Cameron RK. (2002). Age- rhizosphere. J Exp Bot 56: 1729–1739.
related resistance in Arabidopsis is a developmentally Mougel C, Offre P, Ranjard L, Corberand T, Gamalero E,
regulated defense response to Pseudomonas syringae. Robin C et al. (2006). Dynamic of the genetic structure
Plant Cell 14: 479–490. of bacterial and fungal communities at different
Langmead B, Trapnell C, Pop M, Salzberg SL. (2009). developmental stages of Medicago truncatula Gaertn.
Ultrafast and memory-efficient alignment of short cv. Jemalong line J5. New Phytol 170: 165–175.
DNA sequences to the human genome. Genome Biol Murashige T, Skoog F. (1962). A revised medium for rapid
10: R25. growth and bio assays with tobacco tissue cultures.
Lau JA, Lennon JT. (2011). Evolutionary ecology of Physiol Plant 15: 473.
plant-microbe interactions: soil microbial structure Nazoa P, Vidmar JJ, Tranbarger TJ, Mouline K, Damiani I,
alters selection on plant traits. New Phytologist 192: Tillard P et al. (2003). Regulation of the nitrate
215–224. transporter gene AtNRT2.1 in Arabidopsis thaliana:
Lee SH, Ka JO, Cho JC. (2008). Members of the phylum responses to nitrate, amino acids and developmental
Acidobacteria are dominant and metabolically active stage. Plant Mol Biol 52: 689–703.
in rhizosphere soil. FEMS Microbiol Lett 285: 263–269. Neal AL, Ahmad S, Gordon-Weeks R, Ton J. (2012).
Li X, Bergelson J, Chapple C. (2010). The ARABIDOPSIS Benzoxazinoids in root exudates of maize attract
accession Pna-10 is a naturally occurring sng1 pseudomonas putida to the rhizosphere. PLoS One 7:
deletion mutant. Mol Plant 3: 91–100. e35498.
Lidstrom ME, Chistoserdova L. (2002). Plants in the pink: Nicoli A, Zambolim L, TJP Junior, Vieira RF, Teixeira H,
cytokinin production by methylobacterium. J Bacteriol JES Carneiro. (2012). Resistance of advanced common
184: 1818. bean lines to Fusarium root rot. Trop plant pathol 37:
Lin L, Ge HM, Yan T, Qin YH, Tan RX. (2012). Thaxtomin 393–398.
A-deficient endophytic Streptomyces sp. enhances Novoa R, Loomis RS. (1981). Nitrogen and plant production.
plant disease resistance to pathogenic Streptomyces Plant and Soil 58: 177–204.
scabies. Planta 236: 1849–1861. Oksanen J, Blanchet G, Kindt R, Legendre P, Minchin PR,
Lundberg DS, Lebeis SL, Paredes SH, Yourstone S, O’Hara RB et al. (2012). vegan: Community Ecology
Gehring J, Malfatti S et al. (2012). Defining the core Package. R package version 2.0-4.
Arabidopsis thaliana root microbiome. Nature 488: Poretsky RS, Gifford S, Rinta-Kanto J, Vila-Costa M,
86–90. Moran MA. (2009). Analyzing gene expression from

The ISME Journal

 Rhizosphere microbiome assemblage with plant age
JM Chaparro et al
803
marine microbial communities using environmental the rhizosphere microbiome of plants. Isme J 7:
transcriptomics. J Vis Exp 24: e1086. 2248–2258.
Pradella S, Hans A, Sproer C, Reichenbach H, Gerth K, Urich T, Lanzen A, Qi J, Huson DH, Schleper C, Schuster SC.
Beyer S. (2002). Characterisation, genome size and (2008). Simultaneous assessment of soil microbial
genetic manipulation of the myxobacterium Sorangium community structure and function through analysis
cellulosum So ce56. Arch Microbiol 178: 484–492. of the meta-transcriptome. PLoS One 3: e2527.
Prasanna R, Jaiswal P, Nayak S, Sood A, Kaushik BD. (2009). van der Heijden MGA, Bardgett RD, Van Straalen NM.
Cyanobacterial diversity in the rhizosphere of rice and (2008). The unseen majority: soil microbes as drivers
its ecological significance. Indian J Microbiol 49: 89–97. of plant diversity and productivity in terrestrial
Rogers EE, Glazebrook J, Ausubel FN. (1996). Mode of ecosystems. Ecol Lett 11: 296–310.
action of the Arabidopsis thaliana phytoalexin Van Spanning RJM, Delgado MJ, Richardson DJ. (2005).
camalexin and its role in Arabidopsis-pathogen The nitrogen cycle: denitrification and its relationship
interactions. Mol Plant Microbe In 9: 748–757. to N2 fixation. In: Werner D, Newton WE (eds)
Rossato L, Laine P, Ourry A. (2001). Nitrogen storage and Nitrogen Fixation in Agriculture, Forestry, Ecology,
remobilization in Brassica napus L. during the growth and the Environment. Springer: Netherlands,
cycle: nitrogen fluxes within the plant and changes in pp 277–342.
soluble protein patterns. J Exp Bot 52: 1655–1663. Wang Q, Garrity GM, Tiedje JM, Cole JR. (2007). Naive
Rusterucci C, Zhao Z, Haines K, Mellersh D, Neumann M, Bayesian classifier for rapid assignment of rRNA
Cameron RK. (2005). Age-related resistance to sequences into the new bacterial taxonomy. App
Pseudomonas syringae pv. tomato is associated with Environ Microb 73: 5261–5267.
the transition to flowering in Arabidopsis and is Ward NL, Challacombe JF, Janssen PH, Henrissat B,
effective against Peronospora parasitica. Physiol Mol Coutinho PM, Wu M et al. (2009). Three genomes
Plant P 66: 222. from the phylum Acidobacteria provide insight into
Schatz A, Bugie E, Waksman SA. (2005). Streptomycin, a the lifestyles of these microorganisms in soils. Appl
substance exhibiting antibiotic activity against Environ Microbiol 75: 2046–2056.
gram-positive and gram-negative bacteria. 1944. Clin Xiong X, Frank DN, Robertson CE, Hung SS, Markle J,
Orthop Relat Res 3–6. Canty AJ et al. (2012). Generation and analysis of a
Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, mouse intestinal metatranscriptome through Illumina
Hollister EB et al. (2009). Introducing mothur: open- based RNA-sequencing. PLoS One 7: e36009.
source, platform-independent, community-supported Xu Y, Wang G, Jin J, Liu J, Zhang Q, Liu X. (2009). Bacterial
software for describing and comparing microbial communities in soybean rhizosphere in response to
communities. Appl Environ Microbiol 75: 7537–7541. soil type, soybean genotype, and their growth stage.
Selvakumar G, Panneerselvam P, Ganeshamurthy AN, Soil Biol Biochem 41: 919.
Maheshwari DK. (2012). Bacterial mediated allevia- Yamada T, Letunic I, Okuda S, Kanehisa M, Bork P. (2011).
tion of abiotic stress in crops. In: Maheshwari DK (ed) iPath2.0: interactive pathway explorer. Nucleic Acids
Bacteria in Agrobiology: Stress Management. Springer: Res 39: W412–W415.
New York, NY, USA, pp 205–224. Yousuf B, Keshri J, Mishra A, Jha B. (2012). Application of
Shade A, Gregory Caporaso J, Handelsman J, Knight R, targeted metagenomics to explore abundance and
Fierer N. (2013). A meta-analysis of changes in diversity of CO(2)-fixing bacterial community using
bacterial and archaeal communities with time. Isme J cbbL gene from the rhizosphere of Arachis hypogaea.
7: 1493–1506. Gene 506: 18–24.
Shaw CH. (1991). Swimming against the tide: Chemotaxis Zahran HH. (1999). Rhizobium-legume symbiosis and
in Agrobacterium. BioEssays 13: 25–29. nitrogen fixation under severe conditions and in an
Shi S, Richardson AE, O’Callaghan M, DeAngelis KM, arid climate. Microbiol Mol Biol Rev 63: 968–989.
Jones EE, Stewart A et al. (2011). Effects of selected Zaki K, Misaghi IJ, Heydari A. (1998). Control of Cotton
root exudate components on soil bacterial commu- Seedling Damping-off in the Field by Burkholderia
nities. FEMS Microbiol Ecol 77: 600–610. (Pseudomonas) cepacia. Plant Disease 82: 291–293.
Stacey G, Sanjuan J, Luka S, Dockendorff T, Carlson RW. Zamioudis C, Pieterse CM. (2012). Modulation of host
(1995). Signal exchange in the Bradyrhizobium- immunity by beneficial microbes. Mol Plant Microbe
soybean symbiosis. Soil Biol and Biochem 27: 473. Interact 25: 139–150.
Sy A, Giraud E, Jourand P, Garcia N, Willems A, Zhang J, Subramanian S, Stacey G, Yu O. (2009). Flavones
De Lajudie P et al. (2001). Methylotrophic Methylo- and flavonols play distinct critical roles during
bacterium bacteria nodulate and fix nitrogen in nodulation of Medicago truncatula by Sinorhizobium
symbiosis with legumes. J Bacteriol 183: 214–220. meliloti. Plant J 57: 171–183.
Tokala RK, Strap JL, Jung CM, Crawford DL, Salove MH, Zhang Y, Lu Z, Liu S, Yang Y, He Z, Ren Z et al. (2013).
Deobald LA et al. (2002). Novel plant-microbe rhizo- Geochip-based analysis of microbial communities in
sphere interaction involving Streptomyces lydicus alpine meadow soils in the Qinghai-Tibetan plateau.
WYEC108 and the pea plant (Pisum sativum). Appl BMC Microbiol 13: 72.
Environ Microbiol 68: 2161–2171. Zolla G, Badri DV, Bakker MG, Manter DK, Vivanco JM.
Turner TR, Ramakrishnan K, Walshaw J, Heavens D, (2013). Soil microbiomes vary in their ability to
Alston M, Swarbreck D et al. (2013). Comparative confer drought tolerance to Arabidopsis. App Soil
metatranscriptomics reveals kingdom level changes in Ecol 68: 1.

Supplementary Information accompanies this paper on The ISME Journal website (http://www.nature.com/ismej)

The ISME Journal