Exopolysaccharide II Production Is Regulated by Salt in the ...

14 downloads 0 Views 202KB Size Report
Jul 3, 1997 - We generated mutants that are nonmucoid in either stan- .... Becker, A., S. Rüberg, H. Küster, A. A. Roxlau, M. Keller, T. Ivashina, H. Cheng ...
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Mar. 1998, p. 1024–1028 0099-2240/98/$04.0010 Copyright © 1998, American Society for Microbiology

Vol. 64, No. 3

Exopolysaccharide II Production Is Regulated by Salt in the Halotolerant Strain Rhizobium meliloti EFB1 JAVIER LLORET,1 BRANDE B. H. WULFF,1 JOSE M. RUBIO,1 J. ALLAN DOWNIE,2 ILDEFONSO BONILLA,1 AND RAFAEL RIVILLA1,2* Departamento de Biologı´a, Universidad Auto ´noma de Madrid, 28049 Madrid, Spain,1 and Department of Genetics, John Innes Centre, Norwich Research Park, Norwich, NR4 7UH, United Kingdom2 Received 3 July 1997/Accepted 9 December 1997

The halotolerant strain Rhizobium meliloti EFB1 modifies the production of extracellular polysaccharides in response to salt. EFB1 colonies grown in the presence of 0.3 M NaCl show a decrease in mucoidy, and in saltsupplemented liquid medium this organism produces 40% less exopolysaccharides. We isolated transposoninduced mutant that, when grown in the absence of salt, had a colony morphology (nonmucoid) similar to the colony morphology of the wild type grown in the presence of salt. Calcofluor fluorescence, proton nuclear magnetic resonance spectroscopy, and genetic analysis of the mutant indicated that galactoglucan, which is not produced under normal conditions by other R. meliloti strains, is produced by strain EFB1 and that production of this compound decreases when the organism is grown in the presence of salt. The mutant was found to be affected in a genetic region highly homologous to genes for galactoglucan production in R. meliloti Rm2011 (expE genes). However, sequence divergence occurs in a putative expE promoter region. A transcriptional fusion of the promoter with lacZ demonstrated that, unlike R. meliloti Rm2011, galactoglucan is produced constitutively by EFB1 and that its expression is reduced 10-fold during exponential growth in the presence of salt. Bacterial polysaccharides are necessary for a functional Rhizobium-legume symbiosis. Exopolysaccharide (EPS), lipopolysaccharide (LPS), capsular polysaccharide, and cyclic b-(132)glucan play essential roles in the formation of the infection thread and in nodule development, although the precise functions of these molecules are still being investigated (19). Several reports have shown that these molecules are also important for the adaptation and survival of Rhizobium strains under different environmental conditions at both the free-living and symbiotic stages. Cyclic b-(132)-glucan is necessary for hypoosmotic adaptation (10), and alterations in LPS have been observed to occur in response to environmental changes, such as pO2, low-pH, ionic, and osmotic stresses, and in response to the presence of plant inducers of nod genes (18, 23, 25, 27, 34, 36). One of these inducers, genistein, has been shown to alter the composition of the EPS in Rhizobium fredii (9). Rhizobium meliloti SU47 can produce two EPSs, a succinoglycan (EPS I) and a galactoglucan (EPS II) (22). Most physiological and genetics studies have been performed with derivatives of this strain, which under normal growth conditions produces only EPS I (14, 38). Galactoglucan is produced under phosphate limitation conditions (39) or in genetic backgrounds in which a mutation in either of two regulatory loci, mucR (20, 38) and expR (14), has occurred. On TY medium colonies have a compact (dry) morphology when only EPS I is produced and a mucoid morphology when both EPSs are produced or only EPS II is produced. R. meliloti EFB1 differs from SU47 in that EFB1 is very mucoid under these growth conditions, suggesting that EPS II is produced. EPSs are essential for the establishment of a functional symbiosis between R. meliloti and its host plant, alfalfa. Mutants of SU47 that do not produce EPS I form empty nonfunc-

tional nodules (21), although functional nodule formation can be restored by the production of EPS II (14, 38). It has been shown recently that a low-molecular-weight fraction of EPS II is responsible for this effect and that this fraction acts at picogram levels, indicating that the EPS may function as a signal for nodule invasion (15). The genetic regions implicated in the production of the two EPSs are not linked on the second megaplasmid of R. meliloti SU47 (6). The genes for EPS I production are designated exo genes, and most of them are located in a 24-kb cluster (22). The genes for EPS II production, the exp gene cluster, are organized in five complementation groups which comprise 22 genes. These genes have been sequenced recently (2). Previously, we showed that different changes in the LPS of the halotolerant organism R. meliloti EFB1 are induced by osmotic pressure and by salinity (23). Furthermore, a different colony morphology was observed in salt-supplemented medium, and the difference was not due to osmotic stress. Here, we show that EPS production is also affected by salt in strain EFB1 and that the difference in the production of EPS II in the presence and absence of salt is regulated at the transcriptional level. MATERIALS AND METHODS Strains, plasmids, and media. The bacterial strains and plasmids used in this study are listed in Table 1. R. meliloti was grown at 28°C in TY (3), YM (35), GYM (10), or M9 (29) medium containing 1% mannitol as a carbon source and 5 mM potassium glutamate as a nitrogen source. When appropriate, media were supplemented with calcofluor (0.02%) and/or 0.3 M NaCl. Escherichia coli strains were grown at 37°C in Luria-Bertani medium. Antibiotics were added at the following concentrations when required: streptomycin, 200 mg/ml; tetracycline, 10 mg/ml; kanamycin, 30 mg/ml; and ampicillin, 200 mg/ml. DNA manipulations. DNA preparation, restriction endonuclease digestion, cloning, visualization, and E. coli transformation were carried out by using previously described protocols (29). Southern blotting and hybridization were performed with a nonradioactive detection kit, and a chemiluminescence method was used to detect hybridization bands according to the instructions of the manufacturer (Boehringer Mannheim Biochemicals). An EFB1 gene bank was constructed by cloning genomic DNA partially digested with EcoRI into cosmid pLAFR3. The size of genomic DNA was determined by centrifugation in a NaCl

* Corresponding author. Mailing address: Departamento de Biologı´a, Facultad de Ciencias, Universidad Auto ´noma de Madrid, E-28049 Madrid, Spain. Phone: 34 1 397 81 77. Fax: 34 1 397 83 44. E-mail: [email protected]. 1024

SALT REGULATION OF EPS II IN R. MELILOTI

VOL. 64, 1998

1025

TABLE 1. Bacterial strains and plasmids Strain or plasmid

Relevant characteristics

Source or reference

R. meliloti strains Rm2011 EFB1 EFB107 EFB1011

Smr derivative of SU47 Wild type, Smr EFB1 expE2::Tn5-lacZ EFB1, Km cassette of Tn5 inserted into the HindIII site of expE2

5 23 This study This study

E. coli strains S17-1 DH5a HB101 JM109

pro recA hsdR, RP4-2 Tc::Mu Km::Tn7 integrated in the chromosome endA1 recA1 hsdR17 supE44 thi-1 gyrA96 relA1 D(lacZYA-argF)U169 f80d lacZDM15 thi-1 hsdS20 recA13 ara-14 leuB6 proA2 lacY1 galK2 rpsL20 supE44 recA endA1 gyrA96 thi hsdR17 supE44 D(lac-proAB) relA1

30 B. R. L. Inc. Lab collection Lab collection

Plasmids Bluescript SK1

Ampr, phagemid, M13 derivative, f1 origin of replication

Stratagene, La Jolla, Calif. 33 31 12 32 16 This study

pLAFR3 pSUP102-Tn5-lacZ pRK2013 pMP220 pPH1JI pBG1000 pBG1010 pBG1011 pBS1004 pBS1009 pBS1027 pRO1052 pRO1053

Tetr, host range P-group cloning vector, mobilizable RK2 cosmids Kmr Tcr Cms::Tn5-lacZ Kmr ColE1 replicon with RK2 tra genes Tcr mob1 lacZ IncP Gnr Spr tra1 IncP pLAFR3 cosmid with chromosomal EcoRI fragments of R. meliloti EFB1 complementing EFB107 and EFB1011 pLAFR3 cosmid with 16-kb EcoRI-BamHI fragment complementing EFB1011 pLAFR3 cosmid with EcoRI fragment from pBS1009 Bluescript SK1 with 1.627-kb EcoRI fragment of the exp gene cluster pBS1004 derivative, expE2-Knr, insertion site: HindIII Bluescript SK1 with 2,140-bp XhoI-EcoRI fragment of the exp gene cluster pMP220 with 346-bp SphI fragment from pBS1027 in sense orientation pMP220 with 346-bp SphI fragment from pBS1027 in antisense orientation

gradient, and 25- to 30-kb fragments were ligated with EcoRI-digested pLAFR3. The ligation preparation was packaged in vitro into lambda and was used to transfect E. coli cells. Genetic techniques. Tn5::lacZ mutagenesis was performed by using E. coli S17-1, as described previously (31). The location of insertions was determined by restriction analysis and Southern blotting. Derivatives of pLAFR3 were transferred by triparental mating by using helper plasmid pRK2013, as described previously (7). Bacteria in which marker exchange had occurred were selected by introducing the incompatible plasmid pPH1JI, as described previously (28). All recombinants were checked by Southern hybridization. DNA sequencing and sequence analysis. The 2.1-kb XhoI-EcoRI and 1.6-kb EcoRI fragments from plasmid pBG1000 were subcloned, and both strands were sequenced by the chain termination method by using the DyeDeoxy terminator cycle sequencing kit protocol and an Applied Biosystems automatic sequencer. A homology search and a sequence analysis were performed by using software from the Genetics Computer Group (Madison, Wis.). Nodulation test. Seeds of alfalfa (Medicago sativa cv. Moapa) were surface sterilized with bleach, germinated in the dark, and placed in Leonard jars; perlite was used as the substrate, and FP medium was used as the nutrient solution (11). Germinated seeds were inoculated with ca. 106 bacteria/seed. The Leonard jars were incubated by using a photoperiod consisting of 16 h of light and 8 h of dark at 23 and 16°C, respectively. Nodulation scores were determined 4 weeks after inoculation. Nitrogen fixation was determined by an acetylene reduction assay by using standard protocols. EPS isolation. R. meliloti EFB1 and EFB1011 were grown in TY medium and TY medium containing 0.3 M NaCl for 4 days. The cultures were centrifuged at 12,000 3 g, and the supernatants were decanted and concentrated 10-fold with a rotatory evaporator at 45°C. The concentrated supernatants were dialyzed, and EPS was precipitated with 3 volumes of ethanol. Each pellet was resuspended in H2O, dialyzed against twice-distilled water, and lyophilized. EPS was quantified by the phenol-sulfuric acid method (8). For proton nuclear magnetic resonance (NMR) spectroscopy, 10-mg portions of the lyophilized EPSs were exchanged several times with D2O and sonicated before they were finally dissolved in 99.96% D2O. Proton NMR spectroscopy was performed with a 500-MHz instrument operated by the Servicio Interdepartamental de Investigacio ´n at the Universidad Autonoma de Madrid. The probe temperature was 75°C. b-Galactosidase assay. A 346-bp SphI fragment of pBS1027 containing a putative promoter region between expD2 and expE1 of R. meliloti EFB1 was cloned in both orientations into the promoterless lacZ vector pMP220 (32). EFB1 carrying the resulting plasmids was grown in TY medium and TY medium containing 0.3 M NaCl, and b-galactosidase activity was assayed after 16, 20, 24, and 48 h of growth. b-Galactosidase units were calculated as described by Miller (24).

This This This This This This This

study study study study study study study

Nucleotide sequence accession number. The GenBank accession number for the R. meliloti EFB1 exp region which we sequenced is Y08703.

RESULTS Effect of salt on the colony morphology of R. meliloti EFB1. R. meliloti EFB1 has a different colony morphology when it is grown on salt-containing media (23). As shown in Fig. 1, a decrease in colony mucoidy was observed when EFB1 was grown on TY medium supplemented with 0.3 M NaCl, indicating that salt greatly reduced EPS production. This effect was

FIG. 1. Mucoidy and fluorescence on calcofluor-containing medium of R. meliloti EFB1 (section 1), R. meliloti EFB1011 (section 2), and R. meliloti EFB1011(pBG1000) (section 3). (A) TY medium. (B) TY medium containing 300 mM NaCl.

1026

LLORET ET AL.

FIG. 2. Physical genetic map of the exp region of R. meliloti EFB1. (A) EcoRI restriction map of cosmid pBG1000. Other relevant restriction sites are also indicated. (B) Restriction map of the sequenced 3.7-kb XhoI-EcoRI fragment. Only relevant sites are indicated. (C) Genetic map of the 3.7-kb XhoI-EcoRI fragment. The positions of Tn5::lacZ (solid triangle) and kanamycin resistance cassette (open triangle) insertions are indicated. E, EcoRI; B, BamHI; X, XhoI; H, HindIII; S, SphI.

not observed when TY medium was supplemented with a nonionic osmoticum, such as polyethylene glycol. Salt-induced decreases in mucoidy were independent of the growth medium used, as similar results were obtained when we used other media, such as YM agar, GYM, and M9 media (data not shown). When the production of EPS I was assessed by calcofluor fluorescence, EFB1 produced fluorescence in the presence and absence of salt (Fig. 1). Isolation and genetic analysis of a nonmucoid mutant. A Tn5::lacZ mutagenized population of R. meliloti EFB1 was plated onto TY medium and TY medium containing NaCl, and we identified one mutant (strain EFB107) with nonmucoid morphology in both media among the 5,000 mutants screened. EFB107 was complemented with plasmids from an EFB1 cosmid library constructed in pLAFR3. Two colonies with the mucoid phenotype restored were isolated. Both colonies contained the same cosmid, pBG1000 (Fig. 2A), which was used as a probe and hybridized with EcoRI-digested genomic DNA from EFB1 and EFB107. Hybridization showed that the Tn5:: lacZ insertion was in a 1.6-kb EcoRI band. To demonstrate that the mutation was caused by the Tn5:: lacZ insertion, the 1.6-kb EcoRI band from pBG1000 was cloned in Bluescript SK1 to create pBS1004. A kanamycin resistance cassette from Tn5 was introduced into pBS1004 after partial digestion with HindIII and ligation to create pBS1009. The resulting 5.1-kb EcoRI fragment was subcloned into pLAFR3 to form pBG1011, which was transferred into EFB1. The kanamycin cassette was recombined by using marker exchange, and the resulting strain, designated EFB1011, was nonmucoid (Fig. 1). When pBG1000 was mobilized into EFB1011, mucoidy was restored (Fig. 1). A 16-kb EcoRI-BamHI fragment from pBG1000 subcloned into pLAFR3, pBG1010 (Fig. 2A), was able to complement EFB1011 but not EFB107, indicating that the insertion in EFB107 had a polar effect on downstream genes but the insertion in EFB1011 did not. As pBG1000 can complement both mutants, the gene mutated in these strains is essential for mucoidy production. No other differences between EFB107 and EFB1011 were observed. The mutants were identical to the wild type, EFB1, with regard to growth, nodulation, and nitrogen fixation and produced the same LPS electrophoretic profile as the parental strain (data not shown), indicating that loss of mucoidy was not due to LPS alterations. All of the strains exhibited fluorescence on calcofluor-containing medium in the presence or absence of salt (Fig. 1), indicating that the genetic defect did not affect EPS I production. Sequence analysis. Two adjacent fragments from pBG1000 (pBS1004 and pBS1027) which contained the complete open

APPL. ENVIRON. MICROBIOL.

reading frame (ORF) where insertions 107 and 1011 map were sequenced (Fig. 2B). The sequence, comprising 3,767 bp, contained four ORFs, two of which were truncated. A homology search showed that the sequenced region corresponded to the recently reported genes for galactoglucan (EPS II) synthesis in R. meliloti Rm2011 (85% identity). The deduced products from the sequence were 92.7, 97.7, 84.9, and 91.4% identical to Rm2011 ExpD2, ExpE1, ExpE2, and ExpE3, respectively (Fig. 2C). Both insertions mapped on expE2. The intergenic region between expE1 and expE2 was also very similar (84% identity) in the two strains and contained a putative incomplete Rhizobium-specific intergenic mosaic element 2 (RIME2) sequence (26). However, greater divergence was found in the expD2-expE1 intergenic region; Becker et al. (2) have suggested that the expE promoter is in this region. This region includes an AT-rich segment (62% AT) in both strains that may contain the expE promoter. The level of identity for this AT-rich segment in the two strains is only 68.4% (Fig. 3). This difference might account for the difference in expression of galactoglucan production genes and therefore the difference in production of EPS II in strains Rm2011 and EFB1. EPS production in R. meliloti EFB1 and EFB1011. To confirm that the reduction in mucoidy observed on solid medium was due to a decrease in EPS II production, total EPS was isolated from EFB1 and EFB1011. Table 2 shows the amounts of EPSs secreted in TY medium and TY medium supplemented with 0.3 M NaCl. When EFB1 was grown in a saltcontaining medium, there was a 40% reduction in the quantity of secreted EPSs, which resulted in a loss of mucoidy in the colonies formed by this strain. Mutant EFB1011, which has expE2 disrupted and does not produce EPS II, secreted 58% less EPS than the wild-type strain, indicating that EFB1 produces more galactoglucan than EPS I in TY medium. Galactoglucan production in EFB1 was confirmed by performing a proton NMR analysis (data not shown). Unexpectedly, the spectrum of the EPS secreted by EFB1011 did not have any of the peaks corresponding to the acidic substituents of succinoglycan, although the isolated EPS was still fluorescent on calcofluor-containing medium. Therefore, we could not confirm by NMR analysis that EPS II production was reduced. However, the differences in morphology, the similar fluorescence characteristics on TY medium and TY medium containing NaCl, and the greater reduction in EPS secretion in EFB1 than

FIG. 3. Alignment of the region containing the putative promoter of expE from R. meliloti EFB1 (uppercase letters) with the corresponding region of R. meliloti Rm2011 (lowercase letters). The EFB1 sequence corresponds to nucleotides 290 to 462 from GenBank nucleotide sequence Y08703, and the Rm2011 sequence corresponds to nucleotides 13521 to 13338 from GenBank nucleotide sequence Z79692 (reverse). Both sequences start at the first nucleotide after the predicted stop codon for expD2.

SALT REGULATION OF EPS II IN R. MELILOTI

VOL. 64, 1998

1027

TABLE 2. EPSs secreted by R. meliloti EFB1 and EFB1011 in the presence and absence of salt and b-galactosidase activity expressed by an expE-lacZ transcriptional fusion under the same conditions Medium

TY medium TY medium containing NaClc a b c

Time course expression of expE::lacZ (Miller units)b

Amt of EPS secreted (mg/liter)a EFB1

EFB1011

16 h

20 h

24 h

48 h

368 6 11 213 6 25

155 6 4 114 6 3

8,340 6 87 683 6 84

8,688 6 422 3,403 6 356

8,668 6 277 6,576 6 403

8,940 6 1,047 8,100 6 782

The results are expressed in terms of glucose equivalents. The b-galactosidase activity of the promoter in the antisense orientation was less than 200 Miller units in all cases. TY medium was supplemented with 0.3 M NaCl.

in EFB1011 strongly indicate that the galactoglucan produced by EFB1 is the EPS most affected by a high salt concentration. Expression of galactoglucan production genes in R. meliloti EFB1. A 346-bp SphI fragment containing a putative expE promoter (Fig. 2B) was subcloned into pMP220 in order to create a transcriptional fusion with lacZ. The resulting plasmid, pRO1052, was introduced by conjugation into EFB1, and b-galactosidase activity was measured on TY medium and TY medium containing 0.3 M NaCl. As shown in Table 2, the putative promoter region induced lacZ expression constitutively and at a high rate all along the growth curve of EFB1. However, in the presence of salt the expression pattern was different; expression was more than 10 times lower after 16 h of growth, which corresponds to the midpoint of the exponential phase (23), but expression increased linearly with growth, reaching levels similar to those obtained in the absence of salt at high cell densities. DISCUSSION Modification of extracellular polysaccharides has been described frequently as a response to different environmental and physiological conditions. Most of the reports refer to modifications of the LPS electrophoretic pattern (23, 27, 36) or antigenicity (18, 23, 25, 34). We have previously shown that the LPS of halotolerant strain EFB1 produces a different electrophoretic profile and that a new LPS epitope appears when the strain is grown in salt-supplemented media (23). Here we show that salt also affects EPS production. Unlike the colonies of other R. meliloti strains, which have a nonmucoid (compact) aspect when the organisms are grown under normal laboratory conditions (14, 38), EFB1 colonies are very mucoid, and this mucoidy is severely reduced when the strain is grown on salt-containing media. Compared to SU47 derivatives, R. meliloti EFB1 has a different pattern of EPS production; under normal laboratory growth conditions, EPS I and EPS II are produced simultaneously, and EPS II is the major form of EPS secreted. NMR analysis of the EPSs secreted by EFB1 confirmed that two EPSs are simultaneously produced in this strain. One of these EPSs appears to be very similar to the galactoglucan produced under some circumstances by SU47, but the other produces a spectrum that is different from the spectrum reported for succinoglycan (21). The results presented here indicate that mucoidy in EFB1 on TY medium is due to galactoglucan production. Reductions in mucoidy of EFB1 occurred in all of the growth media tested and were induced by salt. Stimulation of succinoglycan production in R. meliloti SU47 has been reported when the organism is grown in the presence of a moderate NaCl concentration (0.2 M); however, higher concentrations reduce the amount of secreted polysaccharides (4). We generated mutants that are nonmucoid in either standard or salt-supplemented media. These mutants had a colony

morphology similar to that of R. meliloti SU47 derivatives. Loss of mucoidy did not impair symbiotic performance under the conditions tested, indicating that mucoidy does not affect the performance of a strain under laboratory conditions, as expected since EPS II is not necessary for symbiosis when EPS I is produced (14). However, it is possible that in nature mucoidy is an important trait. It should be noted that under stress conditions, such as phosphate limitation, R. meliloti Rm1021 shows a mucoid phenotype (39), and low salt concentrations can be considered stress conditions for R. meliloti EFB1 since this strain was isolated from nodules of a legume growing in a salt marsh, where the salinity level was ca. 0.3 M (23). The relevance of mucoidy for survival and/or symbiotic performance has yet to be investigated in this strain and other strains. Nonmucoid mutants EFB107 and EFB1011 were produced by gene disruption in a genomic region implicated in the production of galactoglucan. Recently, a similar region has been sequenced in R. meliloti Rm2011 (2), and this region comprises 32 kb where 25 ORFs are located. These ORFs are organized in five operons, designated expACGDE. The sequence that we obtained corresponds to the end of the expD complementation unit and the beginning of the expE complementation unit. Both the sequence and the genetic organization are very similar in Rm2011 and EFB1. Insertions causing the mutations map within expE2, which encodes a glucosyl transferase necessary for EPS II production in both strains (2). The homology between the two strains is maximal in the coding regions and in the intergenic region between expE1 and expE2, where a RIME2 sequence appears, as determined by a homology search of databases. RIME sequences have been described as long palindromic regions which might be involved in overall genome regulation (26). Two types have been found in R. meliloti, RIME1 sequences present in all members of the family Rhizobiaceae and RIME2 sequences found exclusively in R. meliloti. RIME regions are thought to be homologs of E. coli bacterial interspersed mosaic element (BIME) and enterobacterial repetitive intergenic consensus (ERIC) sequences (13, 17). Despite the overall homology in the EPS II production genes, clear divergences occur in the intergenic region between expD2 and expE1. This region, situated between two complementation groups, is likely to contain a promoter for the expE operon (2). We found that within the intergenic region there is an AT-rich stretch. The level of identity in this region between Rm2011 and EFB1 is significantly lower than the levels of identity in the coding region and the intergenic region within the operon. A transcriptional fusion with lacZ demonstrated that this region contains a promoter. Furthermore, expression of this fusion showed that galactoglucan production in EFB1 may be regulated at the transcriptional level. Expression is cryptic in SU47 derivatives (14) but constitutive in EFB1. In halotolerant strain EFB1 expression of expE genes is repressed

1028

LLORET ET AL.

by salt, at least until high cell densities are reached. The low level of expression of galactoglucan production genes under salt conditions correlates with the decrease in mucoidy and EPS secretion observed when EFB1 is grown in NaCl-containing media, although some other mechanisms may be present. In mutants of R. meliloti YE-2 which produce EPS II when they are grown in the presence of a low salt concentration succinoglycan is gradually replaced by galactoglucan as the salt concentration is increased (37). Despite the differences in the promoter region between strains SU47 and EFB1, regions of high homology are conserved within the AT-rich stretch. This might reflect common regulatory elements interacting with this promoter. One of them might be MucS, a positive transcriptional activator encoded by a gene in the exp cluster (1). The results that we present here clearly show that strains of the same Rhizobium species isolated from different environments may have different cell envelope compositions and may differ in regulation of the production of the cell envelope components. As no evidence concerning the importance of such differences for adaptation, survival, and symbiotic performance is available yet, ecological studies to investigate the role of EPSs in adaptation to different environments are necessary and may be relevant for deliberate releases of microorganisms in soil. ACKNOWLEDGMENTS We thank Susana Escobedo-Mariconda for technical support and invaluable encouragement and Ana Arraztio for help with computing and discussions. This work was supported by DGYCYT grant PB95-0217-C02-01. B.B.H.W. was supported by Erasmus and Studieuddannelsesstøtte fellowships. R.R. thanks the Federation of European Microbiological Societies for a short-term fellowship. REFERENCES 1. Astete, S. G., and J. A. Leigh. 1996. mucS, a gene involved in activation of galactoglucan (EPS II) synthesis gene expression in Rhizobium meliloti. Mol. Plant Microbe Interact. 9:395–400. 2. Becker, A., S. Ru ¨berg, H. Ku ¨ster, A. A. Roxlau, M. Keller, T. Ivashina, H. Cheng, G. C. Walker, and A. Pu ¨hler. 1997. A 32-kilobase exp gene cluster of Rhizobium meliloti directing the biosynthesis of galactoglucan: genetic organization and properties of the encoded gene products. J. Bacteriol. 179: 1375–1384. 3. Beringer, J. E. 1974. R factor transfer in Rhizobium leguminosarum. J. Gen. Microbiol. 84:188–198. 4. Breedveld, M. W., L. P. T. M. Zevenhuizen, and A. J. B. Zehnder. 1990. Osmotically induced oligo- and polysaccharide synthesis by Rhizobium meliloti SU47. J. Gen. Microbiol. 136:2511–2519. 5. Casse, F., C. Boucher, J. S. Julliot, M. Michel, and J. De´narie. 1979. Identification and characterization of large plasmids in Rhizobium meliloti using agarose gel electrophoresis. J. Gen. Microbiol. 113:229–242. 6. Charles, T. C., and T. M. Finan. 1991. Analysis of a 1600-kilobase Rhizobium meliloti megaplasmid using defined deletions generated in vivo. Genetics 127: 5–20. 7. Ditta, G., T. Schmidhauser, E. Yacobson, P. Lu, X. W. Liang, D. R. Finlay, D. Guiney, and D. R. Helinski. 1985. Plasmid related to the broad host-range vector, pRK290, useful for gene cloning and for monitoring gene expression. Plasmid 13:149–153. 8. Dubois, M., K. A. Gilles, J. K. Hamilton, P. A. Rebers, and F. Smith. 1956. Colorimetric method for determination of sugars and related substances. Anal. Chem. 28:350–356. 9. Dunn, M. F., S. G. Pueppke, and H. B. Krishnan. 1992. The nod gene inducer genistein alters the composition and molecular mass distribution of extracellular polysaccharides produced by Rhizobium fredii USDA193. FEMS Microbiol. Lett. 97:107–112. 10. Dylan, T., D. R. Helinski, and G. S. Ditta. 1990. Hypoosmotic adaptation in Rhizobium meliloti requires b-(132)-glucan. J. Bacteriol. 172:1400–1408. 11. Fahraeus, G. 1957. The infection of clover root hairs by nodule bacteria studied by simple glass technique. J. Gen. Microbiol. 16:374–381. 12. Figurski, D. H., and D. R. Helinski. 1979. Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in

APPL. ENVIRON. MICROBIOL. trans. Proc. Natl. Acad. Sci. USA 76:1648–1652. 13. Gilson, E., W. Saurin, D. Perrin, S. Bachellier, and M. Hofnung. 1991. Palindromic units are part of a new bacterial interspersed mosaic element. Nucleic Acids Res. 19:1375–1383. 14. Glazebrook, J., and G. C. Walker. 1989. A novel exopolysaccharide can function in place of the calcofluor-binding exopolysaccharide in nodulation of alfalfa by Rhizobium meliloti. Cell 56:661–672. 15. Gonza ´lez, J. E., B. L. Reuhs, and G. C. Walker. 1996. Low molecular weight EPS II of Rhizobium meliloti allows nodule invasion in Medicago sativa. Proc. Natl. Acad. Sci. USA 93:8636–8641. 16. Hirsch, P. R., and J. E. Beringer. 1984. A physical map of pPH1JI and pJB4JI. Plasmid 12:139–141. 17. Hulton, C. S., C. F. Higgins, and P. M. Sharp. 1991. ERIC sequences: a novel family of repetitive elements in the genomes of Escherichia coli, Salmonella typhimurium and other enterobacteria. Mol. Microbiol. 5:825–834. 18. Kannenberg, E. L., and N. J. Brewin. 1989. Expression of a cell surface antigen from Rhizobium leguminosarum 3841 is regulated by oxygen and pH. J. Bacteriol. 171:4543–4548. 19. Kannenberg, E. L., and N. J. Brewin. 1994. Host-plant invasion by Rhizobium: the role of cell-surface components. Trends Microbiol. 2:277–283. 20. Keller, M., A. Roxlau, W. M. Weng, M. Schmidt, J. Quandt, K. Niehaus, D. Jording, W. Arnold, and A. Pu ¨hler. 1995. Molecular analysis of the Rhizobium meliloti mucR gene regulating the biosynthesis of the exopolysaccharide succinoglycan and galactoglucan. Mol. Plant Microbe Interact. 8:267–277. 21. Leigh, J. A., E. R. Signer, and G. C. Walker. 1985. Exopolysaccharidedeficient mutants of Rhizobium meliloti that form ineffective nodules. Proc. Natl. Acad. Sci. USA 82:6231–6235. 22. Leigh, J. A., and G. C. Walker. 1994. Exopolysaccharides of Rhizobium: synthesis, regulation and symbiotic function. Trends Genet. 10:63–67. 23. Lloret, J., L. Bolan ˜ os, M. M. Lucas, J. Peart, N. J. Brewin, I. Bonilla, and R. Rivilla. 1995. Ionic and osmotic pressure induce different alterations in the lipopolysaccharide of a Rhizobium meliloti strain. Appl. Environ. Microbiol. 61:3701–3704. 24. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 25. Noel, K. D., D. M. Duelli, H. Tao, and N. J. Brewin. 1996. Antigenic change in the lipopolysaccharide of Rhizobium etli CFN42 induced by exudates of Phaseolus vulgaris. Mol. Plant Microbe Interact. 9:180–186. 26. Osteras, M., J. Stanley, and T. M. Finan. 1995. Identification of Rhizobiumspecific intergenic mosaic elements within an essential two-component regulatory system of Rhizobium species. J. Bacteriol. 177:5485–5494. 27. Reuhs, B. L., J. S. Kim, A. Badgett, and R. W. Carlson. 1994. Production of cell-associated polysaccharides of Rhizobium fredii USDA205 is modulated by apigenin and host root extract. Mol. Plant Microbe Interact. 7:240–247. 28. Ruvkun, G. B., and F. M. Ausubel. 1981. A general method for site-directed mutagenesis in prokaryotes. Nature 289:85–88. 29. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 30. Simon, R., U. Priefer, and A. Pu ¨hler. 1983. A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in gramnegative bacteria. Bio/Technology 1:784–791. 31. Simon, R., J. Quandt, and W. Klipp. 1989. New derivatives of transposon Tn5 suitable for mobilization of replicons, generation of operon fusions and induction of genes in gram-negative bacteria. Gene 80:161–169. 32. Spaink, H. P., R. J. H. Okker, C. A. Wijffelman, E. Pees, and B. J. J. Lugtenberg. 1987. Promoters in the nodulation region of the Rhizobium leguminosarum Sym plasmid pRL1JI. Plant Mol. Biol. 9:27–39. 33. Staskawicz, B., D. Dahlbeck, N. Keen, and C. Napoli. 1987. Molecular characterization of cloned avirulence genes from race 0 and race 1 of Pseudomonas syringae pv. glycinea. J. Bacteriol. 169:5789–5794. 34. Tao, H., N. J. Brewin, and K. D. Noel. 1992. Rhizobium leguminosarum CFN42 lipopolysaccharide antigenic changes induced by environmental conditions. J. Bacteriol. 174:2222–2229. 35. Vincent, J. M. 1970. A manual for the practical study of root-nodule bacteria. IBP handbook no. 15. Blackwell Scientific Publications Ltd., Oxford, United Kingdom. 36. Zahran, H. H., L. A. Ra ¨sa ¨nen, M. Karsisto, and K. Lindstro¨m. 1994. Alteration of lipopolysaccharide and protein profiles in SDS-PAGE of rhizobia by osmotic and heat stress. World J. Microbiol. Biotechnol. 10:100–105. 37. Zevenhuizen, L. P. T. M., and P. Faleschini. 1991. Effect of the concentration of sodium chloride in the medium on the relative proportions of poly- and oligosaccharides excreted by Rhizobium meliloti strain YE-2SL. Carbohydr. Res. 209:203–209. 38. Zhan, H., S. B. Levery, C. C. Lee, and J. A. Leigh. 1989. A second exopolysaccharide of Rhizobium meliloti strain SU47 that can function in root nodule invasion. Proc. Natl. Acad. Sci. USA 86:3055–3059. 39. Zhan, H., C. C. Lee, and J. A. Leigh. 1991. Induction of the second exopolysaccharide (EPSb) in Rhizobium meliloti strain SU47 by low phosphate concentrations. J. Bacteriol. 173:7391–7394.