A single copy of the green fluorescent protein (GFP)-encoding gene gfp-P64L/S65T under the control of the constitutive nptII promoter was introduced in a neutral region of the Sinorhizobium meliloti chromosome, between the genes recA and alaS. Within the same chromosomal region downstream of gfp-P64L/S65T a tetracycline (Tc) resistant cassette was also inserted. Both markers were very stable during at least 40 bacterial generations without any selective pressure. Similarly, the gfp-Tc cassette was stable and functional in all rhizobia that were recovered from alfalfa nodules. The GFP-associated fluorescence derived from the (single copy) chromosomal gfp-P64L/S65T allowed detection of rhizobia during the colonisation of the root, infection thread formation, and nodule development. The gfp-Tc rhizobia showed indistinguishable phenotypes for nodulation, competitiveness, and nitrogen-fixation from the parental strain. The labelling system described here can be used for the stable fluorescent tagging of S. meliloti strains allowing their detection in biologically complex soil environments.
A significant amount of practical and fundamental knowledge on the biology of the rhizobium–legume symbioses has been gathered over the years [1,2], allowing this association to become a model system for the study of other plant–microbe interactions. However, owing to the difficulty of monitoring specific bacterial genotypes in the background of large numbers of heterogeneous microbial populations, most studies on the molecular ecology of rhizobia have been carried out principally in the last decade. The study of the establishment of rhizobia inoculated into natural soils, their rhizosphere colonisation, nodule occupancy, and putative exchange of genetic traits with other soil bacteria all require the use of properly tagged strains for their unequivocal recognition in a biodiverse environment.
The green fluorescent protein (GFP) from Aequorea victoria is being used increasingly as a reporter for gene expression and protein localisation in bacterium–host interactions, as well as for in vivo studies in microbe, plant, insect, and mammalian systems. The ability of GFP to fluoresce in the absence of any added cofactor makes it particularly useful for in situ studies. The use of GFPs and other autofluorescence proteins has already been reported in rhizobia. First, GFP-S65T was cloned in a replicative plasmid and was successfully expressed in Sinorhizobium meliloti cells. However, 40% of the rhizobia recovered from the root nodules had lost the fluorescent marker due to plasmid instability. To ensure that every bacterial cell remained labelled with GFP throughout nodulation without selective pressure, Cheng and Walker constructed a GFP-expressing vector that is stable and replicates in S. meliloti cells in planta. Using a different approach, the stability of the fluorescent marker in S. meliloti was recently achieved by the genomic integration of a GFP-S65T-encoding non-replicative plasmid. The system was successfully used to investigate the expression of exudate-inducible S. meliloti genes in the presence of different plant species. Color variants of fluorescent proteins were also cloned in a stable broad-host-range vector to visualise rhizobia interacting with plants. Although these constructions are useful for studies under controlled laboratory conditions, they should not be used in non-sterile soil assays since the recombinant plasmids (either integrated or not) could be easily dispersed by conjugal transfer from the reference strain to other rhizobia or even to another species [8,9]. The conjugal-transfer origin present in the plasmids might eventually promote genetic mobilisation if compatible helper functions are provided by other replicons acquired from a second strain (i.e. present in the soil).
In this work we present the construction and characterisation of a S. meliloti strain that carries a chromosomal single copy of the gfp-P64L/S65T which proved to be stable, symbiotically neutral, and useful to detect single fluorescent rhizobial cells.
2Materials and methods
2.1Bacterial strains and plasmids
The strains and plasmids used are listed in Table 1. Escherichia coli and S. meliloti strains were grown on Luria–Bertani medium at 37°C, and on TY or YEM at 28°C, respectively. For solid media 15 g of agar per litre of medium was added. The final concentration of antibiotics per millitro of medium was 5 μg gentamycin (Gm) and 6 μg tetracycline (Tc) for E. coli, and 400 μg streptomycin (Sm), 50 μg Gm and 6 μg Tc for S. meliloti.
Table 1. Bacterial strains and plasmids used in this work
Strain or plasmid
E. coli DH5α
recA, ΔlacU169, F80dlac ZDM15
Bethesda Res. Lab.
E. coli S 17–1
E. coli 294 RP4-2-Tc::Mu-Km::Tn7 integrated into the chromosome
Gmr, Tcr, pSM10 carrying in Bam HI (blunt ended by Klenow enzyme) a 2-kb Tcr DNA cassette from pHP45Ω-Tc cloned as Sma I
Gmr, Tcr, pMP5 carrying in Eco RI a 0.75-kb DNA fragment containing the gfp-P64L/S65T cassette from pGreenTIR cloned as Eco RI
2.2DNA manipulation and genetic constructs
Plasmid DNA preparation, restriction enzyme analysis, cloning procedures, and E. coli transformation were performed according to previously established techniques. Southern hybridisation was carried out using DNA probes labelled with digoxigenin. The probes were synthesised by PCR using digoxigenin–dUTP (Boehringer Mannheim) and appropriate primers to amplify the region of interest. For hybridisation, DNA extracted from bacteria was digested and transferred to nitrocellulose membranes (Hybond N; Amersham) as described by Chomczynski. The digoxigenin-labelled DNA probes were hybridised to the membranes at 65°C overnight after the blocking of non-specific binding sites for 1 h at 68°C, using the solutions and experimental conditions specified by Boehringer Mannheim (Catalogue No. 1093 657). For the visualisation of positive bands, the membranes were incubated with an antibody against the digoxigenin ligand and washed, and a final colour reaction was initiated at alkaline pH by the addition of X-phosphate plus nitroblue tetrazolium chloride as specified by the manufacturer.
Transfer of plasmids between E. coli S17–1 and S. meliloti were performed according to Simon et al..
2.4Plant nodulation assays
Medicago sativa seeds (alfalfa, cv CUF101 obtained from Instituto Nacional de Tecnología Agropecuaria, La Pampa, Argentina) were surface-sterilised for 10 min with commercial bleach 20% v/v (NaClO concentration equivalent to 55 g active Cl2 l−1) followed by six washes with sterile distilled water. Surface-sterilised seeds were germinated on water agar (1.5%, w/v). For infection and competition studies 2-day-old seedlings were transferred to gamma-irradiated sterilised plastic growth pouches (Mega Minneapolis International, Minneapolis, MN, USA) containing 10 ml of nitrogen-free Jensen mineral solution, pH 6.7. Three days later, primary roots were inoculated with 106 rhizobia in Jensen medium by dripping 100 μl of a bacterial suspension onto the root from the tip toward the base. The rhizobia were obtained from exponential-phase YEM cultures. The plants were cultured in a growth chamber at 22°C for a 16-h photoperiod. When two different strains were co-inoculated for competition experiments both rhizobia were mixed prior to plant inoculation (106 CFU/100 μl inoculant each). The CFU contained in the inocula were estimated by plate counts.
Root hair curling, infection threads, and nodules were examined under a Carl Zeiss Jena fluorescent microscope either with visible light or UV with a filter set for fluorescence. In this last case a 470–490-nm excitation filter and a 520–560-nm barrier filter were used. Nodule occupancy was scored 4 weeks after inoculation by counting fluorescent nodules. In addition random-selected nodules from the same experiments were removed from the roots, surface-sterilised with 30 vol. H2O2 for 10 min followed by washing with abundant sterile water, and crushed in 10 μl of sterile isotonic solution. Appropriate dilutions were plated on YEM Petri dishes with or without Tc, and the proportion of the antibiotic resistant CFUs was compared to the proportion of fluorescent nodules scored by eye under the microscope.
Nitrogen fixation was estimated through the dry matter of the aerial part of inoculated alfalfa plants cultivated in vermiculite with N-free Jensen medium. The average dry weight for each treatment was obtained using 10 individual plants (mean±S.D. are shown). Statistical analysis was performed using analysis of variance as previously described.
3Results and discussion
3.1Construction of the pMP6, a GFP-Tc derivative pSM10, and its integration into the S. meliloti chromosome
Our first goal was the construction of a shuttle vector for the stable integration of a gfp marker into a neutral site on the S. meliloti chromosome. The vector pMP6 was constructed as outlined in Fig. 1. Using the basic vector pSM10 which carries partial sequences of the chromosomal S. meliloti recA (3′-end) and alaS genes (5′-end), we cloned a variant of gfp and a DNA cassette for Tc resistance (Tcr). The modified gfp-referred to as gfp P64L/S65T-contains the mutations F64L and S65T associated with an increased GFP solubility and a red-shift in the fluorescence [20,21]. As indicated in Fig. 1 the integration vector pSM10 was first digested with Bam HI and the cohesive ends were filled-in with Klenow fragment enzyme. The digested vector was then ligated to an approximately 2.0-kb Sma I DNA fragment carrying the Tcr gene from vector pHP45Ω-Tc. The resultant plasmid designated pMP5 was finally digested with Eco RI and ligated to gfp-P64L/S65T (760 bp) from pGreenTIR to yield plasmid pMP6. For integration of the transgenic markers into the S. meliloti chromosome plasmid pMP6 was transferred from E. coli S17–1 to S. meliloti 2011 by conjugation, and transconjugants that had presumable undergone the marker exchange (non-repetitive sequences bordering the transgenes) were identified by their expected Tcr (from the omegon-cassette), Smr (marker of the recipient bacteria), and Nms (after plasmid segregation) phenotype. The genomic structure was confirmed by Southern transfer analysis.
3.2Stability of the fluorescent phenotype and Tcr of S. meliloti 20MP6
The stability of the markers was analysed after extensive cultivation of strain S. meliloti 20MP6 in TY medium without any selective pressure (no antibiotics were added). The Tcr and the green fluorescence under UV illumination were analysed over the course of four serial batch cultures that spanned approximately 40 rhizobial generations. The frequency of Tc-resistant CFU was assessed by plating appropriate dilutions of the cultures in TY Petri dishes with and without the antibiotic. As shown in Fig. 2, the Tc-resistance phenotype was maintained during the whole period of analysis showing a very high stability of the marker. All colonies were also fluorescent under UV light both in the presence and absence of the antibiotic. The results showed a remarkable stability for the chromosomal construct under the experimental conditions used.
3.3Phenotypic expression of fluorescence by S. meliloti 20MP6 at difference stages of symbiosis
The early colonisation of alfalfa roots and their further infection by S. meliloti 20MP6 were analysed using fluorescence and phase contrast microscopy (Fig. 3). Bacteria could be seen adsorbed and embedded in the mucigel surrounding the root tip (Fig. 3A, B). Infection threads could also be easily detected (Fig. 3C, D). No loss of fluorescence was detected during root invasion by S. meliloti 20MP6 as infection threads or within the developing nodules (Fig. 3E, F). Results showed a continuous phenotypic expression of the GFP, from outside the root to the infected plant tissues at different stages of symbiosis. Clearly, the single chromosomal copy of the gfp P64L/S65T under the control of the strong nptII promoter was sufficient for the detection of free-living or symbiotic rhizobia. Single fluorescent free-living rhizobia could be detected when a bacterial suspension of strain 20MP6 was observed under UV light with 1000× magnification (not shown).
3.4Symbiotic properties of S. meliloti 20MP6: nodulation, competitiveness and nitrogen fixation
Previous reports indicated that the DNA at the intergenic region between the S. meliloti genes recA and alaS can be modified in different ways without any detectable changes in the symbiotic properties of the resultant recombinant rhizobia. Here, we evaluated if the gfp-P64L/S65T and the Tc-resistant marker had any effects on the symbiotic behaviour of strain 20MP6. Results from Fig. 4A showed that a similar number of root nodules on alfalfa plants was induced by strain 20MP6 and by the parental wild-type 2011. Furthermore, as is shown in Fig. 4B the aerial dry weight of alfalfa plants inoculated with S. meliloti 20MP6 was statistically comparable to the dry weight of alfalfa plants inoculated with strain 2011. Because in the growing conditions used no N-sources other than atmospheric N2 were present, dry matter accumulation was directly related to N2-fixation. The results indicated that the chromosomal insertion of the new markers did not significantly affect either nodulation or nitrogen fixation in strain 20MP6. To compare in more detail the nodulation capabilities of parental and recombinant rhizobia we performed a co-inoculation assay (106 CFU of each rhizobia/root 1:1 inoculant ratio of strains 20MP6:2011) and we evaluated root nodule occupancy. From 67 nodules analysed the ratio of fluorescent nodules: non-fluorescent nodules was 0.49:0.51 showing a similar level of competitiveness for strains 20MP6 and 2011. Fluorescent bacteria recovered from nodules were all resistant to Tc. This result suggests that gfp-P64L/S65T and the Tcr cassette are both stable in symbiosis in agreement with their observed stability in culture medium (Fig. 2).
Current studies on molecular bacterial ecology have an increasing demand for the specific detection of particular genotypes in complex and biodiverse environments. We developed in this work a gfp-P64L/S65T-Tc-based labelling system for the specific tagging of the soil bacteria S. meliloti. Although previous work used plasmid-borne versions of the gfp marker in S. meliloti, two main problems arise from the use of broad host range and transmissible replicons to track strains in the soil: the loss of the markers due to plasmid instability, and vector dispersion via conjugal transfer mechanisms. Unfortunately, the extent to which such events occur in a given soil sample are difficult to predict. For this reasons the use of chromosomally integrated markers represent a more suitable choice for those strains that have to be studied in non-sterile and biologically complex systems. In this work, the gfp-P64L/S65T marker was stable over time with appropriate phenotypic expression at different stages of the life cycle of the bacteria. Interestingly, although gfp-P64L/S65T was introduced as a single copy within the rhizobial chromosome, the expression level of GFP supported the detection of single fluorescent bacteria. It was also shown that the recombinant rhizobia did not show any detectable change in symbiosis when compared to parental rhizobia. The delivery plasmid pMP6 constructed here represents a suitable and reliable tool for the fluorescent labelling and detection of specific S. meliloti genotypes in soil.
The research was supported by grants PEI98-377, PIP 2000-2681 (CONICET), PICT 2000-08-09834 (ANPCyT-FONCyT), and FONTAGRO-BID-FTG/RF-01-03-RG. A.L. and D.F.H. are members of the Research Career of CONICET and CICBA, respectively. M.F.D.P. was supported by a postdoctoral fellowship of CONICET. We are grateful to Alfred Pühler and Werner Selbitschka from the University of Bielefeld (Germany) for providing plasmid pSM10.