|Home | About | Journals | Submit | Contact Us | Français|
The nitrogen-fixing symbiont Sinorhizobium meliloti senses and responds to constantly changing environmental conditions as it makes its way through the soil in search of its leguminous plant host, Medicago sativa (alfalfa). As a result, this bacterium regulates various aspects of its physiology in order to respond appropriately to stress, starvation, and competition. For example, exopolysaccharide production, which has been shown to play an important role in the ability of S. meliloti to successfully invade its host, also helps the bacterium withstand osmotic changes and other environmental stresses. In an effort to further elucidate the intricate regulation of this important cell component, we set out to identify genetic factors that may affect its production. Here we characterize novel genes that encode a small protein (EmmA) and a putative two-component system (EmmB-EmmC). A mutation in any of these genes leads to increased production of the symbiotically important exopolysaccharide succinoglycan. In addition, emm mutants display membrane-associated defects, are nonmotile, and are unable to form an optimal symbiosis with alfalfa, suggesting that these novel genes may play a greater role in the overall fitness of S. meliloti both during the free-living stage and in its association with its host.
The nitrogen-fixing symbiotic rhizobia possess a variety of mechanisms that allow them to exist in and adapt to various environmental conditions as they seek out their leguminous plant hosts. Coordination of many cellular processes, including exopolysaccharide production, motility, stress adaptation, and biofilm production, must occur in order for rhizobia to persist in their soil and plant settings (5, 33, 50, 59). In the gram-negative soil bacterium Sinorhizobium meliloti, the production of exopolysaccharides is particularly crucial because it provides the capacity to form an association with Medicago sativa (alfalfa) (27, 33). Indeed, exopolysaccharides have been shown to be essential because they provide protection from desiccation, allow biofilm formation, improve soil structure in the rhizosphere, and facilitate symbiotic associations with the plant host (1, 33, 38, 43, 52, 57).
S. meliloti strain Rm8530 has the ability to produce two distinct exopolysaccharides, succinoglycan and EPS II, and the presence of at least one is necessary to mediate symbiosis with alfalfa (4, 19, 29, 54). Mutants unable to produce both of these exopolysaccharides are impaired in initiation of infection threads (the plant-derived, tube-like structures through which S. meliloti enters the plant to reach a nodule) and nodule invasion and therefore are defective in establishing a proper nitrogen-fixing association with the host (12, 33, 45). The production of EPS II by S. meliloti confers a mucoid phenotype to the wild-type strain that sharply contrasts with the dry-colony morphology of the commonly used laboratory strain Rm1021. The latter strain is unable to produce EPS II due to the presence of an insertion element in the chromosomal gene expR, a global quorum-sensing regulator in S. meliloti that plays a central role in many important cellular processes, including the production of low-molecular-weight succinoglycan and EPS II (25, 31, 38, 46).
Various studies have shed light on the precise role that exopolysaccharides play during the establishment and maintenance of symbiosis between alfalfa and S. meliloti. Evidence indicates that the low-molecular-weight form of either succinoglycan or EPS II is the active component that facilitates plant invasion (3, 28, 45, 58, 60). Each of these exopolysaccharides may promote nodule invasion via distinct mechanisms. For example, succinoglycan appears to be more efficient than EPS II in mediating nodule invasion, as it facilitates colonization of curled root hairs, as well as initiation and extension of the infection threads earlier and in greater abundance during the invasion process (12, 45). On the other hand, EPS II appears to be the major contributor to biofilm production compared to succinoglycan. Strains that produce EPS II, irrespective of the presence of succinoglycan, show a dramatic increase in biofilm formation compared to strains unable to make this polymer (L. Rinaudi and J. E. González, submitted for publication). Additionally, exopolysaccharides have been shown to play a role in stress adaptation and resistance to desiccation (14, 41, 43, 53, 59).
In light of the fact that S. meliloti exopolysaccharides play such a critical role in the soil and plant environments, it is not surprising that their biosynthesis is coordinated with other crucial phenotypes through participation of various central regulatory mechanisms. For example, the transcriptional regulator ExoR and the ExoS-ChvI two-component system control not only succinoglycan production but also motility and biofilm formation (61, 62). A mutation in cbrA, which encodes a stationary phase-induced sensor kinase, leads to overproduction of succinoglycan, a decrease in flagellar biosynthesis, and cell envelope defects (21, 22). In addition to the regulatory mechanisms described above, the ExpR/Sin quorum-sensing system controls EPS II and low-molecular-weight succinoglycan synthesis, as well as many other important cell functions, including motility and chemotaxis, in a quorum-sensing-dependent manner (25, 31, 38, 45, 46).
Because of this complex regulation, the precise role of exopolysaccharides in S. meliloti is inherently difficult to determine. We set out to elucidate other possible genetic factors that may affect the regulation of succinoglycan biosynthesis in isolation of EPS II by using an expR mutant strain (38, 46). Random transposon mutagenesis was performed, and mutants were examined to determine changes in morphology. During our preliminary screening, several mucoid candidates were isolated, which suggested that succinoglycan production was increased. Analysis of one of these mucoid candidates revealed that the disrupted gene, SMb21521::Tn5, is immediately adjacent to genes encoding an uncharacterized two-component system consisting of a putative histidine sensor kinase (SMb21519) and a response regulator (SMb21520). The presence and orientation of these three genes are highly conserved among rhizobial species. A mutation in the SMb21521, SMb21519, or SMb21520 locus, which we designated emmA, emmB, and emmC, respectively, leads to defects in exopolysaccharide production, motility, and membrane stress sensitivity, as well as a decrease in the efficiency of symbiosis with alfalfa. This paper discusses characterization of this novel three-component system, which appears to play a crucial role in the overall maintenance of necessary cellular processes critical to the adaptability and survivability of S. meliloti.
The bacterial strains and plasmids used in this study are listed in Table Table1.1. S. meliloti strains were grown at 30°C in Luria-Bertani (LB) medium supplemented with 2.5 mM MgSO4 and 2.5 mM CaCl2 for routine cultures (LBMC), unless otherwise specified. Escherichia coli strains were grown in LB broth or agar with the appropriate antibiotics. Strains were created by generalized transduction using phage M12 as previously described (23). Growth curve studies were performed by inoculating 2 ml TYC medium (10 g of tryptone/liter, 5 g of yeast extract/liter, 0.4 g of CaCl2/liter) containing the appropriate antibiotics with a single colony of each S. meliloti strain and growing the cultures at 30°C to saturation (2 days). Saturated cultures were then washed and subcultured (1:100) in minimal glutamate mannitol low-phosphate broth (MLP) (50 mM morpholinepropanesulfonic acid [MOPS], 19 mM sodium glutamate, 55 mM mannitol, 0.1 mM K2HPO4 · KH2PO4 [the stock solution contained equal molar concentrations of the two compounds], 1 mM MgSO4, 0.25 mM CaCl2, 0.004 mM biotin; pH 7.0) as previously described (31). Cell density was measured by monitoring the optical density at 600 nm (OD600), and all growth curve experiments were performed in triplicate. Antibiotics were used at the following concentrations: streptomycin, 500 μg/ml; neomycin, 200 μg/ml; hygromycin, 100 μg/ml; and gentamicin, 100 μg/ml (10 μg/ml for E. coli). To observe succinoglycan biosynthesis, LBMC agar plates were supplemented with calcofluor (Fluorescent Brightener 28; New Dragon Co.) at a final concentration of 0.02% buffered with 10 mM HEPES (pH 7.4).
Random transposon mutagenesis of the expR strain was performed as previously described (23). Mutants obtained using the transposon mutagenesis procedure were grown on LBMC agar with the appropriate antibiotics and incubated at 30°C. Candidates were observed to identify differences in exopolysaccharide production based on the mucoid appearance and selected for further analysis. The SMb21521:Tn5 mutation was identified by performing arbitrary PCR with the mutant chromosomal DNA as previously described (15). The PCR products were purified, sequenced (Macrogen USA, Rockville, MD), and then analyzed using a BLAST search. EmmB and EmmC mutants (2011mTn5STM.2.06.D08 and 2011mTn5STM.2.09.C10, respectively) were obtained from A. Becker (47), and the mutations were transduced into the Rm8530 and Rm1021 backgrounds as previously described (23).
Cultures were grown in MLP to an OD600 of 1.0 and centrifuged at 4°C for 15 min at 3,220 × g. The total carbohydrate concentration of the exopolysaccharide-containing supernatant was determined using the anthrone-H2SO4 method (40). The amount of carbohydrate was normalized to the total protein concentration of the corresponding cell pellet, which was determined using the Bio-Rad DC protein assay. Three independent cultures of each strain were used to calculate the average and standard deviation.
Bacterial cultures were grown and RNA was extracted as previously described (31). First-strand cDNA for each strain was prepared with a RETROscript kit from Ambion by using 1 μg of total RNA in each reaction mixture. One microliter of a cDNA reaction mixture was used as a template for real-time PCR. The oligonucleotide sequences used for real-time PCR are listed in Table Table2.2. The SMc00128 gene (34) was used as a control for equal loading in each real-time PCR. Each 25-μl reaction mixture contained 0.3 μM sense oligonucleotide, 0.3 μM antisense oligonucleotide, 0.5× SYBR green 1 (Sigma), and one-half of an Omni Mix HS PCR bead (each PCR bead contains 1.5 U of Taq DNA polymerase, 10 mM Tris-HCl [pH 9.0], 50 mM KCl, 1.5 mM MgCl2, 200 μM deoxynucleoside triphosphate, and stabilizers, including bovine serum albumin). The experiment was performed with a Cepheid Smart Cycler version 2.0c programmed as follows: stage 1, 95°C for 120 s; and stage 2, 40 cycles of 95°C for 15 s and 60°C for 30 s. For reverse transcription (RT)-PCR, 1 μl of cDNA was mixed with 24 μl of a master mixture using a MasterTaq kit (5 prime), and the reaction was performed under the following conditions: 95°C for 5 min, followed by 29 cycles of 95°C for 30 s, 60°C for 30 s, and 72°C for 1 min and then 72°C for 10 min. RT-PCR products were visualized by electrophoresis.
To complement the emmA and emmC mutations, SMb21521 and SMb21920, respectively, were individually cloned into the XbaI site of the arabinose-inducible plasmid pJN105. In S. meliloti, this plasmid is constitutively expressed, and the inducibility by arabinose is marginal; therefore, addition of arabinose for induction is not necessary (42). For emmA, a 700-bp fragment was amplified from wild-type chromosomal DNA by using forward primer 5′-CGCGCGCTCTAGACGATGCCGACGATCGGTGAAA-3′ and reverse primer 5′-CGGCGCGTCTAGAGTCGCCTCGGGAATGAGCAAC-3′, and for emmC, an 856-bp fragment was amplified from wild-type chromosomal DNA by using forward primer 5′-CGCGCGTCTAGACCGCGTCGAGAGCCTCGGCGG −3′ and reverse primer 5′-CGCGCGTCTAGAGAGGTGATCGAAGGCGCCCGCT-3′ (the underlined nucleotides are XbaI restriction enzyme sites). The correct orientation of the SMb21521 or SMb21520 fragment was confirmed using the forward primer for the respective gene in a reaction with the pJN105 mcs Rev primer (5′-GAGCTCCACCGCGGTGGCGGCCGC-3′). The individual plasmids were then separately transformed into E. coli, and candidates were selected with gentamicin. The recombinant plasmids were introduced into the different S. meliloti strains by triparental mating, and transconjugants were selected with neomycin and gentamicin. All strain constructs were confirmed by PCR.
For detergent sensitivity analysis, assays were performed as previously described (22). Briefly, cultures were grown to exponential phase (OD600, 0.5) in LBMC and then diluted to obtain an OD600 of 0.1 in LB medium. Each sample was serially diluted in LB medium, and 50 μl was spread onto LB agar or LB agar containing 0.1% sodium deoxycholate (DOC). After 4 to 5 days of growth at 30°C, the number of CFU was determined. Three independent cultures of each strain were used to calculate the average and standard deviation.
The effect of high osmolarity on growth was assessed as described previously (17). Briefly, strains were grown at 30°C in MLP to an OD600 of 0.3 to 0.4, and then sodium chloride (NaCl) was added to a final concentration of 300 mM. Cell density was measured by monitoring the OD600 at regular time intervals. Three independent cultures of each strain were used to calculate the average and standard deviation.
Plate assays to observe S. meliloti motility were conducted by growing strains in TYC medium at 30°C to saturation. Two microliters of each strain was then stabbed into plates containing 0.04% tryptone, 0.01% yeast extract, 0.01% calcium chloride, and 0.3% agar. The plates were incubated at 30°C for 2 to 3 days and monitored to determine growth away from the point of inoculation.
To detect the presence of flagellin proteins, immunoblotting was performed, with some modifications, as previously described (56). Briefly, wild-type and emmA, emmC, and flaA flaB mutant strains were grown in TYC medium, and 1-ml aliquots of each cell culture at an OD600 of 0.2 were harvested by centrifugation at 20,000 × g for 6 min. The cells were resuspended in 20 μl sodium dodecyl sulfate sample buffer and heated to 100°C for 8 min. Equal amounts of sample proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrophoretically transferred from the gel to a polyvinylidene difluoride membrane (Hybond LFP; Amersham) at 30 V for 2 h at 4°C. The polyvinylidene difluoride membrane was blocked overnight at 4°C with Tris-buffered saline-Tween 20 (pH 7.5). The membrane was probed with anti-Fla polyclonal antibodies at a 1:500 dilution for 2 h at room temperature with agitation, washed three times (10 min each) with fresh Tris-buffered saline-Tween 20 (pH 7.5), and probed with goat anti-rabbit Cy3 fluorescent antibody (Amersham). Fluorescent bands were imaged using a Typhoon scanner.
To test the nodulation and nitrogen fixation abilities of the emmA, emmA exoY, emmC, and emmC exoY mutants, plant assays with the symbiotic host M. sativa were carried out in triplicate experiments using 35 plants in each experiment. S. meliloti cultures were grown to saturation in LBMC with the appropriate antibiotics at 30°C until saturation. A 1:100 dilution in water was used to inoculate 3-day-old plant seedlings on Jensen's agar as previously described (36). The plates were incubated at 25°C and 60% relative humidity with a 16-h light cycle. The plants were monitored weekly, and after 4 weeks the height of each plant was determined and the pink (nitrogen-fixing) and white nodules were enumerated.
In an effort to identify additional genes involved in the regulation of succinoglycan production, random transposon mutagenesis of a strain that produces only this polymer was performed. Our screen revealed several mucoid candidates, suggesting that exopolysaccharide production was increased in these derivatives. The genomic location of the transposon disruption corresponded to a small 459-bp open reading frame annotated as SMb21521 (emmA). This gene is adjacent to and transcribed divergently from genes encoding an uncharacterized putative two-component system containing a sensor kinase (SMb21519) and a response regulator (SMb21520), which we designated EmmB and EmmC, respectively (see Fig. S1 in the supplemental material). Sequence analysis of the derived 152-amino-acid EmmA protein showed that it contains a predicted export signal sequence with an overlapping transmembrane region at its N terminus. Such signal domains are frequently found in proteins located in the periplasm. Bioinformatic analysis of the derived 490-amino-acid EmmB protein revealed two transmembrane domains near its N terminus and a histidine kinase-like ATPase sequence at its C terminus, which is consistent with the structure of membrane-bound sensor kinases. The 219-amino-acid EmmC protein is predicted to contain a receiver domain and a helix-turn-helix DNA-binding domain, which is typical of response regulators. The emm locus appears to be highly conserved only in rhizobial species with homologs of these three genes oriented in a similar fashion (Fig. (Fig.1A).1A). A BLAST search with EmmA revealed no significant homology to any known protein in bacteria other than rhizobia. Likewise, EmmB and EmmC are highly conserved in rhizobial species (Fig. (Fig.1B).1B). Interestingly, no orthologs with significant similarity to EmmA were found in either Agrobacterium tumefaciens or Brucella sp., despite a strong overlap between their genome sequences and those of rhizobia (20, 44). The fact that the emm locus appears to be highly conserved mainly in nitrogen-fixing symbionts suggests that it may play an important and specific role in these organisms.
The emmA mutant was isolated based on the mucoid phenotype in a background strain that lacks an intact expR gene and, therefore, is not capable of producing EPS II (38, 46). No change in the expression of the EPS II biosynthesis genes (exp) was observed (Table (Table3).3). Therefore, it was likely that succinoglycan was the exopolysaccharide causing the mucoid phenotype. Succinoglycan has a unique quality: it is able to bind to the dye calcofluor and fluoresce under UV light (18, 36). As a result, calcofluor has been used extensively to identify and help characterize mutations that affect succinoglycan production (18, 36, 37). When the emmA mutation was transduced either into the wild-type background (Fig. (Fig.2)2) or into a strain with a disrupted expR gene (see Fig. S2 in the supplemental material), very bright fluorescence was observed. The fluorescence was eliminated by disruption of the exoY gene, the first gene in the succinoglycan biosynthetic pathway (49), confirming that the production of this polymer was affected by the emmA mutation. Fluorescence was restored to wild-type levels by introduction of a plasmid containing a constitutively expressed emmA gene (pJNemmA). Since EmmA is not predicted to contain any DNA-binding motifs, we inquired whether the nearby genes, which are predicted to code for a two-component system, might act in conjunction with EmmA and affect succinoglycan production. Thus, the role of these genes was characterized further by observing how individual mutations affected succinoglycan production. Similar to the emmA mutant, an emmC mutant displayed very bright fluorescence when it was streaked on a calcofluor plate, and fluorescence did not occur in the presence of an exoY mutation, which eliminates succinoglycan production (Fig. (Fig.2;2; see Fig. S2 in the supplemental material). Fluorescence was restored to wild-type levels by introduction of a plasmid containing a constitutively expressed emmC gene (pJNemmC), confirming that, similar to a disruption in the emmA gene, a disruption in emmC affects succinoglycan production. We observed a similar calcofluor-bright phenotype for an emmB mutation (data not shown). These results indicate that the emm gene products are involved in succinoglycan production and that their effect is independent of regulation via the quorum-sensing regulator ExpR.
An analysis of the total carbohydrate from emm mutant cultures grown to early stationary phase (OD600, 1.0) was performed to quantify succinoglycan production. Strains that produce only succinoglycan (expR, expR emmA, and expR emmC mutants) were used so that the polymer could be quantified without interference from the abundantly produced polymer EPS II (38, 46). An eightfold increase in the total carbohydrate content was observed for both the expR emmA and expR emmC strains compared to the expR strain, and normal levels of carbohydrate production were restored by introduction of the pJNemmA or pJNemmC complementation plasmid into the appropriate mutant (Fig. (Fig.3).3). A mutation in emmB also resulted in increased carbohydrate production (data not shown). These results suggest that disruption of the emm locus leads to a quantifiable increase in succinoglycan production.
Quantitative real-time PCR studies were performed for the exoY gene in order to determine the effect of the emm locus on succinoglycan production at the level of gene expression (25). Consistent with the results of the total carbohydrate analysis, the expression of exoY was 3.3-fold greater in the emmA mutant than in the wild type (Table (Table3).3). An even greater increase in the expression of exoY was observed in the presence of an emmB or emmC mutation (Table (Table3).3). The calcofluor-bright phenotype, the upregulation of exoY expression, and the overall increase in carbohydrate production in a strain that produces only succinoglycan suggest that the emm locus directly or indirectly plays a role in the biosynthesis of this important exopolysaccharide.
Changes in exopolysaccharide production are often correlated with increased sensitivity to different membrane stressors, such as detergents (22, 35). In keeping with this, we found that a succinoglycan-deficient strain appears to be more sensitive to the membrane stress caused by exposure to DOC (Fig. (Fig.4;4; see Fig. S3 in the supplemental material). If succinoglycan provides a mechanism that protects the cells from potential membrane stresses, then one might expect that a larger amount of this polymer, as observed in emm locus mutants, might increase the cells' ability to withstand environmental stresses. However, mutations in emmA or emmC led to a 7-order-of-magnitude increase in the sensitivity of S. meliloti to detergent stress, and this susceptibility was eliminated when the mutants were complemented with a functional copy of the emmA or emmC gene (Fig. (Fig.4).4). The exoY emmA and exoY emmC double mutants displayed slightly more pronounced sensitivity to DOC than the emmA and emmC single mutants, consistent with the hypothesis that a succinoglycan-deficient strain is compromised. Indeed, we found that an exoY mutant displayed innate sensitivity to increased concentrations of detergent compared to the wild type, indicating that succinoglycan protects against potential membrane stressors (see Fig. S3 in the supplemental material). Similar to these results, a mutation in the putative sensor kinase gene emmB also resulted in increased susceptibility to DOC (data not shown). Despite the previous finding that such mutants overproduce succinoglycan, these data indicate that the emm locus plays a role in maintaining membrane integrity during stress and that this role is independent of succinoglycan production.
Fluctuations in osmolarity can potentially provide a source of stress for S. meliloti as it encounters different soil environments, and physiological abnormalities may be revealed by an inability to withstand osmotic shock (16, 59). The role of the emm locus during osmotic stress was assessed by exposing emmA and emmC mutants to high concentrations of salt (NaCl) and comparing their growth to that of the wild type and other derivatives. No difference in growth was detected in any of the strains grown in minimal medium alone (see Fig. S4 in the supplemental material). As observed for detergent stress, the exoY strain displayed innate sensitivity to osmotic shock (Fig. (Fig.5).5). The presence of the emmA or emmC mutation resulted in a decreased ability to grow in the presence of high salt concentrations in both the wild-type and exoY backgrounds, and this growth defect was eliminated by introduction of the pJNemmA and pJNemmC complementation plasmids, respectively (Fig. (Fig.5).5). Increased sensitivity to changes in osmolarity was also observed for the emmB mutant (data not shown). These data indicate that the emm locus is involved in protecting S. meliloti from membrane stresses, and this ability is optimal in the presence of succinoglycan.
The regulation of bacterial exopolysaccharide biosynthesis and motility are intimately coupled in many species of bacteria, including Ralstonia solanacearum, Vibrio cholerae, and Salmonella enterica (2, 6, 9). In S. meliloti, mutants with mutations in ExoR and ExoS, both of which are involved in the regulation of succinoglycan production, have been shown to have motility defects that include loss of flagellum production (61, 62). Since succinoglycan production was affected in the emm mutants, we tested the abilities of these mutants to move on media containing 0.3% agar. No difference in the size of the colonies was observed between the wild-type strain and an exoY mutant (data not shown). However, both the emmA and emmC mutants displayed a dramatic phenotype by forming small, compact colonies independent of succinoglycan production, like the well-characterized visN motility mutant (56). Complementation of the emmA and emmC mutants with the pJNemmA and pJNemmC plasmids, respectively, restored the motility to wild-type levels (Fig. (Fig.6A).6A). A disruption in emmB also resulted in a defective-motility phenotype similar to that of the emmA and emmC mutants (data not shown). The expression of several key motility genes was analyzed using cultures grown to early log phase (OD600, 0.2), the growth phase at which these genes are maximally induced in the wild-type strain (30, 32). The expression of visN, rem, and two flagellar genes, flbT (VisN-VisR dependent) and flgB (Rem dependent), was analyzed (31, 51, 56). Five- and sixfold decreases in the expression of visN and flbT, respectively, were observed, and even greater decreases (around 14-fold) in the expression of rem and flgB were detected in the emm mutants (Fig. (Fig.6B).6B). To further characterize the effect of the emm locus on flagellum synthesis, the production of flagellin (Fla) protein in the emmA and emmC mutants was compared to that in the wild-type strain using specific antibodies. The Fla protein was detected in the wild-type strain, but the amounts were significantly reduced in the emm mutants and in the flaA flaB flagellar mutant, which was used as a negative control (Fig. (Fig.6C6C).
In S. meliloti, the VisN-VisR pair of transcriptional regulators is at the apex of the flagellar regulon and is responsible for the direct activation of many genes involved in the flagellar assembly and motor functions. VisN-VisR also controls a subset of motility genes via the action of another intermediate regulator, Rem (31, 51, 56). To determine if the motility defect observed in mutants with mutations in the emm locus was due to downregulation of visN-visR, which would result in a decrease in expression of the rest of the motility genes, a plasmid constitutively expressing visN-visR (30) was introduced into the emmA and emmC mutants, and the expression of the motility genes was studied. While visN and the visN-controlled gene flbT were expressed at wild-type levels, the expression of rem and flgB was still significantly downregulated (Fig. (Fig.6B).6B). These results suggest that the emm locus does not affect motility solely through the VisN-VisR regulatory pathway but may have a more global effect on motility gene expression.
Appropriate regulation of the production of exopolysaccharides is necessary for a successful symbiosis between S. meliloti and alfalfa. Mutations in the negative regulators of succinoglycan production exoR and exoX cause overproduction of succinoglycan (48, 63). Although strains with these mutations can induce nodules on plants, they are unable to fix nitrogen (33). Since disruption of the emm locus also causes an increase in succinoglycan biosynthesis, its effect on the S. meliloti-alfalfa symbiosis was determined. The emmA and emmC mutations were analyzed both in the wild type and in the presence of an exoY mutation. In a wild-type background, an exoY mutant invades alfalfa since it still produces the other symbiotically important exopolysaccharide, EPS II.
No significant difference was observed between the abilities of the wild type and either the emmA or emmC mutant to form nodules on alfalfa plants. In plants inoculated with the wild type, the emmA mutant, or the emmC mutant, the majority of the nodules (86, 92, and 88%, respectively) were pink, indicating that nitrogen fixation was occurring, and only a small proportion of the nodules (14, 8, and 12%, respectively) were white (Fig. (Fig.7).7). However, in plants inoculated with the exoY emmA double mutant and in plants inoculated with the exoY emmC double mutant, there were significant decreases in the total number of nodules (33 and 35% fewer nodules, respectively), and the majority of the nodules were empty (67 and 65% were white, respectively). Furthermore, the overall health and size of the plants inoculated with the exoY emmA double mutant or with the exoY emmC double mutant were decreased compared to plants inoculated with the exoY strain. These results indicate that disruption of the emm locus in an already compromised exoY strain renders the bacteria incapable of establishing an efficient symbiosis with alfalfa.
In order to determine the regulatory interactions among the three components of the emm locus, the role of each gene in the expression of the other genes was analyzed. First we studied the effect of a mutation in emmA on emmB and emmC (from which emmA is divergently transcribed). There was no difference in expression of emmB and emmC in an emmA mutant (Table (Table3),3), indicating that EmmA does not have a transcriptional regulatory effect on the emmB and emmC genes. On the other hand, a fourfold decrease in the expression of emmA was observed when either emmB or emmC was disrupted (Table (Table3),3), indicating that the EmmB and EmmC proteins may have a regulatory effect on emmA. The effects of mutations in emmB and emmC were analyzed to explore any potential cross-regulation of these two components. No significant difference was observed in either case, indicating that at least at the transcriptional level, there is minimal cross-regulation of EmmB and EmmC (Table (Table3).3). These results suggest that the putative sensor kinase EmmB and the response regulator EmmC have a small effect on the regulation of emmA and, in addition to the previously described common phenotypes of all emm mutants, support the notion that these three genes appear to act cooperatively at a posttranscriptional level to affect, directly or indirectly, common cellular pathways.
We identified a new locus, emm, which appears to be involved in important cellular processes in the alfalfa symbiont S. meliloti. Here, we show that the products of the three emm genes play a role in succinoglycan production, stress survival, and motility. Additionally, a disruption in any of the emm locus genes results in a significant reduction in the ability of S. meliloti to form an optimal symbiosis with its plant host.
S. meliloti is a member of the order Rhizobiales, a phylogenetically distinct group of gram-negative bacteria whose members also include the plant pathogen A. tumefaciens and the facultative intracellular animal pathogen Brucella sp. Comparative genome analysis of these species revealed considerable similarity in both genome structures and metabolic functions (20, 44). There are several protein homologs in these bacteria, and their roles in the respective species are often quite comparable. For example, the ExoS-ChvI two-component system regulates the production of succinoglycan in S. meliloti and is necessary for symbiosis with alfalfa, while its homologs ChvG-ChvI in A. tumefaciens and BvrR-BvrS in Brucella sp. are crucial for plant tumor formation and intracellular replication, respectively (10, 13, 55). The emm locus appears to play an important role in the S. meliloti-alfalfa symbiosis; however, this three-gene cluster is not present in A. tumefaciens or Brucella sp. Rather, it is highly conserved specifically in the nitrogen-fixing rhizobia, suggesting that these genes have a specialized role in these symbiotic organisms.
The first apparent phenotype of an emm mutant is its overproduction of succinoglycan. While clearly the synthesis of this polymer is important for plant invasion (45), its production has also been associated with protecting S. meliloti against stress (41, 59). Indeed, our observation that succinoglycan-deficient strains appear to be inherently defective in the ability to withstand membrane stress induced by detergent and osmotic pressure agrees with previous reports suggesting that symbiotically important exopolysaccharides contribute to the overall fitness of soil rhizobia (41, 59). Disruption of any of the emm genes resulted in an increase in succinoglycan production and a corresponding increase in the expression of exoY, the first gene in the biosynthetic pathway of this polymer. It seems reasonable to assume that the overproduction of succinoglycan increases the bacterium's ability to withstand stress. On the contrary, S. meliloti with mutations in the emm locus displayed dramatic sensitivity to both detergent stress and osmotic upshock compared to the wild type, and these effects were more pronounced when succinoglycan was not produced. While it is not clear if upregulation of succinoglycan production is directly controlled by any of the emm gene products or is a compensatory mechanism to overcome the apparent susceptibility to membrane stressors, the presence of an intact emm locus appears to be critical to S. meliloti's ability to withstand stress, and this role is independent of the presence of succinoglycan.
The detergent and high-salt assay results indicate that when there are mutations in the emm locus, the bacteria are deficient in the ability to tolerate conditions that place physiological stress on their membranes. Susceptibility to membrane stress is often associated with defects in lipopolysaccharide production (7, 8). However, no differences in lipopolysaccharide composition were detected between an emm mutant and the wild type (data not shown), suggesting that the membrane stress sensitivity observed in our work involves other mechanisms that have not been determined yet.
Another phenotype of strains with mutations in the emm locus was a dramatic motility defect due to downregulation of the expression of the motility genes. Since exopolysaccharide production and motility have been shown to be concurrently regulated in many bacterial systems (61, 62), we initially considered the possibility that the observed motility defect was due to overproduction of succinoglycan. However, exoY derivatives of emm mutants not capable of producing succinoglycan also demonstrated the same motility deficiency. Strains with mutations in the emm locus showed a decrease in the expression of visN, one of the main motility regulators (along with visR) in S. meliloti. Not surprisingly, expression of the downstream motility genes flbT, rem, and flgB was also decreased. Furthermore, very little flagellin protein production was detected in strains with mutations in the emm locus compared to the wild type. The effects that the emm genes have on motility appear to be at more than one level since constitutive expression of visN-visR failed to restore expression of the hierarchically downstream rem gene and did not restore motility. This suggests that emm may have multiple targets in the motility pathway. Therefore, EmmABC is a motility regulator whose action appears to be separate from its function in the regulation of succinoglycan production.
The role of the emm locus appears to be independent of succinoglycan production for all of the phenotypes observed with the exception of plant symbiosis. Although an emm mutation alone did not have a dramatic effect on the ability of S. meliloti to effectively associate with alfalfa, a marked decrease in plant invasion efficiency was observed when such a mutation was combined with exoY in a strain derivative that should be invasion proficient since it produces the other symbiotically important exopolysaccharide, EPS II. If, in addition to facilitating root nodule invasion, succinoglycan protects S. meliloti against changes encountered in the soil and plant environments, then its absence could be particularly detrimental if the cell were otherwise compromised. Indeed, the dramatic sensitivity to physiological stressors such as detergent and high salt of emm mutants were exacerbated when production of succinoglycan was removed. In light of these facts, it is possible that an emm mutation in a succinoglycan-deficient background renders the bacteria highly susceptible to the potential stresses endured during invasion and occupation of the root nodule and, therefore, prevents S. meliloti from forming an optimal symbiosis with alfalfa.
Based on sequence homology, emmB and emmC may encode a two-component system. The EmmA protein is predicted to be secreted and may act in conjunction with the EmmBC pair to sense a change(s) in the immediate environment. Clearly, mutations in these three components result in similar phenotypes, suggesting that the components work in concert. They seem to protect the cell from stress, and their disruption results in increased succinoglycan production, as well as motility defects. Protective mechanisms involving a periplasmic protein, a sensor kinase, and a response regulator have been described previously for S. meliloti. The ExoR-ExoS/ChvI system is a regulatory pathway involved in critical cell processes, including exopolysaccharide production, nutrient utilization, motility, and free-living viability (11). Since exoR, exoS, and chvI mutants display defects similar to those of the emm mutants (62), we were intrigued by the possibility that there is an interaction between this system and Emm. Analysis of the expression of exoR, exoS, and chvI in the emm mutants revealed no differences compared to the wild type (data not shown), suggesting that the phenotypes of emm mutants are not directed through ExoR or the ExoS/ChvI two-component system, at least at the level of gene regulation. Likewise, the expression of the emm genes does not depend on the presence of an intact exoR gene, which is hierarchically above the ExoS/ChvI two-component system (data not shown).
Although it does not appear that there is any cross-regulation of the ExoR-ExoS-ChvI system and the EmmABC system, several parallels are evident. For example, ExoR is a periplasmic protein and has been shown to locate to the periplasm, where it physically interacts with the membrane-bound kinase ExoS (11). Both ExoR and EmmA contain a predicted N-terminal signal sequence, and it is feasible that EmmA locates to the periplasm as well. exoR and exoS-chvI mutants have similar phenotypes, suggesting that ExoR and ExoS-ChvI utilize a common pathway in the regulation of important cell functions (11, 61). Likewise, mutations in any of the emm loci also result in related phenotypes, indicating the common functionality of these loci. The ExoS-ChvI system has recently been shown to positively control transcription of exoR (11). Similarly, a mutation in emmB or emmC resulted in downregulation of emmA. These results suggest that the EmmABC and ExoR-ExoS-ChvI systems may act in parallel but separate fashions as they each regulate important cell processes, such as the production of succinoglycan and motility in response to environmental cues.
Similar phenotypes have also been observed for strains with mutations in the emm locus and strains with mutations in the stationary phase-induced sensor kinase CbrA. A cbrA mutant shows cell envelope defects, overproduction of succinoglycan, a decrease in flagellar biosynthesis, and symbiotic defects (21, 22). It remains to be determined if these two loci act in parallel or are part of the same pathway. It should be noted that the symbiotic defects shown by the emm mutants are milder than those shown by the exoR or cbrA mutants. It is possible that these three systems play similar roles in the adaptation of S. meliloti to changes in the environment but respond to different signals. Identification of the signals should provide more insight into the exact role of these systems during symbiosis.
S. meliloti utilizes various means to thrive and compete in its soil and plant environments. Here we characterized a new locus, emmABC, which appears to play an important role in the overall fitness of S. meliloti. It is tempting to speculate that EmmB-EmmC, in conjunction with EmmA, perceives signals from the environment and modulates gene expression accordingly. Future studies will determine the comprehensive role of the EmmABC system in both the free-living and symbiotic stages of the S. meliloti lifecycle.
We thank the members of our laboratory for helpful discussions and critical reading of the manuscript. We are also grateful to Birgit Scharf for providing the anti-Fla antibody and to Anke Becker for providing the emmB and emmC mutants.
This work was supported by National Science Foundation grant MCB-9733532 and National Institutes of Health grant 1R01GM069925 to J.E.G.
Published ahead of print on 24 July 2009.
†Supplemental material for this article may be found at http://jb.asm.org/.