Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. 2010 Oct 26; 107(43): 18735–18740.
Published online 2010 Oct 20. doi:  10.1073/pnas.1011957107
PMCID: PMC2973005
Plant Biology

R gene-controlled host specificity in the legume–rhizobia symbiosis


Leguminous plants can enter into root nodule symbioses with nitrogen-fixing soil bacteria known as rhizobia. An intriguing but still poorly understood property of the symbiosis is its host specificity, which is controlled at multiple levels involving both rhizobial and host genes. It is widely believed that the host specificity is determined by specific recognition of bacterially derived Nod factors by the cognate host receptor(s). Here we describe the positional cloning of two soybean genes Rj2 and Rfg1 that restrict nodulation with specific strains of Bradyrhizobium japonicum and Sinorhizobium fredii, respectively. We show that Rj2 and Rfg1 are allelic genes encoding a member of the Toll-interleukin receptor/nucleotide-binding site/leucine-rich repeat (TIR-NBS-LRR) class of plant resistance (R) proteins. The involvement of host R genes in the control of genotype-specific infection and nodulation reveals a common recognition mechanism underlying symbiotic and pathogenic host–bacteria interactions and suggests the existence of their cognate avirulence genes derived from rhizobia. This study suggests that establishment of a root nodule symbiosis requires the evasion of plant immune responses triggered by rhizobial effectors.

Keywords: soybean, nodulation, nitrogen fixation, defense

Plants of the legume family, such as alfalfa, clovers, peas, and beans, can make their own nitrogen fertilizer by forming symbioses with a diverse group of nitrogen-fixing soil bacteria known as rhizobia. This cross-kingdom collaboration is characterized by the formation of the root nodule, a specialized plant organ that provides an optimized environment for the bacteria to convert atmospheric nitrogen into ammonia. The legume–rhizobia association is highly specific, such that each rhizobial strain establishes a symbiosis with only a limited set of host plants and vice versa. Despite recent advances in our understanding of the Nod-factor signaling pathway (1, 2), the host control of nodulation specificity remains poorly understood.

The symbiotic specificity is determined by a fine-tuned exchange of molecular signals between a host plant and its bacterial symbiont (3). The best-known bacterially derived signal is the Nod factor, a family of lipo-chitooligosaccharides with various strain-specific chemical decorations (46). The ability to induce the production of bacterial Nod factors in response to host-secreted flavonoids and to subsequently perceive the signal by the cognate host receptor(s) is widely thought to play a key role in defining the host range (3, 710). For example, allelic variation in the pea sym2, a putative Nod-factor receptor, leads to natural polymorphisms in Nod-factor perception and thus symbiotic specificity (7, 8). Moreover, introduction of the putative Nod-factor receptors NFR1 and NFR5 of Lotus japonicus into Medicago truncatula enables nodulation of the transgenic roots by the L. japonicus symbiont Mesorhizobium loti (9, 10).

Besides Nod factors, rhizobia also use surface polysaccharides or secreted proteins to modulate host range (1113). Rhizobial surface polysaccharides (e.g., exopolysaccharides, lipopolysaccharides, and cyclic glucans) have been implicated in facilitating infection thread formation and nodule development (14). It has been proposed that surface polysaccharides play a role in the evasion or suppression of host defense responses, a feature that is shared by pathogenic and symbiotic bacteria (1214). Another striking similarity between pathogenic and symbiotic bacteria is that many, but not all, rhizobial strains also possess a type III secretion system (T3SS) that delivers effectors [so-called nodulation out proteins (Nops)] into the host cells (11). Rhizobial T3SSs modulate the host range in a genotype-specific manner, but are dispensable for rhizobial infection and nodulation. This latter observation reveals an important difference between T3SSs of mutualistic rhizobia and pathogenic bacteria, because T3SSs of pathogens are required for causing disease in susceptible hosts and for eliciting the hypersensitive response in resistant hosts (15). For these rhizobial signals, the host presumably possesses the yet unknown corresponding recognition mechanisms that control the compatibility of the legume–rhizobia interaction. To unravel such mechanisms, it is critical to characterize natural variation in symbiotic specificity within a host species, but such variation remains largely unexplored in the two model legumes M. truncatula and L. japonicus.

Natural variation in symbiotic specificity has been well documented in soybeans since the 1950s (16). Genetic analysis of these naturally occurring variations identified several dominant genes (e.g., Rj2, Rj4, and Rfg1) that restrict nodulation with specific rhizobial strains (1719). The dominant nature of these genes is in contrast to those “loss-of-function” recessive alleles in Nod-factor receptors (e.g., sym2 and sym37 in pea and rj1 in soybean) (7, 20, 21), but resembles gene-for-gene resistance against plant pathogens. In this report, we describe the positional cloning of Rj2 and Rfg1, the two soybean genes identified several decades ago that restrict nodulation with two distinct groups of rhizobial strains (16). In particular, the Rj2 gene restricts nodulation of soybean with specific Bradyrhizobium japonicum strains such as USDA122 (17), whereas Rfg1 prevents nodulation with certain fast-growing Sinorhizobium fredii strains such as USDA257 (18). We show that Rj2 and Rfg1 are allelic genes encoding a member of the Toll-interleukin receptor/nucleotide-binding site/leucine-rich repeat (TIR-NBS-LRR) class of plant resistance proteins (22). This finding reveals a common recognition mechanism underlying symbiotic and pathogenic host–bacteria interactions and bodes well for the existence of their cognate avirulence genes derived from rhizobia. Our study suggests that establishment of a root nodule symbiosis requires the evasion of plant immune responses triggered by rhizobial effectors.

Results and Discussion

Positional Cloning of Rj2 and Rfg1.

The Rj2 locus was previously mapped to the soybean linkage group J (chromosome 16) (2325). We carried out fine mapping of the locus using an F2 population derived from the cross between L76-1988 and L82-2024 (hereafter referred to as L76 and L82, respectively). L76 (Rj2/Rj2, rfg1/rfg1) restricts nodulation with B. japonicum USDA122 but not with S. fredii USDA257, whereas L82 (rj2/rj2, Rfg1/Rfg1) can nodulate with USDA122 but not USDA257 (Fig. 1 A and B). From a total of 1,258 F2 plants inoculated with USDA122, 296 plants nodulated, which fits the 3:1 (nonnodulation to nodulation) ratio (χ2 = 1.45, df = 1, P = 0.23), consistent with restriction of nodulation by USDA122 being controlled by a single dominant gene (17). Genotyping of the 296 nodulated individuals (rj2/rj2) using single-nucleotide polymorphism (SNP) markers allowed us to locate the Rj2 locus within a 47-kb genomic region (Fig. 1C). The 47-kb genomic sequence of Williams 82 (rj2/rj2) contains at least 10 predicted genes (Glyma16g33690–Glyma16g33800) (25). One of these, Glyma16g33780, encodes a member of the TIR-NBS-LRR class of disease resistance proteins. This gene was considered as a strong candidate of the rj2 allele.

Fig. 1.
Map-based cloning of Rj2 and Rfg1. (A) Nodulation phenotypes of L76 (Rj2/Rj2, left) and L82 (rj2/rj2, right) by B. japonicum USDA122. (B) Nodulation phenotypes of L76 (rfg1/rfg1, left) and L82 (Rfg1/Rfg1, right) by S. fredii USDA257. (C) Fine mapping ...

We performed de novo mapping of the Rfg1 locus using an F2 population derived from the cross between Williams 82 (rj2/rj2, Rfg1/Rfg1) and Peking (rj2/rj2, rfg1/rfg1). Preliminary mapping based on 200 F2 individuals suggested that the Rfg1 locus is tightly linked with the Rj2 locus. Thus, we decided to finely map the Rfg1 locus using an F2 population derived from L76 and L82, taking advantage of the available molecular markers developed for mapping Rj2. Phenotyping a total of 1,350 F2 plants inoculated with S. fredii USDA257 identified 323 nodulated plants. The ratio of nonnodulation to nodulation (1,027:323) statistically fits 3:1 (χ2 = 0.83, df = 1, P = 0.36), confirming that resistance to nodulation by USDA257 is conditioned by a single dominant gene (18). Fine mapping using the 323 nodulated F2 plants (rfg1/rfg1) enabled us to delimit the Rfg1 locus within a 69-kb region that encompasses the 47-kb Rj2 candidate gene region (Fig. 1C). Because Glyma16g33780 is the only TIR-NBS-LRR type gene predicted from the 69-kb genomic region, we postulate that Rj2 and Rfg1 are allelic genes. This observation comes as an intriguing surprise, because the two genes were independently identified by their nodulation restriction with two phylogenetically distinct groups of rhizobial species (1618).

Functional Validation of the Candidate Genes.

We validated the candidate genes by complementation tests using Agrobacterium rhizogenes-mediated hairy root transformation (26). The experiments were conducted without antibiotic selection, and thus the resulting hairy roots contained both transgenic and wild type, which can be readily distinguished by examining the expression of the GUSPlus gene in the binary plasmid pCAMBIA1305.1, followed by analysis of transgene expression (Fig. 2C). As shown in Fig. 2, introduction of the candidate Rj2 allele of L76 into L82 (rj2/rj2) resulted in complete block of nodulation on the transgenic roots inoculated with USDA122 (Fig. 2 A and B). From a total of 41 composite transgenic plants that possess both transgenic and wild-type roots, nodules were formed on the wild-type roots but never observed on the transgenic roots. The function of the candidate Rj2 allele in L76 was further confirmed by RNAi-mediated gene silencing. Transformation of an RNAi construct into L76 led to significant reduction of the candidate gene transcripts (Fig. 2F) and enabled nodulation of the transgenic roots by USDA122 (Fig. 2 D and E). From a total of 30 composite plants, a range of 12–30 nodules were formed on the chimeric roots of 28 plants. Fluorescence microscopy confirmed that all of the nodules were from the roots that expressed the red fluorescent marker DsRed1 in the binary plasmid pRedRoot II. Similarly, transferring the candidate Rfg1 allele of Williams 82 into Peking (rfg1/rfg1) (n = 27 plants) and L76 (rfg1/rfg1) (n = 22 plants) abolished nodulation of the transgenic roots by USDA257, whereas nodulation was normal on the nontransgenic hairy roots (Fig. 3). Taken together, we conclude that Rj2 and Rfg1 are allelic genes encoding a member of the TIR-NBS-LRR class of plant disease resistance proteins.

Fig. 2.
Functional validation of Rj2. (A) Introduction of Rj2 (L76) into L82 (rj2/rj2) led to block of nodulation on the transgenic roots by USDA122 (Right), but nodulation was normal on the wild-type roots (Left). (B) Vector control experiment showing nodulation ...
Fig. 3.
Functional validation of Rfg1. Transferring Rfg1 (Williams 82) to L76 (A) and Peking (B) resulted in block of nodulation on the transgenic roots by USDA257. (C) Control experiment showing that the transgenic Rfg1 roots of Peking can nodulate with USDA122. ...

Expression and Allelic Polymorphisms of the Rj2/Rfg1 Locus.

Rj2 and Rfg1 alleles are constitutively expressed in both roots and leaves, and their expression in the root does not markedly change on rhizobial inoculation (Fig. 4). Both Rj2 (L76) and Rfg1 (L82) transcripts are predicted to encode a protein of 1052 aa, consisting of an N-terminal TIR domain, a centrally located NBS domain, and at least eight degenerate LRRs C-terminal to the NBS domain (Fig. 5A). Sequence comparison identified a total of seven amino acid substitutions between the two allelic products (E452K, I490R, Q731E, E736N, P743S, E756D, and R758S) (Fig. 5B). These polymorphisms occur in both NBS and LRR domains and define the genetic basis of specificity differences between the L76 and L82 alleles.

Fig. 4.
Constitutive expression of the Rj2 (rfg1) and rj2 (Rfg1) alleles in the roots of L76 and L82, respectively. RT-PCR was performed using the gene-specific primers (25 cycles). Roots at 0, 1, 2, 3, and 4 d postinoculation (dpi) were collected for RNA isolation. ...
Fig. 5.
Amino acid polymorphisms among the three allelic groups of Rj2 (rfg1), rj2 (rfg1), and rj2 (Rfg1). (A) Domain structure of the TIR-NBS-LRR protein showing the seven substitution sites between the Rj2-L76 and Rfg1-L82 proteins. (B) Allelic comparisons ...

To gain a further understanding of allelic diversity and specificity at the Rj2/Rfg1 locus, we tested a set of 21 soybean lines for their symbiotic compatibility with USDA122 and USDA257, with a particular focus on previously known Rj2/Rj2 and rfg1/rfg1 genotypes (Fig. 5B). Both Rj2 and rfg1 alleles occur with a low frequency in surveyed soybean varieties (16). Of the 21 genotypes, 9 carry an Rj2 (rfg1) allele that allows for nodulation with USDA257 but not with USDA122; 7 possess an rj2 (Rfg1) allele that permits nodulation with USDA122 but not with USDA257; and the remaining 5 contain an rj2 (rfg1) allele that enables nodulation with both USDA122 and USDA257. We did not identify the Rj2 (Rfg1) allele type that prohibits nodulation with both strains. These results suggest that soybean plants carrying an Rj2 or Rfg1 allele commonly show contrasting specificity with USDA122 and USDA257 and thus can form symbiotic interactions with at least one of the two strains. We expect that Rj2 (Rfg1) alleles would be rare, if they do exist, in natural populations.

We sequenced the alleles (cDNAs) of the 21 genotypes. Pairwise comparisons of these allelic products revealed that the aforementioned seven amino acid substitutions between L76 and L82 represent a minimum difference observed between an Rj2 (rfg1) and an rj2 (Rfg1) allele; all other polymorphisms are also present within the two allelic groups and thus are not associated with allelic specificity. As shown in Fig. 5B, the nine Rj2 (rfg1) alleles share a similar protein haplotype at the seven sites (E452, I490, Q731, E736, P743, E756, and R/L758), whereas the six rj2 (Rfg1) alleles share an identical protein haplotype at the seven sites (K452, R490, E731, N736, S743, D756, and S758). Intriguingly, the five rj2 (rfg1) alleles are chimeric of the above two allelic types (K452, R490, Q731, E736, P743, E756, and R758). Therefore, the allelic diversity and specificity likely resulted from intragenic recombination events, even though we do not yet know the ancestral alleles. Sequence comparisons between the Rj2 and rj2 alleles (regardless of the Rfg1 specificity) and between the Rfg1 and rfg1 alleles appear to suggest that the Rj2 function requires a haplotype containing E452 and I490, whereas the Rfg1 function requires a haplotype consisting of E731, N736, S743, D756, and S758 (Fig. 5B). This inference remains to be tested by sequencing more alleles and by domain swaps and functional validation experiments.

Incompatible Rhizobial Strains Can Induce Root Hair Curling but Fail to Infect the Host Carrying the Cognate Resistance (R) Genes.

Numerous rhizobial genes have been implicated in species-specific or genotype-specific nodulation (2729). Our discovery that host R genes are involved in determination of symbiotic specificity is in line with the finding that many rhizobial strains possess a T3SS that delivers effectors into the host cells to modulate host range (11). Some rhizobial effectors are homologous to those secreted by bacterial pathogens, suggesting a similar recognition mechanism between symbiotic and pathogenic host–bacteria interactions (11, 30).

In the incompatible interactions controlled by the Rj2 or Rfg1 genes, the rhizobial strains were able to induce root hair curling and occasionally cause nodule primordium formation (Fig. 6 A and B), but cortical cell division ceased at an early stage due to a lack of infection thread formation (Fig. 6 CF) (31). The failure of infection-thread growth was likely caused by defense responses triggered by the recognition of yet unknown rhizobial effectors by the host R genes (Fig. 6 G and H).

Fig. 6.
Root responses of L82 and L76 inoculated with B. japonicum USDA122. (A and B) USDA122 induces root hair curling on both compatible (L82, A) and incompatible (L76, B) hosts. Photographs were taken 3 d after inoculation. (C) At 8 d after inoculation, USDA122 ...

Role of a T3SS Mutant of S. fredii USDA257 in Modulation of Genotype-Specific Nodulation.

We further examined the effects of a T3SS mutant of USDA257, called DH4, on nodulation properties of nine selected soybean genotypes that represent all three described allelic types (Table 1). The DH4 mutant disrupted the RhcU function and failed to secrete any effector proteins (32). This experiment revealed that the T3SS mutant gained the ability to nodulate the soybean genotypes carrying an Rfg1 allele and maintained the ability to nodulate the plants carrying an rfg1 allele. The mutant strain can increase, decrease, or have no effect on nodule numbers in comparison with the wild-type strain depending on the genetic background (Table 1). Similar results were also reported for other rhizobial strains such as Rhizobium sp. NGR234 (29). These observations suggest that, in contrast to bacterial pathogens, a rhizobial T3SS and its secreted effectors are not required for rhizobial infection and nodulation, but possibly function as a facilitator superimposed on the Nod-factor signaling pathway. In the absence of recognition by the host R genes, the T3SS effectors may play a positive role in facilitating rhizobial infection, but function negatively if perceived by the host R genes. Thus, our study supports the hypothesis that symbiosis development requires the evasion of plant immune responses triggered by rhizobial effectors (11, 13). Nod factors and surface polysaccharides have been postulated to play a role in suppression of defense responses in compatible hosts induced by microbe-associated molecular patterns (MAMPs) (1114, 33), but they appear unable to overcome the effector-triggered immunity (ETI) to establish a symbiosis (34).

Table 1.
Nodulation properties of USDA257 and a T3SS mutant (DH4) on different soybean genotypes

Evolution of Nodulation Restriction R Genes.

Although very few legume species have been thoroughly surveyed for their symbiotic variability, dominant genes that restrict nodulation with specific rhizobial strains have been frequently identified in the tested species (35). Unlike most R genes conferring strain-specific disease resistance, a single Rj2 or Rfg1 allele could restrict nodulation by multiple rhizobial strains (16). The wide prevalence of such genes in natural populations suggests that they are unlikely to be under negative selection. It is possible that some rhizobia and pathogenic bacteria use a common set of “virulence” effectors to facilitate invasion of the host, providing the host with a dilemma between forming symbiosis and mounting disease resistance. Eventually, natural selection would maintain both alleles to balance benefits and costs, depending on ecological conditions. A second scenario would be that some host genotypes have evolved R genes to selectively interact with certain strains but exclude others. For example, the predominance of the Rj4 allele in soybeans may be attributed to its role in protecting the host from nodulation with many Bradyrhizobium elkanii strains that produce rhizobitoxine, a phytotoxin that causes foliar chlorosis on susceptible host plants (16).


We report here that NBS-LRR type disease resistance genes are involved in determination of symbiotic specificity in soybean, through positional cloning of the Rj2 and Rfg1 genes that restrict nodulation with specific strains of B. japonicum and S. fredii, respectively. Our discovery is consistent with recent reports describing rhizobial T3SS and its secreted effectors that play an important role in modulation of host range. It remains to be seen whether multiple or only a few R gene loci are involved in determination of nodulation specificity and whether these loci are conserved in legumes. The Rj2/Rfg1 locus exhibits highly conserved synteny with the orthologous regions in M. truncatula (AC156629) and L. japonicus (AP010004) that also contain the closest Rj2/Rfg1 homologs. This observation may facilitate testing whether rhizobial recognition specificity has an ancient origin during legume evolution. Our finding may also offer unique strategies to enhance symbiotic nitrogen fixation in crop legumes. For example, the nodulation-restrictive R genes may be manipulated so that a host can deterministically interact with rhizobial inoculants with high nitrogen-fixing efficiency and exclude those indigenous strains that are highly competitive but with very low nitrogen-fixing efficiency (16). It will also be interesting to know whether these types of R genes shape the diversity of the rhizobial community in the rhizosphere. Much remains to be learned about how an R gene conditions the beneficial plant–microbe interactions in natural ecosystems.

Materials and Methods

Plant Materials and Nodulation Assay.

The F2 mapping populations were derived from the crosses between L76-1988 (Rj2/Rj2, rfg1/rfg1) and L82-2024 (rj2/rj2, Rfg1/Rfg1) and between Williams 82 (rj2/rj2, Rfg1/Rfg1) and Peking (rj2/rj2, rfg1/rfg1). Plants were grown in sterile vermiculite in a growth chamber programmed for 12 h light at 26 °C and 12 h dark at 23 °C. Roots of 1-wk-old seedlings were inoculated with B. japonicum USDA122 or S. fredii USDA257. Nodulation phenotypes were recorded 2 wk postinoculation.

Complementation Tests and RNAi-Mediated Gene Silencing.

For Rfg1, we used the Rfg1 (rj2) allele of Williams 82 for complementation tests. Briefly, genomic DNA of the BAC clone Gm_WBa0019D20 of Williams 82 was digested with PstI and BmgBI to obtain a 10.9-kb fragment that contains the ~4.9-kb coding region plus ~4.0 kb upstream of the start codon and ~2.0 kb downstream of the stop codon. The genomic fragment was introduced into the binary vector pCAMBIA1305.1 through blunt end cloning. For Rj2, we developed a genomic construct consisting of a synthetic L76-1988 allele on the basis of its allelic sequence of Williams 82. The genomic construct described above was digested with EcoRI to delete a ~2.2-kb fragment that contains all of the polymorphic sites detected between the coding sequences of Williams 82 and L76-1988. The corresponding genomic fragment of L76-1988 was then amplified by PCR and ligated into the digested pCAMBIA1305.1 vector using the In-Fusion Advantage PCR Cloning kits (Clontech). For silencing of Rj2, an RNAi construct was generated by cloning a 405-bp inverted-repeat sequence of the NBS-encoding region into the pRedRoot II vector that contains the red fluorescent marker DsRed1.

Hairy Root Transformation of Soybean.

A. rhizogenes-mediated hairy root transformation was performed on the basis of the protocol described by Kereszt et al. (26). Briefly, the A. rhizogenes strain K599 that contains a binary vector was injected into the cotyledonary node of 1-wk-old seedlings. The infected seedlings were maintained in sterile vermiculite in a growth chamber with 90% humidity. Two to 3 wk after infection, when hairy roots were well developed at the infection site, the main roots were removed, and the composite plants were inoculated with the rhizobial strains USDA122 or USDA257. Nodulation assays were performed 2 wk after inoculation. The transgenic roots were identified through GUS staining of a small portion of the hairy roots in the cases of complementation tests or through detection of the red fluorescent marker DsRed1 using a fluorescent microscope.

Analysis of Gene Expression by RT-PCR.

Total RNA was isolated by the Qiagen Plant RNeasy mini kit. Two micrograms of RNA was used to perform RT reactions using M-MLV reverse transcriptase (Invitrogen) in a 20-μL reaction mixture. Two microliters of the RT reaction was used as a template in a 20-μL PCR solution. The PCR primers were as follows: Rj2-L76 specific, 5′ATGGCAATTCTGTTATGGAAGACTC3′ and 5′CTGGCCTTCCATTAGCTTTG3′; Rfg1-Williams 82-specific, 5′ATGACAATTCTGTTATGGAAGAGTT3′ and 5′CTGGCCTTCCATTAGCTTTG3′; Rj2/Rfg1 specific primers, 5′GGCACCTCATACGGACTGTT3′ and 5′AGGCTCAGAAGTTTTCCACCT3′; and Gm-Actin, 5′GAGCTATGAATTGCCTGATGG3′ and 5′CGTTTCATGAATTCCAGTAGC3′.

Anatomical Analysis.

Soybean seeds were surface sterilized and germinated on 1% water agar at 30 °C for 3 d. The roots of the seedlings were dipped into bacterial suspensions (109 cells/mL) and transferred to autoclaved plastic growth pouches that were premoistened with nitrogen-free nutrient solution. The location of the root tip of each seedling was marked on the surface of the pouch. The plants were grown for 3 d at 24 °C under a 12-h photoperiod. Primary roots extending 5 cm above and 5 cm below the root tip mark were dissected, stained with 0.01% methylene blue, and examined for root hair curling under a dissecting microscope using bright field illumination. Soybean roots harvested at 5 and 8 d after inoculation were cut into small pieces and immediately fixed in 50% ethyl alcohol, 5% glacial acetic acid, and 10% formaldehyde for 24 h at 4 °C. The tissues were dehydrated in a graded ethanol series followed by a graded series of xylene. Following this, the roots were infiltrated with several changes of paraffin at 60 °C and embedded in paraffin. Embedded tissues were sectioned (10 μm thick) with a microtome, stained with hematoxylin and eosin, and examined with bright-field optics.

Electron Microscopy.

Soybean roots inoculated with B. japonicum USDA122 were dissected into 2- to 4-mm pieces with a double-edge razor blade and fixed immediately in 2.5% glutaraldehyde buffered at pH 7.2 with 50 mM sodium phosphate buffer. The tissues were fixed at room temperature for 4 h, washed extensively with five changes of phosphate buffer, and postfixed with 2% aqueous osmium tetroxide for 1 h. Following this, the tissues were rinsed with several changes of buffer, dehydrated in a graded acetone series, and infiltrated with Spurr's resin. Thin sections were cut with a diamond knife, collected on uncoated 200-mesh copper grids, and stained with 0.5% uranyl acetate and 0.4% lead citrate. Stained sections were examined with a JEOL 1200 EX transmission electron microscope at 80 kV.


We thank Douglas Cook, Varma Penmetsa, and Jens Stougaard for helpful comments on the manuscript. We also thank Dr. Gary Stacey (University of Missouri, Columbia, MO) for providing the A. rhizogenes strain K599, Dr. Jean-Michel Ané (University of Wisconsin, Madison, WI) for the pRedRootII vector, and Dr. Peter van Berkum (US Department of Agriculture-Agricultural Research Service, Beltsville, MD) for the B. japonicum strain USDA122. This research was supported by a competitive grant from the US Department of Agriculture–Agriculture and Food Research Initiative (Grant 2009-65300-05663) (to H.Z.).


1. Madsen LH, et al. The molecular network governing nodule organogenesis and infection in the model legume Lotus japonicus. Nat Commun. 2010;1:10. [PMC free article] [PubMed]
2. Oldroyd GE, Harrison MJ, Paszkowski U. Reprogramming plant cells for endosymbiosis. Science. 2009;324:753–754. [PubMed]
3. Perret X, Staehelin C, Broughton WJ. Molecular basis of symbiotic promiscuity. Microbiol Mol Biol Rev. 2000;64:180–201. [PMC free article] [PubMed]
4. Lerouge P, et al. Symbiotic host-specificity of Rhizobium meliloti is determined by a sulphated and acylated glucosamine oligosaccharide signal. Nature. 1990;344:781–784. [PubMed]
5. Bec-Ferté MP, Krishnan HB, Savagnac A, Pueppke SG, Promé JC. Rhizobium fredii synthesizes an array of lipooligosaccharides, including a novel compound with glucose inserted into the backbone of the molecule. FEBS Lett. 1996;393:273–279. [PubMed]
6. Stacey G. Bradyrhizobium japonicum nodulation genetics. FEMS Microbiol Lett. 1995;127:1–9. [PubMed]
7. Geurts R, et al. Sym2 of pea is involved in a nodulation factor-perception mechanism that controls the infection process in the epidermis. Plant Physiol. 1997;115:351–359. [PMC free article] [PubMed]
8. Limpens E, et al. LysM domain receptor kinases regulating rhizobial Nod factor-induced infection. Science. 2003;302:630–633. [PubMed]
9. Radutoiu S, et al. Plant recognition of symbiotic bacteria requires two LysM receptor-like kinases. Nature. 2003;425:585–592. [PubMed]
10. Radutoiu S, et al. LysM domains mediate lipochitin-oligosaccharide recognition and Nfr genes extend the symbiotic host range. EMBO J. 2007;26:3923–3935. [PMC free article] [PubMed]
11. Deakin WJ, Broughton WJ. Symbiotic use of pathogenic strategies: Rhizobial protein secretion systems. Nat Rev Microbiol. 2009;7:312–320. [PubMed]
12. D'Haeze W, Holsters M. Surface polysaccharides enable bacteria to evade plant immunity. Trends Microbiol. 2004;12:555–561. [PubMed]
13. Soto MJ, Domínguez-Ferreras A, Pérez-Mendoza D, Sanjuán J, Olivares J. Mutualism versus pathogenesis: The give-and-take in plant-bacteria interactions. Cell Microbiol. 2009;11:381–388. [PubMed]
14. Jones KM, et al. Differential response of the plant Medicago truncatula to its symbiont Sinorhizobium meliloti or an exopolysaccharide-deficient mutant. Proc Natl Acad Sci USA. 2008;105:704–709. [PMC free article] [PubMed]
15. Büttner D, He SY. Type III protein secretion in plant pathogenic bacteria. Plant Physiol. 2009;150:1656–1664. [PMC free article] [PubMed]
16. Devine TE, Kuykendall LD. Host genetic control of symbiosis in soybean (Glycine max L.) Plant Soil. 1996;186:173–187.
17. Caldwell BE. Inheritance of a strain specific ineffective nodulation in soybeans. Crop Sci. 1966;6:427–428.
18. Trese AT. A single dominant gene in McCall soybean prevents effective nodulation with Rhizobium fredii USDA257. Euphytica. 1995;81:279–282.
19. Vest G, Caldwell BE. Rj4: A gene conditioning ineffective nodulation in soybean. Crop Sci. 1972;12:692.
20. Zhukov V, et al. The pea Sym37 receptor kinase gene controls infection-thread initiation and nodule development. Mol Plant Microbe Interact. 2008;21:1600–1608. [PubMed]
21. Indrasumunar A, et al. Inactivation of duplicated nod factor receptor 5 (NFR5) genes in recessive loss-of-function non-nodulation mutants of allotetraploid soybean (Glycine max L. Merr.) Plant Cell Physiol. 2010;51:201–214. [PubMed]
22. Dangl JL, Jones JDG. Plant pathogens and integrated defence responses to infection. Nature. 2001;411:826–833. [PubMed]
23. Polzin KM, Lohnes DG, Nickell CD, Shoemaker RC. Integration of Rps2, Rmd, and Rj2 into linkage group J of the soybean molecular map. J Hered. 1994;85:300–303.
24. Graham MA, Marek LF, Lohnes D, Cregan P, Shoemaker RC. Expression and genome organization of resistance gene analogs in soybean. Genome. 2000;43:86–93. [PubMed]
25. Schmutz J, et al. Genome sequence of the palaeopolyploid soybean. Nature. 2010;463:178–183. [PubMed]
26. Kereszt A, et al. Agrobacterium rhizogenes-mediated transformation of soybean to study root biology. Nat Protoc. 2007;2:948–952. [PubMed]
27. Sadowsky MJ, Cregan PB, Rodriguez-Quinones F, Keyser HH. Microbial influence on gene-for-gene interactions in legume-Rhizobium symbioses. Plant Soil. 1990;129:53–60.
28. Krishnan HB, et al. Extracellular proteins involved in soybean cultivar-specific nodulation are associated with pilus-like surface appendages and exported by a type III protein secretion system in Sinorhizobium fredii USDA257. Mol Plant Microbe Interact. 2003;16:617–625. [PubMed]
29. Marie C, et al. Characterization of Nops, nodulation outer proteins, secreted via the type III secretion system of NGR234. Mol Plant Microbe Interact. 2003;16:743–751. [PubMed]
30. Kambara K, et al. Rhizobia utilize pathogen-like effector proteins during symbiosis. Mol Microbiol. 2009;71:92–106. [PubMed]
31. Chatterjee A, Balatti PA, Gibbons W, Pueppke SG. Interaction of Rhizobium fredii USDA257 and nodulation mutants derived from it with the agronomically improved soybean cultivar McCall. Planta. 1990;180:303–311. [PubMed]
32. Meinhardt LW, Krishnan HB, Balatti PA, Pueppke SG. Molecular cloning and characterization of a sym plasmid locus that regulates cultivar-specific nodulation of soybean by Rhizobium fredii USDA257. Mol Microbiol. 1993;9:17–29. [PubMed]
33. Shaw SL, Long SR. Nod factor inhibition of reactive oxygen efflux in a host legume. Plant Physiol. 2003;132:2196–2204. [PMC free article] [PubMed]
34. Jones JD, Dangl JL. The plant immune system. Nature. 2006;444:323–329. [PubMed]
35. Parker MA. Mutualism in metapopulations of legumes and rhizobia. Am Nat. 1999;153:S48–S60.

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...