Increased plant tolerance against chrysanthemum yellows phytoplasma (‘Candidatus Phytoplasma asteris’) following double inoculation with Glomus mosseae BEG12 and Pseudomonas putida S1Pf1Rif

Authors

  • R. D’Amelio,

    1. CNR, Istituto di Virologia Vegetale, Strada delle Cacce 73, Torino 10135
    2. Università di Torino, Dipartimento di Valorizzazione e Protezione delle Risorse Agroforestali, Via Leonardo Da Vinci 44, Grugliasco (Torino) 10095
    Search for more papers by this author
  • G. Berta,

    1. Università del Piemonte Orientale ‘Amedeo Avogadro’, Dipartimento di Scienze dell’Ambiente e della Vita, Viale Teresa Michel 11, Alessandria 15121, Italy
    Search for more papers by this author
  • E. Gamalero,

    1. Università del Piemonte Orientale ‘Amedeo Avogadro’, Dipartimento di Scienze dell’Ambiente e della Vita, Viale Teresa Michel 11, Alessandria 15121, Italy
    Search for more papers by this author
  • N. Massa,

    1. Università del Piemonte Orientale ‘Amedeo Avogadro’, Dipartimento di Scienze dell’Ambiente e della Vita, Viale Teresa Michel 11, Alessandria 15121, Italy
    Search for more papers by this author
  • L. Avidano,

    1. Università del Piemonte Orientale ‘Amedeo Avogadro’, Dipartimento di Scienze dell’Ambiente e della Vita, Viale Teresa Michel 11, Alessandria 15121, Italy
    Search for more papers by this author
  • S. Cantamessa,

    1. Università del Piemonte Orientale ‘Amedeo Avogadro’, Dipartimento di Scienze dell’Ambiente e della Vita, Viale Teresa Michel 11, Alessandria 15121, Italy
    Search for more papers by this author
  • G. D’Agostino,

    1. Università del Piemonte Orientale ‘Amedeo Avogadro’, Dipartimento di Scienze dell’Ambiente e della Vita, Viale Teresa Michel 11, Alessandria 15121, Italy
    Search for more papers by this author
  • D. Bosco,

    1. Università di Torino, Dipartimento di Valorizzazione e Protezione delle Risorse Agroforestali, Via Leonardo Da Vinci 44, Grugliasco (Torino) 10095
    Search for more papers by this author
  • C. Marzachì

    Corresponding author
    1. CNR, Istituto di Virologia Vegetale, Strada delle Cacce 73, Torino 10135
    Search for more papers by this author

E-mail address: c.marzachi@ivv.cnr.it

Abstract

The aim of this work was to assess the effects of a combined inoculum of a rhizobacterium and an arbuscular mycorrhizal (AM) fungus on plant responses to phytoplasma infection, and on phytoplasma multiplication and viability in Chrysanthemum carinatum plants infected by chrysanthemum yellows phytoplasma (CY). Combined inoculation with Glomus mosseae BEG12 and Pseudomonas putida S1Pf1Rif resulted in some resistance to phytoplasma infection (about 30%), delayed symptom expression in nonresistant plants, improved growth of the aerial part of the infected plants (+68·1%), and altered root morphology (root tip number: +49·9%; branching degree: +82·8%). Combined inoculation with the two beneficial microorganisms did not alter CY multiplication and viability. In inoculated and infected plants, phytoplasma morphology was typical of senescent cells. A more active and efficient root system in double-inoculated plants probably mediated the effects of the two rhizospheric microorganisms in the infected plants. The practical application of rhizospheric microorganisms for mitigating phytoplasma damage, following evaluation under field conditions, represents an additional tool for the integrated management of phytoplasmosis.

Introduction

Phytoplasmas are wall-less, nonhelical prokaryotes that colonize plant phloem and insects. They are associated with more than 600 diseases, some of great economic importance, in several hundred plant species (Seemüller et al., 1998), with symptoms such as witches’ broom, proliferation, flower malformation (virescence, phyllody), stunting and general decline.

Phytoplasmas are transmitted by phloem-sucking insects, mostly in the families Cicadellidae, Cixiidae and Psyllidae, in a persistent, propagative manner (Weintraub & Beanland, 2006). Control of phytoplasma diseases is mainly achieved by insecticide treatments against the vectors, with deleterious impacts on the environment. In fruit trees, where economic losses may be severe, breeding for resistant varieties and hot water treatment of propagation material are currently implemented in control strategies. The exploitation of the natural recovery of some fruit tree species, such as grapevine and apple, is also suggested by several research groups as a possible tool for mitigating phytoplasma impacts (Romanazzi et al., 2009), although the control of insect vector populations is required to prevent reinfections (Morone et al., 2007). In other cases, such as in the Bois noir infection of grapevine, alternative strategies for controlling the phytoplasma are even more important. In this case, since the known vector of the phytoplasma, Hyalesthes obsoletus (Maixner, 1994), does not breed on grapes, planting of healthy propagation material and management of weeds in the vineyard are the only available measures of controlling the disease.

The potential role of arbuscular mycorrhizal (AM) fungi and rhizosphere bacteria as defence elicitors has been described (van Loon et al., 1998; Akhtar & Siddiqui, 2008). Arbuscular mycorrhizal fungi form mutualistic associations with most terrestrial plants, including agricultural and horticultural crop species. They obtain carbohydrates from the host plant, and, in return, they assist the plant in the acquisition of mineral nutrients (mainly phosphorus) and water (Smith & Read, 2008), and they also modify root architecture and topology (Hodge et al., 2009). This symbiosis not only influences plant nutrition, but it also improves the plant’s ability to overcome biotic (Lingua et al., 2002; Berta et al., 2005) and abiotic (Volante et al., 2005; Gamalero et al., 2010a) stresses. Moreover, the synthesis of volatile compounds (Guerrieri et al., 2004) and the secondary metabolic pathways of plants (Copetta et al., 2006) are affected in AM plants.

Bacteria living on the roots may promote plant growth via a number of mechanisms, including hormone synthesis, phosphate solubilization, nitrogen fixation (Gamalero et al., 2008) and root architecture modifications (Gamalero et al., 2002, 2004, 2008). However, the beneficial effect of these bacteria also relies on the improvement of plant health through the suppression of soilborne diseases. In this context, the production of antibiotics, lytic enzymes and siderophores, associated with an increase in plant tolerance to environmental stresses, play a key role. Besides affecting plant growth and health, bacteria living in the rhizosphere can interact with AM fungi. Stimulation of the AM symbiosis development by rhizospheric bacteria has been reported (Ravnskov & Jakobsen, 1999; Gamalero et al., 2004, 2008; Pivato et al., 2009). Although knowledge of the mechanism involved in this positive plant/bacteria/AM fungi multitrophic interaction is still incomplete, synthesis of the enzyme 1-aminocyclopropane-1-carboxylate (ACC) deaminase, which lowers the levels of ethylene in plants, seems to be involved (Gamalero et al., 2008). On the other hand, bacteria and AM fungi inoculated together can induce synergistic effects on plants (Gamalero et al., 2008).

Increased tolerance against phytoplasmas has been demonstrated in plants inoculated with AM fungi (Lingua et al., 2002; García-Chapa et al., 2004; Kamińska et al., 2010) or with bacteria colonizing plant roots (Gamalero et al., 2010c). The possible synergistic effect on phytoplasma infection of concurrent inoculation with AM fungi and beneficial rhizospheric bacteria has never been addressed. Therefore, the aim of this work was to assess the effects of a mixed inoculum, composed of a rhizobacterium and an AM fungus on plant responses to phytoplasma infection and on phytoplasma multiplication and viability in the model plant Chrysanthemum carinatum infected by ‘Candidatus Phytoplasma asteris’ chrysanthemum yellows (CY).

Materials and methods

Microorganisms

The strain Pseudomonas putida S1Pf1 was isolated from the soil next to a symptomless grapevine (Vitis vinifera cv. Barbera) among plants with clear symptoms of phytoplasma infection (D’Amelio et al., 2007). A spontaneous rifampicin-resistant mutant (S1Pf1Rif) was obtained, grown in Luria-Bertani (LB) broth supplemented with rifampicin (100 μg mL−1) and stored at −80°C in 50% glycerol. This bacterial strain, unable to produce ACC deaminase, was able to produce siderophores and auxin, to solubilize phosphate and to relieve symptoms in chrysanthemums infected by CY (Gamalero et al., 2010b).

The AM fungus Glomus mosseae BEG12 was previously shown to promote plant growth (Gamalero et al., 2002, 2004; Kamińska et al., 2010) and to suppress soilborne diseases caused by Phytophthora parasitica (Trotta et al., 1996), Rhizoctonia solani (Berta et al., 2005; Gamalero et al., 2010c) and phytoplasmas (Lingua et al., 2002). A mycorrhizal inoculum consisting of infected leek roots, sporocarp spores and hyphae of G. mosseae BEG12 on quartz sand was obtained from Agrauxine-Biorize.

Plant growth conditions and treatments

Chrysanthemum carinatum seeds were surface-sterilized by gently shaking in a 50-g L−1 sodium hypochlorite solution for 3 min. The seeds were rinsed six times for 5 min and four times for 20 min in sterile deionized water, placed in Petri dishes on moist sterile filter paper and incubated in the dark at 24°C for 3 days. The efficiency of the sterilization procedure was assessed by placing 20 C. carinatum seeds on nutrient agar (NA, Fluka) for 3 days. Chrysanthemum carinatum plants were grown from sterilized seeds in 8 × 8-cm pots containing a mixture of 0·6–1·2 mm coarse grade quartz sand (Punto Elle) sterilized at 200°C for 2 h. The plants were grown in a greenhouse (25°C, 16-h photoperiod) and watered to saturation with a modified Long Ashton nutrient solution containing 32 μm phosphate three times per week (Trotta et al., 1996).

Inoculation with a commercial inoculum of G. mosseae BEG12 was performed by potting C. carinatum seeds in the sterile mix of quartz and sand together with 100 g L−1 of commercial inoculum.

Pseudomonas putida S1Pf1Rif inoculants were produced on King’s B (KB) agar plates with 100 μg mL−1 rifampicin added at 28°C for 48 h. Bacteria were scraped from the medium, suspended in 0·1 m MgSO4, pelleted by centrifugation (4500 g, 20 min), washed twice and suspended in the same buffer. The bacterial density of the suspension was determined using a calibration curve assessed by measuring optical density at λ = 600 nm, and then adjusted to 1010 colony forming units mL−1 (CFU mL−1). The bacterial inoculum was applied 15 days after sowing by watering the germinated plants with 5 mL of a suspension (1010 CFU mL−1) of P. putida S1Pf1Rif in order to reach the final density of 108 CFU g−1 of sand.

The control plants were sown in the sterile mix of quartz and sand devoid of G. mosseae BEG12 inoculum and watered with 5 mL sterile 0·1 m MgSO4 instead of the bacterial inoculum.

There were four treatments and 10–20 replicates per treatment: control plants (C), CY-infected plants (CY), P. putida S1Pf1Rif- and G. mosseae BEG12-inoculated plants (Pp + Gm) and P. putida S1Pf1Rif- and G. mosseae BEG12-inoculated and CY-infected plants (Pp + Gm + CY). The experiment was repeated three times (a total of 42 and 49 non-inoculated and inoculated plants, respectively) and the results presented here are the mean values of the data obtained from these independent experiments.

Phytoplasma and vector insect

Candidatus Phytoplasma asteris’, strain CY, originally isolated from Argyranthemum frutescens plants in Liguria (Italy) (Conti et al., 1988), was maintained in C. carinatum by vector transmission.

Healthy colonies of Macrosteles quadripunctulatus were reared on potted oat plants inside plexiglass and nylon cages in growth chambers at 25°C, with a 16-h photoperiod, and checked by polymerase chain reaction (PCR) assays to verify the absence of phytoplasma.

For the transmission experiments, third to fifth instar nymphs were fed for 1 week on CY-infected plants, transferred to healthy oat plants for 2 weeks to complete latency, and then singly transferred to 10/15 C. carinatum for each elicitor and control treatment for an inoculation access period (IAP) of 3 days inside glass cylinders. Inoculation with single vectors was performed in order to reduce the concentration of the initial inoculum in the plant. Test plants were exposed to vectors 2 months after sowing. The plants were then treated with insecticides and maintained in the greenhouse for 1 month. Chrysanthemum carinatum exposed to vectors were used as transmission controls and others inoculated with P. putida S1Pf1Rif and G. mosseae BEG12 but not exposed to vectors were used as treatment controls.

Evaluation of plant development and root architecture

Ten plants per treatment were collected 3 months after sowing at 30 days post-inoculation (dpi) and the following parameters were determined: plant, root and shoot fresh weights, root/shoot fresh weight ratio, total root length, total surface area, number of tips, and degree of root branching (represented by root tip number divided by total root length), by using a dedicated Desk Scan II scanner equipped with a special lighting system for root measurements. The digitalized root images were analysed by the Winrhizo software (Régent Instruments).

Bacterial density and mycorrhizal colonization

Bacterial density was assessed 2·5 months after sowing (at 30 dpi). Whole root systems were aseptically cut from five plants, gently washed with sterile water and vortexed for 15 min in MgSO4 buffer (0·1 m). The suspensions obtained were serially diluted and plated on solid KB supplemented with rifampicin (100 μg mL−1). After incubation for 48 h at 25°C, the numbers of CFU g−1 root fresh weight were determined.

The degree of mycorrhizal colonization within the root systems was evaluated according to Trouvelot et al. (1986). At least 30 randomly chosen, 1-cm-long root fragments were cleared for 20 min at 60°C in 100 g L−1 KOH and stained with 10 g L−1 methyl blue in lactic acid. Root fragments were mounted on glass slides and examined microscopically, with magnifications up to ×630. The mycorrhizal frequency (F%) was calculated as the ratio between root fragments colonized by AM fungal mycelium and the total number of root fragments analysed. The mycorrhizal colonization intensity (M%) was evaluated as the amount of root cortex colonized by the fungus relative to the whole root system. Arbuscule and vesicle richness in the whole root system were measured by arbuscule (A%) and vesicle (V%) abundance.

Symptom evaluation

The presence and severity of CY symptoms on the plants were evaluated every 3 days starting from 9 up until 30 days after the beginning of inoculation (dpi), and plants were classified into five classes of severity: 0 = no symptoms, 1 = yellowing of the apex, 2 = yellowing and distortion of the apex, 3 = stunted apex growth, 4 = severe yellowing and dwarfing of the whole plant, 5 = plant death.

Phytoplasma concentration: DNA extraction and quantitative real-time PCR

Samples of 200 mg were collected from the apical leaves of five plants per treatment at 17 and 24 dpi, and total DNA was extracted from 100 mg using the PureLink Plant Total DNA Purification kit (Invitrogen), according to the manufacturer’s protocol. The same plants were sampled on both dates and quantification was repeated twice (experimental repeats 1 and 2). The concentration of DNA in the extracts was measured using a NanoDrop 1000 Spectrophotometer (Thermo Fisher Scientific). Diluted samples (1 ng μL−1 in sterile double-distilled water) were analysed in triplicate by quantitative real-time PCR (Q-PCR) assays. The CY DNA in each sample was measured as the number of CY genome units (GU) ng−1 of plant DNA, as described by Marzachi & Bosco (2005). The DNA from healthy plants and the PCR mixture devoid of a template were used as negative controls. Threshold cycles and standard curves were automatically calculated by the BioRad iCycler software, version 3·06070. The CY and host DNA from the same sample were quantified in the same plate.

Phytoplasma viability: total RNA extraction and quantitative reverse transcription real-time PCR

The remaining 100 mg leaf tissue collected from the five plants per treatment of experimental repeats 2 and 3 for DNA quantification were ground to powder under liquid nitrogen. Total RNA, extracted with Trizol Reagent (Invitrogen) and treated with RQI RNase-free DNase (Promega), was loaded as the template in quantitative reverse transcription real-time PCR (qRT-PCR) assays as described by D’Amelio et al. (2010). In the control reactions, RNA samples were not supplemented with reverse transcriptase to rule out DNA contamination. The RNA from healthy plants and the PCR mixture devoid of the template were used as negative controls. All samples were run in triplicate.

The phytoplasma 16S rDNA transcript copy number per nanogram of total RNA of each sample was determined. The phytoplasma 16S rDNA transcript copy number per 100 mg fresh leaf tissue was then derived. The viability of CY (expressed as 16S rDNA transcripts per phytoplasma cell) was estimated by dividing the transcript copy number by the number of CY cells in 100 mg of the same sample (measured as detailed above).

Microscopic observations

Pieces of the basal, median and apical zones of mature leaves sampled along the entire stem of plants infected with CY or infected with CY and inoculated with P. putida S1Pf1R1 and G. mosseae BEG12 were fixed for 3 h at room temperature in 3% glutaraldehyde in 0·15 m phosphate buffer (pH 7·2) and post-fixed for 2 h at 4°C in 10 g L−1 osmium tetroxide in the same buffer. The samples were stained overnight at 4°C in 10 g L−1 aqueous uranyl acetate, dehydrated in an ethanol series and embedded in Epon-Araldite. Pieces of leaves from the untreated plants were processed in the same way as the controls. Ultra-thin sections cut with a diamond knife in a Leika Ultracut UCT ultramicrotome (Leica Mikrosysteme Gmbh) were stained with lead citrate and examined under a Philips CM 10 transmission electron microscope. About 15 sections were observed for each treatment.

Statistical analysis

The data on plant growth and root architecture obtained from 10 plants per treatment were expressed as mean values. The data on mycorrhizal colonization were subjected to angular transformation before statistical analysis. Nontransformed and transformed values were submitted to analysis of variance and to Fisher’s least significant difference test ( 0·05) using the statview 5·0 statistics package (SAS Institute Inc.).

In order to analyse symptom development in control and elicited plants, the median number of days required by each plant to reach each symptom severity class was calculated and then compared using the non-parametric Mann–Whitney U Test using sigmaplot 11 (Systat Software Inc.).

Raw data (obtained as a log) of the phytoplasma titre were used and the concentration of CY phytoplasma in each plant was expressed as the difference between the logarithmic concentrations of CY and C. carinatum DNAs. In order to compare the phytoplasma titre measured at different dpi in elicited and control plants, a two-way anova for date and treatment was performed (sigmaplot 11·0; Systat Software Inc.).

Raw data of phytoplasma viability (CY phytoplasma transcripts/CY phytoplasma cells) were transformed into logarithms because the standard deviation appeared as a function of the mean. Two-way anova (treatment and date) was performed to compare the CY viability measured on the two sampling dates (17 and 24 dpi) in CY-infected plants inoculated or not with P. putida S1Pf1Rif and G.  mosseae BEG12 (sigmaplot 11·0; Systat Software Inc.).

Results

Bacterial density and mycorrhizal colonization

At the end of the experiments, the density of P. putida S1Pf1Rif was significantly higher in CY-infected plants than in the uninfected ones (5·05 ± 0·74 and 4·01 ± 0·13 log CFU g−1 root fresh weight, respectively; = 0·002), and the root systems were densely colonized by G. mosseae BEG12. Infection with CY did not significantly affect mycorrhizal frequency, colonization intensity or arbuscule abundance (not shown). However, vesicle abundance was significantly reduced (= 0·006) in CY-infected plants (9·7 ± 2·9) compared to the controls (21·7 ± 1·5). No mycorrhizal colonization was observed in control or phytoplasma-infected non-inoculated plants.

Effect of the microorganisms on plant growth, root architecture and symptom development

Plant growth was reduced by CY (−37·3%, Fig. 1a, = 0·0023). Inoculation with the bacterial strain and the AM fungus increased the biomass of infected plants compared to the controls (+60·5%, Fig. 1a, = 0·0305) and rescued the plant growth depression induced by CY (Fig. 1a, = 0·0024). In particular, root and shoot fresh weights in CY-infected plants were reduced by 39·4 and 34·8%, respectively, compared to the controls (Fig. 1b,c, = 0·0070 and P = 0·0146, respectively). The shoot fresh weight of CY-infected plants inoculated with P. putida S1Pf1Rif and the mycorrhizal fungus was significantly higher (< 0·0001) than that of CY-infected plants (+68·1%) (Fig. 1c). By contrast, inoculation with the two beneficial microorganisms did not rescue the CY-induced reduction of root fresh weight (Fig. 1b). The ratio between root and shoot fresh weights was reduced in plants inoculated with S1Pf1Rif and G. mosseae BEG12, infected (−45·1%, = 0·0113) or not with CY (−54·4%, = 0·0001) compared to the controls (Fig. 1d).

Figure 1.

 Effect of microorganisms on growth and root architecture of Chrysanthemum carinatum plants. Plant (a), root (b), shoot fresh weight (c) and root/shoot fresh weight ratio (d) of control plants, plants infected with chrysanthemum yellows (CY) and plants inoculated with Pseudomonas putida S1Pf1Rif and Glomus mosseae BEG12, infected or not infected with CY. Means with the same letter are not significantly different ( 0·05) according to Fisher’s least significant difference test.

All of the root architecture parameters were reduced in plants infected by CY. Inoculation with the strain S1Pf1Rif and the AM fungus in plants infected by the phytoplasma increased the root tip number (+49·9%, = 0·009) and the branching degree (+82·8%, < 0·001) compared to the non-inoculated and infected controls (Table 1).

Table 1.   Total root length (TL), total root surface area (TSA), number of tips (T) and root branching degree (T/TL) in Chrysanthemum carinatum plants inoculated or not with Pseudomonas putida S1Pf1Rif and Glomus mosseae BEG12 (C, Pp + Gm) and infected or not with CY (CY, Pp + Gm + CY)
 TLTSATT/TL
  1. Values are means of 10 plants ± standard errors. Within columns, means with the same letter are not significantly different (P ≤ 0·05) according to Fisher’s least significant difference test. The analyses were performed at the end of the experiments (30 days post-inoculation).

C4414·7 ± 348·9 a560·0 ± 35·0 a5974·8 ± 397·8 a1·4 ± 0·1 b
CY2472·9 ± 96·9 c335·1 ± 16·0 b2503·0 ± 186·3 c1·0 ± 0·1 c
Pp + Gm3371·3 ± 381·1 b526·5 ± 50·0 a3984·3 ± 293·3 b1·2 ± 0·1 bc
Pp + Gm + CY2041·7 ± 131·9 c264·9 ± 19·1 b3751·6 ± 296·1 b1·8 ± 0·2 a

Symptom development was evaluated every 3 days for a period of 30 days, starting from 9 dpi. Pseudomonas putida S1Pf1Rif and G. mosseae BEG12-inoculated and non-inoculated control plants did not develop any symptoms (data not shown). Following exposure to CY-infective leafhoppers, a higher number of elicited plants compared to non-inoculated ones did not develop symptoms (13/49 vs. 3/42; χ2 = 4·605; = 0·032).

The results from the three experimental repeats were pooled for analysis because a one-way anova on ranks did not reveal any differences between the repeats for both control and inoculated plants. The median numbers of days required by elicited plants to reach the different symptom severity classes were higher than for the control plants, although these differences were not significant (not shown).

Chrysanthemum yellows phytoplasma titre, viability and phytoplasma morphology

The phytoplasma titre and viability in CY-infected plants, inoculated or not with the fluorescent pseudomonad and the AM fungus, were measured on apical leaves at 17 and 24 dpi. Two-way anova for treatment and date showed that the CY titre increased significantly over time (= 0·002), whereas no differences were found between elicited and control (= 0·990) plants (Table 2). The interaction between the two factors (inoculation with the elicitors and sampling date) did not affect the CY titre. The numbers of 16S rDNA transcripts per CY cell (as an estimator of CY viability) in inoculated and CY-infected plants, as well as in the controls, are reported in Table 2. Two-way anova for treatment and date showed a highly significant reduction in phytoplasma viability between 17 and 24 dpi (< 0·001), whereas no differences were found between inoculated and control plants (=0·186). The CY viability was unaffected by the interaction between the two factors (inoculation with the elicitors and sampling date).

Table 2.   Mean chrysanthemum yellows (CY) phytoplasma titre (CY cells per nanogram plant DNA ± standard error) and viability (mean copy number of CY 16S rDNA transcripts per CY cell ± standard error) in CY-infected Chrysanthemum carinatum (CY) inoculated or not with Pseudomonas putida S1Pf1Rif and Glomus mosseae BEG12 (Pp + Gm + CY). Plants were sampled at 17 and 24 days post-inoculation (dpi)
Sampling date (dpi)CY cells/ng plant DNA (mean ± SE)CY 16S rDNA transcripts/CY cell
CYPp + Gm + CYCYPp + Gm + CY
  1. Within rows and columns of each experiment, means followed by the same letter do not differ significantly.

17146 500 ± 30 284 a149 973 ± 39 109 a4455 ± 946 A1207 ± 349 B
24324 000 ± 60 482 b521 714 ± 152 437 b1207 ± 349 B1726 ± 430 B

Phytoplasmas were never detected in the plants inoculated with both beneficial microorganisms or the control plants that were not exposed to infective insects (data not shown).

Microscopic observations detected differences in phytoplasma morphology according to plant treatment. No phytoplasma cells were detected in control plants or in those inoculated with the AM fungus and P. putida S1Pf1Rif. The deposition of callose in the pore of the sieve elements and the synthesis of PR protein in the host plant was found in CY-infected plants (Fig. 2a,b). In general, in these plants, the phytoplasmas were densely pressed into the phloem lumen in all of the observed samples. In CY-infected plants inoculated with G. mosseae BEG12 and the bacterial strain S1Pf1Rif (Fig. 2c), the phytoplasma morphology (empty cells and cells with the cytoplasm confined to the periphery) was typical of senescent cells, clearly indicating the occurrence of a strong degenerative process.

Figure 2.

 Phytoplasma morphology. Micrographs of Chrysanthemum carinatum leaves (a, b) infected with chrysanthemum yellows (CY) phytoplasma and (c) infected with CY and inoculated with Pseudomonas putida S1Pf1Rif and Glomus mosseae BEG12. In the phloem of CY-infected control plants, the CY cells are well preserved (a and b), bound by a well-defined unit membrane and contain a matrix of varying electron densities; copious formation of electron-transparent callose pinching off the pores is evident in the sieve plates (b, black arrows). In the phloem of plants infected with CY and inoculated with P. putida S1Pf1Rif and G. mosseae BEG12, few and degenerated forms of CY (c, black arrows) are evident. Bar = 1 μm.

Discussion

This work shows the effects of a combination of two rhizospheric microorganisms on phytoplasma infection of a model plant. Plant growth, root architecture, multiplication of the phytoplasma and its metabolic activity were measured in ‘Candidatus Phytoplasma asteris’ (chrysanthemum yellows strain, CY)-infected C. carinatum plants inoculated with P. putida S1Pf1Rif (Gamalero et al., 2010b) and the AM fungus G. mosseae BEG12, and compared to CY-infected but non-inoculated plants. Previous works demonstrated the activity of P. putida S1Pf1Rif on C. carinatum infected by CY (Gamalero et al., 2010b) and of G. mosseae BEG12 on tomato plants infected by the stolbur phytoplasma (Lingua et al., 2002) and Catharanthus roseus infected by a ‘Ca. Phytoplasma asteris’ strain (Kamińska et al., 2010). Nevertheless, this is the first report dealing with the activity of a combination of two plant-beneficial microorganisms on a phytoplasma-infected plant.

Inoculation with the two rhizospheric microorganisms resulted in a lower number of infected plants following exposure to infective leafhopper vectors. Protection of plants from CY infection was not observed in C. carinatum inoculated with P. putida S1Pf1Rif alone (Gamalero et al., 2010b) or elicited with benzothiadiazole (D’Amelio et al., 2010). In preliminary experiments, inoculation of C. carinatum with G. mosseae BEG12 alone resulted in a reduced number of CY-infected plants following exposure to infective hoppers (authors’ unpublished results). Therefore, it can be speculated that resistance to CY infection can be attributed to the activity of the mycorrhizal fungus. In a different phytoplasma–plant association, the inoculation of periwinkles with G. mosseae BEG12 alone did not affect the number of ‘Ca. Phytoplasma asteris’-infected plants (Kamińska et al., 2010); however, in that experimental setup, the phytoplasma was transmitted by grafting.

The CY infection induced a severe reduction in plant growth, both at the root and the shoot level. Inoculation with the AM fungus and the bacterial strain induced a significant increase in the fresh weight of the aerial part of the plant, and this effect was still present upon CY infection. According to the results presented by Gamalero et al. (2010b) the bacterial strain did not improve plant growth. Therefore, the increase of the shoot biomass in plants inoculated with G. mosseae BEG12 and P. putida S1Pf1Rif may be ascribed to the fungal partner or to the effects resulting from the interactions between the two beneficial microorganisms (Gamalero et al., 2004, 2008). Moreover, the reduction in the fresh weight root/shoot ratio in inoculated plants, irrespective of phytoplasma infection, suggests that the net benefit of the mycorrhizal association largely exceeded its net cost, as already reported by Fitter (1991). Similarly, G. intraradices mycorrhizal infection of pear trees was associated with a significant increase in plant shoot length in pear decline-infected and healthy controls (García-Chapa et al., 2004). The increase in the development of the aerial part of the plant was supported by a larger root system in the double-inoculated plants. Indeed, double-inoculated, CY-infected plants showed not only a larger root system but also a higher degree of root branching (Table 1, T/TL), thus allowing the exploration of a higher volume of soil than other treatments. This difference would account for the increased growth of the double-inoculated CY-infected plants, as measured by plant weight. The CY infection caused a significant reduction in the number of root tips (Table 1, T). This reduction was also present, although to a lesser extent, in the elicited but healthy controls, and the presence of the two elicitors resulted in an increase in the number of root tips in the infected plants. The CY infection also induced a severe reduction in the degree of root branching (Table 1, T/TL) in control plants, but roots of double-inoculated CY-infected plants were more branched than those in any of the other treatments. Treatment of C. carinatum plants with P. putida S1Pf1Rif alone did not rescue the reduced root branching associated with CY infection (Gamalero et al., 2010b), thus suggesting a relevant role of the AM fungus in the promotion of lateral root initiation. In fact, increased branching seems to be a more general effect induced by AM fungi (Berta et al., 1995, 2002, 2005; Hodge et al., 2009), leading to better nutrient absorption, mainly phosphate, from the soil.

Following inoculation with the fluorescent pseudomonad and the AM fungus, CY-infected plants showed a constant, although not significant, delay in symptom development compared to control plants. Although it is evident that symptom severity does not provide a direct measure of the agronomic losses associated with phytoplasma infection, it still represents a measure of the impact of the pathogen on plant health.

Phytoplasma multiplication in apical leaves was not affected by P. putida S1Pf1Rif and G. mosseae BEG12. Consistent with this, single inoculations with P. putida S1Pf1Rif (Gamalero et al., 2010b) or G. mosseae alone (author’s unpublished results) also had no effect on the multiplication of CY. Phytoplasma viability (as determined by the number of 16S rDNA transcripts per CY cell) in the same plant tissues decreased with the progression of the disease and no differences were found between the inoculated and non-inoculated plants. On the other hand, phytoplasma cells in mature leaves were degenerated in double-inoculated plants when observed by TEM; this observation correlates with the delay in symptom development in inoculated plants. Recently, hormonal and transcriptional profile analyses indicated the role of G. mosseae in the jasmonate biosynthetic pathway (Lopez-Raez et al., 2010) in root tissues. It can be speculated that the activation of metabolic pathways involved in plant resistance to biotic stresses is involved in the senescence of phytoplasma cells in fully differentiated and mature leaves, and that has a minor effect on dividing and elongating cells of apical leaves, possibly because of their distance from the roots (Lopez-Raez et al., 2010). This could explain why viable CY cells were found in developing apical leaves of double-inoculated plants in the present study. Therefore, for the two combined microorganisms, an indirect effect on phytoplasma infection and on the relief of phytoplasma damage, perhaps not just related to the improved nutritional status of inoculated C. carinatum, is envisaged.

At the end of the experiments, both the fluorescent pseudomonad and the AM fungus were present and abundant on and inside the plant roots, respectively. Actually, the density and persistence of P. putida S1Pf1Rif was higher in CY-infected plants compared to inoculated healthy controls. On the other hand, only the frequency of vesicles (Table 2, V%), among all of the mycorrhizal colonization parameters, was lowered in CY-infected plants. Vesicles are lipids and cytoplasm-containing hyphal swellings useful when the plant is maturing or is exposed to unfavourable conditions (Smith & Read, 2008), as is the case in phytoplasma infections.

In conclusion, the application of a combination of an AM fungus and a rhizospheric bacterium provided two major results: some level of resistance (about 30%) to phytoplasma infection, and increased tolerance to the disease as measured by delayed symptom expression and improved growth of infected plants. The tolerance induced by the combination of the two microorganisms was probably not just mediated by the improved growth of the plant and might be associated with alterations in the hormonal balance of the inoculated plant (Lopez-Raez et al., 2010).

Based on these results from an experimental pathosystem model, the practical application of rhizospheric microorganisms for mitigating phytoplasma damage should be evaluated under field conditions, especially in the case of perennial crops. This method might represent an additional tool for the integrated management of phytoplasmosis, and it might eventually allow a reduction in insecticide applications for more environmentally friendly control of these diseases.

Acknowledgements

This work was supported by the Piedmont Region within the project ‘Valutazione dell’azione di microrganismi rizosferici ed elicitori di resistenza sull’infezione da fitoplasmi in un sistema modello (CIPE 2006)’.

Ancillary