Evidence for calcium-mediated perception of plant symbiotic signals in aequorin-expressing Mesorhizobium loti

Construction of an inducible reporter system for Ca²⁺measurements in rhizobia

The apoaequorin gene was cloned in the broad host-range expression vector pDB1 [22] under the control of the strong synthetic promoter P_syn, regulated by the lacI^q repressor (see Additional file 1). The pAEQ80 plasmid was mobilized by conjugation into the type strain of M. loti (USDA 3147^T).

Validation of the experimental system

The functioning in M. loti of the pAEQ80 plasmid containing the apoaequorin gene was verified by evaluating the level of aequorin expression in an in vitro reconstitution assay. Light emitted by total soluble protein contained in the lysates from wild-type and aequorin-expressing M. loti cells was monitored after reconstitution of the apoprotein with coelenterazine. The strong luminescence signal detected in protein extracts from M. loti cells containing the apoaequorin construct and induced with IPTG confirmed the efficient level of aequorin expression (see Additional file 2).

We analysed whether the introduced pAEQ80 plasmid (10.5 kb) encoding apoaequorin or the expressed protein could affect bacterial cell growth and the symbiotic performance of M. loti cells. There is no significant effect on bacterial growth kinetics exerted either by the introduced plasmid or apoaequorin expression. Nodulation efficiency of M. loti pAEQ80 cells on the specific plant host Lotus japonicus was checked 4 weeks after bacterial inoculation on roots of seedlings grown on nitrogen-free medium. L. japonicus roots were found to be effectively nodulated by the transformed bacterial strain, with no differences in nodule number (5 ± 1) and morphological parameters in comparison to seedlings inoculated with wild-type M. loti. The presence of bacteria inside nodules was verified by light microscopy (see Additional file 2). Green foliage was indicative of functional symbiosis.

The occurrence in M. loti cells of homeostatic control of the internal Ca²⁺ activity was then verified by preliminary Ca²⁺ measurement assays in a luminometer after in vivo reconstitution of apoaequorin. Unperturbed exponentially growing rhizobial cells showed a steady-state intracellular free Ca²⁺ concentration ([Ca²⁺]_i) residing in the submicromolar range (around 500 nM) (see Additional file 2), demonstrating a tight regulation of [Ca²⁺]_i. No luminescence was detected either in cultures of the non-recombinant strain incubated with coelenterazine or in recombinant cells that had not been exposed to coelenterazine (data not shown), confirming that the recorded signal was due only to Ca²⁺-dependent light emission from aequorin.

Environmental stimuli are sensed through transient [Ca²⁺]_i elevations by M. loti

To further validate the experimental system, abiotic stimuli which are known to trigger [Ca²⁺]_i changes in both plants [23] and cyanobacteria [18, 19] were applied to apoaequorin-expressing M. loti cells. A mechanical perturbation, simulated by the injection of isoosmotic cell culture medium, resulted in a rapid Ca²⁺ transient increase (1.08 ± 0.24 μM) that decayed within 30 sec (Fig. 1A). This Ca²⁺ trace, which is frequently referred to as a "touch response", is often observed after the hand-operated injection of any stimulus [24]. A similar Ca²⁺ response characterized by an enhanced Ca²⁺ peak of 2.14 ± 0.46 μM was triggered by a simple injection of air into the cell suspension with a needle (Fig. 1A).

Cold and hypoosmotic shocks, caused by supplying three volumes of ice-cold medium and distilled water, respectively, induced Ca²⁺ traces with distinct kinetics, e.g. different height of the Ca²⁺ peak (1.36 ± 0.13 μM and 4.41 ± 0.51 μM, respectively) and rate of dissipation of the Ca²⁺ signal (Fig. 1B and 1C). As a control, cells were stimulated with three volumes of growth medium at room temperature, (Fig. 1B) resulting in a Ca²⁺ trace superimposable on that of the touch response (Fig. 1A). These findings eliminate the possible effect of bacterial dilution on changes in Ca²⁺ homeostasis.

Challenge of M. loti with a salinity stress, which has recently been shown to affect symbiosis-related events in Rhizobium tropici [25], resulted in a [Ca²⁺]_i elevation of large amplitude (3.36 ± 0.24 μM) and a specific signature (Fig. 1C).

Variations in the extracellular Ca²⁺ concentration determined the induction of transient Ca²⁺ elevations whose magnitude was dependent on the level of external Ca²⁺. After a rapidly induced increase in [Ca²⁺]_i, the basal Ca²⁺ level was gradually restored with all the applied external Ca²⁺ concentrations (Fig. 1D), confirming a tight internal homeostatic Ca²⁺ control, as previously shown for other bacteria [14, 18].

All the above results indicate that aequorin-expressing M. loti cells comprise a functionally valid system with which to investigate the involvement of Ca²⁺ in intracellular transduction of environmental stimuli.

Host plant root exudates induce in M. loti a Ca²⁺signal required for activation of nodulation genes

Root exudates from the symbiotically compatible legume L. japonicus were collected from 3-week-old seedlings axenically grown in water and applied to M. loti cells. The dose used for Ca²⁺ measurements was in the range that induced significant expression of nodA, nodB, nodC genes in M. loti (Fig. 2A). This concentration was found to trigger a transient [Ca²⁺]_i change characterized by a very rapid increase (1.38 ± 0.23 μM Ca²⁺) followed by a second sustained major Ca²⁺ peak (2.01 ± 0.24 μM) at about 10 min (Fig. 2B), with a slow decay within the considered time interval (30 min). The observed induction of transient [Ca²⁺]_i changes in M. loti cells suggests a Ca²⁺-mediated perception of signalling molecules contained in host plant root exudates.

Flavonoids are components of root exudates that play a prominent role as inducers of structural nod genes in rhizobia. Although flavonoids have been detected in L. japonicus seeds [26], those that specifically activate the expression of nod genes in M. loti have not yet been identified [27, 28]. The most common flavonoids, known as nod gene inducers in other rhizobia (10 μM naringenin, luteolin, daidzein, kaempferol, quercetin dehydrate) were not able to trigger transient Ca²⁺ elevations in M. loti (data not shown). Tetronic acid, an aldonic acid previously reported to promote Nod factor biosynthesis in M. loti [29], was found to induce a detectable Ca²⁺ response (Fig. 2B). The kinetics of the Ca²⁺ trace was similar to that induced by crude root exudates, with a prompt Ca²⁺ spike (1.36 ± 0.16 μM Ca²⁺) and a subsequent flattened dome (maximal Ca²⁺ value of 1.29 ± 0.08 μM reached around 15 min after the elicitor application). Notably, this second phase of the Ca²⁺ transient induced by tetronic acid only partially accounted for the larger Ca²⁺ increase recorded with the whole L. japonicus root exudates (Fig. 2B). Likewise, the level of nod gene expression induced by tetronic acid was found to be lower (though significantly different from the control, P < 0.05) than that generated by total root exudates (Fig. 2A).

Pretreatment of rhizobial cells with the extracellular Ca²⁺ chelator EGTA for 10 min effectively inhibited both the transient Ca²⁺ elevation (Fig. 3A) and nod gene activation (Fig. 3B) induced by L. japonicus root exudates. This indicates that the main source of the observed Ca²⁺ response is the extracellular medium, and that the elevation in [Ca²⁺]_i is required for nod gene induction. Cell viability, monitored by the BacLight Bacterial viability assay, was not altered by incubation with the Ca²⁺ chelator (Fig. 3C). The expression of both constitutive (glutamine synthetase II and 16S rRNA) and inducible (aequorin) genes was not significantly affected by EGTA treatment (Fig. 3D and 3E), ruling out possible general effects of extracellular Ca²⁺ chelation on gene induction.

To check host specificity of the Ca²⁺ signal, metabolite mixtures exuded by the non-host legumes soybean and Vicia sativa subsp. nigra were tested. After an initial rapid and steep Ca²⁺ rise (1.77 ± 0.34 μM), shared also by the response to L. japonicus root exudates, the Ca²⁺ transients triggered by non-host exudates show very different kinetics, such as a slow rate of decay of the Ca²⁺ level (Fig. 4A versus Fig. 2B). Pretreatment with EGTA also blocked these transient Ca²⁺ elevations (data not shown). The distinct Ca²⁺ signature activated by non-host legumes, together with the lack of activation of nod genes (Fig. 4B), suggests the possibility of Ca²⁺-mediated perception by M. loti of molecules other than nod gene inducers, such as non-specific chemoattractants or other signalling molecules, e.g. proteins [30, 31] or plant cell wall fragments released during the detachment of border cells from the root tip [32], activating a different Ca²⁺ signalling pathway. Further confirmation of the specificity of the host plant-induced Ca²⁺ signalling comes from the complete absence of any detectable Ca²⁺ change and nod gene transcriptional activation by root exudates from a non-legume (tomato) (Fig. 4A and 4B).

Chemicals

Native coelenterazine was purchased from Molecular Probes (Leiden, The Netherlands). Molecular biology reagents were purchased from Promega Co. (Madison, WI, USA), Qiagen (Hilden, Germany) Clontech (Mountain View, CA, USA) and Invitrogen (Paisley, UK). Tetronic acid was obtained from Titolchimica (Rovigo, Italy). Flavonoids (naringenin, luteolin, daidzein, quercetin dehydrate) and all other reagents were obtained from Sigma-Aldrich (St. Louis, MO, USA).

Bacterial strains and growth conditions

Mesorhizobium loti strain USDA 3147^T was kindly provided by Peter Van Berkum (USDA, Beltsville MD) and was grown in minimal BIII medium [39] with or without 30 μg/ml kanamycin, as appropriate, at 28°C with shaking (170 rpm). E. coli was grown in LB medium at 37°C.

Cloning of the apoaequorin gene and introduction into M. loti

The terms aequorin and apoaequorin refer to the bioluminescent protein with and without, respectively, the prostethic group coelenterazine. The apoaequorin cassette, given by the apoaequorin cDNA fused to the first 27 nucleotides encoding hemoagglutinin (HA1-AEQ) [40] was amplified by PCR with primers designed to obtain a 5' XbaI site and to leave out the ATG start codon, already present into the P_syn promoter of the expression vector pDB1 [22]. The correct translation frame was maintained by adding a nucleotide between the 5' XbaI site and the apoaequorin gene. The primers used to obtain the apoaequorin cassette were: 5'-CCTACTCTAGATAAGCTTTATGATGTTCCT-3'and 5'TGATAGCATGCGAATTCATCAGTGTTTTAT-3'. PCR was run with the following parameters: 5 min at 94°C as start step; 30 s at 94°C, 30 s at 58°C, 1 s at 72°C for 30 cycle and 5 s at 72°C as a final step using PLATINUM^® Taq DNA polymerase (Invitrogen). To obtain a 3' XbaI site, the amplicon was then cloned into the pCR 2.1 plasmid by using TA Cloning^® technology (Invitrogen), originating p2.1AEQ. Digestion with XbaI of this intermediate plasmid released the HA1-AEQ coding region, which was then ligated into the XbaI site of pDB1 under the control of the strong isopropylβ-D-thiogalactoside (IPTG)-inducible synthetic promoter P_syn. The apoaequorin gene containing construct (pAEQ80, see Additional file 1) was mobilized to M. loti 3147^T from E. coli by triparental conjugation using plasmid pRK2013 as helper [41]. Transconjugants were selected on BIII agar containing 50 μg/ml kanamycin.

Growth kinetics of the recombinant strain

To determine the effect of the plasmid presence and of apoaequorin expression on bacterial cell growth, M. loti wild-type or containing pAEQ80 (plus or minus IPTG) were grown in 30 ml of BIII medium (supplemented or not with 30 μg/ml kanamycin, as appropriate) as described above. Growth was determined by monitoring turbidity at 600 nm.

In vitro L. japonicusnodulation tests

In vitro nodulation studies were carried out as described by [42]. Briefly, seeds of L. japonicus B-129 GIFU were transferred after sterilization on 0.1% Jensen medium solidified with 1% agar. Inoculation with bacterial suspensions of M. loti wild-type or containing pAEQ80 (5·10⁷ cells/root) was carried out 4 days after seed germination. Lotus seedlings, before and after infection, were grown at 24°C with 16 h light and 8 h dark. Growth and nodulation pattern were monitored for 4 weeks after inoculation. Microscopy observations were carried out with a Leica MZ16 stereomicroscope equipped with a DFC 480 photocamera. To check the actual occurrence of bacteria inside the nodules, they were squeezed and the content stained with 5 μg/ml 4',6-diamino-2-phenylindole (DAPI). Samples were observed with a Leica DMR fluorescence microscope. Images were acquired with a Leica IM500 digital camera.

Expression of apoaequorin

A loopful of M. loti USDA 3147^T pAEQ80 grown on BIII plates was used to inoculate 30 ml of BIII medium supplemented with 30 μg/ml kanamycin and 1 mM IPTG and grown at 28°C overnight, until an absorbance at 600 nm of approximately 0.25 was reached (after about 18 h).

In vitroreconstitution of apoaequorin to aequorin

M. loti suspension cultures (300 ml) were grown to mid-exponential phase (A_{600 nm} = 0.25), pelletted by centrifugation at 3000 g for 10 min at 4°C, washed twice with fresh medium, and finally resuspended in 2 ml reconstitution buffer (Tris-HCl 150 mM, EGTA 4 mM, supplemented with 0.8 mM phenylmethylsulfonyl fluoride, pH 8.0). Bacteria were lysed by 3 cycles (30 s each) of sonication at 35 Hz (Fisher Sonic, Artek Farmingdale, NY, USA), each followed by 30 s on ice. Non lysed bacteria were pelletted and discarded by centrifugation (1600 g for 15 min at 4°C). Protein concentration in the supernatant was estimated using the Bio-Rad (Hercules, CA) protein assay according to manufacturer's instructions. Total soluble proteins were resuspended at 1 μg/μl in reconstitution buffer and incubated with 1 mM β-mercaptoethanol and 5 μM coelenterazine for 4 h in the dark at 4°C. Aequorin luminescence was detected from 50 μl of the in vitro aequorin reconstitution mixture, containing 25 μg of total soluble protein diluted 1:2 with the same buffer and integrated for a 200 s time interval after the addition of an equal volume of 100 mM CaCl₂.

In vivoreconstitution of apoaequorin to aequorin

Mid-exponential phase cells (30 ml) were harvested by centrifugation at 2300 g for 15 min at room temperature and the cell pellet was washed twice in 5 ml BIII medium with intermediate centrifugation as described above. Cells were then incubated in BIII medium containing 5 μM coelenterazine in the dark for 1 h 30 min under shaking. After two washes as above, cells were resuspended in BIII medium and allowed to recover for 10 min prior to Ca²⁺ measurement experiments.

Root exudate production

Seeds of Lotus japonicus GIFU ecotype, soybean, Vicia sativa subsp. nigra and tomato were surface sterilized and allowed to germinate for three days on moistened filter paper at 24°C in the dark. Subsequently, seedlings were transferred aseptically on polystyrene grids covered with nylon meshes in sterile plastic containers containing different volumes of sterile H₂O, depending on the seed and seedling size (on average 5 ml of H₂O per seedling). After 3 weeks of germination crude root exudates were collected, filtered and lyophilized. The pellet was resuspended in BIII medium (50 μl per single root exudate) for cell treatments.

Ca²⁺measurements with recombinant aequorin

Aequorin light emission was measured in a purpose-built luminometer. Bacteria (50 μl) were placed, after aequorin reconstitution, in the luminometer chamber in close proximity to a low-noise photomultiplier, with a built-in amplifier discriminator. The output of the discriminator was captured by a THORN-EMI photon counting board (Electron Tubes Limited, Middlesex, UK) and the luminescence data were converted off-line into Ca²⁺ concentration values by using a computer algorithm based on the Ca²⁺ response curve of aequorin [40]. All stimuli were administered to cells by using a light-tight syringe through the luminometer port. The experiments were terminated by lysing the cells with 15% ethanol in a Ca²⁺-rich solution (0.5 M CaCl₂ in H₂O) to discharge the remaining aequorin pool. For experiments performed in the presence of different external Ca²⁺ concentrations, cells were extensively washed and resuspended in buffer A (25 mM Hepes, 125 mM NaCl, 1 mM MgCl₂, pH 7.5), as described by [16]. When needed, cells were pretreated for 10 min with 5 mM EGTA.

Bacterial cell viability assay

Bacterial cell viability was monitored by the LIVE/DEAD^® BacLight™ Bacterial Viability kit (Molecular Probes), according to manufacturer's instructions. This fluorescence-based assay use a mixture of SYTO 9 and propidium iodide stains to distinguish live and dead bacteria. Bacteria with intact cell membranes stain fluorescent green, whereas bacteria with damaged membranes stain fluorescent red. Samples were observed with a Leica 5000B fluorescence microscope. Images were acquired with a Leica 300F digital camera using the Leica Application Suite (LAS) software.

Semi-quantitative RT-PCR experiments

M. loti cells grown to mid-exponential phase and treated as for Ca²⁺ measurement experiments (see above) were incubated for 1 h with plant root exudates, tetronic acid or cell culture medium only (as control). To stabilize RNA, bacteria were treated with the RNA protect Bacteria Reagent (Qiagen). Bacterial cell wall was then lysed with 1 μg/ml lysozyme (Sigma) in TE buffer. Total RNA was first extracted using RNeasy Mini kit (Qiagen) and, after DNAse I treatment (Promega), quantified. RNA (5 μg) was primed with Random Decamers (Ambion), reverse transcribed with PowerScript Reverse Transcriptase (Clontech) and diluted 1:5. 5 μl of diluted first-strand cDNA were used as a template in a 50 μl PCR reaction solution. Reverse transcription (RT)-PCR was performed with 5 μl diluted first-strand cDNA. The oligonucleotide primers were designed against nodA, nodB, nodC and glutamine synthetase II (GSII) sequences from M. loti [43] and the aequorin gene (aeq) from Aequorea victoria [44], using Primer 3 software. To amplify 16S rRNA gene, Y1 and Y2 primers were used [45].

The thermal cycler was programmed with the following parameters: 20 s at 94°C, 30 s at 68°C and Advantage 2 Polymerase mix (Clontech) was used as Taq polymerase. PCR reactions were allowed to proceed for different number of cycles to determine the exponential phase of amplification. Densitometric analysis of ethidium bromide-stained agarose gels (0.5 μg/ml) was performed using QuantityOne software (Bio-Rad). RT-PCR experiments were conducted in triplicate on three independent experiments. The primer sequences used to obtain amplicons were: 5'-TATGAGCCGACCGGAGCCTTTAAT-3' and 5'-CCGTATAGACCGAGTTCAGCGACAA-3' for nodA, 5'-ATACTCGATGTGCTGGCGCAAAAT-3' and 5'-GCCTGGTTCGCCTCAAATACTTCAC-3' for nodB, 5'-CCACCTTACGATCCTGATGCTGAAA-3' and 5'-CAATATTCTGGCCAATCACGTCCAA-3' for nodC, 5'-ACCGAGACTTACGGCATCGACATC-3' and 5'-GCGACGCCATAGCTAAACTTGTTCC-3' for GSII, 5'-TAACCTTGGAGCAACACCTGAGCAA-3'

5'-ATACGGATGAGCGTTGGTTCGTTTT-3'for aequorin, Y1 (5'-TGGCTCAGAACGAACGCTGGCGGC-3') and Y2 (5'-CCCACTGCTGCCTCCCGTAGGAGT-3') for 16S rRNA. Amplicons were sequenced by BMR Genomics (Padova, Italy).