ABSTRACT
Many bacterial species contain multiple copies of the genes that encode the chaperone GroEL and its cochaperone, GroES, including all of the fully sequenced root-nodulating bacteria that interact symbiotically with legumes to generate fixed nitrogen. In particular, in Sinorhizobium meliloti there are four groESL operons and one groEL gene. To uncover functional redundancies of these genes during growth and symbiosis, we attempted to construct strains containing all combinations of groEL mutations. Although a double groEL1 groEL2 mutant cannot be constructed, we demonstrate that the quadruple groEL1 groESL3 groEL4 groESL5 and groEL2 groESL3 groEL4 groESL5 mutants are viable. Therefore, like E. coli and other species, S. meliloti requires only one groEL gene for viability, and either groEL1 or groEL2 will suffice. The groEL1 groESL5 double mutant is more severely affected for growth at both 30°C and 40°C than the single mutants, suggesting overlapping functions in stress response. During symbiosis the quadruple groEL2 groESL3 groEL4 groESL5 mutant acts like the wild type, but the quadruple groEL1 groESL3 groEL4 groESL5 mutant acts like the groEL1 single mutant, which cannot fully induce nod gene expression and forms ineffective nodules. Therefore, the only groEL gene required for symbiosis is groEL1. However, we show that the other groE genes are expressed in the nodule at lower levels, suggesting minor roles during symbiosis. Combining our data with other data, we conclude that groESL1 encodes the housekeeping GroEL/GroES chaperone and that groESL5 is specialized for stress response.
The groESL operon encodes the chaperone GroEL and its cochaperone, GroES, which function as a multimeric complex that binds protein substrates and enables them to fold properly. Many bacterial species have only one groESL operon, and in Escherichia coli the single copy is required for viability at temperatures as low as 17°C (9). Other bacterial species, however, have more than one groESL operon and additional groEL genes. The reason for maintaining multiple copies has not been fully determined for any species. One possibility is that the genes may be differentially regulated to provide GroEL/GroES at different times or at different levels. Such regulation has been observed in a number of species (5, 12, 16, 25, 26, 30, 41). A second possibility is that the genes may encode proteins with different substrate specificities. Although the substrates of the GroEL/GroES complexes in species with multiple groESL operons have not been determined, there is some evidence consistent with this hypothesis. For example, in Rhizobium leguminosarum the three GroEL proteins have different in vitro properties for folding one substrate (15), and in Sinorhizobium meliloti GroEL3 is not able to functionally replace GroEL1 (5). A third possibility is that the proteins may be specialized for functions that do not include protein folding. For example, in E. coli GroEL is able to bind folded σ32, decreasing σ32-dependent transcription (21).
We are particularly interested in the role of multiple groE genes in the root-nodulating bacteria of the Rhizobiales. These bacteria interact symbiotically with partner legume species by inducing the formation of nodules, colonizing the nodules, and then fixing nitrogen for the host plant. Multiple groE copies have been found in all of the fully sequenced genomes of root nodulators: Bradyrhizobium japonicum (24), Mesorhizobium loti (23), Rhizobium etli (18), R. leguminosarum (47), and S. meliloti (14). In two cases groE has been connected to symbiosis. In B. japonicum the groESL3 operon is regulated along with nitrogen fixation genes, and a groEL3 groEL4 double mutant is unable to fix nitrogen (Fix−) (13). In S. meliloti groEL1 is required for full induction of nodulation genes and nitrogen fixation (35).
S. meliloti has five groE loci in the genome: groESL1 and groEL4 are located on the chromosome, groESL2 and groESL3 are located on the pSyma megaplasmid, and groESL5 is located on the pSymb megaplasmid (14). Only the groESL1 locus has been identified in mutant screens. Originally groESL1 was discovered in a screen for reduced nod gene expression (35). The nod genes encode enzymes that produce Nod factor, which elicits nodule formation. The genes are controlled by several related transcription factors (NodD1, NodD2, and NodD3), some of which require plant inducers for activity. Biochemical studies have demonstrated that GroEL copurifies with NodD1 and NodD3, and GroEL/GroES modulates NodD activity (35, 46). However, the studies did not address which GroEL/GroES complexes are involved. The groESL1 locus was also identified in a screen for genes required for the production of N-acyl homoserine lactones used in quorum sensing (27). The deficiency in N-acyl homoserine lactone production may be due to a direct interaction of GroEL/GroES with the TraR regulator.
All of the single S. meliloti groEL mutants are viable (5, 30, 34, 35), but groEL1 and groEL2 cannot be disrupted at the same time (34). The effect of groE mutations on growth rate has been determined only for the groEL1 mutant, which has a longer doubling time at 30°C than the wild type (27, 35). The effect of groE mutations on symbiosis has been determined for all five groEL mutants, and only groESL1 is associated with symbiotic defects (5, 30, 34, 35). groEL1 mutants are delayed in nodulation and are unable to fix nitrogen (35). Interestingly, groEL2, but not groEL3, can substitute for groEL1 during symbiosis if expressed at high levels (5, 35).
Previous work has demonstrated that all of the S. meliloti groE genes are expressed during free-living growth in rich and minimal media, with groESL1 expressed at high levels and the others expressed at low levels (5, 30). Transcription of only groESL1 and groESL5 increases upon heat shock (30). Two regulatory systems that bacteria use for controlling genes in response to heat stress are the RpoH sigma factor (19, 20), which directs transcription from specific promoters, and the HrcA repressor (32, 40, 42), which binds to a cis-acting element called CIRCE (for “controlling inverted repeat of chaperone function”) (48). S. meliloti has two genes that are known to encode RpoH sigma factors (38, 39). RpoH2 does not control any of the groE genes, and RpoH1 controls only groESL5 (5, 30). The S. meliloti genome contains one gene that is predicted to encode HrcA (14), and putative CIRCE elements are located upstream of groESL1 and groESL2. However, the functionality of this CIRCE/HrcA system has not been demonstrated.
Our goal was to uncover functional redundancies of the groEL genes in S. meliloti during growth and symbiosis by attempting to construct strains containing all combinations of groE mutations. In this paper we demonstrate that S. meliloti cells require only one groEL gene for viability, and either groEL1 or groEL2 will suffice. However, only groEL1 is necessary and sufficient for symbiosis. Although the roles of groEL2, groEL3, and groEL4 are still unclear, we present evidence that groEL1 and groEL5 have overlapping functions and suggest that groESL5 is specialized for stress response.
MATERIALS AND METHODS
Strains and growth conditions.The bacterial strains used in this study were isogenic to the wild-type strain Rm1021 (28). The groEL mutations in strains JO138 (groEL1Δ::gus-aph) and JO60 (groEL2Δ::gus-aph) (34) were transduced into Rm1021 to remove the plasmid pPH1JI, which had been used for homogenotization, generating AB249 and AB247. AF14 (groESL3Δ::tet) (5), VO3193 (groEL4Δ) (5), and NI001 (groESL5Δ::aacC1) (30) have been previously described. Multiple groE mutants were constructed by generalized transduction using N3 phage. Transcriptional groEL-gfp-gus fusions were located in the chromosome of AB140 (groEL1::pAB11), AB129 (groEL2::pAB10), AB145 (groEL3::pAB12), AB147 (groEL4::pAB13), and AB150 (groEL5::pAB14) in a manner such that the groEL gene is not disrupted (5). The multicopy plasmid pO57.33 (nodD1 nodABC′-′lacZ) (35) was introduced into S. meliloti cells by triparental conjugation (17). Cells were grown in LB medium supplemented with MgSO4 and CaCl2 (LB/MC medium) (17) at 30°C, unless otherwise indicated.
Plant assays. Medicago sativa plants were grown on nitrogen-free medium at pH 6.0 and inoculated with S. meliloti cells as previously described (37). Melilotus alba and Melilotus officinalis plants were grown similarly, except the medium was at pH 6.5. Nodulation was determined at 3 weeks postinoculation, and nitrogen fixation was determined at 6 weeks postinoculation. Fix+ bacteria result in tall, green plants with pink nodules, whereas Fix− bacteria result in stunted, chlorotic plants with white nodules.
Western blot analysis.To obtain samples for Western blot analysis, cells were grown overnight at 30°C in LB/MC medium with streptomycin, diluted to an optical density at 595 nm (OD595) of 0.1, and grown to mid-log phase (0.6 ≤ OD595 ≤ 0.8). Cultures were then split and grown for an additional hour at 30°C or heat shocked for an hour at 42°C, after which cells were harvested and frozen at −80°C. Cells were resuspended in 1× phosphate-buffered saline at 0.1 ml per OD595 unit and disrupted by sonication. The resulting extracts were combined with 2× Laemmli sample buffer, and equal volumes were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose. Blots were probed with a 1:5,000 dilution of rabbit polyclonal antibodies to E. coli GroEL (Stressgen), followed by a 1:15,000 dilution of anti-rabbit horseradish peroxidase-conjugated secondary antibody, developed with enhanced chemiluminescence reagents (Pierce), and imaged with a Fujifilm LAS-3000 imaging system.
Assays of β-glucuronidase activity.To quantitate β-glucuronidase activity in free-living bacteria, cells were grown in LB/MC medium, harvested at the indicated times, and frozen at −80°C. The cells were then permeabilized with lysozyme (200 μg ml−1, 37°C for 10 min), and β-glucuronidase activity was assayed with p-nitrophenyl-β-d-glucuronide, as described previously (22). β-Glucuronidase activity is expressed in nanomoles per minute per OD595 unit times 1,000.
To visualize β-glucuronidase activity in planta, nodules were sectioned and stained as described previously (44). To quantitate the activity, we adapted a previously published protocol to isolate total bacteria from alfalfa nodules (11). For each assay, approximately 50 nodules were harvested from plants 3 weeks postinoculation and placed in an Eppendorf tube on ice containing 250 μl of MMS (11). The nodules were crushed with a pestle, the volume was increased to 1 ml, and large plant debris was removed by filtration through a 70-μm cell strainer. Additional plant debris was removed by spinning the sample at 100 × g for 5 min at 4°C and removing the supernatant. The nodule bacteria were then pelleted by spinning the sample at 2,200 × g for 5 min at 4°C and resuspended in 100 μl of cold MMS. A portion of each sample was used to determine the OD595 as a measurement of cell density, and the remaining material was frozen at −80°C. Depending on the level of activity of the fusion, 20 to 50 μl of cells were assayed as described above.
Assay of β-galactosidase activity.Cells containing pO57.33 (nodD1 nodABC′-′lacZ) (35) were grown overnight and then diluted to an OD595 of 0.05 in LB/MC medium containing streptomycin and spectinomycin. After 4 h of growth, each culture was split and a 1/1,000 volume of the solvent N,N-dimethyl formamide or 3 mM luteolin was added to induce NodD1 to activate nodC-lacZ expression. The cells were collected at 10 h and frozen at −80°C until being assayed for activity. The cells were permeabilized with lysozyme (200 μg ml−1, 30°C for 10 min), and β-galactosidase activity was assayed as described by Miller (29).
RESULTS AND DISCUSSION
groE requirements during free-living growth.Previous work has shown that all five groE operons can be disrupted (5, 30, 34, 35), but a groEL1 groEL2 double mutant is not viable (34). To uncover functional redundancies among other groEL genes besides groEL1 and groEL2, we attempted to construct all of the possible double, triple, and quadruple mutants. We confirmed that the groEL1 groEL2 double mutant cannot be constructed but were able to construct strains containing all other combinations of mutations. Since the two quadruple mutants are viable, one of either groEL1 or groEL2 is necessary and sufficient for growth. Therefore, S. meliloti is like all other bacterial species with multiple groEL genes tested so far in requiring only one of the groEL genes for growth under nonstress conditions (26, 36, 41, 43).
To determine whether the mutations affected growth under free-living conditions, we compared the growth of the single, double, triple, and quadruple groE mutants to the wild type in LB/MC medium at 30° and 40°C. At 30°C (Fig. 1A) we found that most of the groE mutants grew like the wild type. The exceptions were that among the single mutants, the groEL1 mutant displayed a slight but reproducible growth defect, as shown previously (27, 35), and among the multiple mutants, strains containing mutations in both groEL1 and groESL5 displayed a slightly more pronounced growth defect. In particular, these strains exhibited a longer lag phase and doubling time, although they reached the same maximum cell density. At 40°C (Fig. 1B) we found that all of the groE single mutants had a growth defect, reaching lower cell densities than the wild type, with the groEL1 mutant always being the most affected. The groEL1 groESL3 double mutant had a slightly larger growth defect than either single mutant. Strains containing mutations in both groEL1 and groESL5 exhibited a severe growth defect. All other double, triple, and quadruple mutants displayed growth phenotypes similar to the single mutants (data not shown). In summary, although most of the groE mutants exhibit only minor growth defects, if any, the double groEL1 groESL5 mutant is temperature sensitive for growth.
Comparison of the growth of single and multiple groE mutants with the wild type as measured with the OD595. Cells were grown in LB/MC medium with streptomycin at 30°C (A) and 40°C (B). Strains are Rm1021 (wild type; filled circles), AB249 (groEL1; open circles), AB247 (groEL2; open squares), AF14 (groESL3; open diamonds), VO3193 (groEL4; open triangles), NI001 (groESL5; open inverted triangles), AB221 (groEL1 groESL3; filled triangles), AB219 (groEL1 groESL5; filled diamonds), AB257 (groEL1 groESL3 groEL4 groESL5; plus signs), and AB238 (groEL2 groESL3 groEL4 groESL5; crosses). The experiment was repeated three times with essentially identical results, and the panels show data from one experiment.
In E. coli, groESL is required for growth at low temperatures (9), and the activity of GroEL/GroES in part determines the lower temperature limit at which the bacteria can grow (10). In addition, a mutation in another chaperone-encoding gene, dnaK, renders the cells both cold sensitive and temperature sensitive for growth (6). Therefore, to determine whether groE mutants in S. meliloti are cold sensitive, we grew the quadruple mutants at 20°, 15°, and 10°C. Neither quadruple mutant was cold sensitive (data not shown).
We subjected groE mutant strains to Western analysis using polyclonal antibodies to the E. coli protein (Fig. 2). At 30°C we obtained a single band for GroEL in wild-type cells. This band is predominantly due to GroEL1, as shown by the groEL2 groESL3 groEL4 groESL5 quadruple mutant, but also includes GroEL2, as shown by the groEL1 groESL3 groEL4 groESL5 quadruple mutant. After subjecting cells to heat shock at 42°C, the levels of GroEL1, but not GroEL2, increased, which is consistent with data on transcription (30). In wild-type cells, a second band of lower molecular weight appeared following heat shock. Production of the second band was dependent on the RpoH1 sigma factor (data not shown). Because groESL5 is the only groE locus controlled by RpoH1 (5, 30) and transcription of groESL5 increases upon heat shock (30), we hypothesized that the second band corresponded to the GroEL5 protein. Consistent with this hypothesis, the second band was not produced in cells containing the groESL5 deletion. Interestingly, the GroEL5 band was observed in the groEL1 mutant even at 30°C, indicating that GroEL5 production increases when GroEL1 is absent. The effect is specific to the groEL1 mutation because the band was not present at 30°C in the triple groEL2 groESL3 groEL4 mutant strain.
Western analysis of GroEL in groE mutant strains. Cells were grown to mid-log phase in LB/MC medium with streptomycin at 30°C. Cultures were split and grown for an additional hour at 30°C (−) or at the heat-shock (HS) temperature of 42°C (+) before being processed for Western analysis using a polyclonal antibody to E. coli GroEL. The strains were Rm1021 (wild type), NI001 (groESL5), AB249 (groEL1), AB219 (groEL1 groESL5), AB243 (groEL2 groESL3 groEL4), AB238 (groEL2 groESL3 groEL4 groESL5), and AB257 (groEL1 groESL3 groEL4 groESL5). The figure shows one representative blot out of three experiments.
Western analysis did not allow us to determine whether production of GroEL2, GroEL3, or GroEL4 was increased in the groEL1 mutant, because the signals from these proteins are most likely masked by the high levels of GroEL1. To resolve this issue and confirm our findings for GroEL5, we transduced groEL-gus transcriptional fusions (5) into the groEL1 mutant (Table 1). During the exponential and stationary phases of growth (6 and 24 h, respectively), expression of groEL2 and groEL5 was significantly increased in the groEL1 mutant compared to the wild type. Expression of groEL4 was significantly increased in the groEL1 mutant only at 24 h, suggesting that the effect on groEL4 expression depends on growth phase. Expression of groEL3 was not affected by the groEL1 mutation. Therefore, the loss of GroEL1 results in upregulation of all of the other groE genes except groESL3.
Effect of groEL1 mutation on groEL-gus gene expressiona
groE requirements during symbiosis.To determine which combinations of groEL genes are important for symbiosis, we tested the effects of groE mutations on the expression of nod genes and the formation of effective nodules. groEL1 was previously shown to be required for full induction of nod gene expression in response to the plant inducer luteolin (35). To determine whether other groE genes play a role in nod gene expression, we introduced a multicopy plasmid containing nodD1 and nodC-lacZ into wild-type cells and the single and quadruple mutants. As shown in Table 2, the groEL1 mutant and the groEL1 groESL3 groEL4 groESL5 quadruple mutant displayed lower expression of nodC-lacZ than the wild type both in the absence and the presence of the plant inducer luteolin. In addition, the amount of induction caused by the addition of luteolin was reduced. In contrast, single groEL2, groESL3, groEL4, and groESL5 mutants, as well as the groEL2 groESL3 groEL4 groESL5 quadruple mutant, displayed full nodC-lacZ expression. Therefore, only groEL1 is necessary and sufficient for full induction of the nod genes.
Effects of groE mutations on nodC-lacZ gene expression
In terms of the formation of effective nodules, among the single groE mutants only groEL1 mutants have a symbiotic defect, resulting in Fix− nodules (5, 30, 34, 35). To uncover redundant functions, we inoculated alfalfa (Medicago sativa) plants with the triple and quadruple mutants and observed the plants for nodule formation and nitrogen fixation. Any mutant that contained the groEL1 mutation formed Fix− nodules. All other mutants were similar to the wild type in the ability to nodulate and fix nitrogen. To test for host-specific effects in other genera, we inoculated white sweet clover (Melilotus alba) and yellow sweet clover (Melilotus officinalis) plants with the groEL1 single mutant and the quadruple mutants. As with alfalfa, the groEL1 mutant and groEL1 groESL3 groEL4 groESL5 quadruple mutant elicited Fix− nodules, whereas the groEL2 groESL3 groEL4 groESL5 quadruple mutant elicited Fix+ nodules. Therefore, the only groEL gene required for symbiosis is groEL1.
Expression of each of the groE loci within nodules on a variety of hosts has been detected in one or more of the global transcript or protein analyses (1, 3, 4, 7, 8, 33). To directly compare levels of gene expression within alfalfa nodules, we inoculated plants with bacteria containing a matched set of groEL-gus transcriptional fusions (5). As shown in Fig. 3, all five groEL genes are expressed within the nodule, although at different levels. To quantitate expression, we harvested bacteria from nodules and determined β-glucuronidase activity. Similar to results obtained under free-living conditions (5), groEL1-gus was expressed at high levels; groEL2-gus, groEL4-gus, and groEL5-gus were expressed at low levels; and groEL3-gus was expressed at very low levels. Therefore, although only groEL1 is required for effective nodules, all of the groE genes are expressed during symbiosis to some degree.
Expression of groE genes within the nodule. Plants were inoculated with bacteria containing the following groEL-gus reporter fusions: groEL1 (A), groEL2 (B), groEL3 (C), groEL4 (D), and groEL5 (E). Nodules were harvested 3 weeks postinfection, hand sectioned, and stained with 5-bromo-4-chloro-3-indolyl-β-d-glucuronic acid for β-glucuronidase activity. Ten nodules were examined per strain in each of three experiments, and the panels show one representative nodule. The tip of the nodule is on the left, and the root-proximal portion is on the right. The average level of β-glucuronidase activity (nmol per min per OD595 unit × 1,000) is indicated under each image. The line in panel A represents 1 mm. Strains, from left to right, are AB140, AB129, AB145, AB147, and AB150.
Conclusions.We conclude that groESL1 encodes the major housekeeping GroEL/GroES chaperone in S. meliloti for the following reasons. First, groESL1 is located on the chromosome near many of the same neighboring genes as the single groESL operon in the closely related bacterium Agrobacterium tumefaciens (14, 45). Second, groESL1 is expressed at much higher levels than the other groE genes during growth in culture and during symbiosis (reference 5 and this study). Third, groEL1 is sufficient for both growth in culture and successful symbiosis (this study). In addition to the housekeeping role, groESL1 is also partially controlled by heat shock (30), indicating a role in stress response. This control is independent of RpoH1 and RpoH2 (5, 30) but may depend on a CIRCE/HrcA regulatory system (32), since a putative CIRCE element is located upstream of groESL1.
groESL5 is probably specialized for stress response, since gene expression (30) and protein production (this study) are induced by heat shock and groESL5 is the only groE locus controlled by RpoH1 (5, 30). In addition, we have shown that groESL5 is upregulated in the absence of groEL1. We postulate that the loss of the major housekeeping chaperone results in unfolded proteins that trigger groESL5 expression. The partially overlapping function of groEL1 and groEL5 can be observed by the synergistic effect of the two mutations on growth at both 30° and 40°C.
The roles of the other groE genes are still unclear. groESL1 and groESL2 encode very similar proteins (two amino acid differences for GroES and one amino acid difference for GroEL). Either groEL1 or groEL2 is sufficient during growth (this study), and groEL2 can substitute for groEL1 during symbiosis if present on a multicopy plasmid (35). This suggests that the proteins are interchangeable but that groESL2 is normally not expressed at levels high enough to be sufficient for symbiosis. As with groESL1, groESL2 is preceded by a putative CIRCE element, although heat shock control has not been observed (30). We have shown that the groEL2 gene is upregulated in the absence of groEL1. Given that transcriptional repression by the CIRCE/HrcA system in other bacteria depends on the levels of GroEL (2, 31), the increase in groEL2 transcription could be mediated through its putative CIRCE element. Why does groESL2 exist? Outside of the open reading frames and CIRCE elements, the groESL1 and groESL2 DNA sequences are quite different, which would be consistent with differential regulation. We speculate that groESL2 is expressed at high levels under some unknown condition when groESL1 is not expressed well or in addition to groESL1 when larger amounts of GroEL/GroES are needed.
The roles of groESL3 and groEL4 remain unknown. Presumably the genes produce GroES and GroEL under different conditions, encode chaperones that fold different ranges of substrates, and/or encode proteins specialized for nonfolding functions. Previously we have shown that groESL3, which encodes the most divergent of the GroEL/GroES homologs, is unable to functionally replace groEL1 (5). In addition, groESL3 is the only groE locus that is not upregulated in response to the loss of groEL1. These results would be consistent with different substrate specificities or different functions. Experiments to compare the substrate profiles of the different GroEL proteins may be illuminating.
ACKNOWLEDGMENTS
This work was supported by award 2001-35319-10902 from the NRI Competitive Grants Program/USDA to V.O. and by an Andrew Mellon Predoctoral Fellowship from the University of Pittsburgh to A.N.B.
FOOTNOTES
- Received 3 October 2006.
- Accepted 30 November 2006.
- Copyright © 2007 American Society for Microbiology