Skip to content


BMC Genomics

Open Access

Transcriptomic profiling of Burkholderia phymatum STM815, Cupriavidus taiwanensis LMG19424 and Rhizobium mesoamericanum STM3625 in response to Mimosa pudica root exudates illuminates the molecular basis of their nodulation competitiveness and symbiotic evolutionary history

Contributed equally
BMC Genomics201819:105

Received: 21 September 2017

Accepted: 17 January 2018

Published: 30 January 2018



Rhizobial symbionts belong to the classes Alphaproteobacteria and Betaproteobacteria (called “alpha” and “beta”-rhizobia). Most knowledge on the genetic basis of symbiosis is based on model strains belonging to alpha-rhizobia. Mimosa pudica is a legume that offers an excellent opportunity to study the adaptation toward symbiotic nitrogen fixation in beta-rhizobia compared to alpha-rhizobia. In a previous study (Melkonian et al., Environ Microbiol 16:2099–111, 2014) we described the symbiotic competitiveness of M. pudica symbionts belonging to Burkholderia, Cupriavidus and Rhizobium species.


In this article we present a comparative analysis of the transcriptomes (by RNAseq) of B. phymatum STM815 (BP), C. taiwanensis LMG19424 (CT) and R. mesoamericanum STM3625 (RM) in conditions mimicking the early steps of symbiosis (i.e. perception of root exudates). BP exhibited the strongest transcriptome shift both quantitatively and qualitatively, which mirrors its high competitiveness in the early steps of symbiosis and its ancient evolutionary history as a symbiont, while CT had a minimal response which correlates with its status as a younger symbiont (probably via acquisition of symbiotic genes from a Burkholderia ancestor) and RM had a typical response of Alphaproteobacterial rhizospheric bacteria. Interestingly, the upregulation of nodulation genes was the only common response among the three strains; the exception was an up-regulated gene encoding a putative fatty acid hydroxylase, which appears to be a novel symbiotic gene specific to Mimosa symbionts.


The transcriptional response to root exudates was correlated to each strain nodulation competitiveness, with Burkholderia phymatum appearing as the best specialised symbiont of Mimosa pudica.




Rhizobia are a functional group of bacteria able to develop a nitrogen-fixing symbiosis with legumes. Symbiotic associations between legumes and rhizobia usually begin with an exchange of diffusible signals among which (iso)flavonoids present in the root exudates are responsible for the production and secretion of bacterial lipochitooligosaccharides (LCOs), known as Nod factors (NF). The recognition of NF by the roots leads to the formation of the root nodule, wherein bacteria are able to fix atmospheric nitrogen to sustain plant growth in exchange for carbon –containing metabolites [1]. The biosynthesis of NF and nitrogen fixation enzymes are encoded by nod and nif genes respectively [2].

Rhizobial species are mainly distributed among four genera of Alphaproteobacteria (Rhizobium, Sinorhizobium, Mesorhizobium, Bradyrhizobium; and a few species in seven other genera) and two genera of Betaproteobacteria (Burkholderia, Cupriavidus) leading to the proposal of the terms alpha- and beta-rhizobia to distinguish both classes of symbionts [3]. The discovery of beta-rhizobia in 2001 changed dramatically our perspectives on the genetic backgrounds required to host the ability to form symbioses with legumes by extending this possibility to a new subclass of Proteobacteria. If beta-rhizobia appeared controversial at the time of their discovery, it is now well established that Burkholderia species are the main symbionts of several endemic species of the Mimosoideae and Papilinoideae legume subfamilies in Brazil and South Africa [47], which raises questions about the origin of symbiosis and performance of these new symbionts compared to alpha-rhizobia. Many symbiotic species of Burkholderia (recently rearranged in the Paraburkholderia genus [8]) have been described from mimosoids: B. phymatum [9], B. tuberum [10], B. mimosarum [11], B. nodosa [12, 13], B. sabiae [14], B. symbiotica [15], B. diazotrophica [16], B. piptadeniae and B. ribeironis [17], while Cupriavidus symbionts (C. taiwanensis, C. necator, C. pinatubonensis) were detected mostly in invasive Mimosa species (M. pudica, M. pigra, M. diplotricha) in South, Central and North America [10, 18, 19], in Asia (China, India, Taiwan, Philippines, New Caledonia, Papua New Guinea) [2024], but also in native Mimosa and Parapiptadenia species in Uruguay [25, 26].

Diversity studies of alpha-rhizobia showed incongruence between rhizobial house-keeping genes and symbiotic nod genes phylogenies. Phylogenetic analyses have shown that the different rhizobial lineages diverged well before the appearance of legumes [27], so that their bacterial nodulation capacity would have been acquired substantially after the diversification of rhizobia in one or several lineages, and was then spread by lateral gene transfer among the different genera and species [28, 29]. For the Mimosoid Burkholderia legume symbionts, the scenario seems to be different as their nod gene phylogenies mirror that of the Burkholderia species (based on their 16S rRNA and housekeeping genes), which has led several authors to hypothesise an ancient transfer of nodulation genes to a burkholderia ancestor followed by vertical transmission and coevolution with endemic host species [30]. In the case of Cupriavidus, the nodulating species were proposed to be recent symbionts compared to Burkholderia based on phylogenetic and genomic analyses [31], but in the light of recent findings and analyses (nodC gene phylogeny, symbiotic specificity) its origin and evolution may be more complex [5].

Mimosa pudica is a legume that has the rare ability to be nodulated naturally by both alpha (Rhizobium spp.) and beta-rhizobia (Burkholderia, Cupriavidus). Diversity studies showed that proportions between beta- and alpha-rhizobia hosted in nodules of M. pudica vary significantly in different geographical locations. For example, Burkholderia were found to be predominant in nodules of M. pudica in Barro Colorado Island and in French Guiana [10, 32], whereas the proportions of Burkholderia and Cupriavidus symbionts were more equal in Costa Rica, China and India [18, 20, 24], and finally Cupriavidus was found to be predominant in nodules of M. pudica in Taiwan and in New Caledonia [23, 33]. Several parameters can help explain the proportions of symbionts in nodules of M. pudica in worldwide diversity studies, but soil characteristics, especially the presence of combined nitrogen affect strain competitiveness [34]. In a previous study [35], we made a large survey of strain nodulation competitiveness between Burkholderia, Cupriavidus and Rhizobium species and showed that B. phymatum and B. tuberum sv. mimosae are the most competitive species on Mimosa pudica, but Cupriavidus taiwanensis could compete with Burkholderia when tested on specific M. pudica genotypes (on variety unijuga from Taiwan where Cupriavidus dominates in nodules of this variety). In the case of Rhizobium species, all species were poorly competitive, despite the fact that R. mesoamericanum are frequently found in diversity studies, although in small proportions, such as in Costa Rica [18], Mexico [36], New Caledonia [23] and French Guiana [10], which raises questions about their capacity to be maintained as M. pudica symbionts. The only case of a Rhizobium species dominating in the nodules of M. pudica is R. altiplani in central Brazil [37].

Several molecular factors linked to nodulation competitiveness have been described in rhizobia, including i) the nodulation gene set [38], ii) rhizobial motility [39] and specific adhesiveness to the plant roots [40], iii) adjustment of rhizobial cell-surface characteristics (EPS and LPS) [41], iv) induction of type III or type IV secretion system expression [42] and v) induction of pSym conjugal transfer genes [43]. Until now, all these studies have been performed on alpha rhizobia, which leaves many questions about the specific adaptations that might be found in beta-rhizobia which could explain their competitiveness superiority over alpha-rhizobia in Mimosa species. Beyond the activation of nod and nif genes, no studies have been so far conducted concerning the mechanisms of competition for nodulation or on the symbiotic process at the molecular level in beta-rhizobia.

In this study, we investigated the molecular bases of competitiveness and symbiosis between three representative and well-studied M. pudica-nodulating symbiotic strains of Burkholderia (B. phymatum STM815), Cupriavidus (C. taiwanensis LMG19424) and Rhizobium (R. mesoamericanum STM3625) by studying their transcriptomic response to root exudates as a mimicking condition of early symbiotic events. Our aim was to analyse the different molecular strategies of each strain (alpha- or beta-rhizobial), to nodulate Mimosa pudica in order i) to reveal commonalities and differences in the molecular dialogues of alpha and beta-rhizobia nodulating the same host, and ii) to reveal the factors underpinning the superiority of B. phymatum in terms of competitiveness compared to the other two symbionts.

Results & discussion

Symbiotic and genomic features of the three model rhizobia

Three symbionts of M. pudica belonging either to alpha-rhizobia (R. mesoamericanum STM3625 (RM)) or to beta-rhizobia (B. phymatum STM815 (BP) and C. taiwanensis LMG19424 (CT)) were used in this transcriptomic study. The symbiotic features (efficiency, nodulation competitiveness) and genome characteristics of each strain are given in Table 1. BP is characterized by a larger genome compared to the two other rhizobia (with an additional 2.2 Mb). The number of chromosomes and plasmids varies among the strains, but all harbor a symbiotic plasmid of similar size (500 to 600 kb). The core genome is composed of 1687 genes (Additional file 1: Figure S1), and, as expected, BP and CT have more additional orthologs in common with each other (1466 among the two Betaproteobacteria) than with RM. BP has 4651 specific genes compared to the other two strains. The symbiotic plasmids in all three rhizobia shared poor syntenies (Additional file 1: Figure S2), with almost only the nod and nif gene regions being syntenic.
Table 1

Strains with symbiotic and genomic features


Symbiotic features

G size (Mb)

Chromosomes (Mb)

Plasmids (size in kb)


Nodulates Mimosa species; highly competitive on M. pudica.


2 chr (3.4 and 2.6)

1 mgpl (1.9) and a pSym (0.59)

CT LMG19424

Nodulates Mimosa pudica; poorly competitive for nodulation on Brazilian ecotypes.


2 chr (3.4 and 2.5)

1 pSym (557)

RM STM3625

Nodulates Mimosa pudica. Poorly competitive compared to BP and CT on M. pudica.


1 chr (4.1)

1 mgpl (1.5) and 3 pl (0.55 (pSym), 0.108, 0.087)

Abbreviations: BP Burkholderia phymatum, CT Cupriavidus taiwanensis, RM Rhizobium mesoamericanum, G genome, mgpl megaplasmid, pl plasmid, chr chromosome, pSym symbiotic plasmid. Genome references: Moulin et al., [144] (BP), Amadou et al., [31] (CT), Moulin et al., [145] (RM). Reference for nodulation competitiveness: Melkonian et al., [35]. Experiments from this study were performed with similar commercial seeds from the Melkonian et al. study. Abbreviations: G genome, Mb megabase, chr chromosome, mgpl megaplasmid, pl plasmid, pSym symbiotic plasmid

Analysis of RNAseq data and their validation by RT-qPCR

We generated RNAseq data from triplicates of cultures of BP, CT and RM at mid-log exponential phase of growth, induced or not by Root Exudates (RE) of Mimosa pudica. The general characteristics of RNAseq data are presented in Additional file 1: Figure S3. The number of reads that passed the initial quality filter (see Mat&Methods) varied from 32 to 111 million. Between 98.16 and 99.78% of reads mapped at least once on each corresponding genome. After in silico depletion of remaining rRNA reads, 3.5 to 18.4 million reads corresponding to mRNA were obtained per sample, representing on average 13 (for BP and CT) to 25% (for RM) of total RNAseq reads.

Box plots of transformed log10 (Read Counts) were used to compare the distributions of our RNAseq data between conditions and between strains and to visualize the effect of normalization. As shown in Fig. 1a, the median and intra-quartile range were equal between the two conditions in each bacterium, but different between bacteria. While BP and CT data distributions were comparable, they differed from those of RM as indicated by box sizes and box plot whiskers. Also, the median reads count per CDS was higher in RM (log value 2.49; 314 reads) than in BP (log value 2.05; 112 reads) or CT (log value 2.196; 157 reads). The box plots representing the distribution of log2 fold change per strain showed further the differences in the response toward RE of the three rhizobia (Fig. 1b). The interquartile values, whisker sizes (+/− 1.5 of the interquartile values) and variance values show that the addition of RE to the culture medium had an important effect on gene expression levels of BP, but were smaller in the case of CT and had a very small effect in the case of RM.
Fig. 1

a-b. Boxplots of Read counts number (Log10) per gene per condition for each bacterium (a), and of Log2 (Fold Change) of gene regulation per bacterium (b). BP, CT and RM: RNAseq data analysis from bacteria in broth culture sampled a 4.5, 4.5 and 6 h, respectively (corresponding to mid-exponential growth phase); BP-I, CT-I and RM-I are the same as previously mentioned, but mixed with Mimosa pudica root exudates

Differentially Expressed Genes (DEG) were selected following the procedure indicated in Methods. For the selected cut-off value (pVal ≤0.01 and fold change < 0.66 (down) or > 1.5 (up)), the number of DEG was 1620, 422 and 237 for BP, CT and RM, respectively. We also compared the number of DEG depending on pVal values and fold change cut-off (Additional file 1: Figure S4), and this showed that the differences in the number of DEG genes was low when varying pVal from 0.05 to 0.001 for a same FC.

To validate the robustness of RNA-seq data, we compared them with the expression of a selection of genes analysed by RT-qPCR. Both specific (5 BP, 11 CT, 7 RM) and common DEG to all three rhizobia (13), up- or down-regulated genes, and those localized on different replicons, were chosen. The characteristics of genes tested are listed in Additional file 1: Figure S5. Two constitutively expressed reference genes (uppS and hisB) were used for qPCR normalisation. RT-qPCR values of genes plotted vs. respective RNA-seq data showed a strong correlation of gene expression between the two methods (BP: 0.82, CT: 0.77, RM: 0.92), validating our RNAseq data (Additional file 1: Figure S6).

Transcriptomic profiles of the three model rhizobia - general overview

The three rhizobia showed different responses to RE at mid-log growth phase. Differences were observed for the number of differentially regulated genes (Table 2), FoldChange values (Fig. 1b), and contribution of each replicon to the transcriptional response (Fig. 2). BP displayed the most complex response with 1620 differentially expressed genes which represent 19.2% of the BP genome. Moreover, BP also showed the greatest number of highly expressed genes, the highest proportion of down-regulated genes (796 down-regulated genes, representing 49.2% of all differentially expressed genes), as well as the highest Fold Change values (FC ≥ 4, not shown) (Fig. 1b). The CT and RM responses were proportionally 4- and 7-fold lower (422 and 237 differentially regulated CDS, corresponding to 7.05 and 3.62% of the CT and RM genomes, respectively). These weaker responses were not proportional to the size of bacterial genomes and should not result from differences in RNAseq data quality. The contribution of replicons to the pool of differentially expressed genes was specific for each strain (Fig. 2). The percentage of induced and repressed genes was the highest in BP representing, on average, 9% per replicon, with an exception for the pSym plasmid. The BP pSym seemed highly stimulated by the induction with RE as we detected a transcriptional induction of 20.5% (129 CDSs) and repression of 1.9% of pSym genes. In the case of CT and RM, up-regulated genes represented, on average, 3.7 and 2% per replicon, respectively. This proportion was lower for down-regulated genes representing 3 and 0.6% for CT and RM, respectively. In contrast to BP, the percentage of induced genes for the CT and RM pSyms did not differ from those of their corresponding chromosomes (Fig. 2). Indeed, only 15 and 19 genes localized on the CT and RM pSyms, respectively, were induced (Fig. 2 and Additional file 1: Figure S2).
Table 2

Distribution of RE-regulated genes among replicons





% of up

% of down







 BURPHK1 (chr)






 BURPHK2 (chr)






 BURPHP1 (pl)






 BURPHP2 (pSym)












 RALTA_A (chr)






 RALTA_B (chr)






 pRALTA (pSym)












 BN77_v1 (chr)






 BN77_p1 (pl)






 BN77_p2 (pSym)






 BN77_p3 (pl)






 BN77_p4 (pl)






 BN77_p5 (pl)






 BN77_p6 (pl)






BP B. phymatum STM815 genome, CT C. taiwanensis LMG19424 genome, RM R. mesoamericanum STM3625 genome, Chr chromosome, pl plasmid, pSym symbiotic plasmid. Replicon names are according to their genbank annotation

Fig. 2

% of DEG per genomes and replicons of bacteria induced by M. pudica root exudates. BP (B. phymatum STM815), CT (C. taiwanensis LMG19424) and RM (R. mesoamericanum STM3625). Replicons names according to Table 2

Among the 1687 orthologous genes constituting the core genome of the three rhizobia, 365 genes showed a differential expression in at least one of the strains (Fig. 3a). We distinguished the different categories of regulation in the center of Fig. 3a. Nine genes were similarly regulated (all up-regulated), with eight genes located on pSym (mainly nod genes), and one (flgG) on the chromosome of each bacterium. Common response of beta-rhizobia (both from core genome set and specific BP/CT orthologs) includes 51 genes (29 up-regulated; 16 down-regulated, 6 in opposite directions) distributed among all replicons in BP and in CT, while 23 genes were found to be regulated in common between BP and RM (10 up, 9 down, 4 in opposite directions) and only three genes between CT and RM (all in opposite directions) (Fig. 3a, Additional file 1: Table S1). We represented the different shared orthologs between replicons in a circular view (RCircos) in Fig. 3b. Many regulated gene orthologs were shared between BP and CT (51 blue links in Fig. 3b), while few were shared between all (9 orthologs, green links) and between BP-RM (23 red links) and RM-CT (3 pink links).
Fig. 3

a-b. Venn diagram of DEG orthologs among the 3 rhizobia (a) and their circular representation (b). On Venn diagram (a), numbers in black indicate the orthologs between bacterial genomes (at intersect) or their specific gene set. Numbers in colors indicate the number of up-regulated genes (red) or downregulated genes (green) among the orthologs at each intersection of the Venn diagram. Numbers in purple indicate gene orthologs regulated in opposite direction between strains. The circular representation (b) was made with RCircos and represent the bacterial replicons (outer circles), their up (red) and down (green) root exudate-regulated genes (log2(fold change) values), as well as links between co-regulated orthologs (green links, between all three rhizobia; blue lines between BP and CT (beta-rhizobia); red links between RM-BP; pink lines: RM-CT). Note the high number of blue lines depicting the high number of shared regulated genes between beta-rhizobial orthologs, while few were shared by all (green links) or between RM-CT and BP-RM (pink and red links)

Concerning the functional classification of DEG, we observed differences among the number of processes affected in each rhizobial species (Fig. 4). However, similarities between rhizobia could be observed when comparing the most affected functional classes (Fig. 4) as Energy metabolism (Biop 6), Transport and binding (Biop 7) and Chemotaxis and motility (Biop 15.2) classes that reflect the typical bacterial responses to the presence of RE components/substrates [4446]. Rhizobia showed also changes in Regulatory functions (Biop 12), Cell envelope (Biop 14), Excrete (Biop 16.4) and Symbiosis (Biop 16.5) functional classes, which can be linked to processes to interaction with the host plant. All these results suggest that BP reacted strongly to the presence of M. pudica RE and deeply reorganized its metabolism and physiology, while CT and RM adjusted the energy metabolism and cellular functions.
Fig. 4

COG functional categories of differentially expressed genes after bacterial treatment by root exudates. Up-regulated COG categories are in red, downregulated in green

RE-regulated functions in the three rhizobia

A general overview of the regulated functions in each rhizobium is presented in Table 3 and all expression data are compiled in Additional file 1: Table S1. In the next sections, we present an analysis of the common and specific responses of each rhizobia to the presence of root exudates.
Table 3

Overview of DEG functional classification in each rhizobium

For each gene functional category, a log2(fold change) is indicated and colored in red if up regulated (> 1.5) or green if down-regulated (<−1.5). P values of DEG are all < 0.01 (cut-off value) (see Additional file 1: Table S1 for values). The gene ID in the genome is indicated below each fold change. NR means Not Regulated by root exudates (genes involved are present but not regulated), while a minus sign (−) indicates the absence of orthologs of such genes in the genome

Common transcriptomic responses of the three rhizobia: Nodulation genes

Nine gene orthologs were detected as commonly regulated in BP, CT and RM. They belong to Symbiosis (7 nodulation genes), Fatty acid and phospholipid metabolism (one gene encoding a putative fatty acid hydroxylase) and Chemotaxis and motility (flgG) functional classes (Table 3 and Additional file 1: Table S1). The induction of nod genes was expected since it has been widely described as a consequence of the perception of compatible flavonoid(s) by NodD transcriptional regulators [42]. The up-regulation of genes from the similar nod operons of BP (nodBCIJHASUQT, Fold Change (FC) 13–22×) and CT (nodBCIJHASUQ, FC 3.6–5.6×) was observed, as shown in Fig. 5. Their orthologs in RM were also up-regulated (FC 1.5 – 2×), but RM carries a different set of nodulation genes compared to BP and CT, with additional nodulation gene operons (nodA1BCSUIJHPQ, nodM, nodO2, and 3 nodD copies) of which some were regulated (Fig. 5). However, while BP and CT nod genes were induced and expressed at a relatively high level (FC of 5-22×) it was not the case for RM (FC of 1.3-2×).
Fig. 5

Nodulation gene operons in BP, CT and RM, colored following their regulation level (fold change) in response to root exudates at 4.5, 4.5 and 6 h post induction, respectively. Black triangles indicate the presence of a conserved nodbox (AATNGATTGTTTNGATNNNNNNNAT). The regulation color scale is indicated on the right side of the figure

Aimed at better exploring the regulation of nod gene expression by Mimosa root exudates, a kinetic analysis of their transcription was performed by qPCR for each rhizobium. The relative expression pattern was established for nodB, nodD and nodA (in the case of RM which contains multiple copies, nodA1, nodA2, nodD1 and nodD3 were tested, Additional file 1: Figure S5).

As shown in Fig. 6, each rhizobium displayed different nod gene transcription kinetics in the presence of root exudates. In the case of BP, the maximum of nodA and nodB transcription was observed in the mid-exponential growth phase, at 4.5 hpi (FC 22× to 48×). CT and RM showed a maximum of nodA(1) and nodB up-regulation (13× and 15×, respectively) 1 h after induction. The signal decreased slowly in CT, but in RM it decreased rapidly to a 2.2× FC 2 h later. This result could explain the weak induction level obtained from RNA-seq data for RM nod genes performed at 6 hpi. No induction of nodA2 (BN77v2_p2140039) was observed in parallel. To our knowledge these results show, for the first time, differences in symbiotic responses of rhizobia stimulated in the same way by root exudates. Clearly, both beta-rhizobia responded with important and persisting up-regulation of genes involved in Nod Factor biosynthesis and transport, contrary to RM. We can hypothesize that this difference could contribute to the competitive deficiency of RM in comparison to BP or CT as observed in [35]. Interestingly, a similar response was observed in the case of other alpha-rhizobia. When induced by bean root exudates, Rhizobium tropici strain PRF81 up-regulated the transcription of its nod genes 5 min post induction [47] followed by a decrease of the transcription signal within 15 min. Sinorhizobium fredii NGR234 also showed a rapid induction of nod genes, within 1 h post induction, although no information is available about the persistence of the induction level [48].
Fig. 6

Kinetics of nod (nodA, nodB and nodD) and fatty acid hydroxylase gene expression as measured by RT-qPCR. The uppS gene was used to normalise data, and the other reference (hisB) gene expression profile is shown

Another commonly regulated gene was related to the “Fatty acid and phospholipid metabolism” bioprocess. This gene (BURPHP2_0638, FC 27×; pRALTA_0471, FC 2.4× and BN77v2_p2140043, FC 1.74×) encodes an integral membrane protein containing a putative domain belonging to the “sterol desaturases and fatty acid hydroxylase superfamily”. The three orthologous genes share 80.3 and 78.7% of amino-acid identity (BP-CT and BP-RM, respectively), and less than 40% amino acid identity with orthologs from other bacterial genomes. Further studies are required to infer the function of these pSym-located putative “fatty acid hydroxylases” which showed several interesting features such as i) their up-regulation by root exudates with an expression pattern similar to nod genes as studied by qPCR (Fig. 6); ii) it is preceded by a nod-box sequence in CT (pRALTA_0471) [31], and we also detected a putative nod-box preceding each ortholog in BP (543894–543,880 bp in BURPHP2) and RM (251,926–251,950 bp in BN77v2_p2). This gene may thus represent a new specific nodulation gene for the symbiosis with Mimosa pudica.

The last commonly up-regulated gene in all three rhizobia is flgG (BURPHK1_3188, FC 1.7×; RALTA_B0282, FC 2.4× and RN77v1_1547, FC 2.1×) which is part of the flagellar components operon and encodes a protein of the basal-body rod.

Apart from the nine commonly regulated orthologous genes, we could observe common responses at the functional and biological process levels. Processes such as sensing, chemotaxis, flagella-driven motility and swarming, were all detected as regulated functions in our transcriptomic datasets. These functions are crucial for bacterial colonization of roots and have been found to be induced by root exudates in many rhizospheric bacteria [4446]; differences in the responses to these functions may strongly influence the competitivity of bacterial symbionts [49]. We observed that the general chemotaxis and flagella-driven motility processes seemed to be regulated differently in the three bacteria. In RM we could observe the up-regulation of a whole region (RHI36v1_1529–1540) containing operons coding for chemotaxis (che and mot), flagella biosynthesis (flg, fli, fla, flb, flh) and flagellar motor components (fliGNM and motA). BP also showed a slight up-regulation of operons coding for chemotaxis (che) and flagella biosynthesis operons (flg, fli, flh) (Table 3, Additional file 1: Table S1). In CT, chemotaxis (che) and flagella biosynthesis operons (flg, fli, flh, mot) are regulated at a relatively low level, with an up-regulation for only two CDS (flgA and flgG).

Commonly regulated functions also comprised several histidine kinases, response regulators of two-component systems, methyltransferase of chemotaxis proteins, methyl-accepting chemotaxis proteins and putative chemoreceptors (Table 3, Additional file 1: Table S1). All these potential regulators could reflect differences between rhizobia in their capacity to sense/recognize and activate chemotaxis toward components of M. pudica root exudates.

We also observed that some regulated genes in the three rhizobia were linked to copper and zinc homeostasis in the presence of RE (Table 3). An up-regulation of metal efflux systems was illustrated in BP by the induction of the cop operon (BURPHK1_2204–2208, FC 2.15–4.05×) involved in copper resistance. In CT, many metal transporters were regulated: the copSRABCD operon (RALTA_B1970-B1965, FC 7.9-11×) coding for proteins involved in copper detoxification by extrusion of Cu2+ [50], genes of the czcA2B2C2I2 operon (pRALTA_0640–0645, FC 1.61× and 1.54×) encoding components of an RND-HME efflux complex conferring resistance to Co, Zn and Cd [51], the cupA2 gene (RALTA_A3123, FC 1.8×) encoding copper transporting P-type ATPase and its predicted partner CupZ (RALTA_A3123, FC 1.95×) encoding a copper ion binding metal transporter, and by a zntA gene encoding a cation (Cd)-efflux ATPase (RALTA_B1454, FC 2.16×). Concerning metal transport and/or homeostasis systems we observed in BP the down–regulation of the cutC operon (BURPHK1_0153–0154, FC − 2.0×) coding for copper homeostasis protein (BURPHK1_0154), and in CT the down-regulation of a gene encoding a divalent heavy-metal cations transporter (RALTA_A2241, FC − 3.85×). The induction of copper efflux-encoding genes has already been reported as an adaptation to the rhizosphere in alpha-rhizobia but also in other rhizospheric micro-organisms: R. leguminosarum bv. viciae showed an up-regulation of copA, copC (copper efflux) and actP/cueA (putative P-type ATPase) genes when induced with hesperetin and in the pea rhizosphere [49], while the copSR genes of Pseudomonas syringae [52] and the cueA gene (P1-type ATPase) of P. fluorescens SBW25 [53] were also up-regulated in the sugar beet rhizosphere.

Burkholderia and Cupriavidus common responses: Beta-rhizobia

Despite the fact that Burkholderia and Cupriavidus belong to Betaproteobacteria, they are taxonomically distant, and in this study we detected only 51 commonly regulated orthologous genes between BP and CT. Apart from the nine genes already described as being commonly regulated in the three rhizobia, the other 42 genes represented 13 putative functional classes (Table 3, Additional file 1: Table S1). Among the up-regulated genes, eight genes belonging to the “Energy metabolism” and “Transport and binding” functional classes were associated with putative systems for the catabolism of RE components. One of the genes encodes a putative beta-ketoadipate enol-lactone hydrolase (BURPHK1_0214, FC 1.93× and RALTA_A0090, FC 1.57×) which catalyses a common step in the catabolism of protocatechuate and catechol by bacteria [54]. The importance of protocatechuate catabolism was demonstrated as an important capacity for the growth of Rhizobium leguminosarum bv viciae in the pea rhizosphere [55]. Two other genes, localized in orthologous operons, encode a putative hydroxymethylglutaryl-CoA lyase (BURPHK2_1264, FC 7.86× and RALTA_A1930, FC 2.12×) and a putative caiB/baiF family CoA-transferase (BURPHK2_1263, FC 8.95× and RALTA_A1931, FC 1.98×) involved in the mevalonate pathway for isoprenoid catabolism [56]. We also detected the up-regulation of a three gene operon encoding a sugar ABC transporter system (BURPHK2_2498–2503, FC 1.5–1.8× and RALTA_A2001–2006, FC 4.2–5.3×), a putative glycerol kinase (glpK, BURPHK1_0447 FC 2.58× and RALTA_A2010, FC 3.21×) and a putative sn-glycerol-3-phosphate dehydrogenase (glpD, BURPHK1_0448 FC 6× and RALTA_A2011, FC 3.54×). These genes show homologies to genes described in R. leguminosarum bv. viciae 3841 and involved in glycerol transport and catabolism, mutants of which are affected in their competition for nodulation [57].

Concerning the “amino acids transport and metabolism” bioprocess, two genes were up-regulated. The first one (dadA, BURPHK1_0823, FC 3.81× and RALTA_A0798, FC 5.04×) encodes a D-amino acid dehydrogenase involved in degradation of D-amino acids leading to the formation of corresponding oxoacids that can enter the Krebs citric acid cycle and glycolysis. The second one (BURPHK1_0327, FC 1.78× and RALTA_B0072, FC 1.65×) shares poor identity (25%) with LeuA (encoding a putative 2-isopropylmalate synthase) and may be involved in the metabolism of amino acids. In parallel, we observed a down-regulation of genes coding for branched-chain amino acid transporters, with two operons in BP (BURPHK1_0631–0635, FC − 2 to − 2.85× and BURPHK1_2430–2433, FC − 2.5 to -3×) and two operons in CT (RALTA_A1326–1331, FC − 2.6×; and RALTA_A3113–3117, FC − 2.5×). Branch-chained amino acids have been shown to be important for effective symbioses between BP and Mimosa pudica [58].

We also observed an adjustment of the central metabolism of beta-rhizobia. First with the up-regulation of the aceA gene (BURPHK1_1476, FC 1.64× and RALTA_A1766, FC 11×) encoding an isocitrate lyase, an enzyme of the glyoxylate bypass of the tricarboxylic acid (TCA) cycle. The FC was much higher in CT, which is linked with the up-regulation of two other genes of the glyoxylate bypass in CT (see CT specific response section). Rhizobia use the TCA cycle for the oxidation of organic acids [49], a feature that could be also present in beta-rhizobia. Further, we observed a down-regulation of an operon of eight genes coding for subunits of a NADH-quinone oxidoreductase (nuo operon: BURPHK1_2160–2173, FC-1.8 to − 2.2× and RALTA_A1034–1047, FC − 1.5 to − 1.6×), involved in the generation of the proton force which can be used for the synthesis of ATP, the active transport of various nutrients and the ATP-dependent rotation of the flagella [59].

Bacterial survival in the rhizosphere, apart from adaptation toward nutrient sources, also involves dealing with environmental challenges caused by the presence of oxidative stress and toxic compounds. In our study, both beta-rhizobia showed a common response toward a putative oxidative stress by up-regulating genes coding for antioxidant-proteins classified in the “Detoxification” functional group. These were genes coding for a peroxiredoxin-like protein (BURPHP1_1192, FC 1.62× and RALTA_B2016, 1.58×), a putative thioredoxin reductase (BURPHK2_0718, FC 4.57× and RALTA_A1176, FC 2.1×) and a glutathione S-transferase (BURPHK1_0157, FC 1.86× and RALTA_A1551, FC 1.54×). Induced transcription was also observed in BP for an ahpD-like gene (BURPHP1_1209, FC 5.37×) coding for a putative antioxidant alkyl hydroperoxide reductase, and in CT for two putative universal stress proteins of the uspA-family (RALTA_A0490, FC 1.81× and RALTA_A2972, FC 1.58×). The up-regulation of genes encoding putative peroxiredoxin, thioredoxin, glutathione S-transferase and UspA, was previously observed in different plant-microbe early interactions as R. leguminosarum bv. viciae in the pea rhizosphere [49]. Surprisingly we did not observe this type of response in our alpha-rhizobium (RM) but only in the beta-rhizobia. Concerning other sensing and detoxification systems we detected the up-regulation of a NO dioxygenase (BURPHK2_1027, FC 1.83× and RALTA_A2986, FC 2.8×) which converts NO to N2O or to NO3 [60], involved in nitric oxide (NO) sensing and detoxification. NO has been shown to be an important factor for the establishment of Sinorhizobium-Medicago symbiosis [61].

Finally, both beta-rhizobia showed similar up-regulation of several operons encoding efflux systems. These were in BP, two systems belonging to resistance/nodulation/cell division (RND) family and of unknown function BURPHK1_2604–2606 (FC 2.2–5.2×) and BURPHP2_0566–0568 (FC 4.8–7.3×), the second being located up-stream of the nod operon. In CT three RND systems were up-regulated: a multidrug efflux system mdtABCD (RALTA_A0537–0540, FC 1.73–1.47×), a multidrug transporter (RALTA_A3171–3173, FC 3.69–2.31×) and a putative efflux pump similar to acriflavine/multidrug efflux acrAB and nodT (RALTA_B0936–38, FC 1.5×). Indeed, it has been found that the RND systems might have a relevant role in the interaction of bacteria with their plant hosts, from the first steps of colonization [62] to their survival in plant tissues [63]. It was also demonstrated in alpha-rhizobia that specific RND systems may affect the number of nodules formed on host roots [64], the level of nitrogen fixation in nodules [65] and the rhizosphere competitiveness of rhizobia [49]. Finally, we observed the up-regulation of four genes coding for proteins involved in regulatory mechanisms including a sigma 54 factor, and regulators of the MerR, GntR and LuxR families (Additional file 1: Table S1).

Burkholderia and Rhizobium common responses

Twenty eight orthologous genes were found to be commonly regulated in BP and RM. Among them, 19 were up-regulated (including nod genes) and nine were down-regulated. They were classified into eight functional groups (Additional file 1: Table S1). Among the up-regulated genes belonging to the “Energy metabolism” bioprocess, we observed induction of two systems. The first one is represented by the genes iolB, iolC and mmsA/iolA (Additional file 1: Figure S8). In RM these genes exhibit a high nucleotide identity (from 82.8 to 94.8%) with orthologs in R. leguminosarum and S. meliloti encoding a myo-inositol catabolism system [66]. The capacity of both bacteria to use myo-inositol as a single carbon source was observed on API GN20 galerie tests (Additional file 1: Figure S9), while CT was unable to. The presence of inositol and pinitol (3-O-Methyl-D-chiro-inositol) was detected in M. pudica RE, the last one representing the most concentrated component of RE (Table 4). These genes could thus be involved in catabolism of inositol (and derivative compounds), a key pathway involved in host nodulation competitiveness in S. meliloti [67]. The second “Energy metabolism” bioprocess is represented by four genes (BURPHK2_0654–0657, FC 1.84–2.32× and BN77v1_3275–3278, FC 2.08–2.32×) encoding a phosphate acetyltransferase (Pta), an acetate kinase (AckA) and an enoyl-[acyl-carrier-protein] reductase (FabI). It was described in S. meliloti as being induced by phosphate starvation [68].
Table 4

Composition of M. pudica root exudates and its final concentration when added into culture medium for rhizobium induction


Concentration (μM)

% error

























aspartic acid



gluconic acid



glutamic acid



malic acid












a polyphenolic compounds measured as equivalent of ferulic acid (μ−1). Major compounds concentrations are indicated in bold

We also observed a down-regulation of four genes belonging to the “Transport and binding proteins” bioprocess, encoding putrescine/spermidine ABC transport systems (BURPHP1_0401–0404, FC -2× and BN77v1_3961–3964, FC − 1.75 to − 5.5×). γ-aminobutyric acid (GABA) and putrescine were shown to be present in root exudates of numerous plants and they seem to play a role in nitrogen cycling in both the rhizosphere and within nodules [69].

Cupriavidus and Rhizobium common responses

Apart from the up-regulation of the nodulation gene operon and a fatty acid hydroxylase gene, there was only three orthologous DEG in CT and RM, but they are regulated in opposite directions, including cobU and fliI, and a gene encoding a protein of unknown function. Such absence of overlap in the responses of both symbionts to root exudates mirrors their different adaptation to symbiosis.

Burkholderia phymatum specific responses to RE

For specific responses of each organism, we compiled both RE-regulated genes specific to each rhizobium and those only regulated in one bacterium but with orthologs in the other rhizobia. Here we present the most interesting, in our opinion, responses of BP, CT and RM, in the context of plant-bacterial interactions. The complete list of specific regulated genes of the three rhizobia is presented in Additional file 1: Table S1.

The BP-specific response is represented by 755 genes with assigned function (288 genes up-regulated and 467 genes down regulated, within 27 assigned bioprocess functions) and 557 genes of unknown function.

BP activated the transcription of several genes encoding functions known to be involved in the interaction between bacteria and plants (Table 3): the acdS gene encoding 1-aminocyclopropane-1-carboxylate (ACC) deaminase (BURPHP2_0697, FC 2.24×) and a putative rhizobitoxine biosynthesis operon of seven genes (BURPHP2_0646–0653, FC 2.4–3.4×) (Additional file 1: Figure S10). The first four genes of this operon show high homology (from 52 to 70% of nucleotide identity) to the rtxACD genes involved in the biosynthesis of rhizobitoxine in Bradyrhizobium elkanii USDA 76 [70]. The fact that RtxACD were shown to be essential and sufficient for the biosynthesis of rhizobitoxine [71] suggests that BP could be able to synthetize rhizobitoxine. Both ACC deaminase and rhizobitoxine were shown to play a positive role in establishing symbiosis between rhizobia and their legume hosts by enhancing the nodulation process: rhizobitoxine by inhibiting the activity of the plant ACC synthase involved in ethylene biosynthesis in host roots [72] and ACC deaminase by degradation of ACC, the immediate precursor of ethylene [73]. We also detected the up-regulation of an operon (BURPHP2_0632–631, FC 1.7–1.8×) encoding two putative proteins involved in the biosynthesis of the plant hormone indole-3-acetic acid (IAA). The role of IAA biosynthesis by rhizobia during symbiosis remains unclear. It was demonstrated that in Sinorhizobium sp. NGR234 flavonoids induce IAA production via the transcriptional regulators NodD1 and NodD2 [74], and a strain of R. leguminosarum bv. viciae (strain RD20) overproducing IAA elicits the development of fewer but bigger nodules on vetch, and with an enlarged and more active meristem, compared to nodules infected by the wild-type strain [75]. Another highly up-regulated gene possibly involved in the plant-bacterial interaction is BURPHP2_0639 (FC 19×), which encodes a putative endoglucanase. This gene is located on the pSym, down-stream of the earlier described putative fatty acid hydroxylase (BURPHP2_0638, FC 27×), and represent the fifth most induced gene in BP. It shows 27% amino acid identity to CelB, an endoglucanase of the phytopathogenic Pectobacterium carotovorum LY34 strain and involved in plant cell wall degradation [76]. The only rhizobial cell-wall-degrading enzyme described to be involved in early symbiotic interaction was CelC2 of R. leguminosarum bv. trifolii [77, 78]. Even if our BP endoglucanase shows no homology to CelC2, its strong induction by RE may, however, be a good indicator of its implications for early symbiotic events and makes it a good candidate for further study. We detected orthologs of these genes in the genomes of CT or RM: acdS is present in the RM genome (BN77v2_p270023, FC 1×) and the endoglucanase encoding gene in the CT genome (pRALTA_0250, FC 1×). It is noteworthy that the three above mentioned mechanisms are encoded by genes localised on the pSym. Several nitrogen fixation-related genes were also highly induced in BP: an operon containing conserved fragments of nifZ (BURPHP2_0643, FC 16×) and nifB (BURPHP2_0640, FC 11×), as well as the fixU gene (BURPHP2_0644, FC 11×), and several other nif and fix genes (BURPHP2_0596–0614, FC 1.6–2.2×) (Table 3).

BP specific responses which could contribute to symbiont-host plant interactions concerned the differential expression of secretion systems (Table 3). BP responded to RE by up-regulation of a T4SS (BURPHP2_0345–367, FC 1.47×–4.34) and a T6SS (T6SS-2, BURPHP1_0625–645, FC 1.53×-1.93×), and by down-regulation of another T6SS (T6SS1, BURPHP1_0486–506, FC − 2.0× to − 2.7×). The up-regulated T6SS consist of genes arranged into two modules transcribed in opposite directions (Additional file 1: Figure S11). This T6SS was classified as T6SS-2 on the basis of neighbour-joining phylogenetic analyses of VgrG, Hcp, ClpV, IcmF, and the lysozyme-like protein VC_A0109 [79]. The closest (but incomplete) locus synteny was found with the endophyte B. phytofirmans PsJN [80], the animal pathogen B. mallei ATCC 23344 [81], the causative agent of melioidosis B. pseudomallei K96243 [82], the rice sheath pathogen B. gladioli BSR3 [83] and with the mango plant pathogen Burkholderia sp. TJI49 [84] (Additional file 1: Figure S11A). The first module of this T6SS-2 is composed of the sciHIKN genes, while the second module comprises a gene encoding a putative conserved protein, the vgr gene and a gene encoding a putative chitinase. Hcp (SciK) and Vgr are the essential components of T6SS and current models of T6SS suggest that Hcp and Vgr form a tubular structure of stacked Hcp hexamers terminated by a VgrG trimer, that functions as a perforating device toward the targeted cells [85]. The Vgr protein (encoded by BURPHP1_0643) possesses a lysozyme domain in C-terminal (as predicted by InterProScan). Downstream of the vgr gene another up-regulated gene is present encoding a putative enzyme harbouring muramidase (504–742 aa) and peptidoglycan hydrolase (33–164 aa) catalytic domains. T6SS have been described as important factors in interbacterial species competition for niche occupancy, by causing damage to competitors by perforation [86, 87]. The induction of such T6SS-2 could be an important factor in the high competitivity of BP against its rhizobial competitors [35, 88]. This is, to our knowledge, the first report of the induction by RE of genes encoding a T6SS in rhizobia.

The second BP T6SS (BURPHP1_0486–506), which is down-regulated by RE, shares high identity and synteny with T6SS in Burkholderia symbionts of Mimosa (Burkholderia sp. CCGE1002, JYP251), plant endophytes (B. kururiensis M130) or soil bacteria (B. terrae BS001) (Additional file 1: Figure S11B). Rhizobial T6SS (as T4SS) can have positive or negative effects on host nodulation by suppressing plant immune systems [89]. Although the direct influence of T4SS on symbiosis was demonstrated by mutagenesis in alpha-rhizobia [90, 91], there is only one report of the role of a T6SS in nodulation in R. leguminosarum where the T6SS prevented nodulation on Pisum sativum cv. Rondo [92].

The BP T4SS consist of the Mpf (Mating pair formation) genes virB1 to virB11, as well as the regulator virD4, and can be classified as of the T4ASS type based on its homology to the complex VirB/D4 of the plant pathogen Agrobacterium tumefaciens [93] (Additional file 1: Figure S12). All genes of this T4SS are moderately up-regulated (1.47 x – 4.34 x FC), as does two genes located up-stream (BURPHP2_0369–0370, 2× and 3.5× FC) which code for a putative transcriptional regulator (XRE family) and a putative sensor protein with a PAS domain which could be involved in the T4SS regulation. Four other up-regulated genes (BURPHP2_0345–348, 2.2 x – 2.5 x FC) located down-stream of the T4SS encode a putative Dtr (DNA transfer and replication) module [94] of the conjugative machinery. The presence on the pSym and simultaneous up-regulation of these conserved Dtr and Mpf modules suggest the involvement of the BP T4SS in the conjugative transfer of the pSym.

We also detected a down-regulation of genes encoding putative proteins of the general secretion pathway (Sec-pathway) classified also as Type 2 Secretion System (T2SS) by which proteins are transported across or inserted into the cytoplasmic membrane [95]. These genes are secY (BURPHK1_3037, FC − 2.4×), secE (BURPHK1_3073, FC − 1.56×) and secDF (BURPHK1_2730–2731, FC − 1.5×). A slight decrease of expression was also observed for the genes secA (BURPHK1_2871, FC − 1.3×) and secB (BURPHK1_0294, FC − 1.3×). The Sec translocation system is involved in the secretion of the vast majority of secreted proteins, and it was also shown to be involved in the secretion of toxins in Listeria monocytogenes and Bacillus cereus [96, 97].

Other specific responses of BP concerned the bacterial catabolism and metabolism of aromatic compounds, polyamines and amino acids (Table 3). Our data showed the up-regulation of genes coding for enzymes possibly involved in the catabolism of aromatic compounds: a putative 4-oxalocrotonate tautomerase (BURPHK2_0305, FC 1.74×) involved in the degradation of catechol, a putative mandelate racemase/muconate lactonizing-like protein (BURPHP1_0070, FC 1.95×) involved in aromatic acid catabolism, a dioxygenase annotated as a putative glyoxalase/bleomycin resistance enzyme (BURPHK2_1088, FC 1.65×), a DOPA 4,5-dioxygenase (BURPHK2_2585, FC 2×) involved in aromatic ring-cleaving, and a dehydratase (BURPHK1_1833, FC 1.5×) which could be involved in the degradation of sistosterol-like molecules or fatty acids. Concerning polyamines, several putrescine transporters were down-regulated (see BP & RM common responses) and several genes encoding putative enzymes involved in the degradation of γ-aminobutyric acid/putrescine (puuC, BURPHK2_0522, FC 1.8×; puuDA, BURPHK2_0525–0526, FC 1.4–1.7×; puuC, BURPHK1_2348, FC 1.55×; BURPHP1_1138–1140, FC 2–2.5×) were up-regulated. Another observation concerns the down-regulation of 17 genes encoding enzymes involved in amino acid (leucine/valine, histidine, phenylalanine/tryptophan, lysine, arginine, tyrosine and glutamate) biosynthesis, as well as cysteine, histidine and methionine transporters (Table 3). The down-regulation of different types of amino acid and of polyamine transporters suggests a reorganization of the metabolism of BP to the presence of RE compounds.

Finally, we detected the up-regulation of genes coding for a putative siderophore non-ribosomal peptide synthase (BURPHK2_1036–1049, FC 1.54–2.34×) known to play a pivotal role in iron access competition. The BraI/BraR quorum sensing regulation system (BURPHK2_1465–1467) [98] and Contact-Dependent growth Inhibition Systems (CDI) CdiAB [99] (BURPHP2_0083, 0089, 0113, 0130) were not regulated in our RE-induced conditions.

Cupriavidus taiwanensis specific response

The CT specific response is represented by 186 genes with assigned function (104 genes up-regulated and 82 genes down regulated) and 134 genes of unknown function. The most pronounced CT specific response in terms of number of genes (up- and down-regulated) concerns the bioprocesses “Energy metabolism”, “Transport and binding”, “Regulatory functions” and “Excrete” (Fig. 4).

An up-regulation of genes coding for proteins involved in root attachment was detected in CT: the operons pilVWXYE coding for the biogenesis of the type 4 fimbriae (twitching motility that can be important factors in host colonization [100]) (RALTA_A2329–2324, FC 1.4–2.08×; and RALTA_A2331–2335, FC 0.9–1.9×) and pilQPONM coding for the assembly of the type 4 fimbriae (RALTA_A2895–2899, FC 1.53–1.94×), and a monocistronic gene coding for a PilX-related protein (RALTA_A0505, FC 1.95×). PilX was shown to be required for surface and endophytic colonization of rice as well as twitching motility of Azoarcus sp. BH72 [46], and required for attachment and invasion of human cells by the pathogens Pseudomonas aeruginosa and Neisseria meningitidis [101, 102].

Two putative chromosomal toxin/antitoxin systems (TA) were up-regulated: kluBA (RALTA_A0064–0065) for which only the antitoxin was induced (RALTA_A0065, FC 1.52×), and RALTA_B2161–2162 for which both toxin and antitoxin were induced at a relatively high level (FC 2.13–2.24×). This latter system shows synteny and 41.3% identity with a TA system localized on the pSym (pRALTA_0622–0623, FC 1.43–1.48×). The physiological function of chromosomal TA systems is still controversial. The activation of TA under various stress conditions suggests a function in general stress management and of special importance during the adaptation of rhizosphere bacteria to oligotrophic conditions as well as a role in plant-bacterial interactions [103, 104]. No up-regulation of genes encoding putative toxin/antitoxin systems was observed in BP or RM.

CT up-regulated several genes involved in the glyoxylate bypass pathway: an isocitrate lyase (aceA1, RALTA_A1766, FC 11×) and its paralog (aceA2, RALTA_B0267, FC 1.89×) and a malate synthase (aceB, RALTA_A1760, FC 1.6×). Induction of these genes was also observed for R. leguminosarum in its specific host rhizosphere (pea) and suggests the catabolism of short chain (C2) organic acids [49]. In CT the glyoxylate bypass could also be linked with PHB degradation that would provide more acetyl-CoA [105]. CT up-regulated two out of seven genes encoding intracellular poly(3-hydroxybutyrate) depolymerases (phaZ3 RALTA_B1788, FC 1.54× and phaZ4 RALTA_B2163 FC 1.52×) as well as one gene encoding a beta-hydroxybutyrate dehydrogenase (RALTA_A1256, FC 1.35×). At the same time, CT down-regulated an acetoacetyl-CoA reductase (phaB2, RALTA_A1758, FC − 2.44×) and phasin (phaP2, RALTA_A1759, FC -2×) involved in PHB synthesis. Studies of [106, 107] on the plant pathogens Rhodococcus fascinans and Xanthomonas campestris have suggested an important role of the glyoxylate cycle in plant-microbe interactions, as malate synthase-deficient mutants were impaired in their pathogenic capacities. However, the glyoxylate cycle does not seem to be important during symbiotic interactions in alpha-rhizobia, as isocitrate lyase null mutants of R. leguminosarum and S. meliloti, and malate synthase null mutants of R. tropici and S. meliloti were unaffected in their nodulation capacities [105, 108, 109]. PHB synthesis and degradation is of great importance for rhizobial persistence and survival during starvation [110]. Concerning the glyoxylate cycle and PHB degradation, our results showed an interesting difference between CT and the other two rhizobia, BP and RM.

CT up-regulated two operons involved in nitric oxide (NO) sensing and detoxification: a NO-sensing transcription repressor (nsrR, FC 6.3×) and a NO dioxygenase (hmp, FC 2.8×) which converts NO to N2O or to nitrate (NO3) (RALTA_A2987–2986) [60] and a NO-binding di-iron protein (norA, FC 6.3×) that modulates the NO-responsive expression of the norAB operon by lowering the free cytoplasmic concentration of NO [111] (RALTA_B2089–2087). The up-regulation of two independent NO detoxification systems suggests the synthesis of NO in the CT cells. CT nitrate and nitrite reductases (napCBA1DE and nirSCFDGHJNE, respectively) are not up-regulated and seem to be expressed at a relatively low level, and no gene(s) coding for NO synthase could be identified. Concerning CT nif and fix genes, these are not regulated by RE but are transcribed at a high level (with nifHDK FC at 1×, 25,000 to 29,000 reads per CDS), contrary to BP and RM (which carry two nitrogenase encoding loci) in which nif and fix genes are transcribed at a very low level under both conditions. NO is not known to be a nitrogen fixation by-product, but its presence in nodules is known to inhibit nitrogenase activity [112], so such detoxification systems in CT could be linked to symbiosis. Multidrug efflux systems were also up-regulated in CT, such as mdtABC (RALTA_A0538–0540, FC 1.7×).

CT possesses a specific T3SS (RALTA_B1250–1262) which is not induced by RE and is not expressed (less than 10 reads for each CDS). It was previously shown that its inactivation had no effect on Mimosa pudica nodulation but allowed CT to establish chronic infections and to fix nitrogen in Leucaena leucocephala [113], which is in line with our RNAseq results showing neither induction nor expression of this T3SS under our experimental conditions.

Rhizobium mesoamericanum specific responses

The RM specific response to root exudates is represented by 117 genes with assigned function (97 genes up-regulated and 20 genes down regulated, with 17 assigned functions) and 69 genes of unknown function (Additional file 1: Table S1). The most pronounced RM-specific response in terms of number of genes involved (up- and down-regulated) concerns the “Chemotaxis and motility” (Biop 15.2), “Transport and binding” (Biop 7) and “Cell envelope” bioprocess (Biop 14) (Fig. 4).

The first interesting feature in the RM-specific response is the induction of specific nodulation genes absent from BP and CT, namely nodM, nodO1 and nodO2 (BN77v2_p2180030 FC 2.4×, BN77v2_p290009 FC 1.44×, BN77v2_p2180035 FC 1.83×, respectively) (Fig. 5). The nodM gene encodes a fructose-6-phosphate glutamine aminotransferase and provides glucosamine-6P that are assembled by NodC to form the Nod factor backbone [114, 115]. It is a paralog of the conserved bacterial gene glmS. BP and CT only carry the conserved glmS gene. The nodO1 and nodO2 genes encode calcium-binding proteins. NodO has been shown to insert into liposomes and form cation-selective channels in lipid bilayers and could, therefore, enhance the calcium spiking that is observed in root hairs upon Nod factor binding [116]. In R. leguminosarum bv. viciae, a type I secrection system (PrsDE) allowed the secretion of the NodO protein [117]. Interestingly, orthologs of PrsDE (66 to 69% amino acid identity) located on the RM pSym were also up-regulated by RE (BN77v2_p2120002-p2120004, FC 1.6–2.4×). RM also harbors two copies of the nodA gene, of which only one (nodA1, BN77v2_p280014, FC 1.9), is induced by RE (nodA2 is BN77v2_p2140039, FC 1.05) (Fig. 5).

Concerning chemotaxis and motility, RM up-regulated a gene encoding a chemoreceptor (mcpE, BN77v1_1522) upstream of a whole chemotaxis up-regulated operon cheXYAWRBD (BN77v1_1522–1530, FC 1.5–1.8×) and close to another up-regulated motBCD operon (BN77v1_1559–1563, FC 1.7–1.8). Other methyl-accepting chemotaxis proteins were up-regulated (mcpA, BN77v1_1799 FC 1.8×, BN77v1_1594 FC 1.6, BN77v1_1773 FC 1.6), and one was down-regulated (mcpY, BN77v1_p10178, FC − 1.7×). Rhizobia mediate chemotaxis by sensing molecules from root exudates using chemoreceptors, a phenomenon that allows the bacterium to activate several functions (including motility) to reach its host, and is important for host nodulation competitiveness [118]. We also observed the up-regulation of 31 motility genes (BN77v1_1532–1576, FC 1.5–2) localized downstream of the chemotaxis genes previously described, and involved in flagellar biosynthesis (motor switch flhB, fliFGNM, motA, basal body flgBC, fliE, flgGAIH, flagellin biosynthesis flaABC, flagellar hook flgEL). The up-regulation of these motility genes is in parallel with the downregulation of the csgAB genes encoding curlin biosynthesis enzymes involved in biofilm formation (BN77v1_p11016–11,018, FC − 2.6 to -3×), reflecting a switch of the bacterium to a mobile state.

Another specific response of RM concerns the “Cell envelope” bioprocess, by regulating genes involved in modifications of EPS and LPS composition. We detected the up-regulation of 18 genes encoding putative proteins involved in the biosynthesis of EPSII (BN77v2_p11446–11,463, wgeBCDEFG/wgeA/wgdAB/expG/wgcA/wgaABDEF, FC 1.9×) and of a gene encoding a negative regulator of EPS, PsiB (BN77v2_p2140032, FC 1.8×) [119], while in parallel 15 genes involved in EPSI were downregulated (exsH, exoYF, aceC, pssP, BN77v2_p10959–10,972, FC − 1.5 to -2×), indicating a shift from EPSI to EPSII of the RM EPS composition. We also observed the up-regulation of the operon rfbABCD (BN77v1_p11464–11,467, FC 1.7–1.8×) encoding proteins involved in O-antigen building, and which blocks biosynthesis of a putative galactopyranose mutase (glf, BN77v1_p10763, FC 1.9×), a putative rhamnosyl O-methyltransferase (BN77v1_0913, FC 1.78×) and a putative glycosyl transferase (BN77v1_2672, FC 2.1×). Exopolysaccharides have been shown to play different roles in symbiosis, especially for bacterial propagation in infection threads and suppression of host defenses [120, 121].

Several genes involved in sugar transport and/or catabolism, were up-regulated in RM: a fructose uptake ABC transporter (frcKABC, BN77v1_1335–13, FC 4–4.7×; fructose being the third main component of root exudates), a xylose epimerase (BN77v1_p11238, FC 4×) and an ABC transporter (BN77v1_p10666–10,668, FC 2.4–2.9×). We also detected the up-regulation of three genes (BN77v1_0504–0506, FC 1.53–2.11×) with high homology with the mocDEF (59.7 to 79.3% amino acid identity) from S. meliloti L5–30 and R. leguminosarum bv viciae, species in which these genes are part of the mocCABRDEF operon encoding a rhizopine catabolism pathway [122, 123]. Close to mocDEF was found an homolog of mocB (encoding the rhizopine-binding protein) within an ABC transporter operon (annotated as a galactoside transporter), which was also up-regulated by RE (BN77v1_0783–0785, FC 4–4.7×). The rhizopine catabolic capacity of R. leguminosarum also requires the presence of the myo-inositol degrading enzymes IolD (acetolactate synthase) and IolA (methylmalonate-semialdehyde dehydrogenase) [124]. We also observed in RM an up-regulation of genes involved in myo-inositol degradation (described in the “Burkholderia and Rhizobium common responses” section) and of the idhA gene (BN77v1_3490, FC 2.4×) encoding a putative Idh/MocA family oxidoreductase. Taken together, these data suggest that RM could catabolize rhizopine, an ability that has been shown to provide an advantage for competitive host nodulation in R. leguminosarum and S. meliloti [67, 125]. We also detected the up-regulation of an operon encoding a lipid acid hydrolase harboring a patatin-like domain and a membrane fusion protein (BN77v1_p10807–10,810, FC 2.4–3.4×). Finally, RM up-regulates genes involved in the formation of mechanosensitive channels (BN77v1_p10999, FC 1.9×; BN77v1_3475 FC 1.9×) which have been suggested to play a role in the response to hypo-osmotic shock in the rhizosphere [126].


In this study we compared the transcriptomes of three different rhizobial symbionts of Mimosa pudica, in response to root exudates (RE) at mid-log exponential stage of growth. The main results are illustrated in Fig. 7. We observed that BP has the most substantial response to RE both quantitatively and qualitatively compared to CT and RM, which is in line with its high competitiveness recorded in previous work on Mimosa [29] and more recently on papilionoid legumes [88]. BP harbors and up-regulates in the presence of RE many genes known to be involved in plant-bacterial interactions, present in Alpha and Betaproteobacteria, including ACC deaminase, rhizobitoxin production, IAA biosynthesis, and a T6SS to probably eliminate its competitors. Such high transcriptomic reorganization to the sensing of RE, and the presence of many (almost all known) genes involved in plant growth promotion in BP, mirrors the long history of symbiosis between burkholderias and Mimosa species as inferred by phylogenetic analyses [30]. Such features were not observed in CT, which nodulates Mimosa pudica mainly in areas where it is invasive and not native, and where its main competitors (i.e. burkholderias) are counter-selected by environmental conditions (such as high soil pH or the presence of heavy metals) [5]. CT displays a “minimal” transcriptomic response to RE, reflecting a shorter history of symbiosis with Mimosa pudica [5], and/or the fact that it encounters its host under harsh conditions counter-selecting Burkholderia or Rhizobium spp., thus lowering the selective pressure for genes involved in competitiveness. Indeed, apart from nodulation genes, CT regulates only trophic or stress-related functions (e.g. heavy-metal transporters) when its competitors up-regulate many plant-related or competitiveness genes. Such a minimal response of CT towards symbiosis is correlated with its small symbiotic gene content, which was taken as an advantage for the transfer of symbiosis to its pathogenic relative, Ralstonia solanacearum [127, 128].
Fig. 7

Illustration of the main functions regulated in BP, CT and RM in the presence of root exudates from Mimosa pudica. Functions commonly up-regulated among the 3 rhizobia, or specific, are colored differently (red: in all 3 rhizobia, orange: BP specific, green: CT specific, blue: BP & CT, purple: RM specific, brown: BP&RM. The cross indicates down-regulated functions

The only common response of the three rhizobia to RE is the up-regulation of the canonical nodulation genes, plus an additional fatty acid hydroxylase which is currently under investigation to prove its role in the symbiosis of Mimosa pudica. RM seems to up-regulate its nodulation gene content differently, with an earlier transcriptional response compared to BP and CT, and a decrease of expression after 1 h. RM also carries and up-regulates extra nod genes (nodM, nodO1, nodO2), illustrating a different history of nod gene acquisition and recruitment to symbiosis [10, 129]. RM also has many genes already described in alpha-rhizobia (such as in R. leguminosarum and S. meliloti) known to be playing a role in nodulation and/or competition in the rhizosphere (rhizopine catabolism, biofilm formation, motility …).

However, given we monitored one time point by RNAseq while kinetics by qPCR on nod genes revealed different induction profiles, we cannot exclude that these conclusions drawn from the RNAseq profiling might be different at earlier or later time-points of induction. In future work we will compare these alpha and beta-rhizobia at later steps of the symbiosis process to decipher the symbiotic specificities in both types of symbionts.


Mimosa pudica Root exudate (RE) preparation and composition

Mimosa pudica seeds (B&T world seeds; mix 50/50 var. hispida and unijuga) were surface-sterilized and germinated as described in [10]. For root exudation, M. pudica were grown in 48-deep well microplates (6 mL Storage Plate, Thermo Scientific ABgene) filled with 4 ml of sterile water per well and covered with adhesive PCR sealing foil (Thermo Scientific ABgene). After 1 week of incubation in a tropical plant growth chamber (30 °C, 80% humidity, 16 h day/8 h night) all liquid was recovered from wells, filtered (8 μm) and freeze-dried. The lyophilized RE powder obtained from 3 biological repetitions (4 microplates each) was dissolved in deionized water. A 1× final concentration of RE was used for bacterial treatments.

The composition of RE in terms of amino acids, organic acids (including cafeic, ferulic and sinapic acids) and soluble sugars was determined as previously described [130, 131] using a GC-MS Thermo TSQ Quantum with the silica column (Phenomenex ZB-5MSi (30 m, I.D. 0.25 mm, 0.25 μm)) with the following conditions: 60 to 350 °C gradient temperature during a 31 min run with an injection temperature of 230 °C. The concentration of polyphenols was estimated by the Folin colorimetric method [132]. The composition of RE is given in Table 4.

Bacterial strains and media

Bacterial strains (B. phymatum STM815 (BP), C. taiwanensis LMG19424 (CT), and R. mesoamericanum STM3625 (RM)) were routinely cultivated in YM culture medium [133] at 28 °C and stored at − 80 °C in 20% glycerol. For RNA-seq analysis, bacteria were grown in a modified minimal medium, VSG (mVSG) [134], containing potassium phosphate monobasic (2 g.L− 1), potassium phosphate dibasic (2 g.L− 1), calcium chloride (0.45 mM), iron(III), chloride (37 μM), sodium glutamate (1 g.L− 1), sodium succinate (2 g.L− 1) and vitamins (biotin 0.5 mg.L− 1, nicotinic acid 0.05 mg.L− 1, calcium pantothenate 0.05 mg.L− 1, pyridoxine 0.05 mg.L− 1, thiamine 0.05 mg.L− 1, myo-inositol 5 mg.L− 1). pH was adjusted to 6.8 and the medium was sterilized by autoclaving. BIOLOG analysis of substrate assimilation by the bacterial strains was performed according to the manufacturer (API GN20, Biolog Inc., Hayward, CA, USA).

Sampling of bacteria for RNA-seq

For each bacterium six liquid pre-cultures were grown in 10 mL of mVSG medium till mid-exponential phase and used to inoculate (at OD600nm of 0.1) six 1 L Erlenmeyer flasks containing 200 mL of mVSG medium. For the RE-induced condition, M. pudica root exudates were added after 1 h of growth (1X final concentration). Bacteria were cultivated for 25 h at 28 °C with shaking at 150 rpm, and were sampled at 1, 3, 4.5, 6, 9 and 12 h post inoculation (hpi). At every time point four replicates of each culture were sampled into tubes containing a volume of ice-cold stop buffer (5% phenol in ethanol as described in [135]), and centrifuged for 10 min at 4000 rpm at 4 °C. Supernatants were removed and cells were frozen in liquid nitrogen and stored at − 80 °C.

RNA preparation and sequencing

The mid-exponential growth phase was chosen for RNA sequencing, corresponding to 4.5 hpi for BP and CT and to 6 hpi for RM. Total RNA was purified using the RiboPure™ kit (Ambion, Austin, USA) following the manufacturer’s recommendations, then treated twice with DNAse I (Ambion), precipitated with ethanol and resuspended in RNAse-free water. Nine μg of total RNA was subjected to two successive rRNA removal steps using the Microbe Express™ kit (Ambion) to finally obtain 1 μg of enriched messenger RNA. The quantity and quality of total and enriched RNA were estimated using a NanoDrop ND-1000 and Bioanalyser Agilent 2100 (RNA nanochip), respectively (Additional file 1: Figure S13). All total and enriched messenger RNA samples were characterized by RIN values ranging from 9 to 10 indicating high quality.

Eighteen RNA samples (3 bacteria × 2 conditions × 3 biological repetitions) were sent for synthesis of corresponding cDNA libraries and sequencing to GATC Biotech (Konstanz, Germany). cDNA libraries were prepared according to the manufacturer’s instructions (Illumina TruSeq™ RNA Preparation) and labelled to sequence 2 to 3 libraries per lane (at least 50 million reads per library, read length of 36 bp) on a Genome Analyzer II (Illumina). Raw data sequences were submitted to ArrayExpress ( under accession: E-MTAB-4144.

Statistical and functional analysis of transcriptomic data

RNAseq reads were mapped to CoDing Sequences (CDS) of each genome (extracted from the Microscope annotation platform [136] using CLC-Genomics workbench v5 (CLC Bio, Aarhus, Denmark) with standard mapping options. Only unique mapped reads (UMR) were considered and used to calculate read counts per CDS. Ribosomal RNAs were discarded from next analyses. Estimates of variance-mean dependence in read counts per CDS data (RCD) from the three replicates for each condition, and tests for differential expression based on a model using the negative binomial distribution, were performed using DESeq (R module) [137, 138]. First, a normalization of read numbers for replicates in each condition was performed (by estimates of size factors in DESeq and normalization), then differential expression values (FoldChange) and statistical values (pVal and adjusted pVal) assessing the false discovery rate (FDR) were calculated. Linear regression between means of gene read counts by condition (basemeans) was used to control the bias of treatment effect. Scatter plots, drawn to show the dispersion of read count data for each bacterium, and R2 correlation coefficents (R2 of 0.861, 0.972 and 0.974 for BP, CT and RM, respectively), validated the robustness of our RNA manipulation (Additional file 1: Figure S14A). Box plots of transformed log10 (Read Counts) were used to compare the distributions of our RNAseq data between conditions and between strains and to visualize the effect of normalization.

RE-regulated genes were identified following a two step process. First, a statistical significance cut-off at pVal ≤0.01 was applied (see Additional file 1: Figure S14B, volcano plots of Log2 (Fold change) versus –Log10 (P-Values)). This pVal cut-off can be considered as necessary and sufficient to focus on differentially expressed genes [138], but given the high number of regulated genes we applied a second cut-off for fold change <− 1.5 (down-regulated) or > 1.5 (up-regulated).

For the functional analysis of our data, the analysis of differentially expressed CDS was coupled with a verification of their annotation (considering the context of surroundings regions) and detection of putative operons. The clustering of regulated genes into functional groups was performed using the Bioprocess classification in MicroScope [136], which is based on the Comprehensive Microbial Resource (CMR, [139].

Comparative genomics

Comparative genomics were performed using the Microscope Microbial Genome Annotation & Analysis Platform (MAGE, [140]). Best Bidirectional Hits (BBH) were used to establish the correspondence of orthologous and specific genes between genomes; and provides lists of orthologs between each pair on all three strains. Venn diagrams were used to represent the number of orthologs shared between each genome. The RCircos module (adapted from Circos software [141]) was used to represent shared regulated orthologs.

qPCR analyses

The primers used for quantitative PCR are listed in Additional file 1: Figure S5. The uppS and hisB genes were used as reference genes [43]. For qPCR, 0.5 μg of total RNA was reverse-transcribed with 200 U of Super-Script II (Invitrogen) using random decamer primers (2.5 μM, Ambion) at 25 °C for 10 min and then at 43 °C for 1 h. qPCR were performed on the Stratagene MXP3005P system using Power SYBR green master mix (Applied Biosystems). PCR started with a 10 min incubation at 95 °C, followed by 40 cycles consisting of 30 s at 95 °C and 30 s at 60 °C and 1 min at 72 °C. Primer specificity and the formation of primer dimers were checked by dissociation curves. PCR efficiency was calculated using the linear regression method on the log (fluorescence)-per-cycle-number data using Stratagene MXpro software. Prior to performing correlation analyses, the data were tested for normality using the Shapiro–Wilk test [142]. Because the data were not normally distributed, correlation analyses were done with Spearman’s Rho test [143].







Burkholderia phymatum STM815




Contact Dependent growth Inhibition system


Coding sequence


Comprehensive Microbial Resource


Cupriavidus taiwanensis LMG19424


Differentially expressed genes




Fold change










Mating pair formation


Nod factor




Symbiotic plasmid




Root Exudates


Rhizobium mesoamericanum STM3625




Type 3 secretion system


Type 4 secretion system


Type 6 secretion system


Toxin-Antitoxin system


Vincent Succinate Glutamate



We are grateful to Saïda Danoun and MetaToul plateforme for performing analyses of the composition of root exudates, and to Microscope MAGE plateform for help in genome annotation and analysis of RNAseq data.


This work was funded by French National Agency of Research (ANR) (Project “BETASYM” ANR-09-JCJC-0046). The funder had no role in the design of the study and collection, analysis, or interpretation of data and in writing the manuscript.

Availability of data and materials

The datasets generated and/or analysed during the current study are available in the Array Express repository ( under accession number E-MTAB-4144.

Authors’ contributions

AK: design, experimental work (RNA), functional analysis of RNAseq data, data interpretation, writing and data report in form of the manuscript; RM: experimental work (RNA), raw analysis of RNAseq data, data interpretation, data report; LMi: experimental work (RNA, qPCR), data interpretation; PT: experimental work (qPCR) and data interpretation, LM: design, raw and functional analysis of RNAseq data, data interpretation, writing and data report in form of the manuscript. All authors read and approved the manuscript.

Ethics approval

This study has not directly involved humans or animals. Plants used in this study were cultivated from seeds purchased at B&T World Seeds company (

Consent for publication

Not applicable

Competing interests

The authors declare that they have no competing interests

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

Authors’ Affiliations

IRD, Cirad, University of Montpellier, IPME, Montpellier, France
IRD, UMR LSTM, Campus de Baillarguet, Montpellier, France
Present address: Aix Marseille University, University of Avignon, CNRS, IRD, IMBE, Marseille, France


  1. Dénarié J, Debellé F, Promé JC. Rhizobium lipo-chitooligosaccharide nodulation factors: signaling molecules mediating recognition and morphogenesis. Annu Rev Biochem. 1996;65:503–35.PubMedView ArticleGoogle Scholar
  2. Debellé F, Moulin L, Mangin B, Dénarié J, Boivin C. Nod genes and nod signals and the evolution of the Rhizobium legume symbiosis. Acta Biochim Pol. 2001;48:359–65.PubMedGoogle Scholar
  3. Moulin L, Munive A, Dreyfus B, Boivin-Masson C. Nodulation of legumes by members of the beta-subclass of Proteobacteria. Nature. 2001;411:948–50.PubMedView ArticleGoogle Scholar
  4. Gyaneshwar P, Hirsch AM, Moulin L, Chen W-M, Elliott GN, Bontemps C, et al. Legume-nodulating betaproteobacteria: diversity, host range, and future prospects. Mol Plant-Microbe Interact. 2011;24:1276–88.PubMedView ArticleGoogle Scholar
  5. Moulin L, James EK, Klonowska A, Miana de Faria S, Simon MF. Phylogeny, diversity, geographical distribution, and host range of legume-nodulating betaproteobacteria : what is the role of plant taxonomy ? In: de Bruijn FJ. Biol. nitrogen Fixat. Chichester: Wiley; 2015. p. 177–190.Google Scholar
  6. Lemaire B, Dlodlo O, Chimphango S, Stirton C, Schrire B, Boatwright JS, et al. Symbiotic diversity, specificity and distribution of rhizobia in native legumes of the Core cape subregion (South Africa). FEMS Microbiol Ecol. 2015;91:1–17.PubMedView ArticleGoogle Scholar
  7. Lemaire B, Van Cauwenberghe J, Verstraete B, Chimphango S, Stirton C, Honnay O, et al. Characterization of the papilionoid-Burkholderia interaction in the Fynbos biome: the diversity and distribution of beta-rhizobia nodulating Podalyria calyptrata (Fabaceae, Podalyrieae). Syst Appl Microbiol. 2016;39:41–8.PubMedView ArticleGoogle Scholar
  8. Sawana A, Adeolu M, Gupta RS. Molecular signatures and phylogenomic analysis of the genus Burkholderia: proposal for division of this genus into the emended genus Burkholderia containing pathogenic organisms and a new genus Paraburkholderia gen. Nov. harboring environmental species. Front Genet. 2014;5:429.PubMedPubMed CentralView ArticleGoogle Scholar
  9. Elliott GN, Chen W-M, Chou J-H, Wang H-C, Sheu S-Y, Perin L, et al. Burkholderia phymatum is a highly effective nitrogen-fixing symbiont of Mimosa spp. and fixes nitrogen ex planta. New Phytol. 2007;173:168–80.PubMedView ArticleGoogle Scholar
  10. Mishra RPN, Tisseyre P, Melkonian R, Chaintreuil C, Miché L, Klonowska A, et al. Genetic diversity of Mimosa pudica rhizobial symbionts in soils of French Guiana: investigating the origin and diversity of Burkholderia phymatum and other beta-rhizobia. FEMS Microbiol Ecol. 2012;79:487–503.PubMedView ArticleGoogle Scholar
  11. Chen W-M, James EK, Coenye T, Chou J-H, Barrios E, de Faria SM, et al. Burkholderia mimosarum sp. nov., isolated from root nodules of Mimosa spp. from Taiwan and South America. Int J Syst Evol Microbiol. 2006;56:1847–51.PubMedView ArticleGoogle Scholar
  12. Chen W-M, de Faria SM, James EK, Elliott GN, Lin K-Y, Chou J-H, et al. Burkholderia nodosa sp. nov., isolated from root nodules of the woody Brazilian legumes Mimosa bimucronata and Mimosa scabrella. Int J Syst Evol Microbiol. 2007;57:1055–9.PubMedView ArticleGoogle Scholar
  13. Dall’Agnol RF, Bournaud C, de Faria SM, Béna G, Moulin L, Hungria M. Genetic diversity of symbiotic Paraburkholderia species isolated from nodules of Mimosa pudica (L.) and Phaseolus vulgaris (L.) grown in soils of the Brazilian Atlantic Forest (Mata Atlântica). FEMS Microbiol Ecol. 2017;93Google Scholar
  14. Chen W-M, de Faria SM, Chou J-H, James EK, Elliott GN, Sprent JI, et al. Burkholderia sabiae sp. nov., isolated from root nodules of Mimosa caesalpiniifolia. Int J Syst Evol Microbiol. 2008;58:2174–9.PubMedView ArticleGoogle Scholar
  15. Sheu S-Y, Chou J-H, Bontemps C, Elliott GN, Gross E, James EK, et al. Burkholderia symbiotica sp. nov., isolated from root nodules of Mimosa spp. native to north East Brazil. Int J Syst Evol Microbiol. 2012;62:2272–8.PubMedView ArticleGoogle Scholar
  16. Sheu S, Chou J, Bontemps C, Elliott G, Gross E, Dos Reis JF, et al. Burkholderia diazotrophica sp. nov., isolated from root nodules of Mimosa spp. Int J Syst Evol Microbiol. 2013;63:435–41.PubMedView ArticleGoogle Scholar
  17. Bournaud C, Moulin L, Cnockaert M, de Faria S, Prin Y, Severac D, et al. Paraburkholderia piptadeniae sp. nov. and Paraburkholderia ribeironis sp. nov., two root-nodulating symbiotic species of Piptadenia gonoacantha in Brazil. Int J Syst Evol Microbiol. 2017;67:432–40.PubMedView ArticleGoogle Scholar
  18. Barrett CF, Parker MA. Coexistence of Burkholderia, Cupriavidus, and Rhizobium sp. nodule bacteria on two Mimosa spp. in Costa Rica. Appl Environ Microbiol. 2006;72:1198–206.PubMedPubMed CentralView ArticleGoogle Scholar
  19. Andam CP, Mondo SJ, Parker MA. Monophyly of nodA and nifH genes across Texan and Costa Rican populations of Cupriavidus nodule symbionts. Appl Environ Microbiol. 2007;73:4686–90.PubMedPubMed CentralView ArticleGoogle Scholar
  20. Liu X, Wei S, Wang F, James EK, Guo X, Zagar C, et al. Burkholderia and Cupriavidus spp. are the preferred symbionts of Mimosa spp. in southern China. FEMS Microbiol Ecol. 2012;80:417–26.PubMedView ArticleGoogle Scholar
  21. Chen WM, Laevens S, Lee TM, Coenye T, De Vos P, Mergeay M, et al. Ralstonia taiwanensis sp. nov., isolated from root nodules of Mimosa species and sputum of a cystic fibrosis patient. Int J Syst Evol Microbiol. 2001;51:1729–35.PubMedView ArticleGoogle Scholar
  22. Andrus AD, Andam C, Parker MA. American origin of Cupriavidus bacteria associated with invasive Mimosa legumes in the Philippines. FEMS Microbiol Ecol. 2012;80:747–50.PubMedView ArticleGoogle Scholar
  23. Klonowska A, Chaintreuil C, Tisseyre P, Miché L, Melkonian R, Ducousso M, et al. Biodiversity of Mimosa pudica rhizobial symbionts (Cupriavidus taiwanensis, Rhizobium mesoamericanum) in New Caledonia and their adaptation to heavy metal-rich soils. FEMS Microbiol Ecol. 2012;81:618–35.PubMedView ArticleGoogle Scholar
  24. Gehlot HS, Tak N, Kaushik M, Mitra S, Chen W-M, Poweleit N, et al. An invasive Mimosa in India does not adopt the symbionts of its native relatives. Ann Bot. 2013;112:179–96.PubMedPubMed CentralView ArticleGoogle Scholar
  25. Platero R, James EK, Rios C, Iriarte A, Sandes L, Zabaleta M, et al. Novel Cupriavidus strains isolated from root nodules of native Uruguayan mimosa species. Voordouw G. Appl Environ Microbiol 2016;82:3150–3164.Google Scholar
  26. Taulé C, Zabaleta M, Mareque C, Platero R, Sanjurjo L, Sicardi M, et al. New betaproteobacterial rhizobium strains able to efficiently nodulate Parapiptadenia rigida (Benth.) Brenan. Appl Environ Microbiol. 2012;78:1692–700.PubMedPubMed CentralView ArticleGoogle Scholar
  27. Turner SL, Young JP. The glutamine synthetases of rhizobia: phylogenetics and evolutionary implications. Mol Biol Evol. 2000;17:309–19.PubMedView ArticleGoogle Scholar
  28. Hirsch AM, Lum MR, Downie JA. What makes the rhizobia-legume symbiosis so special? Plant Physiol. 2001;127:1484–92.PubMedPubMed CentralView ArticleGoogle Scholar
  29. Moulin L, Béna G, Boivin-Masson C, Stępkowski T. Phylogenetic analyses of symbiotic nodulation genes support vertical and lateral gene co-transfer within the Bradyrhizobium genus. Mol Phylogenet Evol. 2004;30:720–32.PubMedView ArticleGoogle Scholar
  30. Bontemps C, Elliott GN, Simon MF, Dos Reis Júnior FB, Gross E, Lawton RC, et al. Burkholderia species are ancient symbionts of legumes. Mol Ecol. 2010;19:44–52.PubMedView ArticleGoogle Scholar
  31. Amadou C, Pascal G, Mangenot S, Glew M, Bontemps C, Capela D, et al. Genome sequence of the beta-rhizobium Cupriavidus taiwanensis and comparative genomics of rhizobia. Genome Res. 2008;18:1472–83.PubMedPubMed CentralView ArticleGoogle Scholar
  32. Barrett CF, Parker MA. Prevalence of Burkholderia sp. nodule symbionts on four mimosoid legumes from Barro Colorado Island, Panama. Syst Appl Microbiol. 2005;28:57–65.PubMedView ArticleGoogle Scholar
  33. Chen W, Moulin L, Bontemps C, Vandamme P, Béna G, Boivin-Masson C. Legume symbiotic nitrogen fixation by β-Proteobacteria is widespread in nature. J Bacteriol. 2003;185:7266–72.PubMedPubMed CentralView ArticleGoogle Scholar
  34. Elliott GN, Chou J-H, Chen W-M, Bloemberg GV, Bontemps C, Martínez-Romero E, et al. Burkholderia spp. are the most competitive symbionts of Mimosa, particularly under N-limited conditions. Environ Microbiol. 2009;11:762–78.PubMedView ArticleGoogle Scholar
  35. Melkonian R, Moulin L, Béna G, Tisseyre P, Chaintreuil C, Heulin K, et al. The geographical patterns of symbiont diversity in the invasive legume Mimosa pudica can be explained by the competitiveness of its symbionts and by the host genotype. Environ Microbiol. 2014;16:2099–111.PubMedView ArticleGoogle Scholar
  36. López-López A, Rogel-Hernández MA, Barois I, Ortiz Ceballos AI, Martínez J, Ormeño-Orrillo E, et al. Rhizobium grahamii sp. nov. from Dalea leporina, Leucaena leucocephala, Clitoria ternatea nodules, and Rhizobium mesoamericanum sp. nov. from Phaseolus vulgaris, siratro, cowpea and Mimosa pudica nodules. Int J Syst Evol Microbiol. 2012;62:2264–71.PubMedView ArticleGoogle Scholar
  37. Baraúna AC, Rouws LFM, Simoes-Araujo JL, Dos Reis Junior FB, Iannetta PPM, Maluk M, et al. Rhizobium altiplani sp. nov., isolated from effective nodules on Mimosa pudica growing in untypically alkaline soil in central Brazil. Int J Syst Evol Microbiol. 2016;66:4118–24.PubMedView ArticleGoogle Scholar
  38. Masson-Boivin C, Giraud E, Perret X, Batut J. Establishing nitrogen-fixing symbiosis with legumes: how many rhizobium recipes? Trends Microbiol. 2009;17:458–66.PubMedView ArticleGoogle Scholar
  39. Malek W. The role of motility in the efficiency of nodulation by Rhizobium meliloti. Arch Microbiol. 1992;158:26–8.View ArticleGoogle Scholar
  40. Lodeiro AR, López-García SL, Vázquez TE, Favelukes G. Stimulation of adhesiveness, infectivity, and competitiveness for nodulation of Bradyrhizobium japonicum by its pretreatment with soybean seed lectin. FEMS Microbiol Lett. 2000;188:177–84.PubMedView ArticleGoogle Scholar
  41. Lagares A, Caetano-Anollés G, Niehaus K, Lorenzen J, Ljunggren HD, Pühler A, et al. A Rhizobium meliloti lipopolysaccharide mutant altered in competitiveness for nodulation of alfalfa. J Bacteriol. 1992;174:5941–52.PubMedPubMed CentralView ArticleGoogle Scholar
  42. Perret X, Staehelin C, Broughton WJ. Molecular basis of symbiotic promiscuity. Microbiol Mol Biol Rev. 2000;64:180–201.PubMedPubMed CentralView ArticleGoogle Scholar
  43. Ampe F, Kiss E, Sabourdy F, Batut J. Transcriptome analysis of Sinorhizobium meliloti during symbiosis. Genome Biol. 2003;4:R15.PubMedPubMed CentralView ArticleGoogle Scholar
  44. Mark GL, Dow JM, Kiely PD, Higgins H, Haynes J, Baysse C, et al. Transcriptome profiling of bacterial responses to root exudates identifies genes involved in microbe-plant interactions. Proc Natl Acad Sci U S A. 2005;102:17454–9.PubMedPubMed CentralView ArticleGoogle Scholar
  45. Fan B, Carvalhais LC, Becker A, Fedoseyenko D, von Wirén N, Borriss R. Transcriptomic profiling of Bacillus amyloliquefaciens FZB42 in response to maize root exudates. BMC Microbiol. 2012;12:116.PubMedPubMed CentralView ArticleGoogle Scholar
  46. Shidore T, Dinse T, Öhrlein J, Becker A, Reinhold-Hurek B. Transcriptomic analysis of responses to exudates reveal genes required for rhizosphere competence of the endophyte Azoarcus sp. strain BH72. Environ Microbiol. 2012;14:2775–87.PubMedView ArticleGoogle Scholar
  47. Oliveira LR, Marcelino FC, Barcellos FG, Rodrigues EP, Megías M, Hungria M. The nodC, nodG, and glgX genes of Rhizobium tropici strain PRF 81. Funct Integr Genomics. 2010;10:425–31.PubMedView ArticleGoogle Scholar
  48. Perret X, Freiberg C, Rosenthal A, Broughton WJ, Fellay R. High-resolution transcriptional analysis of the symbiotic plasmid of Rhizobium sp. NGR234. Mol Microbiol. 1999;32:415–25.PubMedView ArticleGoogle Scholar
  49. Ramachandran VK, East AK, Karunakaran R, Downie JA, Poole PS. Adaptation of Rhizobium leguminosarum to pea, alfalfa and sugar beet rhizospheres investigated by comparative transcriptomics. Genome Biol. 2011;12:R106.PubMedPubMed CentralView ArticleGoogle Scholar
  50. Janssen PJ, van Houdt R, Moors H, Monsieurs P, Morin N, Michaux A, et al. The complete genome sequence of Cupriavidus metallidurans strain CH34, a master survivalist in harsh and anthropogenic environments. PLoS One. 2010;5Google Scholar
  51. Monchy S, Benotmane MA, Janssen P, Vallaeys T, Taghavi S, van der Lelie D, et al. Plasmids pMOL28 and pMOL30 of Cupriavidus metallidurans are specialized in the maximal viable response to heavy metals. J Bacteriol. 2007;189:7417–25.PubMedPubMed CentralView ArticleGoogle Scholar
  52. Rainey PB. Adaptation of Pseudomonas fluorescens to the plant rhizosphere. Environ Microbiol. 1999;1:243–57.PubMedView ArticleGoogle Scholar
  53. Zhang X-X, Rainey PB. The role of a P1-type ATPase from Pseudomonas fluorescens SBW25 in copper homeostasis and plant colonization. Mol Plant-Microbe Interact. 2007;20:581–8.PubMedView ArticleGoogle Scholar
  54. Bains J, Kaufman L, Farnell B, Boulanger MJ. A product analog bound form of 3-oxoadipate-enol-lactonase (PcaD) reveals a multifunctional role for the divergent cap domain. J Mol Biol. 2011;406:649–58.PubMedView ArticleGoogle Scholar
  55. Garcia-Fraile P, Seaman JC, Karunakaran R, Edwards A, Poole PS, Downie JA. Arabinose and protocatechuate catabolism genes are important for growth of Rhizobium leguminosarum biovar viciae in the pea rhizosphere. Plant Soil. 2015;390:251–64.PubMedPubMed CentralView ArticleGoogle Scholar
  56. Rohmer M, Knani M, Simonin P, Sutter B, Sahm H. Isoprenoid biosynthesis in bacteria: a novel pathway for the early steps leading to isopentenyl diphosphate. Biochem J. 1993;295(Pt 2):517–24.PubMedPubMed CentralView ArticleGoogle Scholar
  57. Ding H, Yip CB, Geddes BA, Oresnik IJ, Hynes MF. Glycerol utilization by Rhizobium leguminosarum requires an ABC transporter and affects competition for nodulation. Microbiology. 2012;158:1369–78.PubMedView ArticleGoogle Scholar
  58. Chen W-M, Prell J, James EK, Sheu D-S, Sheu S-Y. Biosynthesis of branched-chain amino acids is essential for effective symbioses between betarhizobia and Mimosa pudica. Microbiology. 2012;158:1758–66.PubMedView ArticleGoogle Scholar
  59. Anraku Y, Gennis RB. The aerobic respiratory chain of Escherichia coli. Trends Biochem Sci. 1987;12:262–6.View ArticleGoogle Scholar
  60. Tucker NP, Hicks MG, Clarke TA, Crack JC, Chandra G, Le Brun NE, et al. The transcriptional repressor protein NsrR senses nitric oxide directly via a [2Fe-2S] cluster. PLoS One. 2008;3:e3623.PubMedPubMed CentralView ArticleGoogle Scholar
  61. del Giudice J, Cam Y, Damiani I, Fung-Chat F, Meilhoc E, Bruand C, et al. Nitric oxide is required for an optimal establishment of the Medicago truncatula-Sinorhizobium meliloti symbiosis. New Phytol. 2011;191:405–17.PubMedPubMed CentralView ArticleGoogle Scholar
  62. Espinosa-Urgel M, Salido A, Ramos JL. Genetic analysis of functions involved in adhesion of Pseudomonas putida to seeds. J Bacteriol. 2000;182:2363–9.PubMedPubMed CentralView ArticleGoogle Scholar
  63. Martinez JL, Sánchez MB, Martínez-Solano L, Hernandez A, Garmendia L, Fajardo A, et al. Functional role of bacterial multidrug efflux pumps in microbial natural ecosystems. FEMS Microbiol Rev. 2009;33:430–49.PubMedView ArticleGoogle Scholar
  64. González-Pasayo R, Martínez-Romero E. Multiresistance genes of Rhizobium etli CFN42. Mol Plant-Microbe Interact. 2000;13:572–7.PubMedView ArticleGoogle Scholar
  65. Lindemann A, Koch M, Pessi G, Müller AJ, Balsiger S, Hennecke H, et al. Host-specific symbiotic requirement of BdeAB, a RegR-controlled RND-type efflux system in Bradyrhizobium japonicum. FEMS Microbiol Lett. 2010;312:184–91.PubMedView ArticleGoogle Scholar
  66. Kohler PRA, Choong E-L, Rossbach S. The RpiR-like repressor IolR regulates inositol catabolism in Sinorhizobium meliloti. J Bacteriol. 2011;193:5155–63.PubMedPubMed CentralView ArticleGoogle Scholar
  67. Kohler PRA, Zheng JY, Schoffers E, Rossbach S. Inositol catabolism, a key pathway in Sinorhizobium meliloti for competitive host nodulation. Appl Environ Microbiol. 2010;76:7972–80.PubMedPubMed CentralView ArticleGoogle Scholar
  68. Summers ML, Denton MC, McDermott TR. Genes coding for phosphotransacetylase and acetate kinase in Sinorhizobium meliloti are in an operon that is inducible by phosphate stress and controlled by phoB. J Bacteriol. 1999;181:2217–24.PubMedPubMed CentralGoogle Scholar
  69. Prell J, Bourdès A, Karunakaran R, Lopez-Gomez M, Poole P. Pathway of gamma-aminobutyrate metabolism in Rhizobium leguminosarum 3841 and its role in symbiosis. J Bacteriol. 2009;191:2177–86.PubMedPubMed CentralView ArticleGoogle Scholar
  70. Yasuta T, Okazaki S, Mitsui H, Yuhashi K-I, Ezura H, Minamisawa K. DNA sequence and mutational analysis of rhizobitoxine biosynthesis genes in Bradyrhizobium elkanii. Appl Environ Microbiol. 2001;67:4999–5009.PubMedPubMed CentralView ArticleGoogle Scholar
  71. Okazaki S, Sugawara M, Minamisawa K. Bradyrhizobium elkanii rtxC gene is required for expression of symbiotic phenotypes in the final step of rhizobitoxine biosynthesis. Appl Environ Microbiol. 2004;70:535–41.PubMedPubMed CentralView ArticleGoogle Scholar
  72. Sugawara M, Okazaki S, Nukui N, Ezura H, Mitsui H, Minamisawa K. Rhizobitoxine modulates plant-microbe interactions by ethylene inhibition. Biotechnol Adv. 2006;24:382–8.PubMedView ArticleGoogle Scholar
  73. Ma W, Charles TC, Glick BR. Expression of an exogenous 1-aminocyclopropane-1-carboxylate deaminase gene in Sinorhizobium meliloti increases its ability to nodulate alfalfa. Appl Environ Microbiol. 2004;70:5891–7.PubMedPubMed CentralView ArticleGoogle Scholar
  74. Theunis M, Kobayashi H, Broughton WJ, Prinsen E. Flavonoids, NodD1, NodD2, and nod-box NB15 modulate expression of the y4wEFG locus that is required for indole-3-acetic acid synthesis in Rhizobium sp. strain NGR234. Mol Plant-Microbe Interact. 2004;17:1153–61.PubMedView ArticleGoogle Scholar
  75. Camerini S, Senatore B, Lonardo E, Imperlini E, Bianco C, Moschetti G, et al. Introduction of a novel pathway for IAA biosynthesis to rhizobia alters vetch root nodule development. Arch Microbiol. 2008;190:67–77.PubMedView ArticleGoogle Scholar
  76. Woo Park Y, Tech Lim S, Dae YH. Cloning and characterization of a CMCase gene, celB, of Erwinia carotovora subsp. carotovora LY34 and its comparison to celA. Mol Cells. 1998;8:280–5.PubMedGoogle Scholar
  77. Robledo M, Jiménez-Zurdo JI, Velázquez E, Trujillo ME, Zurdo-Piñeiro JL, Ramírez-Bahena MH, et al. Rhizobium cellulase CelC2 is essential for primary symbiotic infection of legume host roots. Proc Natl Acad Sci U S A. 2008;105:7064–9.PubMedPubMed CentralView ArticleGoogle Scholar
  78. Robledo M, Jiménez-Zurdo JI, Soto MJ, Velázquez E, Dazzo F, Martínez-Molina E, et al. Development of functional symbiotic white clover root hairs and nodules requires tightly regulated production of rhizobial cellulase CelC2. Mol Plant-Microbe Interact. 2011;24:798–807.PubMedView ArticleGoogle Scholar
  79. Angus AA, Agapakis CM, Fong S, Yerrapragada S, Estrada-de los Santos P, Yang P, et al. Plant-associated symbiotic Burkholderia species lack hallmark strategies required in mammalian pathogenesis. PLoS One. 2014;9:e83779.PubMedPubMed CentralView ArticleGoogle Scholar
  80. Mitter B, Petric A, Shin MW, Chain PSG, Hauberg-Lotte L, Reinhold-Hurek B, et al. Comparative genome analysis of Burkholderia phytofirmans PsJN reveals a wide spectrum of endophytic lifestyles based on interaction strategies with host plants. Front Plant Sci Frontiers. 2013;4:120.Google Scholar
  81. Nierman WC, DeShazer D, Kim HS, Tettelin H, Nelson KE, Feldblyum T, et al. Structural flexibility in the Burkholderia mallei genome. Proc Natl Acad Sci U S A. 2004;101:14246–51.PubMedPubMed CentralView ArticleGoogle Scholar
  82. Holden MTG, Titball RW, Peacock SJ, Cerdeño-Tárraga AM, Atkins T, Crossman LC, et al. Genomic plasticity of the causative agent of melioidosis, Burkholderia pseudomallei. Proc Natl Acad Sci U S A. 2004;101:14240–5.PubMedPubMed CentralView ArticleGoogle Scholar
  83. Seo Y-S, Lim J, Choi B-S, Kim H, Goo E, Lee B, et al. Complete genome sequence of Burkholderia gladioli BSR3. J Bacteriol. 2011;193:3149.PubMedPubMed CentralView ArticleGoogle Scholar
  84. Khan A, Asif H, Studholme DJ, Khan IA, Azim MK. Genome characterization of a novel Burkholderia cepacia complex genomovar isolated from dieback affected mango orchards. World J Microbiol Biotechnol. 2013;29:2033–44.PubMedView ArticleGoogle Scholar
  85. Cascales E, Cambillau C. Structural biology of type VI secretion systems. Philos Trans R Soc Lond Ser B Biol Sci. 2012;367:1102–11.View ArticleGoogle Scholar
  86. Bernal P, Allsopp LP, Filloux A, Llamas MA. The Pseudomonas putida T6SS is a plant warden against phytopathogens. ISME J. 2017;11:972–87.PubMedPubMed CentralView ArticleGoogle Scholar
  87. Anderson MC, Vonaesch P, Saffarian A, Marteyn BS, Sansonetti PJ. Shigella sonnei encodes a functional T6SS used for interbacterial competition and niche occupancy. Cell Host Microbe. 2017;21:769–776.e3.PubMedView ArticleGoogle Scholar
  88. Lardi M, de Campos SB, Purtschert G, Eberl L, Pessi G. Competition experiments for legume infection identify Burkholderia phymatum as a highly competitive β-Rhizobium. Front Microbiol. 2017;8:1527.PubMedPubMed CentralView ArticleGoogle Scholar
  89. Nelson MS, Sadowsky MJ. Secretion systems and signal exchange between nitrogen-fixing rhizobia and legumes. Front Plant Sci. 2015;6:491.PubMedPubMed CentralView ArticleGoogle Scholar
  90. Hubber AM, Sullivan JT, Ronson CW. Symbiosis-induced cascade regulation of the Mesorhizobium loti R7A VirB/D4 type IV secretion system. Mol Plant-Microbe Interact. 2007;20:255–61.PubMedView ArticleGoogle Scholar
  91. Sugawara M, Epstein B, Badgley BD, Unno T, Xu L, Reese J, et al. Comparative genomics of the core and accessory genomes of 48 Sinorhizobium strains comprising five genospecies. Genome Biol. 2013;14:R17.PubMedPubMed CentralView ArticleGoogle Scholar
  92. Bladergroen MR, Badelt K, Spaink HP. Infection-blocking genes of a symbiotic Rhizobium leguminosarum strain that are involved in temperature-dependent protein secretion. Mol Plant-Microbe Interact. 2003;16:53–64.PubMedView ArticleGoogle Scholar
  93. Voth DE, Broederdorf LJ, Graham JG. Bacterial type IV secretion systems: versatile virulence machines. Future Microbiol. 2012;7:241–57.PubMedPubMed CentralView ArticleGoogle Scholar
  94. Fernández-López R, Garcillán-Barcia MP, Revilla C, Lázaro M, Vielva L, de la Cruz F. Dynamics of the IncW genetic backbone imply general trends in conjugative plasmid evolution. FEMS Microbiol Rev. 2006;30:942–66.PubMedView ArticleGoogle Scholar
  95. de Keyzer J, van der Does C, Driessen AJM. The bacterial translocase: a dynamic protein channel complex. Cell Mol Life Sci. 2003;60:2034–52.PubMedView ArticleGoogle Scholar
  96. Burg-Golani T, Pozniak Y, Rabinovich L, Sigal N, Nir Paz R, Herskovits AA. Membrane chaperone SecDF plays a role in the secretion of Listeria monocytogenes major virulence factors. J Bacteriol. 2013;195:5262–72.PubMedPubMed CentralView ArticleGoogle Scholar
  97. Vörös A, Simm R, Slamti L, McKay MJ, Hegna IK, Nielsen-LeRoux C, et al. SecDF as part of the sec-translocase facilitates efficient secretion of Bacillus cereus toxins and cell wall-associated proteins. Freitag NE, editor. PLoS One. 2014;9:e103326.Google Scholar
  98. Coutinho BG, Mitter B, Talbi C, Sessitsch A, Bedmar EJ, Halliday N, et al. Regulon studies and in planta role of the BraI/R quorum-sensing system in the plant-beneficial Burkholderia cluster. Appl Environ Microbiol. 2013;79:4421–32.PubMedPubMed CentralView ArticleGoogle Scholar
  99. Garcia EC, Perault AI, Marlatt SA, Cotter PA. Interbacterial signaling via Burkholderia contact-dependent growth inhibition system proteins. Proc Natl Acad Sci. 2016;113:8296–301.PubMedPubMed CentralView ArticleGoogle Scholar
  100. Mattick JS. Type IV Pili and twitching motility. Annu Rev Microbiol. 2002;56:289–314.PubMedView ArticleGoogle Scholar
  101. Giltner CL, Habash M, Burrows LL. Pseudomonas Aeruginosa Minor pilins are incorporated into type IV pili. J Mol Biol. 2010;398:444–61.PubMedView ArticleGoogle Scholar
  102. Hélaine S, Carbonnelle E, Prouvensier L, Beretti J-L, Nassif X, Pelicic V. PilX, a pilus-associated protein essential for bacterial aggregation, is a key to pilus-facilitated attachment of Neisseria meningitidis to human cells. Mol Microbiol. 2004;55:65–77.View ArticleGoogle Scholar
  103. Bodogai M, Ferenczi S, Bashtovyy D, Miclea P, Papp P, Dusha I. The ntrPR operon of Sinorhizobium meliloti is organized and functions as a toxin-antitoxin module. Mol Plant-Microbe Interact. 2006;19:811–22.PubMedView ArticleGoogle Scholar
  104. Miclea PS, Péter M, Végh G, Cinege G, Kiss E, Váró G, et al. Atypical transcriptional regulation and role of a new toxin-antitoxin-like module and its effect on the lipid composition of Bradyrhizobium japonicum. Mol Plant-Microbe Interact. 2010;23:638–50.PubMedView ArticleGoogle Scholar
  105. Trainer MA, Charles TC. The role of PHB metabolism in the symbiosis of rhizobia with legumes. Appl Microbiol Biotechnol. 2006;71:377–86.PubMedView ArticleGoogle Scholar
  106. Vereecke D, Cornelis K, Temmerman W, Jaziri M, Van Montagu M, Holsters M, et al. Chromosomal locus that affects pathogenicity of Rhodococcus fascians. J Bacteriol. 2002;184:1112–20.PubMedPubMed CentralView ArticleGoogle Scholar
  107. Tamir-Ariel D, Navon N, Burdman S. Identification of genes in Xanthomonas campestris pv. vesicatoria induced during its interaction with tomato. J Bacteriol. 2007;189:6359–71.PubMedPubMed CentralView ArticleGoogle Scholar
  108. Dunn MF, Ramirez-Trujillo JA, Hernandez-Lucas I. Major roles of isocitrate lyase and malate synthase in bacterial and fungal pathogenesis. Microbiology. 2009;155:3166–75.PubMedView ArticleGoogle Scholar
  109. Ramírez-Trujillo JA, Encarnación S, Salazar E, de los Santos AG, Dunn MF, Emerich DW, et al. Functional characterization of the Sinorhizobium meliloti acetate metabolism genes aceA, SMc00767, and glcB. J Bacteriol. 2007;189:5875–84.PubMedPubMed CentralView ArticleGoogle Scholar
  110. Ratcliff WC, Denison RF. Bacterial persistence and bet hedging in Sinorhizobium meliloti. Commun Integr Biol. 2011;4:98–100.PubMedPubMed CentralView ArticleGoogle Scholar
  111. Strube K, de Vries S, Cramm R. Formation of a dinitrosyl iron complex by NorA, a nitric oxide-binding di-iron protein from Ralstonia eutropha H16. J Biol Chem. 2007;282:20292–300.PubMedView ArticleGoogle Scholar
  112. Boscari A, Meilhoc E, Castella C, Bruand C, Puppo A, Brouquisse R. Which role for nitric oxide in symbiotic N2-fixing nodules: toxic by-product or useful signaling/metabolic intermediate? Front Plant Sci. 2013;4:384.PubMedPubMed CentralView ArticleGoogle Scholar
  113. Saad MM, Crèvecoeur M, Masson-Boivin C, Perret X. The type 3 protein secretion system of Cupriavidus taiwanensis strain LMG19424 compromises symbiosis with Leucaena leucocephala. Appl Environ Microbiol. 2012;78:7476–9.PubMedPubMed CentralView ArticleGoogle Scholar
  114. Marie C, Barny MA, Downie JA. Rhizobium leguminosarum has two glucosamine synthases, GlmS and NodM, required for nodulation and development of nitrogen-fixing nodules. Mol Microbiol. 1992;6:843–51.PubMedView ArticleGoogle Scholar
  115. Kamst E, Breek CK, Spaink HP. Functional analysis of chimeras derived from the Sinorhizobium meliloti and Mesorhizobium loti nodC genes identifies regions controlling chitin oligosaccharide chain length. Mol Gen Genet. 2000;264:75–81.PubMedView ArticleGoogle Scholar
  116. Miwa H, Sun J, Oldroyd GED, Downie JA. Analysis of nod-factor-induced calcium signaling in root hairs of symbiotically defective mutants of Lotus japonicus. Mol Plant-Microbe Interact. 2006;19:914–23.PubMedView ArticleGoogle Scholar
  117. Finnie C, Hartley NM, Findlay KC, Downie JA. The Rhizobium leguminosarum prsDE genes are required for secretion of several proteins, some of which influence nodulation, symbiotic nitrogen fixation and exopolysaccharide modification. Mol Microbiol. 1997;25:135–46.PubMedView ArticleGoogle Scholar
  118. Caetano-Anollés G, Wall LG, De Micheli AT, Macchi EM, Bauer WD, Favelukes G. Role of motility and chemotaxis in efficiency of nodulation by Rhizobium meliloti. Plant Physiol. 1988;86:1228.PubMedPubMed CentralView ArticleGoogle Scholar
  119. Mimmack ML, Borthakur D, Jones MA, Downie JA, AWB J. The psi operon of Rhizobium leguminosarum biovar phaseoli: identification of two genes whose products are located at the bacterial cell surface. Microbiology. 1994;140:1223–9.PubMedView ArticleGoogle Scholar
  120. Fraysse N, Couderc F, Poinsot V. Surface polysaccharide involvement in establishing the rhizobium-legume symbiosis. Eur J Biochem. 2003;270:1365–80.PubMedView ArticleGoogle Scholar
  121. Skorupska A, Janczarek M, Marczak M, Mazur A, Król J. Rhizobial exopolysaccharides: genetic control and symbiotic functions. Microb Cell Factories. 2006;5:7.View ArticleGoogle Scholar
  122. Saint CP, Wexler M, Murphy PJ, Tempé J, Tate ME, Murphy PJ. Characterization of genes for synthesis and catabolism of a new rhizopine induced in nodules by Rhizobium meliloti Rm220-3: extension of the rhizopine concept. J Bacteriol. 1993;175:5205–15.PubMedPubMed CentralView ArticleGoogle Scholar
  123. Bahar M, de Majnik J, Wexler M, Fry J, Poole PS, Murphy PJ. A model for the catabolism of rhizopine in Rhizobium leguminosarum involves a ferredoxin oxygenase complex and the inositol degradative pathway. Mol Plant-Microbe Interact. 1998;11:1057–68.PubMedView ArticleGoogle Scholar
  124. Poole PS, Blyth A, Reid CJ, Walters K. Myo-Inositol catabolism and catabolite regulation in Rhizobium leguminosarum bv. viciae. Microbiology. 1994;140:2787–95.View ArticleGoogle Scholar
  125. Gordon DM, Ryder MH, Heinrich K, Murphy PJ. An experimental test of the rhizopine concept in Rhizobium meliloti. Appl Environ Microbiol. 1996;62:3991–6.PubMedPubMed CentralGoogle Scholar
  126. Balleza D, Gómez-Lagunas F, Quinto C. Cloning and functional expression of an MscL ortholog from Rhizobium etli: characterization of a mechanosensitive channel. J Membr Biol. 2010;234:13–27.PubMedView ArticleGoogle Scholar
  127. Marchetti M, Capela D, Glew M, Cruveiller S, Chane-Woon-Ming B, Gris C, et al. Experimental evolution of a plant pathogen into a legume symbiont. PLoS Biol. 2010;8:e1000280.PubMedPubMed CentralView ArticleGoogle Scholar
  128. Marchetti M, Jauneau A, Capela D, Remigi P, Gris C, Batut J, et al. Shaping bacterial symbiosis with legumes by experimental evolution. Mol Plant-Microbe Interact. 2014;27:956–64.PubMedView ArticleGoogle Scholar
  129. De Meyer SE, Briscoe L, Martínez-Hidalgo P, Agapakis CM, de Los Santos PE, Seshadri R, et al. Symbiotic Burkholderia species show diverse arrangements of nif/fix and nod genes and lack typical high-affinity cytochrome cbb3 oxidase genes. Mol Plant-Microbe Interact. 2016;29:609–19.PubMedView ArticleGoogle Scholar
  130. Sobolevsky T, Chernetsova E, Revelsky A, Revelsky I, Starostin A, Miller B, et al. Electron ionization mass spectra and their reproducibility for trialkylsilylated derivatives of organic acids, sugars and alcohols. Eur J Mass Spectrom. 2003;9:487.View ArticleGoogle Scholar
  131. Ruiz-Matute AI, Hernández-Hernández O, Rodríguez-Sánchez S, Sanz ML, Martínez-Castro I. Derivatization of carbohydrates for GC and GC–MS analyses. J Chromatogr B. 2011;879:1226–40.View ArticleGoogle Scholar
  132. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem. 1951;193:265–75.PubMedGoogle Scholar
  133. Vincent J. A manual for the pratical study of root-nodule bacteria. I.B.P. Han. Ltd, Oxford: Blackwell Scientific Publications; 1970.Google Scholar
  134. Roche P, Lerouge P, Ponthus C, Promé JC. Structural determination of bacterial nodulation factors involved in the Rhizobium meliloti-alfalfa symbiosis. J Biol Chem. 1991;266:10933–40.PubMedGoogle Scholar
  135. Barnett MJ. A dual-genome Symbiosis chip for coordinate study of signal exchange and development in a prokaryote-host interaction. PNAS. 2004;101:16636–41.PubMedPubMed CentralView ArticleGoogle Scholar
  136. Vallenet D, Belda E, Calteau A, Cruveiller S, Engelen S, Lajus A, et al. MicroScope - an integrated microbial resource for the curation and comparative analysis of genomic and metabolic data. Nucleic Acids Res. 2013;41:D636–47.PubMedView ArticleGoogle Scholar
  137. Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. 2010;11:R106.PubMedPubMed CentralView ArticleGoogle Scholar
  138. Dillies M-A, Rau A, Aubert J, Hennequet-Antier C, Jeanmougin M, Servant N, et al. A comprehensive evaluation of normalization methods for Illumina high-throughput RNA sequencing data analysis. Brief Bioinform. 2013;14:671–83.PubMedView ArticleGoogle Scholar
  139. Davidsen T, Beck E, Ganapathy A, Montgomery R, Zafar N, Yang Q, et al. The comprehensive microbial resource. Nucleic Acids Res. 2010;38:D340–5.PubMedView ArticleGoogle Scholar
  140. Vallenet D, Calteau A, Cruveiller S, Gachet M, Lajus A, Josso A, et al. MicroScope in 2017: an expanding and evolving integrated resource for community expertise of microbial genomes. Nucleic Acids Res. 2017;45:D517–28.PubMedView ArticleGoogle Scholar
  141. Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, et al. Circos: an information aesthetic for comparative genomics. Genome Res. 2009;19:1639–45.PubMedPubMed CentralView ArticleGoogle Scholar
  142. Shapiro SS, Wilk MB. An analysis of variance test for normality (complete samples). Biometrika. 1965;52:591–611.View ArticleGoogle Scholar
  143. Morey JS, Ryan JC, Van Dolah FM. Microarray validation: factors influencing correlation between oligonucleotide microarrays and real-time PCR. Biol Proced Online. 2006;8:175–93.PubMedPubMed CentralView ArticleGoogle Scholar
  144. Moulin L, Klonowska A, Caroline B, Booth K, Vriezen JA, Melkonian R, James EK, Young JP, Bena G, Hauser L, Land M, Kyrpides N, Bruce D, Chain P, Copeland A, Pitluck S, Woyke T, Lizotte-Waniewski M, Bristow J, Riley M. Complete Genome sequence of Burkholderia phymatum STM815(T), a broad host range and efficient nitrogen-fixing symbiont of Mimosa species. Stand Genomic Sci. 2014:9(3);763-74.
  145. Moulin L, Mornico D, Melkonian R, Klonowska A. Draft Genome Sequence of Rhizobium mesoamericanum STM3625, a Nitrogen-Fixing Symbiont of Mimosa pudica Isolated in French Guiana (South America). Genome Announc. 2013:1(1);e00066-12.


© The Author(s). 2018