Alain Sarniguet

Inter-kingdom encounters: recent advances in molecular bacterium-fungus interactions.

Interactions between bacteria and fungi are well known, but it is often underestimated how intimate and decisive such associations can be with respect to behaviour and survival of each participating organism. In this article we review recent advances in molecular bacterium–fungus interactions, combining the data of different model systems. Emphasis is given to the positive or negative consequences these interactions have on the microbe accommodating plants and animals. Intricate mechanisms of antagonism and tolerance have emerged, being as important for the biological control of plants against fungal diseases as for the human body against fungal infections. Bacterial growth promoters of fungal mycelium have been characterized, and these may as well assist plant-fungus mutualism as disease development in animals. Some of the toxins that have been previously associated with fungi are actually produced by endobacteria, and the mechanisms that lie behind the maintenance of such exquisite endosymbioses are fascinating. Bacteria do cause diseases in fungi, and a synergistic action between bacterial toxins and extracellular enzymes is the hallmark of such diseases. The molecular study of bacterium–fungus associations has expanded our view on microbial communication, and this promising field shows now great potentials in medicinal, agricultural and biotechnological applications.

Relationships between take-all, soil conduciveness to the disease, populations of fluorescent pseudomonads and nitrogen fertilizers

Take-all of wheat, caused by Gaeumannomyces graminis var tritici (Ggt), is reduced by ammoniacal fertilizers as compared to nitrate sources. This influence of nitrogen on the disease is only observed on nodal roots at flowering. But soil conduciveness to take-all, as measured in a soil bioassay, is modified earlier. Forty days after nitrogen application at early tillering, the NH4-treated soil became less conducive than the NO3-treated one. When nitrogen applications are done at sowing and at tillering, differences in disease propagation between the two soils are enhanced. Results from four years of experimentation show that when the level of natural soil inoculum is high, disease severity is reduced by ammonium, showing an effect on the parasitic phase of Ggt. At a low level of natural inoculum the effect of the source of nitrogen is mainly observed on the percent of infected plants, indicating that the saprophytic and preparasitic phases are affected. Rhizospheric bacterial populations increase from sowing to tillering, but differences on take-all conduciveness after tillering are not correlated with differences in the amounts of aerobic bacteria or fluorescent pseudomonads isolated from soils treated with different sources of nitrogen. Qualitative changes in fluorescent Pseudomonas spp. populations, like in vitro antagonism, are more likely to explain differences in soil conduciveness to take-all than are quantitative changes in this group. Nevertheless, the introduction of Ggt in a cropped soil leads to a greater increase in fluorescent pseudomonads populations than in total aerobic bacteria.The delay between reducing soil conduciveness and reducing disease in the field with ammonium nitrogen fertilization, the qualitative change of fluorescent pseudomonads populations and the role of necroses in rhizobacteria multiplication, provide information leading to our representation of a dynamic model based on the differentiation of the wheat root system into seminal and nodal roots.

Dominant colonisation of wheat roots by Pseudomonas fluorescens Pf29A and selection of the indigenous microflora in the presence of the take-all fungus

Increases in populations of fluorescent pseudomonads on wheat roots are usually associated with take-all decline, natural control of take-all, a disease caused by the fungus Gaeumannomyces graminis var. tritici (Ggt). Colonisation by Pseudomonas fluorescens strain Pf29A was assessed on the roots of healthy plants and of plants with take-all, and the effect of this bacterium on indigenous populations of fluorescent pseudomonads was studied. The efficacy of Pf29A as an agent for the biocontrol of take-all on five-week-old wheat seedlings was tested in non-sterile conducive soil in a growth chamber. RAPD (random amplification of polymorphic DNA) fingerprinting with a decamer primer was used to monitor strain Pf29A and culturable indigenous rhizoplane populations of fluorescent pseudomonad. Pf29A decreased disease severity and accounted for 44.6% of the culturable fluorescent pseudomonads on healthy plant rhizoplane and 75.8% on diseased plant rhizoplane. Fewer RAPD patterns were obtained when Pf29A was introduced into the soil with Ggt. In the presence of Ggt and necrotic roots, Pf29A became the dominant root coloniser and dramatically changed the diversity and the structure of indigenous fluorescent pseudomonad populations. The results show that Ggt and reduced lesion size on roots can trigger a specific increase in antagonist populations and that the introduction of a biocontrol agent in soil influences the structure of indigenous bacterial populations.

Soil conduciveness to take-all of wheat: Influence of the nitrogen fertilizers on the structure of populations of fluorescent pseudomonads

In a field cropped with wheat, a high and low level of soil conduciveness to take-all were induced by applying a nitrogen fertilizer with either calcium nitrate or ammonium sulphate. From these two soils, two representative populations of fluorescent pseudomonads were tested for their in situ behaviour. Take-all index and root dry weight were assessed on plants cropped in soils infested with Gaeumannomyces graminis var tritici (Ggt) and each bacterized with one of the isolates of fluorescent pseudomonads. The bacteria tested can be split into three groups: antagonists which reduce take-all, deleterious isolates which aggravate the disease and neutral without evident effect on the disease. The predominance of antagonistic fluorescent pseudomonads in the NH4-treated soil and the predominance of deleterious ones in the NO3-treated soil was confirmed after statistical analysis. The microbial impact on take-all must be more considered as the resulting effect of divergent activities of both rhizobacteria types than the only consequences of the presence of antagonistic pseudomonads. All the high cyanogenic pseudomonads were antagonists in situ and were more numerous in the NH4-treated soil than in the NO3-treated soil.

Brassilexin, a novel sulphur-containing phytoalexin from Brassica juncea L., (Cruciferae)

Structure 1 is established for brassilexin, a new sulphur-containing phytoalexin isolated from the leaves of Brassica juncea (Cruciferae), on the basis of spectrographical data (UV, IR, high resolution MS, 13C and 1H NMR).

Role of fungal trehalose and bacterial thiamine in the improved survival and growth of the ectomycorrhizal fungus Laccaria bicolor S238N and the helper bacterium Pseudomonas fluorescens BBc6R8

The mycorrhiza helper bacterial strain Pseudomonas fluorescens BBc6R8 enhances the establishment of Laccaria bicolor S238N ectomycorrhizae by improving the pre-symbiotic growth and survival of the fungus. Nothing is known about the effect of the ectomycorrhizal fungus on the helper bacteria or the molecules that are involved in the interaction. In this study, we have monitored the population density of the helper strain P. fluorescens BBc6R8 in soils inoculated with L. bicolor and in control soils and found that the ectomycorhizal fungus improves the survival of the helper bacteria. We investigated the identity of the fungal and bacterial metabolites involved in this reciprocal growth-promoting effect using a combination of growth measurements, chemoattractant assays, HPLC and in silico genome analyses. We showed that trehalose, a disaccharide that accumulates to high levels in the fungal hyphae, chemoattracted and promoted the growth of the helper bacteria. Meanwhile, P. fluorescens BBc6R8 produced thiamine at concentrations that enhanced the fungal growth in vitro. Altogether our data indicate that the interaction between the two microorganisms is beneficial for both species and relies, at least in part, on trophic mutualism.

Accumulation of a phytoalexin in Brassica spp in relation to a hypersensitive reaction to Leptosphaeria maculans

A phytoalexin was isolated from the leaves of Brassica juncea cv. Aurea inoculated with Leptosphaeria maculans. Its purification and quantification were carried out by reversed-phase and silica HPLC. The molecular formula of this phytoalexin is C9H6N2S. It prevented spore germination and hyphal growth of L. maculans at 12 μg ml−1 in water. Accumulation of this phytoalexin was also obtained with sprays of AgNO3 or CuCl2 solutions. It was not detected in healthy plant tissues, and the effect of cycloheximide suggests that its accumulation is associated with induced plant metabolism. Kinetics of phytoalexin accumulation were compared in B. juncea cv. Aurea (hypersensitive to L. maculans) and B. napus cv. Brutor (susceptible to L. maculans) with CuCl2 as a non-specific elicitor. The phytoalexin was detected in B. juncea 6 h after challenge and in B. napus 18 h after challenge. Moreover, B. juncea always accumulated more phytoalexin than did B. napus (4 to 10 times more 48 h after elicitation).

Evaluation of populations of fluorescent pseudomonads related to decline of take-all patch on turfgrass

Take-all on turfgrass caused by Gaeumannomyces graminis var. avenae (Gga) occurs as patches of yellowish plants. On some patches the central zone was recolonized by the same grass species, Festuca sp., previously damaged by the fungus despite the centrifugal extension of the disease. This disease remission was assimilated to decline. Rhizosphere bacterial counts showed that total population of bacteria was nearly the same in all zones across the patches. However, the ratio of fluorescent Pseudomonas spp./ total bacteria was 1/22, 1/15.4, 1/3.5 and 1/2.9 in the disease free area, the front margin of the patch, in the damaged part of the patch, and in the recolonized central part respectively. Furthermore, in this last mentioned zone, 44 to 82% of the fluorescent Pseudomonas spp. were antagonistic in vitro to Gga, whereas only 12 to 34% from the disease free area were antagonistic. So the development of take-all on turf induced quantitative and qualitative changes in populations of fluorescent pseudomonads. The remission of the disease in the center was correlated to higher amount of antagonistic fluorescent pseudomonads in this part of the patches. This typical patch with the well defined zones can provide a good model for the study of changes in bacterial populations related to the build up of take-all decline.

Effect of Wheat Roots Infected with the Pathogenic Fungus Gaeumannomyces graminis var. tritici on Gene Expression of the Biocontrol Bacterium Pseudomonas fluorescens Pf29Arp

Traits contributing to the competence of biocontrol bacteria to colonize plant roots are often induced in the rhizosphere in response to plant components. These interactions have been studied using the two partners in gnotobiotic systems. However, in nature, beneficial or pathogenic fungi often colonize roots. Influence of these plant-fungus interactions on bacterial behavior remains to be investigated. Here, we have examined the influence of colonization of wheat roots by the take-all fungus Gaeumannomyces graminis var. tritici on gene expression of the biocontrol bacterium Pseudomonas fluorescens Pf29Arp. Bacteria were inoculated onto healthy, early G. graminis var. tritici-colonized and necrotic roots and transcriptomes were compared by shotgun DNA microarray. Pf29Arp decreased disease severity when inoculated before the onset of necrosis. Necrotic roots exerted a broader effect on gene expression compared with early G. graminis var. tritici-colonized and healthy roots. A gene encoding a putative type VI secretion system effector was only induced in necrotic conditions. A common pool of Pf29Arp genes differentially expressed on G. graminis var. tritici-colonized roots was related to carbon metabolism and oxidative stress, with a highest fold-change with necrosis. Overall, the data showed that the association of the pathogenic fungus with the roots strongly altered Pf29Arp adaptation with differences between early and late G. graminis var. tritici infection steps.

Direct and specific assessment of colonisation of wheat rhizoplane by Pseudomonas fluorescens Pf29A

The efficacy of fluorescent pseudomonads as suppressors of soil-borne diseases is linked to their ability to colonise plant roots. Monitoring the dynamics of biocontrol agents in the rhizosphere should improve the irreliability. We designed a pair of Sequenced Characterised Amplified Region (SCAR) primers specific to Pseudomonas fluorescens Pf29A, based on a specific 700 bp RAPD product selected in a previous work. Primer specificity was tested with DNA samples extracted from rhizospheric soil and rhizoplane of wheat plants grown in two different non-sterile soils. We assessed the total population of Pf29A by PCR and the culturable population by counting a tetracycline-resistant Pf29A transformant producing Green Fluorescent Protein (GFP), on selective medium 5 days after inoculation of non-sterile soil. SCAR primers were specific for Pf29A in both soils. We evaluated the limit of detection to 14.2 fg of target DNA, equivalent to 242 Pf29A cells per cm of wheat root. Culturable populations of Pf29A transformant accounted for 13% and 4% of the total populations 5 days after treatment with 103 and 107 CFU of transformed Pf29A per gram of soil. The SCAR derived sequence is a good candidate to develop a strain specific and sensitive PCR-quantification of Pf29A available for population dynamic studies in fields. We confirm that only a small proportion of the total Pf29A rhizosphere population is culturable.