rhizosphere-driven lipopeptide production by different...

COMMUNAUTE FRANCAISE DE BELGIQUE ACADEMIE UNIVERSITAIRE WALLONIE-EUROPE

UNIVERSITE DE L IEGE – GEMBLOUX AGRO-BIO TECH

“ Rhizosphere-driven Lipopeptide Production by Different

Strains of Bacillus spp. as Mechanism Involved in

Biological Control of Plant Pathogens”

Venant NIHORIMBERE

Dissertation originale présentée en vue de l’obtention du grade de Docteur en sciences agronomiques et ingénierie biologique

Promoteurs: Pr. Philippe THONART

Dr. Marc ONGENA

2010

Copyright. Aux termes de la loi belge du 30 juin 1994, sur le droit d'auteur et les droits voisins, seul l'auteur a le droit de reproduire partiellement ou complètement cet ouvrage de quelque façon et forme que ce soit ou d'en autoriser la reproduction partielle ou complète de quelque manière et sous quelque forme que ce soit. Toute photocopie ou reproduction sous autre forme est donc faite en violation de la dite loi et de des modifications ultérieures.

Members of the jury

Mme Marianne SINDIC (president), ULg, Gbx AGRO-BIO TECH

M. Philippe THONART (promotor), ULg, Gbx AGRO-BIO TECH

M. Marc ONGENA (co-promotor), ULg, Gbx AGRO-BIO TECH

M. Daniel PORTETELLE, ULg, Gbx AGRO-BIO TECH

M. Georges LOGNAY, ULg, Gbx AGRO-BIO TECH

Mme Magali DELEU, ULg, Gbx AGRO-BIO TECH

M. Philippe JACQUES, Université des Sciences et Technologies de Lille, France

i

Summary/ Résumé Nihorimbere Venant (2010). “Rhizosphere-driven lipopeptide production by different strains of Bacillus spp. as mechanism involved in biological control of plant pathogens” (Ph.D. Thesis). Belgium, University of Liège - Gembloux Agro-Bio Tech, 151 p., 9 Tabl., 28 Fig. Summary: Some plant-beneficial bacteria with biopesticide potential produce antimicrobial compounds that are tightly involved in multitrophic interactions occurring in the phytosphere among which antagonism toward a diverse range of phytopathogens. However, there is a crucial need for an accurate assessment of antibiotic production rate in this environment colonized by these bacteria. In this context, we conducted the present thesis work with the scope to better understand how two different Bacillus genus strains produce lipopeptides in the rhizosphere of plants. In the first part, one of the strains (B. amyloliquefaciens S499) was selected out of other isolates for its technological traits and in vitro inhibition growth activity of plant pathogenic fungi. We further demonstrated its biocontrol potential in tomato open-field experiments where plantings have been devastated by a local fungus preliminary identified as Fusarium semitectum. In a second part, we combined two mass spectrometry-based approaches (electrospray ionization and imaging) to analyze the pattern of surfactin, iturin and fengycin lipopeptide families produced in planta by strain S499. Our results show that rhizosphere conditions are conducive for surfactin synthesis but not for other types of lipopeptides and that the lipopeptide pattern can be markedly influenced by nutritional factors, biofilm formation and oxygen availability. In a last part, surfactin gene expression (srfA) level was evaluated in situ on tomato root using the reporter gene (LacZ) inserted in B. subtilis strain BGS3. Results showed effective expression of srfA and production of surfactin in biologically important level quantities upon establishment of bacterial population on roots. Our results also demonstrate that BGS3 developing in colonies, efficiently utilizes the main substrates from plant exudates to produce surfactins. The production may also be favored in bacteria growing slowly in the rhizosphere. Globally, this work contributes to better appreciate the impact of some environmental factors on the in situ biosynthesis of lipopeptides by strains of Bacillus which is probably an essential step for improving the level and reliability of their efficacy as biological agents for the control of plant diseases. Nihorimbere Venant (2010). “ Production des lipopeptides dans la rhizosphère par Bacillus en tant que mécanisme impliqué dans le contrôle biologique des agents pathogènes des plantes”(Thèse de Doctorat en anglais). Belgique, Université de Liège - Gembloux Agro-Bio Tech, 151 p., 9 Tabl., 28 Fig. Résumé: Certaines bactéries bénéfiques des plantes à fort potentiel biopesticide produisent des composés antimicrobiens qui sont largement impliqués dans les interactions multi-trophiques qui ont lieu dans la phytosphère et notamment l’antagonisme envers une gamme de phytopathogènes. Actuellement, il est crucial d’évaluer avec précision les taux de productions de ces antibiotiques dans la phytosphère. Dans ce contexte, nous avons réalisé cette thèse ayant comme objectif de mieux comprendre la production de lipopeptides par deux souches de Bacillus dans la rhizosphère. Dans un premier temps, l’une des souches (B. amyloliquefaciens S499) a été sélectionnée parmi une collection d’isolats sur base de son potentiel technologique et de sa capacité à inhiber in vitro la croissance de divers champignons phytopathogènes. De plus, nous avons démontré le potentiel de cette souche en tant qu’agent de biocontrôle lors d’essais en plein champs. Ces derniers ont été réalisés sur des cultures de tomates contaminées par un pathogène local identifié comme étant potentiellement une souche de Fusarium semitectum. Dans un second temps, nous avons combiné deux approches basées sur la spectrométrie de masse (ionisation par électrospray et «imaging») afin d’analyser les profils en surfactines, iturines et fengycines produits par la souche S499 in planta. Nos résultats indiquent que les conditions de la rhizosphère permettent la production de surfactines mais pas celle des autres familles de lipopeptides. De plus, nous avons également montré que la production de ces trois familles d’antibiotiques est fortement influencée par les nutriments disponibles, la formation de biofilm et la pression en oxygène. Dans une dernière partie, l’expression in situ du gène de la surfactine (srfA) dans la rhizosphère de plants de tomate a été évaluée grâce au gène rapporteur (LacZ) inséré dans la souche B. subtilis BGS3. Nos résultats montrent que le gène srfA est effectivement exprimé et entraine la production de quantités biologiquement significatives de surfactine après la colonisation des racines par la bactérie. Nous avons également démontré que BGS3 exploite efficacement les principaux substrats présents dans les exsudats de tomates afin de produire de la surfactine. Les résultats obtenus laissent également supposer que la production de surfactine pourrait être favorisée par la croissance lente de la bactérie dans la rhizosphère. Globalement, ce travail contribue à une meilleure compréhension des facteurs environnementaux influençant la production in situ de lipopeptides par les souches de Bacillus. Ceci est probablement une étape essentielle dans l’amélioration de la fiabilité et l’augmentation de l’intensité de la protection associée à ces agents biologiques de contrôle des phytopathogènes.

ii

Scientific communications

Publications Venant Nihorimbere, Patrick Fickers, Philippe Thonart and Marc Ongena (2009). Ecological

fitness of Bacillus subtilis BGS3 regarding production of the surfactin lipopeptide in the rhizosphere. Environmental Microbiology Reports, 1: 124-130.

Venant Nihorimbere, Marc Ongena, Hélène Cawoy, Yves Brostaux, Pascal Kakana, Emmanuel Jourdan and Philippe Thonart (2010). Beneficial effects of Bacillus subtilis on field-grown tomato in Burundi: Reduction of local Fusarium disease and growth promotion. African Journal of Microbiology Research 4: 1135-1142.

Venant Nihorimbere, Marc Ongena, Maïté Smargiassi and Philippe Thonart (2010). Beneficial effect of the rhizosphere microbial community for plant growth and health. In press to be published in the next issue in the journal of Biotechnology Agronomy Sociology and Environment (BASE). Review.

Venant Nihorimbere, Hélène Cawoy, Alexandre Seyer, Alain Brunelle, Philippe Thonart and Marc Ongena (2010). Bacillus life in soil: impact of rhizosphere factors on lipopeptide signature from the plant beneficial strain B. amyloliquefaciens S499. Submitted in the Journal of International Society for Microbial Ecology.

Nihorimbere Venant, Ongena Marc and Thonart Philippe (2010). In situ production of Bacillus antibiotics related to biocontrol: more evidences needed. In preparation for submission as a review.

Venant Nihorimbere, Marc Ongena, Hélène Cawoy, Yves Brostaux, Pascal Kakana and Philippe Thonart (2010). Bacillus-based biocontrol of Fusarium disease on tomato cultures in Burundi. Communications in Agricultural and Applied Biological Science journal, 74: 645-649.

Marc Ongena, Venant Nihorimbere, Patrick Fickers, Philippe Thonart (2009). Influence of rhizosphere-specific parameters on surfactin production by Bacillus subtilis. IOBC/wprs Bulletin, 43: 317-320.

Marc Ongena, Hélène Cawoy, Maïté Smargiassi, Venant Nihorimbere, Emmanuel Jourdan and Philippe Thonart (2010). Modulation of the lipopeptide pattern secreted by Bacillus subtilis upon colonization of different plant roots. IOBC bulletin, in Press.

Symposium & Congress 61th International Symposium on Crop Protection. Gent, Belgium, May 19th 2009. Bacillus-

based biocontrol of Fusarium disease on tomato cultures in Burundi (Oral presentation).

Bioforum, Liège, Belgium, October 11th 2007. Ecological fitness of plant beneficial Bacillus subtilis strains in soil: influence of rhizosphere-specific parameters on surfactin synthesis (Poster).

9th International Congress of Plant Pathology. Torino, Italy, 24-29 August 2008. Ecological fitness of plant-beneficial Bacillus subtilis strains in soil: Influence of rhizosphere-specific parameters on surfactin synthesis (Poster).

iii

Acknowledgements / Remerciements

La thèse constitue une expérience intense, passionnante et marquante. Elle m’a donné l’occasion de

côtoyer de nombreuses personnes pour qui je ne saurais en quelques lignes exprimer mes profonds

sentiments. First of all, I would like to thank the persons who helped me the most in achieving this

thesis. Je remercie très vivement le Professeur Philippe THONART qui m’a admis dans son

laboratoire de Bioindustries et qui, par la suite, m’a confié à travailler sur le thème de cette thèse de

doctorat. Ses paroles et conseils à l’occasion de nos rencontres sont, on ne peut plus, très

encourageants. Je suis extrêmement reconnaissant au Docteur Pascal KAKANA qui m’a introduit

auprès du Professeur THONART pour qu’il soit promoteur de ma thèse.

Cette thèse a été de plus agréable, vivante, et excitante grâce au Dr. Marc ONGENA. Marc,

thanks very much for all your guidance. You’re very direct, which can be quite confronting. We

discussed regularly the many options in my project, and by thinking in terms of what is required for a

Ph.D. thesis, we achieved a great output. We had a very good collaboration, and fruitful conversations,

also made possible because you also concern somebody’s personal life. Every day you make your

round in the lab to ask how everything was going, and you always noticed when something was not

going okay so well, even when I did not tell you. Not always you would ask about it further, I would

tell you in the end anyway, even when it was something personal. We also had lots of fun if there was

time; we enjoyed and exchanged lots sometimes after lab in the bar. I here want to thank you for being

such a good co-supervisor; also by your enthusiasm, optimism and humor. You are a perfectionist!!!

Uufff, pas encore fatigué, il m’est un honneur et un grand plaisir de t’exprimer ma profonde

reconnaissance et mes sincères remerciements.

Mes remerciements les plus sincères sont aussi à ceux qui m’ont fait l’honneur d’être dans le

comité de lecture de ma thèse en tant que rapporteurs: Prof. Philippe JACQUES, Dr. Magali DELEU,

qui ont bien voulu me faire-part de leurs pertinentes critiques et suggestions, et à ceux qui m’ont fait

l’honneur d’être dans mon jury: Prof. Daniel PORTETELLE, Prof. Georges LOGNAY pour leurs

disponibilités et conseils tout le long des quatres années de mon doctorat. Je remercie également

madame la présidente du Jury Prof. Marianne SINDIC pour sa disponibilité.

Si l’Unité de Bioindustries est un lieu à la fois convivial et pratique, c’est avant tout au Prof.

THONART, qu’elle le doit : je ne serais ni le premier ni le dernier à le reconnaître, mais c’est sincère.

Mais aussi, cet état est plus enthousiamé par les Drs. Jacqueline DESTAIN, Marc ONGENA et Franck

DELVIGNE. Les secrétaires, Mmes CHANET et DEMI, jouant un rôle très essentiel.

Je tiens également à remercier toutes les personnes que j’ai appris à connaître et à apprécier au

cours de ces années passées dans l'Unité de Bioindustries. Merci pour la bonne ambiance et les

échanges, scientifiques et autres. Plus particulièrement, mes remerciements s’adressent à mon groupe

de recherche «Plant-Bacillus-Lipopeptides». Que vous, Hélène, Guillaume et Manu, trouviez ici le

fruit de vos efforts particuliers déployés for the last chechs and finishing touches of the the lay-out of

iv

the thesis. A Michel et Tambi (merci pour le thé du Burundi partagé chaque soir vers 18 h 30 min, où

j'ai peut-être trouvé une partie de l'inspiration pour rédiger cette thèse !!),…et à tous techniciens, les

étudiants, doctorants, stagiaires de notre unité croisés au labo ou dans les bureaux.

I would also like to acknowledge the financers of this Ph.D. project, CTB/BTC, for giving me

all the support required for that. Je n'oublie pas ma famille pour leur soutien constant au cours de

toutes ces années d’études… Of course, last but not least, un merci et un bisou à mes enfants Reine et

Théo, pour tout le courage que vous avez su me donner pour finaliser cette thèse dans les meilleurs

délais (dommage, vous êtes encore petits pour ne pas réaliser ça. You were born respectively in the

first and third year of this thesis (2007 & 2009)! Enfin, un merci tout particulier à toi ma très chère

épouse Gaudence, pour toute la confiance que tu as toujours su mettre en moi.

v

Contents

PREFACE ............................................................................................................................... 1

INTRODUCTION .................................................................................................................... 3

Chapter 1. Beneficial effect of the rhizosphere microbial community for plant growth and

health ..................................................................................................................... 4

Chapter 2. Bacillus lipopeptides: structure, regulation and roles in biocontrol................ 25

Chapter 3. In situ production of Bacillus antibiotics involved in the biocontrol of plant

diseases: more evidences needed ........................................................................ 40

OBJECTIVES OF THE THESIS ......................................................................................... 57

RESULTS ............................................................................................................................. 59

Chapter 4. Beneficial effects of Bacillus subtilis on field-grown tomato in Burundi:

Reduction of local Fusarium disease and growth promotion ............................. 61

Chapter 5. Bacillus life in soil: impact of rhizosphere factors on lipopeptide signature

from the plant beneficial strain B. amyloliquefaciens S499................................ 82

Chapter 6. Ecological fitness of Bacillus subtilis BGS3 regarding production of the

surfactin lipopeptide in the rhizosphere ............................................................ 116

GENERAL DISCUSSION AND FUTURE PROSPECTS............................................... 131

1

PREFACE

The rhizosphere is a very complex environment in which the effects of the plant on soil

microorganisms and the effects of the microorganisms on the plant are interacting and

interdependent. Plant root exudates and breakdown products attracts microbes and feed them.

In turn, the plant may clearly benefits from some microbes. Such interaction may be indeed

essential for plant health not only in view of nutritional requirements and enhancement of

growth but also to control certain plant pathogens. An overview of the fundamental processes

governing rhizosphere microbial ecology is presented in Chapter 1.

On another hand, some microorganisms cause serious diseases to plants leading to

considerable losses in agricultural production. Therefore, plant diseases need to be controlled

to maintain the quality and abundance of food produced by growers to mitigate the problem

of food security faced by the increasing world population. Nowadays, growers still heavily

rely on chemical fertilizers and pesticides to prevent, or control plant diseases. However,

environmental, societal and ethical problems are rising up because of excessive use or misuse

of these agrochemicals. Significant pollution of soils and ground water reservoirs,

accumulation of undesirable chemical residues on products and subsequently in the food

chain, emergence of fungicide-resistant strains of pathogens and, last but not least, very

harmful health concerns for growers. This situation calls for alternative strategies to control

plant infection. Among these alternatives are those related to biological control defined as the

use of living organisms or natural substances isolated thereof to combat plant disease.

Therefore, several groups of scientists are actively engaged in research pertaining to the

identification of biocontrol agents against specific plant pathogens, with special focus on their

use as biopesticides at a commercial scale. Over the past decades, research has repeatedly

demonstrated that phylogenetically diverse microorganisms can act as natural antagonists of

various plant pathogens. In that context, some microorganisms have been successfully

commercialized and are currently marketed as biopesticides including bacteria belonging to

the genera Agrobacterium, Bacillus, Pseudomonas, and Streptomyces and fungi belonging to

the genera Ampelomyces, Candida, Coniothyrium, and Trichoderma.

Among these genera, Bacillus strains are among the best candidates for use as biological

control agents. This is first due to their great potential to produce several different types of

insecticidal and antimicrobial compounds. Secondly, Bacillus species are able to produce

spores that allow them to resist adverse environmental conditions and permit easy formulation

and storage of the commercial products. Among the large panoply of antibiotics synthesized

2

by Bacillus, lipopeptides are more specifically studied in this work because they are tightly

related to the biocontrol activity of the producing strain. The structural traits and cellular

biosynthesis pathways of these molecules are presented in Chapter 2 along with their

biological activities in the context of plant-Bacillus-pathogen interactions. To connect with

the objectives of the thesis, the last chapter of the literature review deals with the

environmental parameters that may influence lipopeptide production in situ by Bacillus spp.

and presents the techniques and limitations for their detection in the rhizosphere.

It should be noted that an important part of this work is devoted to the study of one

particular Bacillus isolate named S499. For better clarity, we want to inform the reader here

that this strain was formerly identified as B. subtilis based on basic bacteriological tests but

further and recent genetic characterization showed that it most probably belongs to the closely

related B. amyloliquefaciens species rather than B. subtilis (see Chapter 5 for details).

3

INTRODUCTION

Introduction-Chapter 1

4

Chapter 1. Beneficial effect of the rhizosphere microbial community for plant growth and health

In Press in Biotechnologie, Agronomie, Société et Environnement Journal

Venant Nihorimbere (1,3†), Marc Ongena (1†), Maïté Smargiassi (1), Philippe Thonart (1,2)

1 Université de Liège, Gembloux Agro-Bio Tech, Centre Wallon de Biologie Industriel,

Passage des déportés, 2, B-5030 Gembloux, Belgique. 2 Université de Liège, Service de Technologie Microbienne, Boulevard du Rectorat, 29, B-40

Liège, Belgique. 3 University of Burundi, Faculty of Agricultural Sciences, PO Box 1550, Bujumbura, Burundi.

†Both authors equally contributed to this article.


5

Beneficial effect of the rhizosphere microbial community for plant growth and health

Abstract

Plant rhizosphere is the soil nearest to the plant root system where roots release large quantity

of metabolites from living root hairs or fibrous root systems. These metabolites act as

chemical signals for motile bacteria to move to the root surface but also represent the main

nutrient sources available to support growth and persistence in the rhizosphere. Some of the

microbes that inhabit this area are bacteria that are able to colonize very efficiently the roots

or the rhizosphere soil of crop plants. These bacteria are referred to as plant growth promoting

rhizobacteria (PGPR). They fulfil important functions for plant growth and health by various

manners. Direct plant growth promotion may result either from improved nutrient acquisition

and/or from hormonal stimulation. Diverse mechanisms are involved in the suppression of

plant pathogens, which is often indirectly connected with plant growth. This paper describes

the different mechanisms commonly used by most PGPR in their natural habitats to influence

plant-growth and health.

Keywords: Rhizosphere, PGPR, Root exudation, Plant-microbe interaction

Effet bénéfique de la communauté microbienne de la rhizosphère sur la croissance et la santé des plantes

Résumé

La rhizosphère est le volume du sol situé au voisinage immédiat des racines des plantes et qui

se caractérise par la présence d’exsudats racinaires (rhizodépôts). Ces exsudats sont utilisés

par la microflore endémique en tant que signaux chimiques en plus d’être un substrat nutritif

disponible pour la croissance et le développement de ces microorganismes dans la

rhizosphère. Certaines de ces bactéries du sol, appelées PGPRs (Plant Growth Promoting

Rhizobacteria), sont capables de coloniser les racines ou bien encore la rhizosphère, mais à la

différence des autres bactéries rhizosphériques elles ont, en retour, un effet bénéfique sur la

plante. Cet effet bénéfique peut être direct, ou indirect. La promotion directe de la croissance

est le résultat du pouvoir d’acquisition des nutriments ou de la stimulation des hormones de la

plante. D’autres mécanismes indirects, mais le plus souvent liés à la croissance des plantes,

sont impliqués dans la réduction/suppression des pathogènes des plantes. Cet article décrit les

différents mécanismes mis en jeu par les PGPRs dans leur environnement naturel pour

influencer favorablement la croissance et la santé des plantes.

Mots-clés: Rhizosphère, PGPR, Exsudation racinaire, Interaction plantes-microorganismes


6

Introduction

According to a general view the rhizosphere includes plant roots and the surrounding soil.

This is a wide and wise definition, already coined more than hundred years ago by Hiltner

(1904). In that particular environment, very important and intensive interactions take place

between the plant, soil, and microfauna. Biochemical interactions and exchanges of signal

molecules between plants and soil microbes have been described and reviewed (Pinton et al.,

2007). The rhizosphere inhabiting microorganisms compete for water, nutrients and space and

sometimes improve their competitiveness by developing an intimate association with plant

(Hartmann et al., 2009). These microorganisms play important roles in the growth and

ecological fitness of their host. An understanding of the basic principles of rhizosphere

microbial ecology, including the function and diversity of microorganisms that reside there, is

necessary before soil microbial technology can be applied in the rhizosphere. Here we review

different mechanisms commonly used by the beneficial rhizosphere bacteria to influence

plant-growth and health in the natural environment.

The rhizosphere effect

During seed germination and seedling growth, the developing plant interacts with a range of

microorganisms present in the surrounding soil. As seeds germinate and roots grow through

the soil, the release of organic material provides the driving force for the development of

active microbial populations in a zone that includes plant root and surrounding soil in a few

mm of thickness. This phenomenon is referred as the rhizosphere effect (Morgan and Whipps,

2001).

Broadly, there are three distinct components recognized in the rhizosphere; the

rhizosphere per se (soil), the rhizoplane, and the root itself. The rhizosphere is thus the zone

of soil influenced by roots through the release of substrates that affect microbial activity. The

rhizoplane is the root surface, including the strongly adhering soil particles. The root itself is a

part of the system, because certain endophytic microorganisms are able to colonize inner root

tissues (Bowen and Rovira, 1999). The rhizosphere effect can thus be viewed as the creation

of a dynamic environment where microbes can develop and interact.


7

Main characteristics of the root exudation process

Root exudation is the release of organic compounds from living plant roots into the

surrounding soil; it is a ubiquitous phenomenon (Moszer et al., 2002). Roots release

compounds via at least two potential mechanisms, and the rates of exudation sensu stricto

vary widely among species and environmental conditions (Kochian et al., 2005). Exudates are

transported across the cellular membrane and secreted into the surrounding rhizosphere. Plant

products are also released from roots border cells and root border-like cells which separate

from border as they grow (Bais et al., 2006). However, it is important to note that it is very

difficult to identify root exudates with respect to the chemical composition and the

concentration in the soil because of methodological difficulties (Stolp, 1988). At the moment

of exudation and thereafter, the organic materials are subject to microbial attack, and thus

cannot be enriched and separated from the roots in the natural environments. Data on the

nature and quantity of root exudates have been obtained from sterile hydroponic cultures; but

the results, however, are difficult to extrapolate to the natural conditions (Stolp, 1988). In this

context, root exudation has been quantified by measuring the production of labelled CO2 in

the rhizosphere of 14C-labelled plants, and it has been estimated that 12-40% of the total

amount of carbohydrates produced by photosynthesis is released into the soil surrounding

roots (Davis et al., 1999). Root exudates are mainly composed of water soluble sugars,

organic acids, and amino acids, but also contain hormones, vitamins, amino compounds,

phenolics and sugar phosphate esters (Uren, 2001).

Release of these low molecular weight compounds is a passive process along the steep

concentration gradient which usually exists between the cytoplasm of intact root cells

(millimolar range) and the external (soil) solution (micromolar range). Direct or passive

diffusion through the lipid bilayer of the plasma membrane is determined by membrane

permeability, which depends on the physiological state of the root cell and on the polarity of

the compounds, facilitating the permeation of lipophilic exudates (Rudrappan et al., 2007).

The efficiency of the exudation process may thus be enhanced by stress factors affecting

membrane integrity such as nutrient deficiency, temperature extremes, or exudation stress

(Ratnayale et al., 1978).

It is assumed that both the qualitative and quantitative compositions of root exudates are

affected by various environmental factors, including pH, soil type, oxygen status, light

intensity, soil temperature, nutrient availability and the presence of microorganisms. These


8

factors may have a greater impact on root exudation than differences due to the plant species

(Singh and Mukerji, 2006).

The proportion of C released from roots has been estimated to as much as 50% in the

young plants (Whipps, 1990) but less in plants grown to maturity in the field (Jensen, 1993).

The nature of exudates may also vary according to the growth stage of the plant. For instance,

there are more carboxylates and root mucilage at the six leaf stage than earlier. On the other

hand, N is also of considerable importance to nutrient cycling, usually as NH4+, NO3-

(Wacquant et al., 1989), amino acids (Boulter et al., 1966), cell lysates, sloughed roots, and

other root-derived debris. It is estimated at the maturity that the rhizodeposition of N

amounted to 20% of the total plant N (Jensen, 1996). Root exudation is also largely dependent

on the nutritional status of the plant regarding oligoelements. Low concentrations of some

nutrients such as K+, Na+ and Mg++ readily stimulate the activity of major enzymes of the

glycolytic pathway, namely phosphofructokinase and pyruvate kinase, which together

regulate glycolysis in plant cells (Plaxton, 1996). Individual micronutrients are similarly

important components of major enzymes, which regulate all biological processes in plants. It

is clear from these considerations that low nutrient availability can constraint plant growth in

many environments of the world, especially the tropics where soils are extremely deficient in

these oligoelement nutrients (Pinton et al., 2007). Some species typically exude organic acid

anions in response to P and Fe deficiency or phytosiderophores due to Fe and Zn deficiency

(Haynes, 1990).

The rhizosphere-inhabiting microflora

Diversity

The rhizosphere microflora include bacteria, fungi, nematodes, protozoa, algae and

microarthrops (Raaijmakers and Weller, 2001). Of the soil microbes, 98% cannot be cultured.

Their identification, characterization and the description of their role are therefore particularly

difficult. Recently, nucleic acid based techniques including analysis of DNA and rRNA

molecules from soil samples have revealed enormous diversity in the rhizosphere inhabiting

microbial flora (Suzuki et al., 2006). The molecular methods used for soil microbial diversity

are covered in the review by Nannipieri and collaborators (2003). The number of microbial

species present in soil may vary from thousands to millions. Many studies indeed suggest that

the Proteobacteria and the Actinobacteria form the most common of the dominant populations

(>1%, usually much more) found in the rhizosphere of many different plant species (Singh et

al., 2007). These groups contain many ‘cultured’ members. They are the most studied of the


9

rhizobacteria, and as such, contain the majority of the organisms investigated, both as

beneficial microbial inoculants and as pathogens.

The specific content of root exudates may create a niche that influences which

microorganisms are to colonize the rhizosphere, thereby altering the composition and

diversity of microorganisms colonizing the rhizosphere in a plant specific manner (Grayston

et al., 1998). Plant species, plant developmental stage and soil type have thus been indicated

as major factors determining the composition of rhizosphere microbial communities

(Broeckling et al., 2008). That said, the extent to which the above-cited factors contribute to

microbial communities is not fully understood and there are several contrasting reports in the

literature indicating either plant or soil type as dominant factor (Nunan et al., 2005). Owing to

the above statement, it can be generalized that the diversity and predominance of rhizosphere

microbial population depend on a number of abiotic and biotic factors prevailing in that

particular ecological niche (Figure 1).

Population level

Studies based on the use of growth media steadily showed that bacterial populations residing

in the rhizosphere are several orders of magnitude larger than those residing in bulk soils.

Rhizosphere bacteria concentration can reach between 1010 and 1012 cells per gram of soil

(Foster, 1988), and they are transferred to various associated environments including plants,

foods, animals, marine and freshwater habitats (Buée et al., 2009). Only few groups of these

bacteria are considered to be soilborne, probably because non-spore forming bacteria cannot

survive well in soil for long periods.


10

Plant species

Root colonization

Root exudates

Humidity

Temperature variationMicrofauna

soil type, pH, …

Photoperiod

Rhizobacteria

Figure 1. Ecological factors influencing the root exudation process and thereby rhizosphere colonization by beneficial rhizobacteria – Facteurs écologiques influençant le processus d’exsudation racinaire et par conséquent la colonisation de la rhizosphère par les rhizobactéries du sol.


11

The effect of root exudates depends on the distance that they can diffuse away from

rhizoplane (Gupta and Mukerji, 2002). Bacterial communities are not uniformly distributed

along root axes, and differ between root zones. Distinct bacterial community compositions are

obtained by molecular fingerprints in different root zones, like those of emerging roots and

root tips, elongating roots, sites of emergence of lateral roots, and older roots (Yang and

Crowley, 2000). It has been proposed that populations residing in the rhizosphere oscillate

along root axes in a wave-like fashion (Semenov et al., 1999). Accordingly, bacterial

communities temporarily profit from the nutrients released by younger roots in the root hair

zones, and wave-like fluctuations in bacterial cell numbers can be explained by death and

lysis of bacterial cells upon starvation when nutrients become depleted, followed by cell

divisions in surviving and thus viable populations as promoted by the release of nutrients

from dead and decaying cells (Semenov et al., 1999). Bacterial communities in rhizosphere

soils are thus not static, but will fluctuate over time in different root zones.

The rhizosphere as a battle field

The number and diversity of microorganisms are related to the quantity and quality of the

exudates but also to the outcome of the microbial interactions that occur in the rhizosphere

(Somers et al., 2004). Soil biota (bacteria, fungi, micro-fauna and the plant root) are

themselves embedded in food webs and thus interactions with consumers or predators in the

microbial as well as macro- and mesofaunal world are important to understand rhizosphere

processes. A high number of soil microbes attained properties enabling them to interact more

efficiently with roots and withstand the quite challenging conditions of rhizosphere life. The

rhizosphere inhabiting microorganisms compete each other for water, nutrients and space and

sometimes improve their competitiveness by developing an intimate association with plant.

This process can be regarded as an ongoing process of micro-evolution in low-nutrient

environments, which are quite common in natural ecosystems (Schloter et al., 2000).

Plant-bacteria interactions in the rhizosphere

Microorganisms present in the rhizosphere play important roles in ecological fitness of their

plant host. Important microbial processes that are expected to occur in the rhizosphere include

pathogenesis and its counterpart, plant protection/growth promotion, as well as the production

of antibiotics, geochemical cycling of minerals and plant colonization (Kent and Triplett,

2002). Plant-microbe interactions may thus be considered beneficial, neutral, or harmful to

the plant, depending on the specific microorganisms and plants involved and on the prevailing

environmental conditions (Bais et al., 2006). Exploring these microorganisms by unravelling


12

their possible relationships with plants has launched a new and fascinating area of

investigations in the rhizosphere research.

Pathogenic interactions

Roots exudates can attract beneficial organisms (see below), but they can also be equally

attractive to pathogenic populations (Schroth and Hildebrand, 1964), that many express

virulence on only a limited number of host species. Many pathogenic organisms, bacteria as

well as fungi, have coevolved with plants and show a high degree of host specificity

(Raaijmakers et al., 2009). In nature however, plant disease is the exception rather than the

rule because the conditions that are optimized for the plant growth may not be favourable for

pathogens (Paulitz and Bélanger, 2001).

Plants are not defenceless. In fact, it is estimated that only about 2% of the known

fungal species are able to colonize plants and cause disease (Buchanan et al., 2000). Even

though plants are in permanent contact with potential pathogens such as fungi, bacteria or

viruses, successful infection is rarely established. Such a general resistance against most

pathogens has been named “horizontal resistance” or “non-host-resistance” (Heath, 1981).

This reflects the fact that the plant is not a suitable target for infection by a specific pathogen

due to preformed, passive resistance mechanisms resulting in “basic incompatibility”. These

resistance mechanisms comprise structural barriers and toxic compounds that are present in

the unaffected, healthy plant and limit successful infection to specialized pathogens that have

the ability to overcome these factors and therefore exhibit “basic compatibility”. If contact is

nevertheless established with the plant tissue, pathogens are often confronted with preformed

chemical components named phytoanticipins (van Etten et al., 1994). This term comprises a

variety of compounds produced by different biosynthetic pathways which possess

antimicrobial properties. These low molecular weight secondary metabolites are mainly

stored in inactive form in the vacuoles or organelles and are released upon destruction of the

cells. Since destroying the integrity of the plant tissue is part of the colonization strategy by

fungi, phytoanticipins represent an important resistance mechanism against these pathogens.

However, in some instances, pathogens can overcome the pre-formed barriers and

develop virulent infection processes leading to plant disease. Plant diseases play a direct role

in the destruction of natural resources in agriculture. In particular, soilborne pathogens cause

important losses, fungi being the most aggressive. The extent of their harmful effects ranges

from mild symptoms to catastrophes where large fields planted with agricultural crops are

destroyed. Thus, they are major and chronic threats to food production and ecosystem


13

stability worldwide. Common and well investigated bacterial agents include Gram-negative

bacteria Erwinia carotovora, Pseudomonas, Ralstonia spp. and the Gram-positive bacterium

Streptomyces scabies. The fungal and oomycete phytopathogens include members of

Fusarium, Phytophthora, Pythium, Rhizopus, Rhizoctonia and Verticillium (Tournas and

Katsoudas, 2005). From the forest pathogens, among the most important are the filamentous

fungi Heterobasidion and Armillariella (Asiegbu and Nahalkova, 2005), and Phytophthora

spp. (Rizzo et al., 2005).

Beneficial microorganisms and modes of action

Plant-beneficial microbial interactions can be roughly divided into three categories. First,

those microorganisms that, in association with plants, are responsible for its nutrition (i.e.,

microorganisms that can increase the supply of mineral nutrients to the plant). In this case,

while most may not directly interact with the plant, their effects on soil biotic and abiotic

parameters certainly have an impact on plant growth. Second, there is a group of

microorganisms that stimulate plant growth indirectly by preventing the growth or activity of

pathogens. Such microorganisms are referred to as biocontrol agents, and they have been well

documented. A third group involves those microorganisms responsible for direct growth

promotion, for example, by production of phytohormones. There has been a large body of

literature describing potential uses of plant associated bacteria as agents stimulating plant

growth and managing soil and plant fitness (Welbaum et al., 2004). On another hand,

apparently neutral interactions are found extensively in the rhizosphere of all crop plants.

Saprophytic microorganisms are responsible for many vital soil processes, such as

decomposition of organic residues in soil and associated soil nutrient mineralization or

turnover processes. Whereas these organisms do not appear to benefit or harm the plant

directly (hence the term neutral), their presence is obviously vital for soil dynamic, and their

absence would clearly influence plant health and productivity (Brimecombe et al., 2007).

Rhizosphere-living bacteria that exert a global beneficial effect on plant growth are

referred as plant growth promoting rhizobacteria (PGPR) (Kloepper and Schroth, 1978). The

number of bacterial species identified as PGPR increased recently as a result of the numerous

studies covering a wider range of plant species and because of the advances made in bacterial

taxonomy and the progress in our understanding of the different mechanisms of action of

PGPR. Presently, PGPR include representatives from very diverse bacterial taxa (Lucy et al.,

2004) and in the following sections we are not giving a thorough description of all the genera

and species of PGPR, but rather a few examples to illustrate the diversity and modes of action


14

of these beneficial bacteria. Diverse PGPR strains have been used successfully for crop

inoculations. These comprise members of the bacterial genera Azospirillum (Cassán and

García Salamone, 2008), Bacillus (Jacobsen et al., 2004), Pseudomonas (Loper and Gross,

2007), Rhizobium (Long, 2001), Serratia (De Vleeschauwer and Höfte, 2007),

Stenotrophomonas (Ryan et al., 2009), and Streptomyces (Schrey and Tarkka, 2008). Some

fungi belonging to the genera Ampelomyces, Coniothyrium, and Trichoderma have also been

described to be beneficial for the host plant (Harman et al., 2004). The modes of action of

PGPR involve complex mechanisms to promote plant growth, development and protection.

Important among them are biofertilization (increasing the availability of nutrients to plant),

phytostimulation (plant growth promoting, usually by the production of phytohormones) and

biocontrol (controlling diseases, mainly by the production of antibiotics and antifungal

metabolites, lytic enzymes and induction of plant defense responses). Pseudomonas and

Bacillus genera are the most commonly investigated PGPR, and often the dominating

bacterial groups in the rhizosphere (Morgan et al., 2005). One has to mention that, in many

cases of individual beneficial plant–microbe interactions, several mechanisms are involved

(Müller et al., 2009). Ad planta, direct mechanisms of plant growth promotion are difficult to

differentiate from disease suppression and the relative importance on a specific mechanism

can vary within different pathosystems (Chet and Chernin, 2002).

Colonization

In all successful plant–microbe interactions, the competence to colonize plant habitats is

important (Lugtenberg et al., 2002; Kamilova et al., 2005). Single bacterial cells can attach to

surfaces and, after cell division and proliferation, form dense aggregates commonly referred

to as macrocolonies or biofilms. Steps of colonization include attraction, recognition,

adherence, invasion (only endophytes and pathogens), colonization and growth, and several

strategies to establish interactions. Plant roots initiate crosstalk with soil microbes by

producing signals that are recognized by the microbes, which in turn produce signals that

initiate colonization (Berg, 2009). PGPR reach root surfaces by active motility facilitated by

flagella and are guided by chemotactic responses (Pinton et al., 2007). This implies that

PGPR competence highly depends either on their abilities to take advantage of a specific

environment or on their abilities to adapt to changing conditions or plant species. As an

example, when strain S499 of Bacillus subtilis was applied to plant seedlings, it showed more

distinct but effective colonization of the root system of two distinct plants (Figure 2). In most

cases, the population of many PGPR inoculants actually declines progressively in time after


15

inoculation from 107-109 cells per gram dry soil to 105-106 cells per gram dry soil after 2-3

weeks (DeFlaun and Gerba, 1993). Nevertheless this population threshold is often sufficient

to provide beneficial effects (Raaijmakers et al., 2002). Rhizosphere competence of

biocontrol agents thus involves effective root colonization combined with the ability to

survive and proliferate along growing plant roots over a large time period, in the presence of

the indigenous microflora (Weller, 1988; Lugtenberg and Dekkers, 1999).

a b c d

Figure 2. Root colonization by strain S499 of Bacillus subitilis - Colonisation racinaire par la souche de Bacillus subtilis S499. Microscopy visualization of root of – Visualisation par microscopie des racines de, a: treated salad – salade traitée, b: untreated salad – salade non traitée, c: treated tomato – tomate traitée, and d: untreated tomato – tomate non traitée. The arrow in a and c indicates the biofilm formation and the cross-surrounded in b and d indicates absence of biofilm formation by the strain on the root system – La flèche en a et c indique la formation de biofilm et la croix entourée en b et d indique l’absence de formation de biofilm sur le système racinaire par la souche. Surface-sterilized tomato and salad seeds were suspended in bacterial suspension of 108 cells ml-1 and germinated for two weeks in the conditioned chamber and in gelified sterile plant nutrient medium as defined by Murashige and Skoog (1962) – Les graines de tomate et de salade stérilisées sont suspendues dans une solution bactérienne de 108 cellules par ml et sont mises à germer pendant deux semaines dans une chambre d’air conditionnée sur un milieu nutritive pour plantes mis au point par Murashige et Skoog (1962). Source: results obtained in our laboratory - Source: résultats obtenus dans notre laboratoire.

Pathogen inhibition

Bacteria and fungi live around roots and feed on root exudates and dead root cells.

Competition between microbial species in this area is stiff. In the battle for establishment and

persistence in the niche, bacteria use several strategies.

Antagonism

Root colonization not only results in high PGPR population densities on the root system, it

also functions as the delivery system of antagonistic metabolites that are involved in direct

inhibition of plant pathogens (Shoda, 2000; Raaijmakers et al., 2002). It includes antibiosis

i.e. the inhibition of microbial growth by diffusible antibiotics and volatile organic

compounds, toxins, and biosurfactants, and parasitism that may involve production of

extracellular cell wall-degrading enzymes such as chitinases and β-1,3-glucanase (Compant et

al., 2005; Haas and Défago, 2005). The degradation of pathogenicity factors of the pathogen

such as toxins by the beneficial organism has also been reported as protective mechanism

(Haas and Défago, 2005). To demonstrate the role of antibiotics in biocontrol, mutants

impaired in biosynthesis or over-producing mutants have been used together with, in some


16

cases, the use of reporter genes or probes to show efficient production of the compound in the

rhizosphere. As example, Bacillus subtilis strains produce a variety of powerful antifungal

metabolites, e.g., zwittermicin-A, kanosamine and lipopeptides from the surfactin, iturin and

fengycin families (Emmert and Handelsman, 1999; Ongena and Thonart, 2006). Dunne and

collaborators (2000) showed that overproduction of extracellular protease in the mutant

strains of Stenotrophomonas maltophilia W81 resulted in improved biocontrol of Pythium

ultimum. Excretion of chitinases and glucanases by species of Trichoderma and Streptomyces

has also been shown to play an important role in mycoparasitism of phytopathogenic fungi

(Whipps, 2001).

Competition

Competition for resources such as nutrients and oxygen occurs generally in soil between soil-

inhabiting organisms. For biocontrol purpose, it occurs when the antagonist directly competes

with pathogens for these resources. Root inhabiting microorganisms compete for suitable sites

at the root surfaces. Competition for nutrients, especially for carbon, is assumed to be

responsible for the well-known phenomenon of fungistasis characterizing the inhibition of

fungal spore germination in soil (Alabouvette et al., 2006). Given the relative abundance of

substrates in the rhizosphere, the efficiency of nutrient uptake and catabolism by bacteria is a

key factor in competitiveness (Chin-A-Woeng et al., 2003). The capacity for rapid growth

when substrates are encountered is not the only factor affecting rhizosphere competence, as

rhizobacteria deploy many other metabolic strategies. For example, the capacity for

extracellular conversion of glucose to gluconic acid and 2-ketogluconic acid enables some

bacteria, including several species of Pseudomonas to sequester glucose effectively and gives

a competitive advantage over microorganisms that lack the ability to utilise these compounds

(Gottschalk, 1986).

Competition for trace elements, such as iron, copper, zinc, manganese etc., also occurs

in soils. For example, iron is an essential growth element for all living organisms and the

scarcity of its bio-available form in soil habitats results in a furious competition (Loper and

Henkels, 1997). Siderophores, low molecular weight compounds with high iron affinity, are

produced by some microorganisms (also by most biocontrol agents) to solubilize and

competitively acquire ferric ion under iron-limiting conditions, thereby making iron

unavailable to other soil microorganisms which cannot grow for lack of it (Loper and

Henkels, 1997; Haas and Défago, 2005). The bacterium that originally synthesized the

siderophores takes up the iron siderophore complex by using a receptor that is specific to the


17

complex and is located in the outer cell membrane. Suppression of soilborne plant pathogens

by siderophore producing Pseudomonads has been reported in some instances (Loper, 1988;

Weger et al., 1988; Buysens et al., 1996).

Induced resistance

Plant-associated bacteria can reduce the activity of pathogenic microorganisms not only

through microbial antagonisms, but also by activating the plant to better defend itself, a

phenomenon termed “induced systemic resistance” (ISR) (Shoda, 2000; van Loon, 2007).

Sometimes, the mechanism of ISR elicited by PGPR overlaps partly with that of pathogen-

induced systemic acquired resistance (SAR). Both ISR and SAR represent a state of enhanced

basal resistance of the plant that depends on signalling compounds such as jasmonic acid,

ethylene and salicylic acid (van Loon, 2007). Expression of natural defense reaction against

stresses from biotic or abiotic origin is exhibited by all plants, such as (i) physical stresses

(heat or frost), (ii) inoculation by pathogenic or non-pathogenic organisms, (iii) chemical

molecules from natural or synthetic origins (Alabouvette et al., 2006). Early recognition of

the aggressor by the plant is one of the mechanisms involved in elicitation of plant defense

reactions (Lugtenberg et al., 2002). Recognition of the aggressor immediately initiates a

cascade of molecular signals and the transcription of many genes, which eventually results in

the production of defence molecules by the host plant (van Loon, 2000). Such defence

molecules include phytoalexins, pathogenesis-related (PR) proteins (such as chitinases, β-1,3-

glucanases, proteinase inhibitors, etc.) and lignin for reinforcement of cell walls (van Loon,

2000). In fact, cell wall thickenings, wall appositions or rapid death of the injured plant cells

resulting in necrosis of the immediate adjacent tissues are barriers which cut the pathogen off

its nutrients and contribute to slowing down of the fungus progressive invasion (Lugtenberg

et al., 2002; Alabouvette et al., 2006).

Plant growth promotion

Phytostimulation

Phytostimulation enhances plant growth in a direct way. In the processes of plant growth,

phytohormones (e.g., production of indole-3-acetic acid (IAA), auxins, cytokinins, and

gibberellins) play an important role. These hormones can be synthesized by the plant

themselves but also by their associated microorganisms such as Azospirillum spp., besides

having nitrogen-fixing ability (Steenhoudt and Vanderleyden, 2000). Species of Pseudomonas

and Bacillus can produce as yet not well characterized phytohormones or growth regulators

that cause crops to have greater amounts of fine roots which have the effect of increasing the


18

absorptive surface of plant roots for uptake of water and nutrients. The phytohormones they

produce include indole-acetic acid, cytokinins, gibberellins and inhibitors of ethylene

production. Indole-3-acetic acid is a phytohormone which is known to be involved in root

initiation, cell division, and cell enlargement (Salisbury, 1994). This hormone is very

commonly produced by PGPRs (Barazani and Friedman, 2001). Auxins are quantitatively the

most abundant phytohormones secreted by Azospirillum, and it is generally agreed that their

production, rather than nitrogen-fixation, is the major factor responsible for the stimulation of

rooting and, hence, enhanced plant growth (Bloemberg and Lugtenberg, 2001). Furthermore,

plant-associated bacteria can influence the hormonal balance of the plant. Ethylene is an

important example to show that the balance is most important for the effect of hormones: at

low levels, it can promote plant growth in several plant species including Arabidopsis

thaliana, while it is normally considered as an inhibitor of plant growth and known as a

senescence hormone (Pierik et al., 2006).

The general effect on the plant can be direct, that is through plant growth promotion, or

indirect, that is through improving plant nutrition via the better development of the roots, and

it is difficult to distinguish between them. The elevation of root IAA level in lodgepole pine

plantlets, inoculated with Paenibacillus polymyxa, and, of dihydroxyzeatin riboside root

concentration in plants inoculated with Pseudomonas fluorescens (Fuentes-Ramirez and

Caballero-Mellado, 2005), might be attributed to the induction of plant hormone synthesis by

the bacteria. However, the uptake of bacterial synthesized phytohormones can not be

excluded, since both P. polymyxa and Pseudomonas produce IAA and cytokinins in vitro

(Fuentes-Ramirez and Caballero-Mellado, 2005).

Biofertilization

The mechanisms by which PGPR increase crop performance is not well understood. There are

several PGPR inoculants currently commercialized that seem to promote growth through at

least one mechanism; suppression of plant disease (termed bioprotectants), phytohormone

production (termed biostimulants), or improved nutrient acquisition (termed biofertilizers).

The mode of action of PGPR by biofertilizers act either, directly by helping to provide

nutrient to the host plant, or indirectly by positively influencing root growth and morphology

or by aiding other beneficial symbiotic relationships (Vessey, 2003). The most prominent

example is bacterial nitrogen fixation. The symbiosis between rhizobia and its legume host

plants is an important example for plant growth-promoting rhizobacteria (PGPR). Bacteria of

this group metabolize root exudates (carbohydrates) and in turn provide nitrogen to the plant


19

for amino acid synthesis. The ability to fix nitrogen also occurs in free-living bacteria like

Azospirillum, Burkholderia, and Stenotrophomonas (Dobbelare et al., 2003). Biofertilization

accounts for approximately 65% of the nitrogen supply to crops worldwide (Bloemberg and

Lugtenberg, 2001). Another nutrient is sulfate, which can be provided to the plant via

oxidation by bacteria (Banerjee and Yesmin, 2002). Bacteria may contribute to plant nutrition

by liberating phosphorous from organic compounds such as phytates and thus indirectly

promote plant growth (Unno et al., 2005). Azospirillum treatment resulted in enhancement of

root growth and activities (e.g., acidification of the root surroundings) that increases

phosphorous and other macroelements and microelements uptake (Dobbelaere and Okon,

2007). Recently, De Werra and collaborators (2009) showed that the ability of Pseudomonas

fluorescens CHA0 to acidify its environment and to solubilize mineral phosphate is strongly

dependent on its ability to produce gluconic acid.

Conclusions

The rhizosphere is the zone of soil surrounding a plant root where the biology and chemistry

of the soil are influenced by the root. As plant roots grow through soil they mostly release

water soluble compounds such as amino acids, sugars and organic acids that supply food for

the microorganisms. High levels of exudates in the rhizosphere attract a plethora of

microorganisms to a larger extend than elsewhere in the soil. The composition and pattern of

root exudates affect microbial activity and population numbers. Plant species, plant

developmental stage and soil type have been indicated as major factors determining the

composition of rhizosphere microbial communities (Broeckling et al., 2008). As shown in

many studies, there is no general decision about the key player: the diversity and

predominance of rhizosphere microbial population depend on a number of abiotic and biotic

factors of a particular ecological niche.

A better understanding of the basic principles of the rhizosphere ecology, including the

function and diversity of inhabiting microorganisms is on the way but further knowledge is

necessary to optimize soil microbial technology to the benefit of plant-growth and health in

the natural environment. In sum, this can constitute overwhelming evidence indicating that an

ever exploitation of plant growth promoting rhizobateria (PGPR) can be a true success story

in sustainable agriculture. As a consequence, current production methods in agriculture, e.g.,

the improper use of chemical pesticides and fertilizers creating a long list of environmental

and health problems, will reduce.


20

Acknowledgements Venant Nihorimbere is recipient of a grant from the Belgian Technical Cooperation (BTC/CTB). Marc Ongena is Research Associate at the F.R.S.-FNRS (National Funds for Scientific Research, Belgium).

References

Alabouvette C., Olivain C. & Steinberg C., 2006. Biological control of plant diseases: the European situation. Eur. J. Plant Pathol., 114, 329–341.

Asiegbu F.O. & Nahalkova J.L.G., 2005. Pathogen-inducible cDNAs from the interactionof the root rot fungus Heterobasidion annosum with Scots pine (Pinus sylvestris L.). Plant Sc., 168, 365–372.

Bais H.P., Weir T.L., Perry L.G., Gilroy S. & Vivanco J.M., 2006. The role of root exudates in rhizosphere interactions with plants and other organisms. Ann. Rev. Plant Biol., 57, 233-266.

Banerjee M. & Yesmin L., 2002. Sulfur-oxidizing plant growth promoting Rhizobacteria for enhanced canola performance. US Patent.

Barazani O. & Friedman J., 2001. Allelopathic bacteria and their impact on higher plants. Crit. Rev. Microbiol., 27, 41–55.

Berg G., 2009. Plant–microbe interactions promoting plant growth and health: perspectives for controlled use of microorganisms in agriculture. Appl. Microbiol. Biotechnol., 84, 11–18.

Bloemberg G.V. & Lugtenberg B.J.J., 2001. Molecular basis of plant growth promotion and biocontrol by rhizobacteria. Curr. Opin. Plant Biol., 4, 343–350.

Broeckling CD, Broz AK, Bergelson J, Manter DK, Vivanco JM. (2008). Root exudates regulate soil fungal community composition and diversity. Appl. Environ. Microbiol. 74, 738–744.

Boulter D., Jeremy J.J. & Wilding M., 1966. Amino acids liberated into the culture medium by pea seedling roots. Plant Soil, 24, 121-127.

Bowen G. & Rovira A., 1999. The rhizosphere and its management to improve plant growth. Adv. Agron., 66, 1-102.

Brimecombe M.J., De Leij F.A.A.M. & LynchJ.M., 2007. Rhizodeposition and microbial populations. In: Pinton R., Veranini Z. & Nannipieri P., eds. The rhizosphere biochemistry and organic substances at the soil-plant interface. Boca Raton, FL: CRC Press. pp. 73–109.

Broeckling C.D., Broz A.K., Bergelson J., Manter D.K. & Vivanco J.M., 2008. Root exudates regulate soil fungal community composition and diversity. Appl. Environ. Microbiol., 74, 738–744.

Buchanan R., Gruissem W. & Jones R.L., 2000. Biochemistry and molecular biology of plants. Rockville, MD, USA, American Society of Plant Biologists.

Buée M., De Boer W., Martin F., van Overbeek L. & Jurkevitch E., 2009. The rhizosphere zoo: An overview of plant-associated communities of microorganisms, including phages, bacteria, archaea, and fungi, and of some of their structuring factors. Plant Soil, 321, 189-212.

Buysens S., Heungens K., Poppe J. & Höfte M., 1996. Involvement of pyochelin and pyoverdin in suppression of Pythium-induced damping-off of tomato by Pseudomonas aeruginosa 7NSK2. Appl. Environ. Microbiol., 62, 865–871.

Cassán F. & García Salamone I., 2008. Azospirillum sp: cell physiology, plant response, agronomic and environmental research in Argentina. Asociación Argentina de Microbiología, Buenos Aires.

Chet I. & Chernin L., 2002. Biocontrol, microbial agents in soil. In: Bitton G, eds. Encyclopedia of environmental microbiology. New York, Willey, 450–465.

Chin-A-Woeng T.F.C., Bloemberg G.V. & Lugtenberg B.J., 2003. Phenazines and their role in biocontrol by Pseudomonas bacteria. New Phytol., 157, 503–523.

Compant S., Duffy B., Nowak J., Clement C. & Barka E.A., 2005. Use of plant growth-promoting bacteria for biocontrol of plant diseases: principles, mechanisms of action, and future prospects. Appl. Environ. Microbiol., 71, 4951–4959.

Davis D.A., Lynch H.C., & Varley J. 1999. The production of surfactin in batch culture by Bacillus subtilis ATCC 21332 is strongly influenced by the conditions of nitrogen metabolism. Enzyme Microb. Tech., 25, 322-329.


21

De Vleeschauwer D. & Höfte M., 2007. Using Serratia plymuthica to control fungal pathogens of plants. CAB Reviews, 2, 046. Doi: 10.1079/PAVANNR20072046.

De Werra P., Péchy-Tarr M., Keel C. & Maurhofer M. (2009). Role of gluconic acid production in the regulation of biocontrol traits of Pseudomonas fluorescens CHA0. Appl. Environ. Microbiol., 75, 4162–4174.

DeFlaun M.F. & Gerba C.P., 1993. Monitoring recombinant DNA microorganisms and viruses in soil. In: Metting FBJ, ed. Soil microbial ecology: Application in agricultural and environmental management. Washington, Marcel Dekker, Inc., 131-150.

Dobbelaere S. & Okon Y., 2007. The plant growth promoting effects and plant responses. In: Elmerich CWE Newton, eds. Nitrogen fixation: origins, applications and research progress. Associative and endophytic nitrogen-fixing bacteria and cyanobacterial associations. Heidelberg, Springer, 145–170.

Dobbelare S., Vanderleydern J. & Okon Y., 2003. Plant-growth promoting effects of diazotrophs in the rhizosphere. Crit. Rev. Plant Sci., 22, 107–149.

Dunne C., Moenne-Loccoz Y., de Bruijn F.J. & O’Gara F. (2000). Overproduction of an inducible extracellular serine protease improves biological control of Pythium ultimum by Stenotrophomonas maltophilia strain W81. Microbiology, 146, 2069–2078.

Emmert E.A.B. & Handelsman J., 1999. Biocontrol of plant disease: a (Gram-) positive perspective. FEMS Microbiol. Lett., 171, 1–9.

Foster R.C., 1988. Microenvironments of soil microorganisms. Biol. Fertil. Soils, 6, 189-203. Fuentes-Ramirez L.E. & Caballero-Mellado J., 2005. Bacterial biofertlizers. In: Siddiqui ZA, ed.

PGPR: Biocontrol and biofertilization The Netherlands, Springer, 143-172. Gottschalk G., 1986. Bacterial metabolism. Springer-Verlag Berlin and Heidelberg GmbH & Co. K. Grayston S.J., Wang S.Q., Campbell C.D. & Edwards A.C., 1998. Selective influence of plant species

on microbial diversity in the rhizosphere. Soil Biol. Biochem., 30, 369-378. Gupta R. & Mukerji K.G., 2002. Root exudate-biology. In: Mukerji K.G., Manaharachary C &

Chamola, B.P., eds. Techniques in mychorrhizal studies. Kluwer Academic Publishers, Dordrecht, The Neederland, 103-131.

Haas D. & Défago G., 2005. Biological control of soilborne pathogens by fluorescent Pseudomonads. Nat. Rev. Microbiol., 3, 307–319.

Harman G.E., Howell C.R., Viterbo A., Chet I. & Lorito M., 2004. Trichoderma species—opportunistic, avirulent plant symbionts. Nat. Rev. Microbiol., 2, 43–56.

Hartmann A., Schmid M., van Tuinen D. & Berg G., 2009. Plant-driven selection of microbes. Plant Soil, 321, 235-257.

Haynes R.J., 1990. Active ion uptake and maintenance of cation anion balance: A critical examination of their role in regulating rhizosphere pH. Plant Soil, 126, 247–264.

Heath M.C., 1981. A generalized concept of host-parasite specificity. Phytopathology, 71, 1121-1123. Hiltner L. 1904. Über neuere Erfahrungen und Probleme auf dem Gebiete der Bodenbakteriologie

unter besonderer Berücksichtigung der Gründüngung und Brache. Arb. Deutsch.Land.Ges., 98: 59–78.

Jacobsen B.J., Zidack N.K. & Larson B.J., 2004. The role of Bacillus-based biological control agents in integrated pest management systems: plant diseases. Phytopathology, 94, 1272–1275.

Jensen B., 1993. Rhizodeposition by 14C02-pulse-labelled spring barley grown in small field plots on sandy loam. Soil Biol. Biochem., 25, 1553-1559.

Jensen E.S., 1996. Rhizodeposition of N by pea and barley and its effects on soil N dynamics Soil Biol. Biochem., 28, 65-71.

Kamilova F., Validov S., Azarova T., Mulders I. & Lugtenberg B., 2005. Enrichment for enhanced competitive plant root tip colonizers selects for a new class of biocontrol bacteria. Environ. Microbiol., 7, 1809–1817.

Kent A.D. & Triplett E.W., 2002. Microbial communities and their interactions in soil and rhizosphere ecosystems. Ann. Rev. Microbiol., 56, 211-236.

Kloepper J.W. & Schroth M.N., 1978. Plant growth promoting rhizobacteria on radish. In: Phytobacteriologic SdPVe, eds. Proceedings of the Fourth Conference Plant Pathogenic Bacteria. Angers, INRA, 879–882.


22

Kochian L., Piñeros M. & Hoekenga O., 2005. The physiology, genetics and molecular biology of plant aluminum resistance and toxicity. Plant Soil., 274, 175–195.

Long S.R., 2001. Genes and signals in the Rhizobium-legume symbiosis. Plant Physiol., 125, 69-72. Loper J.E., 1988. Role of fluorescent siderophore production in biological control of Pythium ultirnum

by a Pseudomonas fluorescens strain. Phytopathology, 78, 166-172. Loper J.E. & Gross H., 2007. Genomic analysis of antifungal metabolite production by Pseudomonas

fluorescens Pf-5. Eur. J. Plant Pathol., 119, 265–278. Loper J.E. & Henkels M.D., 1997. Availability of iron to Pseudomonas fluorescens in rhizosphere and

bulk soil evaluated with an ice nucleation reporter gene. Appl. Environ. Microbiol., 63, 99–105. Lucy M., Reed E. & Glick B.R., 2004. Applications of free living plant growth promoting

rhizobacteria. A. Van Leeuw. Microb., 86, 1-25. Lugtenberg B.J.J., Chin-A-Woeng T.F.C. & Bloemberg G.V., 2002. Microbe–plant interactions:

principles and mechanisms. A. Van Leeuw. Microb., 81, 373–383. Lugtenberg B.J.J. & Dekkers L.C., 1999. What makes Pseudomonas bacteria rhizosphere competent.

Environ. Microbiol., 1, 9-13. Morgan J.A., Bending G.D. & White P.J., 2005. Biological costs and benefits to plant-microbe

interactions in the rhizosphere. J. Exp. Bot., 56, 1729–1739. Morgan J.A.W. & Whipps J.M., 2001. Methodological approaches to the study of rhizosphere carbon

flow and microbial population dynamics. Rhizosphere, 373-409. Moszer I., Jones L.M., Moreira S., Fabry C. & Danchin A. 2002. Subtilist: the reference database for

the Bacillus subtilis genome. Nucleic Acids Res. 30, 62-65. Müller H., Westendorf C., Leitner E., Chernin L., Riedel K., Schmidt S., Eberl L. & Berg G., 2009.

Quorum-sensing effects in the antagonistic rhizosphere bacterium Serratia plymuthica HRO-C48. FEMS Microbiol. Ecol., 67, 468–467.

Nannipieri P., Ascher J., Ceccherini M.T., Landi L., Pietramellara G. & Renella G., 2003. Microbial diversity and soil functions. Eur. J. Soil Sci. 54: 655–670.

Nunan N., Daniell T.J., Singh B.K., Papert A., McNicol J.W. & Prosser J.I., 2005. Links between plant and rhizoplane bacterial communities in grassland soils, characterized using molecular techniques. Appl. Environ. Microbiol., 71, 6784-6792.

Ongena M. & Thonart P., 2006. Resistance induced in plants by non-pathogenic microorganisms: elicitation and defense responses. In: Teixeira da Silva JA, ed. Floriculture, Ornamental and Plant Biotechnology: Advances and Topical Issues. London, Global Science Books, 447-463.

Paulitz T.C. & Bélanger R.R., 2001. Biological control in greenhouse systems. Annu. Rev. Microbiol., 39, 103-133.

Pierik R., Tholen D., Poorter H., Visser E.J.W. & Voesenek L.A.C.J., 2006. The Janus face of ethylene: Growth inhibition and stimulation. Trends Plant Sci, 11, 176–183.

Pinton R., Veranini Z. & Nannipieri P., 2007. The rhizosphere biochemistry and organic substances at the soil-plant interface. New York, Taylor & Francis Group, LLC

Plaxton W.C., 1996. The organization and regulation of plant glycolysis. Annu. Rev. Plant Physiol. Plant Mol. Biol., 47, 185–214.

Raaijmakers J.M., Paulitz T.C., Steinberg C., Alabouvette C. & Moënne-Loccoz Y., 2009. The rhizosphere: a playground and battlefield for soilborne pathogens and beneficial microorganisms. Plant Soil, 321, 341-361.

Raaijmakers J.M., Vlami M. & de Souza J.T., 2002. Antibiotic production by bacterial biocontrol agents. A. Van Leeuw. Microb., 81, 537–547.

Raaijmakers J.M. & Weller D.M., 2001. Exploiting genotypic diversity of 2,4-diacetylphloroglucinol-producing Pseudomonas spp.: Characterization of superior root-colonizing P. fluorescens strain Q8r1-96. Appl. Environ. Microbiol., 67, 2545-54.

Ratnayale M., Leonard R.T. & Menge A., 1978. Root exudation in relation to supply of phosphorus and its possible relevance to mycorrhizal infection. New Phytol, 81, 543–552.

Rizzo D.M., Garbelotto M. & Hansen E.A., 2005. Phytophthora ramorum: integrative research and management of an emerging pathogen in California and Oregon forests. Annu. Rev. Phytopathol., 43, 309–335.


23

Rudrappan T., Quinn W.J., Stanley-Wall N.R. & Bais H.P., 2007. A degradation product of the salicylic acid pathway triggers oxidative stress resulting in down-regulation of Bacillus subtilis biofilm formation on Arabidopsis thaliana roots. Planta, 226, 283-297.

Ryan R.P., Monchy S., Cardinale M., Taghavi S., Crossman L., Avison M.B., Berg G., van der Lelie D. & Dow J.M., 2009. Versatility and adaptation of bacteria from the genus Stenotrophomonas. Nat. Microbiol. Rev., 7, 514–525.

Salisbury F.B., 1994. The role of plant hormones. In: Wilkinson RE, eds. Plant–Environment Interaction. New York, Dekker, 39–81.

Schloter M., Lebuhn M., Heulin T. & Hartmann A., 2000. Ecology and evolution of bacterial microdiversity. FEMS Microbiol. Rev., 24, 647–660.

Schrey S.D. & Tarkka M.T., 2008. Friends and foes: Streptomycetes as modulators of plant disease and symbiosis. A. Van Leeuw. Microb., 94, 11–19.

Schroth M.N. & Hildebrand D.C., 1964. Influence of plant exudates on root-infecting fungi. Annu. Rev. Phytopathology, 2, 101-132.

Semenov A.M., van Bruggen A.H.C. & Zelenev V.V., 1999. Moving waves of bacterial populations and total organic carbon along roots of wheat. Microbial Ecol., 37, 116–128.

Shoda M., 2000. Bacterial control of plant diseases. J. Biosci. Bioeng., 89, 515–521. Singh G. & Mukerji K.G., 2006. Root exudates as determinant of rhizospheric microbial biodiversity.

In: Mukerji K.G, Manoharachary C. & Singh J., eds. Microbial activity in the rhizosphere. Germany, Springer-Verlag Berlin Heidelberg, 39-53.

Singh S., Ladha J.K., Gupta R.K., Bhushan L., Rao A.N., Sivaprasad S. & Singh P.P., 2007. Evaluation of mulching, intercropping with Sesbania and herbicide use for weed management in dry-seeded rice (Oryza sativa L.). Crop Protect., 26, 518-524.

Somers E., Vanderleyden J. & Srinivasan M., 2004. Rhizosphere bacterial signalling: a love parade beneath our feet. Crit. Rev. Microbiol., 30, 205–235.

Steenhoudt O. & Vanderleyden J., 2000. Azospirillum, a free-living nitrogen-fixing bacterium closely associated with grasses: genetic, biochemical and ecological aspects. FEMS Microbiol. Rev., 24, 487–506.

Stolp H., 1988. Microbial ecology: organisms, habitats, activities. New York, Cambridge university press

Suzuki M., Takahashi M., Tsukamoto T., Watanabe S., Matsuhashi S., Yazaki J., Kishimoto N., Kikuchi S., Nakanishi H., Mori S. & Nishizawa N.K., 2006. Biosynthesis and secretion of mugineic acid family phytosiderophores in zinc-deficient barley. Plant J., 48, 85–97.

Tournas V.H. & Katsoudas E., 2005. Mould and yeast flora in fresh berries, grapes and citrus fruits. Int. J. Food Microbiol., 105, 11-17.

Unno Y., Okubo K., Wasaki J., Shinano T. & Osaki M., 2005. Plant growth promotion abilities and microscale bacterial dynamics in the rhizosphere of lupin analysed by phytate utilization ability. Environ. Microbiol., 7, 396–404.

Uren N.C., 2001. Types, amounts and possible functions of compounds released into the rhizosphere by soil-grown plants. In: Pinton R, Varanini Z. & Nannipieri P., eds. The Rhizosphere. New York, Marcel Dekker, 19-40.

van Etten H.D., Mansfield J.W., Bailey J.A. & Farmer E.E., 1994. Two classes of plant antibiotics: Phytoalexins versus Phytoanticipins. Plant Cell, 6, 1191-1192.

van Loon J.C., 2000. Induced resistance. In: Slusarenko A.J. & FRSSav L.J.C., eds. Mechanisms of resistance to plant diseases. Dordrecht, NL., Kluwer Academic publishers, 521–574.

van Loon L.C., 2007. Plant responses to plant growth promoting bacteria. Eur. J. Plant Pathol., 119, 243–254.

Vessey J.K., 2003. Plant growth promoting rhizobacteria as biofertilizers. Plant Soil, 255, 571-586. Wacquant J.P., Ouknider M. & Jacquard P., 1989. Evidence for a periodic excretion of nitrogen by

roots of grass-legume associations. Plant Soil, 116, 57-68. Weger L.A., Arendonk J.J.C.M., Recourt K., Hofstad G.A.J.M., Weisbeek P.J. & Lugtenberg B.,

1988. Siderophore-mediated uptake of Fe3+ by the plant growth-stimulating Pseudomonas putida strain WCS358 and by other rhizosphere microorganisms. J. Bacteriol., 170, 4693-4698.

Welbaum G., Sturz A.V., Dong Z. & Nowak J., 2004. Fertilizing soil microorganisms to improve productivity of agroecosystems. Crit. Rev. Plant Sci., 23, 175–93.


24

Weller D.M., 1988. Biological control of soilborne plant pathogens in the rhizosphere with bacteria. Annu. Rev. Phytopathol., 26, 379–407.

Whipps J., 2001. Microbial interactions and biocontrol in the rhizosphere. J. Exp. Bot., 52, 487–511. Whipps J.M., 1990. Carbon economy. In: Lynch J.M., eds. The Rhizosphere. Chichester, Wiley &

Son, 59-87. Yang C.-H. & Crowley D.E., 2000. Rhizosphere microbial community structure in relation to root

location and plant iron nutritional status. Appl. Environ. Microbiol., 66, 345-351.


25

Chapter 2. Bacillus lipopeptides: structure, regulation and roles in biocontrol


26

Abstract

Lipopeptides (LPs) from plant-associated bacteria display versatile functions in the ecology of

the producing strains and notably in interactions with co-existing organisms, including

bacteria, fungi, oomycetes, protozoan predators and plants. The proposed primary mode of

action of LPs is pore formation in membranes leading to an imbalance in transmembrane ion

fluxes and cell death. LPs have demonstrated to play a crucial role in the antagonism

developed towards phytopathogens sharing the same microenvironment and thus in the

biocontrol potential of the strains. Beside their antimicrobial properties, LPs are also involved

in root colonization as well as in the systemic stimulation of host plant immune system. We

thus summarized here the current knowledge of LPs biocontrol activities.

A variety of structures

Cyclic lipopeptides (LPs) are synthesized by non-ribosomal peptide synthetases (NRPS) or

hybrid polyketide synthases/non-ribosomal peptide synthetases (PKS/NRPS). These modular

proteins are megaenzymes organized in iterative functional units called modules that catalyze

the different reactions leading to polyketide or peptide transformation (Finking and Marahiel,

2004). Such biosynthetic systems lead to a remarkable heterogeneity among the LPs products

which vary in the type and sequence of amino acid residues, the nature of the peptide

cyclisation and in the length and branching of the fatty acid chain (Ongena and Jacques,

2008). LPs are synthesized by various genera of plant-associated bacterial such as

Streptomyces, Pseudomonas and Bacillus (Pirri et al., 2009).

Surfactins Iturins Fengycins

Figure 1: Structures of representative members of the three families of Bacillus LPs. Source: Raaijmakers et al.

(2010).


27

The three main Bacillus LP families are surfactin, iturin and fengycin. They encompass

structural variants depending on the genetic background of the considered strain and in

several case environmental growth conditions (Figure 1 and Table 1).

Table 1: Updated list of Bacillus LPs identified so far*. Amino-acid residues in bold are involved in cyclisation.

Number

of aa Fatty acids Peptidic sequence

surfactin 7 C12-C14 Glu Leu Leu Val Asp Leu Leu

surfactin 7 C12-C14 Glu Leu Leu Val Asp Leu Val

surfactin 7 C12-C14 Glu Leu Leu Val Asp Leu Ile

bamylocin 7 C13 Glu Leu Met Leu Pro Leu Leu

pumilacidin A 7 C12-C14 Glu Leu Leu Leu Asp Leu Val

pumilacidin B 7 C12-C14 Glu Leu Leu Leu Asp Leu Ile

lichenysin 7 C12-C14 Gln Leu Leu Val Asp Leu Val

lichenisyn 7 C12-C14 Gln Leu Leu Val Asp Leu Ile

kurstakin 7 C10-C12 Thr Gly Ala Ser His Gln Gln

iturin A 7 C14-C16 Asn Tyr Asn Gln Pro Asn Ser

iturin C 7 C14-C16 Asp Tyr Asn Gln Pro Asn Ser

mycosubtilin 7 C16-C17 Asn Tyr Asn Gln Pro Ser Asn

bacillomycin D 7 C14-C16 Asn Tyr Asn Pro Glu Ser Thr

bacillomycin F 7 C16-C17 Asn Tyr Asn Gln Pro Asn Thr

bacillomycin L 7 C14-C16 Asp Tyr Asn Ser Gln Ser Thr

fengycin A 10 C15-C17 Glu Orn Tyr aThr Glu Ala Pro Gln Tyr Ile

fengycin B 10 C15-C17 Glu Orn Tyr aThr Glu Val Pro Gln Tyr Ile

plipastatin A 10 C15-C17 Glu Orn Tyr aThr Glu Ala Pro Gln Tyr Ile

plipastatin B 10 C15-C17 Glu Orn Tyr aThr Glu Val Pro Gln Tyr Ile

polymyxin 10 Dad Thr Dab Dab Dab Phe Leu Dab Dab Thr

*: Table adapted from works of Fernando et al. (2005) and others.

Surfactins are heptapeptides interlinked with a β-hydroxy fatty acid to form a cyclic lactone

ring structure. The group of iturins encompasses 7 variants including bacillomycins and

mycosubtilin. All are heptapeptides linked to a β-amino fatty acid chain with a length from

C14 to C17. The third family comprises fengycins A and B, which are also called plipastatins.

These molecules are lipodecapeptides with an internal lactone ring in the peptidic moiety and

with a β-hydroxy fatty acid chain (C14 to C18) that can be saturated or unsaturated. Beside

these three main families, other classes of bioactive lipopeptides synthesized by Bacillus

species have been identified such as kurstakins isolated from B. thuringiensis (Hathout et al.,


28

2000), polymyxins, the dodecacyclopeptide bacitracins and the surfactin-ressembling

bamylocin A (Lee et al., 2007).

Cellular regulatory network of LP biosynthesis

According to Duitman and collaborators (2007), there is no auto-induction process in the

biosynthesis of surfactins or iturins. However, there is relatively few information on the

regulatory aspects at the molecular level of the biosynthesis of Bacillus LPs compared to the

well-characterized structural and functional organization of their biosynthetic genes.

Regulation of srfA is typically multi-component and take part of a complex regulatory

cascade involved in multiple basal cellular processes that governs several differentiation

pathways in B. subtilis. This regulation at the cellular level is schematized in Figure 2A.

Transcription of the srfA gene is under the primary control of the quorum-sensing pheromone

ComX which is secreted with the help of ComQ in the culture medium. ComP senses the

accumulation of the ComX and at critical ComX concentration phosphorylates itself.

Subsequently autophosphorylated ComP phosphorylates ComA. ComA-P activates

expression of surfactin operon and ComS which is an open reading frame encoded within de

srfA mRNA. ComS leads to the autoactivation of ComK. In addition, several activators such

as ComK itself and DegU and repressors including CodY, AbrB and Rok are involved in

ComK activation. By analogy with the ComA/ComP regulon, DegU is a two-component

response regulator that controls many cellular processes in Bacillus (Dahl et al., 1992; Steil et

al. 2003). The second type of pheromones involved in surfactin expression is the group of Phr

peptides: PhrC, PhrF, PhrG and PhrH. They are probably first synthesized as small proteins

which are secreted, processed in pentapeptides and then internalised by an oligopeptide

permease (Opp). The phrH gene constitutes an operon with a rapH gene. Boths genes are thus

activated by ComK. Products of phr and rap genes are interacting regulators factors that

modulate the phosphorylation state of DNA specific response regulators. The phrC, phrF and

phrG genes need the sigmaH (Spo0H) form of RNA polymerase to be transcribed. RapC,

RapF and, in certain circumstances RapG and RapH, are inhibitors of the DNA binding of

ComA-P. ComA-P activates rapC and rapF. PhrC, PhrF, PhrG and PhrH inhibit their cognate

Rap proteins. RapG and RapH inhibit DNA binding of DegU. RghR is a rapG, rapH and

phrH repressor.

Expression of the mycosubtilin genes is under the influence of AbrB, one of the main

transition state regulators in B. subtilis but additional regulators may be involved (Figure 2B).


29

Figure 2. Regulation of surfactin (A) and iturin (B) biosynthesis in Bacillus. Source: Jacques (2011).

As for surfactin, the pheromone PhrC is seemingly involved in the regulation of mycosubtilin

biosynthesis (Duitman et al., 2007). For bacillomycin D, another member of the iturin family,


30

the synthesis is activated by interaction of the promoter with DegU in the early stationary

growth phase and DegQ is also involved (Koumoutsi et al., 2007). To our knowledge, little

information is available about the regulation of fengycin biosynthesis, although it was

demonstrated that expression of the plipastatin operon ppsABCDE requires at least degQ in

addition to the sfp gene (Tsuge et al., 2007).

A variety of biocontrol-related functions

Involvement in root colonization

The ability of B. subtilis to efficiently colonize surfaces of plant roots is a prerequisite for

phytostimulation. The first step of translocation on surfaces like roots, the spreading, can be

achieved through several ways. The probably most studied is the cell motility in colonies,

swarming, which involves differentiation of vegetative cells into hyperflagellated “swarmer

cells” (Fraser and Hughes, 1999). This swarming process allows an easier colony spreading

but also an improved antimicrobial substances resistance. The second step for rhizosphere

competence is linked to the capability to form sessile, highly structured and antimicrobial

resistant multicellular communities. Microbial populations such as plant-associated bacteria

evolve and behave as structured communities called biofilms (on solid surfaces) or pellicles

(at air/liquid interfaces) that could adhere to root and on soil particle surfaces (Danhorn and

Fuqua, 2007). Several studies showed that LPs could be involved at different levels in the

complex network linking motility and biofilms/pellicles/fruiting-bodies formation.

Surfactin is thought to act by the aggregation of the cells into dendrites and by the

coordination of their advance throughout the swarm front (Julkowska et al., 2005). The LPs

surfactin and mycosubtilin have been shown to be implicated in a flagella-independent

surface motility of B. subtilis (Kinsinger et al., 2003; Leclère et al., 2006). This surfactant-

induced spreading is most likely due to a reduction of frictions between cells and surface in

combination with a surface tension-driven flow. A low surface tension which could be

reached with strong surfactants such as surfactin and mycosubtilin was demonstrated to be

sufficient to facilitate microbial colonization (Leclère et al., 2006). This could explain why, in

some strains but not all (Figure 3), surfactin production is necessary but not sufficient for

swarming, in which other factors like genes swrABC and efp are additionally involved

(Kearns et al., 2004). Results obtained with B. subtilis A1/3 and B. amyloliquefaciens FZB42

(Figure 3) have shown that surfactin but no other lipopeptides produced by the strain was

required for the formation of biofilms and pellicles indicating that surfactins may still serve

specific developmental functions (Hofemeister et al., 2004). More conclusively in the view of


31

biocontrol, the production of surfactin was demonstrated to be essential for biofilm formation

and colonization of Arabidopsis roots by the strain B. subtilis 6051 and that biocontrol

exhibited against P. syringae is linked to the formation of this antibiotic at the root surface

(Bais et al., 2004).

FZB42 wtSrf+

Fen+

Bac+

AK3 deriv Srf+

Fen-

Bac-

CH1 deriv Srf-

Fen+

Bac-

CH2 derivSrf-

Fen-

Bac+

Figure 3: Illustration of the importance of surfactin for biofilm formation (up, as revealed by the blue coloration of cell aggregates in microtiter plate wells) and for motility (down, as revealed by the spreading ability on soft agar plates) of B. amyloliquefaciens FZB42. Source: Raaijmakers et al. (2010).

Finally, genes that mediate production of surfactin (srfAA and sfp) were shown to be required

for the erection of fruiting-bodies (Branda et al., 2001), at least in part by their ability to lower

the surface tension of water.

Involvement in direct antagonism

Each family of Bacillus LPs displays specific antibiotic activities and may thus be

differentially involved in the antagonism of the various plant pathogens. Reports suggesting a

role for LPs in antagonism based on in vitro studies (such as illustrated in Figure 4A) are

numerous but there are only few studies that associate biocontrol activity with LPs production

in planta. In the case of soilborne diseases, iturin A produced by B. subtilis RB14 was

involved in damping-off of tomato (a seedling disease) caused by Rhizoctonia solani (Asaka

and Shoda, 1996). Overexpression of mycosubtilin in B. subtilis ATCC 6633 also led to a

significant reduction of seedling infection by Pythium aphanidermatum (Leclère et al., 2005).

As examples in control of phyllosphere diseases, a contribution of both iturins and fengycins

was recently shown in the antagonism of B. subtilis toward Podosphaera fusca infecting

melon leaves (Figure 4B) (Romero et al., 2007). This was notably demonstrated by showing

the strong inhibitory effect of these LPs on P. fusca conidia germination, and by recovering

LPs from bacterial-treated leaves and using LP-deficient transformants. In the protection of


32

post harvest diseases, the strain Bacillus subtilis strain GA1 which efficiently produces LPs

from the three families and notably a wide variety of fengycins, protected wounded apple

fruits against gray mold disease caused by Botrytis cinerea. The role of fengycins was

demonstrated by the very effective disease control provided by treatment of fruits with LPs-

enriched extracts and by in situ detection of fengycins in inhibitory amounts (Touré et al.,

2004). To further illustrate the broad range of fungal targets, fengycins were also reported for

their antagonistic activity against Fusarium graminearum (Wang et al., 2007), and iturins for

their inhibitory effect towards the anthracnose-causing agent Colletotrichum demiatium

(Hiradate et al., 2002), Penicillium roqueforti (Chitara et al, 2003), Aspergillus flavus (Moyne

et al., 2001), Rhizoctonia solani (Yu et al., 2002), wood-staining fungi (Velmurugan et al.,

2009) and nematophagous fungi (Li et al., 2007). In some instances, the fungitoxic activity

was clearly related to the permeabilization of spore/conidia therefore inhibiting germination

or alternatively to hyphal cell perturbation.

- LP + LP

- LP + LP

A

B

C

s s

s s

Figure 4: Antagonistic effect of LPs at various stages of the development of phytopathogenic fungi illustrated by global growth reduction on solid medium (up), hyphal-induced swelling and collapse and conidia lysis (down) occurring within minutes after addition of LPs. Results obtained in our laboratory.

As revealed by transmission electron miscroscopy techniques, both phenomena most probably

result from membrane damaging by the LPs (Chitarra et al., 2003, Romero et al., 2007;

Etchegaray et al., 2008). LPs from Pseudomonas such as viscosinamide, massetolide,


33

putisolvin and orfamide also have significant impact on pathogenic fungi and oomycetes

either through their ability to lyse zoospores (Figure 4C) or to inhibit hyphal growth

(Raaijmakers et al., 2010). A few studies have revealed some insecticide activity of LPs from

B. subtilis. Surfactin and iturin were described for their antagonistic effect against fruit fly

Drosophila melanogaster (Assie et al., 2002) and LPs contained in a crude extract were

efficient at inhibiting the development of larvae of the mosquito Culex quinquefasciatus (Das

and Mukherjee, 2006). Thought that active doses are quite high (approx. 200 µM) and that

mechanisms underpinning such biocidal effect have not been investigated,, treatment with

LPs are presented as possible alternative to the use of the endotoxin producer B. thuringenesis

in the biocontrol of insect disease for which this bacterium is not efficient.

Like other antimicrobial peptides, LPs are not only membrane disruptive but can also

directly or indirectly act on intracellular targets and inhibit some functions such as enzymatic

activity. An example is the inhibition of phospholipase A2 in target fungal cells

(Vanittanakom et al., 1986). Whether this inhibition results from a direct interference with

enzyme structure/stability within the plasma membrane or from some repression of associated

gene expression is not clear. However, that may have some important metabolic consequences

since products released by this enzyme from membrane phospholipids are involved in

multiple cellular processes among which the synthesis of bioactive products (Köhler et al.,

2006). In the recent report on inhibition of toxic fumonisin B1 synthesis in Fusarium

verticillioides, it has been suggested that some products deriving from phospholipase A2

activity may be somewhere involved in the down-regulation of toxin genes transcriptions

observed upon treatment with fengycin (Hu et al., 2009). Whether the action of fengycin in

the inhibition of toxin production by Fusarium is direct or not, results from such studies allow

to extend application of LPs or their producing strains to the food safety sector. Such potential

has also recently been illustrated by the involvement of iturin in the inhibition of aflatoxin

producing Aspergillus isolates (Cho et al., 2009).

Bacterial LPs as elicitors of plant defense mechanisms

Another well established way for some beneficial rhizobacteria to provide plant protective

effect is through the stimulation of the plant immune system (Bakker et al., 2007). These

isolates are indeed able to reduce disease through the stimulation of a priming state in host

plant which allows an accelerated activation of defense responses upon pathogen attack,

leading to an enhanced resistance to the attacker encountered (Conrath et al., 2006). This

induced systemic phenomenon (ISR) can be globally viewed as a three-step process involving


34

sequentially, i) the perception by plant cells of elicitors produced by the inducing agents that

initiates the phenomenon, ii) signal transduction that is needed to propagate the induced state

systemically through the plant and iii) expression of defense mechanisms stricto sensu that

limit or inhibit pathogen penetration into the host tissues.

The potential of LPs as plant resistance inducers was demonstrated in 2007 for two

different molecules synthesized by Pseudomonas and Bacillus. Tran and collaborators

showed that massetolide A produced by Pseudomonas fluorescens retains ISR-eliciting

activity in tomato plants for the control of Phytophthora infestans, the causal agent of late

blight. It was demonstrated by testing the purified compound and by testing mutants impaired

in their production, in typical ISR assays where the inducing agent and the pathogen remain

spacially separated (Figure 5).

C LPs

Figure 5: Illustration of the ISR phenomenon based on disease reduction observed on leaves of various plants after treatment at the root level with lipopeptide (over)producing Bacillus isolates. Source: results obtained in our laboratory. C: control plants, LPs: treated plants. Results obtained in our labortory.

In bean, pure surfactins and fengycins to a lower extent but not iturins provided a significant

induced protective effect similar to the one induced by living cells of the producing strain (B.

amyloliquefaciens S499). In a complementary approach, experiments conducted on bean and

tomato showed that overexpression of both surfactin and fengycin biosynthetic genes in the

naturally poor producer B. subtilis strain 168 was associated with a significant increase in the

potential of the derivatives to induce resistance (Ongena et al., 2007). Moreover, the

macroscopic disease reduction induced by the surfactin overproducer was associated with


35

defense-related metabolic changes in the host plant tissues (Ongena et al., 2007). Treatment

of tobacco cell suspensions with surfactin but not fengycin nor iturin, also induced some

defence-related early events such as phosphorylation, Ca2+-dependent extracellular

alkalinisation and oxidative burst without causing any significant cell death (Jourdan et al.,

2009). The same study also showed that perception of surfactin by tobacco cells resulted in

the stimulation of defense pathways initiated by the enzymes phenylalanine ammonia lyase

(PAL) and lipoxygenase (LOX). It is thus clear that the three lipopeptide families retain

differential abilities to stimulate defense reactions in different plant species.

Other applications for Bacillus LPs

Because of their potential to disturb the integrity of the biological membranes, cyclic

lipopeptides secreted by Bacillus and Pseudomonas are potentially toxic for auxiliary

microflora and higher microorganisms. However, based on some recent data reported here, it

is now obvious that LPs cannot be simply viewed as molecular pneumatic drills creating

irreversible pores, cytosolic leakage and cell death. Despite similar global structures and

apparent physico-chemical traits, not all LPs do provoke lysis of a given cell target and it also

became clear that some lipopeptides may induce some modifications in basic cell processes

without causing any detrimental leakage in the plasma membrane.

These compounds thus retain other natural functions beside those directly involved in

biocontrol of plant diseases. These properties schematized in Figure 6 not only favour the

developmental processes of the producing isolates in their natural habitat in terms of growth

and niche colonization but may also strongly influence the interactions with the other

organisms sharing the microenvironment.

Besides plants, surfactins could also interfere with several stages of the immune

processes in animals. They indeed display lipopolysaccharide-binding and neutralizing

activities (Hwang et al., 2005; Takahashi et al., 2006) and antitumor activity. This activity is

based on apoptosis induction via cellular Ca2+ augmentation, induction of MAPK and related

up/down regulation of protein factors associated with apoptosis (Wang et al., 2007). Another

in depth study demonstrated anti-proliferative activity of surfactin on human colon carcinoma

cell line related to induction of apoptosis and cell cycle arrest (Kim et al., 2007). Surfactin

also retains some anti-inflammatory (Huang et al., 2007) properties and can also inhibit

platelet aggregation (Kim et al., 2006). Finally these LPs could inhibit AMPc

phosphodiesterase, alkaline phosphatase activities (Bortolato et al., 1997) and either inhibit or


36

enhance phospholipases A2 depending on their origin (Kim et al., 1998). These activities

open the way to interesting biomedical and pharmaceutical applications (Pirri et al., 2009).

Colonization, biofilm

Antiviral

Antiprotozoal

Antibacterial

Antioomycete Antifungal

Chelation,solubilization

MotilityVirulence,Immunization

Figure 6. Natural functions of lipopeptides secreted by plant-associated bacteria with special emphasis on those involved in biological control of plant diseases (in bold). Source: figure adapted from Raaijmakers et al. (2010).

Finally, some of these surface-active LPs are also considered as among the most

powerful biosurfactants isolated so far. They retain exceptional emulsifying/foaming

properties and high potential for the solubilization of hydrophobic compounds. LPs have

thereby received considerable attention for various environmental applications such as

bioremediation of soils polluted by oils or recalcitrant compounds.


37

References

Asaka, O. and M. Shoda (1996). "Biocontrol of Rhizoctonia solani damping-off of tomato with Bacillus subtilis RB14." Appl. Environ. Microbiol. 62: 4081-4085.

Assie, L. K., M. Deleu, L. Arnaud, M. Paquot, P. Thonart, C. Gaspar and E. Haubruge (2002). "Insecticide activity of surfactins and iturins from a biopesticide Bacillus subtilis Cohn (S499 strain)." Meded Rijksuniv Gent Fak Landbouwkd Toegep Biol. Wet. 67: 647–655.

Bais, H. P., R. Fall and J. M. Vivanco (2004). "Biocontrol of Bacillus subtilis against infection of Arabidopsis roots by Pseudomonas syringae is facilitated by biofilm formation and surfactin production." Plant Physiol. 134: 307–319.

Bakker, P. A. H. M., C. M. J. Pieterse and L. C. Van Loon (2007). "Induced systemic resistance by fluorescent Pseudomonas spp." Phytopathology 97: 239-243.

Bortolato, M., F. Besson and R. B. (1997). "Inhibition of alkaline phosphatise by surfactin, a natural chelating lipopeptide from Bacillus subtilis." Biotechnol. Lett. 19: 433–435.

Branda, S. S., J. E. Gonzalez-Pastor, S. Ben-Yehuda, R. Losick and R. Kolter (2001). "Fruiting body formation by Bacillus subtilis." P. Natl. Acad. Sci. U.S.A. 98: 11621–11626.

Chitarra, G. S., P. Breeuwer, M. J. R. Nout, A. C. Van Aelst, F. M. Rombouts and T. Abee (2003). "An antifungal compound produced by Bacillus subtilis YM 10–20 inhibits germination of Penicillium roqueforti conidiospores." J. Appl. Microbiol. 94: 159–166.

Cho, K. M., R. K. Math, S. Y. Hong, S. M. A. Islam, D. K. Mandanna, J. J. Cho, M. G. Yun, J. M. Kim and H. D. Yun (2009). "Iturin produced by Bacillus pumilus HY1 from korean soybean sauce (kanjang) inhibits growth of aflatoxin producing fungi." Food Control 20: 402–406.

Conrath, U., G. Beckers, V. Flors, P. Garcia-Agustin, G. Jakab, F. Mauch, M. Newman, C. Pieterse, B. Poinssot, M. Pozo, A. Pugin, U. Schaffrath, J. Ton, D. Wendehenne, L. Zimmerli and B. Mauch-Mani (2006). "Priming: getting ready for battle." Mol. Plant Microbe Interact. 19: 1062-1071.

Dahl, M. K., T. Msadek, F. Kunst and G. Rapoport (1992). "The phosphorylation state of the DegU response regulator acts as a molecular switch allowing either degradative enzyme-synthesis or expression of genetic competence in Bacillus subtilis." J. Biol. Chem. 267: 14509–14514.

Danhorn, T. and C. Fuqua (2007). "Biofilm formation by plant associated bacteria." Annu. Rev. Microbiol. 61: 401–422.

Das, K. and A. K. Mukherjee (2005). "Characterization of biochemical properties and biological activities of biosurfactants produced by Pseudomonas aeruginosa mucoid and non-mucoid strains." 69: 192–195.

Duitman, E. H., D. Wyczawski, L. G. Boven, G. Venema, O. P. Kuipers and L. W. Hamoen (2007). "Novel methods for genetic transformation of natural Bacillus subtilis isolates used to study the regulation of the mycosubtilin and surfactin synthetases." Appl. Environ. Microb. 73: 3490–3496.

Etchegaray, A., C. de Castro Bueno, I. S. de Melo, S. M. Tsai, M. de Fátima Fiore, M. E. Silva-Stenico, L. A. B. de Moraes and O. Teschke (2008). "Effect of a highly concentrated lipopeptide extract of Bacillus subtilis on fungal and bacterial cells." Arch. Microbiol. 190: 611–622.

Fernando, W. G. D., S. Nakkeeran and Y. Zhang (2005). "Biosynthesis of antibiotics by PGPR and its relation in biocontrol of plant diseases." In: Z. A. Siddiqui (ed.), PGPR: Biocontrol and biofertilization. Springer, The Netherlands, 67-109.

Finking, R. and M. A. Marahiel (2004). "Biosynthesis of nonribosomal peptides1." Annu. Rev. Microbiol. 58: 453-488.

Fraser, G. M. and C. Hughes (1999). "Swarming motility." Curr. Opin. Microbiol. 2: 630–635. Hathout, Y., Y. P. Ho, V. Ryzhov, P. Demirev and C. Fenselau (2000). "Kurstakins: a new class of

lipopeptides isolated from Bacillus thuringiensis." J. Nat. Prod. 63: 1492–1496. Hiradate, S., S. Yoshida, H. Sugie, H. Yada and Y. Fujii (2002). "Mulberry anthracnose antagonists

(iturins) produced by Bacillus amyloliquefaciens RC-2." Phytochemistry 61: 693–698. Hofemeister, J., B. Conrad and B. Adler (2004). "Genetic analysis of the biosynthesis of non-

ribosomal peptide- and polyketidelike antibiotics, iron uptake and biofilm formation by Bacillus subtilis A1/3." Mol. Genet. Genomics 272: 363–378.


38

Hu, X., G. Van der Auwera, S. Timmery, L. Zhu and J. Mahillon (2009). "Distribution, diversity, and potential mobility of extrachromosomal elements related to the Bacillus anthracis pXO1 and pXO2 virulence plasmids." Appl. Environ. Microbiol. 75: 3016-3028.

Huang, X. Q., Z. X. Lu, X. M. Bie, F. X. Lu, H. Z. Zhao and S. J. Yang (2007). "Optimization of inactivation of endospores of Bacillus cereus by antimicrobial lipopeptides from Bacillus subtilis fmbj strains using a response surface method." Appl. Microbiol. Biot. 74: 454–461.

Hwang, M. S. H., R. L. Morgan, S. F. Sarkar, P. W. Wang and D. S. Guttman (2005). "Phylogenetic characterization of virulence and resistance phenotypes of Pseudomonas syringae." Appl. Environ. Microb. 71: 5182–5191.

Jacques P. (2011). Surfactin and other lipopeptides from Bacillus sp. In: Gloria Soberon-Chavez (Ed.), Biosurfactant: from to gene to applications. Springer-Verlag, Berlin, 57-92.

Jourdan, E., G. Henry, F. Duby, J. Dommes, J. P. Barthelemy, P. Thonart and M. Ongena (2009). "Insights into the defense-related events occurring in plant cells following perception of surfactin-type lipopeptide from Bacillus subtilis." Mol. Plant Microbe Interact. 22: 456–468.

Julkowska, D., M. Obuchowski, I. B. Holland and S. J. Seror (2005). "Comparative analysis of the development of swarming communities of Bacillus subtilis 168 and a natural wild type: critical effects of surfactin and the composition of the medium." J. Bacteriol. 187: 65–76.

Kearns, D. B., F. Chu, R. Rudner and R. Losick (2004). "Genes governing swarming in Bacillus subtilis and evidence for a phase variation mechanism controlling surface motility." Mol. Microbiol. 52: 357–369.

Kim, J. J., D. J. Krupa and R. F. Thompson (1998). "Inhibitory cerebello-olivary projections and blocking effect in classical conditioning." Science 279: 570–573.

Kim, S. D., S. K. Park, J. Y. Cho, H. J. Park, J. H. Lim, H. I. Yun, S. C. Park, K. Y. Lee, S. K. Kim and M. H. Rhee (2006). "Surfactin C inhibits platelet aggregation." J. Pharm. Pharmacol. 58: 867–870.

Kim, S. Y., J. Y. Kim, S. H. Kim, H. J. Bae, H. Yi, S. H. Yoon, B. S. Koo, M. Kwon, J. Y. Cho, C. E. Lee and S. Hong (2007). "Surfactin from Bacillus subtilis displays anti-proliferative effect via apoptosis induction, cell cycle arrest and survival signaling suppression." FEBS Lett. 581: 865–871.

Kinsinger, R. F., M. C. Shirk and R. Fall (2003). "Rapid surface motility in Bacillus subtilis is dependent on extracellular surfactin and potassium ion." J. Bacteriol. 185: 5627–5631.

Köhler, K., A. Meister, B. Dobner, S. Drescher, F. Ziethe, A. Blume. (2006). Temperature-Dependent Aggregation Behavior of Symmetric Long-Chain Bolaamphiphiles at the Air-Water Interface. Langmuir 22: 2668-2675.

Koumoutsi, A., X. Chen, J. Vater and R. Borriss (2007). "DegU and YczE positively regulate the synthesis of bacillomycin D by Bacillus amyloliquefaciens strain FZB42". Appl. Environ. Microb. 73: 6953–6964.

Leclère, V., M. Bechet, A. Adam, J. S. Guez, B. Wathelet, M. Ongena, P. Thonart, F. Gancel, M. Chollet-Imbert and P. Jacques (2005). "Mycosubtilin overproduction by Bacillus subtilis BBG100 enhances the organism's antagonistic and biocontrol activities". Appl. Environ. Microbiol. 71: 4577-4584.

Leclère, V., R. Marti, M. Bechet, P. Fickers and P. Jacques (2006). "The lipopeptides mycosubtilin and surfactin enhance spreading of Bacillus subtilis strains by their surface-active properties." Arch. Microbiol. 186: 475–483.

Lee, S. C., S. H. Kim, I. H. Park, S. Y. Chung and Y. L. Choi (2007). "Isolation and structural analysis of bamylocin A, novel lipopeptide from Bacillus amyloliquefaciens LP03 having antagonistic and crude oil-emulsifying activity." Arch. Microbiol. 188: 307–312.

Li, L., M. H. Mo, Q. Qu, H. Luo and K. Q. Zhang (2007). "Compounds inhibitory to nematophagous fungi produced by Bacillus sp strain H6 isolated from fungistatic soil." Eur. J. Plant Pathol. 117: 329–340.

Moyne, A. L., R. Shelby, T. E. Cleveland and S. Tuzun (2001). "Bacillomycin D: an iturin with antifungal activity against Aspergillus flavus." J. Appl. Microbiol. 90: 622–629.

Ongena, M. and P. Jacques (2008). " Bacillus lipopeptides: versatile weapons for plant disease biocontrol." Trends Microbiol. 16: 115-125.


39

Ongena, M., E. Jourdan, A. Adam, M. Paquot, A. Brans, B. Joris, J.-L. Arpigny and P. Thonart (2007). "Surfactin and fengycin lipopeptides of Bacillus subtilis as elicitors of induced systemic resistance in plants." Environ. Microbiol. 9: 1084–1090.

Pirri, G., A. Giuliani, S. F. Nicoletto, L. Pizzuto and A. C. Rinaldi (2009). "Lipopeptides as anti-infectives: a practical perspective." Cent. Eur. J. Biol. 4: 258–273.

Raaijmakers, J. M., I. De Bruijn, O. Nybroe and M. Ongena (2010). "Natural functions of lipopeptides from Bacillus and Pseudomonas: more than surfactants and antibiotics." FEMS Microbiol. Rev. 34: 1037-1062.

Romero, D., A. de Vicente, R. H. Rakotoaly, S. E. Dufour, J.-W. Veening, E. Arrebola, F. M. Cazorla, O. P. Kuipers, M. Paquot and A. Pérez-García (2007). "The iturin and fengycin families of lipopeptides are key factors in antagonism of Bacillus subtilis toward Podosphaera fusca." Mol. Plant Microbe Interact. 20: 430-440.

Steil, L., T. Hoffmann, I. Budde, U. Volker and E. Bremer (2003). "Genome-wide transcriptional profiling analysis of adaptation of Bacillus subtilis to high salinity." J. Bacteriol. 185: 6358–6370.

Takahashi, T., O. Ohino, Y. Ikeda, Y. Homma, M. Igarashi and K. Umuzawa (2006). "Inhibition of lipopolysaccharide activity by a bacterial cyclic lipopeptide surfactin." J. Antibiot. (Tokyo) 59: 35-43.

Touré, Y., M. Ongena, P. Jacques, A. Guiro and P. Thonart (2004). "Role of lipopeptides produced by Bacillus subtilis GA1 in the reduction of grey mould disease caused by Botrytis cinerea on apple. J Appl Microbiol 96: 1151–1160." J. Appl. Microbiol. 96: 1151–1160.

Tran H., H. Tran, A. Ficke, T. Asiimwe, M. Höfte and J. M. Raaijmakers (2007). Role of the cyclic lipopeptide massetolide A in biological control of Phytophthora infestans and in colonization of tomato plants by Pseudomonas fluorescens. New Phytol. 175: 731–742.

Tsuge, K., K. Matsui and M. Itaya (2007). "Production of the non-ribosomal peptide plipastatin in Bacillus subtilis regulated by three relevant gene blocks assembled in a single movable DNA segment." J. Biotechnol. 129: 592–603.

Vanittanakom, N., W. Loeffler, U. Koch and G. Jung (1986). "Fengycin-a novel antifungal lipopeptide antibiotic produced by Bacillus subtilis F-29-3." J. Antibiot. (Tokyo) 39: 888–901.

Velmurugan, N., M. S. Choi, S. S. Han and Y. S. Lee (2009). "Evaluation of antagonistic activities of Bacillus subtilis and Bacillus licheniformis against wood-staining fungi: in vitro and in vivo experiments." J. Microbiol. 47: 385–392.

Wang, J., J. Liu, H. Chen and J. Yao (2007). "Characterization of Fusarium graminearum inhibitory lipopeptide from Bacillus subtilis IB." Appl. Microbiol. Biot. 76: 889–894.

Yu, G. Y., J. B. Sinclair, G. L. Hartman and B. L. Bertagnolli (2002). "Production of iturin A by Bacillus amyloliquefaciens suppressing Rhizoctonia solani." Soil Biol. Biochem. 34: 955–963.


40

Chapter 3. In situ production of Bacillus antibiotics involved in the biocontrol of plant diseases: more evidences

needed

Nihorimbere Venant, Thonart Philippe and Ongena Marc

To be submitted as review paper after including our contribution to the topic (Chapters 5 and 6)


41

Abstract

Interest in biological control of plant pathogens has been stimulated in recent years by trends

in agriculture towards greater sustainability and public concern about the use of hazardous

pesticides. There is now unequivocal evidence that antimicrobial compounds play a key role

in the suppression of various soilborne plant pathogens by antagonistic microorganisms. The

significance of antibiotics in biocontrol, often has been questioned because of the indirect

nature of the supporting evidence and the perceived constraints to antibiotic production in

rhizosphere environments. Reporter gene systems and bio-analytical techniques have

demonstrated that antibiotics are effectively produced in the rhizosphere of a variety of host

plants, although in limited quantities. Several abiotic factors such as oxygen, temperature,

specific carbon and nitrogen sources, and microelements have been shown to influence

antibiotic production by bacteria biocontrol agents. Among the biotic factors that may play a

determinative role in antibiotic production are the plant host, the pathogen, the indigenous

microflora, and the cell density of the producing strain. Alternatives to antimicrobials, such as

biocontrol agents of Bacillus spp., are being developed and marketed but slowly because of

their higher production cost. A discussion of how and why these antibiotics are produced in

rhizosphere by soil bacterial biocontrol agents is given.

Bacillus as biopesticide

Biological control through the use of natural antagonistic microorganisms has emerged as a

promising alternative to chemical pesticides for more rational and safe crop management

(Montesinos, 2007). There is a large body of literature reporting the beneficial effect of

rhizosphere-associated bacteria for the stimulation of plant growth and for the reduction of

diseases caused by phytopathogens (Lucy et al., 2004; Lugtenberg and Kamilova, 2009).

Most of the bacterial strains exploited to date as biopesticides belong to the genera

Agrobacterium, Bacillus and Pseudomonas (Flavel, 2005). By taking benefits from the

nutrients exuded by the host plant, beneficial strains efficiently colonize root systems and the

surrounding soil layer (rhizosphere). In turn, they beneficially influence the plant. From a

very global point of view, improvement in plant health/productivity may be mediated via two

different mechanisms e.g. protection against pest and pathogens and promotion of host

nutrition and growth. The beneficial protective effect of these bacterial agents may rely on

different mechanisms. Antibiosis through the production of antifungal metabolites and

antibiotics is probably the best known and most important mechanism used by biocontrol

bacteria to limit pathogen invasion in host plant tissues. Beside this direct growth inhibitory


42

activity, other widely recognized mechanisms involved in biocontrol mediated by rhizosphere

bacteria are: competition for an ecological niche/substrate, production of inhibitory

allelochemicals, and induction of systemic resistance (ISR) in host plants to a broad spectrum

of pathogens (Figure 1) (Mukerji et al., 2006).

Direct antagonism through the production of highly active antibiotics is supposed to

play a pivotal role in the multiptrophic interactions taking place in the phytosphere and

thereby in the global protective effect against phytopathogens (Stallings, 1954; Thomashow et

al., 2008). The exploitation of these natural antagonistic interactions has been a driving force

in research on the biological control of plant pathogens over the past century. Some species of

the Bacillus genus such as B. subtilis or B. amyloliquefaciens retain the potential to produce

many different antibiotics (Koumousti et al., 2004; Stein, 2005). Several strains belonging to

these species were reported effective for the biocontrol of multiple plant diseases caused by

soilborne (Chen et al., 2009) or post-harvest pathogens (Kotan et al., 2009). The production

of antimicrobial agents thus appears to be an important criterion in the evaluation of the

potential use of Bacillus strains as biocontrol.

Figure 1. Beneficial effects of Bacillus subtilis on plants

They are numerous reports suggesting a role for some specific antibiotics in antagonism

based only on in vitro antimicrobial activity displayed by the producing strain. However, it

just highlights some intrinsic potential of that strain to synthesize antibiotics under laboratory

conditions. One needs to consider at least two other approaches to provide evidence for the

direct involvement of antibiotic production in beneficial bacteria-mediated disease


43

suppression. First by using either biosynthesis suppressed mutants that are no longer able to

prevent phytopathogens to cause damage in plants or antibiotic-overproducer isolates that

protect plants against pathogens more effectively than the wild type strain. Second, by

showing that the antibiotic applied in a pure form is also effective at reducing disease.

Furthermore, providing evidence for an efficient in situ production is a third approach that has

to be combined with the others to unambiguously demonstrate the involvement of a particular

compound in the global biocontrol activity of the producing strain. In soil, rhizosphere-

specific factors may strongly impact the physiology of the biocontrol strain developing on

roots. It is thus crucial to demonstrate that the antibiotic is actually secreted and may

accumulate in the microenvironment in biologically relevant quantities.

This review aims at showing that only a very limited number of studies have

successfully linked antibiotic-based biocontrol with efficient in planta production of the

compound. Despite many recent conceptual and technical advances, the investigation of

antibiotic gene expression during interaction with plants and of their secretion by Bacillus

isolates growing in the phytosphere is still in its infancy. This is however crucial for our

global knowledge of Bacillus fitness in its ecological niche regarding antibiotics synthesis.

More particularly, it is also important to understand the biotrophic interactions occurring in

the micro-environment which is pivotal for optimizing biocontrol strategies using these

organisms.

The Bacillus antibiotic arsenal to combat plant diseases

In the case of Bacillus, members of multiple species such as B. amyloliquefaciens, B. subtilis,

B. cereus, B. licheniformis, B. megaterium, B. mycoides, and B. pumilus have been reported

for developing antibiotic-based antagonism toward a number of phytopathogens (Choudhary

and Jorhi, 2009). Bacillus subtilis has an average of 4-5% of its genome devoted to antibiotic

synthesis and has the potential to produce more than two dozen structurally diverse

antimicrobial compounds (Stein, 2005; Ongena and Jacques, 2008). In strain FZB42, which is

proposed as paradigm for plant-associated Bacillus amyloliquefaciens, an even larger part of

the genome (~8%) is seemingly involved in antibiotic synthesis.

Among the vast array of biologically active molecules synthesized by Bacillus, some

have been reported for their inhibitory activity against plant pathogens (Figure 2). B. cereus

UW85 produces two fungistatic antibiotics, zwittermicin A and kanosamine, that are

suggested to contribute to the suppression of damping-off disease of alfalfa caused by


44

Phytophthora medicaginis (Silo-Suh et al., 1994). Zwittermicin A may also control the fruit

rot of cucumber (Smith et al., 1993) and suppress other plant diseases (Silo-Suh et al., 1998).

A B C

D

Figure 2: Some antibiotics (other than lipopeptides) from Bacillus with proven or putative biocontrol-related activity. A: 3,3-Neotrehalosadiamine, B: Bacilysin, C: Amicoumacin, D: Ericin S. The figure reproduced from Stein (2005).

Polyketides are synthesized via PKS modularly organized assembly lines that require specific

domains and follows the same logic as in NRP systems involved in the synthesis of peptidic

compounds (Chapter 2). Bacillaene, difficidin and macrolactin are polyketides displaying

broad spectrum antibacterial activities and that may be involved in the biocontrol activity of

the producing strain such as in the case of fire blight, a serious disease of orchard trees caused

by Erwinia amylovora (Chen et al., 2009). The phosphono-oligopeptide rhizocticin produced

by B. subtilis displays some antifungal and nematicidal but no bactericidal properties.

Peptide compounds represent the predominant class of Bacillus antibiotics. They are of

various sizes, may be composed entirely of amino acids but some contain other residues. They

are synthesized either by ribosomal or non-ribosomal mechanisms. Cyclic or linear

oligopeptides, basic peptides and aminoglycoside antibiotics usually occur (Stein, 2005). Low

molecular weight and hydrophobic or cyclic structures, with unusual constituents like D-

amino acids, are also common characteristics of peptide antibiotics normally synthesized by

Bacillus. Moreover, they are generally resistant to hydrolysis by peptidases and proteases of

animal and plant origins (Katz and Demain, 1977). Bacillus brevis (Brevibacillus brevis) and

Bacillus polymyxa (PaeniBacillus polymyxa) produce gramicidin S and polymyxin B peptide

antibiotics that strongly inhibited in vitro the phytopathogen Botrytis cinerea at the various

stages of germination, growth and extra-cellular enzyme production. These antibiotics also


45

exhibited significant biocontrol activity under natural field conditions against the Botrytis

grey mould on strawberry (Haggag, 2008). More simple molecules such as the dipeptide

bacilysin (L-Ala linked to the non-proteinogenic amino acid L-anticapsin) also retain strong

bactericidal effect and are seemingly involved in the control of some phytopathogens (Chen et

al., 2009). Another group of peptide antibiotics usually produced by Bacillus subtilis is

composed by lanthionines with inter-residual thioether bonds that are formed through post-

translational modification of ribosomally synthesized precursor peptides (Stein, 2005). These

compounds display strong antibacterial properties mostly against gram-positive bacteria but

their involvement in the biocontrol activity of plant-associated Bacillus isolates has not been

clearly demonstrated so far. However, some lantibiotics (subtilin and ericin) also display other

functions than pure antibiotic activity. They are autoregulated via two-component regulatory

systems and function as pheromones for quorum-sensing in some B. subtilis strains (Stein,

2005).

Lipopeptides as crucial ingredients for biocontrol

As described in Chapter 2, a major class of peptide antibiotics in the context of biological

control are lipopeptides (LPs) of the surfactin, iturin and fengycin families. According to our

current knowledge, these lipopeptides are by far the most important antibiotic compounds

involved in multitrophic interactions occurring in the plant microenvironment and thereby

they are probably the most investigated for their role in the global biocontrol activity of

specific strains from Bacillus species. They are composed of a lipid tail linked to a cyclised or

partially cyclised oligopeptide and can be synthesized by various species of Bacillus. Some of

these surface-active compounds are considered as among the most powerful biosurfactants

isolated so far. LPs actually display versatile functions in the ecology of the producing strains

in the phytosphere and notably in interactions with co-existing organisms, including bacteria,

fungi, oomycetes, protozoan, predators and plants. When produced by plant beneficial

rhizobacteria, LPs have thus demonstrated to play a crucial role in the antagonism developed

towards phytopathogens and thus in the biocontrol potential of the strains (Chapter 2). In this

case, the proposed primary mode of action of LPs is pore formation in membranes leading to

an imbalance in transmembrane ion fluxes and cell death (Baltz, 2009). Beside their

antimicrobial properties, LPs are also involved in root colonization as well as in the systemic

stimulation of host plant immune system. The current knowledge of LPs activities directly

involved in biocontrol is summarized in Figure 3. Beside these traits, LPs may also have

different roles in the development and survival of B. subtilis strains in their natural habitat.


46

These include increasing of the surface area and bioavailability of hydrophobic water-

insoluble substrates, heavy metal binding, bacterial pathogenesis, quorum-sensing, motility

and biofilm formation (Mulligan, 2005; Ron and Rosenberg, 2001; Raajmakers et al., 2010).

Surfactins Iturins Fengycins

Powerful biosurfactants,

Virucidal activity,

Bactericidal activity,

Strong fungitoxic activity (yeastand filamentous fungi),

Strong fungitoxic activity(filamentous fungi)

Root colonization(biofilm formation, surface motility)

Direct antagonism toward phytopathogens

Involvement in plant systemic resistance,elicitation

Figure 3: Specific biocontrol-related activities for the three families of Bacillus lipopeptides.

Linking basic cellular regulatory processes and environmental conditions affecting lipopeptide production Even if few information is currently available concerning regulation of fengycin biosynthesis,

it has become clear that LP production by Bacillus is modulated at the gene level by complex

regulatory events (presented in more details in Chapter 2 for surfactin and iturin). As a global

picture, regulation of surfactin genes is typically multi-component and takes part of a complex

regulatory cascade involved in multiple basal cellular central processes and that governs

several differentiation pathways in B. subtilis. LP production is growth-phase dependent and

may be linked to population-driven responses such as quorum-sensing via the ComX

pheromone in the case of surfactin. Two-component response regulators that control many

cellular processes like transition state are also involved in fine tuning the expression of

biosynthesis genes for both three families.

On another hand, plants roots usually release specific soluble carbon compounds as well

as pH and redox-modulating factors, complexing agents (siderophores, phenols,

carboxylates), antimicrobials (antibiotics, quorum-sensing inhibitors) and specific stimulatory

compounds. They may also physically shape specific habitat conditions and the combination


47

of all these events results in the so-called rhizosphere effect. In other words, plants usually

create a specific environment for microbial development in the soil surrounding roots

(Chapter 1). This may results in a selection of particular microbial populations and/or in a

drastic modification of the cellular physiology of established species associated with roots

(Hartmann et al., 2009). This rhizosphere effect depends both qualitatively and quantitatively

on the plant species, cultivar and even developement stage. Other environmental factors

inherent to the soil ecology are the co-existence of an auxiliary microflora sharing the

ecosystem, the soil type and heterogeneity and other abiotic parameters such as humidity, pH,

temperature or dissolved oxygen availability. All these rhizosphere factors may potentially

impact on the expression of biocontrol traits by plant beneficial strains and corollary on the

biosynthesis of specific metabolites responsible for these activities like Bacillus lipopeptides.

The influence of some environmental factors may be quite easily estimated under well-

controlled laboratory conditions in flask-based cultures or by using bioreactors in fed-batch or

chemostat modes. By contrast with batch culture where the exhaustion of a specific nutrient

terminates the exponential growth phase, biomass concentration in the chemostat culture is

usually controlled and maintained stable in concentration by the permanent limitation of a

single defined nutrient (soluble iron in this case). It allows the study of metabolite synthesis

upon fixed growth rate and constant cellular physiological state. Here are summarized some

works performed in vitro which provided some valuable information about the impact of

rhizosphere-specific factors on the secretion of Bacillus LPs.

Carbon and energy resources are scarce in bulk soil and, consequently, metabolic levels

in microorganisms are low. Microorganisms in the phytosphere depend on substrates liberated

from the root or shoot for their growth and consequently, the host plant greatly influences the

quantity and composition of indigenous microorganism populations as well as the expression

of antibiotic biosynthetic genes. This was illustrated by Milner and collaborators (1996) who

showed that production of the antibiotic kanosamine in Bacillus cereus was enhanced by

more than 300% by the addition of alfalfa seedling exudates to the culture medium. The

surfactin-apparented lychenysin also accumulates differentially in the culture broth upon

growth in various media differing in C source (Li et al., 2008). Akpa and collaborators (2001)

have also demonstrated the influence of the nature of carbon source on the pattern of

surfactins but not mycosubtilins produced by B.subtilis. This low flexibility in the

mycosubtilin homologue eventail could be related to the rather high selectivity of the domain

responsible for fixation and loading of the fatty acid in the mycosubtilin synthetase (Hansen et

al., 2007). Nutrient status may also be viewed in terms of availability of some essential


48

oligoelements and improvement of surfactin production was obtained upon elevated Mg, K,

Mn as well as Fe cations concentrations in the medium (Wei and Chu, 2002; Wei et al., 2004;

2007).

At least for surfactin, a dependence of Bacillus LP synthesis to the nitrogen metabolism

has been evidenced as the production clearly increased under dissimilatory nitrate metabolism

compared to the aerobic conditions of ammonium consumption (Davis et al., 1999). This

study also showed the influence of oxygen concentration in the surrounding medium on LP

production rate since nitrate may be assimilated in both aerobic and anaerobic conditions.

However, the question remains open in how far these two parameters (N status and O2) are

indissociable in Bacillus cells to globally influence LP production through the form of

nitrogen metabolism or alternatively if they may trigger their own regulatory processes

independently. Nevertheless, a positive influence of O2 limitation on production of surfactin

has also been demonstrated by modifying the oxygenation method (Lee et al., 2007) or by

modulating the oxygen transfer rate in bioreactor (Yeh et al., 2006). Such a process appears

not to be adapted to Bacillus subtilis S499, which produced higher surfactin yield in better

aeration conditions (Hbid et al., 1996; Jacques et al., 1999). Production of lichenysin by

Bacillus licheniformis was also conducted in anaerobic conditions. By contrast, mycosubtilin

expression is oxygen dependent (Guez et al., 2008). These results should be considered with

the view that a low oxygen status prevails in the rhizosphere of many soils (Hojberg et al.,

1999).

On another hand, root-colonizing rhizobacteria such as Bacillus or Pseudomonas

typically develop in microcolonies or form biofilms and their actual physiology is thus

probably not related to a planktonic state. Few studies have accurately evaluated the

consequences of growing on solid support on LP synthesis by B. subtilis but it is clear that at

least surfactin and iturin productions are efficient in swarming/spreading or colony-forming

cells (Leenders et al., 1999; Debois et al., 2008). In another approach, Gancel and

collaborators (2009) have also demonstrated an enhancement of surfactin production upon

cell immobilization on polymer particles.

In the same context of basic cellular processes, Bacillus cells on roots most obviously

evolve in a nutrient-starved state that would drive cell physiology to slow growth rate. Such a

slow-growth-imposed state may be more favorable for surfactin synthesis compared to higher

growth rates based on the expression level of the surfactin srfA biosynthetic genes in Bacillus

subtilis (Kakana, 2005).


49

Other physico-chemical parameters such as soil moisture, temperature and pH may also

drastically influence LP synthesis in Bacillus (Georgakopoulos et al., 1994; Cosby et al.,

1998; Fickers et al., 2009). Higher temperatures (37 °C) favoured surfactin synthesis of

Bacillus subtilis RB14 isolated from compost (Ohno et al., 1995) and ATCC6633, but not of

Bacillus subtilis S499 (Jacques et al., 1999). A 30-fold increase in mycosubtilin production

was observed when the temperature was decreased from 37 °C to 25 °C. This was observed

for both strain ATCC6633 and its derivative BBG100, which is a constitutive mycosubtilin

overproducer. However, the results suggest that the observed phenotype might originate from

a higher mycosubtilin synthetase turnover at lower temperature rather than reflecting changes

in the regulatory mechanisms per se (Fickers et al., 2008). The relative abundance of the

various co-produced mycosubtilin homologues was also affected by temperature. Surfactin

gene expression is pH-dependent.

Many rhizosphere-related abiotic parameters may thus impact the biosynthesis of

lipopeptides from Bacillus spp. This is certainly due to the complex regulation of lipopeptide

operon expression involving several pleiotropic regulators. However, intracellular pools of

synthetase substrates (amino acid residues and fatty chain) and turn-over of the enzymes have

also to be taken into account.

Evidence for in situ lipopeptide production by Bacillus: so few examples!

Studies conducted in vitro may thus be useful to evaluate the influence of a number of factors

prevailing in the natural environment. However, some conditions such as very specific soil

characteristics or exogenous micro- or macroflora with putative competing effect, are very

difficult to re-create in artificial systems. The influence of some parameters inherent to the

rhizosphere is thus merely not possible to assess under artificial conditions without

introducing some bias. The study of antibiotic production by Bacillus in natural conditions for

plant cultivation is thus necessary and retains the advantage of being conducted in an

integrated system where all the parameters evoked above are interplaying. Results presented

here deal almost exclusively with lipopeptides, illustrating our inability to find any work

reporting consistent study of in situ production of other Bacillus antibiotics in the

phytosphere.

Direct detection

As an illustration of the analytical power of the most recent and optimized mass spectrometry

technologies, matrix assisted laser desorption ionization-time of flight MS was used to detect

and identify the whole array of Bacillus LPs and their biosynthetic intermediates, not only in


50

culture filtrate but also at the whole cell level in a very rapid and sensitive way (Vater et al.,

2002; 2009). A qualitative and quantitative image of the surfactin homologue patterns

produced by Bacillus subtilis under various experimental conditions has also been obtained by

coupling HPLC with electrospray ionization MS (Li et al., 2008). Other studies that used

MALDI-TOF mass spectrometry to very precisely compare the lipopeptide products

synthesized by various B. subtilis strains also confirm that the potential to produce surfactins

and iturins is widespread in this species (Athukorala et al., 2009; Stein, 2008; Price et al.,

2007). Also time of flight secondary ion mass spectrometry (TOF-SIMS) imaging appears to

be a yet emerging but extremely powerful technology to study the production of lipidic

molecules by microbes developing at the surface of solid supports. It was recently used by

Debois and collaborators (2008) to study the heterogeneity of surfactin secretion in a

swarming pattern of Bacillus cells.

HPLC and mass spectrometry or a combination thereof have also been used successfully

for identification and quantification of antibiotics directly in rhizosphere extracts. However,

by contrast with the multiple examples available for antibiotics from other bacterial genera

such as phenazines or DAPG formed by Pseudomonas sp., only few biocontrol-related

antibiotics synthesized by Bacillus species have been studied for their in situ production.

Surfactin and/or iturin were recovered in biologically relevant quantities from cucumber roots

surrounding soil inoculated with Bacillus strain QST173 (Kinsela et al., 2009) and from

tomato rhizosphere inoculated with strain BGS3 (Ongena et al., 2007) and strain RB14-C

(Asaka and Shoda, 1996). Fengycins and iturins were also occasionally detected in the

phyllosphere and on fruits, illustrating that growth conditions on these plant tissues are

conducive for their biosynthesis at least by some strains (Romero et al., 2007; Touré et al.,

2004). In all cases, accumulation of LPs in or around plant tissues was actually correlated

with disease suppression.

For Bacillus isolates, there is a competitive advantage for efficient co-production of

surfactin, fengycin and iturin with specific roles and targets in the context of biocontrol of

plant diseases (Figure 2) (Ongena and Jacques, 2008; Raaijmakers et al., 2010). Several

strains efficiently co-produces the three lipopeptide families under well defined and optimized

in vitro conditions. However, almost nothing is known about the actual physiology of Bacillus

cells developing on plant and how far it may impact on the lipopeptide signature secreted in

situ. Also, whether in situ productivities may generate rhizospheric LP concentrations

sufficient to provide any biological effect remain questionable. These molecules are generally

active at 2-30 µM depending on the role and target organism considered. Interestingly,


51

quantities corresponding to micromolar concentrations of both surfactin and iturin were

detected in the cucumber rhizosphere colonized by strain QST713 (Kinsella et al., 2009) but

additional examples of consistent in vivo LPs secretion by other beneficial Bacillus strains are

required.

Inherent difficulties for evaluation of actual antibiotic concentration in soil

The limited number of reports on in situ production of bacterial antibiotics may reflect the

inherent difficulties in detecting or quantifying these small-size compounds in complex

environments like rhizosphere. This is particularly true for molecules such as LPs that can

tightly be adsorbed on the surface of soil particles because of their amphiphilic character or

because of unspecific, rapid and irreversible embedment in lipid structures at the surface or in

the cytosol of target and non-target organisms present in the diffusion zone. Moreover, the

stability of LPs and other peptide antibiotics in complex environments is questionable. The

cyclic nature and the alternance of L/D configured amino acids not only provide some

conformational advantages but should also enhance the level of resistance to degradation by

proteolytic enzymes from other organisms sharing the same niche. However, several works

obviously suggest that such decreased susceptibility to enzymatic degradation is only limited.

For Bacillus, Asaka and Shoda (1996) showed that iturin levels declined within 25 days in

sterile soil, while surfactin levels remained stable. The authors suggested that iturin may have

been degraded by microorganisms or leached from soil by watering, or was not detectable due

to irreversible binding to soil particles (Asaka and Shoda, 1996). It was also shown that an

endoprotease from Staphylococcus aureus hydrolyzed surfactin, thereby generating a linear

peptide (Grangemard et al., 1999). In addition, Nitschke and Pastore (2004) suggested that

protease activity contributed to a decrease of the surfactin levels in culture medium of B.

subtilis strain 21332. It globally means that antibiotic amounts that can be recovered from the

phytosphere may be limited and one needs to optimize the extraction procedures and use the

most powerful and sensitive detection methods for an accurate estimation of the quantities

that may be actually secreted by root colonizing Bacillus.

Gene expression

On another hand, the detection of antibiotic biosynthetic genes provides insight into the

distribution of antibiotic producers in nature and serves as a first indication that antibiotics

may be present (Thomashow et al., 2008). However, PCR-based molecular detection of such

antibiotic producers has inherent limitations keeping in mind that those genes may be present

but not all readily expressed under in vitro or natural conditions (Athukorala et al., 2009;


52

Joshi and McSpadden Gardener, 2006). In that context, the demonstration of an efficient

expression of biosynthetic genes in the environment is more informative and must also be

considered in support to direct detection of the antibiotic to provide conclusive evidence for in

situ production. Reporter gene systems are widely used as a marker to monitor populations of

introduced strains and occasionally to provide information on the transcriptional activity of

specific biocontrol-related antibiotic genes (Loper and Lindow, 1997; Thomashow et al.,

2008). However, it was mainly done for other bacterial genera and to our best knowledge,

there was no published work concerning Bacillus using some gene reporter system to provide

information on the transcriptional activity of genes coding for LPs on roots yet tools do exist.

Unstable variants of green fluorescent protein (gfp) have been used successfully as reporters,

and the corresponding gfp genes are fused to bacterial genes of interest, allowing gene

expression to be observed in situ (Leveau and Lindow, 2002; Haas and Keel, 2003). In

contrast to conventional reporters such as the stable enzyme β-galactosidase, which can only

monitor increases in gene expression, unstable gfp can report both “on” and “off” states. All

together, these molecular approaches are fundamentally indirect and subject to limitations, but

they can provide a degree of sensitivity not achieved by direct analysis.

Next to these genetic approaches, immunological assays have been proposed as sensitive

approach to detect and quantify antibiotics in situ, especially in plant associated environments

where relatively low amounts are produced and where plant-derived compounds may interfere

with chemical detection. For example, immunological detection of Bacillus LPs is still in its

infancy, but has been successfully adopted to monitor in situ production of the LP

syringopeptin by Pseudomonas pv. lachrymans, the causal agent of angular leaf spot. The

competitive ELISA assay developed by Fogliano and colleagues appeared to be

approximatively 100 times more sensitive than HPLC analysis and did not require extraction

of plant material with organic solvents (Fogliano et al., 1999). Specific antibodies will also be

highly instrumental to study the localization of LPs or other antibiotics in complex

environments or to monitor their fate and stability after application to soil, plant tissues or

other habitats.

Conclusion

The overall lack of information on LP gene expression during plant interaction is in part due

to the fact that Bacillus has been primarily studied for its biochemical and genetic traits. By

contrast with some other plant beneficial bacteria such as Pseudomonas, research on Bacillus

as a biocontrol agent has focussed mainly on the protective effect under practical use in the


53

field or greenhouse. Several Bacillus strains have great potential for biological control of

various plant diseases. However, there is still crucial lack of data on the level of antibiotic

production in the natural environments. This would be also helpful for a better understanding

of the complex biotrophic interactions occurring in the soil micro-environment but also more

globally, for Bacillus agricultural fitness.

References

Akpa E., Jacques P., Wathelet B., Paquot M., Fuchs R., Budzikiewicz H. & Thonart P. (2001). Influence of culture conditions on lipopeptide production by Bacillus subtilis. Appl. Bioch. Biotechn. 93: 551-561.

Asaka O. & Shoda M. (1996). Biocontrol of Rhizoctonia solani damping-off of tomato with Bacillus subtilis RB14. Appl. Environ. Microbiol. 62: 4081-4085.

Athukorala S.N.P., Dilantha F.W.G. & Rashid K.Y. (2009). Identification of antifungal antibiotics of Bacillus species isolated from different microhabitats using polymerase chain reaction and MALDI-TOF mass spectrometry. Can. J. Microbiol. 55: 1021-1032.

Baltz R.H. (2009). Daptomycin: mechanisms of action and resistance, and biosynthetic engineering. Curr. Opin. Chem. Biol. 13: 144–151.

Chen X., Koumoutsi A., Scholz R., Schneider K., Vater J., Sussmuth R., Piel J. & Borriss R. (2009). Genome analysis of Bacillus amyloliquefaciens FZB42 reveals its potential for biocontrol of plant pathogens. J. Biotechn. 140: 27-37.

Choudhary D.K. & Johri B.N. (2009). Interactions of Bacillus spp. and plants- with special reference to induced systemic resistance (ISR). Microbiol. Res. 164: 493-513.

Cosby W.M., Vollenbroich D., Lee O.H. & Zuber P. (1998). Altered srf expression in Bacillus subtilis resulting from changes in culture pH is dependent on the Spo0K oligopeptide permease and the comQX system of extracellular control. J. Bacteriol. 180: 1438-1445.

Davis D.A., Lynch H.C. & Varley J. (1999). The production of surfactin in batch culture by Bacillus subtilis ATCC 21332 is strongly influenced by the conditions of nitrogen metabolism. Enzyme Microb. Tech. 25: 322-329.

Debois D., Hamze K., Guérineau V., Le Caër J.P., Holland I.B., Lopes P., Ouazzani J, Séror SJ, Brunelle A & Laprévote O. (2008). In situ localisation and quantification of surfactins in a Bacillus subtilis swarming community by imaging mass spectrometry. Proteomics 8: 3682–3691.

Fickers P., Guez J-S., Damblon C., Leclère V., Béchet M., Jacques P. & Joris B. (2009). High-level biosynthesis of the anteiso-C17 isoform of the antibiotic mycosubtilin in Bacillus subtilis and characterization of its candidacidal activity. Appl. Environ. Microbiol. 75: 4636-4640.

Fickers P., Leclère V., Guez J., Bechet M., Coucheney F., Joris B. & Jacques P. (2008). Temperature dependence of mycosubtilin homologue production in Bacillus subtilis ATCC6633. Res. Microbiol. 159: 449–457.

Fogliano V., Gallo M., Vinale F., Ritieni A., Randazzo G., Greco G., Lops R. & Graniti A. (1999). Immunological detection of syringopeptins produced by Pseudomonas syringae pv. lachrymans. Physiol. Mol. Plant Pathol. 55: 255–261.

Fravel D.R. (2005). Commercialization and implementation of biocontrol. Annu. Rev. Phytopathol. 43: 337-359.

Gancel F., Montastruc L., Liu T., Zhao L. & Nikov L. (2009). Lipopeptide overproduction by cell immobilization on iron-enriched light polymer particles. Process Biochem. 44: 975-978.

Georgakopoulos D.G., Hendson M., Panopoulos N.J. & Schroth M.N. (1994). Analysis of expression of a phenazine biosynthesis locus of Pseudomonas aureofaciens PGS12 on seeds with a mutant carrying a phenazine biosynthesis locus-ice nucleation reporter gene fusion. Appl. Environ. Microbiol. 60: 4573-4579.


54

Grangemard I., Bonmatin J.-M., Bernillon J., Das B.C., & Peypoux F. (1999). Lichenysins G, a novel family of lipopeptide biosurfactants from Bacillus licheniformis IM 1307. J Antibiotics 52:363–373.

Guez J.S., Müller C.H., Danze P.M., Büchs J. & Jacques P. (2008). Respiration activity monitoring system (RAMOS), an efficient tool to study the influence of the oxygen transfer rate on the synthesis of lipopeptide by Bacillus subtilis ATCC6633. J. Biotechn. 134: 121-126.

Haas D. & Keel C. (2003). Regulation of antibiotic production in root-colonizing Peudomonas spp. and relevance for biological control of plant disease. Annu. Rev. Phytopathol. 41: 117-153.

Haggag W.M. (2008). Isolation of bioactive antibiotic peptides from Bacillus brevis and Bacillus polymyxa against Botrytis grey mould in strawberry. Arch. Phytopath. Plant Protect. 41: 477-491.

Hansen D.B., Bumpus S.B., Aron Z.D., Kelleher N.L. & Walsh C.T. (2007). The loading module of mycosubtilin: an adenylation domain with fatty acid selectivity. J. Am. Chem. Soc. 129: 6366–6367.

Hartmann A., Schmid M., Van Tuinen D. & Berg G. (2009). Plant-driven selection of microbes. Plant Soil 321: 235–257.

Hhid C. (1996). Contribution à l'étude de la relation entre la structure des lipopeptides de Bacillus subtilis et leurs activités hémolytique et antifongique. Ph.D. Thesis. University of Liège, Liège, Belgium.

Højberg O., Schnider U., Winteler H.V., Sørensen J. & Haas D. (1999). Oxygen-sensing reporter strain of Pseudomonas fluorescens for monitoring the distribution of low-oxygen habitats in soil. Appl. Environ. Microbiol. 65: 4085–4093.

Jacques P., Hbid C., Destain J., Razafindralambo H., Paquot M., De Pauw E. & Thonart P. (1999). Optimization of biosurfactant lipopeptide production from Bacillus subtilis S499 by Plackett–Burman design. Appl. Biochem. Biotechnol. 77: 223–233.

Joshi R. & McSpadden Gardener B.B. (2006). Identification and characterization of novel genetic markers associated with biological control activities in Bacillus subtilis. Phytopathology 96: 145-154.

Kakana P. (2005). Etude de la biosynthèse des lipopeptides par Bacillus subtilis et de leurs potentialités d’utilisation dans la lutte biologique au Burundi. Ph.D. Thesis. Agricultural University of Gembloux, Gembloux, Belgium.

Katz E. & Demain A.L. (1977). The peptide antibiotics in Bacillus: chemistry, biogenesis and possible functions. Bacteriol. Rev. 41: 449-474.

Kinsella K., Schulthess C.P., Morris T.F. & Stuart J.D. (2009). Rapid quantification of Bacillus subtilis antibiotics in the rhizosphere. Soil Biol. Biochem. 41: 374–379.

Kotan R., Dikbas N. & Bostan H. (2009). Biological control of post harvest disease caused by Aspergillus flavus on stored lemon fruits. Afr. J. Biotechn. 8: 209-214.

Koumoutsi A., Chen X.-H., Henne A., Liesegang H., Hitzeroth G., Franke P., Vater J. & Borriss R. (2004). Structural and functional characterization of gene clusters directing nonribosomal synthesis of bioactive cyclic lipopeptide in Bacillus amyloliquefaciens strain FZB42. J. Bacteriol. 186: 1084–1096.

Lee S.C., Kim S.H., Park I.H., Chung S.Y. & Choi Y.L. (2007). Isolation and structural analysis of bamylocin A, novel lipopeptide from Bacillus amyloliquefaciens LP03 having antagonistic and crude oil-emulsifying activity. Arch. Microbiol. 188: 307–312.

Leenders F., Stein T.H., Kablitz B., Franke P. & Vater J. (1999). Rapid typing of Bacillus subtilis strains by their secondary metabolites using matrix-assisted laser desorption/ionization mass spectrometry of intact cells. Rapid Commun. Mass Spectrom. 13: 943–949.

Leveau J.H.J. & Lindow S.E. (2002). Bioreporters in microbial ecology. Curr. Opin. Microbiol. 5: 259–265.

Li Y.M., Haddad N.I.A., Yang S.Z. & Mu B.Z. (2008). Variants of lipopeptides produced by Bacillus licheniformis HSN221 in different medium components evaluated by a rapid method ESI-MS. Int. J. Pept. Res. Ther. 14: 229-235.

Loper J.E. & Lindow S.E. (1997). Reporter gene systems useful in evaluating gene expression by soil- and plant-associated bacteria. In: Hurst CJ, Knudsen, G.R., McInerney, M.J., Stetzenbach, L.D., Walter, M.V., eds. Manual of Environmental Microbiology. ASM Press, DC, Washington.


55

Lucy M., Reed E. & Glick B.R. (2004). Application of free living plant growth promoting rhizobacteria. A. Van Leeuw. Microb. 86: 1-25.

Lugtenberg B. & Kamilova F. (2009). Plant growth-promoting rhizobacteria. Annu. Rev. Microbiol. 63: 541–556.

Milner J.L., Stohl E.A. & Handelsman J. (1996). Zwittermicin A resistance gene from Bacillus cereus. J. Bacteriol. 78: 4266-4272.

Montesinos E. (2007). Antimicrobial peptides and plant disease control. FEMS Microbiol. Lett. 270: 1-11.

Mukerji K.G., Manoharachary C. & Singh J. (2006). Microbial Activity in the Rhizosphere. Springer-Verlag, Berlin Heidelberg.

Mulligan C.N. (2005). Environmental applications for biosurfactants. Environ. Pollut. 133: 183–198. Nitschke M. & Pastore G.M. (2004). Biosurfactant production by Bacillus subtilis using cassava-

processing effluent. Appl. Biochem. Biotechn. 112: 163-172. Ohno A., Ano T. & Shoda M. (1995). Effect of temperature on production of lipopeptide antibiotics,

iturin A and surfactin, by a dual producer, Bacillus subtilis RB14, in solid-state fermentation. J. Ferment. Bioeng. 80:517–519.

Ongena M. & Jacques P. (2008). Bacillus lipopeptides: versatile weapons for plant disease biocontrol. Trends Microbiol. 16: 115-125.

Ongena M., Jourdan E., Adam A., Paquot M., Brans A., Joris B., Arpigny J.-L. & Thonart P. (2007). Surfactin and fengycin lipopeptides of Bacillus subtilis as elicitors of induced systemic resistance in plants. Environ. Microbiol. 9: 1084–1090.

Price N.P.J., Rooney A.P., Swezey J.M., Perry E. & Cohan F.M. (2007). Mass spectrometric analysis of lipopeptides from Bacillus strains isolated from diverse geographical locations. FEMS Microbiol. Lett. 271: 83-89.

Raaijmakers J.M., de Bruijn I., Nybroe O. & Ongena M. (2010). Natural functions of lipopeptides from Bacillus and Pseudomonas: more than surfactants and antibiotics. FEMS Microbiol. Rev. 34: 1034-1062.

Romero D., de Vicente A., Rakotoaly R.H., Dufour S.E., Veening J.-W., Arrebola E., Cazorla F.M., Kuipers O.P., Paquot M. & Pérez-García A. (2007). The iturin and fengycin families of lipopeptides are key factors in antagonism of Bacillus subtilis toward Podosphaera fusca. Mol. Plant Microbe Interact. 20: 430-440.

Ron E.Z. & Rosenberg E. (2001). Natural roles of biosurfactants. Environ. Microbiol. 3: 229–236. Silo-Suh L.A., Lethbridge B.J., Raffel S.J., He H., Clardy J. & Handelsman J. (1994). Biological

activities of two fungistatic antibiotics produced by Bacillus cereus UW85. Appl. Environ. Microbiol. 60: 2023-2030.

Silo-Suh L.A., Stabb E.V., Raffel S.J. & Handelsman J. (1998). Target range of zwittermycin A, an aminopolyol antibiotic from Bacillus cereus. Curr. Microbiol. 37: 6-11.

Smith K., Havey M. & Handelsman J. (1993). Suppression of cottony leak of cucumber with Bacillus cereus strain UW85. Plant Dis. 77: 139-142.

Stallings J.H. (1954). Soil produced antibiotics-plant disease and insect control. Bacteriol. Rev. 18: 131–146.

Stein T. (2005). Bacillus subtilis antibiotics: structures, synhesis and specifics functions. Mol. Microbiol. 56: 845-847.

Stein T. (2008). Whole-cell matrix-assisted laser desorption/ionization mass spectrometry for rapid identification of bacteriocin/lantibiotic-producing bacteria. Rapid Commun. Mass Sp. 22: 1146-1152.

Thomashow L.S., Bonsall R.F. & Weller D.M. (2008). Detection of antibiotics produced by soil and rhizosphere microbes in situ. In: Karlovsky P, eds. Secondary Metabolites in Soil Ecology. Springer-Verlag, Berlin Heidelberg.

Touré Y., Ongena M., Jacques P., Guiro A. & Thonart P. (2004). Role of lipopeptides produced by Bacillus subtilis GA1 in the reduction of grey mould disease caused by Botrytis cinerea on apple. J. Appl. Microbiol. 96: 1151–1160.

Vater J., Kablitz B., Wilde C., Franke P., Mehta N. & Cameotra S.S. (2002). Matrix-assisted laser desorption ionization-time of flight mass spectrometry of lipopeptide biosurfactant in whole


56

cells and culture filtrates of Bacillus subtilis C-1 isolated from petroleum sludge. Appl. Environ. Microb. 68: 6210-6219.

Vater J., Wilde C. & Kell H. (2009). In situ detection of the intermediates in the biosynthesis of surfactin, a lipoheptapeptide from Bacillus subtilis OKB 105, by whole-cell cell matrix-assisted laser desorption/ionization time-of-flight mass spectrometry in combination with mutant analysis. Rapid Commun. Mass Sp. 23: 1493–1498.

Wei S.C., Shun C.T., Tsai-Wu J.J., Wu C.H., Sheu J.C., Wang C.Y. & Wong J.M. (2004). Microsatellite instability and hMLH1 and hMSH2 gene expression in Taiwanese hereditary nonpolyposis colorectal cancer. J. Formos. Med. Assoc. 103: 331-6.

Wei Y.H. & Chu I.M. (2002). Mn2+ improves surfactin production by Bacillus subtilis. Biotechnol. Lett. 24: 479–482.

Wei Y.H., Laia C.C. & JChangb O.S. (2007). Using Taguchi experimental design methods to optimize trace element composition for enhanced surfactin production by Bacillus subtilis ATCC 21332. Process Biochem. 42: 40-45.

Yeh M.S., Wei Y.H. & Chang J.S. (2006). Bioreactor design for enhanced carrier-assisted surfactin production with Bacillus subtilis. Process Biochem. 41: 1799-1805.

57

OBJECTIVES OF THE THESIS

58

OBJECTIVES OF THE THESIS

As stated in the introducive chapters, Bacillus has been primarily studied for its biochemical

and genetic traits. By contrast with some other plant beneficial bacteria such as Pseudomonas,

research on Bacillus as a biocontrol agent has focussed mainly on the efficacy of the

protective effect under practical use in the field or greenhouse. However, despite many recent

conceptual and technical advances, the investigation of antibiotic gene expression during

interaction with plants and of their secretion by Bacillus isolates growing in the phytosphere

is still in its infancy. Thus, there is still a crucial lack of data on the antibiotics production by

these bacteria in the natural environment. The main objective of our thesis is to contribute at

filling this gap of knowledge.

In a first part, we wanted to evaluate the potential of strain S499, selected out of more

than forty other soil isolates, as a microbial biocontrol agent. To that end, we tested S499 for

its ability to control an important disease under field conditions in Burundi (Chapter 4 in

Results section). In the next part of this work, we wanted to gain some insight into how

lipopeptide patterns of surfactin, iturin and fengycin families are produced by B.

amyloliquefaciens strain S499 developing on roots of tomato plants and assert environmental

factors that inherently influence their production (Chapter 5). Further exploring surfactin gene

expression and production upon growth of Bacillus on tomato roots under greenhouse

conditions was our last objective. Another strain, Bacillus subtilis BGS3, containing a

reporter gene system that helps strain discrimination among many other bacteria, was used.

59

RESULTS

Results-Chapter 4

60

Foreword on Chapter 4

Tomato is one of the most important crops food worldwide. However, this plant productivity

is very low in Burundi because of, amongst other factors, primitive agricultural practices most

of which are meant for subsistence farming, led to continuous reduction in yields. This

situation is worsened by increased infection by phytopathogenic fungi such as Fusarium,

Pythium spp. and others causing root rot in a wide variety of crops especially tomato. Efforts

to combat these problems using fungicides and chemical fertilizers were not only

unsuccessful but are also unaffordable to the local farmers. In contrast to research on the

suppression of diseases with chemical pesticides, to date no study has yet addressed the

introduction of antagonistic microorganisms in the plant rhizosphere in Burundi.

The first objective of our thesis was thus to test selected rhizobacteria, with potential

antimicrobial activity as biocontrol agents to protect tomato. In this context, we first screened

for Bacillus-like strains isolated from tomato rhizosphere samples collected in local fields in

Burundi. Their antagonistic potential against some fungi well known for their severe

pathogenicity towards many economically important crops was compared to the one of B.

subtilis/amyloliquefaciens strain S499. However, none of them displayed greater potential and

we decided to test S499 for its biocontrol activity in field trials because of the scientific and

technological background in our laboratory. More specifically, the main part of this chapter

will describe the beneficial effect of tomato treatment with spores suspensions from strain

S499 on plant growth and tolerance to an uncommon fungal isolate under agricultural field

conditions in Burundi.

Results-Chapter 4

61

Chapter 4. Beneficial effects of Bacillus subtilis on field-grown tomato in Burundi: Reduction of local Fusarium

disease and growth promotion

Published in African Journal of Microbiology Research (2010) 4 (11): 1135-1142.

(except data concerning Bacillus strain screening in the first section of Material and

method and in second section of the Results)

Venant Nihorimbere1,3†, Marc Ongena1†, Hélène Cawoy1, Yves Brostaux2, Pascal Kakana3 and

Philippe Thonart1

1 Walloon Centre of Industrial Biology, Gembloux Agricultural University, Passage des

déportés, 2, B-5030 Gembloux, Belgium. 2 Statistics, informatics and applied mathematics, Gembloux Agricultural University, Passage

des déportés, 2, B-5030 Gembloux, Belgium. 3 Faculty of Agronomy, University of Burundi, PO Box 1550, Bujumbura, Burundi.

†Both authors equally contributed to this article.

Results-Chapter 4

62

Abstract

In this work, we first determined the in vitro antagonistic activity of 44 native strains of

Bacillus spp. isolated from the tomato rhizosphere against some fungi well known for their

severe pathogenicity towards many economically important crops throughout the world. The

bacterial isolates showed differential levels of antagonistic activity against the

phytopathogenic fungi tested in vitro on PDA medium. Four of them were selected to be

stronger inhibitors of at least two fungi but globally, none displayed a superior potential

compared to strain S499 which is commonly used in our laboratory. The biocontrol potential

of this S499 strain was thus further evaluated on tomato in open field sites in low altitude area

of the plain of Imbo in Burundi. This strain was tested in order to reduce the impact of an

important fungal disease giving rise to large losses in local plantings. The causing pathogen

was isolated from diseased leaves at different locations in the fields and identified as

Fusarium most probably related to the semitectum species according to the fermentation

profile, morphology and gene homology. Results of assays performed in two successive years

on the same site indicate that bacterial treatment on seeds significantly increased growth and

fruit yield of tomato plants and also provided a high level of protection against the disease

caused by the Fusarium pathogen. This is the first reported study on this disease and based on

the collected data, Bacillus subtilis S499 may represent an effective solution as biocontrol

agent where other, chemical, options have failed.

Keywords: Bacillus subtilis, biocontrol, Fusarium disease, tomato plants

Results-Chapter 4

63

Introduction

Root colonizing bacteria that exert beneficial effects on plant development through direct or

indirect mechanisms have been defined as Plant Growth-Promoting Rhizobacteria (PGPR).

These bacteria include several genera such as Acinetobacter, Agrobacterium, Arthrobacter,

Azospirillum, Bacillus, Bradyrhizobium, Frankia, Pseudomonas, Rhizobium, Serratia, and

Thiobacillus (Flavel, 2005) but the most well-known and commonly-used strains belong to

the Pseudomonas and Bacillus genera (Kloepper et al., 2004; Haas and Défago, 2005). In

addition to promoting plant growth, PGPR are also employed to control plant pathogens,

enhance efficiency of fertilizers, and degrade xenobiotic compounds (Clayet-Marcel et al.,

2001; Choudary and Johri, 2008; Berg, 2009).

Some Bacillus strains with PGPR activity are among the most exploited bacteria as

biocontrol agents against plant diseases (Flavel, 2005). Bacillus isolates are considered to be

safe microorganisms and hold the remarkable abilities of synthesizing a vast array of

beneficial substances (Stein, 2005) for agronomic and industrial purposes. They produce

endospores which warrant their prevalence under different environmental conditions and

facilitate its long-term storage and easy development of reliable formulations. One of the most

commonly used and well-studied organisms, the rhizobacterium Bacillus subtilis, has an

average of 4-5% of its genome devoted to antibiotic synthesis and has the potential to produce

more than two dozen structurally diverse antimicrobial compounds (Stein, 2005). Among

these antimicrobial compounds, cyclic lipopeptides (LPs) of the surfactin, iturin and fengycin

(or plipastatin) families have well-recognized potential uses in biotechnology and

pharmaceutical applications because of their surfactant properties (Ongena and Jacques,

2008). Several mechanisms have been postulated to explain how these beneficial

rhizobacteria stimulate plant growth. Some strains secrete antibiotics that can directly inhibit

the growth of fungal, oomycete, and bacterial plant pathogens (Mckeen et al., 1986; Touré et

al., 2004; Romero et al., 2007). Some other strains of B. subtilis suppress diseases by

inducing host defenses (Kloepper et al., 2004). However, it is likely that the most effective

biological control strains act via multiple mechanisms. For example, B. subtilis strain S499

can inhibit fungal pathogens directly through antibiosis, but it also was found to induce

resistance to foliar pathogens when it was applied to the plant root (Ongena et al., 2005a; b).

Such induction of enhanced defensive capacity can be systemic as seed-treatment with

bacteria at the time of seeding was shown to trigger protective effects on above-ground parts

(van Loon et al., 1998). The induced resistance constitutes an increase on the level of basal

Results-Chapter 4

64

resistance to several pathogens simultaneously, which is one of the benefits under natural

conditions where multiple pathogens exist (van Loon and Glick, 2004).

Tomato is intensively cultivated in Burundi, mainly in low-altitude regions of the

country (800 to 1200 m), which yields three quarters of the national production of tomato

fruits. Recently, however, this plant has been experiencing significant losses from pathogenic

fungal attacks. The causal agent(s) in this case adversely affects plant survival and growth

before flowering. The resulting crop losses are increasing and the global fruit production in

the last five years only represents 52% of the average yield recorded in the preceding five-

year period (1997-2002) (Bitoga J.P., Institut des Sciences Agronomiques du Burundi,

Burundi, unpublished). The disease symptoms in this case begin with a change in the colour

of the leaves from green to yellowish brown. When the infection progresses, defoliation

occurs as does the darkening of the vessels followed by plant death. This disease was not

known in Burundi tomato cultures. In the region, the control of tomato diseases has been most

exclusively based on the application of chemical pesticides, but there is no real efficient

solution for local farmers to reduce disease impact. In that context and according to present

and future regulations on the use of many chemical fungicides, and considering that phyto-

pharmaceutical treatment must prevent environment pollution, the use of biological agents to

essentially control the fungus that devastates tomato plantings in Burundi can be assayed.

In the work reported here, we have isolated the most predominant fungal pathogen from

diseased tomato plants and have performed an initial putative characterization of this isolate.

The Bacillus subtilis strain S499 has then been selected among other Bacillus isolates based,

first on its high antagonistic activity against the growth of the fungus and second, on its

known biocontrol potential. Indeed, strain S499 efficiently colonizes plant roots and has been

already demonstrated to induce resistance in tomato and other plants under greenhouse

conditions (Ongena et al., 2005a; b). However, no studies have been attempted to date in

order to determine its biocontrol potential on plants cultivated in the open field. In this

context, the objective of the study was to evaluate the beneficial effect of tomato seed

treatment with an aqueous suspension of spores from B. subtilis strain S499 on plant growth

and tolerance to an uncommon fungal isolate under agricultural field conditions in Burundi.

Results-Chapter 4

65

Materials and Methods

In vitro primary screening for antagonistic activities against phytopathogens *1

Screening for in vitro antagonism was performed on Potato Dextrose Agar (PDA) medium.

Bacillus isolates were tested for their ability to inhibit the growth of fungal pathogens:

Botrytis cinerea, Pythium ultimum, Fusarium oxysporum, Cladosporium cucumerinum and

Aspergillus niger. A bacterial culture at an approximate concentration of 108 cfu/ml was

spotted with a sterile toothpick on two edges opposites one of another of the Petri dish, and a

small mycelial section of each fungus comprised of only a few mycelial sections (6-mm

diameter) was deposited at the centre at approx. 3 cm from the isolate, ± 24 h in advance

when needed depending on the fungus growth-rate. After 4 days of incubation at 25-28 °C,

the inhibition effect on fungal growth was evaluated. The radii of the fungal colony towards

and away from the bacterial colony were assessed after the incubation period. All in vitro

antagonism assays were made in duplicates. To make a selection of the best biocontrol

isolatespreference was given to the isolats with inhibition abilities against more than one

pathogen.

Microbial strains and inoculum preparation

B. subtilis strain S499 was isolated from soil in Congo by Dr. L. Delcambe (Centre National

de Production et d’Etude de Substances d’Origine Microbienne, Liège, Belgium). The

bacterium was maintained on plate count agar (PCA medium; Becton, Dickinson and Co., Le

pont de Claix, France) at 4 °C before experimental use. For long-term storage, it was

conserved at −80 °C in cryotubes, according to the manufacturer’s recommendations

(Microbank, Prolab Diagnostic). Bacterial spore inoculum used in this study was provided by

the society Artechno S.A. after growing the S499 strain under optimized industrial conditions

in a 2000-L bioreactor. The fermentation was stopped at the time of almost full sporulation,

centrifuged and lyophilized to yield a highly concentrated stable powder. This product was

resuspended in sterile distilled water to obtain the final desired spore concentration to

inoculate tomato seeds.

The fungal pathogen of tomato plants was isolated from different fields in the intensive

agricultural zone in the plain of Imbo (Bujumbura-Burundi). The isolate was obtained from

infected parts of leaves and tested for characterization firstly by the laboratory of mycology at

the University of Louvain-La-Neuve (Belgium) and secondly by Scientific Institute for Public

Health in Brussels (Belgium). The technique of slide culture was used (Rodrigues and

* 1 Unpublished section

Results-Chapter 4

66

Menezes, 2005), which allows the direct microscopic observation of morphological structures

of taxonomic value. This technique consisted of inoculating the fungus at the sides of small

cube of agar (1 cm2) maintained in the center or at the edge of a slide and covered with a glass

cover. The slide cultures were kept on a support to avoid direct contact with the humid base of

the Petri dish. After incubation for 48 hours, the microcultures were examined in preparation

with Amman blue. For short-term maintenance, a piece of agar with mycelium measuring

about 1 cm2 was transferred top-down to a new plate approximately every three weeks under

sterile conditions. The plates were sealed with parafilm and kept at room temperature in the

dark for several days until the medium was completely covered by the fungus. The plates

were then stored at 4 °C. The pathogen inoculum was prepared by harvesting both micro- and

macroconidia from 10-day-old cultures in sterile peptone water (1 g/l Bactopeptone, 9 g/l

NaCl, 0.02% Tween 80). After removing mycelial debris by filtration through cheese cloth,

the suspension was centrifuged for 5 min at 5,000 g and the conidia were resuspended in an

adequate volume of sterile distilled water to obtain the desired final concentration of 6x105

conidia/ml, determined microscopically by the use of a Bürker counting cell.

Biocontrol experiments

Field experiments were conducted during the dry tomato-growing season from April to

September for two consecutive years. Tomato seeds (Lycopersion esculentum L. cv Merveille

des Marchés) were dipped in the various bacterial spore suspensions or in distilled water

(control) for twenty minutes immediately before sowing. In addition, for each treatment, 100

ml of a spore suspension at the same concentration as the one used to treat seeds was poured

on the soil surface surrounding each seed in every planting. Four week-old tomato plants were

pathogen-inoculated by depositing three drops of the conidia suspension on three different

leaves of each plant. Two replications (parcels), each consisting of 30 tomato plants, were

used for every bacterial treatment with specific spore concentration and arranged in a

randomized design (single row plot). Two weeks after leaf inoculation, plants in each

treatment were rated for disease severity on the basis of necrosis zones (lesion diameter).

The different treatments were compared by ANOVA (P < 0.05, Minitab software) and

data from experiments with the same set-up in the two different years were pooled for

analysis, as interactions between experiment and treatment were not significant. Means from

the different treatments were compared using Newman and Keuls' test (least significant

difference at α = 0.05).

Results-Chapter 4

67

In vitro antagonism

Antifungal activities of multiple Bacillus strains were tested in plate bioassays. Inhibition of

pathogen growth was estimated on PDA medium poured into 9.0-cm Petri dishes. Mycelial

plugs (5.0 mm) of the fungus were deposited in the center of the plates, approximately 3.5 cm

from bacterial colonies. The antagonism developed by strain S499 was tested by streaking the

bacterium on the edge of the plates (Figure 2). Plates were incubated at 25 °C and fungal

growth inhibition was rated after 4 days.

Results

Fusarium isolate identification

Six different isolates of Fusarium–resembling fungal pathogens were isolated from diseased

leaves of tomato plants cultivated in different fields of the agricultural zone in the plain of

Imbo (Bujumbura-Burundi). One of these isolates was recovered from almost all tomato

plantings and was submitted to preliminary characterization after single spore isolation on

PDA. On this medium, the fungus grows rapidly, with abundant aerial mycelium that turns

from a white/ yellowish coloration to dark brown as the culture ages. In two separate

experiments, the fungal isolate was grown at different temperatures in vitro for four days by

placing mycelium discs (6 mm diameter) on PDA. The optimal temperature for mycelium

growth is between 20-25 °C and no or very little growth of mycelium was observed at 15 °C

and 35 °C.

A first identification of the fungal isolate was made in collaboration with the laboratory

of mycology at the University of Louvain-la-Neuve (Belgium). Conidiophores and conidia are

produced after 9-11 days of incubation at 19-22 °C in darkness. Microscopic observation

revealed two types of conidia. The primary ones formed in aerial mycelium on conidiophores

in polyphialides. These conidia were slightly sickle-shaped, thin-walled and without a

pedicellate basal cell. They had three to six septate. The second type of conidia were

abundant, mostly with no septate, ellipsoidal to cylindrical, slightly curved or straight,

occurring in false head (Figure 1).

Results-Chapter 4

68

Figure 1 Microscopic observations of 10-day old conidia of Fusarium isolated from leaves of diseased tomato plants in Burundi field. a: macroconidia, b: microconidia. The bar is equal to 5µm.

The fungal isolate causes blight symptoms and foliar necrosis. Under appropriate conditions,

infected leaves turn brown about 10 days after inoculation with a conidial suspension (Figure

3A). As the disease symptoms progress, host plants generally die before the flouring stage.

Morphology, fermentation profile and a first set of gene sequencing (calmodulin and

elongation factor) indicated that this isolate probably belongs to the genus Fusarium. It may

correspond to the semitectum species but this should be confirmed by further genetic

characterization.

Selection of Bacillus strain S499 as antagonist used for field experiments *2

A total of forty-four isolates from the rhizosphere of tomato plant that were cultivated in the

plain of Imbo (Burundi) were selected as Bacillus spp. on glucose yeast peptone agar

medium. According to preliminary characterization, all these isolates may belong to Bacillus

genus based on the Bergey’s Manual of determinative Bacteriology (Holt, 1994). Microscopic

studies revealed that the cells were rod-like, motile, Gram-positive, catalase positive, and

spore-forming under aerobic conditions.

To carry out the antagonistic activity test in vitro in the primary screening, the isolates

were assayed on two media viz. LB and PDA agar, and growth was almost similar but PDA

agar medium was chosen for further assays because its white color allows more clarity of the

inhibition zone. The strength of fungal growth inhibition depends on the rhizobacterial isolate

antimicrobial activity. To allow comparisons, growth alterations (expressed in mm) were

evaluated for all fungus-bacterium combinations on the fourth day of growth (not shown).

The determination of the strains’antagonistic behaviour towards fungi crucially depends upon

the dual bacterial-fungus simultaneous growth rate. To determine a procedure for assessing

the antagonism between these organisms, many attempts were assayed to first inoculate either

the fungus or the antagonistic agent a few days early or late. Therefore, the bacterial isolates

* 2Unpublished section

Results-Chapter 4

69

were inoculated on the medium 24h before the strain of Pythium ultimum, 60 h after Botrytis

cinerea, 48 h after Fusarium oxysporum, 24 h after Cladosporium cucumerinum and

simultaneously with Aspergillus niger. The antagonistic isolates were clearly discerned by

limited growth or complete absence of fungal mycelium in the inhibition zone surrounding a

bacterial colony. All the investigated isolates showed individual inhibition patterns and were

able to inhibit, even weakly, at least one of the tested fungi (Table 1). Then, we further

selected the most efficient strains based on the criterion of isolates showing capacity of

mycelia inhibition of at least two fungi with a 3 mm clearance zone towards and away from

the bacterial colony. Among the 44 bacterial isolates, only four of them (Table 2) behaved

very comparably to the strain S499 of B. subtilis considered in our case as antifungal model

because previous studies had shown this strain to exhibit strong antagonizing potential against

some pathogenic fungi (Ongena et al., 2007; Nihorimbere et al., 2009). In the present study,

the bacterial isolates inhibited the fungi growth at different degrees from 3 to 8 mm (Table 2).

Such variation in the extent of mycelial inhibition can be influenced by the nature of bacterial

isolates. Landa and collaborators (1997), for example, indicated that the ability of four

bacteria isolates to inhibit different strains of F. oxysporum differed significantly. This

suggests that taxonomically distinct isolates may exert their antifungal potential through

different modes of action and/or produce different types of antifungal metabolites (William

and Ash, 1996). Here, the modes of action were not elucidated. However, early works have

correlated the production of glucanolytic and proteolytic enzymes with antagonism in

PaeniBacillus polymyxa, B. pumilis and Bacillus spp. (Nielsen and Sorensen, 1997). The

production of HCN, siderophores, ammonia, lipase and chitinase in growth medium by

Pseudomonas corrugata was considered contributing to the antagonistic activity of the

bacterium against Alternaria alternate and Fusarium oxysporum (Trivedi et al., 2008). It is

likely that the above proposed mechanisms be involved in antagonistic behaviour of the

strains, but we believe that some other mechanisms might be operating behind antagonism.

The isolates that most effectively inhibited fungal growth on the dual culture experiment

resulted in such a major inhibition zone that there was no physical contact with the pathogens

suggesting that the rhizobacteria could be producing certain metabolites (Montealegre et al.,

2003; Ahmed et al., 2007). Many of these specialized compounds, such as antibiotics, are

either liquid or solid at room temperature, and little is known about volatiles (with molecular

masses less than 300 Da, low polarity, and a high vapor pressure) that can act as antibiotics

and cause growth inhibition or have more deleterious effects on organisms. Indeed, the

microbial world synthesizes and emits many volatile compounds that can inhibit pathogen

Results-Chapter 4

70

growth (Dickschat et al., 2005). At present, the biologically active volatiles causing the

inhibition are not known because many volatiles have not been detected or identified (Kai et

al., 2007).

Table 1. Bacterial isolates screening for fungal growth inhibition

Bacillus isolates

Otrytis cinerea

Pythium ultimum

Fusarium oxysporum

Cladosporium cucumerinum

Aspergilus niger

BG152 + - - + + BG2511 - - - + + BG2512 - - - + + BG252 + - - + +

BG2571a + + - + + BG257b + + - + - BG311 + + + - + BG312 - - - - + BG3361 - - - + - BG3362 + - - - + PD2511 + - - + + PD2512 + - - + - PD31 - - + + + PD32 - - - + - PD35a + - - - - PD35b - - - + - PD35c + + - - - PD35d - + - - - PD35e - - - - + PD36 - - - - - PD361 + - - + - PD362 - - - + + PD6571 + - - + + PD6572 + - - + + PD6573 + + + - + PD661 - - - + - PD662 + + - + +

PD6652a + + - - - PD6652b + - - - +

PD71 + - - - - PD72 + - - - - PD76 - - - + - PD81 - - - + - PD82 - - - + - PD9 - - - + -

KD15 + - - + - KD151 - - - + - KD152 - - - + + KD161 - - - + - KD162 + + + - + KD25 + - - - + KD261 + - - - - KD35 - - - - +

KD4621 - - - + - KD1622 - - - + -

S499 (ctrl)* + - - + + +: Halo formation; -: No halo formation; *: control strain

Antifungal effects of organic volatiles were previously shown to inhibit germination or

mycelial growth of S. sclerotiorum (Fernando et al., 2005), and unidentified compounds from

Results-Chapter 4

71

Bacillus subtilis cause structural deformation of pathogenic fungi (Chaurasia et al., 2005).

The approach herein investigated (phytopathogens’inhibition) provided an opportunity to

select effective biocontrol strains capable of antagonizing tested fungi that are susceptible to

damage crops in the same environment. We believe that the isolates showing the highest

ability to reduce the pathogens’ growths in the in vitro dual assay can also be efficient

antagonists against phytopathogens of many crops in the in vivo biological control tests.

However, it is known that in vitro assays have certain limitations in that biocontrol

efficiencies may not be equally expressed under gnotobiotic (axenic) and in vivo conditions

(Inam-ul-Haq et al., 2003). Tests based on in vitro mycelial inhibition and root colonization

do not always correlate with biocontrol efficacy under natural conditions (Williams and

Asher, 1996).

Table 2. In vitro assay of bacterial isolates most inhibiting towards some fungi

Isolates Botrytis cinerea

Pythium ultimum

Fusarium oxysporum

Cladosporium cucumerinum

Aspergilus niger

BG311 + + + + + BG2571a + +++ - ++ + PD6572 ++ - - + + BG252 + - - + ++

S499(ctrl)* ++ - - +++ ++ Clearance zone: +3 mm; ++: 5 mm; +++: 8 mm; _: negative reaction; *: ctrl: control strain

Dual assay method allowed determination of the antifungal activity of the native

isolates, and showed higher activity of four isolates among 44 isolated. These isolates should

be further identified and later investigated for their efficacy in greenhouse and in field. On the

basis of this preliminary study on biocontrol, it is deduced that the rhizosphere of tomato in

the plain of Imbo harbours beneficial microorganisms with the potential to suppress soilborne

diseases caused by some common fungi especially encountered in tropical regions such as

Fusarium, Pythium, Apergillus, Botrytis spp. and possibly other soilborne pathogens. We thus

selected the strain S499 for the next experimentations because it presents a little bit high

antagonistic activity as compared to the four local strains which are potent fungal growth

inhibitors.

Fusarium disease reduction upon treatment with B. subtilis

As a first step, a dozen of Bacillus strains either isolated from local soil or originating from

suppressive soils in other regions of central Africa, were tested for their in vitro antagonistic

potential towards growth of the pathogen. The B. subtilis strain S499 from Congo was among

the best with this respect (Figure 2) and was selected because of its high spore-forming

capability and based on its previously well established biocontrol activity against tomato,

Results-Chapter 4

72

bean and cucumber diseases. A formulated powder containing high concentration of viable

endospores was obtained from strain S499 by pilot-scale fermentation and lyophilisation.

Using this product, biocontrol assays were performed in two successive years on the same site

by treating tomato seeds with spore suspensions of S499 before sowing.

Figure 2 In vitro growth inhibition of the pathogen Fusarium semitectum caused by Bacillus subtilis S499 on PDA medium. The bacterium and the fungus were inoculated at the same time and the antagonism was scored after incubation of the plates for 4 days at 25 °C.

The disease was introduced artificially by inoculating the Fusarium pathogen on the leaves of

21-day-old plants. Disease incidence in control and Bacillus-treated plants was compared on

the basis of the diameter of spreading lesions observed two weeks later. In these assays, we

also evaluated the dose-efficacy relationship for the biocontrol agent. Four spore

concentrations were used to inoculate tomato seeds in separate parcels. Control plants were

not inoculated and as these control plants did not present any necrosis or symptoms at the

time of assessment, we assume that there has been no cross contamination of the fungus

between parcels. Based on the diameter of lesions on leaves, plants inoculated with strain

S499 showed significantly lower infection rates compared to the non-treated, inoculated

plants (Figure 3). Interestingly, the four inoculum concentrations were all efficient at the same

level to reduce the fungal pathogen ingress and in all cases, a disease reduction of up to 65-

70% was observed.

Colonization of the rhizosphere is required for a PGPR strain to consistently influence

plant growth and health (Chin-A-Woeng et al., 2003). In order to support these biocontrol

data, the evolution of S499 populations was evaluated on roots of tomato plants grown under

these fields' conditions. For the four different inocula tested, the population level was assessed

by agar plate count 30, 49 and 73 days after inoculation.

Results-Chapter 4

73

0

1

2

3

4

5

6

7

C 10exp5 10exp6 10exp7 10exp8

Me

an

dia

me

ter

of

ne

cro

sis

(cm

)

Bacterial inoculum concentration (CFU/ml)

a b c A

B

Figure 3 A, Disease symptoms on infected leaves of untreated tomato plants collected from the field two weeks after pathogen inoculation (b) compared to plants treated with strain S499 prior to infection(c) and to treated, uninfected control plants (a). B, Disease incidence rated as mean diameter of necrosis (cm) measured two weeks after fungal inoculation on leaves of plants germinated from bacterized and non-bacterized seeds with various bacterial inoculum concentrations and seeded in separate parcels. Infection rate data from two parcels per treatment with 30 plants each were submitted to statistical analysis. For infection, the first three leaves of each plant were infected with three drops of a fungal spore suspension containing 6 x 105 cfu/ml. As the analysis of variance and the multiple comparison tests did not reveal any statistical difference between results from two successive years for similar treatments, these data were pooled and the results are thus means and standards errors calculated from two independent experiments.

Discrimination between B. subtilis S499 and soil microflora was based on colony

morphology. Populations observed for the different inocula were found to progressively

converge and decrease, stabilizing around 2x106 (Table 3).

Table 3 Root colonization of tomato plants grown under field conditions in separate parcels from seeds treated with different spore suspension concentrations of B. subtilis S499. Values represent the mean and standard error calculated from at least three plants randomly collected in each parcel.

Inoculum (cfu/g) 30 days 49 days 73 days

105 3.9 ±4.2 x 107 5.3 ±1.4 x 106 1.8 ±0.6 x 106

106 2.3 ±2.2 x 107 7.3 ±14.6 x 106 2.6 ±2.6 x 106

107 4.8 ±0.6 x 107 1.6 ±15 x 107 2.1 ±12 x 106

108 8.4 ±1.4 x 106 4.5 ±0.6 x 106 1.6 ±1.0 x 106

Results-Chapter 4

74

Effect of B. subtilis on tomato plant growth and fruit yield

In parallel to plant protection assays, some additional parcels were used to evaluate the plant

growth-promoting effect of strain S499. To this end, we only used non-treated tomatoes and

plants treated with various concentrations of Bacillus but without pathogen inoculation. The

first parameter used to evaluate plant growth promotion was the height of the plants at 62

days after sowing. Our results show that treatment with B. subtilis S499 significantly

increases plant height at this stage. At 62 days, water control plants had only reached an

average size of 44 cm while plants inoculated with the bacterium suspensions reached an

average height of 70 cm representing an increase in shoot height of more than 59%. Plants

inoculated with a suspension at 107 CFU/ml were significantly higher than those inoculated

with other inoculum concentrations, but with only less than 5 cm difference. No statistical

difference was recorded between plants inoculated with suspensions at 105, 106 and 108

CFU/ml (Figure 4A). These results indicate that the inoculation of tomato seeds with a spore

suspension of S499 at the low 105 CFU/ml concentration is sufficient to significantly increase

plant growth.

To further evaluate the growth stimulation potential of B. subtilis S499, fruit yield was

measured. At 72 days post-sowing, we first determined the percentage of plants bearing three

to four fruits in the various parcels treated with different inoculum concentrations. At least

60% of the tomato plants bacterized with the various inoculum densities had three to four

fruits but only five percents of the untreated control plants had reached this stage (Figure 4B).

A few days later, the total fruit yield per plant was also compared. Weight measurements of

fruits on treated plants were much higher (at least 14-fold increase) compared to controls

(Figure 4C). Concerning both the quantity and weight of fruits, these highly significant

statistical differences in comparison to untreated tomatoes was observed regardless of the

inoculum concentration tested.

Results-Chapter 4

75

0

10

20

30

40

50

60

70

80

90

He

igh

t o

f to

ma

to p

lan

ts (

cm)

0

10

20

30

40

50

60

70

80

90

% t

om

ato

pla

nts

wit

h 3

-4 f

ruit

s

0

20

40

60

80

100

120

C 10exp5 10exp6 10exp7 10exp8

To

ma

to f

ruit

we

igh

t (g

)

Bacterial inoculum concentration (CFU/ml)

A

B

C

Figure 4 Effect of strain S499 on tomato growth under field conditions. This effect was evaluated within each treatment by (A) Plant shoot height measured 62 days after sowing, (B) Percentage of tomato plants bearing at least three fruits 72 days after sowing; (C) Tomato fruit weight (g) 78 days after sowing. Plants were sown in separate parcels with two parcels per treatment and 30 plants per parcel. Seed-treatments were performed with various concentrations of S499 spores and compared to untreated control plants (C). Data were treated statistically as described earlier and mean values and standard errors were calculated from 10 plants selected randomly among 30 in every parcel.

Results-Chapter 4

76

Discussion

Six fungal isolates were obtained from diseased tomato plants from different fields in the

plain of Imbo (Bujumbura-Burundi). One isolate stood out by having the greatest frequency,

with 98.7% in all the tomato plantings. Although further work is required, microscopic

observation, cultural characteristics and preliminary gene homology analyses indicated that

the fungus is of the Fusarium genus and may belong to the semitectum species (Bokshi et al.,

2003). Fusarium is relatively common within subtropical and tropical environments and some

species can be pathogenic to several plant species (Nelson et al., 1983; Bokshi et al., 2003;

Kloepper et al., 2004). In this work, the fungus was isolated from leaves of tomato plants

displaying brown symptoms and foliar necrosis. It is possibly this same pathogen which

caused great losses by devastating young plants and subsequently affected the local farming

economy. This isolate has not been previously described as an infectious agent attacking

aerial parts of tomato plantings in Burundi. By contrast, F. semitectum is well known as a

major cause of seedling diseases on cotton and other crops in many subtropical African

countries (Abd-Elsalam et al., 2003; Kloepper et al., 2004) and was also shown to be

moderately virulent on Acacia Koa seedlings (Dudley et al., 2007). Its high pathogenicity

may be due to the fact that it produces some mycotoxins that may adversely affect plants

(Logrieco et al., 2002). It is also interesting to note that this strain we isolated grows well in

vitro at 20-25 °C. This optimal temperature range corresponds to the average values

prevailing locally in the natural conditions in Burundi.

Strain S499 was selected out of a range of B. subtilis isolates and tested for its efficacy

at controlling the disease caused by this locally important fungal isolate. Results revealed a

highly significant disease reduction following the treatment of seeds with bacterial spore

suspensions in comparison with the control plants. Tomato plants germinated from bacterized

seeds show strong resistance of about five-fold less necrosis diameter as compared with

controls (Figure 3). Strain S499 has been extensively studied in our laboratory and

demonstrated to be effective as biocontrol agent in greenhouse conditions (Ongena et al.,

2005a; b). This strain efficiently produces antibiotic compounds, among which lipopeptides

that may act through direct antagonism toward various fungi and oomycetes and/or by

enhancing the host plant defensive capacity (Ongena and Jacques, 2008). Various bacterial

spore suspensions were used and their effects on disease suppression were shown to be

similar suggesting a low importance of the inoculum concentrations used. This may be

explained by the fact that the population levels corresponding to the four inocula tested

Results-Chapter 4

77

became almost similar after a few weeks post-inoculation. These results also illustrate the

colonization potential and thus, the rhizosphere fitness, of strain S499 that maintained a

certain steady level of population for a prolonged period even if some decrease in cell

densities were observed over time in a prior phase. Such progressive decline of PGPR

populations after introduction at the root level has already been reported and the following

stable phase probably corresponds to a resident phase where the population size is restricted

by space and/or nutrients availability (Bashan, 1998; Di Mattia et al., 2002; Espinosa-Urgel et

al., 2002). From a more practical point of view, the use of low inoculum densities (105 cfu ml-

1) is also economically relevant for treating this kind of disease in open fields.

Possible colonization of leaf tissues by S499 or migration of the strain through the plant

resulting from seed treatment was also assessed by counting colonies on the basis of their

typical morphology. Bacillus cells were not detected, suggesting that the bacterium remained

within the root system and demonstrating that the disease suppression was due to the

induction of resistance in the host plant as the beneficial strain and the pathogen are localized

on different plant organs. Moreover, the bacterial populations established on the roots are in

the range of the threshold commonly required to trigger plant systemic resistance. The ISR

triggering potential of strain S499 has already been observed on tomato and other plants

(Ongena et al., 2005a; b) but the experiments conducted in this work are the first

demonstration of ISR-based biocontrol activity by S499 under field conditions. In previous

works, it has also been evidenced that lipopeptides of the surfactin and fengycin families do

play a crucial role in the elicitation of this induced systemic resistance phenomenon (Ongena

et al., 2007; Jourdan et al., 2009). Treatment with pure compounds and/or with overproducing

derivatives triggered ISR in bean and tomato plants (Ongena et al, 2007). We have also

recently demonstrated that surfactin genes in B. subtilis are readily expressed in vitro

(Nihorimbere et al., 2009) and that the compound is efficiently produced by S499 cells

developing in the rhizosphere of tomato plants (V. Nihorimbere, University of Liège,

unpublished results). On the basis of these data, the production of lipopeptides may also be

involved in ISR-based disease protection by S499 under field conditions.

In parallel to plant protection, some additional parcels were reserved to evaluate the

strain for its plant growth promoting effect sensu stricto. Results show that tomato shoot

heights were markedly increased by approx. 59% upon seed-treatment with all the S499

inoculum densities tested (Figure 4A). Based on two additional parameters, trial assessment

approx. 70 days after sowing also revealed a clearly enhanced fruit yield for treated-plants as

compared with the controls. All the inoculum concentrations tested improved similarly the

Results-Chapter 4

78

number of fruits as well as the total fruit weight per plant (Figures 4B and C). The beneficial

effect of rhizobacteria in general and of Bacillus species in particular, on plant development

putatively rely on diverse mechanisms that may be involved concomitantly (Ping and Boland,

2004; Berg, 2009). Plant growth may be directly stimulated through the bacterial production

of auxins, cytokinins, gibberellins or hormone-like compounds. The hormonal balance of the

plant can also be influenced indirectly by associated microorganisms as exemplified by those

producing ACCdeaminase able to degrade the precursor of ethylene that plays a pivotal role in

various developmental and stress tolerance functions (Glick, 2005). Besides these

mechanisms, plant-associated bacteria can also improve nutrient acquisition by the plant,

especially via nitrogen fixation or via solubilization of phosphorous, iron and other elements.

Other compounds, such as the volatile acetoin and butandiol emitted by some Bacillus strains

are also involved in the stimulation of plant growth (Ryu et al., 2003). We don’t know yet

what traits are involved in the case of S499 but experiments are being performed to evaluate

the potential of the strain at producing phytohormones, volatile organics, siderophores and

specific enzyme activities such as ACCdeaminase and phytase with the aim to better

understand its broad-spectrum growth promotion activity as it was also observed with oat,

broad bean and maize (data not shown).

In conclusion, this work illustrates the effectiveness of one particular B. subtilis strain at

providing beneficial effects on health of tomato plants cultivated in open fields in Burundi.

Based on the level of both growth-promoting effect and disease reduction, our results suggest

that a formulated B. subtilis product can be utilized as inoculum at low spore concentration to

help local farmers at combating a new Fusarium disease devastating their tomato fields. The

efficacy of inocula at low spore concentrations used for seed treatment, leads to conclude that

such a Bacillus-based product may represent for local growers a low cost solution or

alternative to the use of chemicals for the control of Fusarium diseases in this region.

Acknowledgements V. Nihorimbere is recipient of a grant from the Belgian Technical Cooperation (BTC/CTB). M. Ongena and E. Jourdan are respectively Research Associate and Post-doctoral Researcher at the Fonds de la Recherche Scientifique–FNRS in Belgium. H. Cawoy’s Ph.D. thesis is supported by a grant from the Fonds pour la formation à la recherche dans l’industrie et dans l’agriculture (F.R.I.A.). The authors thank the society Artechno s.a. for provinding the B. subtilis powder.

Results-Chapter 4

79

References

Abd-Elsalam KA, Schnieder F, Asran-Amal A, Khalil MS, Verreet JA. (2003). Intra-species genomic groups in Fusarium semitectum and their correlation with origin and cultural characteristics. PflKrankk. 110:409-418.

Ahmed IH, Labuschagne N, Korsten L. (2007). Screening rhizobacteria for biological control of Fusarium root and crown rot of sorghum en Ethiopia. Biol. control 40: 96-107.

Bashan Y. (1998). Inoculants of plant growth-promoting bacteria for use in agriculture. Biotechnol. Adv. 16:729-770.

Berg G. (2009). Plant-microbe interactions promoting plant growth and health: perspectives for controlled use of microorganisms in agriculture. Appl. Microbiol. Biotechnol. 84:11-18.

Bokshi AI, Morris SC, Deverall BJ. (2003). Effects of benzothiadiazole and acetylsalicylic acid on beta-1,3-glucanase activity and disease resistance in potato. Plant Pathol. 52:22-27.

Chaurasia BA, Pandey L, Palni MS, Trivedi P, Kumar B, Colvin N. (2005). Diffusible and volatile compounds produced by an antagonistic Bacillus subtilis strain cause control of root-knot nematodes. Crop Prot. 26: 1644-1654.

Chin-A-Woeng TFC, Bloemberg GV, Lugtenberg BJJ. (2003). Phenazines and their role in biocontrol by Pseudomonas bacteria. New Phytol. 157: 503–523.

Choudary D, Johri B. (2008). Interactions of Bacillus spp. and plants-with special reference to induced systemic resistance (ISR). Microbiol. Res. doi:10.1016/j.micres.2008.08.007.

Clayet-Marcel J, Larcher M, Bartrand H, Rapior S, Pinochet X. (2001). Plant growth enhancement by rhizobacteria. In: JF Morot-Gaudry (ed) Nitrogen assimilation by plant: physiological, biochemical, and molecular aspects. Science publishers Inc., Enfield.

Di Mattia E, Grego S, Cacciari I. (2002). Eco-physiological characterization of soil bacterial populations in different states of growth. Microb. Ecol. 43:34-43.

Dickschat JS, Martens R, Brinkhoff T, MS, Schulz S. (2005). Volatiles released by a Streptomyces species isolated from the North Sea. Chem. Biodivers. 2: 837–865.

Dudley NS, James RL, Sniezko RA, Yeh A. (2007). Pathogenicity of four Fusarium species on Acacia koa seedlings. Forest Health Protection Report - Northern Region, USDA Forest Service 9 pp.

Espinosa-Urgel M, Kolter R, Ramos JL. (2002). Root colonization by Pseudomonas putida: love at first sight. Microbiology 148:1-3.

Fernando WG, Ramarathnam R, Krishnamoorthy AS, Savchuk SC. (2005). Identification and use of potential bacterial organic antifungal volatiles in biocontrol. Soil Biol. Biochem. 37: 955–964.

Fravel DR. (2005). Commercialization and implementation of biocontrol. Annu. Rev. Phytopathol. 43:337-359.

Glick BR. (2005). Modulation of plant ethylene levels by the bacterial enzyme ACC deaminase. FEMS Microbiol. Lett. 251:1-7.

Haas H, Défago G. (2005). Biological control of soilborne pathogens by fluorescent Pseudomonads. Nat. Rev. Microbiol. 3:307-319.

Holt JG. (1994). Bergey’s manual of systematic bacteriology, Vol 1 NR Krieg. Inam-ul-Haq M, Javed M, Ahmad R, Rehman A. (2003). Evaluation of different strains of

Pseudomonas fluorescens for the biocontrol of Fusarium wilt of chickpea. Pakistan J. Plant Pathol. 2: 65-74.

Jourdan E, Henry G, Duby F, Dommes J, Barthelemy JP, Thonart P, Ongena M. (2009). Insights into the defense-related events occurring in plant cells following perception of surfactin-type lipopeptide from Bacillus subtilis. Mol. Plant-Microbe Interact. 22:456-468.

Kai M, Effmert U, Berg G, Piechulla B. (2007). Volatiles of bacterial antagonists inhibit mycelial growth of the plant pathogen Rhizoctonia solani. Arch. Microbiol. 187: 351–360.

Kloepper JW, Ryu CM, Zhang SA. (2004). Induced systemic resistance and promotion of plant growth by Bacillus spp. Phytopathology 94:1259-1266.

Landa BB, Hervas A, Bethiol W, Jimenez-Diaz DM. (1997). Antagonistic activity of bacteria from the chickpea rhizosphere against Fusarium oxysporum f.sp.cicersi. Phytoparasitica 25: 305-318.

Logrieco A, Mule G, Moretti A, Bottalico A. (2002). Toxigenic Fusarium species and mycotoxins associated with maize ear rot in Europe. Eur. J. Plant Pathol. 108:598-609.

Results-Chapter 4

80

Mckeen CD, Reilly CC, Pusey PL. (1986). Production and partial characterization of antifungal substances antagonistic to Monilinia fructicola from Bacillus subtilis. Phytopathology 76:136-139.

Montealegre RJ, Reyes R, Pérez ML, Herrera R, Silva P, Besoain X. (2003). Selection of bioantagonistic bacteria to be used in biological control of Rhizoctonia solani in tomato. Electron. J. Biotechnol. 6: 115-127.

Nelson PE, Toussoun TA, Marasas WFO. (1983). Fusarium species. An illustrated manual for identification. Pennsylvania State University Press, University Park, PA.

Nielsen P, Sorensen J. (1997). Multi-target and medium independent fungal antagonism by hydrolytic enzymes in Paenibacillus polymyxa and Bacillus pumilis strains from barley rhizosphere. FEMS Microbiol Ecol 22: 183–192.

Nihorimbere V, Fickers P, Thonart P, Ongena M. (2009). Ecological fitness of Bacillus subtilis BGS3 regarding production of the surfactin lipopeptide in the rhizosphere. Environ. Microbiol. Report 1:124-130.

Ongena M, Adam A, Jourdan E, Paquot M, Brans A, Joris B, Arpigny JL, Thonart P. (2007). Surfactin and fengycin lipopeptides of Bacillus subtilis as elicitors of induced systemic resistance in plants. Environ. Microbiol. 9:1084-1090.

Ongena M, Duby F, Jourdan E, Beaudry T, Jadin V, Dommes J, Thonart P. (2005a). Bacillus subtilis M4 decreases plant susceptibility towards fungal pathogens by increasing host resistance associated with differential gene expression. Appl. Microbiol. Biotechnol. 67:692-698.

Ongena M, Jacques P (2008). Bacillus lipopeptides: versatile weapons for plant disease biocontrol. Trends Microbiol. 16:115-125.

Ongena M, Jacques P, Toure Y, Destain J, Jabrane A, Thonart P. (2005b). Involvement of fengycin-type lipopeptides in the multifaceted biocontrol potential of Bacillus subtilis. Appl. Microbiol. Biotechnol. 69:29-38.

Ping LY, Boland W. (2004). Signals from the underground: bacterial volatiles promote growth in Arabidopsis. Trends Plant Sci. 9:263-266.

Romero D, de Vicente A, Rakotoaly RH, Dufour SE, Veening J-W, Arrebola E, Cazorla FM, Kuipers OP, Paquot M, Pérez-García A. (2007). The iturin and fengycin families of lipopeptides are key factors in antagonism of Bacillus subtilis toward Podosphaera fusca. Mol. Plant Microbe Interact. 20: 430-440.

Rodrigues AAC, Menezes M. (2005). Indentification and pathogenic characterization of endophytic Fusarium species from cowpea seeds. Mycopathologia 159:79-85.

Ryu CM, Farag MA, Hu CH, Reddy MS, Wei HX, Pare PW, Kloepper JW. (2003). Bacterial volatiles promote growth in Arabidopsis. Proc. Natl. Acad. Sci. USA 100:4927-4932.

Stein T. (2005). Bacillus subtilis antibiotics: structures, synhesis and specifics functions. Mol. Microbiol. 56:845-847.

Touré Y, Ongena M, Jacques P, Guiro A, Thonart P. (2004). Role of lipopeptides produced by Bacillus subtilis GA1 in the reduction of grey mould disease caused by Botrytis cinerea on apple. J. Appl. Microbiol. 96:1151-1160.

Trivedi P, Pandey A, Palni LM. (2008). In vitro evaluation of antagonistic properties of Pseudomonas corrugata. Microbiol. Res. 163: 329–336.

van Loon LC, Bakker PAHM, Pieterse CMJ. (1998). Systemic resistance induced by rhizosphere bacteria. Ann. Rev. Phytopathol. 36:453-483.

van Loon LC, Glick BR. (2004). Increased plant fitness by rhizobacteria. In: H Sandermann (ed) Molecular ecotoxicology of plants. Springer, Berlin.

Williams GE, Asher MJC. (1996). Selection of rhizobacteria for the control of Pythium ultimum and Aphanomyces cochlioides on sugar-beet seedlings. Crop prot. 15: 479-486.

Results-Chapter 5

81

Foreword on chapter 5

Altogether, the results presented in Chapter 4 demonstrated the effectiveness of S499 strain at

controlling disease in the tomato open field in Burundi. On another hand, it is known that

lipopeptide antibiotics are the main antimicrobial components produced by this strain, and

these molecules may play an important role in the global antagonism developed towards

phytopathogens thereby contributing to the efficacy of some specific strains at reducing plant

disease incidence. The objective of the work reported within Chapter 5 is to evaluate whether

S499 strain (re-identified as B. amyloliquefaciens) produces or not these antibiotics of the

lipopeptides families of surfactins, iturins, and fengycins in situ on tomato plant root and in

vitro versus different parameters known to have influence on the cell culture development.

Results-Chapter 5

82

Chapter 5. Bacillus life in soil: impact of rhizosphere factors on lipopeptide signature from the plant beneficial strain

B. amyloliquefaciens S499

Submitted in International Society M icrobial Ecology Journal Venant Nihorimbere1, Hélène Cawoy1, Alexandre Seyer2, Alain Brunelle2, Philippe Thonart1, Marc Ongena1 1 Walloon Center for Industrial Biology, University of Liège/Gembloux Agro-Bio Tech, Gembloux B-5030, Belgium 2 Centre de recherche de Gif, Institut de Chimie des Substances Naturelles, CNRS, Avenue de la Terrasse, F-91198 Gif-sur-Yvette, France Running title: Bacillus lipopeptide produced in planta

Results-Chapter 5

83

Abstract

Through the so-called rhizosphere effect, plants usually create a specific environment in the

soil surrounding roots thereby influencing the production of biologically active compounds by

plant-associated microorganisms. In that context, lipopeptides represent a major class of

antibiotics involved in multiple facets of the tritrophic interactions between plant, pathogen

and beneficial Bacillus strains. In this work, we combined two mass spectrometry-based

approaches (electrospray ionization and imaging) to analyze the pattern of surfactin, iturin and

fengycin families produced in vivo by B. amyloliquefaciens strain S499. Our results show that

rhizosphere conditions are conducive for surfactin synthesis but not for other types of

lipopeptides that are actually secreted upon laboratory growth in artificial optimized medium.

We also show that the LP pattern secreted by S499 can be markedly influenced by nutritional

factors, biofilm formation and oxygen availability. From a qualitative point of view, it also

clearly appears that the relative proportions of homologues with long fatty acid chains are

higher in LP pattern produced in planta by strain S499. This is of biological relevance since

these LP forms were proven to be biologically more active regarding their haemolytic,

antiviral and antimicrobial activities as well as their potential to trigger plant defense

reactions. Globally, this work contributes to better appreciate the impact of some

environmental factors on the in situ biosynthesis of lipopeptides by strains of Bacillus which

is probably an essential step for improving the level and reliability of their efficacy as

biological agents for the control of plant diseases.

Keywords: Bacillus/ in situ/ lipopeptides/ production

Introduction

Bacterial lipopeptides (LPs) are composed of a lipid tail linked to a short linear or cyclic

oligopeptide and can be synthesized by various genera and species such as Streptomyces,

Pseudomonas and Bacillus. LPs are synthesized by non-ribosomal peptide synthetases or

hybrid polyketide synthases/non-ribosomal peptide synthetases. These modular proteins are

megaenzymes organized in iterative functional units called modules that catalyze the different

reactions leading to polyketide or peptide transformation (Finking and Marahiel, 2004). Such

biosynthetic systems lead to a remarkable heterogeneity among the LPs products which vary

in the type and sequence of amino acid residues, the nature of the peptide cyclisation and in

the nature, length and branching of the fatty acid chain (Raaijmakers et al., 2006; Ongena and

Jacques, 2008). Cyclic lipopeptides secreted by Bacillus strains (LPs) also encompass

Results-Chapter 5

84

structural variants depending on the genetic background of the considered strain. Although

other classes have been identified recently (Hagelin et al., 2007; Lee et al., 2007), Bacillus

LPs may be classified in three main families, surfactins, iturins and fengycins (Ongena and

Jacques, 2008). Surfactins are heptapeptides interlinked with a β-hydroxy fatty acid to form a

cyclic lactone ring structure. The group of iturins encompasses 7 variants including

bacillomycins and mycosubtilin. Iturins are also heptapeptides but linked to a β-amino fatty

acid chain with a length that usually varies from C14 to C17. The third family comprises

fengycins A and B, which are also called plipastatins. These molecules are lipodecapeptides

with an internal lactone ring in the peptidic moiety and with a β-hydroxy fatty acid chain (C14

to C18) that can be saturated or unsaturated.

Some of these bacterial LPs are considered as among the most powerful biosurfactants

isolated so far. They combine high effects on surface tension and low critical micellar

concentrations and thus retain exceptional emulsifying/foaming properties and high potential

for the solubilization of hydrophobic compounds. LPs have also received considerable

attention for their antimicrobial, cytotoxic, antitumor and immunosuppressant properties.

There is thus a growing interest in the study and application of these molecules for various

environmental and pharmaceutical applications (Pirri et al., 2009). LPs also display versatile

functions in the ecology of the producing strains and notably in interactions with co-existing

organisms, including bacteria, fungi, oomycetes, protozoan predators and plants in the case of

soil isolates (reviewed in Raaijmakers et al., 2010). When produced by plant beneficial

rhizobacteria, LPs may play an important role in the global antagonism developed towards

phytopathogens thereby contributing to the efficacy of some specific strains at reducing plant

disease incidence. Such a biocontrol potential mostly relies on efficient root colonization,

direct pathogen inhibition and possible stimulation of the host plant immune system. As a

matter of fact, LPs from Bacillus have been demonstrated to be involved in these three

mechanisms (Ongena and Jacques, 2008; Raaijmakers et al., 2010).

They are numerous reports suggesting a role for LPs in biocontrol relying on in vitro

antimicrobial activity. However, additional approaches are necessary to correlate biocontrol

activity with an efficient production of LPs since rhizosphere conditions can greatly modulate

antibiotic gene expression compared to laboratory. These approaches include the use of non-

producing or over-producing derivatives with respective loss or increase in the biocontrol

activity and/or the study of in situ LP gene expression (fusion of promoters of LP operons to

reporter genes such as lacZ, xylE, gusA, luxAB or gfp). Also necessary is the detection at

active concentrations of the antibiotic compound in the microenvironment which has been

Results-Chapter 5

85

inoculated with the producer strain. However, very few studies have been devoted to

investigate LP gene expression during interaction with plants and only a very limited number

of works have successfully detected LPs secreted by Bacillus isolates growing in the

phytosphere.

In that context, the present study was initiated with the first aim to perform an in-depth

characterization of the LP pattern produced by B. amyloliquefaciens strain S499 developing

on roots of tomato plants. Strain S499 has been selected because this natural soil isolate co-

produces in vitro both three LP families in relatively high amounts under optimized in vitro

conditions and retains some biocontrol activity in various pathosystems. On that basis, our

main goal is to better appreciate to what extent its LP production may be modulated both

qualitatively and quantitatively by some rhizosphere-specific factors.

Materials and methods

Strain identification

Strain identification was performed in the laboratory of microbiology at the University of

Gent in Belgium. Beside basic bacteriological tests, fatty acid analysis was performed on

whole-cells by gas-chromatography following the recommendation of the commercial

identification system MIDI (Microbial identification System, Inc., Delaware U.S.A.). The

profile was compared with the MIDI identification database TSBA50 V.5.0 and with the in-

house database containing several profiles of reference strains.

The applicability of the gyrA sequence for species discrimination within Bacillus subtilis

complex has been reported by Chun and Bae (2000). Total DNA was prepared according to

the protocol of Niemann and collaborators (1997). The gyrA gene was amplified by PCR

using the following primers: gyrA-forward, CAG TCA GGA AAT GCG TAG GTC CTT

Sequence (5´-3´) and gyrA-reverse, CAA GGT AAT GCT CCT GGC ATT GCT. The gyrA

fragment corresponds to B. subtilis gyrA numbering 43-1070. The PCR amplified gyrA was

purified using the NucleoFast® Terminator 96 PCR Clean-up Kit (Machinerey-Nagel, Düren,

Germany). Sequencing reactions were performed using the BigDye® Terminator Cycle

Sequencing Kit (Applied Biosystes, Foster City, CA, USA) and purified using montage ™

SEQ96 Sequencing Reaction Cleanup Kit (Millipore, Bedford, MA, USA). Sequencing was

performed using an ABI prism® 3100 Genetic Analyser (Applied Biosystems, Foster City,

CA, USA).

Results-Chapter 5

86

Extraction and LC-MS analyses of lipopeptides from tomato hydroponic cultures

Surface-sterilized tomato seeds (treated or not with 10 µl of Bacillus cell suspension prepared

as described below) were germinated for 5 days at room temperature in the dark and on

gelified sterile nutrient medium containing 5mM Ca(NO3)2, 5 mM KNO3, 2 mM MgSO4, 1

mM KH2PO4 and micronutrients. Germinated seedlings were transferred in sterilized 50-ml

tubes filled with nutrient solution, inoculated with strain S499 and incubated at 25 ± 2 °C in

the greenhouse with a 16 h-photoperiod. For inoculum preparation, B. amyloliquefaciens

strain S499 was grown on gelified Luria Bertani (LB) medium for 24 h. Cells were

resuspended in NaCl 0.85% and the concentration was determined by measuring turbidity at

600 nm. A final concentration of 5x107 cells/ml in the nutrient solution was used.

To evaluate in situ production of LPs, we used plantlets of approx. four weeks old.

Fresh roots (approx. 2 g RFW) were carefully separated from plants and cut into 2-3 cm

segments. The tissues were immersed in 10 ml of 90%ACN, 0.1% HCl, 0.9% Triton-X 100,

ultrasonicated for 2 min and extracted for 2h under regular vortexing. This extract was

combined with the hydroponic liquid which was collected from same plant after 4 weeks and

filtered on Whatman Paper Grade 1 to remove macroscopic insoluble material. This mix was

loaded on C18 solid-phase extraction cartridges (900 mg, Alltech), washed with 20% ACN

and lipopeptides were desorbed with 1 ml 100% ACN. The material was then concentrated 5

times using speed vacuum.

The resulting samples were analyzed by reverse phase HPLC (HPLC Waters Alliance

2695/ diode array detector) coupled with single quad mass spectrometer, (Waters SQD mass

analyser) on a X-terra MS (Waters) 150*2.1 mm, 3.5 mm column. Surfactins were eluted in

the isochratic mode (78% acetonitrile in water acidified with 0.1% formic acid) at a constant

flow rate of 0.5 ml min -1 and 40 °C. Iturins and fengycins were selectively desorbed by using

ACN gradients from 35% to 65% in 35 min and from 40% to 60% in 40 min, respectively. In

both cases, a lag time of 5 min was set after injection of the samples on the column.

Compounds were first identified on the basis of their retention times compared with authentic

standards (supplied by AIBI asbl). The identity of each LP homologue in every family was

confirmed on the basis of the masses detected in the SQD by setting electrospray ionization

source conditions in the MS as source temperature, 130 °C; desolvation temperature, 250 °C

and nitrogen flow at 500 l h-1. Specific cone voltages were used for each family and were 70

V, 60 V and 100 V for surfactins, iturins and fengycins, respectively. The positive ion mode

was used for analysis of both three families because also in the case of surfactins, a higher

signal/background ration was obtained compared to negative ion recording. Depending on the

Results-Chapter 5

87

experiment and purpose, LP samples were analyzed either in the SIR (Single Ion Recording)

or the SCAN mode. In the first case, masses of both proton, sodium and potassium adducts of

the various LP forms were used for optimal detection and in the second case, the scanned

mass range was limited to 150 mu spanning over all the masses of homologues within each

family. The amounts were calculated on the basis of the corresponding peak area in the SIR or

SCAN traces. We used 95%-pure standards as methanolic solutions with serial dilutions to

0.1 µg/ml.

LP analyses by imaging MS

Sterilized tomato seeds were immersed in S499 strain suspension at concentration of 108 cfu

for 5 minutes. The seeds were gently harvested from the suspension to be transferred on agar

plates containing sterile plant nutrient solution (PNS) as described above. The plates were

incubated at 30 °C for approx. two weeks.

Globally, the method used for sample processing is the same as described by Debois and

collaborators (2008), as well as detailed characteristics and settings for the mass spectrometer.

Briefly, lipopeptide material present at the surface of the plate was directly transferred onto

polished silicon wafer by applying a manual press for 30 s. The wafer was then allowed to dry

for 30 min under vacuum at room temperature before introducing in the mass spectrometer. A

TOF-SIMS IV (Ion-TOF, Münster, Germany) reflectron-type TOF mass spectrometer was

used for the MS imaging experiments. The primary ion source is a bismuth LMIG which

delivers Bi3+ cluster ions. The secondary ions are accelerated to a kinetic energy of 2 keV and

are postaccelerated to 10 keV before hitting a hybrid detector. Details concerning image

processing are given in the legend of Figure 3. Methanolic solutions (1 mg/ml or serial

dilutions of 0.3 mg/ml) of authentic and 95%-pure standards (AIBI asbl) were prepared and

10-µl drops were deposited on the same gelified medium as the one used for plantlet grow and

colonization. After incubation at room temperature for 2 h, the plates were processed for

imaging as described for those used for analyzing LP production by S499. Core sampling in

gelified medium were performed with a Pasteur Pipette and lipopeptides contained in the 200

µl agar piece were then extracted by adding 100 µl acidified acetonitrile/water 80/20 v/v. The

mixture was then ultrasonicated for one minute, extracted under regular stirring for 30 min

and incubated overnight at – 20 °C. Extracts were then centrifuged at 14000 r.p.m. for 10

minutes and the supernatant was analysed in LC-MS as described above.

Results-Chapter 5

88

Growth conditions for LP production by S499 cells in liquid cultures

All the S499 cultures were performed in 12-well microplates by dispensing 2.5 ml of medium

per well. Plates were incubated under agitation on a rotary shaker for 72h at 28 °C. In all

cases, the pH of the different media was adjusted to 6.8 before filter-sterilization. Over the

incubation time, bacterial growth was monitored 24 h and 48 h after inoculation by measuring

turbidity (OD) at 600 nm. At the end, cultures were centrifuged (25 min, 13000 g) and

lipopeptides were extracted from supernatant samples using C18 solid-phase extraction

cartridges before analysis by LC-MS as described above. Three wells per plate were used for

each medium but corresponding supernatants were combined to yield a unique sample

submitted to extraction and LC-MS analysis. However, the experiments were repeated at least

three times allowing to calculate means values and standard deviations mentioned in the

results.

To test LP production in the presence of natural exudates, we collected the nutrient

solution of non-inoculated tomato plants grown for four weeks in the greenhouse. This

medium was supplemented with 2 g/l of (NH4)2SO4 to compensate for the lack of available

nitrogen. In order to evaluate the production of LPs in various carbon sources, the individual

substrates were added at a final concentration of 5 g l -1 to a minimal medium composed by

MgSO4 7H2O 0.5 g l -1, K2HPO 4 1.0 g l -1, KCl 0.5 g l -1, Yeast extract 1.0 g l -1, Fe2(SO4)3

1.2 mg l -1, MnSO4 0.4 mg l -1, CuSO4 1.6 mg l -1, (NH4)2SO4 1 g l -1.

LP production in function of oxygen concentration

These experiments were conducted in 2-L bioreactor using a medium described in

Nihorimbere and colleagues (2009) and recomposed on the basis of tomato root exudate

composition as described by Kamilova and collaborators (2006). The fermentor is equipped

with automated pH control maintained with H3PO4 (3N) and KOH 50% at a value of 6.8. The

fermentor working volume of 1.5 l was inoculated with a 16 h old preculture (25 ml) prepared

in the same medium. The temperature was maintained at 30 °C. For aerobic conditions,

aeration rate and agitation were fixed respectively at 0.5 V.V.M. (volume of air per volume of

medium per minute) and 150 r.p.m (rotation per minute). In these conditions, the oxygen

concentration transiently decreased to zero during a few hours corresponding to the mid-/late

exponential growth phase but oxygen was not limiting for the remaining part of the culture.

Oxygen-depleted condition was created by replacing air with nitrogen for flushing the reactor

through the sparger until pO2 value reached zero. This nearly-anaerobic condition was

maintained over the culture time by subsurface injection of nitrogen at a low flow rate.

Results-Chapter 5

89

Results

Strain identification

The Bacillus strain S499 used in this study was formerly identified as B. subtilis based on

basic bacteriological tests (cell morphology, gram stain, oxidase and catalase reactions) and

on API fermentation profile. The whole-cell fatty acid profiling confirmed that strain S499

belongs to the Bacillus subtilis complex but gyrA sequencing revealed that its closest relatives

are two strains of B. amyloliquefaciens, LMG 12384 and LMG 17599 with respectively

99.8% and 98.1% of sequence similarity (supplementary data, Figure 1S). These data show

that strain S499 most probably belongs to the closely related B. amyloliquefaciens species

rather than B. subtilis.

LC-MS determination of LP pattern produced by strain S499 in vitro

The first objective of the work was to obtain an in-depth characterization of the lipopeptide

pattern synthesized by strain S499 upon growth in erlenmeyer flasks in a culture medium

defined by Jacques and collaborators (1999) for high LP productivity (OM medium). To that

end, we have optimized an LC-MS method providing both sensitive detection and reliable

identification/quantification of the three families of LPs secreted in the medium. As detailed

in the Material and Methods section, HPLC parameters and settings of the electrospray source

in the mass spectrometer were adapted to allow the resolution and optimal detection of most

of the peaks corresponding to the various homologues within each LP family produced by the

strain (Figure 1). By using the SCAN mode of detection (over mass range of 950 mu to 1150

mu for iturin and surfactin and from 1400 mu to 1550 mu for fengycin) in the spectrometer,

detection limits for surfactin, iturin and fengycin were respectively 2.5 ng, 10 ng and 25 ng

injected on the HPLC column. This sensitivity could be approx. 5 times enhanced by using

the single ion recording (SIR) mode which allows to markedly reducing the signal to

background noise ratio. In most cases, specific LC-MS methods were thus used for surfactin,

iturin and fengycin analyses. In some instances, a single method allowing the simultaneous

measurement of both three families was used as it.

Results-Chapter 5

90

I

F

S DAD

MS

MS

MS

DAD

DAD

I1

I3I2

I4

S1

S2S3

S4

S5

S6

S7

S8 S9S10

I5

F3

F7F5

F6F8

F9F10

F13

F19F4F14

F15 F16

Figure 1. Typical LC-MS separation and identification of the three lipopeptide families (I, iturins; F, fengycins; S,surfactins) produced by strain S499 in optimized medium. DAD, HPLC chromatograms obtained with diode arraydetection of material eluted under family-specific conditions. MS, corresponding signal from the mass spectrometerupon Single Ion Recording detection using masses specific for each homologue within each family.

As shown in Figure 1, the various surfactin homologues could be easily identified on the

basis of their retention times in the UV-Vis trace. However, an accurate identification of iturin

and fengycin homologues simply based on the chromatogram generated by the diode array

detector is not possible due to contaminating material produced by the strain that co-eluates.

Coupling with mass spectrometry is thus necessary for peak identification but also to

significantly improve the sensitivity. Homologues corresponding to the peaks in Figure 1 are

listed in Table 1. No additional MS signal in the range 900-1550 mu were recorded upon LC-

MS analysis of crude S499 supernatant strongly suggesting that other LP variants putatively

Results-Chapter 5

91

synthesized by Bacillus (Ongena and Jacques, 2008) are not formed by the strain (data not

shown).

Table 1. Assignement of peaks detected by LC-ESI-MS in Figure 1 and corresponding to surfactins, fengycins and iturins produced by strain S499 in optimized medium. Surfactin (A, Leu in position 7) and fengycin (A, Ala-7 and B, Val-7) variants were identified on the basis of typical MS fragments and amino-acid composition (data not shown). However our data do not allow to discriminate between iturin A and mycosubtilin which only vary in the order of the last two Asn and Ser residues of the peptide. Only protonated forms are mentioned but in most cases, sodium and potassium adducts were also detected for the various homologues. n1 means that there is a double bond in the fatty acid chain of the fengycin homologue.

Peak m/z [M+H]+ Assignement S1 994.8 Surfactin A C12 S2 1008.8 Surfactin A C13 S3 994.8 Surfactin A C12 S4 1022.8 Surfactin A C14 S5 1022.8 Surfactin A C14 S6 1008.8 Surfactin A C13 S7 1036.8 Surfactin A C15 S8 1022.8 Surfactin A C14 S9 1050.8 Surfactin A C16 I1 1043.7 Iturin A/mycosubtilin C14 I2 1057.7 Iturin A/mycosubtilin C15 I3 1057.7 Iturin A/mycosubtilin C15 I4 1071.7 Iturin A/mycosubtilin C16 I5 1071.7 Iturin A/mycosubtilin C16 F4 1435.9 Fengycin A C14 F5 1449.9 Fengycin A C15 F6 1449.9 Fengycin A C15 F7 1463.9 Fengycin B C14 F8 1463.9 Fengycin A C16 F9 1477,9 Fengycin B C15 F10 1477.9 Fengycin A C17 F13 1491,7 Fengycin B C16 F14 1477.9 Fengycin A C17 F15 1505,7 Fengycin B C17 F16 1467.8 Fengycin A C15 n1 F19 1491,7 Fengycin A C18

LP pattern of strain S499 developing on roots of hydroponic plants

In a first approach to investigate in situ LP production, experiments were conducted on

tomato plants grown in a hydroponic culture mode to facilitate colonization study and

quantification of the metabolites secreted into the surrounding environment. Efficient root

colonization is a key event required for optimal expression of biocontrol activity of beneficial

rhizobacteria acting through competition for substrate, antibiosis or plant resistance triggering

(Lugtenberg et al., 2001). Rhizobacterial populations must reach a threshold level sufficient

to produce the active metabolites in sufficient amounts. Plate counts of bacterial cells

adhering to tomato roots were performed 22 days after inoculation and Bacillus colonies were

detected on the basis of their typical morphology at concentrations up to 5x105 cfu/g root.

Results-Chapter 5

92

However, S499 populations could not be accurately quantified due to the lack of natural

antibiotic resistance or antibiotic-resistant derivative.

At the same time, lipopeptide recovery was performed by combining material extracted

from the nutrient solution with the one recovered by extraction of root tissues using detergent.

This is justified by the fact that surfactins secreted by colonizing Bacillus cells may readily

insert into the membrane structure of the plant cells as suggested by the rapid disappearance

of the lipopeptide from the medium when supplemented in hydroponically-grown tomatoes or

when added to tobacco cell suspension where more than 90% of the added surfactin

embedded within minutes into the plasma membrane (data not shown). LC-MS analyses using

specific methods in the SIR mode were performed on 100-times concentrated extracts in order

to determine their precise content in surfactins, iturins and fengycins. A high variability was

observed in the LP quantities recovered from one plant to another and data presented in

Figure 2 are means calculated from 4 independent cultures using each time the material

collected from three plants as a single sample. These data show that members of the surfactin

family were mainly detected while iturins were present but in low concentrations and

fengycins are seemingly not produced under these conditions. In this work, we could not

determine precisely the S499 population on roots but considering a biomass of 5x105 cell/g, it

corresponds to surfactin and iturin productivities of 26 µg/108 cells and 1.7 µg/108 cells,

respectively. It also corresponds to surfactin and iturin concentrations in the hydroponic

medium of approx. 120 nM and 8 nM, respectively.

Results-Chapter 5

93

itu

srf

fen

Rel

ativ

e ab

und

ance

Elution time

Figure 2. Representative ESI mass spectrometric signal recorder during LC-MS analysis of lipopeptidesproduced byB. amyloliquefacienscolonizing roots of hydroponically-grown tomato plants. Detection wasperformed in the SCAN mode (m/z range 900-1550). Elution time periods for every LP family are indicated.Every single peak represents a specific homologue (srf, surfactin; itu, iturin; fen, fengycin). The relativeproportions of total surfactins, iturins and fengycins were calculated using authentic standards and are shownin the boxed graph. Data and standard errors were calculatedfrom 4 samples obtained in 4 differentexperiments but each consisting of approx.2 g fresh weight of combined root materialcoming from 3 plants.

Mass spectrometry imaging of LP pattern

Recent developments in Time-Of-Flight Secondary Ion Mass Spectrometry (TOF-SIMS)

imaging allow the two-dimensional analysis of some lipidic compounds in various biological

samples and it was successfully applied for determining the distribution of surfactins secreted

by B. subtilis in a swarming pattern (Debois et al., 2008). In a second approach to evaluate in

situ LP production by strain S499, we thus wanted to use this imaging technique to analyze

the lipopeptides secreted by cells developing on roots of tomato plantlets grown in a

gnotobiotic system. Disinfected tomato seeds were germinated on gelified medium, bacterized

and colonization was observed after a two-week incubation period. As shown in Figure 3,

S499 readily colonizes the root system by forming macro-colonies that cover many root hairs.

Mean value for S499 population established on roots at that time was 2.3 x 107 CFU/g root

FW. Under these conditions, no significant root-associated microflora was observed on

control plantlets after two weeks of incubation (detection limit of 102 CFU/g root FW).

Lipopeptides secreted by these S499 cells and present at the surface of the gelified

medium were then analyzed by TOF-SIMS after transfer onto a silicon wafer. Scanning over

Results-Chapter 5

94

a small area around S499-coated root hair tips revealed the relative amounts of LP from the

three families (Figure 3d).

a b c

S I F

d

e

*

*

*

*

**

****

f

S C12 S C13 S C15

Figure 3. Imaging MS of LPs produced by strain S499 colonizing roots of tomato plantlets grown on gelified medium in Petriplates. A, microscope image of the main root and root hairs colonized by S499 cells forming macrocolonies. B, microscopeimage of a part of the transferred colonization pattern transferred onto silicon wafer. C, v ideo image of the scanned zonerecorded with an integrated camera inside the TOF-SIMS apparatus. D, High-definition scan of the lipopeptide distribution atthe extremity of biofilm-embedded root hairs. TOF-SIMS images of the sum of surfactin (S), iturins (I) and fengycins (F)ionsare shown. The colour scale indicates the relative amounts of total surfactin/iturin/fengycin ions detected. E, Relativedistribution of surfactin homologues in the same pattern. F, corresponding partial mass spectra of surfactins/iturins (left) andfengycins (right). Similar images and results were obtained by analyzing two other root zones from plantlets inoculated andgrown independently.

Results-Chapter 5

95

Based on total counts in the MS detector corresponding to the more abundant homologues of

the three families (surfactin C14, iturin C15 and fengycin C16), the quantity of surfactin is

respectively 11 times and 34 times higher than iturin and fengycin (Figure 3f). Analysis of

concentrated solutions of authentic standards showed a detection limit of approx. 3 µg/ml.

Almost the same results were obtained upon scanning a similar small area around the

extremity of root hairs in an independent experiment (data not shown). TOF-SIMS imaging

over a larger surface (5 x 5 mm) of the colonized-root system also revealed that surfactins are

produced in much higher quantities compared to iturins and fengycins, respectively 14 times

and 84 times (supplementary data, Figure 2S). In all cases, surfactins are much more abundant

in the periphery than on biofilm suggesting that these hydrophobic LP do not accumulate on

or around the producing cells but readily diffuse into the gelified medium as demonstrated for

swarming cells (Debois et al., 2008).

TOF-SIMS imaging thus provides useful information to study in situ production of LPs

but this technique does not allow quantitative measurements and only molecules present at the

surface are detected. Moreover, the analysis of pure LP solutions (10 µl-drop deposit on the

same gelified medium) in the same conditions also revealed that fengycin detection is 5 times

less sensitive either because they are less efficiently ionized in the source of the mass

spectrometer or because they are less efficiently extracted from the surface of the medium to

adsorb onto the silicium wafer or both. In order to better determine the actual concentrations

of the three types of LPs, we have extracted the compounds contained in agar cylinders (5

mm diameter) collected by core-sampling from the medium in the close vicinity of S499-

colonized roots. Analysis of these extracts by LC-ESI-MS also revealed much higher relative

quantities for surfactin compared to the two other LP families as mentioned in Table 2. It is

not possible to express the quantities measured in term of productivity but it corresponds to

quite high concentrations that were may be generated by S499 macrocolonies locally around

the roots. Surfactins are dominant but fengycins and iturins are also present in concentrations

from 2 to 9 µM (supplementary data, Figure 3S).

Results-Chapter 5

96

Table 2: Comparison of the lipopeptide patterns produced by strain S499 in various conditions. RZ (root zone), LPs produced by cells colonizing roots of tomato plantlets grown in Petri plates and extracted from the gelified medium surrounding roots after core sampling. NE, RE, OM, LPs quantified from liquid cultures in natural exudates, recomposed exudates (Kamilova et al., 2006) and optimized medium (Jacques et al., 1999) respectively. OM, b is LP production under conditions conducive for biofilm formation (growth in 96-well plates) where most of the cells are in the pellicle by contrast with OM, p which means LP production by planktonic cells in agitated flask. O2+ and O2- mean LP production following growth in well aerated and in nitrogen flushed and oxygen depleted bioreactor respectively.

Surfactin Iturin Fengycin Total

LPs

Conc.

(mg/l)

Rel. prop.

(%)

Conc.

(mg/l)

Rel. prop.

(%)

Conc.

(mg/l)

Rel. prop.

(%)

Prod.

µg/108

cfu

RZ* 58 83 9 12 3 4 nc

NE 32 ± 7 58 15 ± 8 27 8 ± 6 14 147

RE 364 ± 112 61 175 ± 65 27 70 ± 24 11 304

OM, b 179 ± 46 80 27 ± 18 13 17 ± 11 7 nc

OM, p 628 ± 108 51 321 ± 55 26 272 ± 71 22 436

RE, O2 +* 556 66 212 25 73 8 341

RE, O2 -* 252 67 108 29 10 2 740

Data are means and standard errors calculated from at least four repeats of the experiment otherwise stated (*) representing average values from two independent assays. nc: means non calculated due to non accurate biomass quantification.

LPs secreted upon growth in the presence of root exudates

In order to test more specifically the effect of the nutritional status imposed by the plant on

LP production, the S499 strain was grown in agitated liquid culture in the presence of root

exudates as sole carbon source (NE medium). These exudates were collected from hydroponic

tomato plants after 23-25 days of growth in nutrient solution under controlled conditions.

S499 growth was obviously substrate-limited since low biomass levels (OD600 between 0.35

and 0.45) were observed. It resulted in much lower concentrations of surfactins, iturins and

fengycins in the medium compared to those observed in OM (Table 2). From a quantitative

point of view, the secretion thus appears much less efficient than in optimized conditions.

However, in term of productivity, LP synthesis efficiency in natural exudates corresponds to

the third of the one observed in OM and thus globally, these data confirm that exudate

components are conducive for LP synthesis. Qualitatively, the relative proportion of iturins

compared to surfactins is similar to the one observed in optimized medium but the formation

of fengycins appears to be significantly less efficient.

Results-Chapter 5

97

LPs formed in the presence of specific carbon sources

Strain S499 was also cultivated in a minimal medium (RE medium) containing as sole carbon

sources the sugars, organic acids and amino acids typically found in the tomato exudates and

added in the proportions determined by Kamilova and collaborators (2006). Increased levels

of biomass and LP synthesis (Table 2) compared to natural exudates can be explained by

higher concentrations of nutrients in the medium. The relative proportions of the three LP

families produced in this medium are similar to those observed in NE.

The main representatives of these substrates were tested individually for their potential

to support growth of strain S499 and lipopeptide synthesis. Most of these C sources supported

growth to an almost similar level (data not shown) but it clearly appears from Figure 4 that

surfactin synthesis is significantly higher in the presence of organic acids as unique C sources

compared to sugars. Thought it may vary in function of various plant factors, succinic and

citric acids are the more abundant substrates released by uninoculated tomato roots and also

represent adequate nutrient sources utilizable by S499 for surfactin production (Figure 4A).

By contrast, iturin productivity is slightly higher upon growth in the presence of sugars

compared to organic acids (Figure 4B) and fengycins that are only poorly produced in the

presence of citrate, succinate or malate (Figure 4C).

Results-Chapter 5

98

A

C

B

Figure 4. Lipopeptide productivities in the presence of various carbon sources typically found in tomato exudates. A, surfactins; B, iturins; C, fengycins. Data are mean values and standard errors calculated from five independent experiments.

Results-Chapter 5

99

LP production upon biofilm formation and oxygen limitation

When grown in the same optimized medium as used for flask cultures but in 96-well plates,

strain S499 readily forms pellicle at the surface of the liquid medium. We wanted to evaluate

the effect of such a pellicle formation on the pattern of LP produced since the macro-colonies

developing on roots may be considered as biofilm-related structures (Bais et al., 2004; Ramey

et al., 2004). The results presented in Table 2 show that surfactin synthesis in S499 cells

forming pellicles is very effective compared to iturins and fengycins. The relative proportion

of the three LP families produced under these conditions is similar to the one observed upon

root colonization. However, it significantly differs from the one secreted upon growth of S499

as planktonic cells in optimized liquid medium where iturins and fengycins are secreted in

much higher amounts corresponding to approx. half of the surfactin produced (Table 2).

The effect of oxygen concentration was evaluated by cultivating the strain in a

bioreactor upon two different aeration rates. Final cell densities were much lower in the non-

aerated culture but a significant biomass was measured, showing that strain S499 could

develop under limited availability or absence of oxygen molecular as a terminal electron

acceptor. Some Bacillus strains can utilize nitrate as an alternative electron acceptor (Mach et

al., 1988; Glaser et al., 1995) or are able to grow anaerobically on minimal media in the

absence of terminal electron acceptors (Hoffmann et al., 1988; Nakano et al., 1988; Ramos et

al., 2000). As shown in Table 2, strain S499 produces iturin, surfactin and fengycin

lipopeptides upon growth in the two conditions of oxygen availability tested. In terms of total

productivity, LPs are more efficiently produced under oxygen-starved conditions but the

relative production rate of the three families is similar in both conditions of aeration.

Influence of temperature and pH on LP production

To test the effect of temperature, cells were grown until late stationary phase in Erlenmeyer

agitated flaks in RE medium and LP concentrations are determined by LC-MS (Table 3). The

production of total amount of iturins and surfactins increased with temperature while the

production of fengycins is higher at 25 °C compared to 20 °C but then decreased beyond that

temperature. It appears that cell population was not significantly different among the tested

temperatures (not shown).

The effect of pH on LP production was tested on B. amyloliquefaciens S499 cultured in

fermentor with automatic regulation allowing the medium to be maintained at different

specific pH values (Table 3). Not surprisingly, optimal growth of the strain was observed at

pH 7 but lipopeptide production under this condition was very low and only traces of

Results-Chapter 5

100

surfactin could be detected. Very low biomass level was observed by growing S499 at a

controlled pH value of 8 and no LP were detected. Significant amounts of surfactin and iturin

could only be measured in culture supernatant after growth at pH continuously adjusted to 5.7

but the quantities recovered were still fairly reduced compared to those observed upon growth

in bioreactor in the same conditions but without pH control (RE, O2 + in Table 1).

Table 3. Lipopeptide productivities (µg/108 cfu) in function of temperature and pH. Each value is an average of results obtained from two independent experiments with similar results. OD: Optical density, Srf: surfactins, Itu: Iturins, Fen: Fengycins, ND: not detected.

Parameters OD 600nm Srf Itu Fen 20 °C 1.4 510 245.46 98.46 25 °C 1.15 862 367.31 231 30 °C 1.32 1295.2 683.17 84.8 5.7 0.43 187 58 ND 7.0 2.75 1.2 ND ND 8.0 0.17 ND ND ND

Changes in the relative proportions of LP homologues

Under laboratory growth conditions in OM medium, strain S499 usually produces a

mixture of C12/C13/C14/C15/C16 surfactin homologues in the proportion of 7/17/48/32/1 in %.

Interestingly, the proportion of C15 tremendously increased in the pattern of surfactins that are

secreted by the strain evolving on roots of hydroponic tomato plants (Figure 5). As a matter of

fact, higher levels of C15 surfactins were also detected after growth in liquid cultures in the

presence of natural root exudates. Globally, SC14 and SC15 together represent more than

80% of the total surfactins produced in vivo which is in agreement with imaging data showing

much higher MS signals corresponding to C15 surfactins compared to SC12 and SC13 (Figure

3e). However, this increase is difficult to interpret on a nutritional basis since organic acids as

sole C substrates are only slightly more conducive for C15 synthesis than sugars (Figure 5).

On another hand, higher proportions of long chain fengycins (mainly C18) are also seemingly

produced by the strain upon growth in the presence of natural exudates compared to those

formed in the OM medium but globally, these changes are much less obvious than in the case

of surfactins and these minor differences have to be confirmed in additional experiments.

Concerning iturins, the C16 homologue is usually produced in most of the media tested in

laboratory conditions but was not detected in LP extracts prepared from S499-colonized

plants. In these in vivo samples, iturin C15 represents more than 85% of the total and thus

much higher quantities compared to the other conditions (Figure 5). However, given the small

amounts detected in the hydroponic systems, other assays are also required. In bioreactor

cultures performed to evaluate the influence of aeration rate, the relative proportions of

Results-Chapter 5

101

C12/C13/C14/C15/C16 surfactin homologues were respectively 2/45/14/28/11 and 6/9/62/19/2 in

aerobic and anaerobic conditions.

S

I

F

OM In situ NE Glu Malic

Figure 5. Distribution of homologues varying in the length of the fatty acid chains within the three families of lipopeptides(surfactins (S), iturins (I) and fengycins (F)) produced byS499 after growth upon various conditions and as determinedby LC-ESI-MS. OM., LPs produced in optimized medium (Jacqueset al. 1999);In situ, in hydroponic cultures; NE, in natural exudates;Glu and Malic, in the presence of glucose and malic acid as sole carbon sources. Data are mean values calculated from 3independent cultures.

Thus higher quantities of C15 and C16 homologues were observed upon aerated conditions but

the global proportions of surfactins with fatty-acid chains higher than C13 is greater upon

growth under oxygen-depleted conditions (Figure 6B). There was no significant difference in

homologues of iturins and fengycins produced by the strain in different aeration conditions

(results not shown).

Concerning surfactins and iturins, there is no marked differences between the relative

proportions of individual homologue produced considering the temperature of growth. By

contrast, the fengycin homologues with short chain were abundant at 20 °C whereas at 25 °C

and 30 °C, C16 and C18 fengycins dominate with 50% and 33% of total amounts, respectively

(Figure 6A).

Results-Chapter 5

102

20 C 25 C 30 C

S

I

F

S

O2 O2 d

Figure 6. Relative proportions of homologues varying in length of fatty acid chain within the threefamilies of LPs vis-à-vis temperature of growth (20 C, 25 C, 30 C) and oxygen availability (O2:oxygen supplied, O2d: oxygen depleted). Data are mean values obtained from two replications of oneconclusive experiment.

A

B

Discussion

Through the so-called rhizosphere effect, plants create a specific environment for microbial

development in the soil surrounding roots. The influence of rhizosphere conditions on the

inhabiting microflora is being extensively studied at the population level because of the

Results-Chapter 5

103

interest in engineering this environment in order to improve the efficacy of beneficial

microbes as biocontrol agents. Much less is known about the influence of rhizosphere factors

on the biosynthesis of specific metabolites responsible for biocontrol activity of these

beneficial strains. In this context, the present work was initiated to accurately identify and

quantify lipopeptides produced by Bacillus in situ. The combination of HPLC and mass

spectrometry techniques allowed us to determine the whole pattern of LPs from the surfactin,

iturin and fengycin families produced by the biocontrol strain B amyloliquefaciens S499

evolving on roots. Multiple studies used optimized mass spectrometry techniques to

characterize or compare the LP patterns produced by Bacillus isolates under laboratory

conditions (Vater et al., 2002; Price et al., 2007; Stein, 2008; Athukorala et al., 2009; Vater et

al., 2009). However, only few works published up to now deled with LPs secreted in the

natural environment in general and in the rhizosphere in particular. Surfactin and/or iturins

were recovered from cucumber roots surrounding soil inoculated with Bacillus strain QST713

(Kinsella et al., 2009) and from tomato rhizosphere inoculated with strain BGS3 (Ongena et

al., 2007; Nihorimbere et al., 2009) and strain RB14-C (Asaka and Shoda, 1996). Fengycins

and Iturins were also occasionally detected in the phyllosphere and on fruits, (Touré et al.,

2004; Romero et al., 2007). However, in all these studies, only some LP variants or even

homologues were analyzed. This work is, to our knowledge, the first report dealing with

determination of the whole array of surfactin, iturin and fengycin homologues that may be co-

produced in the soil by a natural strain of Bacillus.

B. amyloliquefaciens S499 retain biocontrol potential and efficiently co-produces the

three lipopeptide families under well defined and optimized in vitro conditions. However, LC-

ESI-MS analyses of the LPs secreted by S499 cells colonizing tomato roots revealed much

higher relative quantities of surfactins. Fengycins and iturins were detected in much smaller

amounts or only traces (Figure 1). This strongly suggests that the physiology of S499 cells

developing on plant tissues is conducive for surfactin synthesis but not for an efficient

production of the two other types of lipopeptides. We previously used the LacZ reporter

system in B. subtilis strain BGS3 to provide a first evidence that surfactin genes may be

readily expressed under rhizosphere conditions (Nihorimbere et al., 2009). However, BGS3

derives from the B. subtilis 168 Marburg strain which can no longer be viewed as natural

isolate since it has been acclimated to laboratory conditions for several decades and exposed

to X-rays in the mid-1940’s.

Such in vivo lipopeptide signature is of biological relevance since the three LPs families

actually display distinct functions in the ecology of the producing strains in the phytosphere

Results-Chapter 5

104

and notably in interactions with co-existing organisms, including bacteria, fungi, oomycetes,

protozoan predators and plants (Ongena and Jacques, 2008; Raaijmakers et al., 2010). LPs

play a crucial role in the antagonism developed towards fungal pathogens but it mainly

concerns iturins and fengycins while surfactins has been demonstrated to be mostly

antibacterial and antiviral. Beside their antimicrobial properties, surfactin and to a lower

extent iturins, are also involved in root colonization and surfactin has been identified as

elicitor of the systemic stimulation of host plant immune system (Ongena and Jacques, 2008).

Beside these biocontrol-related traits, LPs may also have different roles in the development

and survival of B. subtilis strains in their natural habitat. It includes increasing of the surface

area and bioavailability of hydrophobic water-insoluble substrates, heavy metal binding,

bacterial pathogenesis, quorum-sensing, motility and biofilm formation (Ron and Rosenberg,

2001; Mulligan, 2005; Raaijmakers et al., 2010). Intrinsically, there is a competitive

advantage for Bacillus isolates to co-produce surfactin, fengycin and iturin with specific roles

and targets in the context of biocontrol of plant diseases. However, our results strongly

suggest that in situ expression of the related biosynthesis genes may considerably differ from

one family to another. This could not be demonstrated experimentally because we were

unable to generate reporter derivatives of strain S499 which is highly recalcitrant to

transformation like many natural B. subtilis/amyloliquefaciens isolates. Investigating the

differential expression of srf, fen and myc genes under natural soil conditions in S499 or in

another natural isolate co-producing the three families, is one of our main future prospects.

Such a surfactin-enriched LP pattern produced by strain S499 in the tomato rhizosphere

under non sterile conditions was also observed by TOF-SIMS imaging after root colonization

of tomato plantlets on gelified medium (Figure 3). Under such sterile conditions, S499 cells

can readily grow without having to compete with auxiliary microflora. Seemingly, the

presence of other microorganisms sharing the same ecological niche does not significantly

impact on the LP pattern secreted by S499 since surfactins again represent the large majority

of the lipopeptidic compounds produced in the gnotobiotic system. From a technological

point of view, this imaging MS technique appears to be a yet emerging but extremely

powerful method to study the production of lipidic molecules by microbes developing on

solid supports. It was successfully used by Debois and collaborators (2008) to study the

heterogeneity of surfactin secretion in a swarming pattern of Bacillus cells. In that study,

fentomole amounts of C15 surfactin could be detected. Surfactin as major LP produced in

planta by S499 is a trend that was also observed in preliminary experiments with corn and

lettuce. However, the LP signature may also be strongly influenced by the host plant species

Results-Chapter 5

105

since almost no production of surfactin was observed upon growth on cucumber and soy

(Ongena et al., unpublished data).

It is clear from our results (Figure 4) that the S499 LP pattern can be markedly

influenced by nutritional factors such as carbon source. The proportions of

surfactin/iturin/fengycin readily differed in the presence of sugars or organic acids as sole

carbon source. Much lower relative quantities of iturins and fengycins are produced by S499

in the presence of the last substrates. It may partly explain the lower production levels of

these two LPs in planta because organic acids are the main components of tomato exudates

(Kamilova et al., 2006). However, the proportions of iturins and fengycins secreted upon

growth in natural exudates collected from hydroponically-grown tomatoes were higher (Table

2). So, the nutritional context in term of carbon source created by the plant exudation may

favour surfactin synthesis to some extent but is not sufficient by itself to explain the very high

proportion of this LP secreted in situ compared with the two other families. The effect of the

non-polar fraction of natural exudates (hydrophobic material eluted from C18 cartridge with

pure methanol and potentially containing phenolics, flavonoides, etc...) on LP production was

also tested. However, supplementing the RE medium with a concentrated extract of this non-

polar fraction had no effect on the pattern of LPs both qualitatively and quantitatively (data

not shown).

The oxygen status can also affect the physiology of both the host plant and the

colonizing bacterium and thus alter the interactions. In agreement with previous works (Hbid

et al., 1996; Jacques et al., 1999), final LP concentrations in the supernatant were higher in

well aerated cultures. However, results from the present study show that productivities for

both three LP families were quite higher under oxygen starvation conditions (Table 2).

Oxygen depletion is common in many soils and especially in the rhizosphere, notably due to

root and microbe-associated consumption (Højberg et al., 1999). However, the bacterium can

obviously adapt to such conditions and our results suggest that strain S499 can well develop

and produce LPs on roots where oxygen is not abundant. It agrees with previous work

performed with another strain demonstrating that srfA expression and surfactin production

still remain very effective in oxygen-starved culture conditions (Nihorimbere et al., 2009). A

positive influence of O2 limitation on surfactin production has also been demonstrated by

modifying the oxygenation method (Lee et al., 2001) or by modulating the oxygen transfer

rate in bioreactor (Yeh et al., 2006). The introduction of oxygen limitation, which redirects

the energy flux into product synthesis, has led to the highest productivity of surfactin by

Bacillus subtilis C9 (Kim et al., 1997). That said, our bioreactor experiments also revealed

Results-Chapter 5

106

that the level of oxygen availability during growth of S499 has no marked effect on the

relative proportions of the three LP families (Table 2). Therefore, this parameter does not

deserve further consideration for explaining the lipopeptide signature in planta.

By contrast, our results showed that surfactin synthesis in S499 cells forming biofilm is

very effective compared to iturins and fengycins with relative proportions of the three

families similar to the ones observed in rhizosphere extracts (Table 2). In situ, root-colonizing

rhizobacteria such as Bacillus typically develop in macrocolonies or form biofilms and their

actual physiology is thus probably not related to a planktonic state. Few studies have

accurately evaluated the consequences of growing on solid support on LP synthesis but some

clues arose from mass spectrometry studies of LPs produced by B. subtilis cells developing on

gelified media showing that patterns of surfactins and fengycins may be qualitatively and

quantitatively modified compared to cells living freely in liquid cultures (Leenders et al.,

1999; Nihorimbere et al., 2009). In another approach, Gancel and collaborators (2009) have

also demonstrated an enhancement of surfactin production upon cell immobilization on

polymer particles. So living in root-adhering macrocolonies may readily influence the pattern

of LPs secreted by Bacillus cells.

In batch cultures, the synthesis of the three LP families occurs at different growth

stages. Surfactin typically accumulates in the medium during the late exponential phase while

the syntheses of fengycins and iturins are delayed to the stationary phase. It means that

expression of the corresponding genes is somewhere related to the growth rate of the

bacterium. Such effect of growth rate on surfactins’ synthesis was confirmed by previous

experiments performed in chemostat bioreactors which showed that srfA expression

significantly decreased with increasing µ (Nihorimbere et al., 2009). A similar relationship

between µ and srfA expression was observed with another Bacillus subtilis isolate derived

from the surfactin over-producing strain JH642 (Kakana, 2005). Interestingly, the

development of established bacterial populations in the rhizosphere is actually restricted by

nutrient availability due to low root exudation rate (Lugtenberg et al., 2001; Bais et al., 2006).

Bacillus cells on roots can thus be in a nutrient-starved state driving cell physiology to slow

growth rate which is apparently conducive for surfactin but not fengycin or iturin synthesis.

The effect of temperature on LP production by Bacillus has not been studied

extensively. It seems globally that the expression of synthesis genes is not tremendously

influenced by this parameter (Ohno et al., 1995; Fickers et al., 2008) but, changes in

production level were actually observed suggesting that protein turn-over or some

downstream regulatory processes may be affected (Fickers et al., 2008). Higher temperatures

Results-Chapter 5

107

(37 °C) favoured surfactin production by B. subtilis strains RB14 (Ohno et al., 1995) and

ATCC6633 but not by strain S499 for which increasing temperature from 30 °C to 40 °C is

detrimental for LP production (Jacques et al., 1999). In our experiments, surfactin and iturin

productions are higher at 30 °C than 20 °C but we did not test higher temperatures. This is

however contrasting with the trends observed for two others Bacillus strains for which iturin

production clearly decreased with higher temperatures (Fickers et al., 2008). So the incidence

of temperature on LP synthesis seems to be strain-specific and may be related to the synthetic

pathways of fatty acids forming cellular membranes. The fatty acid composition of these

membranes is altered by varying the growth temperature to maintain the proper membrane

rigidity at a given temperature (Phae et al., 1990). It has been reported that the relative

amount of long-chain fatty acids increases with increasing temperature for five Bacillus

strains (Kaneda, 1991) as well as Clostridia (Huang et al., 1993), mainly because the longer

the chain of a fatty acid is, the higher the melting temperature of the fatty acid becomes.

Lipopeptide production by strain S499 in bioreactor under controlled pH was very low.

Significant amounts of surfactins but not iturins or fengycins, could be detected only at pH

5.7 and these results have to be taken with care. However it is in agreement with the fact that

surfactin gene expression is pH-dependent and also agrees with a previous study conducted

with the same strain showing a clear negative effect of raising the pH value from 6 to 8 on

biosurfactant production (Akpa et al., 2001). More generally, it has been reported that many

soil-bacterial strains are known to produce metabolites at maximum when they are cultured at

pH below neutrality (Abushady et al., 2005). For instance, rhamnolipids production by

Pseudomonas sp. was at its maximum in a pH range of 6.0-6.5 and decreased sharply above

pH 7 (Guerra–Santos et al., 1984). Globally, it appears from these data that an efficient

synthesis of LPs may be favoured in acidic to neutral environments, which is often the case

for rhizosphere due to multiple processes related notably to plant exudation and microflora

activity (Hinsinger et al., 2009).

Whether in situ productivities may generate rhizospheric LP concentrations sufficient to

provide any biological effect remains questionable. LPs are generally active at 2-30 µM

depending on the role and target organism considered but data mentioned in Figure 2

correspond to approx. 0.12 µM surfactins and 0.008 µM iturins in the nutrient solution. In the

previous study conducted with B. subtilis BGS3, we observed a LP concentration of 1.8 µM

in the hydroponic medium surrounding tomato roots (Nihorimbere et al., 2009). Even higher

quantities of both surfactin and iturin were detected in the cucumber rhizosphere colonized by

strain QST713 (Kinsella et al., 2009). Possible enzymatic degradation by auxiliary microflora

Results-Chapter 5

108

and incomplete extraction may contribute to underestimate the actual abundance of LPs in the

vicinity of the root system in the soil on the basis of results presented here. Another

possibility is that S499 retains only poor ability to produce LPs while developing on roots.

However, under in vitro conditions, this isolate produces a quite high level of surfactin

compared to other B. subtilis/amyloliquefaciens strains (data not shown). Moreover, LC-ESI-

MS measurements of LPs secreted into the gelified medium by S499 cells in plate

experiments (Table 2) revealed that surfactin concentration may locally reach 50 µM

concentrations within the diffusion zone close to the colonized roots. Because LP destruction

by other organisms cannot occur and dilution effect is reduced under these conditions

compared to the hydroponic system, such 50 µM concentration probably better reflects the

amounts of surfactin that may be actually secreted by S499 evolving on tomato roots.

Concentration of surfactins in the tomato rhizosphere may thus be compatible with biological

activity as inducer of resistance in the host plant (2-10 µM) (Ongena et al., 2007; Jourdan et

al., 2009) but obviously, iturin and fengycin amounts by S499 are not sufficient for antifungal

activity against soilborne phytopathogens.

From a qualitative point of view, it also clearly appears that the relative proportions of

the different homologues within the surfactin and iturin families may drastically vary in

function of the growth conditions. For instance, the C15 homologue by itself represents more

than 75% of the total surfactins produced by cells colonizing tomato roots. It largely differs

from the one observed upon growth of S499 as planktonic cells in the optimized laboratory

culture medium (Figure 5). Such an in vivo LP signature is interesting since, in some

instances, homologues with the longest fatty acid chains were proven to be biologically more

active. This does not only concern their potential to trigger plant defense reactions (Jourdan et

al., 2009), but also its haemolytic, antiviral and antimicrobial activities (Kracht et al., 1999;

Bonmatin et al., 2003; Dufour et al., 2005). Such an important role for the length of the acyl

chain may be related to the fact that it readily inserts into phospholipid bilayers (Carrillo et

al., 2003; Eeman et al., 2006; Heerklotz and Seelig, 2007). Synthesis of C15 surfactins is

seemingly favoured in the presence of organics acids compared to sugars (Figure 5). Akpa et

al. (2001) have also demonstrated some influence of the nature of carbon source on the

pattern of surfactins produced by B.subtilis. The surfactin-apparented lychenisin also

accumulates differently in the culture broth upon growth in various media differing in C

source (Li et al., 2008).

In conclusion, the success of some isolates of the Bacillus genus as biopesticides

certainly rely on the large part of their genome devoted to antibiotic synthesis (Stein, 2005).

Results-Chapter 5

109

We show here that lipopeptides may be readily produced by Bacillus colonizing roots. This

reinforces the importance of these compounds for rhizosphere fitness and biocontrol activity.

However, the in planta lipopeptide signature does not reflect the broad panoply of LPs that

may be produced in vitro in an appropriate medium. It means that promising strains selected

on the basis of an efficient co-production of surfactins, fengycins and iturins with specific

roles and targets, have to be further tested for their potential to secrete the same LPs pattern

under natural rhizosphere conditions. This would provide a competitive advantage for such

Bacillus isolates with potential multi-faceted biocontrol activity.

Acknowledgements This work received financial support from the Program F.R.F.C. n° 2.4.628.10F and Crédit aux chercheurs n° 1.5.192.08F (National Funds for Scientific Research, F.R.S.-FNRS, Belgium). V. Nihorimbere is a recipient of a grant from the Belgian Technical Cooperation (BTC/CTB). M. Ongena is a Research Associate at the National Funds for Scientific Research and H. Cawoy's PhD thesis is supported by a grant from the Fonds pour la formation à la Recherche dans l'Industrie et dans l'Agriculture (F.R.I.A.).

References

Abushady H, Bashandy A, Aziz N, Ibrahim H. (2005). Molecular Characterization of Bacillus subtilis Surfactin Producing Strain and the Factors Affecting its Production. Int J Agri Biol 7: 337–344.

Akpa E, Jacques P, Wathelet B, Paquot M, Fuchs R, Budzikiewicz H, Thonart P. (2001). Influence of culture conditions on lipopeptide production by Bacillus subtilis. Appl Bioch Biotechn 93: 551–561.

Asaka O, Shoda M. (1996). Biocontrol of Rhizoctonia solani damping-off of tomato with Bacillus subtilis RB14. Appl Environ Microbiol 62: 4081–4085.

Athukorala SNP, Dilantha FWG, Rashid KY. (2009). Identification of antifungal antibiotics of Bacillus species isolated from different microhabitats using polymerase chain reaction and MALDI-TOF mass spectrometry. Can J Microbiol 55: 1021–1032.

Bais H, Fall R, Vivanco J. (2004). Biocontrol of Bacillus subtilis against infection of Arabidopsis roots by Pseudomonas syringae is facilitated by biofilm formation and surfactin production. J Plant Physiol 134: 307–319.

Bais H, Weir T, Perry LG, Gilroy S, Vivanco J. (2006). The role of root exudates in rhizosphere interactions with plants and other organisms. Annu Rev Plant Biol 57: 233–266.

Bonmatin J, Laprevote O, Peypoux F. (2003). Diversity among microbial cyclic lipopeptides: iturins and surfactins. Activity-structure relationships to design new bioactive agents. Comb Chem High T Scr 6: 541–556.

Carrillo C, Teruel J, Aranda F, Ortiz A. (2003). Molecular mechanism of membrane permeabilization by the peptide antibiotic surfactin. Biochim Biophys Acta 1611: 91–97.

Chun J, Bae K. (2000). Phylogenetic analysis of Bacillus subtilis and related taxa based on partial gyrA gene sequences. Anton Leeuw Int J G 78: 123–127.

Debois D, Hamze K, Guérineau V, Le Caër J-P, Holland IB, Lopes P, Ouazzani J, Séror SJ, Brunelle A, Laprévote O. (2008). In situ localisation and quantification of surfactins in a Bacillus subtilis swarming community by imaging mass spectrometry. Proteomics 8: 3682–3691.

Dufour S, Deleu M, Nott K, Wathelet B, Thonart P, Paquot M. (2005). Hemolytic activity of new linear surfactin analogs in relation to their physico-chemical properties. Biochim Biophys Acta– Gen Subjects 1726: 87–95.

Eeman M, Berquand A, Dufrene Y, Paquot M, Dufour S, Deleu M. (2006). Penetration of surfactin into phospholipid monolayers: nanoscale interfacial organization. Langmuir 22: 11337–11345.

Results-Chapter 5

110

Fickers P, Leclère V, Guez J-S, Béchet M, Coucheney F, Joris B, Jacques P. (2008). Temperature dependence of mycosubtilin homologue production in Bacillus subtilis ATCC6633. Res Microbiol 159: 449–457.

Finking R, Marahiel M. (2004). Biosynthesis of nonribosomal peptides1. Annu Rev Microbiol 58: 453–488.

Gancel F, Montastruc L, Liu T, Zhao L, Nikov I. (2009). Lipopeptide overproduction by cell immobilization on iron-enriched light polymer particles. Process Biochem 44: 975–978.

Glaser P, Danchin A, Kunst F, Zuber P, MM N. (1995). Identification and isolation of a gene required for nitrate assimilation and anaerobic growth of Bacillus subtilis. J Bacteriol 177: 1112–1115.

Guerra–Santos L, Kappeli O, Fiechter A. (1984). P. aeruginosa biosurfactant production in continuous culture with glucose as carbon source. Appl Environ Microbiol 48: 301–305.

Hagelin G, Indrevoll B, Hoeg-Jensen T. (2007). Use of synthetic analogues in confirmation of structure of the peptide antibiotics maltacines. Int J Mass Spectrom 268: 254–264.

Hbid C, Jacques P, Razafindralambo H, Kalounda M, Meurisse E, Paquot M, Thonart P. (1996). Influence of the production of two lipopeptides, Iturin A and Surfactin SI, on oxygen transfer during Bacillus subtilis fermentation. Appl Biochem Biotech 57/58: 571–579.

Heerklotz H, Seelig J. (2007). Leakage and lysis of lipid membranes induced by the lipopeptide surfactin. Eur Biophys J 36: 305–314.

Hinsinger P, Bengough A, Vetterlein D, Young I. (2009). Rhizosphere: biophysics, biogeochemistry and ecological relevance. Plant Soil 321: 117–152.

Hoffmann T, Frankenberg N, Marino M, Jahn D. (1988). Ammonification in Bacillus subtilis utilizing dissimilatory nitrite reductase is dependent on resDE. J Bacteriol 80: 186–189.

Højberg O, Schnider U, Winteler HV, Sørensen J, Haas D. (1999). Oxygen-sensing reporter strain of Pseudomonas fluorescens for monitoring the distribution of low-oxygen habitats in soil. Appl Environ Microbiol 65: 4085–4093.

Huang Y, Koba K, Horrobin D, Sugano M. (1993). Interrelationship between dietary protein, cholesterol, and n-6 polyunsaturated fatty acid metabolism. Prog Lipid Res 32: 123–137.

Jacques P, Hbid C, Destain J, Razafindralambo H, Paquot M, De Pauw E, Thonart P. (1999). Optimization of biosurfactant lipopeptide production from Bacillus subtilis S499 by Plackett–Burman design. Appl Biochem Biotech 77: 223–233.

Jourdan E, Henry G, Duby F, Dommes J, Barthelemy JP, Thonart P, Ongena M. (2009). Insights into the defense-related events occurring in plant cells following perception of surfactin-type lipopeptide from Bacillus subtilis. Mol Plant Microbe Interact 22: 456–468.

Kakana P (2005). Etude de la biosynthèse des lipopeptides par Bacillus subtilis et de leurs potentialités d’utilisation dans la lutte biologique au Burundi. Gembloux, Agricultural University of Gembloux. Ph.D. Thesis.

Kamilova F, Kravchenko L, Shaposhnikov A, Makarova N, Lugtenberg B. (2006). Effects of the tomato pathogen Fusarium oxysporum f. sp. radicis-lycopersici and of the biocontrol bacterium Pseudomonas fluorescens WCS365 on the composition of organic acids and sugars in tomato root exudate. Mol Plant-Microbe Interact 9: 1121–1126.

Kaneda T. (1991). Iso- and anteiso-fatty acids in bacteria: biosynthesis, function, and taxonomic significance. Microbiol Rev 55: 288-302.

Kim H, Byung D, Chang H, Hyun H, Hee M, Tohoru K, Yoshiki T. (1997). Production and properties of a lipopeptide biosurfactant from B. subtilis C9. J Ferment Bioengin 84: 41–46.

Kinsella K, Schulthess CP, Morris TF, Stuart JD. (2009). Rapid quantification of Bacillus subtilis antibiotics in the rhizosphere. Soil Biol Biochem 41: 374–379.

Kracht M, Rokos H, Ozel M, Kowall M, Pauli G, Vater J. (1999). Antiviral and hemolytic activities of surfactin isoforms and their methyl ester derivatives. J Antibiot 52: 613–619.

Lee K, Jun K, Kim W, Paik H. (2001). Partial characterization of polyfermenticin SCD, a newly identified bacteriocin of Bacillus polyfermenticus. Lett Appl Microbiol 32: 146–151.

Lee S, Kim S, Park I, Chung S, Choi Y. (2007). Isolation and structural analysis of bamylocin A, novel lipopeptide from Bacillus amyloliquefaciens LP03 having antagonistic and crude oil-emulsifying activity. Arch Microbiol 188: 307–312.

Results-Chapter 5

111

Leenders F, Stein TH, Kablitz B, Franke P, Vater J. (1999). Rapid typing of Bacillus subtilis strains by their secondary metabolites using matrix-assisted laser desorption/ionization mass spectrometry of intact cells. Rapid Commun Mass Spectrom 13: 943–949.

Li YM, Haddad NIA, Yang SZ, Mu BZ. (2008). Variants of lipopeptides produced by Bacillus licheniformis HSN221 in different medium components evaluated by a rapid method ESI-MS. Int J Pept Res Ther 14: 229–235.

Lugtenberg BJJ, Dekkers L, Bloemberg GV. (2001). Molecular determinants of rhizosphere colonization by Pseudomonas. Annu Rev Phytopathol 39: 461–490.

Mach H, Hecker M, Mach F. (1988). Physiological studies on cAMP synthesis in Bacillus subtilis. FEMS Microbiol Lett 52: 89–192.

Mulligan CN. (2005). Environmental applications for biosurfactants. Environ Pollut 133: 183–198. Nakano M, Marahiel M, Zuber P. (1988). Identification of a genetic locus required for biosynthesis of

the lipopeptide antibiotic surfactin in Bacillus subtilis. J Bacteriol 170: 5662–5668. Niemann S, Pühler A, Tichy HV, Simon R, Selbitschka W. (1997). Evaluation of the resolving power

of three different DNA fingerprinting methods to discriminate among isolates of a natural Rhizobium meliloti population. J Appl Microbiol 82: 477–484.

Nihorimbere V, Fickers P, Thonart P, Ongena M. (2009). Ecological fitness of Bacillus subtilis BGS3 regarding production of the surfactin lipopeptides in the rhizosphere. Environ Microbiol Report 1: 134–130.

Ohno A, Ano T, Shoda M. (1995). Effect of temperature on production of lipopeptide antibiotics, iturin A and surfactin by a dual producer, Bacillus subtilis RB14, in solid-state fermentation. J Ferment Bioeng 80: 517–519

Ongena M, Jacques P. (2008). Bacillus lipopeptides: versatile weapons for plant disease biocontrol. Trends Microbiol 16: 115–125.

Ongena M, Jourdan E, Adam A, Paquot M, Brans A, Joris B, Arpigny J-L, Thonart P. (2007). Surfactin and fengycin lipopeptides of Bacillus subtilis as elicitors of induced systemic resistance in plants. Environ Microbiol 9(4): 1084–1090.

Phae CG, Shoda M, Kubota H. (1990). Suppressive effect of Bacillus subtilis and its products on phytopathogenic microorganisms. J Ferment Bioeng 69: l-7.

Pirri G, Giuliani A, Nicoletto S, Pizzuto L, Rinaldi A. (2009). Lipopeptides as anti-infectives: a practical perspective. Cent Eur J Biol 4: 258–273.

Price NPJ, Rooney AP, Swezey JM, Perry E, Cohan FM. (2007). Mass spectrometric analysis of lipopeptides from Bacillus strains isolated from diverse geographical locations. FEMS Microbiol Lett 271: 83–89.

Raaijmakers J, de Bruijn I, de Kock M. (2006). Cyclic lipopeptide production by plant-associated Pseudomonas spp. Diversity, activity, biosynthesis, and regulation. Mol Plant Microbe Interact 19: 699–710.

Raaijmakers J, De Bruijn I, Nybroe O, Ongena M. (2010). Natural functions of lipopeptides from Bacillus and Pseudomonas: more than surfactants and antibiotics. FEMS Microbiol Rev 34: 1037–1062.

Ramey B, Koutsoudis M, von Bodman S, Fuqua C. (2004). Biofilm formation in plant–microbe associations. Curr Opin Microbiol 7: 602–609.

Ramos H, Hoffmann T, Marino M, Nedjari H, Presecan-Siedel E, Dreesen O, Glaser P. (2000). Fermentative Metabolism of Bacillus subtilis: Physiology and Regulation of Gene Expression. J Bacteriol 182: 3072–3080.

Romero D, de Vicente A, Rakotoaly RH, Dufour SE, Veening J-W, Arrebola E, Cazorla FM, Kuipers OP, Paquot M, Pérez-García A. (2007). The Iturin and Fengycin Families of Lipopeptides Are Key Factors in Antagonism of Bacillus subtilis Toward Podosphaera fusca. Mol Plant Microbe Interact 20: 430–440.

Ron EZ, Rosenberg E. (2001). Natural roles of biosurfactants. Environ Microbiol 3: 229–236. Stein T. (2005). Bacillus subtilis antibiotics: structures, synhesis and specifics functions. Mol

Microbiol 56: 845–847. Stein T. (2008). Whole-cell matrix-assisted laser desorption/ionization mass spectrometry for rapid

identification of bacteriocin/lantibiotic-producing bacteria. Rapid Commun Mass Sp 22: 1146–1152.

Results-Chapter 5

112

Touré Y, Ongena M, Jacques P, Guiro A, Thonart P. (2004). Role of lipopeptides produced by Bacillus subtilis GA1 in the reduction of grey mould disease caused by Botrytis cinerea on apple. J Appl Microbiol 96: 1151–1160.

Vater J, Kablitz B, Wilde C, Franke P, Mehta N, Cameotra SS. (2002). Matrix-assisted laser desorption ionization-time of flight mass spectrometry of lipopeptide biosurfactant in whole cells and culture filtrates of Bacillus subtilis C-1 isolated from petroleum sludge. Appl Environ Microb 68: 6210–6219.

Vater J, Wilde C, Kell H. (2009). In situ detection of the intermediates in the biosynthesis of surfactin, a lipoheptapeptide from Bacillus subtilis OKB 105, by whole-cell cell matrix-assisted laser desorption/ionization time-of-flight mass spectrometry in combination with mutant analysis. Rapid Commun Mass Sp 23: 1493–1498.

Yeh MS, Wei YH, Chang JS. (2006). Bioreactor Design For Enhanced Carrier-Assisted Surfactin Production With Bacillus Subtilis. Process Biochem 41: 1799–1805.

Results-Chapter 5

113

Supplementary data

Figure 1S

Figure 2S

Results-Chapter 5

114

Surfactin

1 mg/mL

Iturine

1 mg/mL

Fengycin

1 mg/mL

Figure 3S

Results-Chapter 6

115

Foreword on Chapter 6

In Chapter 5, we have studied the pattern of production of the three families of lipopeptides

by strain S499 of B. amyloliquefaciens cultured in situ on tomato plants in greenhouse and in

vitro versus different environmental parameters. This strain showed to produce the three

families of lipopeptides but in very different relative amounts; and, surfactins were by far

more efficiently secreted compared to iturins and fengycins in the conditions assayed.

Moreover, surfactins are probably the Bacillus LPs displaying the largest eventail of

biocontrol-related activities. It is why we demonstrated sensu stricto its potential factories

involved in biological activities. We wanted to further study in situ production of this

lipopeptide by Bacillus by coupling direct detection of the compound and genetic expression

(Chapter 6). To that end, we had to use another Bacillus strain BGS3 that contains a reporter

gene system LacZ which helps to evaluate the surfactin gene expression. BGS3 strain was

constructed by integrating into the strain BC25 of Bacillus amyloliquefaciens the promoter of

surfactins (psrfA) obtained from B. amyloliquefaciens S499 genomic DNA.

Results-Chapter 6

116

Chapter 6. Ecological fitness of Bacillus subtilis BGS3 regarding production of the surfactin lipopeptide in the

rhizosphere

Published in Environmental Microbiology Report (2009) 1(2), 124-130.

Nihorimbere Venanta, Fickers Patrickb, Thonart Philippea, and Ongena Marca

a Centre Wallon de Biologie Industrielle, Unité de Bioindustries, Gembloux University of

Agricultural Sciences, B-5030 Gembloux. b Centre d'Ingénierie des Protéines, University of Liège, B-4000 Liège, Belgium.

Running title: Bacillus lipopeptide synthesis in the rhizosphere

The first two authors contributed equally to the paper.

Results-Chapter 6

117

Abstract

Cyclic lipopeptides and particularly surfactins produced by Bacillus species retain

antibacterial, antiviral, biofilm-forming and plant resistance-inducing activities. In most cases,

their role in biological control of plant diseases was evoked on the basis of in vitro assays or

by using non-producing/ overproducing mutants but there is a need for more direct evidence

of an efficient lipopeptide biosynthesis in the rhizosphere. In this work, we coupled LC-MS

quantification of the lipopeptides secreted by cells colonizing tomato plants with the use of

psrfA–lacZ reporter system integrated within the BGS3 chromosome to study the expression

of the surfactin operon in planta. Results showed that a higher level of psrfA induction was

observed upon the establishment of a stable BGS3 population on roots and surfactins

extracted from the rhizosphere were produced in biologically significant quantities. Our

results also demonstrate that BGS3 efficiently utilizes the main substrates from plant exudates

to produce surfactins. This synthesis is also efficient in cells forming colonies and the

production may be favoured in bacteria developing slowly in the rhizosphere. This provides a

first understanding of how environmental factors may influence lipopeptide production by

beneficial Bacillus strains.

Introduction

Among the panoply of antimicrobial compounds that can be produced by Bacillus species,

cyclic lipopeptides of the surfactin, iturin and fengycin families retain potential

biotechnological applications (Mulligan, 2005; Mukherjee et al., 2006) but are also tightly

involved in most of the mechanisms described to date that explain the beneficial effect of

specific strains to control plant diseases (Bais et al., 2004; Ongena and Jacques, 2008). They

can facilitate root colonization, act as antagonists by directly inhibiting phytopathogens and

reinforce the resistance potential of the host plant via the stimulation of the so-called induced

systemic resistance (ISR) phenomenon. This broad range of activities is due to the physico-

chemical properties of these non-ribosomally synthesized amphiphilic molecules. The srf

operon encodes genes involved in surfactin biosynthesis. Members of this group are

heptapeptides interlinked with a b-hydroxy fatty acid to form a cyclic lactone ring structure.

The length of this fatty acid chain may vary from C12 to C15 and different homologous

compounds are thus usually co-produced (Peypoux et al., 1999). Plant roots exude a large

range of potentially valuable small-molecular-weight compounds into the rhizosphere that

creates a highly dynamic front for interactions between roots and soil microflora (Bais et al.,

2006). Some of these exuded molecules act as chemical signals for motile bacteria to move to

Results-Chapter 6

118

the root surface but also represent the main nutrient sources available to support growth and

persistence of rhizobacteria (Hirsch et al., 2003). The host plant thus imposes a specific

nutritional context but the complexity of microbial communities sharing the ecosystem and

the variability of physico-chemical parameters inherent to the soil such as mineral content,

pH, temperature and oxygen availability may also drastically influence rhizobacterial growth

and production of biocontrol determinants as demonstrated with various pseudomonads

(Duffy and Défago, 1999; Van Rij et al., 2004; Dubern and Bloemberg 2006; Ongena et al.,

2008). Understanding which and how environmental factors may modulate the production of

biocontrol determinants is a key to improve the level and reliability of plant disease reduction

through the use of rhizobacteria. However, in the case of Bacillus antibiotics and particularly

lipopeptides, most of the data suggesting their role in biocontrol derive from antagonism

assays performed in vitro or by using non-producing or overproducing derivatives and

correlation with respective loss or increase in the biocontrol activity (Ongena and Jacques,

2008). Comparatively, studies providing direct evidence through the demonstration of an

efficient lipopeptide production in situ are scarce (Asaka and Shoda, 1996; Touré et al., 2004;

Romero et al., 2007). In this context, the study reported here was conducted with the first

objective to demonstrate efficient expression of biosynthetic genes and surfactin secretion in

the rhizosphere and second to evaluate the influence of some rhizosphere-specific parameters

on lipopeptide synthesis. To this end, we used the LacZ gene reporter system in Bacillus

subtilis BGS3, not only to provide information on the transcriptional activity of specific

antibiotics in situ but also to easily monitor populations of the introduced strain among the

whole root-associated microflora. The BGS3 derivative was generated from the surfactin-

overproducing isolate BC25 previously demonstrated to provide a protective effect on tomato

via the stimulation of systemic resistance in the host plant (Ongena et al., 2007).

Results and discussion

Colonization and surfactin production in the rhizosphere

Expression of biosynthetic genes and surfactin secretion by the BGS3 derivative was first

verified during growth in an optimized medium in batch-type bioreactor with control of

fermentation parameters. Surfactins production clearly occurred within the 5-6 next hours and

reached a maximum after 18 h of cultivation corresponding to the transition from exponential

to stationary growth phase which is in agreement with results from other studies (Figure S1 in

Supporting information) (Lin et al., 1998; Koumoutsi et al., 2004). Biomass and surfactins

concentrations were similar to the parental BC25 strain showing that the genetic

Results-Chapter 6

119

transformation did not significantly affect the basal metabolic functions or those related to the

synthesis of surfactins (not shown).

0

1

2

3

4

5

6

0

5

10

15

20

25

0 5 10 15 20 25 30

Relative ββ ββ-gal activity U

/105

cells)

Cel

ls x

10

6/g

of r

oot

Days post-inoculation

colonization BGS3 relative srfA gene expression

Figure 1. Evolution of population and expression of surfactin operon in B.subtilis BGS3 cells colonizing tomato roots. Bacillus subtilis BGS3 was constructed as follow: psrfA promoter was PCR amplified from B. subtilis S499 genomic DNA using primer MO01b (5′-CCT CAT GCC TAT TCT TGA AGC CAT GTA TG-3′) and MO02b (5′-GGA TCC TAT TTC CAT ATT GTC ATA CCT CCC CTA ATC-3′) and cloned into pGEM-T Easy vector (Promega, Madisson, WI) to generate pMO01. After correct sequence verification with universal primer T7 and SP6, psrfA promoter was rescued from pMO01 by BamHI–EcoRI digestion and cloned into pDG1663 (Guérout-fleury et al., 1996) at the corresponding site to yield pMO02. This final construct was then integrated into B. subtilis BC25 (Ongena et al., 2007) at the thrc locus to give rise to strain BGS3. Ectopic integration was verified by threonine auxothrophy on minimal medium, resistance to erythromycine and sensibility to spectinomycin as described elsewhere (Guérout-fleury et al., 1996). Surface-sterilized tomato seeds were germinated for 5 days at room temperature in the dark and in gelified sterile nutrient medium consisting of 5 mM Ca(NO3)2, 5 mM KNO3, 2 mM MgSO4, 1 mM KH2PO4 and micronutrients. They were transferred in sterilized 50 ml tubes filled with nutrient solution, inoculated with strain BGS3 and incubated at 25 ± 2 °C in the greenhouse with a 16 h photoperiod alternating sunlight and fluorescent light. For inoculum preparation, the bacterium was grown on solidified Luria–Bertani (LB) medium for 24 h and harvested cells were re-suspended in NaCl 0.85% and the concentration was determined by measuring turbidity at 600 nm. BGS3 cells were inoculated at a final concentration of 108 cells ml -1 nutrient solution. Root colonization by the BGS3 strain was determined by plate counts on LB medium on the basis of typical morphology of the colonies and blue coloration upon growth on X-Gal-supplemented medium. The use of both criteria was necessary to reliably discriminate BGS3 cells from the contaminating bacterial microflora. Three plants were used to obtain values at each time point and data represent mean values from two experiments. In situ surfactin gene expression was evaluated on samples of approximately 0.2 g of root material that was suspended in buffer Z containing lysozyme for lysis of root-adhering cells and further incubated for 60 min at 40 °C at pH 7 in the presence of β-gal substrate. srfA expression was determined by A 420 after 60 min of incubation at 40 °C. For determination of both colonization and srfA promoter gene expression (on the same plants), root samples from three different plants were used at each time point in one assay and data mentioned in the figure represent mean values calculated from two independent experiments that yielded similar results. Relative β-galactosidase activity was obtained by subtracting values obtained for untreated control plants from those measured for BGS3-inoculated tomatoes.

In vivo experiments were conducted on tomato plants grown in a hydroponic culture

mode to facilitate colonization study and quantification of surfactins secreted into the

surrounding environment. The colonization process observed for BGS3 was quite similar to

Results-Chapter 6

120

the one reported for other plant beneficial rhizobacteria (Figure 1) (Rainey, 1999; Espinosa-

Urgel et al., 2002) with a first step of bacterial cell attraction/adhesion followed by an

efficient colonization phase and by a further decrease of the population to reach a stable level

at approx. 5 x 106 cell/g roots (Figure 1). Such steady-state phase most probably corresponds

to a resident phase where the population size is restricted by space and/or nutrients

availability and is thus limited by plant growth and root exudation rate.

0

2

4

6

8

10

12

14

16

18

20

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

batch 0,1 0,21 0,33

ββ ββ-gal activity (U)C

ells

x 1

08/m

l

biomass srfA gene expression

µ= µ= µ=

Figure 2. Influence of growth rate of BGS3 cells on expression of surfactin synthetase gene (srfA) and biomass production. Chemostat cultures were performed in a 2 l laboratory glass bioreactor with a working volume of 1.5 l and equipped with automated pH control, a dissolved oxygen tension electrode and a magnetic coupled stirrer (BiostatB, B. Braun Biotech. International, Melsungen, Germany). The temperature was maintained at 30 °C and pH was regulated at 7 by addition of H3PO4 (3N). Aeration rate and agitation were fixed respectively at 0.5 V.V.M. (volume of air per volume of medium per minute) and 600 r.p.m.. By contrast with batch culture where the exhaustion of a specific nutrient terminates the exponential growth phase, biomass concentration in the chemostat culture is usually controlled by the permanent limitation of a single defined nutrient (peptone as source of threonine in this case). It allows the study of metabolite synthesis upon fixed growth rate and constant cellular physiological state. Fresh medium (containing per liter bacto-peptone 1 g, (NH4)2SO4 1 g, sucrose 10g, KH2PO4 1.9g, CuSO4 0.001 mg, FeCl3 0.005 mg, NaMoO4 0.004 mg, MnSO4 3.6 mg, KI 0.002 mg, MgSO4 3.6 mg, ZnSO4 0.014 mg, H3BO3 0.01 mg, citric acid 10 mg) was fed into the reactor using a peristaltic pump to control the dilution rate. The volume of the culture was maintained constant by the use of an overflow device. Cultures were inoculated with a 2% volume of 16 h-subcultures realized in the same medium and the bioreactor was switched to continuous mode after an initial batch start-up. Silicone antifoam was added to control extensive foam formation. The different growth rates (µ) of 0.1 h-1, 0.21 h-1, 0.33 h-1, were respectively considered as low, intermediate and relatively high in this medium on the basis of the growth kinetic observed in batch culture in the same conditions. Samples were harvested at steady state for cell density determination (OD600 nm) and β-galactosidase activity measurement on cell extract prepared by lysozyme treatment and centrifugation (Fickers et al., 2008). One unit of β-galactosidase activity was defined as the amount of enzyme that produces 1 nmol of o-nitrophenol min-1 at 37 °C per OD 600 unit. Data are average values (± standard error) obtained from surfactin concentrations in three independent cultures performed for the study of one specific µ. Surfactin concentration in every independent culture was calculated on the basis of three measurements at 1 h intervals during steady state.

The level of srfA gene expression in root-colonizing BGS3 cells remained almost

stable during the first four days after inoculation, but sharply increased from day 10 to reach a

maximum at day 21 (Figure 1). A high and constant relative expression of srfA gene per cell

unit in the rhizosphere was then observed until the end of the experiment corresponding to the

period where BGS3 cells maintained a stable population. The development of established

Results-Chapter 6

121

bacterial populations in the rhizosphere is actually restricted by nutrient availability owing to

limited root exudation rate (Lugtenberg et al., 2001; Bais et al., 2006). Bacillus cells on roots

can thus be in a nutrient-starved state that would drive cell physiology to slow growth rate

that is apparently conducive for surfactin synthesis. Such effect of growth rate on surfactin

synthesis was confirmed by experiments in chemostat bioreactors that allows the study of

metabolite synthesis upon fixed growth rate (µ) and constant cellular physiological state.

Biomass levels obtained at steady state in the different conditions of growth rates were similar

at approx. 2.5 x 108 cells/ml (Figure 2). In contrast, srfA expression significantly decreased

with increasing µ and the residual β-galactosidase activity measured at µ=0.33 h-1 represented

only 6% of the activity observed at µ=0.1 h-1 (Figure 2). A similar relationship between µ and

srfA expression was observed with another B. subtilis isolate derived from the surfactin over-

producing JH642 strain (not shown).

In parallel to the gene expression determination, a reliable recovery and quantification

of lipopeptides secreted in the microenvironment by the root-colonizing strain is essential. To

this end, we had to combine material extracted from the nutrient solution with the one

recovered by repeated extraction of root tissues [1 g of samples extracted with 4 ml of

MeOH/H2O/Triton X-100, 50/50/0.9 (v/v/v) solution]. This could be explained by the fact

that secreted surfactins are rapidly embedded in the membrane structure of the target

organism as suggested by the rapid disappearance of the lipopeptides from the nutrient

solution when supplemented in hydroponically grown tomatoes (not shown). In two different

experiments, rhizosphere samples were collected after a period of 20 days post inoculation

and LC-MS analyses (see footnote of Table 1 for details) of their content in surfactin revealed

a mean value of 320 ± 80 µg/108 cells. In our conditions, considering an average biomass of

7.5x106 cell/g of root (Figure 1) and 2 g of root tissue per plant, it corresponds to approx. 1.8

µM in the nutrient solution. Importantly, such a surfactin concentration is in the range of the

one necessary to induce ISR on the same plant (5 µM) (Ongena et al., 2007). It indicates that

the lipopeptide may be produced in biologically significant quantities in the rhizosphere of the

host plant.

Results-Chapter 6

122

Surfactin biosynthesis in tomato exudates

We further tested srfA expression and surfactins production upon growth of strain BGS3 in

the root-exudates collected from tomato plantlets after 23-25 days of growth in nutrient

solution under sterile conditions. Freshly collected exudates supported growth of strain BGS3

(Table 1). The srfA gene expression and lipopeptides secretion into the medium were also

effective under these conditions (Table 1), confirming that BGS3 cells can efficiently use

exsudate components for surfactin biosynthesis. This was also demonstrated by cultivating

BGS3 in a minimal medium containing as sole carbon sources the sugars, organic acids and

amino acids typically found in the tomato exudates and added in the proportions established

by Kamilova and colleagues (2006) (Table 1). Increased levels of biomass and surfactin

synthesis compared to natural exudates can be explained by higher concentrations of nutrients

in the medium.

Table 1. Surfactin biosynthesis upon growth of strain BGS3 in tomato root exudatesa.

Biomass

(cells x 108/ml) srfA expressionb

(β-gal U) Surfactin prod.c (µg/108 cells)

Natural exudatesd, liquide 0.34 ± 0.09 17 ± 3 14 ± 4

Recomposed exudatesf, liquid 1.40 ± 0.07 25 ± 8 34 ± 9

Natural exudates, solide 0.37 ± 0.05 14 ± 4 4 ± 2

Recomposed exudates, solid 1.28 ± 0.09 18 ± 7 5 ± 1 a Cultures were realized in 150 ml agitated erlenmeyer flasks (130 r.p.m., 28 °C) filled with 50 ml of medium. In all cases, data are mean values and standard error from four independent experiments. b See legend of Figure 2 for methods of quantification. Cell densities are those corresponding to the maximal OD measured 48 h after incubation. Values for gene expression also correspond to the maximal β-gal activities measured during the culture, usually 24 h after inoculation. c Supernatant samples were collected after 48 h of growth, loaded on C18 solid-phase extraction cartridges (900 mg, Alltech) and surfactins were desorbed with 100% ACN. The resulting samples were analyzed by reverse phase HPLC coupled with single quad mass spectrometer (HPLC Waters Alliance 2695/ diode array detector, coupled with waters SQD mass analyser) on a X-terra MS (Waters) 150*2.1 mm, 3.5 mm column. Surfactins were eluted in the isochratic mode (78% acetonitrile in water acidified with 0.1% formic acid) at a constant flow rate of 0.3 ml min -1 and 40 °C. Compounds were first identified on the basis of their retention times compared with purified standards and the amounts were calculated on the basis of the corresponding peak area (max plot). The identity of each homologue was confirmed on the basis of the masses detected in the SQD by setting electrospray ionization (positive ion) conditions in the MS as source temperature, 130 °C; desolvation temperature, 250 °C; nitrogen flow, 500 l h-1; cone voltage, 70 V. d Filter-sterilized tomato exudates freshly collected from non-inoculated plants grown for 23 days in the greenhouse (see legend of Figure 1). These natural exudates were supplemented with 2 g l-1 (NH4)2SO4 to compensate for the lack of available nitrogen. e Production by BGS3 upon growth as planktonic cells in liquid cultures or as colonies on the same but solidified media (bactoagar at 1.5%). f Medium composed of exudates artificially recomposed [containing 10 g l-1 glucose equivalent in (i) sugars 5 g l-1: glucose 34%, fructose 57%, maltose 8%, ribose 0.75%; (ii) organic acids 4.5 g l-1: citrate 77%, succinate 19%, malate 2%, fumarate 0.5%; (iii) amino acids 0.5 g l-1: glutamate 23%, aspartate 18%, leucine 12%, isoleucine 10%, lysine 9%, (serine + aspartame) 5.5%, (arginine + threonine) 5.5%, (glutamate + glycine) 5.5%, histidine 5.5%, phenylalanine 5.5%] on the basis of data from Kamilova and colleagues (2006).

Results-Chapter 6

123

When tested individually, the main representatives of typical root exudates supported

growth to an almost similar (data not shown) but srfA gene expression and surfactin secretion

were significantly higher in the presence of organic acids and amino acids as unique C

sources (Figure 3A).Aspartate and glutamate are the major amino acid residues found in

tomato exudates but quantitatively, they represent minor components. Although it may vary in

function of various plant factors, succinic and citric acids are the more abundant substrates in

tomato roots and also represent adequate nutrient sources utilizable by BGS3 for surfactin

production.

Rhizosphere-colonizing bacteria typically aggregate in microcolonies on the root

surface (Ramey et al., 2004) and their actual physiology is probably not related to a

planktonic state. We therefore performed BGS3 cultures in 12-well microplates on the various

media described above but gelified by adding 1.5% bactoagar in order to test for lipopeptide

synthesis by immobilized cells. Growth on X-Gal-supplemented media containing either

natural or recomposed exudates resulted in both cases in the formation of blue colonies with

b-galactosidase-positive phenotype (not shown). This first visual assessment was supported

by measurements of the gene expression within the cells scrapped off the wells. The values

obtained for surfactin gene expression are in the range of those measured in liquid cultures in

the presence of both kinds of exudates (Table 1). In contrast, we observed reduced surfactin

production rates. The secreted amounts could be underestimated due to diffusion limitation of

the lipopeptide throughout the biofilm into the environment or because of a less efficient

recovery from the gelified matrix than from the liquid medium. On the other hand, differential

surfactin biosynthesis observed following growth on solid media with individual carbon

sources correlates well with those from liquid cultures with higher levels of gene expression

in the presence of organic and amino acids than the three sugars tested (Figure 3B). Globally,

these results demonstrate that surfactin production is also readily effective in BGS3 cells

developing on solid surface according to the fact that it actually occurs in biofilm-related

structures formed on roots (Bais et al., 2004; Ramey et al., 2004).

Results-Chapter 6

124

0

1

2

3

4

5

6

7

0

5

10

15

20

25

30

Glu Fru Malt Succ Cit Asp Glu ExR

Surfactin µg/10

8cellsββ ββ-

gal a

ctiv

ity (

U)

0

10

20

30

40

50

60

70

0

5

10

15

20

25

Glu Fru Mal Suc Cit Asp Glu ExR

Surfactin µg/10

8cellsββ ββ-

gal a

ctiv

ity (

U)

A

B

0

20

40

60

80

100

Glu Fru Mal Succ Cit Asp Glu ExR

Rel

ativ

e pr

opor

tion

(%)

C12 C13 C14 C15C

Glu Fru Malt Suc Cit Asp Glut ExR



Figure 3. Surfactin synthesis in the presence of various carbon sources typically found in tomato exudates. Lipopeptide gene expression (black columns) and production (white columns) were determined both in liquid cultures (A) in flasks under the conditions described in Table 1 and on the same gelified media (B). The proportions of the various homologues differing in the length of the fatty acid chain from C12 to C15 and produced in the liquid cultures are shown in (C). The substrates (glucose 10 g l -1, fructose 10 g l -1, maltose 5 g l -1, succinate 15 g l -1, citrate 10 g l -1, aspartame 10 g l -1, glutamate 15 g l -1, recomposed exudates containing 10 g l -1 glucose equivalent, for details see footnote of Table 1) were tested by adding a concentrated solution to a minimal medium composed by MgSO4 7H2O 0.5 g l -1, K 2 HPO 4 1.0 g l -1, KCl 0.5 g l -1, Yeast extract 1.0 g l -1, Fe2(SO4)3 1.2 mg l -1, MnSO4 0.4 mg l -1, CuSO4 1.6 mg l -1, (NH4)2SO4 1 g l -1. In all cases, pH was adjusted to 7 ± 0.2 with NaOH or HCl 0.1 N before sterilization. Cultures on solid media were performed in 12-well

Results-Chapter 6

125

microplates in a volume of 1.5 ml per well. Three wells per plate were used for each carbon source. Bacterial growth in the presence of the various substrates was monitored 24 and 48 h after inoculation by measuring turbidity (OD) at 600 nm after scrapping cells for the surface of the medium and re-suspension to homogeneity. The same cells were used for β-gal activity determination. Surfactin produced in each well was extracted from the gelified medium by adding the same volume of HCl 4 N (1.5 ml) and heating for 10 min at 80 °C. The resulting solution was diluted to 10 ml with distilled water and extracted on C18 cartridge to yield semi-purified surfactins in pure methanol (1 ml) used for LC-MS analysis (see footnote of Table 1). In every experiment and for each specific substrate, surfactin concentration and β-gal activity were determined in three replicates (wells) in the same plate and first means were calculated from these data. Results presented in the figure are average values (± standard error) calculated from those first means obtained in three independent experiments.

On another hand, qualitative analyses of lipopeptide patterns by LC-MS revealed that

the C14 and C15 homologues together represent more than 80% of the total surfactins produced

in most media. It also clearly appeared that the relative proportions of the different

homologues may vary in function of the C source and the synthesis of C15 surfactins is

seemingly favoured in the presence of organic acids and amino acids compared to the sugar

group (Figure 3C). The C15 homologue is also by far (58%) the main form produced by BGS3

in natural exudates. This is of biological relevance since surfactins with the longest fatty acid

chains appear to be the more active, not only for antimicrobial activity (Kracht et al., 1999)

but also for their potential to trigger plant defense reactions (Jourdan et al., 2009). It partly

relies on the fact that a more hydrophobic aliphatic chain in the molecule induces greater

surface active properties and membrane destabilization/perturbation (Eeman et al, 2006).

These results suggest that the nutritional status encountered in the rhizosphere could also

somewhat influence qualitatively the production of surfactin homologues by B. subtilis.

Influence of immobilization*3

These results with planktonic cells provide usefull indications but have to be considered in the

context of soil where rhizobacteria aggregate in microcolonies on the root surface. We

therefore performed BGS3 cultures in twelve-well microplates on the various media described

above but gelified by adding 1.5% bactoagar in order to test for lipopeptide synthesis by

immobilized cells. Growth on X-Gal-supplemented media containing either natural or

recomposed exudates resulted in both cases in the formation of blue colonies with β-gal

positive phenotype (not shown). This first visual assessment was supported by measurements

of the gene expression within the cells scrapped-off the wells. The values obtained for

surfactin gene expression are in the range of those measured in liquid cultures in the presence

of both kinds of exudates (Table 1). By contrast, we observed reduced surfactin production

rates suggesting some unidentified regulatory process at the post-transcriptional level. The

secreted amounts could also be underestimated due to diffusion limitation of the lipopeptide

3 Unpublished results

Results-Chapter 6

126

throughout the biofilm into the environment or because of a less efficient recovery from the

gelified matrix than from the liquid medium. On the other hand, differential surfactin

biosynthesis observed following growth on solid media with individual carbon sources

correlate well with those from liquid cultures with higher levels of gene expression in the

presence of organic and amino acids than the three sugars tested (Figure 3B). Globally, these

results demonstrate that surfactin production is also readily effective in BGS3 cells

developing in macrocolonies on solid support thereby confirming that it actually occurs in

biofilm-related structures formed on roots (Bais et al., 2004).

Influence of aeration*4

Oxygen availability is among the physico-chemical parameters inherent to the soil that can

influence the fitness of PGPRs notably regarding the production of biocontrol metabolites in

the rhizosphere. The effect of oxygen concentration in the growth medium on the synthesis of

surfactin was investigated by comparing srfA gene expression and lipopeptide production by

strain BGS3 upon grown in different aerated conditions. Results presented in Table 1 show

that the strain readily adapts its respiratory metabolism to very low oxygen content of the

medium and the final biomass was not drastically affected upon growth in sealed and nitrogen

flushed flask compared with aerated cultures (Table 2). BGS3 cells may thus use nitrate as an

alternative electron acceptor like source of nutrients to grow (Ghiglione et al., 2000).

Interestingly, srfA expression and surfactin production are lower but still remain very

effective in such oxygen-starved culture conditions suggesting that the low oxygen

concentrations in the rhizosphere is not detrimental to lipopeptide synthesis by BGS3 cells

evolving on roots. This observation is relevant in the context of rhizosphere where decreasing

oxygen concentration gradient occurs in poorly aerated soils due to root and microbial

respiration (Højberg et al., 1999).

4 Unpublished results

Results-Chapter 6

127

Table 2. Effect of aeration on surfactin production by strain BGS3 in an optimized medium (Jacques et al.

1999).

Biomass

(cells x 108/ml) srfA expression (β-gal /108 cells)

Surfactin prod. (µg/108 cells)

Bioreactora 5.1 44 550

Aerated flaskb 5.6 ± 0.3 31 ± 9 660 ± 70

Anaerobic flaskc 3.7 ± 0.5 18 ± 4 480 ± 40 a Batch cultures in 2-liter bioreactor with aeration rate and agitation fixed respectively at 1 VVM and 300 rpm. b Erlenmeyers (150 ml filled with 50 ml of medium) with cotton wool that normally allows oxygen exchange. The second flask was simply closed without pushing air out. This allows the culture to start in the presence of oxygen that will be fast limited by the lack of oxygen exchange. This was named sealed flask. c Anaerobic culture conditions created in flasks closed hermetically and medium flushed with nitrogen to eliminate oxygen. Data are from three independent replicate cultures.

Article-supporting information : Gene expression and production of surfactin in artificial medium

Figure S1. Gene expression and production of surfactin in artificial medium. Time-course profiles of cell growth, dissolved oxygen level, srfA expression and surfactin production during batch fermentation of Bacillus subtilis BGS3 under an aeration rate of 1.0 V.V.M. and an agitation rate of 300 r.p.m.

Results-Chapter 6

128

Conclusion

Previous in vitro experiments have provided some insights on the effect of various sugars and

nitrogen substrates and of some abiotic conditions such as temperature and pH on lipopeptide

synthesis by Bacillus strains (Peypoux et al., 1999; Kim, et al., 2004; Guez et al., 2008).

However, the study of antimicrobial compound synthesis in the rhizosphere is more

challenging. To our knowledge, this work provides a first demonstration of efficient

lipopeptide gene expression in the rhizosphere during Bacillus-plant interaction that is

associated with relevant surfactin production in the context of biocontrol both quantitatively

and qualitatively. Our data show that important rhizosphere-specific factors such as particular

nutritional status, slow growth rate and development in aggregated and immobilized

structures are conducive for surfactin synthesis. In other experiments, we also observed that

srfA expression and surfactin production still remain very effective in oxygen-starved culture

conditions (V. Nihorimbere, unpublished) suggesting that the low oxygen status prevailing in

the rhizosphere (Højberg et al., 1999) is not detrimental to lipopeptide synthesis by BGS3 cells

evolving on roots. Reduced O2 availability and slow growth were also reported to stimulate

the production of the cyclic lipopeptide putisolvin by Pseudomonas putida (Dubern and

Bloemberg, 2006). Although the influence of many other factors deserves further

investigation and other Bacillus strains have to be studied, these results contribute to fill the

gap in the global knowledge of Bacillus fitness in the rhizosphere regarding antibiotic

synthesis, a crucial point for optimizing biocontrol strate gies using this organism.

Acknowledgments This work received financial support from the Program F.R.F.C. No. 2.4624.06 and Crédit aux chercheurs No.1.5.192.08F (National Funds for Scientific Research, F.R.S.-FNRS, Belgium). V. Nihorimbere is recipient of a grant from the Belgian Technical Cooperation (BTC/CTB). M. Ongena and P. Fickers are respectively research associate and post-doctoral researcher at the F.R.S.-FNRS.

Results-Chapter 6

129

References

Asaka, O., and Shoda, M. (1996) Biocontrol of Rhizoctonia solani damping-off of tomato with Bacillus subtilis RB14. Appl Environ Microbiol 62: 4081–4085.

Bais, H.P., Fall, R., and Vivanco, J.M. (2004) Biocontrol of Bacillus subtilis against infection of Arabidopsis roots by Pseudomonas syringae is facilitated by biofilm formation and surfactin production. Plant Physiol 134: 307–319.

Bais, H.P., Weir, T.L., Perry, L., Gilroy, S., and Vivanco, J.M. (2006) The role of root exudates in rhizosphere interations with plants and other organisms. Annu Rev Plant Biol 57: 233–266.

Dubern, J.F., and Bloemberg, G.V. (2006) Influence of environmental conditions on putisolvins I and II production in Pseudomonas putida strain PCL1445. FEMS Microbiol Lett 263: 169–175.

Duffy, B.K., and Défago, G. (1999) Environmental factors modulating antibiotic and siderophore biosynthesis by Pseudomonas fluorescens biocontrol strains. Appl Environ Microbiol 65: 2429–2438.

Eeman, M., Berquand, A., Dufrene, Y.F., Paquot, M., Dufour, S., and Deleu, M. (2006) Penetration of surfactin into phospholipid monolayers: nanoscale interfacial organization. Langmuir 22: 11337–11345.

Espinosa-Urgel, M., Kolter, R., and Ramos, J.L. (2002) Root colonization by Pseudomonas putida: love at first sight. Microbiol UK 148: 1–3.

Fickers, P., Leclère, V., Guez, J., Bechet, M., Coucheney, F., Joris, B., and Jacques, P. (2008) Temperature dependence of mycosubtilin homologue production in Bacillus subtilis ATCC6633. Res Microbiol 159: 449–457.

Ghiglione, J.F., Gourbiere, F., Potier, P., Philippot, L., and Lensi, R. (2000). Role of respiratory nitrate reductase in ability of Peudomonas fluorescens YT101 to colonize the rhizosphere of maise. Appl Eviron Microbiol 66: 4012–4016.

Guérout-fleury, A., Frandsen, N., and Stragier, P. (1996) Plasmid for ectopic integration in Bacillus subtillis. Gene 180: 57–61.

Guez, J.S., Muller, C.H., Danze, P.M., Buchs, J., and Jacques, P. (2008) Respiration activity monitoring system (RAMOS), an efficient tool to study the influence of the oxygen transfer rate on the synthesis of lipopeptide by Bacillus subtilis ATCC6633. J Biotechnol 134: 121–126.

Hirsch, A.M., Bauer, W.D., Bird, D.M., Cullimore, J., Tyler, B., and Yoder, J.I. (2003) Molecular signals and receptors: controlling rhizosphere interactions between plants and other organisms. Ecology 84: 858–868.

Højberg, O., Schnider, U., Winteler, H.V., Sørensen, J., and Haas, D. (1999) Oxygen-sensing reporter strain of Pseudomonas fluorescens for monitoring the distribution of low-oxygen habitats in soil. Appl Environ Microbiol 65: 4085–4093.

Jacques P, Hbid C, Destain J, Razafindralambo H, Paquot M, De Pauw E, Thonart P. (1999). Optimization of biosurfactant lipopeptide production from Bacillus subtilis S499 by Plackett–Burman design. Appl Biochem Biotech 77: 223–233.

Jourdan, E., Henry, G., Duby, F., Dommes, J., Barthélemy, J.P., Thonart, P., and Ongena, M. (2009) Insights into the defense-related events occurring in plant cells following perceptions of surfactin–type lipopeptide from Bacillus subtilis. Mol Plant Microbe Interact 22: 456-468.

Kamilova, F., Kravchenko, L.V., Shaposhnikov, A.I., Makarova, N., and Lugtenberg, B. (2006) Effects of the tomato pathogen Fusarium oxysporum f. sp. radicis-lycopersici and of the biocontrol bacterium Pseudomonas fluorescens WCS365 on the composition of organic acids and sugars in tomato root exudate. Mol Plant Microbe Interact 19: 1121–1126.

Kim, P.I., Bai, H., Bai, D., Chae, H., Chung, S., Kim, Y., et al. (2004) Purification and characterization of a lipopeptide produced by Bacillus thuringiensis CMB26. J Appl Microbiol 97: 942–949.

Koumoutsi, A., Chen, X.H., Henne, A., Liesegang, H., Hitzeroth, G., Franke, P., et al. (2004) Structural and functional characterization of gene clusters directing nonribosomal synthesis of bioactive cyclic lipopeptides in Bacillus amyloliquefaciens strain FZB42. J Bacteriol 186: 1084–1096.

Results-Chapter 6

130

Kracht, M., Rokos, H., Ozel, M., Kowall, M., Pauli, G., and Vater, J. (1999) Antiviral and hemolytic activities of surfactin isoforms and their methyl ester derivatives. J Antibiot (Tokyo) 52: 613–619.

Lin, S.C., Lin, K.G., Lo, C.C., and Lin, Y.M. (1998) Enhanced biosurfactant production by a Bacillus licheniformis mutant. Enzyme Microb Technol 23: 267–273.

Lugtenberg, B.J.J., Dekkers, L., and Bloemberg, G.V. (2001) Molecular determinants of rhizosphere colonization by Pseudomonas. Annu Rev Phytopathol 39: 461–490.

Mukherjee, S., Dasa, P., and Sen, R. (2006) Towards commercial production of microbial surfactants. Trends Biotechnol 24: 509–515.

Mulligan, C.N. (2005) Environmental applications for biosurfactants. Environ Pollut 133: 183–198. Ongena, M., and Jacques, P. (2008) Bacillus lipopeptides: versatile weapons for plant disease

biocontrol. Trends Microbiol 16: 115–125. Ongena, M., Adam, A., Jourdan, E., Paquot, M., Brans, A., Joris, B., et al. (2007) Surfactin and

fengycin lipopeptides of Bacillus subtilis as elicitors of induced systemic resistance in plants. Environ Microbiol 9: 1084–1090.

Ongena, M., Jourdan, E., Adam, A., Schaefer, M., Budzikiewicz, H., and Thonart, P. (2008) Amino acids, iron, and growth rate as key factors influencing production of the Pseudomonas putida BTP1 benzylamine derivative involved in systemic resistance induction in different plants. Microb Ecol 55: 280–292.

Peypoux, F., Bonmatin, J.M., and Wallach, J. (1999) Recent trends in the biochemistry of surfactin. Appl Microbiol Biotechnol 51: 553–563.

Rainey, P.B. (1999) Adaptation of Pseudomonas fluorescens to the plant rhizosphere. Environ Microbiol 1: 243–257.

Ramey, B.E., Koutsoudis, M., von Bodman, S.B., and Fuqua, C. (2004) Biofilm formation in plant–microbe associations. Curr Opin Microbiol 7: 602–609.

Romero, D., de Vicente, A., Rakotoaly, R.H., Dufour, S.E., Veening, J.W., Arrebola, E., et al. (2007) The iturin and fengycin families of lipopeptides are key factors in antagonism of Bacillus subtilis toward Podosphaera fusca. Mol Plant Microbe Interact 20: 430–440.

Touré, Y., Ongena, M., Jacques, P., Guiro, A., and Thonart, P. (2004) Role of lipopeptides produced by Bacillus subtilis GA1 in the reduction of grey mould disease caused by Botrytis cinerea on apple. J Appl Microbiol 96: 1151–1160.

Van Rij, E.T., Wesselink, M., Chin-A-Woeng, T.F.C., Bloemberg, G.V., and Lugtenberg, B.J.J. (2004) Influence of environmental conditions on the production of phenazine-1-carboxamide by Pseudomonas chlororaphis PCL1391. Mol Plant Microbe Interact 17: 557–566.

131

GENERAL DISCUSSION AND FUTURE PROSPECTS

General discussion and future prospects

132

GENERAL DISCUSSION AND FUTURE PROSPECTS

B. amyloliquefaciens strain S499 as biocontrol agent

During the past decades, microbial biopesticides have gained increased attention and interest

for developing environmentally friendly approaches to control plant diseases by reducing the

use of chemicals (Copping and Menn, 2000). In that context, isolation of new bacterial strains

with promising biocontrol potential is necessary to enlarge the still limited range of

marketable products. In the first part of this thesis (Chapter 4), we have selected Bacillus

amyloliquefaciens strain S499 out of more than forty other isolates to conduct open-field

experiments in order to evaluate its efficacy to control a disease that is devastating tomato

plantings in Burundi. This disease is caused by a fungal isolate belonging to the Fusarium

semitectum species based on preliminary characterization. This fungus has not been

previously described as an infectious agent attacking aerial parts of tomato plantings but is

well known as a major cause of seedling diseases on cotton and other crops in many

subtropical african countries (Zhang et al., 1996; Abd-Elsalam et al., 2003). Our results

showed that soil treatment with strain S499 significantly reduced leaf infection rates. Fungal

disease reduction by strain S499 has already been observed on tomato and other plants in

greenhouse system (Jourdan et al., 2009) but our experiments are the first demonstration of

biocontrol activity for that strain against Fusarium infection under field conditions.

In addition, the same treatments significantly increased growth and fruit yield of tomato,

strongly suggesting that S499 strain may also retain plant growth promoting activity besides

disease reduction. A better understanding of such growth promoting activity of the bacterium

deserves further investigation but it may arise from various phenomena such as enhancement

of nutrient (essential oligoelements) availability, production of hormone-like substances or

inhibition of the endogenous soil microflora that is not pathogenic per se but may trigger

adverse effects on plant growth (McKellar and Nelson, 2003; McSpadden Gardener, 2010).

The last phenomenon may be explained by the secretion of antimicrobials by S499 that

directly affect the development of auxiliary microflora or by competition for the limited

sources of nutrients provided through root exudation. For example, diverse seed-colonizing

bacteria can consume nutrients that are released into the soil during germination thereby

suppressing pathogen germination and growth (Pal and McSpadden Gardener, 2006). In both

cases, these effects can only be supported by the fact that S499 strain successfully colonized

the plant root system and rhizosphere. Accurate cell counts on root samples could not be

performed during these field experiments but other results obtained in our research (Chapter


133

6) together with those obtained in the laboratory have actually illustrated the good root-

colonizing potential of S499 (Ongena et al., 2002; Ongena et al., 2007; Ongena et al., 2010).

The success of an introduced biocontrol agent in suppressing disease can be influenced

by the conduciveness of the environment to the disease but also by the successful

establishment and proliferation of the antagonist in the root zone (Landa, 2004; Chatterton

and Punja, 2010). In our assays, the use of inoculum densities as low as 105 cfu ml-1 showed

to be relevant to sufficiently reduce the disease. It is clear that biocontrol was not significantly

improved by using higher Bacillus inoculum concentrations suggesting that populations tend

to equilibrate after a certain time. It is important since the success of biocontrol also depends

on the preparation of appropriate formulations of the antagonist and from the commercial

side, it is also crucial to use the active ingredient at the lowest concentration.

Also regarding growth and antibiotic lipopeptide synthesis, strain B. amyloliquefaciens

S499 is well adapted to soil temperature range between 20 °C and 25 °C (Chapter 5)

prevailing in (sub)tropical regions from which it was isolated. Additional experiments are

obviously required but taken together; our results suggest that this strain can be proposed for

biological product to be used in Burundi and other subtropical regions, not only for tomato

but also for other crops.

Strain S499 acting via ISR

Some threshold in root colonization rate by the introduced biocontrol agent is also a

prerequisite for disease control activity. Thought that pathogen inhibition might be attributed

to other mechanisms such as the production of enzymes, antibiotics or siderophores,

geographic separation of the beneficial strain and the pathogen in our field trials strongly

suggests that disease suppression was due to induction of resistance in the host plant. Such an

ISR-triggering potential of strain S499 has already been observed on tomato and other plants

in greenhouse systems (Ongena et al., 2002; Ongena et al., 2005b; Ongena et al., 2007;

Choudhary and Johri, 2009; Jourdan et al., 2009). More globally, systemic induction of plant

defenses is currently viewed as an important mechanism by which antagonistic strains, among

which Bacillus isolates inoculated in soil diminish foliar diseases (Harman et al., 2004; Liu et

al., 2008; Huang et al., 2010). ISR-based biocontrol strategies are promising and our trials

may be added to the limited number of those that were successfully performed under field

conditions using Bacillus as inducing agent.


134

Linking ISR potential with surfactin production in vivo

Although a lot still needs to be discovered, some information is available concerning the

molecular mechanisms underlying ISR triggered by beneficial rhizobacteria. For instance, a

previous work realized in the laboratory has demonstrated that surfactin and to a lower extent

fengycin lipopeptides act as signals emitted by Bacillus to trigger ISR (Ongena et al., 2007).

Surfactin plays a predominant role and represent the main ingredient secreted by strain S499

retaining plant defense eliciting potential. These molecules can thus be considered as a novel

class of microbial-associated molecular patterns perceived by plant cells to activate immune

reaction (Jourdan et al., 2009). Together with the volatile compound 2,3-butandiol (Ryu et al.,

2004), these lipopeptides are the sole determinants for ISR elicitation identified from Bacillus

spp so far.

That said, this ISR eliciting activity of surfactin has been demonstrated by treating

plants with the pure compound and by using over-producing mutants (Ongena et al., 2007).

However, the demonstration of an efficient surfactin production by S499 cells colonizing

roots is crucial to provide conclusive and direct evidence for the implication of this LP in ISR

triggering. In that context, the main goal of this thesis was to better evaluate whether these

lipopeptides are actually produced in the rhizosphere of plants colonized at the root level by

Bacillus. This is a challenging issue because of the small amounts of LPs potentially secreted

and/or the difficulties to quantify them in a complex mixture relative to other organic

compounds present in the environment. Moreover, a major challenge is the reproducibility of

the extraction/detection efficacy protocol. Nevertheless, in this thesis work, we combined

different approaches based on electrospray and imaging mass spectrometry techniques to

demonstrate that surfactin is actually secreted by strain S499 evolving in the tomato

rhizosphere (Chapter 5). Efficient synthesis of this LP in planta was further supported by

using a psrfA-LacZ reporter system inserted in strain BGS3 which is a surfactin overproducer

that protected bean and tomato plants against Botrytis disease via ISR (Ongena et al. 2007).

By contrast with S499, the BGS3 isolate derives from the laboratory “acclimated” isolate B.

subtilis 168 and therefore, is no longer considered as “natural”. It is nevertheless a nice model

strain and results showed that a significant level of psrfA induction was observed upon the

establishment of a stable BGS3 population on roots and surfactins extracted from the

rhizosphere were produced in biologically important quantities (Chapter 6). From both studies

with S499 and BGS3, it also appears that not only the nature of tomato root exudates but also

several other physiological and environmental parameters inherent to Bacillus life in the

rhizosphere are conducive for surfactin biosynthesis. More precisely, oxygen limitation,


135

adhesion on solid surface, biofilm formation or reduced growth rate are absolutely not

detrimental to surfactin production.

Surfactin in the rhizosphere: the right form at the right concentration

From a qualitative point of view, our results also revealed that the relative proportions of

surfactin homologues produced by cells colonizing tomato roots is clearly different compared

to those observed upon growth of S499 as planktonic cells in an optimized laboratory culture

medium (Chapter 5). The C15 form represents more than 75% of the total surfactins secreted

by S499 in planta and long-fatty acid chain surfactins are also predominantly produced by

BGS3 colonizing roots (Chapter 6). Such an in vivo LP signature is interesting in the context

of ISR since homologues with the longest fatty acid chains were proven to be the more active

regarding their potential to trigger plant defense reactions (Jourdan et al., 2009). Such an

important role for the length of the acyl chain is most probably related to the fact that it

readily inserts into the plasma membrane of plant cells. However, this insertion is limited and

does not lead to pore-formation and phytotoxicity at concentrations below 20µM (Jourdan et

al., 2009; Henry et al., unpublished results).

Whether in situ productivities may generate rhizospheric surfactin concentrations

sufficient to provide any biological effect remains questionable. This LP is active in the range

2-10 µM for inducing plant defense reactions and triggering ISR. However, quantities

recovered from the hydroponic medium surrounding colonized tomato roots in our

experiments ranged from approx. 0.12 µM to 1.8 µM (Chapters 5 and 6). Possible enzymatic

degradation by auxiliary microflora and/or incomplete extraction may contribute to

underestimate the actual abundance of these LPs in the vicinity of the root system. In a more

artificial but gnotobiotic system, measurements of LPs secreted into the surrounding medium

by S499 cells colonizing tomato plantlets revealed that surfactin concentration may locally

reach 50 µM (Chapter 5). Such micromolar concentrations probably better reflects the

amounts of surfactin that may be actually secreted by Bacillus evolving on roots and it may

thus be compatible with biological activity as inducer of resistance in the host plant (Ongena

et al., 2007; Jourdan et al., 2009). This is supported by the higher surfactin quantities that

were detected in the cucumber rhizosphere colonized by strain QST713 (Kinsella et al.,

2009).


136

Surfactin production in the rhizosphere: other advantages!

Several studies showed that surfactins could be involved in motility and biofilm formation

that are essential traits for plant-associated bacteria in rhizosphere competence and root

colonization process (Danhorn and Fuqua, 2007). For instance, surfactin is necessary but not

always sufficient for surface motility of B. subtilis (Kinsinger et al., 2003; Kearns et al., 2004;

Julkowska et al., 2005). Results obtained with at least two different B. subtilis and B.

amyloliquefaciens strains have shown that surfactin but no other lipopeptides was required for

the formation of biofilms and pellicles (Hofemeister et al., 2004). The production of surfactin

was also demonstrated to be essential for biofilm formation and colonization of Arabidopsis

roots by the strain B. subtilis 6051 (Bais et al., 2004). The same study also showed that

surfactin may also play some role in direct antagonism developed towards bacterial plant

pathogens and the molecule also retains some antiviral activity. Occasionally, surfactin was

also described for its antagonistic effect against insect larvae (Assie et al., 2002).

Additionally, this LP may act as signalling molecule in quorum-sensing and in inducing

cannibalism and matrix formation in biofilms. Beside biocontrol-related traits, surfactins thus

also retain different roles in the development and survival of B. subtilis strains in their natural

habitat (see Raaijmakers et al., 2010 for recent review).

Global pattern of lipopeptides produced in vivo by S499

In vitro, B. amyloliquefaciens S499 co-produces efficiently the three families of

lipopeptides retaining distinct biological activities (Chapter 2). From a practical point of view,

the strain should therefore display some multi-faceted biocontrol activity in different

environmental sites against many plant pathogens. However, only surfactin is produced in

biologically significant amounts in planta (Chapter 5). Iturins and fengycins are also secreted

by S499 colonizing tomato roots but in much lower amounts that are probably not sufficient

to support consistent antifungal activity. It means that these two LPs would not play a crucial

role in antagonism that has to be potentially developed at the root level towards fungal

pathogens. As surfactins are mostly antibacterial and antiviral, our results strongly suggest

that strain S499 would not be very efficient at protecting tomato plants against diseases

caused by soilborne fungal/oomycete phytopathogens.

Such a differential production of LP families strongly suggests that in situ expression of

the related biosynthesis genes may also considerably differ. This could not be demonstrated

experimentally because of the lack of reporter derivatives of strain S499 which is highly

recalcitrant to transformation like many natural B. subtilis/amyloliquefaciens isolates.


137

Investigating the differential expression of srf, fen and myc genes under natural soil conditions

in S499 or in another natural isolate co-producing the three families, should represent one of

the main future prospects.

Whether such differential production of the three lipopeptide families is specific for

S499 or not is difficult to establish given the rather limited information available in the

literature concerning in situ synthesis of these compounds by other Bacillus isolates.

Mizumoto and Shoda (2007) have also demonstrated that iturins in comparison with

surfactins are less associated to young seedlings in cucumber. By contrast, a few other studies

have shown that lipopeptides from the three families were occasionally detected in the

phytosphere, illustrating that natural conditions are conducive for their biosynthesis, at least in

some cases (Ongena and Jacques, 2008). However, these studies were conducted with

different Bacillus strains on different crops (and organs) and a more consistent information

can only be obtain by comparing LP production by various isolates colonizing the same

model plant in the same conditions of growth. This is clearly one of the main perspectives of

this work.

Another appealing perspective of this thesis is to establish whether the surfactin-

enriched LP pattern secreted by S499 in situ is conserved upon colonization of other plants

than tomato. Preliminary experiments suggest that a similar trend occurs with corn and lettuce

but the LP signature may also be strongly influenced by the host plant species since almost no

production of surfactin was observed upon growth on cucumber and soy (Ongena et al.,

unpublished data). The influence of many other rhizosphere-related factors such as the

auxiliary microflora, the soil type and heterogeneity or the plant developmental stage, on

Bacillus LP signature also deserves further investigation.

Concluding comments

In light of the initial great optimism (Copping and Menn, 2000), and in spite of

extensive research efforts, it is paradoxical that progress in achieving commercial, large-scale

usage of biological control has been slow. The reasons for this are partly of commercial and

legislative character. Cost-effective production, good storage shelf-life, worked out

application procedures, marketing or legislative requirements for registration may explain

limitations in product’s distribution and selling. Another serious constraint is that, in

comparison to the conventional pesticides that usually have simple knockout effects, the field

efficacy of many biological products, which most often are more dependent on environmental

and other circumstances, is inherently more variable. Because many successful antagonists


138

exert their activity by secreting antibiotic metabolites, an obvious approach is to optimize this

process of antibiotic production in the field. To maximize it, a fundamental understanding of

their production in their natural niche is imperative.

In this work, we have inoculated Bacillus strains in situ on tomato cultivated in

greenhouse in hydroponic mode. We first assumed that root exudates represent food-web for

bacterial fitness in the rhizosphere regarding antibiotic production. Our second hypothesis

was that the lipopeptides biosynthesis in situ may be modulated by multiple rhizosphere-

related parameters in addition to the substrates. The influence of all these environmental

factors on the lipopeptide pattern production has been evaluated qualitatively and

quantitatively. Globally, it was shown that the nutritive and physico-chemical conditions of

bacterial cells growing in the rhizosphere on the root system are conducive for efficient

surfactin lipopeptide synthesis. Thus, these results provide a first demonstration of efficient

related-gene expression and surfactin production in the rhizosphere during Bacillus-plant

interaction.

From a more general and ecological point of view, it is often speculated that the frequent

occurrence of B. subtilis in its natural environment might be due to the selective advantage

conferred by the panoply of bioactive metabolites that it may produce (Stein, 2005). However,

even if some B. subtilis or amyloliquefaciens strains are well equipped genetically to produce

a vast array of antibiotics (Chen et al., 2009), our results suggest that only a limited part of

this antibiotic-devoted genetic background may be readily expressed in situ. Thus the results

from this work also substantially contribute to enhance our knowledge of Bacillus fitness in

natural living conditions regarding its antibiotic producing potential.

References

Abd-Elsalam KA, Schnieder F, Asran-Amal A, Khalil MS, Verreet JA. (2003). Intra-species genomic groups in Fusarium semitectum and their correlation with origin and cultural characteristics. Z Pflanzenk Pflanzen-J Plant Dis Protect 110: 409-418.

Assie LK, Deleu M, Arnaud L, Paquot M, Thonart P, Gaspar C, Haubruge E. (2002). Insecticide activity of surfactins and iturins from a biopesticide Bacillus subtilis Cohn (S499 strain). Meded Rijksuniv Gent Fak Landbouwkd Toegep Biol Wet 67: 647–655.

Bais H, Fall R, Vivanco J. (2004). Biocontrol of Bacillus subtilis against infection of Arabidopsis roots by Pseudomonas syringae is facilitated by biofilm formation and surfactin production. J Plant Physiol 134: 307–319.

Chatterton S, Punja ZK. (2010). Factors influencing colonization of cucumber roots by Clonostachys rosea f. catenulata, a biological disease control agent. Biocontrol Sci Techn 20: 37-55.

Chen X, Koumoutsi A, Scholz R, Schneider K, Vater J, Sussmuth R, Piel J, Borriss R. (2009). Genome analysis of Bacillus amyloliquefaciens FZB42 reveals its potential for biocontrol of plant pathogens. J Biotechnol 140: 27-37.

Choudhary DK, Johri BN. (2009). Interactions of Bacillus spp. and plants- With special reference to induced systemic resistance (ISR). Microbiol Res 164: 493-513.


139

Copping LG, Menn JJ. (2000). Biopesticides: a review of their action,applications and efficacy. Pest Manag Sci 56: 651-676.

Danhorn T, Fuqua C. (2007). Biofilm formation by plant associated bacteria. Annu Rev Microbiol 61: 401–422.

Harman GE, Howell CR, Viterbo A, Chet I, Lorito M. (2004). Trichoderma species-opportunistic, avirulent plant symbionts. Nat Rev Microbiol 2: 43-56.

Hofemeister J, Conrad B, Adler B. (2004). Genetic analysis of the biosynthesis of non-ribosomal peptide- and polyketidelike antibiotics, iron uptake and biofilm formation by Bacillus subtilis A1/3. Mol Genet Genomics 272: 363–378.

Huang C, Yang K, Liu Y, Lin Y, Chen C. (2010). Suppression of southern corn leaf blight by a plant growth-promoting rhizobacterium Bacillus cereus C1L. Ann Appl Biol 157: 45–53.

Jourdan E, Henry G, Duby F, Dommes J, Barthelemy JP, Thonart P, Ongena M. (2009). Insights into the defense-related events occurring in plant cells following perception of surfactin-type lipopeptide from Bacillus subtilis. Mol Plant Microbe Interact 22: 456–468.

Julkowska D, Obuchowski M, Holland I, Seror S. (2005). Comparative analysis of the development of swarming communities of Bacillus subtilis 168 and a natural wild type: critical effects of surfactin and the composition of the medium. J Bacteriol 187: 65-76.

Kearns DB, Chu F, Rudner R, Losick R. (2004). Genes governing swarming in Bacillus subtilis and evidence for a phase variation mechanism controlling surface motility. Mol Microbiol 52: 357–369.

Kinsella K, Schulthess CP, Morris TF, Stuart JD. (2009). Rapid quantification of Bacillus subtilis antibiotics in the rhizosphere. Soil Biol Biochem 41: 374–379.

Kinsinger R, Shirk M, Fall R. (2003). Rapid surface motility in Bacillus subtilis is dependent on extracellular surfactin and potassium ion. J Bacteriol 185: 5627–5631.

Landa BB. (2004). Influence of temperature on plant-rhizobacteria interactions related to biocontrol potential for suppression of Fusarium wilt of chickpea. Plant Pathol 53: 341-352.

Liu Y, Huang C, Chen C. (2008). Evidence of induced systemic resistance against Botrytis elliptica in Lily. Phytopathology 98: 830-836.

McKellar ME, Nelson EB. (2003). Compost-induced suppression of Pythium damping-off is mediated by fatty-acid-metabolizing seed-colonizing microbial communities. Appl Environ Microbiol 69: 452-460.

McSpadden Gardener B. (2010). Biocontrol of plant pathogens and plant growth promotion by Bacillus. In: U. Gisi et al. (eds.). Recent developments in management of plant diseases, Plant Pathology in the 21st Century 1. Springer Science.

Mizumoto S, Shoda M. (2007). Medium optimization of antifungal lipopeptide, iturin A, production by Bacillus subtilis in solid-state fermentation by response surface methodology. Appl Microbiol Biotechnol 76: 101-108.

Ongena M, Cawoy H, Smargiassi M, Nihorimbere V, Jourdan E, Thonart P. (2010). Modulation of the lipopeptide pattern secreted by Bacillus subtilis upon colonization of different plant roots. IOBC bulletin, in Press.

Ongena M, Jacques P. (2008). Bacillus lipopeptides: versatile weapons for plant disease biocontrol. Trends Microbiol 16: 115-125.

Ongena M, Jacques P, de Pauw E, Thonart P. (2002). Synthesis of peptide-modified pyoverdins by a fluorescent Pseudomonas strain grown in isoleucine-supplemented medium. Lett Pept Sci 8: 21–27.

Ongena M, Jacques P, Touré Y, Destain J, Jabrane A, Thonart P. (2005b). Involvement of fengycin-type lipopeptides in the multifaceted biocontrol potential of Bacillus subtilis. Appl Microbiol Biotechnol 69: 29-38.

Ongena M, Jourdan E, Adam A, Paquot M, Brans A, Joris B, Arpigny J-L, Thonart P. (2007). Surfactin and fengycin lipopeptides of Bacillus subtilis as elicitors of induced systemic resistance in plants. Environ Microbiol 9: 1084–1090.

Pal K, McSpadden-Gardener B. (2006). Biological control of plant pathogens. The Plant Health Instructor. DOI:10.1094/PHI-A-2006-1117-02.


140

Raaijmakers J, De Bruijn I, Nybroe O, Ongena M. (2010). Natural functions of lipopeptides from Bacillus and Pseudomonas: more than surfactants and antibiotics. FEMS Microbiol Rev 34: 1037-1062.

Ryu CM, Farag MA, Hu CH, Reddy MS, Kloepper JW, Pare PW. (2004). Bacterial volatiles induce systemic resistance in Arabidopsis. Plant Physiol 134: 1017-1026.

Stein T. (2005). Bacillus subtilis antibiotics: structures, synhesis and specifics functions. Mol Microbiol 56: 845-847.

Zhang JX, Howell CR, Starr JL, Wheeler MH. (1996). Frequency of isolation and the pathogenicity of Fusarium species associated with roots of healthy cotton seedlings. Mycol Res 100: 747-752.

rhizosphere-driven lipopeptide production by different...

Documents