|Home | About | Journals | Submit | Contact Us | Français|
Sinorhizobium meliloti cells store excess carbon as intracellular poly-3-hydroxybutyrate (PHB) granules that assist survival under fluctuating nutritional conditions. PHB granule-associated proteins (phasins) are proposed to regulate PHB synthesis and granule formation. Although the enzymology and genetics of PHB metabolism in S. meliloti have been well characterized, phasins have not yet been described for this organism. Comparison of the protein profiles of the wild type and a PHB synthesis mutant revealed two major proteins absent from the mutant. These were identified by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) as being encoded by the SMc00777 (phaP1) and SMc02111 (phaP2) genes. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of proteins associated with PHB granules followed by MALDI-TOF confirmed that PhaP1 and PhaP2 were the two major phasins. Double mutants were defective in PHB production, while single mutants still produced PHB, and unlike PHB synthesis mutants that have reduced exopolysaccharide, the double mutants had higher exopolysaccharide levels. Medicago truncatula plants inoculated with the double mutant exhibited reduced shoot dry weight (SDW), although there was no corresponding reduction in nitrogen fixation activity. Whether the phasins are involved in a metabolic regulatory response or whether the reduced SDW is due to a reduction in assimilation of fixed nitrogen rather than a reduction in nitrogen fixation activity remains to be established.
The alfalfa root nodule symbiont Sinorhizobium meliloti stores excess carbon as intracellular poly-3-hydroxybutyrate (PHB) granules as its main carbon storage compound. Mutant analysis has demonstrated that PHB metabolism plays a role in rhizobium-legume symbiosis (19, 38, 39, 43), although the metabolic role of the PHB cycle during nitrogen fixation is still not completely understood (28). While the enzymology and genetics of PHB biosynthesis have been studied extensively with various bacteria (35), less is known about the regulation of this process in S. meliloti. So far, the following two major types of PHB accumulation effectors have been investigated in several bacteria: (i) the granule-associated proteins, or phasins, encoded by phaP genes, which bind to PHB granules and promote PHB synthesis; and (ii) a regulator, encoded by phaR (15). PhaR was first designated AniA in rhizobia because of its expression under anaerobic growth conditions (27). Although the function of aniA has still to be clarified, Povolo and Casella provided evidence that AniA, in partitioning carbon flow in cells, affects not only PHB production but also the production of extracellularly polymeric substances and nitrogen fixation in S. meliloti Rm41 (27). In Rhizobium etli, this protein has been proposed to be involved in directing carbon flow (8, 9). Phasins have not yet been described for rhizobia.
Phasins are characterized by low molecular masses (mostly between 11 and 25 kDa), have an amphiphilic character and a high affinity for polyhydroxyalkanoate (PHA) inclusions, and can comprise a significant fraction of total cell proteins (13, 44). Phasins and their structural genes, phaP, have been found in various bacteria. These proteins have been shown to play a major role in the synthesis and degradation of PHB and in the formation of PHB granules (26). For example, Ralstonia eutropha H16 has four phasin genes, namely, phaP1, phaP2, phaP3, and phaP4. These genes are all expressed (24, 33), but only PhaP1, the major phasin, appears to influence PHB accumulation (25). Moreover, York et al. demonstrated that PhaP promotes PHB synthesis by regulating the surface/volume ratio of PHB granules or by interacting with PHB synthase, and the levels of PhaP generally parallel levels of PHB in cells (45). Methylobacterium extorquens AM1 has two major phasins, and mutations in their genes result in defective PHB production and also in inhibited growth on C2 compounds, while not affecting growth on C1 or multicarbon compounds (15). Phasins appear to be present in all PHA-synthesizing bacteria, and even though they generally are not conserved in sequence, they are believed to fulfill the same functions, binding to PHA granules and promoting PHA granule formation in a manner that is still poorly understood (14).
In this study, we identified two major proteins associated with PHB granules, namely, PhaP1, encoded by SMc00777 (phaP1), and PhaP2, encoded by SMc02111 (phaP2), in S. meliloti Rm1021. To understand the functions of phaP1 and phaP2, mutations in these genes were generated. The effects of the phaP mutations on PHB formation and accumulation were investigated. Furthermore, we also investigated the effects of mutation of these genes on nodulation and nitrogen fixation.
Strains and plasmids used in this study are listed in Table Table1.1. S. meliloti strains were cultured in TY (2) or YMB (37) medium at 30°C. Antibiotics were used at the following concentrations: 100 μg ampicillin (Am) ml−1, 20 μg kanamycin (Km) ml−1, 200 μg neomycin (Nm) ml−1, 20 μg chloramphenicol (Cm) ml−1, and 200 μg streptomycin (Sm) ml−1. Escherichia coli strains were grown in Luria-Bertani (LB) medium (22). Antibiotics for E. coli were used at the following concentrations: 20 μg Km ml−1 and 20 μg Cm ml−1. M9 minimal medium with various carbon sources, each at a final concentration of 15 mM, was prepared as described previously (3, 4). Sucrose was added to the medium at 5% (wt/vol), when required. Media were solidified by the addition of 1.5% (wt/vol) agar.
Wild-type Rm1021 and phbC mutant Rm11105 cells were grown in YMB for 3 days, after which they were well within stationary phase. Cells were sonicated with a Sonifier150 instrument (Branson Ultrasonics Corporation, Danbury, CT) with a microtip in an ice bath. A sonicate from approximately 2 × 107 cells was prepared according to the protocol provided with ProteomeLab PF 2D kits (kit recorder no. 390977; Beckman Coulter, Inc.). Proteomic maps were generated with a Proteomelab PF 2D system as described previously (34).
PHB granules were isolated by a modification of the method described by Preusting et al. (29). Cells were harvested from 3-day-old 250-ml YMB cultures, washed, and resuspended in 10 ml 100 mM potassium phosphate buffer (pH 7.5). After three passages through a French press (110 × 106 Pa), 5 ml of the lysate was loaded on a discontinuous linear sucrose gradient (1 to 2 M) consisting of 8 ml each of 2, 1.66, 1.33, and 1 M sucrose in 10 mM Tris-HCl (pH 8.0) in an ultracentrifuge tube (Beckman Instruments, Inc.). After 15 h of centrifugation (Beckman SW 28 rotor; 4°C) at 26,000 rpm, the granules were removed from the gradient, washed twice with 10 mM Tris-HCl (pH 8.0), and then stored at −80°C.
Pelleted granules were resuspended in gel loading buffer. After 5 min of incubation in loading buffer at 100°C, the granule-associated proteins were separated by 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as described by Laemmli (16). Proteins were stained with Coomassie brilliant blue R-250 (41). Protein bands were digested in the gel with trypsin (Promega, Madison, WI). Before measuring the mass of the peptide mixture, the peptides were purified using a C18 ZipTip according to the manufacturer's instructions (Millipore, Bedford, MA). Purified peptides were eluted with saturated α-cyano-4-hydroxycinnamic acid in 50% acetonitrile-0.5% trifluoroacetic acid for matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS) and were analyzed with a MicroMass M@LDI instrument (Waters Corporation, Milford, MA). Peptide mass fingerprints were analyzed using ProteinLynx global server software (Waters Corporation, Milford, MA), and a custom peptide mass fingerprint database was created using S. meliloti genome sequence data (12; http://bioinfo.genopole-toulouse.prd.fr/annotation/iANT/bacteria/rhime/).
Cloning procedures, including DNA isolation, restriction digestion, ligation, and transformation, were performed as described by Sambrook and Russell (30). Conjugal mating with DH5α(pRK600) as the helper strain was performed as previously described (5). ΦM12 generalized transduction was carried out as described by Finan et al. (10).
To construct phaP1, phaP2, and phbC mutants, data from the S. meliloti genome sequence were used to design primers specific for the predicted phaP1 (SMc00777), phaP2 (SMc02111), and phbC (SMc00296) genes of S. meliloti Rm1021. The amplified DNA fragments were confirmed by DNA sequencing at the Virginia Bioinformatics Institute Core Laboratory Facility (https://www.vbi.vt.edu/core/).
To disrupt the S. meliloti phaP1 gene, a 301-bp internal piece from the phaP1 gene was amplified from S. meliloti Rm1021 chromosomal DNA by using the following primers: phaP1-F, 5′-TACCAGGAAGACCGAAGACG-3′; and phaP1-R, 5′-GGTCTGCAGTTCGACGAGTT-3′. PCR was performed by using 30 cycles of 30 s at 95°C, 30 s at 58°C, and 30 s at 72°C. The PCR fragment was purified, cloned into T-Easy vector, digested with EcoRI, and cloned into the vector pK19mob, creating pXS001. The resulting plasmid, pXS001, was transferred into Rm1021 by triparental mating from DH5α, using DH5α(pRK600) as a helper, where the cointegrate was made by single recombination. S. meliloti clones carrying the chromosomal cointegrate were selected by being plated on TY plates containing Sm and Nm. The presence of the insertion was confirmed by PCR followed by sequencing, and the corresponding strain was designated SB100.
The unmarked S. meliloti phaP2 deletion mutant was generated by crossover PCR (36). In brief, PCR on the Rm1021 genomic DNA template was used to generate fragments to either side of the phaP2 sequence targeted for deletion. For amplification of the N-terminal end of phaP2, primers A (5′-TCTATCTCGGCGACGAATTT-3′) and B (5′-GCCGTCGACGAGCGAGAGGCACGGGTGTCTCCTTGTGACGG-3′) were used, resulting in an amplification product of 413 bp. For amplification of the C-terminal end of phaP2, primers C (5′-TGCCTCTCGCTCGTCGACGGCCCATCAGCCTCATCGCTATT-3′) and D (5′-CAAACTCGGCTTCTTGGTCT-3′) were used, resulting in an amplification product of 459 bp. Primers B and C contained 21-bp complementary sequences at their 5′ ends (underlined). PCR products resulting from these two amplifications were gel purified and combined as a template for a PCR using primer pairs A and D. A fragment of the desired 748-bp size was obtained and ligated into the pGEM-T Easy vector (Promega Corporation, Madison, WI), and the sequence was confirmed to be correct; the fragment was then subcloned from this construct into pK19mobsacB (31) as an EcoRI fragment to make pXS002. Gene replacement in Rm1021 was carried out by first introducing pXS002 by triparental mating, selecting for Smr Nmr single-crossover cointegrants. Selection for double-crossover events was carried out on TY agar containing 5% sucrose, followed by screening for Nm sensitivity. The presence of the deletion was confirmed by PCR and DNA sequencing. The corresponding strain was designated SB104.
To disrupt the S. meliloti phbC gene by insertion using vector pTH1703 (Gmr) (7), a 669-bp DNA fragment internal to phbC was PCR amplified with primers phbC-F (5′-ACGGACACCAGCAAGTTC-3′) and phbC-R (5′-CCAGTAAAGCAGGTCGAAGG-3′). The PCR product was ligated into the pGEM-T Easy vector to give plasmid pXS1, and the NotI fragment from pXS1 was subcloned into pTH1703 vector to generate the plasmid pXS003, which was verified by DNA sequencing. Integration of pXS003 by homologous recombination into the S. meliloti genome resulted in the gentamicin-resistant phbC mutant strain SB110. The presence of the insertion was confirmed by PCR followed by sequencing.
To construct the phaP1 phaP2 double mutant, the Nm resistance marker carried by SB100 (Rm1021 phaP1) was transduced into SB104 (Rm1021 phaP2) to give SB108. To generate the triple mutant SB119 (Rm1021 phaP1 phaP2 phbC), the gentamicin resistance marker from SB110 (Rm1201 phbC) was transduced into SB108 (Rm1021 phaP1 phaP2).
The concentration of PHB was determined from cells grown in YMB medium, using a spectrophotometric technique (17) as modified by Peoples and Sinskey (23). The exopolysaccharide (EPS) content was determined as previously described (20), using cultures grown in YMB for 4 days at 30°C. Glycogen was extracted from cells and determined by anthrone assay according to the procedure of Chun and Yin (6).
Cells were washed, suspended in 0.1 M sodium phosphate buffer (SPB; pH 6.8), fixed with 2% (vol/vol) glutaraldehyde in 0.1 M SPB at 4°C for 1 h, and washed in 0.1 M SPB. Subsequently, they were postfixed in 0.5% (wt/vol) osmium tetroxide (OsO4) in 0.1 M SPB at room temperature overnight, dehydrated, and then embedded in a Poly/Bed 812 resin (Polysciences Inc., Warrington, PA). Sections with a thickness of 70 to 90 nm were made with a Leica Ultracut UCT diamond knife (Leica Inc., Deerfield, IL) and placed on a 200-mesh copper grid. Imaging was performed with a Zeiss 10CA transmission electron microscope (Carl Zeiss Inc., Oberkochen, Germany) at a 60-kV accelerating voltage and at room temperature.
Symbiotic phenotype assays were performed with Medicago truncatula Jemalong (line A17) and alfalfa (Medicago sativa cv. Iroquois) plants in growth pouches and growth chambers as described before (39). Shoot dry weights (SDW), numbers of nodules, nodule dry weights, and nitrogen fixation activities (acetylene reduction) were determined 5 weeks after inoculation.
All data presented for PHB, EPS, glycogen, nodule numbers, acetylene reduction assays, plant dry weights, and nodule dry weights are given as means with standard errors. The significance of the results was assessed using Student's t test.
Proteome maps of S. meliloti soluble proteins were generated by a ProteomeLab PF 2D system, which separated proteins according to pI in a first-dimension chromatofocusing step and according to hydrophobicity in a reversed-phase step in the second dimension. The data were displayed as a UV/pI map by using ProteoVue* software, which displays the pI fractions as lanes and the second-dimension retention time as the vertical position in individual lanes. The fractionated proteomes of Rm1021 and Rm11105 were compared using DeltaVue* software (Fig. (Fig.1).1). Two major bands were present in Rm1021 but absent in Rm11105 (see the red bands in lane 1 and 2, with retention times of 15.4 min and 17.5 min, respectively). These two bands were selected for MALDI-TOF MS analysis, which predicted peptide masses that were in agreement with the amino acid sequences deduced from the putative genes for the SMc00777 (15.9 kDa) and SMc02111 (12.7 kDa) proteins from S. meliloti Rm1021.
PHB granules were isolated from the wild-type strain Rm1021. In control experiments, no granules could be isolated from the phbC mutant. The granule-associated proteins were solubilized and separated by SDS-PAGE. Two proteins, one of approximately 16 kDa and another of 13 kDa, were observed as predominant granule-associated proteins (Fig. (Fig.2).2). These proteins were isolated from the gel and digested with trypsin. The molecular masses of the resulting fragments were subsequently determined by MALDI-TOF, which predicted the SMc00777- and SMc02111-encoded proteins. Based on this analysis, SMc00777 and SMc02111 were designated phaP1 and phaP2, respectively. The encoded PhaP1 and PhaP2 proteins fall within COG 5490, which includes a number of experimentally determined PHA granule-associated proteins from M. extorquens AM1 (15), Magnetospirillum gryphiswaldense (32), and Alcaligenes eutrophus (42). Members of COG 5490 are found exclusively within the Alphaproteobacteria and Betaproteobacteria.
We also analyzed PHB granules from the phaP mutants. Granules were isolated from YMB-grown cultures and subjected to SDS-PAGE. In the phaP1 mutant, only the 13-kDa band was present, while in the phaP2 mutant, only the 16-kDa band was seen. These data confirmed that the products of phaP1 and phaP2 are the major phasins in S. meliloti Rm1021. No PHB granules could be isolated from the double and triple mutants (Fig. (Fig.22).
To determine the growth behavior of strains and their ability to synthesize and accumulate PHB, EPS, and glycogen, the phaP mutants and wild-type strains were cultivated under conditions permissive for PHB accumulation in liquid YMB medium. The wild-type strain and the phaP2 single mutant showed similar growth kinetics (Table (Table2).2). However, the phaP1 single mutant, the double mutant SB108 (phaP1 phaP2), and the triple mutant SB119 (phaP1 phaP2 phbC) grew slower than the wild type and the phaP2 mutant. Lower PHB levels were determined for the phaP1 and phaP2 single mutants, while PHB was not detected in the double mutant (phaP1 phaP2) or the triple mutant (phaP1 phaP2 phbC) (Table (Table2).2). Surprisingly, unlike the phbC mutant, which had undetectable EPS production, the PHB granule-associated protein mutants produced more EPS than did the wild-type strain, and the effect was more dramatic in the phaP1 and double mutants (Table (Table2).2). The phaP mutations also resulted in significantly more glycogen production than that in the wild type or the phbC single mutant. Moreover, the production of glycogen was significantly higher in the double mutant (phaP1 phaP2) than it was in the single mutants in YMB medium (Table (Table22).
Carbon source utilization patterns were determined on M9 agar with either succinate, fructose, sucrose, glucose, acetoacetate, d-3-hydroxybutyrate, or pyruvate as the sole carbon source. All strains had similar growth behaviors on fructose, sucrose, glucose, and pyruvate. On succinate, the phbC and phaP1 single mutants, the double mutant (phaP1 phaP2), and the triple mutant (phaP1 phaP2 phbC) grew slightly slower, forming smaller colonies after 4 days of growth than those of the wild type. Consistent with previous reports, the growth of the phbC mutants was severely reduced on d-3-hydroxybutyrate, and no growth was observed on acetoacetate. However, this phenomenon was not observed for the phaP mutants that did not contain a phbC mutation, which exhibited growth similar to that of the wild-type strain on these two carbon sources.
Cells of wild-type Rm1021 and the phaP mutants were also analyzed by electron microscopy. Similar to the wild-type strain, both single mutants (phaP1 and phaP2) still exhibited PHB granules in the cytoplasm. As expected from the PHB assay, similar to the case for the phbC mutant, no PHB granules were detectable in the cytoplasm of the double mutant (phaP1 phaP2) (Fig. (Fig.3).3). Merodiploid complementation experiments confirmed that these phenotypes were a direct result of the mutations (data not shown).
To address the question of whether phaP mutations influence the effectiveness of symbiosis, we performed plant inoculation experiments, using M. truncatula and alfalfa as hosts. Consistent with previous studies (39), M. truncatula plants inoculated with the phbC mutants displayed SDW similar to that of the uninoculated control, and these plants exhibited severely reduced levels of acetylene reduction activity (ARA) and numbers of nodules (Fig. (Fig.4).4). The phaP single and double mutants also had similarly reduced levels of SDW and nodule numbers, but they had intermediate levels of ARA activity. There was no clear distinction between the phaP single mutants and the double mutant; all three mutants appeared to be affected similarly. On alfalfa, inoculation with each of the mutants resulted in nodule-bearing plants with similar SDW and ARA to those of plants inoculated with the wild type (data not shown).
During an experimental search for proteins that are differentially present in PHB synthase mutants, we identified two proteins that copurify with PHB granules and appear to be the major granule-associated proteins. The genes encoding these two proteins, SMc00777 and SMc02111, are carried in monocistronic operons and are not located near any other genes that have been implicated in PHB synthesis. The genes encode predicted proteins of sizes consistent with the physical measurements of the identified proteins. Both proteins had previously been assigned to the same COG group (COG 5490), and it appears that members of this group are found within many Alphaproteobacteria, with at least one example (42) in the Betaproteobacteria. A number of these have been determined by experimentation to be PHA granule-associated proteins, e.g., those from M. extorquens AM1 (15) and M. gryphiswaldense (32).
We found that the PhaP proteins are required for PHB granule accumulation in S. meliloti. Our studies have not addressed whether the granule-associated proteins are involved in PHB granule degradation. The functions of the PhaP proteins are apparently redundant, as PHB granules are detectable when either one of the genes is mutated but not in the double mutant. The specific role of these proteins in granule formation will be the basis of future investigations.
An intriguing result was the increase in EPS synthesis in the phaP mutants under growth conditions that yield reduced levels of EPS synthesis in other PHB synthesis mutants (1). Whether this increased EPS is succinoglycan, galactoglucan, or some other EPS is not known at this point. Similarly, the phaP mutants did not exhibit the expected acetoacetate and d-3-hydroxybutyrate utilization deficiencies of other PHB accumulation mutants (39). Previous studies have demonstrated a role for AniA (PhaR) in the regulation of EPS production in S. meliloti; aniA mutants produce more EPS than the wild-type strain does (27). Based primarily on data from studies of R. eutropha, a model was recently proposed for the regulation of PHB granule formation by PhaR (25), incorporating a role for PhaP. In this model, phaP expression is under the transcriptional repression control of PhaR. When PHB granules are first formed, PhaR binding to these granules results in reduced concentrations of PhaR in the cytoplasm, yielding increased phaP synthesis. The increased PhaP concentration then results in displacement of PhaR from the surfaces of the granules, increasing the cytoplasmic concentration of PhaR and thus reducing phaP transcription. Investigation of such a model for PHB accumulation in S. meliloti will first require demonstration that phaP transcription is under the control of AniA (PhaR).
Our electron microscopy results match the PHB quantitation assay results. The double mutant (phaP1 phaP2) and the phbC mutant lack PHB. However, there appears to be some sort of inclusion in the double mutant (Fig. (Fig.3E)3E) that is missing in the phbC mutant (Fig. (Fig.3B).3B). Since the double mutant produced a significantly larger amount of glycogen than that produced by the phbC mutant in YMB medium (Table (Table2),2), we assume that the inclusion in the double mutant could be glycogen. However, the carbon flow, especially a correlation between PHB and glycogen in rhizobia, is far from clear (1, 9, 27, 38, 39). We are pursuing further studies to clarify the role of phasins in regulating storage polymer synthesis and carbon metabolism in S. meliloti.
In R. eutropha, production of PhaP protein is dependent upon the production of PHB (44). By analogy, one might expect that the unusual phenotypes of S. meliloti phaP mutants would be dependent on the presence of a wild-type phbC gene. In order to rule out the possibility that the phaP mutants have characteristics that are independent of PHB production in S. meliloti, we constructed a triple mutant (phaP phaP2 phbC). Phenotypic assays showed that the triple mutant has a longer doubling time than the phbC mutant and retains high glycogen production. These results indicate that the phaP mutants have defects that are independent of PHB production.
Although PHB is not accumulated in S. meliloti bacteroids within M. truncatula nodules, we have recently shown that phbC mutants produce nodules with greatly reduced N2 fixation activity (39). There was some question as to whether this reduction was due to reduced levels of infection related to deficiencies in production of symbiotically effective EPS. Since the phaP mutants exhibited reductions in PHB granule formation without concomitant reductions in EPS synthesis, we felt that investigation of the symbiotic phenotype of these mutants would contribute to the clarification of this issue. The nodulation phenotype exhibited by the phaP mutant nodules suggested a metabolic defect related to bacteroid nitrogen metabolism. Similar to pea nodules elicited by aap/bra amino acid transport mutants of Rhizobium leguminosarum (18), these nodules exhibited substantially reduced SDW, although ARA was only moderately reduced. This contrasts with the reduction in SDW of phbC mutant nodules, which is accompanied by a similar reduction in ARA. While the S. meliloti phbC phenotype could potentially be due to a reduction in EPS that might influence the infection process, this is not the case for the phaP mutants that do not have reduced EPS levels. It would be interesting to investigate whether phasins play a regulatory role in amino acid cycling in bacteroids within M. truncatula nodules. Moreover, we note that similar to the phbC mutant, phaP mutants formed productive symbioses on the host plant alfalfa. Differences observed between M. truncatula and alfalfa emphasize the potential for host-dependent effects that should not be ignored.
We are grateful to T. M. Finan for providing pTH1703 prior to publication. We thank Chunhong Mao and Endang Purwantini for critical readings of the manuscript. We appreciate Vladimir Shulaev and Joel Shuman for technical assistance with gas chromatography and Kathy Lowe for assistance with electron microscopy.
This work was supported by the Virginia Bioinformatics Institute. Work in the lab of T.C.C. was supported by NSERC.
Published ahead of print on 5 October 2007.