Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

2.1 Overview of Rhizobium–Legume Symbiosis

2.1.1 Ecological and Agricultural Importance of Symbiotic Nitrogen Fixation

The availability of reduced nitrogenous compounds is a major limiting factor in plant growth and agricultural productivity. The microbiological process that converts atmospheric dinitrogen (N2) into a plant-accessible species is known as biological nitrogen fixation (BNF). BNF reduces the degree of the requirement for external input of chemical N fertilizers to replenish soil N and improve internal resources (Peoples et al. 1995a; Vance 2001; Herridge et al. 2008). Total global N2 fixation from BNF has been estimated to 100–290 million tones N/year, with approximately 50–70 million tones N/year in agricultural systems, compared with 83 million tones N fixed industrially in fertilizer production.

Among the wide range of bacteria that have the ability to reduce N2 to ammonia, the most important are the symbiotic systems of leguminous plants and rhizobial species belonging to β-proteobacteria of the genera Azorhizobium, Bradyrhizobium, Mesorhizobium, Rhizobium, and Sinorhizobium (Ensifer), collectively called rhizobia (Perret et al. 2000; Jones et al. 2007; Franche et al. 2009). Recently, several new species of N2-fixing microsymbionts, such as, Methylobacterium (Sy et al. 2001), Devosia (Rivas et al. 2002), Herbaspirillum (Valverde et al. 2003), Ochrobactrum (Zurdo-Piñeiro et al. 2007), Phyllobacterium (Valverde et al. 2005), and members of the β-proteobacteria such as Burkholderia (Moulin et al. 2001) and Cupriavidus (Ralstonia) (Chen et al. 2001) have been described. A successful interaction between legume plants and rhizobia leads to the formation of nodules on the roots or shoots. Bacteria in the form of bacteroids reside inside nodules and fix atmospheric N into ammonia (Perret et al. 2000; Gibson et al. 2008). The reduced nitrogenous compounds are transported into the host plant in exchange for organic acids. The symbiotic systems are a major source of nitrogen in most legumes with an average of 80% of N derived from BNF (Vance 2001; Graham and Vance 2003). There are estimates that the rhizobial symbioses with 18,000 legume species (Masson-Boivin et al. 2009), including more than 100 agriculturally important legumes spanning all the geographical regions, contribute nearly half of the annual quantity of BNF in soil ecosystems (Graham and Vance 2003). Rotations of legumes with other non-nitrogen-fixing plants enrich the soil with fixed N and increase the productivity and sustainability of agricultural systems. There is evidence that nitrogen derived from legume sources are less susceptible to losses than chemical fertilizer N, which in long term results in the build-up of a reserve of readily mineralizable organic N. The use of BNF in agriculture provides a renewable source of N to supplement or replace fertilizer N and arrests the decline of soil N fertility (Peoples et al. 1995a, b).

2.1.2 Rhizobial Genomes: Background for BNF and Source of Biodiversity

A common feature of the rhizobial genomes is that genes involved in nodulation and N2-fixation are clustered on symbiotic plasmid (pSym), or incorporated into the chromosome as symbiotic islands (Palacios and Newton 2005). The architecture of rhizobial genomes may directly underlie the great genetic and physiological variation of rhizobial strains, resulting in a large diversity of populations. Rhizobial genomes are large (e.g., R. leguminosarum bv. viciae–7.7 Mb; R. etli–6.5 Mb; Sinorhizobium meliloti–6.7 Mb) and are composed of chromosomal core and plasmids, which comprise up to 50% of total genome (Galibert et al. 2001; Gonzalez et al. 2006; Young et al. 2006). Comparative analyses of rhizobial genomes revealed their mosaic structure: regions showing high degree of conserved synteny are separated by other sequences (Guo et al. 2003; Król et al. 2007). There is evidence that such genomes are dynamic structures, where recombination events are very frequent, permanently creating new versions of individual replicons (Brom et al. 1991; Guo et al. 2003). Moreover, not only the symbiotic plasmid but also considerable fractions of the nonsymbiotic plasmid pool present in rhizobial cells are necessary for establishing an effective Rhizobium–legume symbiosis (Brom et al. 1992; Mercado-Blanco and Toro 1996; Galibert et al. 2001). On the one hand, the transfer of symbiotic plasmids between strains in the rhizosphere was evidenced (Broughton et al. 1987), and the lateral transfer of genes was postulated to be a considerable force in rhizobial evolution and diversification (Souza et al. 1992). On the other hand, some data indicate that pSym plasmid transfer frequency is not as high in the field as under laboratory conditions (Wernegreen et al. 1997). Thus, the lateral transfer of genes by plasmid exchange is responsible for an emerging diversity within narrow genetic subdivisions, while main rhizobial genera are quite “reproductively isolated” (Wernegreen and Riley 1999; Bailly et al. 2007).

2.1.3 Signaling in Rhizobium–Legume Symbioses

Rhizobia nodulate wide range of legume plants. Some of them, such as Rhizobium sp. NGR234, are extremely promiscuous and are able to nodulate many different host plants (over 112 hosts) (Pueppke and Broughton 1999), while others, such as R. leguminosarum bv. trifolii, have a very narrow host range and nodulates only clover (Trifolium spp.) plants. Its close relative, R. leguminosarum bv. viciae, nodulates pea (Pisum spp.), vetch (Vicia spp), lentil (Lens spp.), and sweet pea (Lathyrus spp.) (Perret et al. 2000). The specificity of symbiotic interactions is achieved by exchange of molecular signals. In the early steps of symbiosis, a diverse array of compounds is exuded into the rhizosphere, including flavonoids, isoflavonoids, and non-flavonoid inducers (Fig. 2.1). These compounds are chemoattractants for rhizobia (Caetano-Anollés et al. 1988; Dharmatilake and Bauer 1992), influence bacterial growth, and induce the expression of nodulation genes (nod genes) (Peters et al. 1986; Hungria and Stacey 1997) (for detail see Sect. 2 and 3). As a result of nod genes expression, biosynthesis of specific lipochitin oligosaccharides called nodulation factors (Nod factors or LCOs) occurs (Lerouge et al. 1990). Nod factors are structurally diverse and a single rhizobial strain may produce a range of these metabolites (Spaink et al. 1991, 1995).

Fig 2.1
figure 1

The events involved in Rhizobium–legume symbiosis. Roots of legume plant (a) exude flavonoid compounds (b), which are perceived by bacterial NodD regulatory protein (c) that activate set of nod genes resulting in the production of Nod factor (d). Nod factors secreted by rhizobia mediating root hair curling (e) and initiation of nodule primordium (f). Rhizobia attach to the root hairs and infection threads (it) are formed (g). Infection threads penetrate plant tissues, invade nodule primordium, and develops nodule (h, i). (h) Rhizobia tagged with gfp inside young nodules (i) rhizobia tagged with gus in mature nodule

Full size image

Nodulation genes have been classified into three categories. First, the common nodulation genes (nodABC) that are found in all bacteria including the β-proteobacteria (Moulin et al. 2001) with only one exception, that is, photosynthetic bradyrhizobia (Giraud et al. 2007). These are essential for nodulation and mutations in these genes lead to Nod− phenotype (Jacobs et al. 1985; Debellé et al. 1986). The encoded NodC is responsible for biosynthesis of β-1,4-linked N-acetyl-d-glucosamine trimeric to hexameric backbone, while NodB deacetylates glucosamine at the nonreducing end, following which NodA acylates free amino group of the terminal glucosamine (Spaink 2000; D’Haeze and Holsters 2002). Second, the host-specificity nodulation genes (nodFE, nodH, nodG, nodPQ, and several others) whose products modify the N-acylglucosamine backbone by adding species-specific substituents, which are considered the main factors determining the host range of microsymbionts, and influence the rate of nodule formation (Debellé et al. 1986; Horvath et al. 1986; Schwedock and Long 1989). At the reducing-terminal residue, l-fucosyl, 2-O-Me-fucosyl, 4-O-Ac-fucosyl, acetyl, or sulphate esters are present, and at the nonreducing-terminal residues, N-methyl, O-acetyl, and O-carbamoyl are found, respectively. A fatty acyl chain of varying length and with varying degree of unsaturation is attached to the nonreducing end. These moieties controlled by the host-specificity nod gene products make the Nod factors specific for target plant hosts (Dénarié et al. 1996; Spaink 2000; D’Haeze and Holsters 2002). However, variation in Nod factor structure does not fully explain host-range specificity. For example, Nod factors produced by R. etli and Mesorhizobium loti have the same structure but nodulate distinct host legumes (e.g., Phaseolus spp., and Lotus spp.) as reported by Cardenas et al. (1995). This indicates that the elements of the molecular dialogue between the legume plant and rhizobia are more complex and that Nod factor is not the sole signal specifying the host range (Perret et al. 2000; Somers et al. 2004).

The third class of nod genes is a family of regulatory nodD genes (Spaink 2000). NodD proteins belong to the LysR family of transcriptional regulators (Hong et al. 1987; Fisher et al. 1988; Kondorosi et al. 1989). NodD, in complex with a flavonoid, binds conservative sequences upstream of nod operons, called nod-boxes, acting as transcriptional activator of several nod, nol, and noe gene promoters (Peck et al. 2006). Sinorhizobium meliloti synthesizes four NodD proteins (NodD1, NodD2, NodD3, and SyrM) that interact with different plant flavonoid signals. Some rhizobium species use other sensor-activator systems to control the host range. For instance, B. japonicum possesses NodV–NodW, a two-component system that is a positive regulator of nod genes responding to isoflavone signals. NodV and nodW are essential for the nodulation of Macroptilium atropurpureum, Vigna radiata, and V. unguiculata, but contribute only marginally to the symbiosis with Glycine max (Gőttfert et al. 1990; Sanjuan et al. 1994). Furthermore, several nod regulons, which are positively regulated by NodD protein, can be negatively regulated by NolR (Kondorosi et al. 1991; Cren et al. 1995). NolR binds to the promoter regions of nod genes and prevents their expression. Expression of nolR is negatively regulated by luteolin – a specific nod gene inducer in S. meliloti (Cren et al. 1995).

For optimal nodulation, proper level of nod gene expression is required and special mechanisms down-regulate the expression of nod genes in S. meliloti and R. leguminosarum bv. viciae (Somers et al. 2004; Perret et al. 2000; Hogg et al. 2002; Peck et al. 2006). In R. leguminosarum bv. viciae bacteroids, syntheses of NodA, NodI, NodE, and NodO proteins were reduced at least 14-fold compared with free-living bacteria, whereas the level of NodD protein was reduced only threefold (Schlaman et al. 1991). A decreased amount of NodD was also found in a strain harboring multiple copies of nodD. The in situ RNA hybridization of Pisum sativum and Vicia hirsuta nodules showed that transcription of inducible nod genes was switched off by unknown regulatory mechanism before the bacteria differentiated into bacteroids (Schlaman et al. 1991). Moreover, the concentration of Nod factors in the rhizosphere is modulated by plant root hydrolases (Minic et al. 1998; Ovtsyna et al. 2000). Tetrameric NodD binds to 49 bp nod-boxes even in the absence of flavonoids (Feng et al. 2003). However, compatible flavonoids are required to induce the changes in DNA topology at the location of NodD binding in the promoter nod gene, thereby allowing RNA polymerase to initiate gene transcription (Chen et al. 2005). In S. meliloti, only luteolin is capable of activating in vivo nod gene transcription (Peck et al. 2006). Noninducing flavonoids, such as naringenin, eriodictyol, and daidzein, also stimulate an increase in the DNA-binding affinity of NodD1 to nod gene promoters but only luteolin is capable of promoting the topological changes necessary for nod gene induction. This is consistent with the hypothesis that noninducing flavonoids are acting as competitive inhibitors of inducing flavonoids, preventing NodD1 from activating nod gene transcription (Peck et al. 2006).

Flavonoids and nod-boxes also regulate other functions, such as: (1) N2-fixation (Dombrecht et al. 2002), (2) synthesis and/or modification of polysaccharides (Mimmack et al. 1994; Wielbo et al. 2004a), (3) rhizopine catabolism (Rossbach et al. 1994), (4) synthesis of hopanoids (Kannenberg et al. 1995), or (5) synthesis of transcriptional regulators that modulate the synthesis of Tts1 and SyrM2 in Rhizobium sp. NGR234 (Kobayashi et al. 2004). The nodDABC genes, independently of flavonoids, are required for the establishment of the three-dimensional structure of a biofilm formed by S. meliloti, which is enhanced by the flavonoid luteolin – the inducer of nod genes (Fujishige et al. 2008). The core Nod factors facilitate bacterial adhesion to the roots until, in the presence of plant flavonoid inducers, a sufficient concentration of the host-specific Nod factors is reached and plant developmental processes are initiated (Fujishige et al. 2008; Faure et al. 2009).

Host-specific Nod factors are perceived by the plant via LysM-type receptor kinases and a complex signal transduction cascade that triggers early plant responses, such as intra and extracellular alkalinization, membrane depolymerization, calcium spiking, deformation of root hairs, initiation of cortical cell division, infection thread growth, and nodule primordia formation (D’Haeze and Holsters 2002; Oldroyd and Downie 2004; Jones et al. 2007). Recently, several genes of the Nod factor-signaling cascade have been identified and cloned from Medicago truncatula and Lotus japonicus model legumes (Geurts et al. 2005; Oldroyd and Downie 2004, 2006). Nod factors are biologically active at very low concentration (10−9–10−12 M). Rhizobia enter the roots at the sites where root hair cell walls are hydrolysed and may produce either hydrolytic enzymes or use plant mechanisms (Perret et al. 2000). They invade the roots through tubular structures called infection threads. It has also been found that Nod factors reduce the salicylic acid level in the roots to help in the suppression of host defence responses and ensure successful infection by rhizobia (Martínez-Abarca et al. 1998).

Purified lipochitooligosaccharides (LCOs) are sufficient to induce root hair curling, reinitiation of cell division, and in some cases, elicitation of nodule-like structures (Dénarié and Cullimore 1993; Stokkermans et al. 1995; Heidstra and Bisseling 1996; Gibson et al. 2008). Furthermore, Nod factors and other chitin oligosaccharides have been known to have developmental effects on nonleguminous plants, such as carrot (Daucus carota) (de Jong et al. 1993), tobacco (Nicotiana tabacum) (Schmidt et al. 1993), and Norway spruce Picea abies (Dyachok et al. 2000), in the absence of auxin and cytokinin. Global transcriptome analyses of host plants revealed significant changes in the expression of a wide range of genes involved in various aspects of the symbiotic interaction, such as bacterial infection, nodule formation and function, and defense responses (Fedorova et al. 2002; Mitra and Long 2004; Yahyaoui et al. 2004; Barnett and Fisher 2006).

Rhizobia multiply near the tip of growing infection threads, and after reaching the nodule primordium, they are released into the plant cells. Simultaneously, plant-derived peribacteroid membranes (PBM) forming symbiosomes encapsulate them. Inside symbiosomes, they differentiate into bacteroids, which reduces N to NH +4 (Brewin 2004; Gage 2004). Numerous rhizobial signaling molecules are required for initiation and elongation of the infection threads and nodule development. They include symbiotically relevant cell-surface polysaccharides such as capsular polysaccharides (CPS), exopolysaccharides (EPS), cyclic β-(1,2)-glucans, and lipopolysaccharides (LPS). Such polymeric compounds play an important role in plants forming indeterminate type of nodules with a persistent meristem such as Vicia, Medicago, Pisum, or Trifolium (Fraysse et al. 2003; Becker et al. 2005; Skorupska et al. 2006). However, plants that form determinate type of nodules do not have such a requirement (Hotter and Scott 1991). Several possible roles for bacterial polysaccharides in symbiosis are considered: (1) a mechanistic role in protecting bacteria against environmental stresses; (2) acting as signaling molecules triggering plant developmental response; (3) promotion of infection thread initiation and development and suppression of plant defense responses (Fraysse et al. 2003; Skorupska et al. 2006).

Recently, it has been reported that genomes of photosynthetic Bradyrhizobium strains (BTAi1 and ORS278) that induce stem and root nodules on aquatic Aeschynomene legume plants do not contain the common nodulation genes nodABC that are indispensable for Nod factor backbone synthesis, indicating that this group of rhizobia uses mechanism other than Nod factor strategy to enter into symbiosis (Giraud et al. 2007). Purine derivatives, such as cytokinins produced by some strains of B. japonicum and R. leguminosarum, have been considered as a potentially important signal triggering nodule formation (Giraud et al. 2007; Frugier et al. 2008). On the one hand, even though rhizobially derived cytokinins have not been found essential for Nod-factor-dependent symbiosis, yet they seem to be important for Nod-factor-independent nodulation in Aeschynomene. On the other hand, recent data indicate diverse functions of plant cytokinins in symbiosis, such as mediating root susceptibility to rhizobial infection and nodule organogenesis. Cytokinins act downstream of early Nod-factor signaling playing a crucial role in redifferentiation of cortical cells after the induction of Nod-factor-dependent pathways (Lohar et al. 2004; Frugier et al. 2008).

The host plant controls localization, shape, anatomy as well as the infection of the nodules. In general, there are two different kinds of nodules elicited by rhizobia: determinate and indeterminate. The elongated indeterminate nodules have a persistent meristem that continually gives rise to new nodule cells that are subsequently infected by rhizobia. In this type of nodules, the developmental stages can be observed from the meristem at the nodule tip to the older senescent and saprophytic zones near the root (Vasse et al. 1990; Timmers et al. 2000). Bacteroids within indeterminate nodules undergo a terminal differentiation program and grow outside the host cells (Mergaert et al. 2006), while the bacteria released from the saprophytic zone are viable and grow in the rhizosphere and infect legume roots. Determinate nodules are formed generally by tropical legumes (e.g., Glycine max, Vicia faba, and Lotus japonicus) and are round due to lack of persistent meristem and do not display developmental zones (Perret et al. 2000; Gage 2004). Bacteroids within determinate nodules can dedifferentiate after release from nodules and grow in soil (Mergaert et al. 2006). These two types of nodules also differ significantly in their C and N metabolism (White et al. 2007).

2.1.4 Nitrogen Fixation

Inside nodules, rhizobia enclosed in PBM are transformed into bacteroids capable of N2-fixation. This process is catalyzed by oxygen-sensitive molybdenum nitrogenase found in all rhizobia. The nitrogenase catalyzes the following reaction:

$$ {{\text{N}}_2} + {\text{16MgATP}} + {8}{{\text{e}}^{-} } + {8}{{\text{H}}^{+} } \to {\text{2N}}{{\text{H}}_3} + {\text{16MgADP}} + {\text{Pi}} + {{\text{H}}_2} $$

In this reaction, 16–18 molecules of ATP are required for the reduction of N2 molecule to two molecules of NH3 and 2H+ to H2. Nitrogenase is extremely O2 sensitive and is rapidly inactivated in aerobic environment. The low O2 concentration (3–22 nM) in the nodule is achieved by the high concentration of O2-binding heme protein – leghemoglobin, and the symbiosome membrane diffusion barrier (Kaminski et al. 1998; Patriarca et al. 2002). Energetically, this is a very expensive process, which explains the inhibition of nodulation by the presence of fixed nitrogen (White et al. 2007).

Rhizobia are equipped with several nif genes responsible for the N2 fixation process: structural gene (nifH) encoding dinitrogenase reductase, also designated Fe-protein, and nifD, nifK genes encoding α and β subunits of dinitrogenase, respectively, that form functional complexes with FeMo cofactor, also named FeMo protein. nifB, nifE, and nifN genes encode molecular scaffold for the assembly of FeMo cofactor. Prosthetic groups containing 4Fe–4S clusters are covalently bound to MoFe protein bridging the α and β subunits. The 4Fe–4S group is linked also to the Fe protein (Fisher 1994; Newton 2007). The nifA gene performs a regulatory role in expression of nif genes; NifA functions in low oxygen tension. In rhizobia, besides nif genes with significant homology to Klebsiella pneumoniae nif genes (Ruvkun and Ausubel 1980), several other fix genes involved in nitrogen fixation have been found. Generally, rhizobia show high plasticity not only in the gene composition of nif regulon but also in the mode of nif genes regulation (Fisher 1994; Masson-Boivin et al. 2009).

Carbon sources for N2-fixation are supplied by the host plant in the form of photosynthates (sucrose) that are catabolized to C4 dicarboxylic acids and transported to the bacteroids to provide energy for N2-fixation. Ammonium (and alanine) as a product is exported back to the plant through the symbiosome membrane and is further assimilated into glutamine or asparagine in the plant cytosol in indeterminate nodules. In determinate nodules, these amino acids are converted in the uninfected cells, specialized in further nitrogen assimilation into ureides or amides that are transported from the nodules to the shoots (Lodwig and Poole 2003; White et al. 2007).

2.2 Flavonoids Compounds Inducing Nod Factor Production and Enhancing Symbiosis

Plants exude large quantities of sugars (from mono and simple polysaccharides to high molecular weight polysaccharides), acids (aliphatic as well as aromatic acids), amino acids, amines, and many other low molecular weight compounds such as flavonoids, steroids, alkaloids, vitamins, and growth regulators (Gaworzewska and Carlile 1982; Knee et al. 2001; Bertin et al. 2003). All these compounds may affect bacterial soil populations. Some of them, such as sugars, acids, and aminoacids, serve as C and energy sources for microorganisms (Jaeger et al. 1999). Others such as flavonoids are involved in signal exchange between the two symbionts or in anti-pathogen plant defense system (Phillips and Kapulnik 1995). Moreover, some components of plant root exudates have been shown to interfere with quorum sensing-dependent bacterial communication systems (Teplitski et al. 2000). On the basis of these observations, it has been suggested that plants possess a great potential for “soil engineering” and may select for the most suitable bacteria (Simms and Taylor 2002). On the other hand, rhizosphere microorganisms can enhance root exudation of C and energy sources or flavonoids, suggesting the presence of “feedback” in plant–Rhizobium interactions related to bacterial nutrition (Lodwig and Poole 2003; Phillips et al. 2004).

Flavonoids are the most important components of plant root exudates for successful Rhizobium–legume relationships (Table 2.1). The flavonoid backbone is synthesized by condensation of 4-coumaryl-CoA provided by chalcone synthase (CHS) (Dixon and Paiva 1995). Several modifications of this structure yield different classes of flavonoids: flavanones, flavones, isoflavonoids, coumestans, chalcones, and anthocyanidines (Harborne and Williams 2000). More than 4,000 flavonoids are synthesized in vascular plants and released into the rhizosphere (Perret et al. 2000). The production spectrum of these substances may vary with the physiological state and age of the plant (Schlaman et al. 1998; Harborne and Williams 2001). Great amounts of flavonoids are released near the root hair zone; a site of rhizobium infection (Peters and Long 1988; Hartwig et al. 1990; Zuanazzi et al. 1998). Flavonoids play important roles in different stages of nodulation. First, they act as nod gene inducers by activation of NodD protein (Recourt et al. 1991; Hungria and Stacey 1997; Brencic and Winans 2005). For example, luteolin and 4, 7′-dihydroxyflavone (Dhf) are the inducers for S. meliloti (Caetano-Anollés et al. 1988; Hartwig et al. 1990; Peck et al. 2006), isoflavone genistein induces nod gene expression in B. japonicum but inhibits S. meliloti nod gene expression (Kosslak et al. 1987), and hespertin and naringenin are potent inducers of R. leguminosarum bv. viciae (Firmin et al. 1986; Begum et al. 2001). They induce Nod factor synthesis in the infection threads (Sharma and Signer 1990) and act in auxin transport regulation and initiation of nodule primordia cell division (Mathesius et al. 1998; Zhang et al. 2009). Recently, different roles of flavonoids in development of determinate and indeterminate nodules have been reported (Wasson et al. 2006; Subramanian et al. 2007; Zhang et al. 2009). For instance, total silencing of flavonoid biosynthesis in M. truncatula forming indeterminate nodules leads to a near complete loss of nodulation by S. meliloti, whereas flavone-deficiency results in reduced nodulation. Isoflavone-deficient roots are nodulated normally, indicating that isoflavones are not crucial in M. truncatula nodulation (Zhang et al. 2009). Addition of 7,4′-dihydroxyflavone but not flavonol kaempferol (an inhibitor of auxin transport) to flavones-deficient roots can completely restore nodulation. On the basis of these observations, Zhang et al. (2009) proposed the sequence of flavonoid involvement during interaction of M. truncatula with S. meliloti. It has been suggested that the essential role of Dhf is not only as the primary inducer but also in the sustained Nod-factor induction in the infection threads. The sustained Nod-factor induction in turn leads to the accumulation of flavonol kaempferol and accumulation of kaempferol subsequently leading to the localized auxin transport inhibition resulting in cell division, nodule initiation, and development (Zhang et al. 2009). In contrast, silencing of isoflavone biosynthesis in soybean roots forming determinate nodules lead to reduced nodulation and increased auxin transport suggesting the essential role of isoflavones during soybean nodulation. In this case, the isoflavone might be responsible for the sustained induction of bacterial nod genes and Nod-factor biosynthesis inside the roots (Subramanian et al. 2006).

Table 2.1 Substitution patterns of plant flavonoids. Deavours et al. (2006)
Full size table

There are several examples of flavonoids, which have been used to increase nod-gene transcription and promote legume growth. For example, bean nodulation by R. leguminosarum bv. phaseoli or R. tropici was enhanced by the addition of quercetin and malvidin glucoside (Hungria and Phillips 1993); luteolin added to certain alfalfa cultivars significantly increased nodulation (Kapulnik et al. 1987); pretreatment of B. japonicum with genistein increased nodulation, total protein yield, and grain yield of soybean under laboratory (Zhang and Smith 1995) and field conditions (Zhang and Smith 1996); preinduction of R. leguminosarum with flavanones hespertin and naringenin, alone or in mixture, stimulated nodulation and plant dry matter accumulation of pea and lentil plants in comparison to uninduced R. leguminosarum cells in controlled environment growth chamber conditions (Begum et al. 2001).

Seed exudates, which are mixture of flavonoids, are economically more justified when used as exogenous nod genes inducers, although they also contain flavonoid-inhibitors of nod gene expression. Evidence exists that compounds sharing structural similarities with daidzein or genistein are the most effective inhibitors of nod gene induction in B. japonicum. The inhibitors of nod genes may act competitively against inducers at a common target site, such as the NodD protein (Kosslak et al. 1990). However, under laboratory conditions, the clover and bean exudates were more potent inducers of nodA gene of R. leguminosarum than the specific flavonoids alone, indicating the possibility of synergistic effects of nod-gene-activating compounds (Maj et al. 2010). Several authors demonstrated that stimulation of genes with combinations of multiple inducers resulted in better nod gene induction and might be advantageous for early symbiotic interactions (Cooper 2004, 2007). This would not exclude the specific interaction between NodD protein and individual flavonoid (Peck et al. 2006).

The potential of seed exudates as nod-gene inducers could be exploited in the symbiotic activation of inoculants before their use as biofertilizers. As an example, preincubation of R. leguminosarum bv. trifolii strains with clover seed exudate increased fresh mass of shoots and increased nodule numbers in a strain-specific manner under laboratory conditions (Maj et al. 2010). Preactivation of inoculant strains with flavonoids might increase competitiveness in the soil as well as legume productivity; in the case of B. japonicum preactivated with genistein, soybean yields increased by 10–40%, and the seasonal levels of N2 fixation enhanced by 35% (Zhang and Smith 2002). Similarly, field pea and lentil plants displayed increased nodulation and biomass production when inoculated with R. leguminosarum preinduced with hesperetin (Begum et al. 2001). Currently, flavonoids are used commercially to promote Rhizobium–legume symbioses and N2 fixation in agricultural practices (Hungria and Stacey 1997; Mabood et al. 2008). For example, genistein and daidzein, inducers of B. japonicum nod genes, are used in commercial inoculants under the name SoyaSignal. Using SoyaSignal technology to early-planted soybean increased yields by 10% depending on the soybean genotype potential (Leibovitch et al. 2001). These and other associated data thus suggest that the knowledge of the essential role of flavonoids in Rhizobium–legume symbioses could be exploited to enhance the nodulation of economically important legume crops by exogenous addition of synthetic or natural flavonoid compounds. However, the practical application of flavonoids or other signaling molecules in a complex soil environment is much more difficult, and hence, the anticipated results may be different in different ecological niches.

2.3 Non-Flavonoid Signaling Molecules Influencing Legume Productivity

Several non-flavonoid plant factors, such as jasmonates, aldonic acids (erythronic acid and tetronic acid), betains, xanthones, and simple phenolic compounds, can induce the expression of nod genes. The common features of such inducers are that they act at higher concentrations than flavonoids and can induce nod genes and enhance Nod factors production in several legume plants (Cooper 2007). Jasmonic acid (JA) and its ester, methyl jasmonate (MeJA), generally known as jasmonates, are derivatives of linolenic acid and are biosynthesized in plants via the octadecanoic pathway. They are important signal molecules involved in induced disease resistance and mediate many physiological activities, such as environmental stress responses, root growth promotion, or inter-plant communication in plants. Jasmonic acid exogenously applied to the growth medium at high concentration (100 μM) decreases the number of nodules on S. meliloti inoculated Medicago truncatula roots. At such dose rate, JA decreases the responsiveness of calcium spiking to Nod factor, whereas at low concentrations (10–50 μM), it modifies the calcium signal by decreasing the frequency of spiking. Modulation of calcium signaling might have an important role in the initiation of colonization where number of calcium spikes is critical for triggering the Nod factor-signaling pathway (Sun et al. 2006; Miwa et al. 2006; Gutjahr and Paszkowski 2009).

The direct effect of JA and MeJA on induction of nod genes leading to increased Nod factor production has been described for R. leguminosarum and B. japonicum (Rosas et al. 1998; Mabood et al. 2006). In the case of B. japonicum-soybean symbiosis, jasmonates and genistein at 50 μM and 20 μM, respectively, applied alone or together with a Bradyrhizobium promoted nodulation and N fixation under controlled and field conditions, at both optimal and suboptimal root zone temperatures. In the absence of these compounds, at suboptimal root zone temperature, nodulation, nitrogen fixation, and plant growth were inhibited (Mabood and Smith 2005; Mabood et al. 2008). The synergistic effect of jasmonate and flavonoids (naringenin, genistein) on nodulation, N2 fixation, and biomass production was also observed for R. leguminosarum-bean (Phaseolus vulgaris L.) symbiosis (Poustini et al. 2007). These results suggest that both inducers, jasmonates and flavonoids, utilize different receptors for signal transduction, and a concomitant activation of different regulatory mechanisms enhances the transcription of nod genes and Nod factor production (Mabood et al. 2006).

Within the variety of flavonoids, isoflavonoids, and other compounds secreted by lupin (Lupinus albus) roots into the rhizosphere, major proportion is composed of aldonic acids, that is erythronic acid (4-C sugar acids) and its analog tetronic acid. These compounds act as nod gene inducers of R. lupini, M. loti, and S. meliloti. Both aldonic acids in mM concentrations stimulated the expression of nodC gene in rhizobia. In addition, lupiwighteone, a genistein monoprenyl, added to cultures together with either aldonic acids exerted a synergistic effect on nod genes induction of R. lupini. Synergistic effect of luteolin and tetronic acid (but not erytronic acid) on nod induction was also observed for S. meliloti. Concomitantly with nod gene induction, the increase in LCO production was observed in the presence of both aldonic acids in R. lupini cultures, and of tetronic acid in M. loti and S. meliloti (Gagnon and Ibrahim 1998).

The next group of non-flavonoid inducers studied to date is betaines (stachydrine and trigonelline), N-methylated derivatives of aspartic acid and ornithine, secreted in large amounts from germinating alfalfa (Medicago sativa L.) seeds (Goldmann et al. 1991; Phillips et al. 1992). Stachydrine and trigonelline transcriptionally activated only the gene encoding the NodD2 protein but no apparent activation was reported for NodD1, which was activated by luteolin (Phillips et al. 1992). Betaine inducers, similarly to aldonic acids, function at a 103-fold higher concentration than required for nod induction with flavonoids. In S. meliloti-alfalfa symbiosis, using non-flavonoid molecules that are synthesized via a metabolic pathway distinct from those for flavonoid synthesis could be beneficial in the case of flavonoid soil depletion. Also, the good solubility of stachydrine and trigonelline in water may allow them to diffuse more easily than flavonoids within the soil, increasing the availability of these inducers (Phillips et al. 1992). The nod genes of B. japonicum can also be induced by xanthones (Yuen et al. 1995).

2.4 Nod Factor-Enhancement of Bacterium–Plant Symbiosis

In early studies, the manipulation of a number of common nod genes led to a reduced nodulation of Vicia faba inoculated with R. leguminosarum harboring a multicopy plasmid carrying nodABC genes (Knight et al. 1986). Following this, attempts were undertaken to improve nodulation and N2-fixation in alfalfa plants by amplifying specific regions of the symbiotic plasmids of S. meliloti strains 41 and 1021 (Castillo et al. 1999). Amplified DNA fragments containing nodD1 regulatory gene, the common nodulation genes (nodABC) and nifN gene essential for N2-fixation, were introduced into S. meliloti genome by homologous recombination. Derivatives of S. meliloti with a moderately increased copy number of symbiotic genes (2.5–3 copies) showed better symbiotic properties and promoted plant growth under controlled conditions. When the number of copies of symbiotic genes was set to about seven, the nodulation and N2-fixation decreased (Castillo et al. 1999). These results suggested that the manipulation of structural or regulatory nod genes in rhizobia to increase their symbiotic activity is possible but only to some extent.

The key role of the Nod factor in early steps of symbiosis and its mitogenic and morphogenic activity leading to the formation of nodule primordium sparked several attempts to employ it to enhance legume productivity (Spaink et al. 1991). “Hormone-like” effect of purified B. japonicum Nod factor (nod Bj-V C18:1,MeFuc) was observed in legume (soybean) and nonlegume plants (corn) (Souleimanov et al. 2002). The application of Nod factor at concentrations from 10−7 to 10−9 M stimulated biomass accumulation and enhanced growth and architecture of roots in host and nonhost plants indicating that the perception of LCO signal is conserved among a variety of species. This was further confirmed by identifying the genes of Nod factor cascade essential for mycorrhiza or an endosymbiosis in many higher plant species (Geurts et al. 2005). Further experiments showed that B. japonicum Nod factor stimulated germination of soybean and a variety of economically important plants belonging to diverse families under laboratory and field conditions (Prithiviraj et al. 2003). Stimulation of seed germination and seedling growth of several members of angiosperms suggests that specific receptors might also exist on seed surface or the developing embryonic roots, and Nod factor-induced genes may be present in genomes of numerous leguminous or nonleguminous plants (Prithiviraj et al. 2003). The chitin pentamer did not elicit such responses, which demonstrated that the structure of Nod factor plays a role in specificity of its biological activity. For legumes forming indeterminate nodules, Macchiavelli and Brelles-Mariño (2004) observed a noticeable increase in nodule numbers after treating the seeds of M. truncatula with submicromolar concentration of S. meliloti LCOs before inoculation. Moreover, clover seeds treated with specific LCOs before planting displayed a significantly enhanced nodulation and clover growth under competitive conditions in the presence of a mixture of chemical signals in the soil. Under these conditions, the symbiotic activity and competitiveness of R. leguminosarum bv. trifolii test strain were not improved suggesting that mitogenic activity of LCOs was solely responsible for the observed effects (Maj et al. 2009). Similar effect was observed for B. japonicum and soybean, which produce determinate nodules. Commercially, the addition of Nod factors into inoculants of B. japonicum to promote nodulation and plant growth of soybean has been applied by Nitragin Inc (Mabood et al. 2008).

2.5 Competitiveness in Natural Populations and Its Effect on Legume Crops

From among the several desirable characteristics in rhizobial strains that can be used as biofertilizers, the most essential are: (a) the ability to form nodules and to fix nitrogen in the host plant in a range of environmental conditions, and (b) the ability to compete for nodulation with indigenous rhizobial population (Brockwell et al. 1995). The soil is an exceedingly complex habitat inhabited by heterogenous microbial communities, which interact with each other via different chemical compounds secreted into the soil. Among these microbial populations, rhizobia constitute “common soil inhabitants” in all climatic zones, from arctic to the tropics, and are commonly found in different types of soils (Robertson et al. 1995; Andrade et al. 2002; Fagerli and Svenning 2005). The correlation between the type of soils and the quantity of rhizobia have shown that the number of R. leguminosarum, Sinorhizobium, or Bradyrhizobium in most cases was 102–105 cells/g soil (Andrade et al. 2002; Martyniuk et al. 2005). The extensive variability in the soil environment may induce the emergence of diversity in rhizobial populations. Moreover, rhizobial diversity may be influenced by soil properties, such as the availability of N, P, Ca, the acidity, chemical stresses, or by agricultural management regimes (Palmer and Young 2000; Andrade et al. 2002; Laguerre et al. 2006). However, most of the rhizobial biodiversity studies were done with strains collected from a single or a few neighboring geographical regions. They revealed differences in allele frequency of selected genes (e.g., 16S rRNA, nodD, nodEF, nifDK) and sequences (16–23S rRNA ITS), which go in parallel with metabolic differences, for example the enzymatic profile or carbon/energy substrate utilization (Louvrier et al. 1996; Mutch and Young 2004; Silva et al. 2007). Further analysis of these traits allowed clustering the strains and finding correlation between genetic content and origin of the strains. Still, there is evidence that large scale biodiversity also exist within populations of rhizobia colonizing single plants (Wielbo et al. 2010).

Unlike soil, root nodules of legumes form a microenvironment not accessible to all rhizobia inhabiting rhizosphere probably because of the following: (1) the relatively high plant host-microsymbiont specificity as the nodules can be induced and consequently colonized only by rhizobia, which recognize and exchange suitable molecular signals with the host; and (2) variation in secreted Nod factors (Perret et al. 2000; Jones et al. 2007). Moreover, such rhizobia are not exposed to plant defense mechanisms, which are activated if bacteria are not recognized as symbionts, for example as a result of changes in lipopolysaccharide or exopolysaccharide structure (Campbell et al. 2002; Wielbo et al. 2004b). Rhizobia recognized as “suitable microsymbionts” are subjected to a selection process during plant tissue invasion and colonization in two ways – they are exposed to intense competition from other strains, and they are under some selective pressure from the plant host.

The effect of the plant host on the structure and composition of rhizobial population is not conclusive, but there are reports that legumes favor particular symbiotic genotypes of rhizobia (Mutch and Young 2004; Rangin et al. 2008). This relationship seems to be more complex, and even slight differences in genotype or developmental stage of plants may have an influence on the constitution of rhizobial populations in root nodules (Depret and Laguerre 2008). The molecular adjustment of the microsymbionts to their host may be observed as prevalence of particular symbiotic genotypes (versions of nod genes cluster) in rhizobia isolated from root nodules (Laguerre et al. 2003) or dependence between the susceptibility of rhizobia for plant-derived flavonoid induction and competitive abilities of microsymbionts (Maj et al. 2010). For this reason, hypotheses about coevolution in Rhizobium–legume symbiosis have been proposed (Doyle 1998) and recently modified, emphasizing the effect of plants on the evolution of bacteria (Martínez-Romero 2009).

Independently of the plant host selection pressure on rhizobial populations, a lot of bacterial traits are involved in the process of competition between the individual strains, which run a race to the nodules (Vlassak and Vanderleyden 1997). The “external” environmental conditions, such as soil acidity, salinity, and nutrient availability strongly affect the vegetative growth of rhizobia in soil, thus setting up the initial conditions for rhizobial competition. A lot of data are available concerning the effect of single, defined factors such as acid tolerance (Vinuesa et al. 2003), presence of small cryptic plasmid (Bromfield et al. 1985), production of vitamins (Streit et al. 1996), and rhizopines (Murphy et al. 1987) affecting the competitive properties of rhizobia. The ability for utilization of specific C and energy sources, such as rhamnose (Oresnik et al. 1999) and homoserine (Hynes and O’Connel 1990), as well as the ability to metabolize the most variable set of substrates (including acids and amino acids) (Wielbo et al. 2007) have been proved to be important determinants of competitiveness. Moreover, the direct strain-to-strain antagonistic effect should also be taken into consideration as the effect of bacteriocin production on strain competitiveness was also reported (Robleto et al. 1998; Oresnik et al. 1999).

The competition between rhizobial strains does not vanish following the relocation of the bacteria from the soil into plant and remain present after the root colonization stage, possibly inside the infection threads (Duodu et al. 2009). Because rhizobia are immobile in the infection thread, the rate of bacterial growth inside can determine the rate of infection thread proliferation and subsequent nodule occupancy (Hoang et al. 2008; Duodu et al. 2009). Presence of multiple bacterial strains inside a single infection thread was shown by Stuurman et al. (2000) and Gage (2004). The C and N exchange between bacteroids and plant cells suggests that metabolic traits of rhizobia may also be important during this endophytic part of bacterial life cycle (Prell and Poole 2006; White et al. 2007). From “bacterial point of view,” the aim of competition is to reach nodule compartments, which later on serve as a place for growth and massive multiplication. In indeterminate nodules, such a compartment is called saprophytic zone and is an ecological niche where rhizobia take advantage of the interaction with their plant host, escape plant controls, and their morphology, and some of the physiological traits become similar to these characteristic for a saprophytic (nonsymbiotic) stage (Timmers et al. 2000; Wielbo et al. 2009). In summary, the better the competitive abilities of a strain relative to the autochthonous strains of a local population, the more are the chances for colonization and multiplication in the nodules. And consequently, for the “return to the soil” after plant’s vegetative period, albeit in higher number, which may lead to a dominance in the population. The success of a strain is also dependent on the diversity and total viable cell number of the local population. In rhizobia-rich soils, an introduced strain may be quickly dissipated into the autochthonic population, and its “half-life” may not exceed 1–2 years (Jensen and Sorensen 1987). On the other hand, when the soil is depleted for rhizobia, the persistence of the introduced strain may exceed a few years, and the strain’s good competitive abilities may enable a progressive elimination of rivals (Svenning et al. 2001).

One of the strategies that have been postulated a century ago and that has been a common practice for years to enhance legume nodulation and N2 fixation is the introduction of rhizobial inoculants into the cultivated soil (Martínez-Romero 2003). Inoculation has enhanced plant growth and yield in the cases where even the specific rhizobia were absent or inefficient (Streeter 1994; Brockwell et al. 1995; Giller and Cadisch 1995). On the other hand, strain effectiveness and competitiveness are traits not linked to each other, and bacteria introduced as biofertilizer for target host plants might be outcompeted by autochthonous rhizobia abundant in the fields with little end success (Vlassak and Vanderleyden 1997; Burgos et al. 1999). Therefore, there is a need to investigate the strain competitiveness for nodulation as part of the process of converting a “potentially useful strain” into a “commercial inoculant strain”. As discussed earlier, numerous individual traits affecting the competitiveness were identified and some recommendations for improving strain competitiveness have been made (Maier and Triplett 1996), and even genetically engineered strains with competitive abilities targeted in a specific manner were constructed (van Dillewijn et al. 2001). Moreover, promising mathematical models describing nodulation competitiveness were formulated (Beattie et al. 1989). On the other hand, original optimistic attempt to identify and clone the “nodulation competitiveness genes” had to be rejected, with the idea of “spontaneous genetic changes,” which render the strains competitive (Beattie and Handelsman 1993) gaining prominence. The advances in understanding about individual metabolic traits responsible for competitiveness (Murphy et al. 1987; Hynes and O’Connel 1990; Streit et al. 1996; Robleto et al. 1998; Oresnik et al. 1999) has replaced the previous simple or very inaccurate theories, and shed some light on the network of factors involved. In this context, the role of point mutations underlying diversity and thus increasing the adaptive potentials of microbial species was confirmed (de Weert et al. 2004), resulting in inclusion of this phenomenon in the list of factors affecting competitiveness. It is possible that such small-scale evolution and/or recently discovered factors may explain some unsolved problems, for example, why strains selected for a particular trait (e.g., acid tolerant) and competitive under laboratory conditions are outnumbered in the “appropriate” (i.e., low pH) soils by indigenous strains (Gemel and Roughley 1993). The practical consequence of such issue requires a continuous search for identifying effective and competitive strains. Nowadays, more efforts are made in countries with weak agricultural practice (Africa or Southerneastern Asia) or with poor fertility of soils (Australia) (Lupwayi et al. 1997; Slattery and Pearce 2002), while in North America and European Union, such work is still conducted despite the availability of numerous patented and industrially-made rhizobial inoculants. They are often coupled with the investigation of problems relating to agricultural use of the inoculants: bacterial carriers, terms of storage, time and method of application, etc. (Brockwell and Bottomley 1995; Herridge et al. 2002), which may strongly affect the viability and metabolic status of bacteria, and thus exert considerable effect on inoculant competitiveness and legume improvement.

2.6 Conclusions

Research to date has shown that the productivity of Rhizobium–legume symbiosis can be enhanced by manipulating bacterial and plant signals, selecting well-adapted bacteria, so that they could be introduced into soil or by modifying plant and bacterial activities. However, the complexity of competitiveness of rhizobia and plethora of factors influencing this phenomenon, which allow them to multiply and out-compete autochthonous rhizobia, still remains to be elucidated. The advanced global analyses of cellular state through approaches such as genomics, transcriptomics, and metabolomics have however made it possible to investigate whole microbial communities, which in turn, has enhanced further the knowledge of diversity on the genetic and metabolic levels. These are likely to result in better formulation of the inoculants applied as biofertilizers in legume production across different ecological regions. All strategies that positively influence legume biomass and global fixed N are important for sustainable agriculture especially when soils lack specific rhizobia or when their number is low. Nowadays, despite a myriad of “traditional” rhizobial inoculants available in the market, new-formulated biofertilizers supplemented with flavonoids or Nod factors needs to be designed and developed and further tested for the promotion of legumes productivity.