Identification of the S. meliloti phaZ Open Reading Frame and Construction of an S. meliloti phaZ mutant

The phaZ gene was identified as a 1272 bp open reading frame SMc02770 in the S. meliloti genome sequence [20] by comparison to phaZ of Cupriavidus necator[13]. The amino acid sequences of these two proteins share 51% identity. Interestingly, like phaZ of C. necator, the PhaZ protein of S. meliloti does not possess a Gly-X-Ser-X-Gly lipase box motif [21] that is characteristic of many extracellular PHB depolymerases. The absence of this motif implies that these intracellular PhaZ homologues may use a different active site structure to extracellular PHB depolymerases. Primers were designed to internal regions of phaZ to amplify a fragment (from S35 to F292) by PCR, and the resultant 835 bp fragment was cloned into pGEM^®-T Easy (Promega) to generate pAZ101. An internal disruption of the cloned phaZ fragment was generated by introducing a ΩSmSp cassette as a Cfr91 fragment into the unique KpnI site at 299 bp to yield pAZ102. The phaZ::ΩSmSp was subsequently excised as an EcoRI fragment and subcloned into pK19mobsacB to give pAZ103. pAZ103 was introduced into S. meliloti Rm5000 by triparental mating using E. coli MT616 as a helper strain. Single recombinants were identified by selecting for Rf ^R , Sm ^R , Sp ^R transconjugants. Putative double recombinants were identified by plating onto TY Sm Sp Sucrose (5%). Subsequent screening for loss of vector-encoded Nm ^R confirmed the loss of pK19mobsacB. The resultant Rf ^R , Sm ^R , Sp ^R , Nm ^S phaZ mutant was designated Rm11417. The mutagenesis was confirmed by Southern blot using the phaZ PCR product as a probe. The probe hybridized to a 1.55 kb EcoRI fragment of genomic DNA in the wild-type strain Rm5000, and to a 3.55 kb fragment in Rm11417, confirming the presence of the 2 kb ΩSmSp cassette (data not shown). This mutation was transduced into Rm1021 using the ϕ M12 phage by standard techniques [22] and the resultant mutant was designated Rm11430.

Cloning of phaZ gene for complementation assays

Primers Smc02770F 5'-CCTAAGCTTATGTTCTACCAGCTTTACGAGATGAAC-3' and Smc02770R 5'-CGAAAGCTTTTAGTGATGGTGATGGTGATGGGCCGACTTGCCGCCCTTG-3' were designed to the 5' and 3' regions of SMc02770, incorporating HindIII sites into the 5' and 3' ends as well as a 3' terminal His tag. The PCR product was cloned as a HindIII fragment into pRK7813 and the resultant construct was named pMA157. This construct was introduced into Rm11430 by triparental conjugation using MT616 as the mobilizer strain.

Growth Phenotype of Rm11430 and ability to survive long-term carbon starvation

The ability of the phaZ mutant strain to withstand long-term carbon starvation was tested, relative to both Rm1021 and Rm11105, by incubation for 4 weeks in M9 liquid medium with no added carbon source. Cells were grown to late-log in YMB and washed twice in M9. A 1:50 dilution was used to inoculate 75 ml of M9 salts. Starting cfu/ml was determined immediately following inoculation by serial dilution of a 1 ml aliquot. Starting cultures typically contained approximately 2 × 10⁵ cfu/ml. These starting values were each given a relative value of 1. 1 ml samples were removed at 7 day intervals and serial dilutions were used to determine cfu/ml. Values presented are the averages of 3 independent cultures. The data in Figure. 1 show that the ability to synthesize and/or break down PHB has a significant impact on long-term survival in the absence of an exogenous carbon source. The wild-type strain Rm1021 is capable of increasing cell density during the early stages of starvation, presumably by degrading readily mobilizable intracellular carbon stores, a pattern which is not seen in either the phaZ or phaC mutants.

PHB accumulation

Effect on expression of succinoglycan synthesis genes

Symbiotic phenotype of Rm11430 and bacteroid PHB accumulation

Unlike bacteroids of determinate nodules, bacteroids of S. meliloti do not accumulate PHB during symbiosis (reviewed in [4]). Interestingly, a mutant of R. leguminosarum unable to cycle amino acids between the bacteroid and plants, showed apparent accumulation of PHB in the bacteroid within pea indeterminate nodules [11]. This suggests that the pathway for PHB metabolism can function within bacteroids of indeterminate nodules; however accumulation of PHB only occurs under extreme circumstances for example, when carbon is in excess and bacteroid metabolism is limited by the availability of a key nutrient. To confirm that S. meliloti bacteroids are capable of PHB synthesis and accumulation, alfalfa nodules induced by Rm11430 were prepared, sectioned and analyzed by TEM. Figure 2(a) clearly shows that bacteroids of Rm11430 accumulate PHB during symbiosis, with numerous, electron-transparent, PHB granules visible within the cytoplasm of the bacteroids when viewed by TEM. This is in contrast to bacteroids of Rm1021, shown in Figure 2(b), which demonstrate a notable absence of PHB.

Symbiotic assays with the host plant alfalfa revealed no significant difference between the phaZ mutant Rm11430 and the wild-type strain Rm1021. Plants inoculated with Rm11430 had an average shoot dry mass (SDM) of 10.56 mg compared to 10.80 mg for plants inoculated with Rm1021, both of which were significantly different to the uninoculated controls, which had an average SDM of 4.16 mg. This is interesting since it suggests that PHB accumulation, as confirmed in Figure 2, does not occur at the expense of symbiotic effectiveness.

Competitiveness for nodule occupancy of Rm11430

The role of EPS in the establishment of nitrogen-fixing symbioses between S. meliloti and M. sativa has long been acknowledged [29], but the precise mechanism of interaction remains elusive. Mutants unable to synthesize EPS are characteristically Fix^-. The observation that phaC and phaB mutants of S. meliloti are still able to establish successful symbioses [24] suggests that synthesis of succinoglycan in these mutants, albeit at a reduced level, is still sufficient to facilitate nodulation. This is consistent with previous reports which suggest that the production of small amounts of low-molecular-weight (LMW) EPS is sufficient to establish a successful symbiosis [29]. Indeed, it is conceivable that the competition defect observed in phaC mutants of S. meliloti may be due to extremely low levels of succinoglycan production. The phaC mutant may produce sufficient succinoglycan to establish an effective symbiosis but, assuming that the succinoglycan itself is playing a role in signalling during early nodulation, not enough to allow it to compete with strains producing higher levels of the EPS. Interestingly, the phaZ mutant demonstrates wild-type competitiveness and is able to out-compete both the phaC and bdhA mutants for nodulation. It is conceivable that another metabolic pathway that is dependent on D-3-HB metabolism may play a role in nodulation competitiveness. It is noteworthy that, although it has higher succinoglycan production than Rm1021, the phaZ mutant was not more competitive than the wild-type strain. While it is tempting to speculate that there may be a critical level of succinoglycan, above which, further gains in competitiveness are not seen, further information regarding the synthesis of succinoglycan during the infection process is still needed. Studies are currently underway in our lab to investigate this possibility further.

It is conceivable that, when PHB synthesis is inhibited, intermediates required for succinoglycan are not synthesized efficiently. It is also possible that, in the absence of a functional PHB synthesis pathway, enzymes required for succinoglycan may be inhibited or down-regulated. Furthermore, it has been suggested that acetyl phosphate may provide a regulatory link between PHB and succinoglycan synthesis [30]. Studies in the thermophilic cyanobacterium Synechococcus sp. strain MA19, have shown that acetyl phosphate is involved in the post-translational regulation of PHB synthase in vitro, and that this regulation is concentration-dependent [30]. As well, that study revealed that the enzyme phosphotransacetylase, which converts acetyl-CoA to acetyl phosphate, is only active under PHB-accumulating conditions [30]. In E. coli, acetyl phosphate is known to act as a global signal which acts through two-component regulatory signals [31], perhaps by direct phosphorylation of the response regulator [32] itself. Furthermore, the ChvI protein, of the S. meliloti ExoS-ChvI two-component regulatory system, is able to autophosphorylate in the presence of acetyl phosphate in vitro[33]. Since PHB synthesis mutants may excrete excess acetyl-CoA, levels of acetyl phosphate will likely be low under these conditions. Therefore, intracellular levels of acetyl phosphate may be an important factor in the ExoS-ChvI-dependent regulation of succinoglycan synthesis.

Bacterial strains, plasmids, growth media and conditions

Genetics and molecular biology techniques

Bacterial conjugations, ϕ M12 transductions and homogenotizations were carried out as described previously [22]. DNA manipulations were performed using standard techniques [45]. DNA probes for Southern blot analyses were labelled with digoxygenin (DIG) using the DIG High-Prime Kit (Roche Diagnostics Canada) according to manufacturer's instructions. Southern blots were performed using standard techniques [45]. PCR was carried out by standard techniques [45] using KOD Polymerase (Novagen, Canada).

Construction of exoF::TnphoA fusion

To generate plasmid-borne exoF::TnphoA fusions, plasmid pD82, a cosmid clone carrying the S. meliloti exoF gene and surrounding region of the genome [26], was introduced into the S. meliloti exoF::TnphoA fusion strain Rm8369 [27]. This construct was subsequently transferred into E. coli strain MT607, by triparental conjugation using E. coli strain MT616 as the mobilizer. Transconjugants were selected on LB KmTc, and the nature of the fusion was confirmed by testing for inability to confer YMA mucoidy on the exoF::TnphoA mutant Rm7055. The resulting construct was named pD82 exoF::TnphoA.

Biochemical assays

Alkaline phosphatase activity of exoF::TnphoA fusions in S. meliloti strains was measured according to the method of Brinkmann and Beckwith [46]. Cells were grown to an OD₆₀₀ of 0.7. 1 ml of culture was washed twice in 1 M Tris-HCl (pH 8.0), and resuspended in 1 ml 1 M Tris-HCl (pH 8.0). The OD₆₀₀ of this cell suspension was then measured. Following a 10 min equilibration period at 37°C, 50 μ l of 4 mg/ml p-nitrophenyl phosphate (NPP) was added to start the reaction. The reaction was allowed to continue for 11 min at 37°C before being stopped by the addition of 50 μ l of 1 M K₂HPO₄. The cells were pelleted and 50 μ l of the supernatant was diluted in 450 μ l of 1 M Tris-HCl (pH 8.0) and OD₄₂₀ was measured. Units (U) of alkaline phosphatase activity were calculated using the formula:

(1)

Assuming a molar coefficient of 16,000 for p-nitrophenyl phosphate, 1 U is equal to 0.062 nmol of NPP hydrolyzed per min at a cell OD₆₀₀ of 1. Therefore:

(2)

For PHB assays, 50 ml cultures were grown at 30°C to stationary phase in YMB. Cells were harvested and washed in 0.85% NaCl solution before resuspension in 50 ml 0.85% NaCl. PHB was extracted from a 2 ml fraction of this suspension and the remaining 48 ml was used for cell dry weight determination by incubation of the pellet at 60°C until the pellet was dry and no further loss in mass was recorded. PHB content was measured by the method of Law and Slepecky [47] and expressed as a percentage of total cell dry weight. All glassware was washed in hot chloroform and rinsed in ethanol before use, to eliminate plasticizers. A standard curve was constructed by dissolving known quantities of PHB (Sigma) in hot chloroform to a final volume of 1 ml. The chloroform was allowed to evaporate before addition of 10 ml of H₂SO₄ and PHB was processed as described elsewhere [47].

Carbon starvation assay

Saturated TY cultures were washed twice to remove traces of nutrients, and were subcultured 1:50 into carbon-free M9 medium. These cultures were incubated at 30°C, shaking at 180 rpm. Viable cell counts were monitored at weekly intervals by plating on TY agar. Samples at t = 0 were each given a relative value of 1, and all subsequent samples are compared to this starting value. Values recorded are the means from triplicate cultures.

Nodulation studies

Medicago sativa cv. Iroquois seeds were surface-sterilized in 95% ethanol for 2 min followed by 15 min in 0.5% sodium hypochlorite. After several washes in sterile dH2O, seeds were germinated in the dark on sterile water agar plates at room temperature for approximately 36 hours. Seedlings were transferred to modified Leonard assemblies containing sterilized vermiculite soaked in Jensen's N-free plant nutrient solution [48]. Five seedlings were planted in each jar and inoculated with 5 ml of 1:50 dilution of saturated TY culture. The assemblies were placed in a growth chamber (Conviron CMP3244, Model # EF7, Controlled Environments Ltd., Winnipeg) with 16 h, 25°C day/8 h, 20°C night and light intensity of 300 μ moles m^-2s^-1. For shoot dry weight determination, plants were harvested approximately 5 weeks post-inoculation and the shoots separated from the roots. The shoots were transferred to brown paper bags and incubated at 60°C until no further loss in mass was recorded. Shoot dry weight is expressed as mg^-1 plant^-1. Nodule occupancy competitiveness was assayed in modified Leonard assemblies as described above. Inoculants consisted of wild-type and mutant cultures mixed in 1:1 and 1:9 ratios, or mutant cultures mixed in a 1:1 ratio. Plants were harvested four weeks post-inoculation and nodules were collected. Nodules were surface-sterilized with 1% sodium hypochlorite (15 min), washed twice with LB, and then squashed in a few drops of TY containing 0.3 M sucrose. The resultant suspension was streaked on TY. Four colonies isolated from each nodule were screened for the appropriate antibiotic-resistance marker. The bacterial population within each nodule was thus scored as either consisting of one strain or a mixture of two strains.

Electron microscopy

M. sativa plants were harvested 28-30 days post-infection. Roots were washed to remove traces of vermiculite, and the nodules were transferred into primary fixative (4% formaldehyde, 1% glutaraldehyde in 80 mM HEPES pH 7.0) and cut into small pieces. The samples were subjected to 4 cycles of vacuum infiltration (2 mins per cycle) and were left overnight at 4°C. Following infiltration, the nodules were washed thoroughly in sterile water, and stained for 4 hours in 1% OsO₄. The nodules were washed again in water and dehydrated through a gradient of acetone. The nodules were embedded in epon araldite resin and transferred to BEEM capsules for 48 hours at 60°C. Ultrathin sections were cut using a Reichert Ultracut E microtome, and were stained with uranyl acetate and lead citrate using standard techniques [49]. Samples were analyzed in a Philips CM10 transmission electron microscope at an accelerating voltage of 60 kV.