Microdiversity of Burkholderiales associated with mycorrhizal and nonmycorrhizal roots of Medicago truncatula

R E S E A R C H A R T I C L E

MicrodiversityofBurkholderiales associatedwithmycorrhizal andnonmycorrhizal rootsofMedicago truncatulaPierre Offre1, Barbara Pivato1,2, Sylvie Mazurier1, Severine Siblot1, Graziella Berta2, PhilippeLemanceau1 & Christophe Mougel1

1INRA, Universite de Bourgogne, UMR1229 ‘Microbiologie du Sol et de l’Environnement’, CMSE, BP, Dijon, France; and2Universita del Piemonte Orientale ‘Amedeo Avogadro’, Dipartimento di Scienze dell’Ambiente e della Vita, Alessandria, Italy

Correspondence: Philippe Lemanceau,

INRA, Universite de Bourgogne, UMR1229

‘Microbiologie du Sol et de l’Environnement’,

CMSE, 17 rue Sully, BP 86510, F-21065 Dijon,

France. Tel.: 133 3 80 69 30 56; fax: 133 3

80 69 32 24; e-mail: [email protected]

Received 15 September 2007; revised 25 March

2008; accepted 2 April 2008.

First published online 28 May 2008.

DOI:10.1111/j.1574-6941.2008.00504.x

Editor: Jim Prosser

Keywords

arbuscular mycorrhizas; rhizosphere; bacterial

diversity; Comamonadaceae ;

Oxalobacteraceae .

Abstract

The genetic diversity of bacterial communities associated with mycorrhizal and

nonmycorrhizal roots of Medicago truncatula was characterized by two ap-

proaches. Firstly, phylogenetic analysis was performed on 164 partial 16S rRNA

gene–intergenic spacer (IGS) sequences from operational taxonomic units pre-

viously shown to be preferentially associated with mycorrhizal roots. These

sequences were distributed into three branches corresponding to Comamonada-

ceae, Oxalobacteraceae and Rubrivivax subgroups. Most sequences were obtained

from mycorrhizal roots, indicating the preferential association of the correspond-

ing families with mycorrhizal roots. A second phylogenetic analysis was performed

on the partial 16S rRNA gene–IGS sequences of 173 isolates among a large

collection of isolates, from mycorrhizal and nonmycorrhizal roots, belonging to

Comamonadaceae and Oxalobacteraceae on the basis of their positive hybridization

with a partial 16S rRNA gene–IGS probe obtained in this study. Sequence analysis

confirmed the affiliation of 166 isolates to Comamonadaceae and seven to

Oxalobacteraceae. Oxalobacteraceae isolates were more abundant in mycorrhizal

(five) than in nonmycorrhizal (two) roots, whereas Comamonadaceae isolates were

more abundant in nonmycorrhizal (109) than mycorrhizal roots (57). Further

analysis of Comamonadaceae isolates by BOX-PCR showed that the genetic

structure of culturable populations belonging to this family differed significantly

in mycorrhizal and nonmycorrhizal roots, as indicated by distributions in different

BOX types, differences being significantly explained by BOX types only including

isolates from mycorrhizal roots. These data are discussed in an ecological context.

Introduction

Arbuscular mycorrhizas (AM) are symbiotic associations

between most land plants (Wang & Qiu, 2006) and obligate

biotrophic AM fungal species belonging to the Glomeromy-

cota phylum (Schußler et al., 2001). The AM association

represents an ancient symbiosis with fossil evidence dating

back 400 millions years (Remy et al., 1994). The roots of

earliest land plants contained arbuscular mycorrhizal struc-

tures and the symbiosis of ancestral plants with the Glomer-

omycota may have enabled plants to colonize land

(Brundrett, 2002). AM have a central position in terrestrial

nutrient-cycling processes and are of particular interest

because of their positive effects on plant growth and health

(Smith & Read, 1997).

Because of the long coevolution of plants and AM fungi,

reciprocal interactions based on feedbacks between them

have been proposed (Bever et al., 2002). This proposal is

based on the observation of differential responses of plant

species to individual isolates and species of AM fungi

(Streitwolf-Engel et al., 1997) leading to variations in plant

composition and productivity according to AM fungal

diversity and identity (van der Heijden et al., 1998; Klirono-

mos et al., 2000). In return plant species, even those that are

closely related, differentially affect the AM fungal commu-

nity associated with mycorrhizal roots (Vandenkoornhuyse

et al., 2003; Pivato et al., 2007).

Coevolution of plants and AM fungi is not expected

to have occurred independently from the associated

bacterial communities. However, information on bacteria

FEMS Microbiol Ecol 65 (2008) 180–192c� 2008 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

preferentially associated with mycorrhizal roots appears to

be scarce. Few studies have addressed the relationship

between free-living bacteria and AM fungi (Meyer & Linder-

man, 1986; Andrade et al., 1997; Marschner et al., 2001;

Artursson et al., 2005). Until recently (Offre et al., 2007),

none of these studies has identified taxa explaining varia-

tions of bacterial community structure associated with AM.

Furthermore, these studies did not take into account the

natural diversity of AM fungi, because they were based

on inoculation of selected strains of Glomeromycota.

Progress in our knowledge on bacteria associated with

mycorrhizal roots is required. Indeed, from an evolutionary

point of view, one may expect that these bacteria would at

least be nondeleterious or even beneficial to mycorrhization,

as has been shown for a fluorescent pseudomonad repre-

sentative of populations preferentially associated with

ectomycorrhizas (Frey et al., 1997), and with Paenibacillus

sp. strain B2, isolated from the mycorrhizosphere of

Sorghum bicolor–Glomus mosseae and promoting AM (Budi

et al., 1999).

Based on these statements, a research was undertaken to

characterize bacteria preferentially associated with mycor-

rhizal roots. The first step of this research was to compare

the bacterial communities associated with mycorrhizal and

nonmycorrhizal plant genotypes when cultivated in a fallow

soil from the Mediterranean area (Mas d’Imbert, France),

corresponding to the diversification zone of annual Medics.

The genetic structure of these communities, assessed by

Automated-Ribosomal Intergenic Spacer Analysis (A-RISA)

from DNA directly extracted from root tissues, appeared to

differ significantly. The molecular markers [partial 16S

rRNA gene and 16S–23S intergenic spacer (IGS) sequences]

explaining these differences were cloned and sequenced. The

corresponding sequences, grouped into operational taxo-

nomic units (OTUs), showed a high level of similarity with

partial 16S rRNA gene sequences of Methylibium petrolei-

philum and Collimonas fungivorans belonging to Comamo-

nadaceae and Oxalobacteraceae (Burkholderiales),

respectively (Offre et al., 2007).

The second step and the aim of this study were to

characterize further the diversity of Burkholderiales preferen-

tially associated with mycorrhizal roots using molecular and

cultivation-based approaches. The molecular approach was

based on phylogenetic analysis of partial 16S rRNA gene–IGS

sequences belonging to OTUs identified previously (Offre

et al., 2007). The cultivation-based approach involved analy-

sis of a large collection of bacterial isolates from root tissues of

Medicago truncatula J5 and its AM defective mutant TRV25,

cultivated in the Mas d’Imbert soil. Among them, isolates

belonging to Comamonadaceae and Oxalobacteraceae were

expected to be preferentially associated with mycorrhizal

roots (Offre et al., 2007) and were detected by dot-blot

hybridization with a polynucleotide probe. The genetic

diversity of these isolates was characterized by analysing their

partial 16S rRNA gene–IGS sequences and bacterial repetitive

BOX element (BOX)-PCR fingerprinting.

Materials and methods

Plant growth conditions and bacterial isolation

Experiments were conducted with M. truncatula Gaertn. cv.

Jemalong line J5 (Myc1/Nod1) and its symbiosis-defective

mutant TRV25 (Myc�/Nod�) (Sagan et al., 1995). Seeds of

the two plant genotypes were scarified and surface-sterilized

by gently shaking in 98% sulphuric acid for 2 min, 95%

ethanol for 5 min and 3.5% sodium hypochlorite solution

for 10 min and rinsed successively six times for 5 min in

sterile demineralized water. Sterilized seeds were germinated

on a 0.7% (w/v) water agar plate at 25 1C for 48 h. Twelve

plants were cultivated per M. truncatula genotype by sowing

one germinated seed per cylindrical polyvinyl chloride

container (diameter: 9 cm, height: 20 cm) filled with 1.5 L

of a silt-clay loam soil (Mas d’Imbert, France). They were

cultivated in a growth chamber at a 16 : 8 h light/dark

photoperiod, a 23 : 18 1C light/dark thermoperiod, photo-

synthetic active radiation (PAR) = 500 mE m�2 s�1 at pot

height and 55% relative humidity. Water was added daily

to maintain humidity at 55% of the water-holding capacity.

Containers were randomly distributed in the growth cham-

ber and moved each day. The plants were sampled after 34

days, corresponding to the appearance of shoot ramifica-

tions of order 2 (Mougel et al., 2006).

Bacteria were isolated from four root samples per plant

genotype, each sample resulting from pooling three root

systems. Adhering soil was removed from the roots by

shaking them twice in a flask containing 100 mL of sterile

distilled water at 200 r.p.m. for 10 min (Agitests, Bioblock

Scientific, Illkirch, France). The washed roots were ground

with an Ultra-Turax (Janke and Kunkel GmbH and Co.KG,

Staufen, Germany) in 50 mL of sterile distilled water. Root

suspensions were dilution-plated on 0.1� tryptic soy agar

(TSA) (Forbes et al., 1998) supplemented with 200 mg mL�1

of cycloheximide (Sigma, St Louis, MO) and incubated for 5

days at 25 1C. A total of 2412 bacterial colonies were isolated:

1200 from M. truncatula J5 and 1212 from M. truncatula

TRV25 (Table 1). Isolates were subjected to single-colony

isolation and stored at � 80 1C in tryptic soy broth (TSB)

containing 20% glycerol.

Sequences of 16S rRNA gene--IGS

One hundred and sixty-four partial sequences of 16S rRNA

genes and 16S–23S IGS obtained previously from DNA

extracted from root tissues of M. truncatula were analysed.

These sequences were chosen as belonging to OTUs pre-

viously shown to be preferentially associated with

FEMS Microbiol Ecol 65 (2008) 180–192 c� 2008 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

181Bacterial diversity associated with mycorrhizal roots

mycorrhizal roots (Offre et al., 2007). Ninety-five partial

sequences of 16S rRNA gene (465 bp) showed 99% similarity

to that of C. fungivorans (gi: 33145982) plus IGS (515 bp), 24

partial sequences of 16S rRNA gene (465 bp) 99% similarity to

that of C. fungivorans (gi: 33145982) plus IGS (705 bp) and 45

partial sequences of 16S rRNA gene (465 bp) 97% similarity to

that of M. petroleiphilum (gi: 124265193) plus IGS (705 bp)

(Table 2). The accession number of a representative sequence

(sequence-type) of each type of partial 16S rRNA gene–IGS

sequence is given in Table 2.

Reference bacterial strains

The reference strains used in this study are listed in Table 3.

Reference strain C. fungivorans Ter331 (de Boer et al., 2004)

was grown at 25 1C on chitin/yeast agar (de Boer et al.,

2004), Pseudomonas fluorescens C7R12 (Eparvier et al.,

1991) at 25 1C on solid King’s B medium (KB) (King et al.,

1954) and Variovorax sp. MD1 (Devers et al., 2005) at 28 1C

on modified mineral salt agar (MSA) (Devers et al., 2005).

Genomic DNA of reference strains Agrobacterium tumefa-

ciens C58 (Goodner et al., 2001), Sinorhizobium meliloti

ATCC 9930 (Galibert et al., 2001), Bradyrhizobium japoni-

cum USDA110 (Kaneko et al., 2002) and Escherichia

coli S.17-1 (Simon et al., 1983) were obtained from a

previous study (Mazurier et al., 2006). Genomic DNA of

reference strain Xanthomonas campestris pv. vesicatoria

(Stall et al., 1986) was generously provided by Lionel

Ranjard (UMR MSE, INRA Dijon, France) and those of

Paenibacillus sp. strain B2 (Budi et al., 1999), Staphylococcus

epidermidis RP62A (Gill et al., 2005) and Listeria innocua

CIP 8012 (Roccourt et al., 1982) by Alain Hartmann

(UMR MSE).

Cell lysis

Heat-lysed bacterial suspensions used for PCR analyses were

prepared from cultures grown on 0.1� TSA for 5 days at

Table 1. Isolates from mycorrhizal (Medicago truncatula J5) and non-

mycorrhizal (M. truncatula TRV25) roots tested in this study

Origin

Total

Medicago

truncatula J5

Medicago

truncatula

TRV25

Total number of isolates 1200 1212 2412

Isolates with positive 16S rRNA

gene–IGS PCR

703 815 1518

Isolates with positive hybridization

(probe AM396929)

258 286 544

Isolates with analysed 16S rRNA

gene–IGS sequence

207 273 480

Isolates with partial 16S rRNA

gene highly similar (99%)

to that of Collimonas fungivorans

(GenBank accession no. 33145982)

5 2 7

Isolates with partial 16S rRNA

gene highly similar (97%) to

that of Methylibium petroleiphilum

(GenBank accession no. 6063088)

57 109 166

Table 2. Partial sequences, analysed in this study, of 16S rRNA gene–IGS previously shown to belong to OTUs preferentially associated with mycorrhizal

roots of Medicago truncatula (Offre et al., 2007)

No. of

sequences

16S rRNA gene sequenceIGS Accession number of

representative sequence ReferencesSize (bp) Closest relative in GenBank (accession no.) Similarity (%) Size (bp)

95 465 Collimonas fungivorans (33145982) 99 515 AM396929 Offre et al. (2007)

24 465 Collimonas fungivorans (33145982) 99 705 AM396926 Offre et al. (2007)

45 465 Methylibium petroleiphilum (6063088) 97 705 AM396923 Offre et al. (2007)

Table 3. Reference bacterial strains used in this study

Strain Family (phylum) References

Agrobacterium tumefaciens C58 Rhizobiaceae (Alphaproteobacteria) Goodner et al. (2001)

Bradyrhizobium japonicum USDA110 Bradyrhizobiaceae (Alphaproteobacteria) Kaneko et al. (2002)

Collimonas fungivorans Ter331 Oxalobacteraceae (Betaproteobacteria) de Boer et al. (2004)

Escherichia coli S.17-1 Enterobacteriaceae (Gammaproteobacteria) Simon et al. (1983)

Listeria innocua CIP 8012 Listeriaceae (Firmicutes) Roccourt et al. (1982)

Paenibacillus sp. B2 Paenibacillaceae (Firmicutes) Budi et al. (1999)

Pseudomonas fluorescens C7R12 Pseudomonadaceae (Gammaproteobacteria) Eparvier et al. (1991)

Sinorhizobium meliloti ATCC 9930 Rhizobiaceae (Alphaproteobacteria) Galibert et al. (2001)

Staphylococcus epidermidis RP62A Staphylococcaceae (Firmicutes) Gill et al. (2005)

Variovorax sp. MD1 Comamonadaceae (Betaproteobacteria) Devers et al. (2005)

Xanthomonas campestris pv. vesicatoria Xanthomonadaceae (Gammaproteobacteria) Stall et al. (1986)


182 P. Offre et al.

25 1C for bacterial isolates, and as described above for the

reference strains. One colony per bacterial isolate/strain was

suspended in 100mL of lysis solution [0.05 M NaOH, 0.25%

sodium dodecyl sulphate (SDS)] and incubated for 15 min

in boiling water. The resulting suspension was centrifuged

for 1 min at 15 294 g and the supernatant was diluted 50-fold

in sterile water.

PCR amplification and sequencing of 16S rRNAgene--IGS

Sequences of partial 16S rRNA gene–IGS from DNA extracted

previously from root tissues (Table 2) and from bacterial

isolates were amplified by PCR with the primers 1055F

(50-ATGGCTGTCGTCAGCTT-30) and L-D-Bact-132-a-

A-18 (50-CCGGGTTTCCCCATTCGG-30) (Normand et al.,

1996). Reactions were performed in a total volume of 25mL,

by mixing 1 ngmL�1 of DNA, 1� of buffer containing MgCl2[10 mM Tris-HCl (pH 9), 50 mM KCl, 1.5 mM MgCl2, 0.1%

Triton X-100, 0.2 mg mL�1 bovine serum albumin solution

(BSA)] (Q-BIOgene, Illkirch, France), 200mM of dNTPs

(Q-BIOgene), 1mM of each primer and 1.5 U of Taq DNA

polymerase (Q-BIOgene). Each reaction was performed in a

thermal cycler (MJ Research PTC-0225 DNA Engine Tetrad,

Bio-rad, Marne la Coquette, France). PCR conditions were as

follows: a first step at 72 1C for 6 min, a second step at 95 1C

for 5 min, followed by 30 cycles of 94 1C for 1 min, 55 1C for

1 min and 72 1C for 1 min, before a final elongation step for

5 min at 72 1C. PCR products were checked on 2% (w/v)

agarose gels in TAE buffer [40 mM Tris (pH 7.8), 20 mM

acetic acid, 2 mM EDTA] for size determination and were

sequenced using the 1055F primer.

Sequencing of partial 16S rRNA gene–IGS was performed by

the GENOME express company (Genome Express, Meylan,

France) with a standard sequencing protocol. Sequences pro-

vided as text files were used for sequence analysis and phyloge-

netic inference. Newly obtained partial sequences of 16S rRNA

gene–IGS were deposited in GenBank under the following

accession numbers: EU074784, EU074785, EU074786,

EU074787, EU074788, and EU074789.

Development of probe and hybridizationconditions

Hybridizations were performed on Pall Biodyne Plus mem-

branes (VWR France, Fontenay-sous-Bois, France). Ten

nanograms of PCR product was spotted onto the membrane

after denaturation for 10 min in a boiling water bath using a

dot-blot vacuum manifold (Schleicher & Schuell, Dassel,

Germany). DNA was fixed on a membrane for 30 min at

80 1C. Attempts were made to prepare probes from the

sequences AM396929 and AM396923 (Table 2), but were

only successful with the sequence AM396929. The probe was

obtained by labelling this sequence with digoxigenin by PCR

performed in the same reaction mixture as described above,

except for the final volume (50 mL) and for the dNTPs,

which were used at the following concentrations: 5 mM of

dATP, 5 mM of dGTP, 5mM of dCTP, 0.5mM of dTTP and

0.2 mM of Dig-11-dUTP (Roche, Molecular Biochemicals,

Meylan, France). DNA hybridization was performed over-

night at 68 1C in 10 mL of hybridization buffer [sodium

saline citrate (SSC) 5� , 0.1% (w/v) sodium lauroyl sarco-

sinate, 0.02% (w/v) SDS and 0.5% (v/v) blocking reagent

(Roche)] containing 10 ng mL�1 of the AM396929 probe.

The washing steps were performed applying high-stringency

conditions: twice for 5 min at room temperature in 0.1%

SDS and SSC 2� , twice for 15 min at 68 1C in 0.1% SDS

and SSC 0.2� and once for 15 min at 68 1C in 0.1% SDS

and SSC 0.02� . Probe detection was performed using the

DIG Luminescent Detection kit (Roche), following the

manufacturer’s instructions.

Phylogenetic analyses

Sequences of partial 16S rRNA gene–IGS from DNA extracted

previously from root tissues (Table 2) and those from bacterial

isolates hybridizing with the AM396929 probe were used as

queries in Basic Alignment Search Tool (BLAST) searches for

standard nucleotide–nucleotide BLAST (Altschul et al., 1990).

Nucleotide sequences corresponding to queries and the most

closely relative sequences identified by BLAST analysis were

applied to construct multiple sequence alignments as deter-

mined by CLUSTAL W (Thompson et al., 1994). Optimization of

the multiple sequence alignments was performed using the

graphical editor SeaView (Galtier et al., 1996).

Phylogenetic inferences were performed using the PHYLO_WIN

graphical interface (Galtier et al., 1996). Phylogenies were

determined by the neighbour-joining method (Saitou & Nei,

1987) using the Kimura ‘two-parameters’ correction (Kimura,

1980) with the pairwise gap removal option. In order to

estimate tree node validity, the result of 1000 bootstrapped data

sets was determined. Reconstructed trees were drawn and

evaluated using the NJPLOT program (Perriere & Gouy, 1996).

BOX-PCR

Isolates having both a positive hybridization with the

AM396929 probe and a high similarity (at least equal to 97%)

of their 16S rRNA gene with the corresponding partial

sequence (465 bp) of C. fungivorans (gi: 33145982) and of

M. petroleiphilum (gi: 6063088) were selected for repetitive

sequences PCR (rep-PCR) characterization. Rep-PCR finger-

printing was achieved using the primer BOXA1R (50-CTACG

GCAAGGCGACGCTGACG-30) (Versalovic et al., 1994).

BOX-PCR reactions were performed in a total volume of



25mL, by mixing 3mL of the diluted heat-lysed bacterial

suspension, 5mL of 5� Gitschier buffer [83 mM (NH4)2SO4,

335 mM Tris-HCl (pH 8.8), 33.5 mM MgCl2, 33.5mM EDTA,

150 mM b-mercaptoethanol], 0.4mL of 10 mg mL�1 BSA

(Q-BIOgene), 2.5mL of dimethyl sulphoxide, 11.7mL of sterile

distilled water, 25mM each of dATP, dCTP, dGTP, and dTTP,

1mM of BOXA1R primer (Eurogentec, Seraing, Belgium) and

2.25 U of Taq DNA polymerase (Q-BIOgene). Amplification

reactions were processed in a Perkin Elmer GeneAmp PCR

System 9600 thermal cycler (Applied Biosystems, Foster City,

CA). Cycling conditions for BOX-PCR were as follows: 95 1C

for 2 min and then 35 cycles of 94 1C for 3 s, 92 1C for 30 s,

50 1C for 1 min and 65 1C for 8 min, before a final elongation

for 8 min at 65 1C. Five microlitres of amplified products were

separated on 1.5% (w/v) agarose gels in TAE buffer for 15 h at

40 V. A 1-kb DNA ladder (Invitrogen SARL, Cergy Pontoise,

France) was included in the gels as a size marker. The gels were

stained with ethidium bromide (1 mg mL�1) and DNA band-

ing patterns were visualized under UV (BIOCAPT software

version 99.0385, Vilbert Lourmat, Marne La Vallee, France).

BOX-PCR reactions were performed in triplicate to ensure that

only reliable bands were scored. Visual analysis was performed

to group together similar BOX-PCR fingerprints with subse-

quent further visual analysis allowing, for each BOX-PCR

fingerprint, establishment of a binary data matrix scoring 45

bands as being either present (1) or absent (0). A similarity

matrix was calculated using the pairwise Nei & Li (1979)

similarity coefficient, and an arbitrary similarity coefficient of

0.54 was chosen to delineate BOX types. Cluster analysis was

performed by the unweighted pair group method using

arithmetic averages (UPGMA).

The distributions of isolates from M. truncatula J5 and

TRV25 in these BOX types were compared using Pearson w2

test and the probability that the two distributions differed

significantly was calculated using the general Monte Carlo

algorithm. Computations were performed with the STATXACT

software version 3 (CYTEL Software Corporation, Cam-

bridge). BOX types explaining differences between the two

distributions were identified by analysis of contingency-

table cell contribution to the w2 value.

Results

Diversity of partial 16S rRNA gene--IGSsequences belonging to OTUs preferentiallyassociated with mycorrhizal roots ofM. truncatula

The diversity of the 164 partial sequences of 16S rRNA

gene–IGS described in Table 2 was characterized by phylo-

genetic analyses performed separately on the partial 16S

rRNA gene (Fig. 1a) and IGS (Fig. 1b) sequences. The partial

16S rRNA gene sequences analysed were distributed into

three monophyletic branches, supported by a significant

bootstrap value (100), corresponding to the Rubrivivax

subgroup, the Comamonadaceae and the Oxalobacteraceae

families, all belonging to Burkholderiales (Betaproteobacter-

ia) (Fig. 1a). The Rubrivivax subgroup branch encompassed

36 AM396923-type sequences, showing 97% similarity to

that of M. petroleiphilum (gi: 6063088) (Table 2); 32

originated from M. truncatula J5 and four from TRV25.

The Comamonadaceae branch encompassed nine

AM396923-type sequences, eight from J5 and one from

TRV25. Finally, the Oxalobacteraceae branch encompassed

all sequences (119) showing 99% similarity to C. fungivorans

(gi: 33145982) either associated with IGS (515 bp) corre-

sponding to AM396929-type sequences (95), 65 from J5 and

30 from TRV25, or with IGS (705 bp) corresponding to

AM396926-type sequences (24), 20 from J5 and four from

TRV25.

IGS sequences were distributed into four branches (W, X,

Y and Z) supported by a high bootstrap value (100)

(Fig. 1b). Thirty-one AM396923-type sequences were in-

cluded in branch W, 29 from M. truncatula J5 and two from

TRV25. Three AM396923-type sequences were included in

branch X: two from J5 and one from TRV25. Nine

AM396923-type sequences were included in branch Y: eight

from J5 and one from TRV25. Finally, 95 AM396929-type

sequences, 65 from J5 and 30 from TRV25, and 23

AM396926-type sequences, 20 from J5 and 3 from TRV25,

were included in branch Z (Fig. 1b). IGS branches W and

Fig. 1. Phylogenetic analyses of (i) 95 partial 16S rRNA gene–IGS sequences (980 bp) including partial 16S rRNA gene sequences having 99% similarity

to Collimonas fungivorans (gi: 33145982) (AM396929-type sequences, Table 3), (ii) 24 (1170 bp) including 16S rRNA gene having 99% similarity to C.

fungivorans (gi: 33145982) (AM396926-type sequences) and (iii) 45 (1170 bp) including 16S rRNA gene having 97% similarity to Methylibium

petroleiphilum (gi: 6063088) (AM396923-type sequences). Phylogenetic distances were determined by the neighbour-joining analysis using the Kimura

‘two-parameters’ correction with the pairwise gap removal option. The sequence types (AM396929, AM396926, and AM396923) are indicated in bold

and their numbers in brackets. Branching robustness is expressed as the percentage of reliability after 1000 bootstraps’ resampling (values 4 75 are

indicated). The scale bar represents the expected nucleotide replacement per site. (a) Relationships between partial 16S rRNA gene sequences (465 bp)

from (i) AM396929-, AM396926- and AM396923-type sequences and (ii) sequences extracted from GenBank of Burkholderiales (Betaproteobacteria)

belonging to Comamonadaceae and Oxalobacteraceae families and the Rubrivivax subgroup. (b) Relationships between IGS sequences from AM396929

(515 bp)-, AM396926 (705 bp)- and AM396923 (705 bp)-type sequences. Partial 16S rRNA gene sequences and corresponding IGS related to the Rubrivivax

subgroup, Comamonadaceae and Oxalobacteraceae are highlighted with and respectively. Sequences belonging to DNA extracted

previously from M. truncatula J5 and TRV25 roots (Offre et al., 2007) are indicated by and respectively.


184 P. Offre et al.

X included sequence types corresponding to partial 16S

rRNA gene sequences belonging to the Rubrivivax subgroup,

and Y and Z IGS branches to partial 16S rRNA gene

sequences belonging to Comamonadaceae and Oxalobacter-

aceae, respectively (Fig. 1a). Three sequences were not

clustered in any IGS branch (Fig. 1b): two AM396923-type

sequences corresponding to the 16S rRNA gene Rubrivivax

subgroup branch (Fig. 1a) and one AM396926-type

Aquaspirillum arcticum (gi: 21068882)Duganella zoogloeoides (gi: 1944498)

Sinobacter plicatus (gi: 62240118)

Aquaspirillum anulus (gi: 21068885)Giesbergeria kuznetsovii (gi: 60117041)Rhodoferax ferrireducens (gi: 74023592)

Variovorax paradoxus (gi: 82504534)Variovorax ginsengisoli (gi: 84453033)Variovorax dokdonensis (gi: 74423019)Curvibacter delicatus (gi: 3462592)Simplicispira metamorpha (gi: 3462593)Comamonas aquatica (gi: 20513173)

Rubrivivax benzolyticum (gi: 61657864)Methylibium petroleiphilum (gi: 6063088)Rubrivivax gelatinosus (gi: 303828)

Leptothrix mobilis (gi: 1263136)Ideonella dechloratans (gi: 577726)

Leptothrix discophora (gi: 501166)Sphaerotilus natans (gi: 47280)

100

100

100

0.01

Herbaspirillum putei (gi: 39652274)Herbaspirillum huttiense (gi: 4165372)

Collimonas fungivorans (gi: 33145982)Herminiimonas aquatilis (gi: 75754588)

Duganella violaceinigra (gi: 57996778)Massilia timonae (gi: 4204854)

Oxalobacteraceae

Comamonadaceae 100

100

100

100

0.1(a) (b)

AM396923-type (32)

AM396923-type (4)

AM396923-type (8)AM396923-type (1)

AM396929-type (65) &AM396926-type (20)

AM396929-type (30) &AM396926-type (4)

Rubrivivaxsub-group

W

X

Y

Z

AM396923-type (29)

AM396923-type (2)

AM396923-type (1)AM396923-type (2)

AM396923-type (1)AM396923-type (8)

AM396923-type (1)AM396923-type (1)

AM396929-type (65) &AM396926-type (20)

AM396929-type (30) &AM396926-type (3)

AM396926-type (1)



sequence corresponding to the 16S rRNA gene Oxalobacter-

aceae branch (Fig. 1a).

Root isolates belonging to Comamonadaceaeand Oxalobacteraceae

Bacterial isolates expected to be preferentially associated

with mycorrhizal roots were selected on the basis of their

affiliation to Comamonadaceae and Oxalobacteraceae by

dot-blot hybridization and by analysis of their partial 16S

rRNA gene–IGS sequences.

Attempts were made to prepare probes from an

AM396929-type sequence (Table 2) to target sequences

belonging to the Oxalobacteraceae branch (Fig. 1a) and from

two AM396923-type sequences to target those belonging to

the Rubrivivax subgroup and Comamonadaceae branches

(Fig. 1a). However, these attempts were only successful with

the AM396929-type sequence (Table 2). The specificity of the

AM396929-probe was assessed by recording (Fig. 2): (1) a

positive signal when hybridized onto: AM396929-type

sequences (1–3) and partial 16S rRNA gene–IGS from

C. fungivorans Ter331 (4), all belonging to Oxalobacteraceae;

AM396923-type sequences (1–3) and partial 16S rRNA

gene–IGS from Variovorax sp. MD1 (4), all belonging

to Comamonadaceae; one AM396923-type sequence (3)

belonging to the Rubrivivax subgroup; partial 16S rRNA

gene–IGS from A. tumefaciens C58 (1), B. japonicum

USDA110 (2) and S. meliloti ATCC 9930 (3), all belonging

to Alphaproteobacteria; partial 16S rRNA gene–IGS from X.

campestris pv. vesicatoria (3) belonging to Gammaproteobac-

teria; and partial 16S rRNA gene–IGS from S. epidermidis

RP62A (2), belonging to Firmicutes and (2) a negative signal

when hybridized onto two AM396923-type sequences (1–2)

belonging to the Rubrivivax subgroup, onto partial 16S rRNA

gene–IGS from P. fluorescens C7R12 (1) and E. coli S.17–1 (2),

all belonging to Gammaproteobacteria, and onto partial 16S

rRNA gene–IGS from Paenibacillus sp. strain B2 (1) and L.

innocua CIP 8012 (3), all belonging to Firmicutes. Taken

together, these observations indicate that the AM396929

probe was only partially specific because, besides sequences

from Oxalobacteraceae, a positive hybridization was recorded

with those from Comamonadaceae, with one from the

Rubrivivax subgroup and with those from Alphaproteobacter-

ia; a positive signal, although weaker, was also recorded

with one sequence from Gammaproteobacteria and from

Firmicutes.

Partial 16S rRNA gene–IGS sequences were successfully

amplified for 1518 of the 2412 isolates tested (Table 1). A

positive hybridization of partial 16S rRNA gene–IGS se-

quences with the AM396929 probe was obtained for 544

isolates (Table 1). Partial 16S rRNA gene–IGS was sequenced

for 480 of these isolates and 373 of them were analysed.

Taking into account the lack of specificity of the AM396929

probe, partial 16S rRNA gene–IGS of these 373 isolates was

compared with those available in GenBank. Partial 16S

rRNA gene–IGS of seven of these isolates showed a high

level of similarity (99%) to C. fungivorans (gi: 33145982)

belonging to Oxalobacteraceae, that of 166 isolates a high

Fig. 2. Dot-blot analyses of the specificity of the AM396929 probe. The

AM396929 probe was deposited onto (i) 10 ng of partial 16S rRNA

gene–IGS sequences analysed in this study (1–3) and Collimonas fungi-

vorans Ter 331 (4), all belonging to the Oxalobacteraceae, (ii) 10 ng of

partial 16S rRNA gene–IGS sequences analysed in this study (1–3) and of

Variovorax sp. DM1 (4), all belonging to Comamonadaceae, (iii) 10 ng of

partial 16S rRNA gene–IGS sequences analysed in this study (1–3)

belonging to the Rubrivivax subgroup, (iv) 10 ng of partial 16S rRNA

gene–IGS sequences of Agrobacterium tumefaciens C58 (1), Bradyrhi-

zobium japonicum USDA110 (2) and Sinorhizobium meliloti ATCC 9930

(3) all belonging to Alphaproteobacteria, (v) 10 ng of partial 16S rRNA

gene–IGS of Pseudomonas fluorescens C7R12 (1), Escherichia coli S.17-1

(2) and Xanthomonas campestris pv. vesicatoria (3) all belonging to

Gammaproteobacteria, and (vi) 10 ng of partial 16S rRNA gene–IGS

sequences of Paenibacillus sp. strain B2 (1), Staphylococcus epidermis

RP62A (2) and Listeria innocua CIP 8012 (3) all belonging to Firmicutes.

Fig. 3. Phylogenetic analyses of partial 16S rRNA gene–IGS sequences from 62 and 111 isolates from Medicago truncatula J5 ( ) and TRV25 ( ),

respectively. Phylogenetic distances were determined by neighbour-joining analysis using the Kimura ‘two-parameters’ correction with the pairwise gap

removal option. The sequence number for each M. truncatula genotype is indicated in brackets. Branching robustness is expressed as the percentage of

reliability after 1000 bootstraps’ resampling (values 4 75 were indicated). The scale bar represents the expected nucleotide replacement per site. (a)

Relationships between partial 16S rRNA gene sequences from isolates of the present study and extracted from GenBank of Burkholderiales

(Betaproteobacteria) belonging to Comamonadaceae and Oxalobacteraceae families. (b) Relationships between corresponding IGS sequences. Partial 16S

rRNA gene and corresponding IGS related to the Comamonadaceae and Oxalobacteraceae are highlighted with and respectively.


186 P. Offre et al.

level of similarity (97%) to M. petroleiphilum (gi: 6063088)

belonging to Comamonadaceae (Table 1), that of 135 isolates

a high level of similarity (4 97%) to Alphaproteobacteria,

that of 32 isolates a high level of similarity (4 97%) to

Betaproteobacteria other than Oxalobacteraceae and

Comamonadaceae, and finally that of 33 isolates a high

level of similarity (4 97%) to Gammaproteobacteria

(data not shown). These data are consistent with those of

the hybridization test performed with the AM396929-probe

and the reference strains of Oxalobacteraceae, Comamona-

daceae, Alphaproteobacteria and Gammaproteobacteria. In

contrast, the absence of sequences belonging to the

Massilia timonae (gi: 4204854)

Janthinobacterium lividum (gi: 2832894)

Aquaspirillum arcticum (gi: 21068882)Duganella zoogloeoides (gi: 1944498)

Sinobacter plicatus (gi: 62240118)

Herbaspirillum putei (gi: 39652274)Herbaspirillum huttiense (gi: 4165372)

Collimonas fungivorans (gi: 33145982)Herminiimonas aquatilis (gi: 75754588)

Duganella violaceinigra (gi: 57996778)Naxibacter alkalitolerans (gi: 50657801)

Variovorax paradoxus (gi: 82504534)

Variovorax ginsengisoli (gi: 84453033)Variovorax dokdonensis (gi: 74423019)

Ottowia thiooxydans (gi: 27848582)Curvibacter gracilis (gi: 39652273)

Ramlibacter tataouinensis (gi: 16332373)Xylophilus ampelinus (gi: 3462594)Hylemonella gracilis (gi: 31441942)

Polaromonas ginsengisoli (gi: 84453030)Acidovorax konjaci (gi: 3462596)

Diaphorobacter nitroreducens (gi: 18148396)Polaromonas aquatica (gi: 68051130)

Verminephrobacter eiseniae (gi: 84620107)Brachymonas petroleovorans (gi: 33415959)0.01

(29)

(21)

(36)

(2)(5)

Oxalobacteraceae

Comamonadaceae

0.1

(109)

(57)

(2)

(5)

(a)

(b)

100

87

100

(80)

100100

Z

Y



Rubrivivax subgroup and Firmicutes among those of the 373

tested isolates was not in accordance with hybridization of

the AM396929-probe with reference strains belonging to

these groups.

Phylogenetic diversity of partial 16S rRNAgene--IGS sequences from root isolates

The 173 partial 16S rRNA gene–IGS sequences of the isolates

with a high level of similarity to Oxalobacteraceae (7) and

Comamonadaceae (166) plus the most closely related se-

quences identified by BLAST analysis were analysed through

phylogenetic inference using the neighbour-joining method.

Phylogenetic analyses were performed separately on the

corresponding partial 16S rRNA gene (Fig. 3a) and IGS

(Fig. 3b) sequences.

The partial 16S rRNA gene sequences analysed were

distributed into two monophyletic branches, supported by

a significant bootstrap value (100), corresponding to the

Comamonadaceae and the Oxalobacteraceae families (Fig.

3a). The Comamonadaceae branch encompassed all the

sequences (166) showing 97% similarity to M. petroleiphi-

lum (gi: 6063088), 57 from M. truncatula J5 and 109 from

TRV25 (Fig. 3a). Within this branch, partial 16S rRNA gene

sequences of the 166 isolates were distributed into the same

subbranch, supported by a significant bootstrap (87), in-

cluding Variovorax paradoxus (gi: 82504534), Variovorax

ginsengisoli (gi: 84453033) and Variovorax dokdonensis (gi:

74423019). The Oxalobacteraceae branch encompassed all

the sequences (7) showing 99% similarity to C. fungivorans

(gi: 33145982), five from J5 and two from TRV25 (Fig. 3a).

IGS sequences were distributed into two branches (Y and

Z) supported by a high bootstrap value (100) (Fig. 3b). All

the sequences (166) showing 97% similarity to M. petrolei-

philum (gi: 6063088), 57 from M. truncatula J5 and 109

from TRV25, were included in branch Y (Fig. 3b). All the

sequences (7) showing 99% similarity to C. fungivorans (gi:

33145982), five from M. truncatula J5 and two from TRV25,

were included in branch Z (Fig. 3b). IGS branches Y and Z

included sequences corresponding to partial 16S rRNA gene

sequences belonging to Comamonadaceae and Oxalobacter-

aceae, respectively (Fig. 3a).

Genetic diversity of root isolates based onBOX-PCR

The genetic diversity of the 173 root isolates was further

analysed by BOX-PCR (Fig. 4). Amplification products

yielded complex genomic fingerprints consisting of

fragments ranging in size from 506 to 4072 bp (data not

shown).

Twenty-five BOX types (indicated by letters A–Y), en-

compassing the majority (74.1%) of the isolates, were

delineated. All isolates distributed into BOX types belonged

to Comamonadaceae and to IGS Y branches (Fig. 3), whereas

the remaining isolates belonged to the Comamonadaceae

and Oxalobacteraceae branches and to the IGS Y and Z

branches (Fig. 3). The distributions of the isolates from

M. truncatula J5 and from TRV25 in the BOX types differed

significantly (Po 0.0001), this difference being significantly

explained by types (M, T, U and V) that included only

isolates from M. truncatula J5. Other types (E, G, H, O and

X) only included isolates from TRV25 (Fig. 4).

The partial 16S rRNA gene sequences of a representative

isolate per BOX-type were further subjected to a phyloge-

netic analysis (data not shown). The corresponding

results indicated that isolates belonging to BOX types C, G,

I, K, L, M, N, O, P, Q, R and T and those belonging to

BOX types A, B, D, E, F, H, J, S, U, V, W, X and Y

were grouped into the branches defined in Fig. 3 that

include V. ginsengisoli (gi: 84453033) and V. paradoxus

(gi: 82504534), respectively.

Discussion

The diversity of bacteria from mycorrhizal roots of M.

truncatula wild-type and from roots of a mutant impaired

in the ability to establish AM was compared, both plant

genotypes being cultivated in a fallow soil. This comparison

relied on two complementary strategies based on the diver-

sity analysis of DNA extracted from roots and from bacterial

isolates. The first consisted of phylogenetic analysis of

sequences of partial 16S rRNA gene–IGS, amplified and

cloned from DNA extracted from roots, belonging to OTUs

preferentially associated with mycorrhizal roots (Offre et al.,

2007). Phylogenetic analysis of partial 16S rRNA gene

sequences of AM396923, AM396926 and AM396929 types

indicated that all were affiliated to Burkholderiales

Num

ber

of is

olat

es

0A

5

10

15

20

25

BOX types

YB C D E F G H I J K L M N O P Q R S T U V W X

M. truncatula J5 (Myc / Nod )

M. truncatula TRV25 (Myc /Nod )

Fig. 4. Frequency distributions of bacterial isolates from Medicago

truncatula J5 and TRV25 in the different BOX types. The two distributions

are significantly different (w2, exact P-valueo 0.0001). This distribution

difference is explained by underlined BOX types as identified by analysis

of contingency-table cell contribution to the w2 value.


188 P. Offre et al.

(Betaproteobacteria) (Fig. 1a). They were grouped into three

clusters belonging to the Comamonadaceae, Oxalobactera-

ceae and Rubrivivax subgroup, respectively (Fig. 1a). Dis-

crimination of the sequences from M. truncatula J5 and

from the mutant TRV25 showed that most of the sequences of

the Comamonadaceae (eight out of nine), Oxalobacteraceae

(85 out of 119) and Rubrivivax subgroup (32 out of 36) were

associated with the mycorrhizal roots. These data support our

previous conclusions establishing that these groups explained

variations in genetic structure of the bacterial communities

associated with mycorrhizal and nonmycorrhizal roots (Offre

et al., 2007). To our knowledge, this is the first report

describing bacteria from the Comamonadaceae, Oxalobacter-

aceae and Rubrivivax subgroup as being preferentially asso-

ciated with AM. The phylogeny of IGS from AM396923,

AM396926 and AM396929–type sequences was fully congru-

ent with that of partial 16S rRNA genes for Comamonadaceae

and Oxalobacteraceae (Fig. 1a and b), whereas IGS sequences

belonging to the Rubrivivax subgroup were clustered into

two branches, one including only three sequences. This is

consistent with the fact that IGS sequences are known to

be more variable than 16S rRNA gene sequences (Gurtler &

Stanisich, 1996).

The second strategy was based on characterization of

Burkholderiales isolates from a large collection of isolates

(2412) that gave a positive hybridization with the polynu-

cleotide AM396929-probe. We chose to make a polynucleo-

tide rather than an oligonucleotide probe to avoid

false negatives (Amann & Ludwig, 2000). The difficulty

experienced in raising AM396923-probes targeting the

Comamonadaceae and Rubrivivax subgroup may be related

to the larger size of the corresponding sequence types

(1170 bp) than that of the AM396929-probe targeting

Oxalobacteraceae (980 bp). The specificity test indicated

that, unexpectedly, the AM396929-probe not only targeted

Oxalobacteraceae but also the Comamonadaceae, Alphapro-

teobacteria and, to a lesser extent, the Rubrivivax subgroup,

Gammaproteobacteria and Firmicutes (Fig. 2). This could be

related to the low specificity of the AM396929-probe.

Results of the specificity test of the AM396929-probe are in

agreement with the assignment of the 373 strains, showing a

positive signal with the AM396929-probe, made on the basis

of their partial 16S rRNA gene sequences to the following

groups: Oxalobacteraceae (seven isolates), Comamonadaceae

(166 isolates), Alphaproteobacteria (135 isolates), Betapro-

teobacteria (32 isolates) (other than Oxalobacteraceae and

Comamonadaceae) and Gammaproteobacteria (33 isolates).

However, the absence of sequences belonging to the Rubri-

vivax subgroup and Firmicutes among those of the 373

isolates was not in accordance with positive hybridization

of a reference strain of the Rubrivivax subgroup and one of

Firmicutes. This discrepancy could be related to the fact that

we had access only to culturable bacteria and to a limited

number of isolates. Only the sequences shown to have a high

level of similarity to those of Oxalobacteraceae or Comamo-

nadaceae were subjected to the phylogenetic analysis. Phy-

logenetic analyses of the corresponding sequences

confirmed that 166 of the isolates belonged to Comamona-

daceae (Fig. 3a) and seven to Oxalobacteraceae (Fig. 3a). The

low abundance of bacteria belonging to Oxalobacteraceae

compared with the high frequency of the corresponding

sequence in DNA directly extracted from roots might be

related to the difficulty in growing the corresponding

bacteria. Indeed, among this family, C. fungivorans is known

to be chitinolytic (de Boer et al., 2004) and therefore a

growth medium including chitin would have increased the

probability of isolating such bacteria. Despite their small

number, Oxalobacteraceae isolates were more abundant in

mycorrhizal (five) than in nonmycorrhizal (two) roots,

supporting conclusions made from analysis of DNA directly

extracted from roots showing a higher frequency of

AM396929 (65 out of 119) and AM396926-type (20 out of

119) sequences in mycorrhizal than in nonmycorrhizal roots

(30 and 4 out of 119, respectively) (Fig. 1). Conversely,

bacteria belonging to Comamonadaceae were mostly iso-

lated from nonmycorrhizal (109) than from mycorrhizal

roots (57), contrasting with data obtained from DNA

directly extracted from roots, which showed a higher

frequency of AM396923-type sequences in mycorrhizal

(eight out of nine) than in nonmycorrhizal (one out of

nine) roots (Fig. 1). This discrepancy could be explained by

assessment of the 16S rRNA gene–IGS sequence-type fre-

quency, which was based on similar amounts of PCR

products for A-RISA analysis (Ranjard et al., 2003) for both

plant genotypes, whereas bacterial abundance based on

enumeration of total bacterial cells with orange acridine

(Hobbie et al., 1977) was higher in nonmycorrhizal than in

mycorrhizal roots (data not shown), in agreement with the

higher release of photosynthates in nonmycorrhizal roots

(Bago et al., 2000; Johnson et al., 2002).

The genetic diversity of Comamonadaceae isolates asso-

ciated with mycorrhizal and nonmycorrhizal roots differed

significantly as indicated by their different distribution in

the BOX types (Fig. 4). This difference was explained by

BOX types only including isolates from mycorrhizal roots,

suggesting that the corresponding populations were prefer-

entially associated with AM. Root colonization by fungi,

including AM, has been described as providing ecological

niches favourable for specific bacterial populations

(Andrade et al., 1997; Johansson et al., 2004; de Boer et al.,

2005; Toljander et al., 2007). The preferential association

between Comamonadaceae and Oxalobacteraceae and my-

corrhizal roots has already been described for ectomycorrhi-

zas of Scots pine: Suillus flavilus (Izumi et al., 2006),

Tomentella (Khetmalas et al., 2002) and Tuber borchii (Bar-

bieri et al., 2005) ectomycorrhizal fungi. In addition,



Collimonas (Oxalobacteraceae), previously shown to grow at

the expense of living fungi (de Boer et al., 2004, 2005), was

found to be associated with lichen symbiosis in forest soils

(Mannisto & Haggblom, 2006) and with the oak-Scleroderma

citrinum ectomycorrhizosphere (Uroz et al., 2007).

The preferential association of Comamonadaceae and Ox-

alobacteraceae with AM suggests that AM fungi are favourable

for the survival of these bacteria and reciprocally that, from an

evolution point of view, these bacteria are beneficial to

mycorrhization. This hypothesis is supported by the demon-

stration made by Frey et al. (1997) of the promoting effect on

ectomycorrhization of a bacterial isolate representative of the

bacterial community associated with the mycorrhizosphere of

Laccaria bicolor. Intimate associations between Glomerybacter

bacteria, belonging to Burkholderiales, and various AM fungi

(Gigaspora margarita, Scutellospora persica and Scutellospora

castanea) were also demonstrated, these bacteria being con-

sidered as endosymbionts of AM fungi (Bianciotto et al.,

2003). However, one may not exclude the possibility that some

bacterial populations preferentially associated with mycorrhi-

zal roots may also live at the expense of AM fungi. Among

Oxalobacteraceae, C. fungivorans has been shown to grow at

the expense of soil fungi (de Boer et al., 2004) and Janthino-

bacterium agaricidamnosus to be pathogenic of mushroom

(Lincoln et al., 1999).

The present study has advanced our knowledge of the

diversity of Burkholderiales associated with endomycorrhizal

roots by following an approach based jointly on the diversity

analysis of DNA extracted from roots and from bacterial

isolates. Analysis of root DNA confirmed the preferential

association of Burkholderiales bacteria with mycorrhizal

roots reported previously (Offre et al., 2007). We have

further specified that they belong to the Comamonadaceae

and Oxalobacteraceae families and to the Rubrivivax sub-

group. The results obtained by the culturable approach

suggest that the Oxalobacteraceae are preferentially asso-

ciated with endomycorrhizal roots. This could be further

supported by the RNA-stable isotope probing approach

applied by Vandenkoornhuyse et al. (2007), allowing the

identification of bacterial populations utilizing carbon sub-

strates released by mycorrhizal plant roots.

A discrepancy was found between the molecular and

cultivation-based approaches for Comamonadaceae and

could possibly be related to the different quantitative types

of information yielded. However, BOX-PCR analysis of

these isolates, which provided higher resolution analysis of

the diversity of bacterial isolates than that based on partial

16S rRNA gene sequences, indicated that isolates from

mycorrhizal and nonmycorrhizal roots differed. Comamo-

nadaceae isolates explaining these differences and Oxalobac-

teraceae isolates from mycorrhizal roots are currently being

tested for their possible beneficial effect on mycorrhizal

colonization of roots.

Acknowledgements

This study was supported by programs funded by INRA

Sante des Plantes et Environnement (Rhizosphere Ecology

of Annual Medics program) and the Burgundy regional

project (02511CPO2S188) and a doctoral fellowship to P.O.

B.P. was supported by a doctoral fellowship from Italian

MIUR and a Vinci project grant. The authors are grateful to

D. Pouhair and S. Leclercq for technical assistance, to G. Duc

(UMR LEG, Dijon, France) for providing seeds of M.

truncatula and to W. de Boer (NIOO-KNAW, Centre for

Terrestrial Ecology; Heteren, the Netherlands) and M.

Devers (UMR MSE, Dijon, France) for providing C. fungi-

vorans Ter331 and Variovorax sp. MD1 strains, respectively.

Authors’contribution

P.O. and B.P. contributed equally to this study.

References

Altschul SF, Gish W, Miller W, Myers EW & Lipman DJ (1990)

Basic local alignment search tool. J Mol Biol 215: 403–410.

Amann R & Ludwig W (2000) Ribosomal RNA-targeted nucleic

acid probes for studies in microbial ecology. FEMS Microbiol

Rev 24: 555–565.

Andrade G, Mihara KL, Linderman RG & Bethlenfalvay GJ

(1997) Bacteria from rhizosphere and hyphosphere soils of

different arbuscular-mycorrhizal fungi. Plant Soil 192: 71–79.

Artursson V, Finlay RD & Jansson JK (2005) Combined

bromodeoxyuridine immunocapture and terminal-restriction

fragment length polymorphism analysis highlights differences

in the active soil bacterial metagenome due to Glomus mosseae

inoculation or plant species. Environ Microbiol 7: 1952–1966.

Bago B, Pfeffer PE & Sachar-Hill Y (2000) Carbon metabolism

and transport in arbuscular mycorrhizas. Plant Physiol 124:

949–957.

Barbieri E, Bertini L, Rossi I, Ceccaroli P, Saltarelli R, Guidi C,

Zambonelli A & Stocchi V (2005) New evidence for bacterial

diversity in the ascoma of the ectomycorrhizal fungus Tuber

borchii Vittad. FEMS Microbiol Lett 247: 23–35.

Bever JD, Pringle A & Schultz PA (2002) Dynamics within the

plant-arbuscular mycorrhizal fungal mutualism: testing the

nature of community feedback. Mycorrhizal Ecology

(van der Heijden MGA & Sanders IE, eds), pp. 267–292.

Springer-Verlag, Berlin.

Bianciotto V, Lumini E, Bonfante P & Vandamme P (2003)

‘Candidatus Glomeribacter gigasporarum’ gen. nov., sp. nov.,

an endosymbiont of arbuscular mycorrhizal fungi. Int J Syst

Evol Micr 53: 121–124.

Brundrett MC (2002) Coevolution of roots and mycorrhizas of

land plants. New Phytol 154: 275–304.

Budi SW, van Tuinen D, Martinotti MG & Gianinazzi S (1999)

Isolation from the Sorghum bicolour mycorrhizosphere of a

bacterium compatible with arbuscular mycorrhiza


190 P. Offre et al.

development and antagonistic towards soilborne fungal

pathogens. Appl Environ Microbiol 65: 5148–5150.

de Boer W, Leveau JHJ, Kowalchuk GA, Klein Gunnewiek PJA,

Abeln ECA, Figge MJ, Sjollema K, Janse JD & van Veen JA

(2004) Collimonas fungivorans gen. nov., sp. nov., a chitinolytic

soil bacterium with the ability to grow on living fungal hyphae.

Int J Syst Evol Microbiol 54: 857–864.

de Boer W, Folman LB, Summerbell RC & Boddy L (2005) Living

in a fungal world: impact of fungi on soil bacterial niche

development. FEMS Microbiol Rev 29: 795–811.

Devers M, Henry S, Hartmann A & Martin-Laurent F (2005)

Horizontal gene transfer of atrazine-degrading genes (atz)

from Agrobacterium tumefaciens St96-4 pADP1: Tn5 to

bacteria of maize cultivated soil. Pest Manag Sci 61: 870–880.

Eparvier A, Lemanceau P & Alabouvette C (1991) Population

dynamics of non-pathogenic Fusarium and fluorescent

Pseudomonas strains in rockwool, a substratum for soilless

culture. FEMS Microbiol Ecol 86: 177–184.

Forbes BA, Sahm DE & Weissfeld AS (1998) Bailey & Scott’s

Diagnostic Microbiology, 10th edn. Mosby Inc., St Louis.

Frey P, Frey-Klett P, Garbaye J, Berge O & Heulin T (1997)

Metabolic and genotypic fingerprinting of fluorescent

pseudomonads associated with Douglas fir Laccaria bicolor

mycorrhizosphere. Appl Environ Microbiol 63: 1852–1860.

Galibert F, Finan TM, Long SR et al. (2001) The composite

genome of the legume symbiont Sinorhizobium meliloti.

Science 293: 668–672.

Galtier N, Gouy M & Gautier C (1996) SeaView and Phylo_win,

two graphic tools for sequence alignment and molecular

phylogeny. Comput Appl Biosci 12: 543–548.

Gill SR, Fouts DE, Archer GL et al. (2005) Insights on evolution

of virulence and resistance from the complete genome analysis

of an early methicillin-resistant Staphylococcus aureus strain

and a biofilm-producing methicillin-resistant Staphylococcus

epidermidis strain. J Bacteriol 187: 2426–2438.

Goodner B, Hinkle G, Gattung S et al. (2001) Genome sequence

of the plant pathogen and biotechnology agent Agrobacterium

tumefaciens C58. Science 294: 2323–2328.

Gurtler V & Stanisich VA (1996) New approaches to typing and

identification of bacteria using the 16S–23S rDNA spacer

region. Microbiology 142: 3–16.

Hobbie JE, Daley RJ & Jasper S (1977) Use of nucleopore filters

for counting bacteria by epifluorescence microscopy. Appl

Environ Microbiol 33: 1225–1228.

Izumi H, Anderson IC, Alexander IJ, Killham K & Moore ERB

(2006) Endobacteria in some endomycorrhiza of Scots pine

(Pinus sylvestris). FEMS Microbiol Ecol 56: 34–43.

Johansson JF, Paul LR & Finlay RD (2004) Microbial interactions

in the mycorrhizosphere and their significance for sustainable

agriculture. FEMS Microbiol Ecol 48: 1–13.

Johnson D, Leake JR, Ostle N, Ineson P & Read DJ (2002) In situ

(CO2)-13C pulse labelling of upland grassland demonstrates a

rapid pathway of carbon flux from arbuscular mycorrhizal

mycelia to the soil. New Phytol 153: 327–334.

Kaneko T, Nakamura Y, Sato S et al. (2002) Complete genomic

sequence of nitrogen-fixing symbiotic bacterium

Bradyrhizobium japonicum USDA110. DNA Res 9: 189–197.

Khetmalas MB, Egger KN, Massicotte HB, Tackaberry LE &

Clapperton MJ (2002) Bacterial diversity associated with

subalpine fir (Abies lasiocarpa) ectomycorrhizae following

wildfire and salvage-logging in central British Columbia. Can J

Microbiol 48: 611–625.

Kimura M (1980) A simple model for estimating evolutionary

rates of base substitutions through comparative studies of

nucleotide sequences. J Mol Evol 16: 111–120.

King EO, Ward MK & Raney DE (1954) Two simple media for the

demonstration of pyocyanin and fluorescein. J Lab Clin Med

44: 301–307.

Klironomos JN, McCune J, Hart M & Neville J (2000) The

influence of arbuscular mycorrhizae on the relationship

between plant diversity and productivity. Ecol Lett 3: 137–141.

Lincoln SP, Fermor TR & Tindall BJ (1999) Janthinobacterium

agaricidamnosum sp. nov., a soft rot pathogen of Agaricus

bisporus. Int J Syst Evol Microbiol 49: 1577–1589.

Mannisto MK & Haggblom MM (2006) Characterization of

psychrotolerant heterotrophic bacteria from Finnish Lapland.

Syst Appl Microbiol 29: 229–243.

Marschner P, Crowley DE & Lieberei R (2001) Arbuscular

mycorrhizal infection changes the bacterial 16S rDNA

community composition in the rhizosphere of maize.

Mycorrhiza 11: 297–302.

Mazurier S, Lemunier M, Hartmann A, Siblot S & Lemanceau P

(2006) Conservation of type III secretion system genes in

Bradyrhizobium isolated from soybean. FEMS Microbiol Lett

259: 317–325.

Meyer JR & Linderman RG (1986) Selective influence on

populations of rhizosphere or rhizoplane bacteria and

actinomycetes by mycorrhizas formed by Glomus fascicolatum.

Soil Biol Biochem 18: 191–196.

Mougel C, Offre P, Ranjard L, Corberand T, Gamalero E, Robin C

& Lemanceau P (2006) Dynamic of the genetic structure of

bacterial and fungal communities at different development

stages of Medicago truncatula Gaertn. cv. Jemalong line J5. New

Phytol 170: 165–175.

Nei M & Li W (1979) Mathematical model for studying genetic

variation in terms of restriction endonucleases. Proc Natl Acad

Sci USA 76: 5269–5273.

Normand P, Ponsonnet C, Nesme X, Neyra M & Simonet P

(1996) ITS analysis of prokaryotes. Molecular Microbial

Ecology Manual (Akkermans DL, van Elsas JD &

de Bruijn EI, eds), pp. 1–12. Kluwer Academic Publishers,

Amsterdam.

Offre P, Pivato B, Siblot S, Gamalero E, Corberand T, Lemanceau

P & Mougel C (2007) Identification of bacterial groups

preferentially associated with mycorrhizal roots of Medicago

truncatula. Appl Environ Microbiol 73: 913–921.

Perriere G & Gouy M (1996) WWW-Query: an on-line retrieval

system for biological sequence banks. Biochimie 78: 364–369.



Pivato B, Mazurier S, Lemanceau P, Siblot S, Berta G, Mougel C &

van Tuinen D (2007) Medicago species affect the community

composition of arbuscular mycorrhizal fungi associated with

roots. New Phytol 176: 197–210.

Ranjard L, Lejon DPH, Mougel C, Schehrer L, Merdinoglu D &

Chaussod R (2003) Sampling strategy in molecular microbial

ecology: influence of soil sample size on DNA fingerprinting

analysis of fungal and bacterial communities. Environ

Microbiol 5: 1111–1120.

Remy W, Taylor TN, Hass H & Kerp H (1994) Four hundred-

million-year-old vesicular arbuscular mycorrhizae. Proc Natl

Acad Sci USA 91: 11841–11843.

Roccourt J, Grimont F, Grimont PAD & Seeliger HPR (1982)

DNA relatedness among serovars of Listeria monocytogenes

sensu lato. Curr Microbiol 7: 383–388.

Sagan M, Morandi E, Tarenghi E & Duc G (1995) Selection of

nodulation and mycorrhizal mutants in the model plant

Medicago truncatula (Gaertn.) after g-ray mutagenesis. Plant Sci

111: 63–71.

Saitou N & Nei M (1987) The neighbor-joining method: a new

method for reconstructing phylogenetic trees. Mol Biol Evol 4:

406–425.

Schußler A, Schwarzott D & Walker C (2001) A new fungal

phylum, the Glomeromycota: phylogeny and evolution. Mycol

Res 105: 1413–1421.

Simon R, Prifer U & Puhler A (1983) A broad range mobilization

system for in vivo genetic engineering: transposon mutagenesis

in Gram-negative bacteria. Biotechnology 1: 784–791.

Smith SE & Read DJ (1997) Mycorrhizal Symbiosis. Academic

Press, London.

Stall RE, Loschke DC & Jones JB (1986) Linkage of copper

resistance and avirulence loco on a self-transmissible plasmid

in Xanthomonas camprestris pv. vesicatoria. Phytopathology 76:

240–243.

Streitwolf-Engel R, Boller T, Wiemken A & Sanders IR (1997)

Clonal growth traits of two Prunella species are determined by

co-occurring arbuscular mycorrhizal fungi from a calcareous

grassland. J Ecol 85: 181–191.

Thompson JD, Higgins DG & Gibson TJ (1994) CLUSTAL W:

improving the sensitivity of progressive multiple sequence

alignment through sequence weighting, positions-specific gap

penalties and weight matrix choice. Nucleic Acids Res 22:

4673–4680.

Toljander JF, Bjorn DL, Paul LR, Elfstrand M & Finlay RD (2007)

Influence of arbuscular mycorrhizal mycelial exudates on soil

bacterial growth and community structure. FEMS Microbiol

Ecol 61: 295–304.

Uroz S, Calvaruso C, Turpauld MP, Pierrat JC, Mustin C

& Frey-Klett P (2007) Effect of the mycorrhizosphere

on the genotypic and metabolic diversity of the

bacterial communities involved in mineral

weathering in a forest soil. Appl Environ Microbiol 73:

3019–3027.

Vandenkoornhuyse P, Ridgway P, Watson IJ, Fitter AH & Young

JPW (2003) Co-existing grass species have distinctive

arbuscular mycorrhizal communities. Mol Ecol 12:

3085–3095.

Vandenkoornhuyse P, Mahe S, Ineson P, Staddon P, Ostle N,

Cliquet J-B, Francez A-J, Fitter AH & Youg JPW (2007) Active

root-inhabiting microbes identified by rapid incorporation of

plant-derived carbon into RNA. Proc Natl Acad Sci USA 104:

16970–16975.

van der Heijden MGA, Klironomos JN, Ursic M, Moutioglis P,

Streitwolf-Engel R, Boller T, Wiemken A & Sanders IR (1998)

Mycorrhizal fungal diversity determines plant biodiversity,

ecosystem variability and productivity. Nature 396:

69–72.

Versalovic J, Schneider M, de Bruijn FJ & Lupski JR (1994)

Genomic fingerprinting of bacteria using repetitive sequence-

based polymerase chain reaction. Meth Mol Cell Biol 5:

25–40.

Wang B & Qiu YL (2006) Phylogenetic distribution and evolution

of mycorrhizas in land plants. Mycorrhiza 16: 299–363.


192 P. Offre et al.

Documents

Microdiversity of Burkholderiales associated with mycorrhizal and nonmycorrhizal roots of Medicago truncatula