Highways and Subways a Story of Fungi and Bacteria in Soils

Highways and subways A story of fungi and bacteria in soils

Th esis presented to the University of Neuchâtel, Switzerland

For the Philosophiae doctor degree in Sciences

by Anaële Simon

2016

Th esis director Prof. Dr. Pilar Junier University of Neuchâtel, Switzerland

Th esis co-director Prof. Dr. Eric P. Verrecchia University of Lausanne, Switzerland

Th esis commitee Prof. Dr. Geoff rey M. Gadd, University of Dundee, United Kingdom Dr. Markus Künzler, ETH Zürich, Switzerland Dr. Saskia Bindschedler, University of Neuchâtel, Switzerland

Faculté des sciences Secrétariat-décanat de Faculté Rue Emile-Argand 11 2000 Neuchâtel - Suisse Tél: + 41 (0)32 718 2100 E-mail: [email protected]

IMPRIMATUR POUR THESE DE DOCTORAT

La Faculté des sciences de l'Université de Neuchâtel autorise l'impression de la présente thèse soutenue par

Madame Anaële SIMON

Titre: “Highways and subways – a story of fungi and bacteria in soils”

sur le rapport des membres du jury composé comme suit:

- Prof. Pilar Junier, directrice de thèse, Université de Neuchâtel, Suisse - Prof. Eric P. Verrecchia, co-directeur de thèse, Université de Lausanne, Suisse - Prof. G.M. Gadd, University of Dundee, United Kingdom - Dr Saskia Bindschedler, Université de Neuchâtel, Suisse - Dr Markus Künzler, ETH Zürich, Suisse

Neuchâtel, le 22 avril 2016 Le Doyen, Prof. B. Colbois

Imprimatur pour thèse de doctorat www.unine.ch/sciences

Highways and subways A story of fungi and bacteria in soils

Anaële Simon

Cover and chapter illustrations: Anaële Simon

Th e most dangerous of all behaviors may consist of doing things ‘because we are supposed to’. Marshall B. Rosenberg

Abstract of the thesis

Traditionally, bacteriologists and mycologists have conducted study of their respective fi elds separately, despite of the fact that bacteria and fungi co-exist in almost every type of ecosystem. Amongst the numerous types of interactions possibly occurring between fungi and bacteria, this thesis focuses on fungal highways, a mechanism by which bacteria disperse along fungal hyphae. We investigated various soils, with special focus on soils under the infl uence of the oxalate-carbonate pathway. In this pathway, fungal-driven bacterial dispersal seems to be relevant for bacterial oxalotrophic activity. We developed and validated a device called fungal highway column, intended for the targeted isolation of fungi and bacteria migrating along fungal hyphae. Culture-based and culture-independent approaches allowed for an insight into the diversity of fungi and bacteria able to interact via fungal highways in soils. Our fi ndings propose that migration of bacteria along hyphae might be either extrahyphal (fungal highways) or intrahyphal (fungal subways). We observed that bacterial migration along fungal hyphae is dependent on the culture medium, but we could not determine the element(s) triggering this diff e- rential behavior. Our results show that fungal highways (and subways) should be considered as crucial for the colonization of novel environments by bacteria, and that they play an important role in the structuring of bacterial communities in soils. As well, we demonstrated that fungi present in the oxalate-carbonate pathway are able to dissolve calcium oxalate and to disperse non-oxalotrophic bacteria, thus redefi ning the role of the microbial components in this pathway. Finally, we observed surprising properties of fungi, resembling to nitrogen fi xation, but we could neither prove nor disprove this ability. Th is thesis gives an insight into the complexity of fungal-bacterial interactions, and demonstrates the importance of considering both groups of organisms for the study of their ecology in soils.

VII

Résumé de la thèse

Traditionnellement, bactériologues et mycologues mènent leurs recherches séparément, et ce malgré le fait que bactéries et champignons coexistent dans pratiquement tous les écosystèmes. Parmi les nom- breuses interactions qui peuvent avoir lieu entre champignons et bactéries, cette thèse étudie les autoroutes fongiques, une interaction dans laquelle les bactéries sont capables de se déplacer le long d’hyphes de champignons. Nous avons investigué diff érents sols, avec une attention particulière portée à la voie oxalate-carbonate. Dans cette voie métabolique, il semblerait que la dispersion de bactéries le long d’hyphes fongiques soit importante pour l’activité oxalotrophe des bactéries. Nous avons dé- veloppé un nouvel outil, appelé colonne pour autoroutes fongiques, destiné à isoler spécifi quement des champignons et des bactéries capables de se déplacer le long d’hyphes fongiques. Une approche à la fois cultivable et moléculaire nous a permis d’avoir un aperçu de la diversité des champignons et des bactéries interagissant par autoroutes fongiques dans les sols. Nos résultats suggèrent que la migration de bactéries le long des hyphes peut être soit extra-hyphale (autoroutes fongiques), soit intra-hyphale (métros fongiques). Nous avons observé que le déplacement de bactéries le long des hyphes fongiques est dépendant du milieu de culture, mais nous n’avons pas pu trouver l’élément régissant ce comporte- ment diff érentiel. Nos résultats montrent que les autoroutes - et métros - fongiques doivent être consi- dérés comme cruciaux pour la colonisation de nouveaux environnements par les bactéries. De plus, ces interactions jouent un rôle important dans la structuration des communautés bactériennes dans les sols. Nous avons également démontré que des champignons présents dans la voie oxalate-carbonate sont capables de dissoudre l’oxalate de calcium, ainsi que de disperser des bactéries non-oxalotrophes, redéfi nissant ainsi le rôle de chaque organisme dans la voie oxalate-carbonate. Finalement, nous avons observé une propriété surprenante de certains champignons, que l’ont pourrait relier à de la fi xation d’azote, mais nous n’avons pu ni prouver, ni démentir, cette capacité. Cette thèse permet d’entrevoir la complexité des interactions champignons-bactéries, et démontre qu’il est important de considérer les deux groupes d’organismes lors d’études de leur écologie dans les sols.

Acknowledgements

To Pilar, my thesis director, thank you for your trust, the amount of time you spent for me, the fruitful discussions for each article, and your enthusiasm for my research and my ideas.

To Eric, my thesis co-director, it was a pleasure and such an energy-booster to see your optimism every time I presented you my latest results or hypotheses.

To people in the lab, and especially the three persons that were constantly contributing to my happiness. I hope they will recognize themselves. Th ank you for all these funny moments, they were precious.

To the other crazy, humanistic and poetic people I met at the University of Neuchâtel. For the ‘ape- ro of the third Friday of the month’. For the ‘Friday-epic-discussion-topics’ in the cafeteria during lunchbreak. For the huge amount of stupid jokes. For the blind tests and the sharing of comic books. We had great moments. Th ank you.

To the nice people I met during conferences and courses, and the fruitful and funny times we shared together around the world.

To the NonViolent communication I discovered during this period of time, and the persons who shared these moments with me. Th ey helped me growing up.

To my friends, and my beloved mother and stepfather, for your constant support.

And of course, this research would not have been possible without the fi nancial support of the Swiss National Science Foundation. As well, the Comité Egalité des Chances (Neuchâtel, Switzerland) fi - nanced the fi eld work in Morocco, and the Interuniversity Doctoral Program in Organismal Biology (Neuchâtel, Switzerland) fi nanced my participation to the 15th International Symposium on Microbal Ecology (ISME 15) in Seoul, South Korea.

Table of content

General introduction 1 Context of the present thesis 3 About soils and pedogenesis 3 Colonization of land by microorganisms 4 Element cycling by microorganisms in soils 6 Fungal-bacterial interactions in soils 9 About fungal highways 10 Objectives of this thesis 10 Acknowledgements 11 References 11

Chapter 1 - Fungal highways columns 17 Foreword 19 Manuscript 21 Supplementary information 37 Main fi ndings and perspectives 39

Chapter 2 - Culture-independent approach 41 Foreword 43 Manuscript 45 Supplementary information 59 Additional experiments 93 Main fi ndings and perspectives 95

Chapter 3 - Element cycling and metabolism 97 Foreword 99

Chapter 3a - Fungal oxalotrophy 101 Manuscript 101

Chapter 3b - Fungi and nitrogen fi xation 115 Manuscript 115

XIII Supplementary information 125

Chapter 3c - Migration and metabolism 127 Manuscript 127

Additional experiments 137 Main fi ndings and perspectives 145

General discussion 149 About the methodology 151 Redefi ning the oxalate-carbonate pathway 151 Fungal highways as structuring factors of soil bacterial communities 153 Intrahyphal bacterial migration? 153 Importance of fungal-driven bacterial dispersal in biogeochemical cycling 155 Conclusion and perspectives 155 References 156

Collaborations, presentations and CV 161 Collaborations 163 Presentations 165 Curriculum vitae 167

XIV

General introduction

Context of the present thesis

Th e present thesis is framed within two successive collaborative research projects founded by the Swiss National Science Foundation (SNSF), respectively entitled Th e oxalate carbonate pathway: measuring biological interactions and dynamics in a natural carbon sink ecosystem (2012-2014) and Environmental factors limiting the interaction of bacteria and fungi in the context of the oxalate-carbonate pathway (2014- 2016). In this context, this thesis focuses on fungal-bacterial interactions in one particular habitat, the soil, and the oxalate-carbonate pathway was taken as a concrete example of plant-fungal-bacterial interaction occurring in soils. Th is general introduction will briefl y present soils through pedogenesis, the main roles of fungi and bacteria in soils, alongside with their interactions, and the oxalate-carbonate pathway as a model for the study of fungal-bacterial interactions.

About soils and pedogenesis

Soil is a very particular environment, defi ned as a structured interface between mineral and life. For- med from nude rock, soil is the subject of many changes in its lifetime, until it eventually reaches an equilibrium state. Substrate for the growth of plants, habitat for numerous living organisms, and essential for our nutrition, soil is a rich and complex environment, yet poorly known by the general public. In Switzerland, the study of soils is not part of the educational program in the 11 years of mandatory school (Conférence intercantonale [...], 2016), and during the tree years I was teaching in high schools before the beginning of this thesis, I never met a colleague who was teaching basic pedology, or even mentioning the importance of soils to students. Yet, 2015 was declared as the International Year of Soil by the Food and Agriculture Organization of the United Nations (FAO), with the following key messages (FAO, 2016): healthy soils are the basis for healthy food production; soils are the foundation for ve- getation which is cultivated or managed for feed, fi bre, fuel and medicinal products; soils support our planet’s biodiversity and they host a quarter of the total; soils help to combat and adapt to climate change by playing a key role in the carbon cycle; soils store and fi lter water, improving our resilience to fl oods and droughts; soil is a non-renewable resource; its preservation is essential for food security and our sustainable future.

Pedogenesis begins with the physicochemical alteration of a parental rock, accelerated by subsequent microbiological weathering. Th e development into a fertile soil is linked to two main phenomena: the formation of layered aluminium silicates, more commonly called clays, through hydrolysis of silicate minerals (Gobat et al., 2003, Mueller, 2015), and their association with complex organic matter, forming the so-called clay-humus complex, which is the key element of soil fertility (Th eng, 1982, Gobat

3 Figure 1. Simplifi ed scheme of the evolution of a soil on carbonated (top) and non-carbonated (bottom) bedrock.

et al., 2003). Th e organic matter mainly originates from plants, which deposit organic compounds into the soil through leaf litter, root litter, and root exudates (Ekschmitt et al., 2008). In the soil, these compounds can be either mineralized (i.e. turned into inorganic compounds through physical, chemical or biological processes), or humifi ed (i.e. condensed in larger organic macromolecules). Th ey can also evolve through the action of soil organisms (Gobat et al., 2003). Calcium or iron cations allow for the association between clays and humifi ed organic matter (Th eng, 1982, Gobat et al., 2003). Clay-humus complex is the basis of the well-known granulated texture of a fertile soil. But through the action of water, the link between clay and humus can be weakened. Furthermore, when precipitations overtake evaporation, downwards transfers are predominant, leading to the accumulation of clays and other metal oxides in deeper horizons (Gobat et al., 2003, Laliberte et al., 2013). Upward transfers occur when evaporation overtakes precipitation, but downward transfers can also be compensated by plants, transferring elements from their roots to the leaves, which will return to the soil surface layers through litter deposition (Gobat et al., 2003). Classically, fi ve factors are considered as being crucial in the formation of soils: climate, parental rock on which the soil is built, living organisms, topography, and time (Jenny, 1941). In parallel to this factorial approach, pedogenesis can also be modelled regarding energy, or processes, such as addition and removal of material to the soil, or translocation and transformation of material within a profi le (Minasny et al., 2008). A depiction of the general trends of the evolution of a soil on carbonated and non-carbonated parental substrates is shown in Figure 1.

Colonization of land by microorganisms

Fungi, or more precisely the kingdom of Mycota, can be considered as crucial for life on Earth, especially for the colonization of land by plants, because of their associations with plant roots, called mycorrhizas. In these associations, fungi absorb mineral nutrients from the environment and deliver them to the plants, while the latter provide the fungi with fi xed carbon (Deacon, 2006). Fungi are eukaryotic organisms that present the singular property of growing in the form of fi laments, called hyphae, for-

4 ming a network called mycelium (Deacon, 2006). Some fungi can be single-celled (yeasts), but we will not go further into details about these particular fungal forms, as most of soil fungi are fi lamentous (Tedersoo et al., 2014). In contrast to other fi lamentous organisms, fungi present very polarized, apical cell growth. Th e extension of fungi is limited to a very narrow part of the hyphal tips. Cell wall and membrane synthesis occur only at these parts, called extension zones (Gow, 1995). Th us, fungi are able to translocate cell wall, membrane materials, but also organelles, from distal regions in the hyphae towards the extension zones (Schütte, 1956, Markham, 1995). Likely, fungi are also able to translocate nutrients throughout their mycelium (Boswell et al., 2007). Up to 80 % of plant nitrogen, and even up to 90 % of plant phosphorous can be provided by mycorrhizal fungi (van der Heijden et al., 2015). Th e supply of organic carbon is crucial for fungi, as they are chemo-organo-heterotroph organisms, meaning that they need organic compounds as energy and electron sources. Th ey are also absorbotrophs, secreting enzymes or other compounds in order to degrade complex polymers, and then absorbing the released soluble compounds (Deacon, 2006). About 20 % of the plant photosynthates are allocated to the mycorrhizal fungi. So, mycorrhizas play key roles in phosphorus, nitrogen, and carbon cycling in soils (van der Heijden et al., 2015).

As fungi probably colonized land long before plants, maybe even since late Proterozoic, it is suggested that fungi would have already developed effi cient means of accessing to mineral nutrients when the fi rst land plants, thought to be rootless bryophytes, appeared (Brundrett, 2002). Th us, plants de- veloping associations with fungi might have a clear competitive advantage on the other land plants. Indeed, nowadays 100% of the gymnosperms, 80 % of the angiosperms and 70 % of the pteridophytes are associated with mycorrhizal fungi (de Boer et al., 2005). Furthermore, contemporary bryophytes often contain endophytic hyphae of mycorrhizal fungi, suggesting that the fi rst vascular plants were already colonized by mycorrhizal fungi (Brundrett, 2002). Knowing that colonization of land by plants is probably linked with the beginning of development of soils (de Boer et al., 2005), this fungal-plant association demonstrates the crucial role that fungi might have played in soils from the very beginning, and the enormous infl uence they still have on these ecosystems nowadays. Indeed, fi lamentous fungi are well-suited for the colonization of soils, because their hyphae are able to bridge the air-fi lled voids, thus colonizing water unsaturated soil patches (Wösten, 2001, Ritz & Young, 2004), and their fi lamentous network also allows fungi to participate to the stabilization of the soil structure, alongside with the change of soil properties in the surroundings of hyphae (Boswell et al., 2007).

When bacteria fi rst left aquatic environments, they encountered high pressures such as limited dispersion, and therefore scare resources and other environmental hazards (e.g. variable pH and tem- peratures; Reichenbach et al., 2007, Wu et al., 2014). It has been proposed that the cell wall modifi cation of Gram-positive bacteria (Actinobacteria and Firmicutes) has evolved in response to terrestrial conditions (Battistuzzi & Hedges, 2009). While Actinobacteria are important soil bacteria, phyla Proteobacteria, Acidobacteria and Verrucomicrobia, all Gram-negative, are also very abundant in soils (Zhang & Xu, 2008). Most of the bacteria belonging to these phyla are bearing a paralog polymerase, DnaE2, responsible for increasing GC content (Wu et al., 2014). It has been suggested that this DnaE2 polymerase has played a key role in the colonization of land by bacteria, and that bacteria from aquatic environments would have encountered a mutation from DnaE1 to DnaE2, allowing for GC increase and the expansion of the genome, essential for a successful colonization of land (Wu et al., 2014). Pathogenic and symbiotic bacteria generally do not possess DnaE2, and some of them might be

5 ancient terrestrial bacteria that lost DnaE2.

While mycorrhizal symbioses are considered as crucial for land plants, it has also been suggested that early land plants were also hosting diverse bacteria, such as vitamin-producers, nitrogen-fi xers and methanotrophs, which also certainly benefi ted the plant growth (Knack et al., 2015).

Nowadays, a high diversity of bacteria are coexisting in soils (Reichenbach et al., 2007), and they present multiple metabolic properties (Zhang & Xu, 2008). Most of the bacterial populations are either unculturable, or enter in a viable but unculturable state (Oliver, 2005, Zhang & Xu, 2008), and only the rising of molecular biology methods allowed for a more accurate estimation of bacterial taxonomic diversity in soils (Zhang & Xu, 2008).

Element cycling by microorganisms in soils

Carbon cycle In soils, the carbon cycle is strongly related to biological fi xation. It is estimated that the amount of soil organic carbon is three time higher than the amount of carbon stored in the atmosphere in the form of

CO2 (Gougoulias et al., 2014). As mentioned hereabove, it is the fi xation of atmospheric CO2, which is turned into organic compounds by autotrophic organisms – especially plants – that provides the most carbon in the soil. In addition to plants, bacteria also contribute to CO2 fi xation in soils, through phototropic and non-phototropic activities (Miltner et al., 2005).

Fungi and bacteria are the main decomposers of organic matter in soils (Ekschmitt et al., 2008, Rousk et al., 2009). In addition to the major role they play in the growth and prosperity of plants through mycorrhiza, fungi are also crucial actors for the decay of wood. Th e so-called white-rot fungi are usually considered as being the only organisms able to completely degrade lignin, the second most abundant component of wood after cellulose (Kirk & Farrell, 1987). While some bacteria are also able to degrade lignin, whose importance might be underestimated, the degradation processes seem to be slower than for fungi (Bugg et al., 2011). Furthermore, in temperate and boreal forests, litter decomposition is mainly driven by fungal activity (Hattenschwiler et al., 2005).

In most of the soils, the conditions are unsaturated and oxic, and the decomposition of organic matter by microorganisms leads to the formation of CO2 (Gougoulias et al., 2014). Soils can act as carbon sinks if the microbial decomposition rate decreases. Th is is the case in peatlands, in which carbon taken from the atmosphere by plants through photosynthesis is sequestered in the soil, because of the low decomposition rate (Gorham, 1991). While peatlands cover less than 3% of the world’s land surface, they contain 25 % of all the soil carbon of the world (Moore, 2002). Th ese numbers give an insight in the importance of the decomposition of organic matter triggered by microorganisms in all the other soils in the world.

While carbon sequestration in soils is usually attributed to organic carbon storage, the residence time of organic carbon in soils is lower than the residence time of mineral carbon, and thus precipitation of inorganic carbon such as calcium carbonate (CaCO3) in soils would represent a more effi cient carbon sink (Cailleau et al., 2004). Such a mineral carbon sink, following the so-called oxalate-carbo-

6 Figure 2. Schematic representation of the oxalate-carbonate pathway. a) In oxalotrophic plants, a part of the CO2 up- 2+ taken during photosynthesis is used for the production of oxalic acid (H2C2O4). Th e reaction with Ca forms calcium

oxalate (CaC2O4). b) Calcium oxalate is released to the soil through litter deposition and decay. c) In the soil, oxalotrophic bacteria are able to reach the calcium oxalate by taking the fungal highway. Th e consumption of calcium oxalate

leads to an increase of the soil pH. d) Th is increase of pH induces the precipitaion of calcium carbonate (CaCO3). e) Calcium is originating from bedrock and can be uptaken by plants and other organisms. f) Oxalotrophic plants release calcium oxalate in the soil through root exudates. g) Fungi can be producers of oxalic acid, which reacts with calcium to form calcium oxalate.

nate pathway, has been observed in tropical areas (Verrecchia et al., 2006), and is detailed hereunder (see also Figure 2).

Oxalic acid (H2C2O4) and oxalate minerals, especially calcium oxalate (CaC2O4), are produced by many plants as biosynthetic byproducts related to photosynthesis (Franceschi & Horner, 1980, Naka- ta, 2003). In tropical environments, the oxalogenic tree Milicia excelsa (Iroko tree) uses CO2 uptaken during photosynthesis for the production of calcium oxalate (Cailleau et al., 2011), and thus contains highly biomineralized tissues with calcium oxalate crystals. Secretion of oxalic acid by fungi appears to contribute to the precipitation of other forms of calcium oxalate. But surprisingly, oxalate minerals were never observed in the soil and litter surrounding these trees. Th is absence of oxalate minerals was explained by the presence of oxalotrophic bacteria, which are able to consume the oxalate salts and therefore, lead to an increase of the soil pH and the subsequent precipitation of calcium carbonate (Brais-

7 sant et al., 2004). In addition, it has been demonstrated that a dry season is necessary for an effi cient precipitation of calcium carbonate (Cailleau et al., 2005). Th is precipitation of calcium carbonate can be considered as a carbon sink, because its carbon source is organic (i.e. related to photosynthesis), and the calcium is not originating from a carbonated bedrock (Verrecchia et al., 2006, Cailleau et al., 2011).

Further research about the oxalate-carbonate pathway in mesocosms led to another intriguing observation: in soils amended with calcium oxalate and with oxalotrophic bacteria, the expected raise of pH, consequence of the consumption of calcium oxalate by bacteria, never occurred. But when fungi were added to the system, then the expected raise of pH occurred (Martin et al., 2012). While it was clear that fungi played a major role in this pathway, the nature of their contribution was unknown. In this study, it was hypothesized that the key contribution of fungi to this pathway could be linked to fungal highways (see last section of this general introduction).

Nitrogen cycle Th e nitrogen cycle is also strongly related to biological fi xation. Nitrogen is a very abundant element in the atmosphere (about 78 % of its volume; Atkins, 1992), but in the form of dinitrogen (N2), not available for most of the living organisms. Th us, despite of its abundance, nitrogen is often the limiting factor in terrestrial ecosystem productivity (Levy-Booth et al., 2014). In order to become bioavailable, dinitrogen needs to be fi xed, meaning that it has to be reduced into ammonia (NH3; Boyd & Pe- 2- 3- ters, 2013). Although N2 can also be turned abiotically into reactive NOx species (NO, NO , NO ) through lightning, generating 2.4 % of the global reactive nitrogen, this phenomenon is limited to the troposphere and has no signifi cant infl uence on terrestrial ecosystems (Fowler et al., 2013). It is commonly accepted that biological nitrogen fi xation is exclusively performed by prokaryotic organisms (bacteria and archaea; Leigh, 2000), and nitrogen fi xation in fungi remains unproven (Wainwright, 1988, Laere, 1995). Nitrogenase, the enzyme responsible for nitrogen fi xation, is inac- tivated by oxygen. A way for bacteria to fi x nitrogen in oxic environments is to associate with plants, which provide them microaerobic niches inside their roots (Boyd & Peters, 2013). Biological nitrogen fi xation is very important for agricultural systems, as several crops, and especially legumes, are naturally associated with nitrogen-fi xing bacteria, mostly belonging to the genus Rhizobium (Herridge et al., 2008). + While ammonium (NH4 ) can be used directly as nitrogen source by plants, their growth is usually - greater with nitrate (NO3 ) as nitrogen source (Andrews et al., 2013). Th e oxidation of ammonia to nitrite and nitrate (NH3  HNO2  HNO3), called nitrifi cation, is mainly performed by lithoau- totrophic bacteria. However, heterotrophic organisms - including fungi - are also able to perform nitri- + fi cation. It has been postulated that these organisms would be able to use inorganic pathways (NH4  - - NO2  NO3 ) or organic pathways (RNH2  RNO2  NO3; Koops & Pommerening-Röser 2001). Free amino acids can constitute an important fraction of dissolved nitrogen in soils, and they can be used directly by plants and microorganisms, which can be competing for this nitrogen resource (Chen 2015).

Phosphorus cycle Phosphorus is also one of the most important limiting factors for plant growth in soils (Ziadi et al., 2013). It is present in the bedrocks, but also in the soils, in association with metallic ions, such as

8 Ca, Fe or Al, depending of the soil type, and this phosphorus is not bioavailable (Ziadi et al., 2013, Bünemann, 2015). One small fraction of phosphorus is present in the soil as phosphates, which can be directly absorbed by plant or microbial cells (Ziadi et al., 2013). However, some bacteria and fungi are known for being able to solubilize fi xed phosphate and thus to make it bioavailable (Kucey, 1987, Nautiyal, 1999). Th e pool of organic phosphorus in the soils consists of dead microbial cells and plant debris as well as phosphorus associated with organic matter. Th e enzyme phosphatase allows plants and microorganisms to catalyze the conversion of organic phosphorus into phosphate. As well, the secretion of acids such as citrate, oxalate and malate are thought to solubilize fi xed organic phosphorus (Dodd et al., 2015).

Iron cycle Iron is also often a limiting factor for the growth of plants, but also for microorganisms, because of the insolubility of iron oxides. Iron can be bound to the mineral fraction of the soil, but also to organic matter in soluble or insoluble forms (Mimmo et al., 2014). Many bacteria, as well as plants and fungi, are able to produce siderophores, which are low-molecular mass compounds able to chelate metals, such as Fe, Mo, Mn, Co or Ni, and thus make them available for microbial cells (Ahmed & Holmstrom, 2014). It is also known that microbial siderophores can provide plants with Fe (Crowley, 2006). Th e production of siderophores in soils allows for mineral dissolution of insoluble phases, and thus is implied in the dissolution of rocks, but also in weathering processes (Ahmed & Holmstrom, 2014). However, other microorganisms are able to obtain iron by reduction and acidifi cation at the cell surface, and taking the iron up before it has been re-oxidized (Gadd & Raven, 2010).

Bacteria are also playing a role in the cycling of other nutrients. For example, they are able to solubilize potassium from mineral, thus increasing the availability for plants (Miransari, 2013). Furthermore, some bacteria are able to oxidize sulfuric compounds into sulfates, thus lowering the soil pH, which can have great infl uences in calcareous (i.e. alkaline) soils, because the uptake of most nutrients by plants is optimal at neutral pH (Miransari, 2013).

Fungal-bacterial interactions in soils

Traditionally, mycologists and bacteriologists have conducted their studies separately (Strickland & Rousk, 2010), even though the sharing of a common habitat and common resources has led to numerous and diverse interactions between fungi and bacteria, from mutualism to parasitism (Kobayashi & Crouch, 2009). Physical association between fungi and bacteria can range from disordered polymi- crobial communities, to mixed fungal-bacterial biofi lms, to symbiotic relationships (Frey-Klett et al., 2011). Many reports have been done on the observation of intrahyphal bacteria (Ibrahim et al., 2008, Hoff man & Arnold, 2010, Sato et al., 2010, Desiro et al., 2014). Alongside with physical associations, fungi and bacteria can also interact and communicate via chemical compounds (Frey-Klett et al., 2011). Secretion of antimicrobial substances (antibiotics and toxins) is widespread amongst fungi and bacteria, and contributes to the structure and diversity of microbial communities (Czaran et al., 2002). Signaling molecules, as well as volatile organic compounds, also play an important role in communi-

9 cation between fungi and bacteria (Schmidt et al., 2015). Trophic interactions, from competition for nutrients such as carbon, nitrogen, and iron, to cooperative behaviors, such as feeding on exudates, as well as predator behaviors, also occur between fungi and bacteria (Frey-Klett et al., 2011).

About fungal highways

As already mentioned, the fi lamentous growth of fungi is considered as an advantage in the colonization of soils. In addition to the ability of their hyphae to cross air-fi lled gaps, fungi are modular organisms: hyphae that are separated from the mycelium can grow independently, and the genetic individual can consist in several physiological individuals (Carlile, 1995). Th us, the decay of one part of the mycelium will not aff ect the genetic individual. In addition, as fungi have the ability to translocate elements, it allows for reduced activity in older hyphae (Prosser, 1995), and facilitates the crossing of nutrient-poor areas (de Boer et al., 2005). While an important group of soil bacteria, the actinomycetes, are in a fi lamentous shape (de Boer et al., 2005), most of the bacteria are single-celled, and thus bacterial motility is usually considered as very limited in soils (Or et al., 2007, Ekschmitt et al., 2008). Indeed, most of the bacterial motility types request a liquid layer in order to take place (Harshey, 2003), and thus they are not able to bridge the air-fi lled gaps of soils (de Boer et al., 2005). Alongside with chemical warfare (Czaran et al., 2002), limited dispersal has been suggested as the explanation of the co-existence of a high diversity of bacteria in soils (Reichenbach et al., 2007). Doubt is casted on this statement with the observation of bacterial movement along fungal hyphae (Wong & Griffi n, 1976). Th is interaction is called fungal highways, and it has been shown that it could enhance bacterial dispersion in unsaturated porous media (Kohlmeier et al., 2005). Because of the apical growth of fungi, only motile bacteria should be able to take the fungal highway. It has been hypothesized that these bacteria would be able to swim in the liquid fi lm surrounding hydrophilic hyphae, and that bacterial movement along hydrophobic hyphae would be less eff ective (Kohlmeier et al., 2005). It has also been demonstrated that, when associated to motile bacteria, some non-motile bacteria were also able to migrate along fungal hyphae. Th ese bacteria were called hitchhikers (Warmink et al., 2011). In vitro studies have highlighted potential ecological benefi ts of this dispersal mechanism. For example, an improved biodegradation of soil pollutants has been demonstrated in soil microcosm experiments (Wick et al., 2007), with a chemotactic movement of bacteria toward nutrient sources (Furuno et al., 2010). Mycelium-driven bacteria dispersal has also been suggested as a mechanism contributing to the explanation of the maintenance of fl agella in soil bacteria (Pion et al., 2013). It has been hypothesized that the key role of fungi in the oxalate-carbonate pathway would be to create paths for bacterial dispersion, and thus oxalotrophic bacteria could reach the spots containing calcium oxalate, which would otherwise be inaccessible (Martin et al., 2012).

Objectives of this thesis

Th e principal goals of the present thesis were to better understand fungal highway associations, their implication in the oxalate-carbonate pathway, and to identify the organisms involved.

10 When this thesis started, fungal highways had been only studied in vitro, with artifi cial fungal-bacterial couples (i.e. there was no direct evidence showing that the tested couples were truly interacting in natural ecosystems). Th erefore, we wanted to fi nd a way to collect fungi and bacteria interacting through fungal highways directly from the soil. For this purpose, we developed and validated a novel device, called fungal highway column, which is presented in Chapter 1. As well, this chapter presents a fi rst fi eld research performed with fungal highway columns in the context of the oxalate-carbonate pathway, in a soil under an oxalogenic plant in Morocco. In this research, we focused on a cultivable approach. A general description of the isolated organisms, as well as their migratory abilities, is also presented.

In Chapter 2, we present another fi eld research with fungal highway columns. In this research, we conducted a non-culturable approach, and our aim was to identify interacting fungi and bacteria, and to assess whether associates of fungal highways would diff er according to soil properties such as carbon and nitrogen contents, and C:N ratio.

Chapter 3 is linked to metabolic activities of the organisms isolated during the fi eld work in Mo- rocco presented in Chapter 1. Th us we investigated the ability of the isolated fungi and bacteria to solubilize calcium oxalate (Chapter 3a), to grow in nitrogen-depleted environments (Chapter 3b), to solubilize inorganic phosphorus, and to produce siderophores (Chapter 3c), associated with bacterial migratory abilities.

Each chapter of the present thesis includes: • A foreword, which presents the topic. • A manuscript, which presents the main article linked to the topic (published, submitted or in preparation). • Supplementary information related to the manuscript. • If necessary, additional experiments. • A summary, which presents the main results and perspectives of the chapter.

Acknowledgements

I would like to thank Dylan Tatti (Laboratory of Functional Ecology, University of Neuchâtel, Switzer- land) for his help regarding the complexity of the evolution of soils in time and space.

References

Ahmed E & Holmstrom SJ (2014) Siderophores in environmental research: roles and applications. Microb Biotechnol 7: 196-208. Andrews M, Raven JA & Lea PJ (2013) Do plants need nitrate? Th e mechanisms by which nitrogen form aff ects plants. Ann Appl Biol 163: 174–199. Atkins P (1992) Chimie générale. InterEditions, Bologna, Italy. Battistuzzi FU & Hedges SB (2009) A major clade of prokaryotes with ancient adaptations to life on land. Mol Biol Evol 26: 335-343.

11 Boswell GP, Jacobs H, Ritz K, Gadd GM & Davidson FA (2007) Th e development of fungal networks in complex environments. Bull Math Biol 69: 605-634. Boyd ES & Peters JW (2013) New insights into the evolutionary history of biological nitrogen fi xation. Front Microbiol 4: 201. Braissant O, Cailleau G, Aragno M & Verrecchia EP (2004) Biologically induced mineralization in the tree Milicia excelsa (Moraceae): its causes and consequences to the environment. Geobiology 2: 59-66. Brundrett MC (2002) Coevolution of roots and mycorrhizas of land plants. New Phytol 154: 275-304. Bugg TD, Ahmad M, Hardiman EM & Rahmanpour R (2011) Pathways for degradation of lignin in bacteria and fungi. Nat Prod Rep 28: 1883-1896. Bünemann EK (2015) Assessment of gross and net mineralization rates of soil organic phosphorus – A review. Soil Biol Biochem 89: 82-98. Cailleau G, Braissant O & Verrecchia EP (2004) Biomineralization in plants as a long-term carbon sink. Naturwissenschaften 91: 191-194. Cailleau G, Braissant O & Verrecchia EP (2011) Turning sunlight into stone: the oxalate-carbonate pathway in a tropical tree ecosystem. Biogeosciences 8: 1755-1767. Cailleau G, Braissant O, Dupraz C, Aragno M & Verrecchia EP (2005) Biologically induced accumu-

lations of CaCO3 in orthox soils of Biga, Ivory Coast. Catena 59: 1-17. Carlile MJ (1995) Th e success of the hypha and mycelium. In Gow NAR & Gadd GM, eds. Th e Growing Fungus. Chapman & Hall, London, UK. Chen J, Carrillo Y, Pendall E et al. (2015) Soil microbes compete strongly with plants for soil inorganic

and amino acid nitrogen in a semiarid grassland exposed to elevated CO2 and warming. Ecosystems 18: 867-880. Conférence intercantonale de l’instruction publique de la Suisse romande et du Tessin (2016) Plan d’études romand. www.plandetudes.ch. Last date accessed January 2016. Crowley DA (2006) Microbial siderophores in the plant rhizosphere. In Barton LL & Abadía J, eds. Iron nutrition in plants and rhizospheric microorganisms. Springer, Dordrecht, Th e Netherlands. Czaran TL, Hoekstra RF & Pagie L (2002) Chemical warfare between microbes promotes biodiversity. Proc Natl Acad Sci U S A 99: 786-790. de Boer W, Folman LB, Summerbell RC & Boddy L (2005) Living in a fungal world: impact of fungi on soil bacterial niche development. FEMS Microbiol Rev 29: 795-811. Deacon JW (2006) Fungal Biology, 4th edition. Blackwell Publishing Ltd, Oxford, UK. Desiro A, Salvioli A, Ngonkeu EL et al. (2014) Detection of a novel intracellular microbiome hosted in arbuscular mycorrhizal fungi. ISME J 8: 257-270. Dodd RJ & Sharpley AN (2015) Recognizing the role of soil organic phosphorus in soil fertility and water quality. Resour Conserv Recy 105: 282–293. Ekschmitt K, Kandeler E, Poll C et al. (2008) Soil-carbon preservation through habitat constraints and biological limitations on decomposer activity. J Plant Nutr Soil Sci 171: 27-35. Food and Agriculture Organizaton of the United Nations (2016) www.fao.org. Last date accessed Ja- nuary 2016. Fowler D, Coyle M, Skiba U et al. (2013) Th e global nitrogen cycle in the twenty-fi rst century. Philos Trans R Soc Lond B Biol Sci 368: 20130164. Franceschi VR & Horner HTJ (1980) Oxalate crystals in plants. Bot Rev 46: 361-427. Frey-Klett P, Burlinson P, Deveau A, Barret M, Tarkka M & Sarniguet A (2011) Bacterial-fungal interactions: hyphens between agricultural, clinical, environmental, and food microbiologists. Micro-

12 biol Mol Biol Rev 75: 583-609. Furuno S, Päzolt K, Rabe C, Neu TR, Harms H, Wick LY (2010) Fungal mycelia allow chemotactic dispersal of polycyclic aromatic hydrocarbon-degrading bacteria in water-unsaturated systems. En- viron Microbiol 12: 1391–1398. Gadd GM & Raven JA (2010) Geomicrobiology of eukaryotic microorganisms. Geomicrobiol J 27: 491-519. Gobat J-M, Aragno M & Matthey W (2003) Le sol vivant, deuxième édition revue et augmentée. Presses polytechniques et universitaires romandes, Lausanne, Switzerland. Gorham E (1991) Northern peatlands: Role in the carbon cycle and probable responses to climatic warming. Ecol Appl 1: 182-195. Gougoulias C, Clark JM & Shaw LJ (2014) Th e role of soil microbes in the global carbon cycle: tracking the below-ground microbial processing of plant-derived carbon for manipulating carbon dynamics in agricultural systems. J Sci Food Agric 94: 2362-2371. Gow NAR (1995) Tip growth and polarity. In Gow NAR & Gadd GM, eds. Th e growing fungus. Chapman & Hall, London, UK. Harshey RM (2003) Bacterial motility on a surface: many ways to a common goal. Annu Rev Micro- biol 57: 249-273. Hattenschwiler S, Tiunov AV & Scheu S (2005) Biodiversity and litter decomposition in terrestrial ecosystems. Annu Rev Ecol Evol Syst 36: 191-218. Herridge DF, Peoples MB & Boddey RM (2008) Global inputs of biological nitrogen fi xation in agricultural systems. Plant Soil 311: 1-18. Hoff man MT & Arnold AE (2010) Diverse bacteria inhabit living hyphae of phylogenetically diverse fungal endophytes. Appl Environ Microbiol 76: 4063-4075. Ibrahim AS, Gebremariam T, Liu M et al. (2008) Bacterial endosymbiosis is widely present among zygomycetes but does not contribute to the pathogenesis of mucormycosis. J Infect Dis 198: 1083- 1090. Jenny H (1941) Factors of soil formation, a system of quantitative pedology. Constable and Company, London, UK. Kirk TK & Farrell RL (1987) Enzymatic «combustion»: the microbial degradation of lignin. Annu Rev Microbiol 41: 465-505. Knack JJ, Wilcox LW, Delaux PM et al. (2015) Microbiomes of streptophyte algae and bryophytes suggest that a functional suite of microbiota fostered plant colonization of land. Int J Plant Sci 176: 405-420. Kobayashi DY & Crouch JA (2009) Bacterial/fungal interactions: From pathogens to mutualistic endosymbionts. Annu Rev Phytopathol 47: 63-82. Kohlmeier S, Smits TH, Ford R, Keel C, Harms H & Wick LY (2005) Taking the fungal highway: Mobilization of pollutant-degrading bacteria by fungi. Environ Sci Technol 39: 4640-4646. Koops HP & Pommerening-Röser A (2001) Distribution and ecophysiology of the nitrifying bacteria emphasizing cultured species. FEMS Microbiol Ecol 37: 1-9. Kucey RMN (1987) Increased phosphorus uptake by wheat and fi eld beans inoculated with a phosphorus-solubilizing Penicillium bilaji strain and with vesicular-arbuscular mycorrhizal fungi. Appl Environ Microbiol 53: 2699-2703. Laere AV (1995) Intermediary metabolism. In Gow NAR & Gadd GM, eds. Th e growing fungus. Chapman & Hall, London, UK. Laliberte E, Grace JB, Huston MA, et al. (2013) How does pedogenesis drive plant diversity? Trends

13 Ecol Evol 28: 331-340. Leigh JA (2000) Nitrogen fi xation in methanogens: the archaeal perspective. Curr Issues Mol Biol 2: 125-131. Levy-Booth DJ, Prescott CE & Grayston SJ (2014) Microbial functional genes involved in nitrogen fi xation, nitrifi cation and denitrifi cation in forest ecosystems. Soil Biol Biochem 75: 11-25. Markham P (1995) Organelles of fi lamentous fungi. In Gow NA & Gadd GM, eds. Th e growing fungus. Chapman & Hall, London, UK. Martin G, Guggiari M, Bravo D et al. (2012) Fungi, bacteria and soil pH: the oxalate-carbonate pathway as a model for metabolic interaction. Environ Microbiol 14: 2960-2970. Miltner A, Kopinke F-D, Kindler R, Selesi D, Hartmann A & Kästner M (2005) Non-phototrophic

CO2 fi xation by soil microorganisms. Plant and Soil 269: 193-203. Mimmo T, Del Buono D, Terzano R et al. (2014) Rhizospheric organic compounds in the soil-microorganism-plant system: their role in iron availability. Eur JSoil Sci 65: 629–642. Minasny B, McBratney AB & Salvador-Blanes S (2008) Quantitative models for pedogenesis - A review. Geoderma 144: 140-157. Miransari M (2013) Soil microbes and the availability of soil nutrients. Acta Physiol Plant 35: 3075- 3084. Moore PD (2002) Th e future of cool temperate bogs. Environ Conserv 29. Mueller B (2015) Experimental interactions between clay minerals and bacteria: A review. Pedosphere 25: 799-810. Nakata PA (2003) Advances in our understanding of calcium oxalate crystal formation and function in plants. Plant Science 164: 901-909. Nautiyal CS (1999) An effi cient microbiological growth medium for screening phosphate solubilizing microorganisms. FEMS Microbio Lett 170: 265-270. Oliver JD (2005) Th e viable but nonculturable state in bacteria. J Microbiol 43: 93-100. Or D, Smets BF, Wraith JM, Dechesne A & Friedman SP (2007) Physical constraints aff ecting bacterial habitats and activity in unsaturated porous media – a review. Ad Water Resour 30: 1505-1527. Pion M, Bshary R, Bindschedler S, et al. (2013) Gains of bacterial fl agellar motility in a fungal world. Appl Environ Microbiol 79: 6862–7. Prosser JI (1995) Kinetics of fi lamentous growth and branching. In Gow NA & Gadd GM, eds. Th e growing fungus. Chapman & Hall, London, UK. Reichenbach T, Mobilia M & Frey E (2007) Mobility promotes and jeopardizes biodiversity in rock-pa- per-scissors games. Nature 448: 1046-1049. Ritz K & Young IM (2004) Interactions between soil structure and fungi. Mycologist 18: 52-59. Rousk J, Brookes PC & Baath E (2009) Contrasting soil pH eff ects on fungal and bacterial growth suggest functional redundancy in carbon mineralization. Appl Environ Microbiol 75: 1589-1596. Sato Y, Narisawa K, Tsuruta K et al. (2010) Detection of Betaproteobacteria inside the mycelium of the fungus Mortierella elongata. Microb Environ 25: 321-324. Schmidt R, Etalo DW, de Jager V, et al. (2015) Microbial small talk: Volatiles in fungal-bacterial interactions. Front Microbiol 6: 1495. Schütte KH (1956) Translocation in the fungi. New Phytol 55: 164-182. Strickland MS & Rousk J (2010) Considering fungal:bacterial dominance in soils – Methods, controls, and ecosystem implications. Soil Biol Biochem 42: 1385-1395. Tedersoo L, Bahram M, Põlme S et al. (2014) Global diversity and geography of soil fungi. Science 346.

14 Th eng BKG (1982) Clay-polymer interactions: summary and perspectives. Clays Clay Miner 30: 1-10. van der Heijden MG, Martin FM, Selosse MA & Sanders IR (2015) Mycorrhizal ecology and evolution: the past, the present, and the future. New Phytol 205: 1406-1423. Verrecchia EP, Braissant O & Cailleau G (2006) Th e oxalate-carbonate pathway in soil carbone storage: the role of fungi and oxalotrophic bacteria. Fungi in Biochemical Cycles 12: 289-310. Wainwright M (1988) Metabolic diversity of fungi in relation to growth and mineral cycling in soil - a review. Trans Br Mycol Soc 90: 159-170. Warmink JA, Nazir R, Corten B & van Elsas JD (2011) Hitchhikers on the fungal highway: Th e helper eff ect for bacterial migration via fungal hyphae. Soil Biol Biochem 43: 760-765. Wong PTW & Griffi n DM (1976) Bacterial movement at high matric potentials - II. In fungal colonies. Soil Biol Biochem 8: 219-223. Wick LY, Remer R, Würz B, et al. (2007) Eff ect of fungal hyphae on the access of bacteria to phenanthrene in soil. Environ Sci Technol 41: 500-505. Wösten HAB (2001) Hydrophobins: Multipurpose proteins. Annu Rev Microbiol 55: 625-646. Wu H, Fang Y, Yu J & Zhang Z (2014) Th e quest for a unifi ed view of bacterial land colonization. ISME J 8: 1358-1369. Zhang L & Xu Z (2008) Assessing bacterial diversity in soil. J Soils Sediments 8: 379-388. Ziadi N, Whalen JK, Messiga AJ & Morel C (2013) Assessment and modeling of soil available phosphorus in sustainable cropping systems. Adv Agron 122: 85-126.

Chapter 1

Fungal highway columns

Chapter 1 - Fungal highway columns

Foreword

At the begining of this thesis, observations of fungal highways had only been done in vitro, with artifi cial fungal-bacterial couples (i.e. there was no direct evidence showing that the tested couples were truly interacting in natural ecosystems). Th us, my main interest at this point was to fi nd out whether these associations were eff ectively present in natural environments. For this purpose, I needed to develop a novel device, which would allow for an in situ collection of organisms able to interact via fungal highways. Th erefore, the fi rst step of this thesis consisted in the developement of this device we called fungal highway column. Th e most challenging part of this process was to fi nd a way to select only for bacteria able to disperse along fungal hyphae, and to avoid the presence of other bacteria. As well, this device had to be attractive for fungi.

Th is work led to the publication of an article, which presents the validation tests of these columns, as well as a fi rst fi eld test and the organisms isolated form a Moroccan soil related to the oxalate-carbonate pathway.

Fungal highway columns are central for this thesis, and most of its content is directly related to the columns.

A view from the sampling area, a shepherd and the Atlas Mountains, Morocco, 2012.

Th e one month fi eld work in Sidi-Bennour, Morocco, gave me the opportunity of conducting the process of writing a grant application, as well as to accompany a master student on his fi eld research. Moreover, I discovered a beautiful country and I had the opportunity to meet and discuss with numerous interesting people.

In addition to this fi eld work, diverse analyses were performed at the Department of Environ- mental Microbiology in UFZ Leipzig, Germany, where Dr. Lukas Wick and his team welcomed me during two weeks.

Chapter 1 - Fungal highway columns

Manuscript

Th is section presents the following manuscript, published in FEMS Microbiology Ecology:

Simon A, Bindschedler S, Job D, Wick L, Filippidou S, Kooli W, Verrecchia E, Junier P. 2015. Ex- ploiting the fungal highway: Development of a novel tool for the in situ isolation of bacteria migrating along fungal mycelium. FEMS Microbiology Ecology 91 (11).

Th is article has been selected as the Editor’s Choice article for issue 91.11 of FEMS Microbiology Ecology.

FEMS Microbiology Ecology, 91, 2015, fiv116

doi: 10.1093/femsec/fiv116 Advance Access Publication Date: 2 October 2015 Research Article

RESEARCH ARTICLE Exploiting the fungal highway: development of a novel tool for the in situ isolation of bacteria migrating along fungal mycelium Anaele Simon1, Saskia Bindschedler2,†,DanielJob1,LukasY.Wick2, Sevasti Filippidou1, Wafa M. Kooli1,EricP.Verrecchia3 and Pilar Junier1,∗

1Laboratory of Microbiology, Institute of Biology, University of Neuchatel,ˆ 2000 Neuchatel,ˆ Switzerland, 2Department of Environmental Microbiology, Helmholtz Centre for Environmental Research – UFZ, 04318 Leipzig, Germany and 3Biogeosciences laboratory, Institute of Earth Surface Dynamics, University of Lausanne, 1015 Lausanne, Switzerland

∗Corresponding author: Laboratory of Microbiology, Institute of Biology, University of Neuchatel,ˆ 2000 Neuchatel,ˆ Switzerland. Tel: +41 32 718 22 44; E-mail: [email protected] †Present address: Laboratory of Microbiology, Institute of Biology, University of Neuchatel,ˆ 2000 Neuchatel,ˆ Switzerland. One sentence summary: A novel method to enrich bacteria moving along fungal mycelium. Editor: Wietse de Boer

ABSTRACT

Fungi and bacteria form various associations that are central to numerous environmental processes. In the so-called fungal highway, bacteria disperse along fungal mycelium. We developed a novel tool for the in situ isolation of bacteria moving along fungal hyphae as well as for the recovery of fungi potentially involved in dispersal, both of which are attracted towards a target culture medium. We present the validation and the results of the first in situ test. Couples of fungi and bacteria were isolated from soil. Amongst the enriched organisms, we identified several species of fast-growing fungi (Fusarium sp. and Chaetomium sp.), as well as various potentially associated bacterial groups, including Variovorax soli, Olivibacter soli, Acinetobacter calcoaceticus, and several species of the genera Stenotrophomonas, Achromobacter and Ochrobactrum. Migration of bacteria along fungal hyphae across a discontinuous medium was confirmed in most of the cases. Although the majority of the bacteria for which migration was confirmed were also positive for flagellar motility, not all motile bacteria dispersed using their potential fungal partner. In addition, the importance of hydrophobicity of the fungal mycelial surface was confirmed. Future applications of the columns include targeting different typesof microorganisms and their interactions, either by enrichment or by state of the art molecular biological methods.

Keywords: soil; fungal highways; bacteria; fungi; enrichment; columns

INTRODUCTION Thus, in water-unsaturated soils, active movement of bacteria is strongly limited, principally due to the discontinuity of Many bacteria are able to move by swimming, swarming, twitch- water films (Or et al. 2007). Nevertheless, in vitro experiments ing, sliding, or gliding without pili (Biais 2009;Sunet al. 2011). performed in water-unsaturated media have demonstrated that However, except for the last, these bacterial motility types are fungal mycelia can be used as paths for the active dispersal of only efficient in the presence of a liquid film (Harshey 2003).

23 2 FEMS Microbiology Ecology, 2015, Vol. 91, No. 11

bacteria. Bacteria are shown to be moving in the synaeretic film usefulness of the columns as an easy-to-use tool to advance in formed around the hyphae, which act as fungal highways, as the study of fungal–bacterial interactions in soils. termed by Kohlmeier et al. (2005). Knowing that there can be up to 20 000 km of hyphae per cu- bic meter of soil (Moore, Robson and Trinci 2011), and that fungal MATERIALS AND METHODS hyphae can colonize both water-saturated and air-filled voids Design of a fungal highway column between soil particles (Wosten¨ 2001), mycelium-driven transport might play an important role in regulating the dispersal of The fungal highway columns were designed as a completely bacteria in unsaturated soils. Recent in vitro studies have high- closed device (Ø: 15 mm, height: ∼48 mm) intended to isolate lighted ecological benefits of this dispersal mechanism. For ex- bacteria able to disperse along fungal mycelium towards a tar- ample, an improved biodegradation of soil pollutants has been get culture medium, which can be used for enrichment and iso- demonstrated in soil microcosm experiments (Kohlmeier et al. lation of both bacteria and fungi. 2005;Wicket al. 2007;Banitzet al. 2013). Mycelium-driven bac- Unless otherwise stated, each component was sterilized by terial dispersal has also been suggested as a mechanism con- immersion in 70% ethanol for a duration of 30 min and expo- tributing to the explanation of the maintenance of flagella in soil sure to UV-C light for 30 min before assembling. Culture media ◦ bacteria (Pion et al. 2013; Lee et al. 2014). Likewise, promotion and glass beads were sterilized by autoclaving (21 min at 121 C). of long-term carbon storage and soil fertility through fungal– Each column was assembled under a sterile laminar flow hood, bacterial interactions in the context of the oxalate–carbonate as follows (Fig. 1). A 10-mL polystyrene tube (Fig. 1a; Semadeni, pathway has been shown in soil microcosms (Martin et al. 2012). Ostermundigen, Switzerland), Ø ∼ 15 mm, was cut 4 cm from the However, in spite of its potential ecological importance, ev- top. The middle part of this 4-cm tube was heated with a flame idence of active bacterial movement along mycelium in soils and twisted manually in order to leave an inside aperture of Ø ∼ is still missing. Various attempts have been made to isolate 2 mm; a pierced polyethylene cap (Fig. 1b; Semadeni) was used to selectively soil bacteria moving along hyphae. In a first ap- hold a pre-cut agar-based medium (Fig. 1c; attracting medium) proach, a microcosm model based on a compartmentalized Petri inside the bottom part of the twisted tube to serve as an attract- dish in which Lyophyllum sp. strain Karsten was pre-inoculated ing medium, with the help of a 3 × 3 cm square of a 25-mm allowed for the identification of several bacterial species mesh opening tissue (Fig. 1d; Sefar, Heiden, Switzerland) placed potentially migrating along the hyphae (Warmink, Nazir and van on top of the hollow cap. The attracting medium encompassed Elsas 2009). More recently, mycelium of Pythium ultimum was ar- the diameter of the tube but was irregular, leaving some space to tificially inoculated in soil to be used as a path for theisola- allow movement of fungi or bacteria at the edges. A polyethylene tion of contaminant-degrading bacteria (Furuno et al. 2012)us- cap (Fig. 1e) was used to close the bottom of the column; glass ing an inverted Petri dish method. Following the same idea, a beads (Fig. 1f; Ø 1 mm; Glaswarenfabrik Karl Hecht, Sondheim system without pre-inoculation of filamentous microorganisms v. d. Rhon,¨ Germany) were used to fill the tube, leaving ∼1cm was used to show the dispersal of soil oxalotrophic bacteria on free space at the top of it; a second slice of agar-based medium fungal mycelium ex situ (Bravo et al. 2013). (Fig. 1g; target medium) was placed inside the upper part of the The aim of this study was to develop a method for in situ twisted tube, on top of the glass beads. The top of the tube was collection of both bacterial and fungal strains potentially in- closed with a polyethylene cap (Fig. 1h; Ø 10.5 mm; Semadeni) teracting via fungal highways in natural environments such as and held with Parafilm R (Bemis, Oshkosh, USA). The excess tis- soils. We here describe the design and the validation of this sue was cut with sterile scissors and covered with Parafilm R . method, a novel centimeter-scale device, hereafter called the Colored tape was placed on the tube for labeling. ‘fungal highway column’. The applicability of the system was The basic procedure to use a fungal highway column follows evaluated with microcosm experiments in the laboratory. A fur- these steps (Fig. 2): the bottom cap is removed and the column ther validation step was performed in a field experiment with a is placed on a substrate (e.g. soil or a culture of microorganisms soil under the influence of the oxalate–carbonate pathway. This on solid medium) to be left in place for a duration of several days type of soil was selected because fungal-driven bacterial disper- (see validation below). After removal from the substrate, the col- sal seems to be relevant for oxalatrophic bacterial activity in soil umn is immediately closed with another bottom cap (previously under the influence of the oxalate–carbonate pathway (Martin sterilized in 70% ethanol). Directly, or after several days of in- et al. 2012;Bravoet al. 2013). From the field validation, easily cul- cubation (see validation below), the cap on the top of the col- tivable fungi and bacteria recovered simultaneously in fungal umn is removed in a sterile environment (e.g. a sterile laminar highway columns were identified. For the characterization of the flow hood), and the target culture medium is plated on acul- strains, we considered features displayed by bacteria and fungi ture medium for the isolation of cultivable organisms or, alter- described as part of fungal highway interactions. Previous stud- natively, it can be used directly to extract nucleic acids. ies have postulated that bacteria swim actively inside the liquid film surrounding hydrophilic hyphae (Kohlmeier et al. 2005), Laboratory validation tests and therefore bacterial motility and fungal hydrophobicity were tested. Additionally, as positive selection of bacteria harboring A first validation test to evaluate bacterial dispersal in the the Type III secretion system (TTSS) has been proposed on the presence and absence of a fungal carrier was conducted with mycosphere (Warmink and van Elsas 2008), and bacterial migra- the flagellated bacterium Pseudomonas putida KT2440 producing tion along fungal hyphae has shown a strong positive correlation green fluorescent protein (GFP) constitutively (Table 1). This bac- to the presence of genes encoding TTSS subunits (Warmink and terium is known to disperse along hyphae of Morchella crassipes van Elsas 2009), the presence of a gene marker for TTSS was as- and Trichoderma sp. (Pion et al. 2013). P.putida was inoculated with sessed in the bacterial isolates. Finally, as the isolated organisms M. crassipes or Trichoderma sp. in Petri dishes of malt agar (MA) constitute only potential couples of bacteria and fungi interact- medium, composed of 12 g·L−1 malt (Mycotec SA, La Chaux-de- ing via fungal highways, the ability of each bacterium to migrate Fonds, Switzerland) and 15 g·L−1 agar (Biolife Italiana S.r.l, Mi- along its associated fungus was tested. Our results underpin the lano, Italy). As a control in the absence of a fungal network,

24 Simon et al. 3

Figure 1. Components and general procedure for assembling a fungal highway column. (a) Polystyrene tube cut at length of 4 cm, heated with a flame and twisted. (b) Pierced polyethylene cap. (c) A slice of agar-based medium (the attracting culture medium) placed at the bottom of the tube. (d) The tube is placed on a 25-mm mesh opening tissue, wedging it between the hollow cap and the tube. (e) A polyethylene cap closing the bottom of the column. (f) Glass beads (Ø: 1 mm) filling the tube.g ( ) A second slice of agar-based medium (the target culture medium), placed on top of the glass beads. (h) A polyethylene cap closing the top of the tube.

P. putida was inoculated alone on the same medium. Both ex- tagged P. putida was assessed by fluorescence microscopy (Nikon periments were conducted in triplicate. A slice of each inocu- Corp., Tokyo, Japan). Presence of fungi was assessed by visual lated MA medium was placed at the bottom of a fungal high- inspection. way column and left there to incubate for 6 days. After this in- A second validation test was performed in soil-containing cubation time, the target culture medium of each column was microcosms (Table 1). A microcosm consisted of a polyethy- cut into four pieces. Two pieces were directly transferred onto lene container with 12.5 ± 0.5 g of soil, autoclaved (21 min at a plate of MA medium. Two pieces were put in 1 mL of sterile 121◦C), dried overnight at 60◦C, and watered with 5 mL of ster- physiological water (9 g·L−1 NaCl; Sigma-Aldrich, St Louis, MO, ile deionized water after drying (final water content of 40% of USA) and dispersed by shaking. From this, 100 μL was removed the soil water holding capacity). We used a fungal highway col- and streaked on a plate of nutrient agar (NA) medium, composed umn containing 6 g·L−1 malt MA as attracting and target me- of8g·L−1 nutrient broth (Biolife Italiana S.r.l, Milano, Italy) and dia for each microcosm. The microcosm tests were performed 15 g·L−1 agar (Fig. 3). Plates were incubated for 24 h at room for the assessment of three properties: absence of bacterial temperature. After incubation, presence of colonies of GFP- or fungal contamination in the columns, absence of bacterial

25 4 FEMS Microbiology Ecology, 2015, Vol. 91, No. 11

Figure 2. Description of the steps involved in the use of fungal highway columns. In step 2, soil can be replaced by a culture of microorganisms on solid medium or any other substrate. dispersal in soil without fungi, and bacterial dispersal in the way columns during field use and the effect of transportation presence of fungi. The organisms artificially inoculated in the shaking. The hollow cap at the bottom of three fungal highway soil were the same flagellated bacterium P. putida KT2440 and the columns was filled with a sterile soil amended with P. putida (as fungi M. crassipesandFusarium oxysporum. The fungal highway indicated above) before closing the column with the bottom cap. columns were placed in the soil for a duration of 4 or 10 days, Each column was incubated during 3 days at room temperature. and left for incubation after removal from the soil for a duration After this time, columns were stored for 15 min at 4◦C and left of 0, 8, or 15 days. Each condition was tested in triplicate. again at room temperature. After water condensation had oc- For the soil inoculated with P. putida, cells from an overnight curred inside each column, they were shaken by hand to mimic culture of P. putida on NA medium were collected in sterile phys- random disturbances as likely to happen during transportation iological water, to a final OD550 of 3, which corresponded to 25 × from the field site to the laboratory. The columns were thenin- 106 cells mL−1, of which 100 μL was added to the soil and mixed cubated horizontally at room temperature for 24 h, bringing the with it with a sterile loop. The fungal inoculum was a 7-day-old total time of incubation to 4 days. Target culture medium was (M. crassipes) or 3-day-old (F. oxysporum) culture on MA medium. plated on NA medium (as explained above), and presence of bac- From this culture, two squares (∼3mm3) were cut out and mixed terial colonies was assessed by visual inspection after 24 h. with the soil with a sterile loop. An individual fungal highway The same experiment was performed with P. putida inocu- column was placed in each microcosm for the time indicated in lated in soil-containing microcosms (as described above). An in- Table 1. After being removed from the soil, each fungal highway dividual fungal highway column was placed in each microcosm column was closed and left for incubation as indicated in Table 1. for 2 days, in triplicate. After this time, columns were stored After this incubation time, the target culture medium was cut for 1 h at 4◦C and left again at room temperature. After wa- into four pieces and incubated as explained above (Fig. 3). Plates ter condensation had occurred inside each column, they were were incubated for 48 h at room temperature. After incubation, shaken by hand, and the columns were then incubated for a du- presence of colonies of P. putida was assessed by fluorescence ration of 11 days at room temperature. Target culture medium microscopy, and presence of fungi was assessed visually. was plated on NA medium (as explained above), and presence For the columns placed in microcosms and inoculated with of bacterial colonies was assessed by visual inspection after both fungi and bacteria, a Hill and Smith test, allowing evalua- 24 h. tion of the correlation between qualitative and quantitative data (Hill and Smith 1976), was performed on R (R Core Team 2013)us- Field tests ing the package ade4 (Dray and Dufour 2007; Herve´ 2013). This test was selected because it allows for the correlation of the time Fungal highway columns were used to assess the efficiency of of incubation (in the soil and after removal; a quantitative vari- the device to isolate bacteria dispersing using the fungal high- able) with the presence of both fungi and bacteria on the target ways under natural conditions. A soil in a semi-arid area of Mo- culture medium (a qualitative variable). Pearson’s chi-squared rocco (32◦317 N, 8◦329 W) was selected (see Supplementary test was performed to evaluate the significance of the correla- Table S1 for soil properties). In total, 28 columns were placed tions calculated using the command chisq.test in R. along a depth transect in a Cambisol (WRB 2014)underOp- A third validation test was performed to assess the effect untia ficus-indica (Fig. 4 and Supplementary Table S2). This cal- of water condensation that could occur inside the fungal high- cium oxalate-producing plant (McConn and Nakata 2004)has

26 Simon et al. 5

Table 1. Experimental design and results of the validation tests of the columns performed at the laboratory. Results show recovery of no organisms (N), presence of bacteria (B), fungi (F), or both bacteria and fungi (BF).

Validation Condition Microorganisms inoculated Incubation in Post-incubation Organisms test tested in substrate substrate (days) (days) recovered

P. putida Fungus

Agar substrate Absence of bacterial + N60NNN dispersal in absence of fungi

Presence of bacterial + M. crassipes 6 0 FB FB FB dispersal in + Trichoderma sp. 6 0 FB FB FB presence of fungi

Soil microcosm Absence of N N 4 15 NNNNNN contamination N N 10 15 N N N N C N

Absence of bacterial + N40NNN dispersal in absence + N415NNN of fungi + N100NNN + N1015NNN

Presence of bacterial + M. crassipes 40NNN dispersal in + F. oxysporum 40NNN presence of fungi + M. crassipes 48FFBN + F. oxysporum 48NBB + M. crassipes 415FBBN + F. oxysporum 415NBB + M. crassipes 10 0 N N N + F. oxysporum 10 0 N N N + M. crassipes 10 8 N FB FB + F. oxysporum 10 8 F N B + M. crassipes 10 15 FB N N + F. oxysporum 10 15 F F FB

Transport and Bacterial dispersal in + N40NNN condensation water film, with + N211NNN shaking, in absence of fungi

been considered as a model for pedogenic carbonate accumu- responding to the time of transport back to the laboratory). The lation related to the oxalate–carbonate pathway (Cailleau et al. columns were transported fixed into a sealed container. After 2005), in which fungal-driven dispersal seems to be relevant for this time, the target culture medium of each column was in- the activity of oxalotrophic bacteria (Martin et al. 2012;Bravo oculated on MA and NA medium, as indicated for the labora- et al. 2013). In order to attract fungi but avoid a nutrient shock tory microcosm assays. After 72 h, development of colonies was (Oliver 2005), the columns contained a lower amount of carbon visually checked and individual organisms were isolated for fur- (6 g·L−1 MA). Furthermore, the target culture medium of each ther identification and characterization. −1 column was either supplemented with 4 g·L CaC2O4.H2O (Sigma-Aldrich) or included a 1-mm layer of modified Schlegel Identification of enriched microorganisms medium (Braissant, Verrecchia and Aragno 2002)with4g·L−1

CaC2O4.H2O, in order to attract oxalotrophic bacteria. This com- Cultivable microorganisms isolated from columns containing bination of attracting/target media was tailored to the oxalate– both bacterial and fungal growth were identified by DNA carbonate pathway in order to favor the isolation of oxalotrophic sequencing. For bacteria, DNA extraction was performed ac- organisms. However, the set-up can be modified depending on cording to the instructions of the InnuPREP bacteria DNA kit the specific organisms studied (for example, the same medium (Analytic Jena, Jena, Germany). DNA extracts were quantified us- can be used as both attracting and target medium). Variable ing a QubitR fluorometer (Life Technologies Corp., Carlsbad, CA, times of installation (4, 5 and 8 days) were tested. Most of the USA). DNA concentration ranged from 1.21 to 177 ng·μL−1.PCR columns (22 out of 28 columns) were placed at a depth of 15 cm amplification was performed of a partial fragment in response to observations by Bravo et al. (2013) indicating an of the 16S rRNA gene with primers EUB 9–27f (5- abrupt decrease of oxalotrophic bacteria at depths below 20 cm. AGAAAGGAGGTGATCCAGCC-3) and EUB 1542r (5- However, one column was also placed at 35 cm and five more AGAAAGGAGGTGATCCAGCC-3; Liesack, Weyland and Stacke- at 60 cm depth. After being removed from the soil and closed, brandt 1991). PCR master mix contained (in a final volume the columns were incubated for a duration of 15–37 days (cor- of 50 μL): 10 μL buffer (with 1.5 mM MgCl2), 0.2 mM dNTPs

27 6 FEMS Microbiology Ecology, 2015, Vol. 91, No. 11

Figure 3. Recovery of cultivable microorganisms from a fungal highway column. The target culture medium at the top of the column is cut into four pieces. Two pieces are transferred onto a plate of MA medium, and two pieces are placed in physiological water, of which 100 μL are streaked on a plate of culture medium (e.g. nutrient agar).

Figure 4. Fungal highway columns in soils under Opuntia ficus-indica.(a) Columns placed at various depths in the soil. (b, c) close-up images of the columns placed in soil.

mix, 0.2 mM of each primer and 1 U GoTaq DNA Polymerase 5 min. Purification of PCR products was carried out usinga (Promega Corp, Madison, WI, USA); 1 μLofdilutedDNA MultiScreen PCRμ96 Filter Plate (Millipore AG, Zug, Switzerland). template was added (c. 1.21–2 ng·μL−1 of DNA). PCR was car- PCR products were premixed in 50 μL of sterile nanopure water ried out in a Sensoquest Labcycler thermocycler (Witec AG, and filtered for 15 min, adding 10–25 μL of sterile nanopure Gottingen,¨ Germany), with an initial denaturation at 95◦Cfor water afterwards. PCR products were quantified using a QubitR 5 min, followed by 10 cycles consisting of denaturation at fluorometer (DNA concentrations ranged from 51 to· 86ng μL−1) 95◦C for 30 s, annealing at 60◦C(−0.5◦C per cycle) for 45 s, and and sent for Sanger sequencing to GATC Biotech AG (Konstanz, elongation at 72◦C for 1 min, then 25 cycles consisting of denat- Germany). The sequences were deposited in GenBank under uration at 95◦C for 30 s, annealing at 55◦C for 45 s and elongation accession numbers KT634058-KT634071. Search for similarity at 72◦C for 1 min. Final extension was performed at 72◦Cfor against sequences of the 16S rRNA gene was performed using

28 Simon et al. 7

the combination of BLAST (Basic Local Alignment Search Tool; Altschul et al. 1990) and pairwise sequence alignment, comparing the query sequence with the non-redundant GenBank database using the online services of EzTaxon (Kim et al. 2012). For some of the isolates, direct sequencing of the PCR products failed and thus the amplicon was cloned prior to sequencing. The TOPO TA cloning kit (Thermo Fisher Scientific, Waltham, MA, USA) was used to produce this plasmid in One Shot TOP10F chemically competent Escherichia coli cells (Thermo Fisher Scientific, Waltham, MA, USA), following the manufacturer’s guidelines. Plasmid DNA was extracted with the Wizard Plus SV Miniprep DNA purification system (Promega Corp, Madi- son, WI, USA) following the manufacturer’s instructions. Ampli- fication, purification, sequencing and analysis were conducted as indicated above. For fungi, DNA extraction was performed according to the instructions of the PowerSoilR DNA Isolation Kit (MoBio, Carls- bad, CA, USA) with a bead-beating step of 5 min at 50 beats·s−1 Figure 5. Separated Petri dish used for the confirmation of bacterial dispersal along fungal mycelium. Fungus (a) and bacteria (b) were inoculated at one side (Qiagen, Hilden, Germany). DNA extracts were quantified us- of the separation, and a square of medium (c) was cut off beyond the separation ing a Qubit fluorometer. DNA concentration ranged from 1.14 after 8 days of incubation and observed by scanning electron microscopy (SEM). to 8.03 ng·μL−1. A PCR amplification of the internal transcribed spacer (ITS) region of the 5.8S rRNA gene was performed with primers ITS1 (5-TCCGTAGGTGAACCTGCGG-3)andITS4 Germany), 0.5% NaCl (Panreac Quimica S.L.U., Barcelona, Spain) (5-TCCTCCGCTTATTGATATGC-3; Anderson, Prosser and Camp- and 0.3% agar; swarm agar: 0.8% nutrient broth, 0.5% dextrose bell 2003). PCR master mix contained (in a final volume of 50 μL): (Merck, Darmstadt, Germany) and 0.5% agar. Each bacterium 5 μL buffer, 0.25 mM dNTPs mix, 0.2 mM of each primer and 2.5 U was inoculated in triplicate on a single point at the center of Taq DNA Polymerase (New England Biolabs, Ipswich, MA, USA). a swim and a swarm agar plate. Diameter of the colonies was One microliter of DNA template (c. 1.14–2 ng·μL−1 of DNA) was measured after 48 h. Micrococcus luteus was used as a negative added. The PCR program consisted of an initial denaturation at and P. putida as a positive control. 94◦C for 5 min followed by 30 cycles of denaturation at 94◦Cfor In order to confirm that the migratory bacteria recovered 30 s, annealing at 55◦C for 30 s, and elongation at 68◦Cfor30s. were able to disperse on their associated fungal mycelium, Final extension was performed at 68◦C for 5 min. Purification of we co-cultivated each fungal-bacterial couple on a sepa- PCR products was carried out using a MultiScreen PCRμ96 Filter rated Petri dish (Fig. 5). Fungal and bacterial colonization be- Plate. PCR products were premixed in 50 μL of sterile nanopure yond the separation was assessed using scanning electron water and filtered for a duration of 15 min, adding 10–25 μLof microscopy (SEM). Separated Petri dishes contained either MA sterile nanopure water afterwards. Amplicons were quantified or NA medium separated in the middle by a 3-mm wide strip using a Qubit fluorometer (DNA concentrations ranged from 29 without culture medium that encompassed the entire dimen- to 52 ng·μL−1) and sent for Sanger sequencing to GATC Biotech sion of the Petri dish. This area was created by cutting out a strip AG. The sequences were deposited in GenBank under accession ofmediumfromthemiddleofthePetridish.Acube(∼3mm3)of numbers KT634072-KT634079. The search for similarity against a fungal pre-culture on MA medium was inoculated 1 cm away sequences from the ITS region on 5.8S rRNA gene was performed from the separation. A loop of an overnight bacterial pre-culture using BLAST comparing the query sequence with the Interna- on NA medium was streaked on a 2.5 cm line between the fun- tional Nucleotide Sequence Database (INSD) at UNITE (Unified gus and the separation. After 8 days of incubation at room tem- system for the DNA based fungal species linked to the classifi- perature, a cube (∼3mm3) was cut out beyond the separation cation; Koljalg¨ et al. 2013). and fixed during at least 1 h with an aldehyde fixative (1mL paraformaldehyde 20%, 1 mL glutaraldehyde 25% (Agar scien- Hydrophobicity, motility, bacterial dispersal and Type tific, Stansted, UK), 5 mL sodium cacodylate trihydrate buffer III secretion system 0.2 M (pH 7.4; Merck, Darmstadt, Germany), 3 mL H2O), and during at least 1 h with an osmium fixative (1 mL 4OsO 4% (Merck, Since previous research on fungal–bacterial interactions pro- Darmstadt, Germany), 2 mL sodium cacodylate trihydrate buffer posed that bacterial dispersal along mycelia was most efficient 0.2 M (pH 7.4), 1 mL H2O). Fixed samples were dehydrated in for motile bacteria moving along hydrophilic fungal surfaces successive acetone (VWR international, Radnor, PA, USA) baths

(Kohlmeier et al. 2005), and that bacteria from the mycosphere and dried in CO2 at the critical point (BalTec, Pfaffikon,¨ Switzer- often present a Type III secretion system (Warmink and van El- land). Finally, each sample was placed on a support, and covered sas 2008), we tested these properties on the microorganisms iso- with a 23 nm gold layer (Baltec, Pfaffikon,¨ Switzerland). A scan- lated from the field trials with the fungal highway columns. ning electron microscope (ESEM-FEG XL30, Philips, Amsterdam, In order to measure the hydrophilic properties of the sur- the Netherlands) was used to assess the presence of fungi and face of the fungal mycelia, WhatmanR circular membranes of bacteria in the fixed agar plugs (high-vacuum, 10–15 kV, distance nitrocellulose (ref. NC45) or cellulose acetate (ref. OE67) were 10 mm). placed at the center of MA plates. Each fungus was inoculated In order to confirm that the recovered bacteria were not able in triplicate on each type of filter, and incubated at room tem- to reach the target culture medium of a column in the absence perature for 8 days before measuring the contact angle accord- of fungi, a validation test was performed in soil-containing mi- ing to Smits et al. (2003). Motility of bacteria was measured on crocosms, inoculated with each isolated bacterial strain. A bac- swim and swarm agar (composition according to Deziel, Comeau terial lawn was prepared on NA medium and incubated for 48 h. and Villemur [2001]). Swim agar: 1% tryptone (Merck, Darmstadt, Cells from half of the plate were streaked out and resuspended

29 8 FEMS Microbiology Ecology, 2015, Vol. 91, No. 11

in 1 mL of sterile physiological water. The final 550OD of this so- ing culture medium. The additional incubation is intended to lution varied from 1 to 10. One hundred microliters was added allow more time for the organisms to migrate inside the column to 12.5 g of soil and mixed with it with a sterile loop. An indi- towards the target culture medium. In this study, we focused vidual fungal highway column was placed in each microcosm on cultivable organisms. However, the target culture medium for 5 days. After being removed from the soil, the columns were could also be used directly for DNA extraction for use with high- closed and left for incubation for 15 days. After this incubation sensitivity molecular ecology methods (Fig. 2). time, the target culture medium was cut into four pieces and in- Fungal highway columns are low-cost, low-weight, small de- cubated as explained above (Fig. 3). Plates were incubated for 48 vices that were created to be easy to build and transport. We h at room temperature. After incubation, presence of bacterial did not use any glass component (except for the beads) in or- colonies was visually assessed. der to make the device robust. Furthermore, as the composi- Presence of the hrcR gene, coding for one of the elements of tion of the culture media intended to attract the organisms the Type III secretion system, was detected by PCR amplifica- inside the columns can vary, these devices can be used for tion with primers hrcRF (5-ATCGGGGTGCAGCAGGTRC-3)and targeting different kinds of fungi and bacteria with specific hrcRR (5-CGAACAGCAGCAGCTTKARYG-3; Warmink and van El- metabolic capabilities. sas 2008). We used the same DNA extracts employed in bacterial μ identification. PCR master mix contained (in 25 L of final vol- Validation tests with GFP-tagged Pseudomonas putida ume): 5 μL buffer B, 0.2 mM dNTPs mix, 0.2 mM of each primer, 5 μL enhancer, 0.25 μL BSA and 0.1 U KAPA2G Robust DNA poly- The first validation test was performed in order to determine merase (Kapa Biosystems, Wilmington, MA, USA); 1 μLofDNA if fungal highway columns were efficient at attracting fungi as template was added. The PCR program consisted of an initial de- well as collecting bacteria moving along their mycelium and to naturation at 94◦C for 3 min, followed by 30 cycles consisting of assess that this was not possible in the absence of the fungi. denaturation at 94◦C for 45 s, annealing at 56◦C for 45 s, and ex- In co-cultures of the flagellated GFP-tagged bacterium P. putida tension at 72◦C for 25 s. Final extension was performed at 72◦C with either M. crassipes or Trichoderma sp., the fungi–bacteria cou- for 5 min. Presence of amplified DNA fragments was assessed by ple was recovered from the target culture medium after 6 days a 30 min gel electrophoresis at 100 V on a 1.5% agarose gel (Bio- of incubation, and bacterial fluorescence was observed on the concept, Allschwil, Switzerland). DNA was stained for a duration target medium in all columns. In contrast, when columns were of 30 min in a GelRed bath (Biotium, Hayward, CA, USA) before inoculated with P. putida in the absence of fungi, no bacterium observation with a Transilluminator (VWR International). DNA was recovered from the target culture medium. isolated from Burkholderia terrae (BS001) was used as a positive The second validation test was performed with microcosms control. containing soil in which columns were placed for variable du- rations. No fungal growth was observed in the tests for the absence of fungal or bacterial contamination inside the columns. RESULTS AND DISCUSSION One column, placed for a duration of 10 days into the soil and Design of the fungal highway column left closed for another 15 days of incubation, showed bacterial growth (Table 1). Thus, fungal contamination during the assem- A novel tool (fungal highway column) was designed for the bly of the columns is very unlikely, whereas bacterial contami- collection of fungal–bacterial associations from soils or similar nation occurred in 1/12 of the cases tested under the lengthiest water-unsaturated ecosystems by both attracting fungi and se- incubation conditions. In the tests assessing bacterial dispersal lecting for bacteria moving along their hyphae towards a target in the absence of fungi, no bacterium was recovered in any of the culture medium. It thereby exclude bacterial isolation by other columns (Table 1). These results show that motile bacteria can- means than migratory dispersal along the hyphae (i.e. transport not reach the target culture medium in the absence of fungi. In by soil fauna, turbulent dispersal in the soil air and mobilization the tests assessing bacterial dispersal in the presence of fungi, in continuous films of condensing water). Three elements inthe no growth was observed when the target medium was used for columns ensured this. First, a 25-μm mesh tissue was placed plating without post-incubation after taking the columns from between the soil and the attracting culture medium to avoid the the soil regardless of the time they were left in place (4 or 10 entrance of soil fauna, including mites as their body length gen- days). However, we did not find any significant correlation be- erally corresponds to around 500 μm (Goddard 2002; Tixier et al. tween the time columns spent in soil with respect to the time 2012). Second, glass beads inside the columns prevented bacte- of incubation, and the recovery of organisms (p-value = 0.11 for rial aerial dispersal, in particular of bacterial spores (Yamaguchi, the time in soil, and p-value = 0.12 for the time of incubation). In Ichijo and Sakotani 2012). Third, the columns were shrunk in the the remaining 24 columns that were post-incubated for 8 or 15 middle in order to avoid formation of a continuous water film in days, the bacterium alone was recovered from six columns, the which bacteria could swim when the columns are held horizon- fungus alone was recovered from 4 columns, and both fungus tally (i.e. during transportation or when placed horizontally in a and bacterium were recovered from six columns (Table 1). soil profile). For the third validation test, we tested whether a water film The first slice of culture medium, placed between the tis- produced by condensation could allow motile bacteria to reach sue and the glass beads, was carbon rich intended for attract- the target culture medium in the absence of fungal mycelium. ing fungi into the columns. The second slice of culture medium, Similarly, as the columns may be shaken during transportation placed at the top of the glass bead layers, acted as a target from the field to the laboratory, we tested if severe shaking could medium that could be used later for isolation (Fig. 2). When accidentally bring bacteria to the target culture medium. No bac- columns are placed in a substrate, for example a soil, the lat- terium was recovered from the target culture medium in either ter is in direct contact with the 25-μm mesh tissue and the at- case. For the test with shaking and longer incubation time, no tracting culture medium leads fungi into the column (Supple- bacterium was recovered from the target culture medium. mentary Fig. S1). The column is left on the substrate, in order Overall, the validation tests underpin the usefulness of these to allow fungi and their associated bacteria to reach the attract- columns to prevent bacteria from reaching the target culture

30 Simon et al. 9

medium in the absence of fungal hyphae. Based on the valida- ecological factor in semi-arid soils (Deacon 2006). Indeed, tion tests demonstrating the absence of bacterial colonization in fungal species of the genus Fusarium usually show an r-strategy the absence of fungi, the most likely mechanism explaining the (ruderal strategy, consisting of fast-growing organisms pro- presence of bacteria in the target medium, even in the absence ducing many offspring), but in addition, they can also show of the fungal partner, is the facilitated dispersal by the fungal an S-strategy (stress-tolerance, whether to low water potential mycelium. or to plant phenols in this case) and a C-strategy (combative behavior, or antagonistic propensity; Dix 2012; Deacon 2006). Field tests Although other fungal genera such as Penicillium present a better water-stress tolerance than Fusarium, species of the former A soil located in a semi-arid area of Morocco was chosen for generally show a much lower growth rate. In fact, a Penicillium the first field test of the columns. The latter were placed atdif- strain was isolated in one of the columns without bacteria, and ferent depths along a soil transect under the oxalogenic plant thus was not characterized further. In addition, the presence of Opuntia ficus-indica. After various time of both placement into an easily accessible carbon source in the form of the attracting the soil and incubation, presence of easily cultivable fungi and and target medium could select for opportunistic fungi with an bacteria on the target culture medium of each column was as- r-strategy. This does not mean that other fungi are not able to sessed. As observed in the validation tests in microcosms, bac- enter the columns (and thus bring bacteria along their hyphae), teria and fungi (together or alone) were not recovered from all but the latter appear to have been outcompeted in the culturing the columns. Both fungi and bacteria were isolated from 8 out approach conducted here. of the 28 columns, while bacteria alone were isolated from three Most of the isolated bacteria belonged to groups that are columns, and fungi alone from eight columns. We were not able likely to be associated with soil fungi. These ‘fungiphile’ to cultivate any organism from the remaining nine columns bacteria belong to the orders Rhizobiales (Alphaproteobac- (Supplementary Table S2). Concerning the depth, most of the teria), Burkholderiales (Betaproteobacteria), Xanthomonadales columns were placed at 15 cm and thus the results of the other and Pseudomonadales (Gammaproteobacteria; Folman et al. depths need to be considered with caution. From those columns 2008; Bonfante and Anca 2009; Warmink, Nazir and van Elsas placed at 35 or 60 cm, fungi and bacteria were recovered from 2009). In addition to these groups, we have identified Olivibacter only one column and fungi alone from another. sp. (Bacteroidetes), as well as bacteria from the families Coma- For the columns from which we isolated bacteria alone, it is monadaceae (Variovorax sp.) and Alcaligenaceae (Achromobacter possible that these did use a mycelial path, but that the asso- sp.), which have not been reported in previous studies. The latter ciated fungus did not develop in the target culture medium af- two have been recently identified ex situ from soils under the inter reaching the medium serving as dispersal vector for those fluence of Milicia excelsa (another oxalogenic plant) in Cameroon bacteria. In the case of columns colonized by fungi alone, the (Bravo et al. 2013), as well as in a culture-based survey of bac- processes that impaired bacterial dispersal are unclear, but one terial diversity in three different oxalogenic tree species (Bravo possible explanation is that the challenging trophic and physi- et al. 2015). cal conditions inside the unsaturated part of the column (glass beads) might modify fungal physiology and interaction with bac- Hydrophobicity, motility, bacterial dispersal and Type teria. For example, we have observed in culture experiments that III secretion system under conditions that are favorable to the fungus (for example growth in MA), the number of fungi transporting bacteria de- The hydrophobicity of the fungal mycelium measured by the creases (Table 2). More importantly, our results confirm the abil- contact angle method gave origin to two possible interpretations ity of fungal highway columns to attract fungal–bacterial cou- of the results. As there are discrepancies regarding the interpre- ples onto a target culture medium. Although it is likely that the tation of the contact angle value for the determination of the hy- composition of the culture media in the columns could affect the drophobicity of a surface, we compared two references for the amount and diversity of recovered organisms, evaluating this analysis of our results. According to Chau et al. (2009), all fun- was not the aim of these validation tests in the field. gal mycelia are hydrophilic (hydrophilic surface: contact angle <90◦; hydrophobic surface: contact angle >90◦), but according Identification of microorganisms enriched to Smits et al. (2003), one fungal mycelium is hydrophobic and all the others intermediate (hydrophilic surface: contact angle Among the microorganisms isolated, 12 bacterial strains <30◦; intermediate surface: contact angle between 30◦ and 60◦; (Achromobacter mucicolens, Achromobacter spanius, Acineto- hydrophobic surface: contact angle >60◦). Nonetheless, migra- bacter calcoaceticus, Acinetobacter sp., Ochrobactrum pecoris, tion was demonstrated, supporting a hydrophilic nature. Ochrobactrum sp., Olivibacter soli, Pseudomonas frederiksber- As the growth of fungal hyphae is apical (Deacon 2006), pas- gensis, Stenotrophomonas humi, Stenotrophomonas maltophilia, sive bacterial transport by the fungus (a dispersal mechanism Stenotrophomonas rhizophila,andVariovorax soli) and five fungal dubbed ‘subway’; Kohlmeier et al. 2005) is improbable. Bacterial species (Fusarium chlamydosporum, Fusarium equiseti, Fusarium dispersal along fungal hyphae can thus be expected to be an nygamai, Fusarium oxysporum, and Chaetomium globosum)were active process involving a type of bacterial motility (Kohlmeier identified (Table 2). Three more bacterial strains were isolated, et al. 2005). In fact, swimming was a common feature observed but all our attempts either to sequence directly a PCR product in a diverse group of flagellated bacteria able to disperse along or to clone and sequence their DNA failed. Thereby, we named fungal hyphae (Pion et al. 2013). Therefore, we were expecting them ‘Unidentified bacterium a’, ‘Unidentified bacterium b’and that motile bacteria would be recovered from the target culture ‘Unidentified bacterium c’. The identified fungi corresponded medium of the columns. However, not all the isolates were pos- almost exclusively to Fusarium spp., a genus of fast-growing itive in the swimming and swarming tests in agar media. The and competitive fungi producing diverse mycotoxins (Gutleb, capabilities of the isolated bacteria to swim or swarm varied Morrison and Murk 2002). Moreover, this genus is known to amongst the isolates and also within the same bacterial species contain species that resist water stress, which can be a limiting (Table 2). For example, A. calcoaceticus isolated from column 5

31 10 FEMS Microbiology Ecology, 2015, Vol. 91, No. 11 . Opuntia ficus-indica ––– ––– ––– ––– ––– ––– ––– ––– ––– ––– oxalogenic plant –––– + + + + + + ++ ++ + ++ Confirmation of Migration in a column –– – –– – – –––––– –––––– –– –– –– – + ++++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ Swimming Swarming MA NA (99.3%) – (99.0%) (99.6%) (99.8%) (99.4%)–––––––– (99.9%) – – – – (99.2%) (99.9%) (98.8%) 9.% –––––––– (98.4%) (99.1%) sp. (99.6%) sp. Associated bacteria Motility test TTSS migration in absence of fungus Acinetobacter Stenotrophomonas humi Olivibacter soli Unidentified bacteriumPseudomonas frederiksbergensis a – Ochrobactrum Variovorax soli Achromobacter spanius Stenotrophomonas maltophilia Ochrobactrum pecoris Acinetobacter calcoaceticus Acinetobacter calcoaceticus Achromobacter mucicolens b ) ◦ 58 42 47 40 50 72 Unidentified56 bacterium55 c N/A N/A – N/A N/A N/A (99.8%) chlamydosporum (99.4%) (99.4%) oxysporum (99.8%) Fungus Contact angle ( globosum Chaetomium Fusarium Fusarium Fusarium Fusarium oxysporum (99.8%) Fusarium chlamydosporum 99.8%) Fusarium equiseti Fusarium nygamai (99.8%) chlamydosporum Stenotrophomonas rhizophila (99.6%) Unidentified bacterium b a Identification and properties of organismsthe isolated from eight columns presenting simultaneously fungal andbacterial growth in a soil under the Column Mean value. See Supplementary Table S2 for conditions associated with each number. 1 3 4 5 7 8 18 26 Table 2. a b The percentage of identity to the indicated group is shown in brackets.

32 Simon et al. 11

Figure 6. SEM images of bacteria moving along fungal hyphae. (a) Stenotrophomonas maltophilia on hyphae of Fusarium oxysporum;(b) Ochrobactrum sp. on hyphae of Fusarium oxysporum;(c) Unidentified bacterium b on hyphae of Fusarium chlamydosporum;(d) Stenotrophomonas humi on hyphae of Fusarium equiseti.

was positive in both swimming and swarming tests, whereas grators’. From our isolates, V. soli (column 1) and A. calcoaceticus another isolate from the same species isolated from column 8 (column 26) are potential candidates for this type of commu- was negative in both swimming and swarming tests. In order to nity migrators. Even though the phenomenon of hitchhikers has understand the relevance of motility on the dispersal of the iso- not yet been explored extensively, species competent to migrate lated bacteria along fungal hyphae, we tested the ability of each appear to be TTSS positive, while the community migrators ap- isolated bacterium to migrate along its potentially associated pear not to be. Therefore, we investigated whether the capability fungal mycelium on the media used for isolation. For 10 out of of migrating along fungal mycelium was correlated to the pres- the 14 fungal–bacterial couples tested, migration was observed ence of a molecular marker for TTSS. The Type III secretion sys- (Table 2). On the SEM pictures of positive migratory couples (ex- tem was detected in one bacterial strain (Unidentified bacterium amples are shown in Fig. 6), bacteria surround their associated a; Table 2), but was not observed in any of the other strains, hyphae. In three cases, O. soli, ‘Unidentified bacterium a’, and even those that could migrate along the associated mycelium. A. calcoaceticus, the swimming and swarming test were positive Although negative PCR results are not conclusive since PCR bias but the bacteria did not migrate on the associated fungus. The (e.g. mismatches with the annealing sites) might have influ- opposite was observed in the case of A. mucicolens,whichwas enced the results, our findings suggest that TTSS is not essen- negative in the swimming and swarming test but dispersed on tial in fungal highway interactions, but its role on community themyceliumofF. nygamai. We have no certitude about the phe- migration needs to be investigated further. As TTSS is a needle- nomenon that led to these contradictory results. In the case of shaped structure that generally allows for protein injection into motility, it is possible that the swimming and swarming test can target eukaryotic cells by pathogenic bacteria (Galan et al. 2014), lead to false negatives. For those bacteria that reached the target it might not necessarily have a direct link with fungal highways, culture medium of the columns but did not disperse on the as- in which bacteria actively move along the cells and do not attach sociated mycelium, it is important to mention that the isolated to them. couples represent only potentially associated partners and that To confirm that all the bacteria isolated could not reach the it is possible that these bacteria had reached the medium us- target culture medium of a column without the presence of fun- ing hyphae from a fungal species that was not recovered in the gal mycelium, an additional experiment was conducted. When isolation process. each bacterium was inoculated alone under a column, we were An alternative explanation to the absence of dispersal for not able to recover any of these bacteria from the target culture some of the isolated strains is the phenomenon known as ‘hitch- medium, after 5 days of incubation in the soil and 15 days of hikers’ on the fungal highway (Warmink et al. 2011). Hitchhikers post-incubation, which indicated that the bacteria we recovered corresponded to bacteria able to move along fungal hyphae only from the field columns may not have been able to migrate to- in presence of other motile bacteria acting as ‘community mi- wards the target culture medium in absence of fungi.

33 12 FEMS Microbiology Ecology, 2015, Vol. 91, No. 11

CONCLUSION Bravo D, Cailleau G, Bindschedler S, et al. Isolation of oxalotrophic bacteria able to disperse on fungal mycelium. The validation tests we performed on the fungal highway FEMS Microbiol Lett 2013;348:157–66. columns show that they enable the isolation of bacteria mov- Cailleau G, Braissant O, Dupraz C, et al. Biologically induced ac- ing along fungal hyphae from soils. These columns allowed the cumulations of CaCO3 in orthox soils of Biga, Ivory Coast. evaluation of the diversity of microorganisms potentially inter- Catena 2005;59:1–17. acting through fungal highways in one specific soil. The confir- Chau HW, Si BC, Goh YK, et al. A novel method for identify- mation of migration on fungal hyphae (on a Petri dish) and of ing hydrophobicity on fungal surfaces. Mycol Res 2009;113: the absence of bacterial dispersal in columns without the fungal 1046–52. partner are important steps for the validation of the enriched Deacon JW. Fungal Biology. Oxford, UK: Blackwell Publishing, fungal–bacterial couples. The columns could be the basis for 2006. studying fungal highway-like associations in various natural Deziel E, Comeau Y, Villemur R. Initiation of biofilm formation ecosystems. Furthermore, the design of these columns opens by Pseudomonas aeruginosa 57RP correlates with emergence of new perspectives for targeting different kinds of microorgan- hyperpiliated and highly adherent phenotypic variants de- isms, for example by changing the culture media used as attrac- ficient in swimming, swarming, and twitching motilities. J tor or target, by changing the size of the column and by com- Bacteriol 2001;183:1195–204. bining culture-dependent and molecular ecology methods with Dix NJ. Fungal Ecology. Netherlands: Springer, 2012. higher sensitivity. Dray S, Dufour AB. The ade4 package: implementing the duality diagram for ecologists. J Stat Softw 2007;22:1–20. SUPPLEMENTARY DATA Folman LB, Klein Gunnewiek PJA, Boddy L, et al. Impact of white- rot fungi on numbers and community composition of bacte- Supplementary data is available at FEMSEC online. ria colonizing beech wood from forest soil. FEMS Microbiol Ecol 2008;63:181–91. Furuno S, Remer R, Chatzinotas A, et al. Use of mycelia as paths ACKNOWLEDGEMENTS for the isolation of contaminant-degrading bacteria from soil. Microb Biotechnol 2012;5:142–8. We would like to thank Michael¨ Berthoud (University of Lau- Galan JE, Lara-Tejero M, Marlovits TC, et al. Bacterial type III se- sanne, Switzerland) for the soil analysis and Dr Mohammed cretion systems: specialized nanomachines for protein de- Soleih (Morocco) for accommodation and helpful insights dur- livery into target cells. Annu Rev Microbiol 2014;68:415–38. ing our field campaign. Goddard J. Physician’s Guide to Arthropods of Medical Importance. Boca Raton, FL: CRC Press, 2002. FUNDING Gutleb A, Morrison E, Murk A. Cytotoxicity assays for mycotoxins produced by Fusarium strains: a review. Environ Toxicol Phar- Fieldwork was supported by a grant of the Commission Egalite´ macol 2002;11:309–20. des chances (Neuchatel,ˆ Switzerland) to A.S. This research was Harshey RM. Bacterial motility on a surface: many ways to a supported by the Swiss National Science Foundation through common goal. Annu Rev Microbiol 2003;57:249–73. Grants FN CR22I2-137994/1, FN CR3212-149853/1 and BPLAP2 HerveM.´ RVAideMemoire: Diverse Basic Statistical and 140105 as well as by program topic “Chemicals in the Environ- Graphical Functions (R package version 0.9-27). 2013. ment” (CITE) of the Helmholtz Association. http://CRAN.R-project.org/package=RVAideMemoire (20 March 2015, date last accessed). Conflict of interest . None declared. Hill MO, Smith AJE . Principal component analysis of taxonomic data with multi-state discrete characters. Taxon 1976;25: REFERENCES 249–55. IUSS Working Group WRB. World Reference Base for Soil Resources Altschul SF, Gish W, Miller W, et al. Basic local alignment search 2014: International Soil Classification System for Naming Soils and tool. J Mol Biol 1990;215:403–10. Creating Legends for Soil Maps. World Soil Resources Reports Anderson IC, Prosser JI, Campbell CD. Potential bias of fungal 18S No. 106. Rome: FAO, 2014. http://www.fao.org/3/ai3794e.pdf rDNA and internal transcribed spacer polymerase chain re- (20 March 2015, date last accessed). action primers for estimating fungal biodiversity in soil. En- Kim OS, Cho YJ, Lee K, et al. Introducing EzTaxon: a prokary- viron Microbiol 2003;5:36–47. otic 16S rRNA gene sequence database with phylotypes Banitz T, Johst K, Wick LY, et al. Highways versus pipelines: that represent uncultured species. Int J Syst Evol Microbiol contributions of two fungal transport mechanisms to 2012;62:716–21. efficient bioremediation. Environ Microbiol Rep 2013;5: Kohlmeier S, Smits TH, Ford R, et al. Taking the fungal highway: 211–8. mobilization of pollutant-degrading bacteria by fungi. Envi- Biais N. Pili de type IV, Quand l’union fait la force. Med Sci (Paris) ron Sci Technol 2005;39:4640–6. 2009;25:437–60. Koljalg¨ U, Nilsson RH, Abarenkov K, et al. Towards a unified Bonfante P, Anca IA. Plants, mycorrhizal fungi, and bacteria: a paradigm for sequence-based identification of fungi. Mol Ecol network of interactions. Annu Rev Microbiol 2009;63:363–83. 2013;22:5271–7. Braissant O, Verrecchia EP, Aragno M. Is the contribution of bac- Lee K, Kobayashi N, Watanabe M, et al. Spread and change teria to terrestrial carbon budget greatly underestimated? in stress resistance of Shiga toxin-producing Escherichia Naturwissenschaften 2002;89:366–70. coli O157 on fungal colonies. Microb Biotechnol 2014;7: Bravo D, Braissant O, Cailleau G, et al. Isolation and characteri- 621–9. zation of oxalotrophic bacteria from tropical soils. Arch Mi- Liesack W, Weyland H, Stackebrandt E. Potential risks of gene crobiol 2015;197:65–77. amplification by PCR as determined by 16S rDNA analysis

34 Simon et al. 13

of a mixed-culture of strict barophilic bacteria. Microb Ecol Sun M, Wartel M, Cascales E, et al. Motor-driven intracellular 1991;21:191–8. transport powers bacterial gliding motility. Proc Natl Acad Sci Martin G, Guggiari M, Bravo D, et al. Fungi, bacteria and soil pH: USA2011;108:7559–64. the oxalate-carbonate pathway as a model for metabolic in- Tixier MS, Kreiter S, Douin M, et al. Rates of description teraction. Environ Microbiol 2012;14:2960–70. of Phytoseiidae mite species (Acari: Mesostigmata): space, McConn MM, Nakata PA. Oxalate reduces calcium availabil- time and body size variations. Biodivers Conserv 2012;21: ity in the pads of the prickly pear cactus through forma- 993–1013. tion of calcium oxalate crystals. J Agric Food Chem 2004;52: Warmink JA, Nazir R, Corten B, et al. Hitchhikers on the fungal 1371–4. highway: the helper effect for bacterial migration via fungal Moore D, Robson, GD, Trinci APJ. 21st Century Guidebook to Fungi. hyphae. Soil Biol Biochem 2011;43:760–5. New York: Cambridge University Press, 2011. Warmink JA, Nazir R, van Elsas JD. Universal and species-specific Oliver JD. The viable but nonculturable state in bacteria. JMicro- bacterial ‘fungiphiles’ in the mycospheres of different basid- biol 2005;43:93–100. iomycetous fungi. Environ Microbiol 2009;11:300–12. Or D, Smets BF, Wraith JM, et al. Physical constraints affecting Warmink JA, van Elsas JD. Selection of bacterial populations in bacterial habitats and activity in unsaturated porous media the mycosphere of Laccaria proxima: is type III secretion in- –areview.Adv Water Res 2007;30:1505–27. volved? ISME J 2008;2:887–900. Pion M, Bshary R, Bindschedler S, et al. Gains of bacterial Warmink JA, van Elsas JD. Migratory response of soil bacteria to flagellar motility in a fungal world. Appl Environ Microbiol Lyophillum sp. strain Karsten in soil microcosms Appl Environ 2013;79:6862–7. Microbiol 2009;75:2820–30. R Core Team. R: a language and environment for statistical Wick LY, Remer R, Wurz¨ B, et al. Effect of fungal hyphae on the computing. Vienna, Austria: R Foundation for Statistical access of bacteria to phenanthrene in soil. Environ Sci Technol Computing, 2013. http://www.R-project.org/ (31 October 2007;41:500–5. 2015, date last accessed). Wosten¨ HAB. Hydrophobins: multipurpose proteins. Annu Rev Smits TH, Wick LY, Harms H, et al. Characterization of the sur- Microbiol 2001;55:625–46. face hydrophobicity of filamentous fungi. Environ Microbiol Yamaguchi N, Ichijo T, Sakotani A, et al. Global dispersion of bac- 2003;5:85–91. terial cells on Asian dust. Sci Rep 2012;2:525.

Chapter 1 - Fungal highway columns

Supplementary information

Supplementary Information

Figure S1. Lichen on an Opuntia ficus-indica paddle. This image shows the place were a column was placed, just after its removal. Mycelium is escaping from the lichen towards the attracting medium of the column, indicating that it acted as a strong fungal attractor.

Table S1. Main physicochemical characteristics of the soil (Hypereutric Cambisol) under

Opuntia ficus-indica. Table shows water pH (pHH2O), relative humidity (rH), nitrogen content (N) and total organic carbon (TOC) at various depths in the soil profile (data kindly provided by Michael Berthoud).

Depth [cm] pHH2O rH [%] N [%] TOC [%] 0-6 8.5 1.6 0.23 2.73 6-18 8.3 2.3 0.08 0.93 18-35 7.9 2.2 0.09 1.10 35-45 9.0 5.5 0.06 0.71 45-80 9.1 5.1 0.03 0.26

37 Table S2. Details of the 28 columns placed in the soil. Table shows at which depth the columns were placed in the soil, time of incubation in the soil, time of incubation outside of the soil, cultivable organisms recovered: fungi and bacteria (F+B), fungi only (F), bacteria only (B) and no organisms (n), as well as the composition of the target culture medium: supplemented with CaCO3 (S) or with a layer of Modified Schlegel medium containing CaCO3 (L).

Column Depth Time in Incubation Cultivable Target n° [cm] soil [days] organisms culture [days] medium 1 15 8 17 F+B S 2 15 8 17 F L 3 15 8 21 F+B L 4 15 8 28 F+B S 5 15 8 31 F+B L 6 15 5 23 F S 7 15 5 23 F+B S 8 15 5 23 F+B S 9 15 5 23 n L 10 15 5 23 B S 11 15 5 34 F S 12 15 5 34 F L 13 15 5 34 F S 14 15 5 34 n L 15 15 5 34 F L 16 15 5 37 B L 17 15 5 37 F L 18 15 5 37 F+B L 19 15 5 37 n S 20 15 5 37 B L 21 35 4 15 n L 22 60 4 15 n S 23 15 4 21 n L 24 60 4 21 n L 25 60 4 22 n L 26 60 4 25 F+B S 27 15 4 28 n S 28 60 4 28 F L

38 Chapter 1 - Fungal highway columns

Main fi ndings and perspectives

Main fi ndings

We developed and validated the effi ciency of fungal highway columns, a device allowing for in situ collection of both bacterial and fungal strains potentially interacting through fungal highways in natural environments.

Using the fungal highway columns in a semi-arid soil under the oxalogenic plant Opuntia fi cus-indica, we isolated several associations of cultivable organisms. For the fi rst time, we demonstrated that soil fungi and bacteria are able to interact through fungal highways.

We observed that the bacterial migration along their fungal partner varied according to the culture medium.

We observed that a non-motile bacterium was able to migrate along its associated fungus, and we hypothesized that classical swimming and swarming tests were not suffi cient to assess bacterial motility.

Perspectives

As we performed a culture-dependent approach, it would be interesting to perform a molecular-based approach in order to assess the diversity of organisms isolated in the columns. We also would like to know if the composition of the culture medium inside the columns infl uences the diversity of selected microorganisms. Th is is tested in Chapter 2 of the present thesis.

We would like to know if the isolated fungi and bacteria present specifi c metabolic properties. Th is is presented in Chapters 3 and 4 of the present thesis.

Chapter 2

Culture-independent approach

Chapter 2 - Culture-independent approach

Foreword

In Chapter 1 of the present thesis, I presented the fungal highway columns, and cultivable associates of fungal highways isolated from a semi-arid soil. We observed that migration of these bacteria was diff erent according to the culture medium.

In this section, I will present another fi eld work performed with fungal highway columns. We selected three diff erent soils in Switzerland, with various physico-chemical properties. We used two types of fungal highway columns, containing a rich and a poor culture medium, and conducted a molecular biology approach.

Th e idea was to assess whether associates of fungal highways were diff erent according to soil properties, and if the diff erent culture media inside the columns had an infl uence on the isolated fungal and bacterial communities inside the columns.

A view from one of the sampling locations, Le Creux-du-Van, Switzerland, 2014.

For this work, I received substantial help from a lab technician trainee, Andrej Al-Dourobi. We assembled the columns together, and we walked during hours in the forests in order to fi nd all soil profi les. He performed DNA extractions, DGGEs, and most of the PCRs. As well, Vincent Hervé generated all OTU matrices and taxonomic assignments for the pyrosequencing data. Without their help, this work would have been endless.

Chapter 2 - Culture-independent approach

Manuscript

Th is section presents the following manuscript, submitted in FEMS Microbiology Ecology:

Simon A, Hervé V, Al-Dourobi A, Verrecchia E, Junier P. Highways and subways: identifi cation of bacteria dispersing along fungal hyphae in soils.

Highways and subways: identifi cation of bacteria dispersing along fungal hyphae in soils

Anaele Simon1, Vincent Hervé1,2, Andrej Al-Dourobi1, Eric Verrecchia2, Pilar Junier1 1 Laboratory of Microbiology, Institute of Biology, University of Neuchâtel, Neuchâtel, Switzerland 2 Biogeosciences laboratory, Institute of Earth Surface Dynamics, University of Lausanne, 1015 Lausanne, Switzerland

Corresponding author: Pilar Junier, Laboratory of Microbiology, Institute of Biology, University of Neuchâ- tel, Rue Emile-Argand 11, 2000 Neuchâtel, Switzerland. [email protected]. Tel. +41327182244. Fax. +41327183001.

Abstract

Soils are complex ecosystems in which fungi and bacteria co-exist and interact. Fungal highways are a specifi c kind of interaction by which bacteria use fungal hyphae to disperse in soils. Despite the fact that fungal highways have been studied with some culturable laboratory models, the diversity of fungi and bacteria interacting in this way in soils is still unknown. Fungal highway columns were used as a selective method to study the identity of fungi and bacteria able to migrate along their hyphae in three diff erent soils. Regardless of the soil type, the vast majority of fungi selected inside the columns belonged to the genus Mortierella (phylum Zygomycota). In contrast, a diverse community of bacteria dominated by Firmicutes and Proteobacteria was observed inside the columns. Members of the genus Mortierella are known for hosting bacterial endosymbionts, especially bacteria affi liated to Burkholderiales, a group that was identifi ed alongside Mortierella spp. in the columns. Th is suggests that some bacteria might disperse within fungal hyphae, using those as subways. As extra- and intra-hyphal bacterial dispersal involves preferentially some fungal and bacterial phyla, our fi ndings suggest that this type of interaction must be considered as an important structuring factor of microbial communities in soil.

Introduction dered as impossible for bacteria, whose motility is in consequence limited (de Boer et al., 2005). Another Fungi and bacteria coexist in almost every type of eco- crucial advantage of fungi is their ability to translo- system (Frey-Klett et al., 2011). However, mycologists cate nutrients and carbon inside their hyphae (Schütte, and bacteriologists have traditionally conducted their 1956; Wells et al., 1995a), allowing them to better studies separately (Strickland & Rousk, 2010). Yet, the cross nutrient-poor areas, which are a common feature sharing of a common habitat and common resources in soils (de Boer et al., 2005; Wells et al., 1995b). has led to numerous and diverse interactions between Th e existence of fungal highways, an interaction by fungi and bacteria (Kobayashi & Crouch, 2009). which bacteria are able to migrate along fungal hy- When environments such as soil are considered, phae, casts doubts on the current understanding of fungi seem to be better adapted than most bacteria to limited bacterial dispersal in soils, as the motility of the physical constrains of the habitat. In soil, the fi - those bacteria able to disperse on fungal networks in- lamentous growth of fungi allows them to bridge the creases drastically (Kohlmeier et al., 2005). In a pre- air-fi lled voids, thus colonizing water-unsaturated soil vious study, we have developed a device called fungal patches (Ritz & Young, 2004). Th is is classically consi- highway columns, which allows for the selection of

47 fungi and associated migratory bacteria directly from Chaux-de-Fonds, Switzerland) and 15 g.l-1 agar (Bio- soils (Simon et al., 2015). In this previous study, we fo- life Italiana, Milano, Italy), while the relatively poorer cused on isolating and characterizing easily cultivable culture medium was R2A medium (Rice et al., 2012), organisms in a soil infl uenced by the oxalate-carbo- composed of Yeast extract 0.5 g.l-1, Bacto Peptone 0.5 nate pathway, a biogeochemical cycle for which fungal g.l-1, Casamino acids 0.5 g.l-1, Glucose 0.5 g.l-1, Soluble highways are known to be implicated in soil function starch 0.5 g.l-1, Na-pyruvate 0.3 g.l-1, K2HPO4 0.3 (Martin et al., 2012, Bravo et al., 2013). Although se- g.l-1, MgSO4.7H2O 0.05 g.l-1, purifi ed agar 15 g.l-1. veral fungal-bacterial couples were identifi ed in this Th e columns were placed in one of the superfi cial way, culturing is notably prone to bias (Rappe & Gio- horizons of three diff erent soils located in Switzerland: vannoni, 2003) and thus the diversity of fungal and a Folic Histosol (OFnoz3 horizon, 11-22 cm; 46° 55’ bacterial communities engaging in fungal highways in 59.42’’ N, 6° 43’ 36.24’’ E), a Dystric Cambisol (A ho- soils remains unknown. rizon, 0-5 cm; 47° 0’ 13.31’’ N, 6° 56’ 52.50’’ E), and a In the present study, we used fungal highway co- Fluvisol (Jsca1 horizon, 0-9 cm; 46° 32’ 8.77’’ N, 7° 4’ lumns in order to study the diversity of fungi and bac- 3.82’ E; Baize 2009, Jabiol et al., 2013, WRB 2014). teria interacting through fungal highways compared In each soil, organic carbon content, nitrogen content, to the total microbial community in soils. We selec- C:N ratio, pH, water content, and the amount of or- ted soils with diff erent organic carbon and nitrogen ganic matter were measured (Figure 1). Six columns contents and C:N ratios. Th ese variables were studied were placed in each soil type (3 with LMA medium because it has been observed that limitation of car- and 3 with R2A medium). Th e columns were left in bon aff ects bacterial growth (Demoling et al., 2007), the soil for 7 days. After this time, they were carried while a limited access to nitrogen aff ects fungal growth back to the laboratory, and for each column, total (Boyle, 1998). Likewise, it has been shown that the DNA was extracted from 3 parts: the portion of soil fungal:bacterial ratio is related to the C:N ratio in soils on which the column was placed, the attracting culture (Waring et al., 2013). Th us, by choosing three soils medium in direct contact with the soil, and the tar- varying in these parameters, we expected not only to get culture medium at the top of the column. Th is al- describe the diff erent fungal and bacterial communi- lowed for diff erentiating the pool of organisms present ties able to interact in fungal highways but also to as- in the soil from those able to colonize the attracting sess whether specifi c fungal-bacterial associations vary culture medium, and those able to reach the target accordingly to carbon and nitrogen contents in soils. culture medium. In this third section, only fungi and bacteria able to migrate along fungal hyphae should be present. DNA extraction was performed according Materials and Methods to the instructions of the FastDNA spin kit for soil (MP Biomedicals, Santa Ana, USA) with an additional Sampling, soil properties and DNA extraction bead-beating step of 10 min at 50 beats.s-1 (Qiagen, For the sampling of fungi and bacteria able to migrate Hilden, Germany). on fungal mycelium, we used fungal highway columns. Th ese columns are recently developed devices Sequencing and analysis of taxonomic diversity that allow for targeted isolation and separation of soil In order to identify fungi and bacteria, internal trans- fungi and bacteria moving along fungal hyphae towar- cribed spacer (ITS) and a fragment of 16S rRNA gene ds a target culture medium (see Simon et al. (2015) were targeted, respectively. Due to their low yield, for details on the composition, and validation tests). DNA extracts from the top, respectively the middle In order to assess whether the culture medium placed of the three columns from each medium were pooled, inside the columns has an impact on fungal or bacte- and a step of pre-amplifi cation by PCR was performed rial diversity, we used one set of columns containing in order to obtain enough DNA material for sequen- nutrient-rich and another containing nutrient-poor cing. For fungi, amplifi cation of a partial fragment of culture media as attractor and target media. Th e nu- the ITS region was performed with the primers ITS1F trient-rich culture medium was a low-malt agar (LMA) (5’-CTTGGTCATTTAGAGGAAGTAA-3’) and medium, composed of 6 g.l-1 malt (Mycotec SA, La ITS4 (5’-TCCTCCGCTTATTGATATGC-3’; Ander-

48 Figure 1. Soils analyzed in this study. a) Th e physicochemical properties of the three selected soils, a Fluvisol, a Dystic Cambiosol and a Folic Histosol, are indicated aside an image of the soil profi le used for the installation of the columns. Th e soil horizon in which the columns were placed is represented by the colored bands. b) Geographical location of the three soils in Switzerland. c) Close view of columns in the Fluvisol.

son et al., 2003). For pre-amplifi cation, the master mix per cycle) for 20 s, and elongation at 72 °C for 1 min, contained (in 25 μl of fi nal volume): 5 μl buff er (with then 25 cycles consisting of denaturation at 95 °C for 1.5 mM MgCl2), 0.5 mM dNTPs mix, 0.5 mM of 30 s, annealing at 55 °C for 30 s and elongation at 72 each primer, 1 mg.ml-1 BSA and 0.1 U DNA Polyme- °C for 1 min. Final extension was performed at 72 °C rase (Kapa Biosystems, Inc., Wilmington, USA). 1 μl for 1 min. Final products contained 26 to 250 ng.μl-1 of DNA template was added (≤ 2 ng.μl-1 of DNA). DNA. DNA and the pre-amplifi ed PCR products were PCR was carried out in an Arktik thermocycler (Th er- sent to Eurofi ns Genomics (Ebersberg, Germany) for mo Fisher Scientifi c, Waltham, USA), with an initial amplifi cation using the same primers. Subsequently, denaturation at 95 °C for 2 min, followed by 35 cycles 16S rRNA and ITS amplicons were sequenced in the consisting of denaturation at 95 °C for 30 s, annealing forward direction using 454-technology (Roche, Basel, at 61 °C for 30 s, and elongation at 72 °C for 30 s. Fi- Switzerland). nal extension was performed at 72 °C for 2 min. Final Fungal and bacterial amplicon sequences were ana- products contained 19 to 84 ng.μl-1 DNA. For bac- lyzed independently, using the software Mothur ver- teria, PCR amplifi cation was performed on a partial sion 1.36.1 (Schloss et al., 2009). Bacterial reads were fragment of the 16S rRNA gene using primers EUB processed by following a modifi ed standard operating 9-27f (5’-AGAAAGGAGGTGATCCAGCC-3’) and procedure (Schloss et al., 2011). First, sequencing er- EUB 1542r (5’-AGAAAGGAGGTGATCCAGCC-3’; rors were reduced by implementation of the Ampli- Liesack et al., 1991). For pre-amplifi cation, the master conNoise algorithm and low-quality sequences were mix contained (in 25 μl of fi nal volume): 5 μl buff er removed (minimum length of 220 bp, allowing 1 (with 1.5 mM MgCl2), 0.2 mM dNTPs mix, 0.2 mismatch to the barcode, 2 mismatches to the primer, mM of each primer, 1 mg.ml-1 BSA and 0.5 U DNA and homopolymers no longer than 8 bp). Sequences Polymerase. One microliter of DNA template was were then trimmed to keep only high quality reads (Q added (≤ 2 ng.μl-1 of DNA). PCR was carried out in an ≥ 35). Barcode and primer sequences were removed. Arktik thermocycler, with an initial denaturation at 95 Subsequently, sequences were aligned to the SILVA re- °C for 2 min, followed by 10 cycles consisting of dena- ference database release 119 (Quast et al., 2013) and turation at 95 °C for 30 s, annealing at 60 °C (-0.5 °C preclustered (pre.cluster, diff s=1). Chimeras were re-

49 Figure 2. Summary of the fungal (left) and bacterial (right) richness at phylum level for soil, middle and top of the columns. A color code for the diff erent soil types is indicated below. Th e number of OTUs unique to each soil type and those shared are indicated by Venn diagrams, color-coded in the same way as the soil type. Th e Venn diagrams are presented for each compartment of the columns depicted as an illustration in the middle.

moved using the chimera.uchime Mothur command mated using Mothur (Schloss et al., 2009). All statis- and singletons were excluded. Finally, sequences were tical analyses were computed using R software version classifi ed using the naïve Bayesian classifi er (Wang et 3.2.2 (R Core Team, 2013). To evaluate the co-oc- al., 2007) implemented in Mothur with the SILVA currence probability between the bacterial and fungal reference database release 119 (Quast et al., 2013). OTUs, the Veech probabilist model of species co-oc- Operational taxonomic units (OTUs) were generated currence (Veech, 2013) was applied using the cooccur using the average neighbor algorithm. An OTU was package in R. defi ned at the 97 % sequence similarity level. Regar- ding the fungal reads, they were quality processed with the same parameters as described above. After remo- Results ving chimeras and singletons, the presence of fungal ITS was checked using ITSx version 1.0.11 (Bengts- Soil properties son-Palme et al., 2013) and non-fungal ITS sequences Th ree soils, a Folic Histosol, a Distric Cambisol and were discarded. Subsequently, ITS sequences were a Fluvisol, with diff erent organic carbon and nitrogen pairwise aligned to generate a distance matrix using contents and C:N ratios, were selected to study the di- pairwise.seqs command. Finally, sequences were clas- versity of fungi and bacteria interacting through fungal sifi ed using the naïve Bayesian classifi er (Wang et al., highways. Th e Fluvisol had the lowest relative content 2007) implemented in Mothur with the UNITE v6_ of organic carbon (2 %) and organic matter (5.5 %), sh_dynamic database (Köljalg et al., 2013). Operatio- the lowest C:N ratio (15.02) and the highest pH (7.8). nal taxonomic units (OTUs) were generated using the Th e Folic Histosol had the highest relative content of average neighbor algorithm. An OTU was defi ned at organic carbon (45.17 %) and organic matter (92 %), the 97% sequence similarity level. Taxonomic assign- the highest C:N ratio (44.77) and the lowest pH (3.6). ment was made over 80 % similarity. Th e raw sequence Despite of its high C:N ratio, the Folic Histosol had data have been deposited in the NCBI Sequence Read also the highest relative amount of N (1.01 %). Th e Archive under the accession number SRP069959. values measured for the Dystric Cambisol were inter- Rarefaction curves (calculated from 10 000 itera- mediate between the values obtained for the Folic His- tions) and relative abundances of OTUs were esti- tosol and the Fluvisol (Figure 1).

50 Figure 3. Taxonomical breakdown of the fungal OTUs richness present in the soil only (black), in the soil and in the top of the columns (dark gray), and in the columns only (light gray). Th e comparison is presented in percentage values but the total number of OTUs in each taxonomic group is indicated within the bars. Th e right panel displays the diversity of those fungal taxa found on the top of the columns at genus level. Th e three panels correspond to the three soil types indicated on the left.

Sequencing and analysis of taxonomic diversity After quality fi ltering, chimera and singleton re- After leaving the columns for 7 days on site, the fungal movals, a total of 167 518 fungal sequences were re- and bacterial communities in the soils and those reco- tained, and clustered into 718 OTUs defi ned at a si- vered inside the columns were analyzed. We performed milarity level of 97 %. For the bacterial sequences, a separate DNA extractions of three compartments of total of 126 546 reads were retained and clustered into the columns: (i) soil on which the column was placed, 3229 OTUs. Considering that for the attracting and (ii) attracting culture medium (in direct contact with target media we had to perform a pre-amplifi cation the soil), and (iii) target culture medium at the top of step to obtain enough DNA for sequencing, the analy- the columns (soil, middle and top in Figure 2, respec- sis of the community composition was mainly conduc- tively). Sequencing and taxonomic assignment were ted by comparing the presence/absence of individual performed in order to compare the total fungal and OTUs rather than comparing their relative abundance bacterial communities from each soil with the com- measured in sequence counts. Th is approach gave us munities recovered inside the columns. Sequencing a proxy that allows for comparing the richness of the was performed on fungal ITS amplicons and bacterial fungal and bacterial communities between the soils 16S rRNA gene amplicons. One sample was excluded and between the diff erent compartments of the co- from the analysis (top of the column with LMA me- lumns. Th e matrix of the fungal community and the dium placed in the Folic Histosol) because no sequence taxonomic identifi cation of the fungal ITS sequences passed the quality control. are presented in a Supplementary Table S1 and a sum-

51 Figure 4. Taxonomical breakdown of the bacterial OTUs richness present in the soil only (black), in the soil and in the top of the columns (dark gray), and in the columns only (light gray). Th e comparison is presented in percentage values but the total number of OTUs in each taxonomic group is indicated within the bars. Emphasis on the phyla with the highest proportion of OTUs in the columns (highlighted in the top) is presented in the lower part of the fi gure. Th e phylum Nitrospirae is only represented by the genus Nitrospira and was not included in the lower part. For the top and lower part of the fi gure, the three panels correspond to the three soil types indicated on the left.

mary of the results of the sequences obtained for the were present in the soil/columns from Cambisol, 1458 bacterial 16S rRNA gene is presented in a Supplemen- in the Fluvisol and 795 in the Histosol (Figure S2). tary Table S2. Relative changes in the distribution of diff erent fungal In order to account for biases linked to the type phyla in the three compartments (soil, middle and top of culture medium used inside the columns, we em- of the columns) are shown in Figure 2. Fungal OTUs ployed two diff erent culture media as attracting and richness decreased towards the top of the columns, target media: low-malt (LMA) and R2A agar. Howe- with an average of 259 OTUs in the soil, 25 OTUs in ver, since microbial community composition from the the middle of the columns and 5 OTUs in the top of same soil showed higher similarity than between soils, the columns. Amongst the total 692 fungal OTUs pre- sequences originating from the same soil were pooled sent in the three soils, 6.4 % were shared between the (Figure S1). Cambisol and the Fluvisol, 4.6 % between the Histo- Among the global 718 fungal OTUs, 266 were pre- sol and the Cambisol, 2.9 % between the Histosol and sent in the soil and/or columns from Cambisol, 372 the Fluvisol, and 1.7 % between the 3 soils. Amongst in the Fluvisol and 175 in the Histosol (Figure S2). the total 7 fungal OTUs present in the top of all the Among the global 3229 bacterial OTUs present, 1514 columns, 57.1 % were shared between the columns of

52 Figure 5. Heatmaps representing the relative abundance of the 20 most abundant bacterial OTUs in the soils (left) and at the top of the columns (right). Th e only common OTU to both datasets, assigned to Psychrobacillus genus, is in boxes in the fi gure.

Cambisol and Fluvisol, 42.9 % between the columns of the three soils. Acidobacteria, followed by diff erent of Histosol and Cambisol, 42.9 % between the co- classes of Proteobacteria, dominated the community lumns of Histosol and Fluvisol, and 42.9 % between composition in the three soils. Richness decreased dra- the columns of the three soils. Th e only fungal genus matically in the middle of the columns, with only 43, present at the top of the columns (with the exception 38 and 42 OTUs present in the columns from Cam- of one unclassifi ed OTU) was Mortierella spp., belon- bisol, Fluvisol and Histosol, respectively. Th is was ac- ging to the phylum Zygomycota (Figure 3). companied by a shift in community composition with Th e highest bacterial richness was also observed in the clear increase in the fraction of OTUs assigned to the soil compartment, with an average of 1124 OTUs. Bacilli. Richness of certain bacterial taxonomic groups Amongst the total 3893 bacterial OTUs present in the increased in the top of the columns. While in the top three soils, 4.0 % were shared between the Cambisol of Cambisol and Fluvisol, Proteobacteria dominated and the Fluvisol, 6.7 % between the Histosol and the the communities, in the top of Histosol, Firmicutes Cambisol, 1.1 % between the Histosol and the Fluvi- were clearly dominant. In contrast to the soil, Acido- sol, and 0.8 % between the three soils. Amongst the bacteria were poorly represented in the top of the co- total 283 bacterial OTUs present in the top of all the lumns (Figure 2). columns, 25.1 % were shared between the columns of A more detailed taxonomic breakdown of the com- Cambisol and Fluvisol, 7.8 % between the columns of position for bacteria in the soil and the top of the Histosol and Cambisol, 7.1 % between the columns of columns is presented in Figure 4. OTUs assigned to Histosol and Fluvisol, and 6.4 % between the columns the phyla Firmicutes, Nitrospira, Planctomycetes and

53 Proteobacteria corresponded to the highest proportion ciation in the soils (p-value ≤ 0.001) were not found at of OTUs present in the top of the columns. Within the the top of the columns (Table S4). phylum Firmicutes, the highest proportion of OTUs in the top of the columns corresponded to OTUs affi liated to the genera Bacillus and Clostridium. Most Discussion of these OTUs were present only in the top of the columns, and not in the soils. For the phylum Planc- Th e aim of this study was to assess the diversity of fun- tomycetes, the highest proportion of OTUs in the gi and bacteria engaging in fungal highways in compa- top of the columns corresponded to OTUs affi liated rison to the total microbial community in soils. Th is to the genus Gemmata. In the phylum Proteobacteria, was studied with the help of a device that selects for OTUs assigned to the genera Rhodobacter, Zymomonas, this type of interaction by attracting fungi towards two Burkholderia, Aquabacterium, Rhizobacter, Acinetobac- culture media, one of which (the top medium) can ter, Dokdonella, and Pseudofulvimonas represented a only be reached by fi lamentous organisms after cros- high proportion of the OTUs in the top of the co- sing an unsaturated barrier made of glass beads (Simon lumns, but were mostly absent from the soil. et al., 2015). Previously, we have observed a limited In total, 48 bacterial OTUs were present in the top diversity of culturable fungi able to reach the top of the columns, but not in the soils. Regarding their compartment of the columns. Th ose corresponded to taxonomic assignment, 22 of these OTUs belonged to diff erent strains of Fusarium spp. and one strain of the genera present in the soils, but 15 were assigned to ge- genus Chaeotomium. However, it was unclear if this nera absent from all the soils. Th e 11 remaining OTUs extreme selectivity was due to the culture conditions were unclassifi ed (Table S4). We verifi ed the depth used to recover the microorganisms colonizing the co- of sequencing in order to assess if one of the reasons lumns, or if it indeed refl ected a restricted diversity of for not detecting in the soils those OTUs common fungi able to reach the top compartment. Using now in the columns was insuffi cient coverage. Th e depth a culture-independent approach, we have observed a of sequencing in the three compartments appeared drastic decrease of fungal taxonomic richness in the to be adequate to describe the fungal communities target culture medium. All OTUs present in the top of according to the rarefaction curves obtained for the the columns belonged to the genus Mortierella (phy- three soil types. But this was not true for the bacterial lum Zygomycota), with the exception of one unidenti- communities (Figure S3), and thus the absence of a fi ed OTU (Figure 3). Mortierella spp. are saprotrophic corresponding OTU in soil can be due to insuffi cient fungi (Cannon & Kirk, 2007) that show moderate to sampling. fast-growing abilities (Yadav et al., 2015). Th e fact that Th e 20 most abundant bacterial OTUs in the soils fungi with a fast-growing life style reached preferen- and in the top of the columns are presented in Figure tially the top of the columns is not surprising, but it 5. In the soil, 13 out of the 20 OTUs belonged to the raises the question of the importance of these fungi phylum Acidobacteria. Th e other phyla were Firmi- in soils. Usually, mycorrhizal and white-rot fungi are cutes (2 OTUs), Nitrospirae (1 OTU), and Proteo- the focus of studies on the fungal activity in soils (Fol- bacteria (4 OTUs). Despite the fact that Acidobacteria man et al., 2008, Hoeksema & Classen, 2012, Kohl et were clearly dominant in all the soils, the most abun- al., 2016). Mortierellomycotina and Mucoromycotina dant OTUs diff ered depending on the soil type and (phylum Zygomycetes) represent 10.7 % in a recent this was particularly marked for the Fluvisol. In the assessment of the global soil fungal diversity (Tedersoo top of the columns, 5 OTUs belonged to phylum Fir- et al., 2014), but their contribution to soil function micutes, 11 to Proteobacteria, 3 to Acidobacteria, and remains poorly known. Various strains of Mortierella 1 OTU could not be identifi ed. Only one OTU was have been isolated from pesticide-contaminated soils common between both datasets: Psychrobacillus sp., and showed pesticide degradation potentials (Ellen- from the phylum Firmicutes. gaard-Jensen et al., 2013). Additionally, this genus Finally, we investigated the co-occurrence of fungal has been described as r-strategist (Allison et al., 2009, and bacterial OTUs in our dataset. Th e fungal and Brabcová et al., 2016). As soils are extremely complex bacterial OTUs that were preferentially found in asso- and constantly changing environments (Or et al.,

54 2007), fungi with r-strategies might play an important a translocation of the endosymbiont in the same way ecological role, but this has been poorly assessed so far. fungi translocate their organelles. Th e fact that hyphae Th e high selectivity exerted by the columns for speci- of Mortierella spp. are coenocytic (or barely septate) fi c fungal groups able to reach the top medium clearly might help this process. Th is is consistent with a pre- aff ected the structure of the bacterial communities as- vious study suggesting that endobacteria could be fa- sociated to those. Despite the fact that there was little vored by coenocytic hyphae (Desiro et al., 2014). Our to no overlap among the bacteria in the three soils stu- observations support this, and we propose that this died (1.7 % of the OTUs), the proportion of bacte- occurs because of the possible movement of bacteria rial OTUs shared among the top of the columns from inside the hyphae. In this case, endobacteria will use diff erent soils was higher (42.9 %). Th is indicates that fungal hyphae as ‘subways’ rather than highways for the bacterial community linked to fungal highways is their dispersal in soils. Here, we refer to subways-like more homogeneous than the total soil bacterial com- transport as the dispersal of bacteria within the fun- munity. gal cell, which is diff erent from a previous reference A strong selection occurred already at the bacterial to subways-like dispersal that considered the passive phylum level between the soils and the top of the co- movement of bacteria attached to the fungal cell wall lumns. While phyla Acidobacteria and Actinobacteria (Kohlmeier et al., 2005), which is not consistent with showed a high richness in the soils, they represented the apical growth of fi lamentous fungi (Gow 1995). a very small fraction of the OTUs in the top of the Whether other bacterial groups also detected exclu- columns. A higher proportion of Firmicutes, Nitros- sively in the top of the columns are taking the fungal pirae, Planctomycetes and Proteobacteria was obser- subway for dispersal needs to be studied in the future. ved at the top of the columns. Some genera such as However, it has been demonstrated that Proteobacte- Clostridium (Firmicutes), Rhodobacter and Zymomonas ria (Alpha-, Bêta- and Gammaproteobacteria), as well (Alphaproteobacteria), Burkholderia, Aquabacterium, as Firmicutes, can inhabit fungal hyphae (Hoff man & and Rhizobacter (Bêtaproteobacteria), Acinetobacter, Arnold, 2010), which is consistent with our observa- Dokdonella, and Pseudofulvimonas (Gammaproteobac- tions and suggest that a subways-like dispersal mecha- teria) were present only inside the columns, and not in nism might be widespread. Th e diversity and physiolo- the corresponding soils. Moreover, most of the OTUs gical role of bacterial endosymbionts inside fungal cells present in the top of the columns were not detected in is still poorly known (Jargeat et al., 2004). Subway-like the soils (Figure 4), and only one of the OTUs com- dispersal as a competitive ecological strategy could be mon in soils (Psychrobacillus sp., phylum Firmicutes) one of the factors favoring the endosymbiosis of bacte- could be observed reaching the top medium. ria inside fungal cells. Th e diff erences in the bacterial community composi- In a study in which attachment of bacteria to my- tion between the top of the columns and the soil need corrhizal fungi was used as a proxy for fungal-bacte- to be analyzed. Th e fact that Burkholderia spp. were rial interactions, the genera Bacillus and Acinetobacter found among the groups able to reach the top of the were considered as poorly or not interacting with fungi columns could off er a hint in one of the selection pro- (Scheublin et al., 2010), whereas our study suggests cess occurring in fungal highways. Indeed, Burkholde- that both genera disperse in association to fungal ria spp. are known for comprizing numerous species highways. Th is suggests that our approach provides that are endosymbiotic in fungi (de Boer et al., 2005, novel information on fungal-bacterial interactions. Sato et al., 2010). At the same time, Mortierella spp. Moreover, Bacillus spp. are often considered as antago- are known for hosting bacterial endosymbionts, espe- nist for fungi (Jacob et al., 2015, Torres et al., 2016). cially bacteria affi liated to Burkholderiales (Sato et al., However, it has been shown that the same Bacillus 2010). Th us, we hypothesize that part of the diversity strain can act as antagonist against one fungus, and of Burkholderia spp. present only in the top of the co- show a positive interaction with another (Liu et al., lumns corresponds to endosymbionts of Mortierella 2015). Th erefore, our results underpin the fact that spp. reaching the top of the columns. Th e fact that Bacilli should be considered as potentially interacting these bacteria were not recovered in the soil portion in many ways with fungi, and not only through com- in contact with the columns might be explained by petition.

55 Another aspect analyzed in this study was the eff ect bacterial dispersal (i.e. highways versus subways), and of organic carbon and nitrogen contents and C:N supports the fact that these fungi should be considered ratios in soils on the diversity of fungi and bacteria as crucial actors in soils. interacting through fungal highways (and potentially, subways). At a phylum level, the relative taxonomic composition was similar between the three soils, Acknowledgements for fungi and for bacteria (Figure 2). For fungi, this observation is consistent with the recently described We would like to thank Dylan Tatti and Prof. Jean-Mi- stochastic assembly processes of fungal communities chel Gobat (University of Neuchâtel, Switzerland) for present in upper soil layers (Powell et al., 2015). Des- the supply of the data concerning soils properties. Th is pite the fact that in this study we analyzed soils with research was supported by the Swiss National Science very diff erent organic carbon and nitrogen contents, Foundation through Grants FN CR22I2-137994/1 the same dominant genus was observed in the top and FN CR3212-149853/1. compartment of all the columns and thus its ecological role could be related to an edge for colonization of new environments regardless of the initial environ- References mental conditions. If the diversity of bacteria dispersed in association to this type of fungi is as important as Allison SD, LeBauer DS, Ofrecio MR, Reyes R, Ta suggested by our results, then the eff ects of soil pro- AM, Tran TM. (2009). Low levels of nitrogen addi- perties such as C:N ratio or pH on the composition of tion stimulate decomposition by boreal forest fungi. bacterial communities might be variable and the phy- Soil Biol Biochem 41(2):293-302. siological status of the fungal host might have a higher Anderson IC, Prosser JI, Campbell CD. (2003). Po- infl uence on the spatial distribution and abundance of tential bias of fungal 18S rDNA and internal trans- these bacterial groups in soils. Th is is in accordance cribed spacer polymerase chain reaction primers for with a recent network analysis showing that modules estimating fungal biodiversity in soil. Environ Mi- of co-occurring bacterial and fungal OTUs presented crobiol 5:36-47. contrasting relationship to various soil properties (de Baize D. (2009). Référentiel pédologique 2008. Asso- Menezes et al., 2015). ciation française pour l’étude du sol: Versailles. Bengtsson-Palme J, Ryberg M, Hartmann M et al. While bacterial diversity in soils is usually conside- (2013). Improved software detection and extraction red as being controlled by chemical warfare and li- of ITS1 and ITS2 from ribosomal ITS sequences of mited dispersal (Czaran et al., 2002, Reichenbach et fungi and other eukaryotes for analysis of environ- al., 2007), our study suggests that bacterial dispersal mental sequencing data. Methods Ecol Evol 4:914- along fungal hyphae can not be ignored. Moreover, as 919. this interaction involves preferentially some fungal and Boyle D. (1998). Nutritional factors limiting the bacterial phyla, it acts as an additional structuring fac- growth of Lentinula Edodes and other white-rot fun- tor of microbial communities in soils. Only one fungi in wood. Soil Biol Biochem 30:817-823. gal genus, Mortierella, was detected as dispersal agent Brabcová V, Nováková M, Davidová A, Baldrian P. for bacteria. Th e rapid growing of these fungi might (2016). Dead fungal mycelium in forest soil repre- explain this monopoly, but it does not exclude the im- sents a decomposition hotspot and a habitat for a plication of other fungi in this phenomenon. A way to specifi c microbial community. New Phytol. select for slower-growing fungi in soils is still required Bravo D, Cailleau G, Bindschedler S, Simon A, Job D, to assess their implication in bacterial dispersal, and Verrecchia E, Junier P. (2013). Isolation of oxalotro- thus microbial community assembly. Th e variety of phic bacteria able to disperse on fungal mycelium. bacterial genera present only in the top of the columns, FEMS Microbiol Lett 348:157-66. but not in the soils brings a novel perspective on the Cannon PF, Kirk PM. (2007). Fungal families of the indirect metabolic activities that could be favored by world. CAB International: Wallingford. r-strategists fungi, independently of the mechanism of Czaran TL, Hoekstra RF, Pagie L. (2002). Chemical

56 warfare between microbes promotes biodiversity. Bacillus species as an intrinsic controller of fungal Proc Natl Acad Sci USA 99: 786-790. deterioration of archival documents. International de Boer W, Folman LB, Summerbell RC, Boddy L. Biodeter Biodegr 104:46-52. (2005). Living in a fungal world: impact of fungi on Jargeat P, Cosseau C, Ola’h B, Jauneau A, Bonfante soil bacterial niche development. FEMS Microbiol P, Batut J, Becard G. (2004). Isolation, free-living Rev 29:795-811. capacities, and genome structure of «Candidatus de Menezes AB, Prendergast‐Miller MT, Richardson Glomeribacter gigasporarum,» the endocellular bac- AE et al. (2015). Network analysis reveals that bacterium of the mycorrhizal fungus Gigaspora margari- teria and fungi form modules that correlate inde- ta. J Bacteriol 186:6876-6884. pendently with soil parameters. Environ Microbiol Kobayashi DY, Crouch JA. (2009). Bacterial/fungal 17(8):2677-2689. interactions: from pathogens to mutualistic endo- Demoling F, Figueroa D, Bååth E. (2007). Compari- symbionts. Annu Rev Phytopathol 47:63-82. son of factors limiting bacterial growth in diff erent Kohl L, Lukasiewicz CE, van der Heijden MG. (2016). soils. Soil Biol Biochem 39:2485-2495. Establishment and eff ectiveness of inoculated arbus- Desiro A, Salvioli A, Ngonkeu EL et al. (2014). De- cular mycorrhizal fungi in agricultural soils. Plant tection of a novel intracellular microbiome hosted Cell Environ 39:136-146. in arbuscular mycorrhizal fungi. ISME J 8:257-270. Kohlmeier S, Smits TH, Ford R, Keel C, Harms H, Ellegaard-Jensen L, Aamand J, Kragelund BB, Johnsen Wick LY. (2005). Taking the fungal highway: mo- AH, Rosendahl S. (2013). Strains of the soil fungus bilization of pollutant-degrading bacteria by fungi. Mortierella show diff erent degradation potentials for Environ Sci Technol 39:4640-4646. the phenylurea herbicide diuron. Biodegradation Köljalg U, Nilsson RH, Abarenkov K et al. (2013). 24(6):765-774. Towards a unifi ed paradigm for sequence-based Folman LB, Klein Gunnewiek PJ, Boddy L, de Boer identifi cation of fungi. Mol Ecol 22:5271-5277. W. (2008). Impact of white-rot fungi on numbers Liesack W, Weyland H, Stackebrandt E. (1991). Po- and community composition of bacteria colonizing tential risks of gene amplifi cation by PCR as deter- beech wood from forest soil. FEMS Microbiol Ecol mined by 16S rDNA analysis of a mixed-culture of 63:181-191. strict barophilic bacteria. Microb Ecol 21:191-198. Frey-Klett P, Burlinson P, Deveau A, Barret M, Tarkka Liu C, Sheng J, Chen L et al. (2015). Biocontrol acti- M, Sarniguet A. (2011). Bacterial-fungal interac- vity of Bacillus subtilis isolated from Agaricus bispo- tions: hyphens between agricultural, clinical, en- rus mushroom compost against pathogenic fungi. J vironmental, and food microbiologists. Microbiol Agric Food Chem 63:6009-6018. Mol Biol Rev 75:583-609. Martin G, Guggiari M, Bravo D et al. (2012). Fungi, Gow NAR. (1995). Tip growth and polarity. In Gow bacteria and soil pH: the oxalate-carbonate pathway NAR & Gadd GM (eds). Th e growing fungus. as a model for metabolic interaction. Environ Mi- Chapman & Hall: London, UK, pp 277-299. crobiol 14:2960-2970. Hoeksema JD, Classen AT. (2012). Is plant genetic Or D, Smets BF, Wraith JM, Dechesne A, Friedman control of ectomycorrhizal fungal communities an SP. (2007). Physical constraints aff ecting bacterial untapped source of stable soil carbon in managed habitats and activity in unsaturated porous media – forests? Plant Soil 359:197-204. a review. Adv Water Resour 30:1505-1527. Hoff man MT, Arnold AE. (2010). Diverse bacteria in- Powell JR, Karunaratne S, Campbell CD, Yao H, Ro- habit living hyphae of phylogenetically diverse fun- binson L, Singh BK. (2015). Deterministic processes gal endophytes. Appl Environ Microbiol 76:4063- vary during community assembly for ecologically 4075. dissimilar taxa. Nat Commun 6:8444. Jabiol B, Zanella A, Ponge J-F et al. (2013). A proposal Quast C, Pruesse E, Yilmaz P et al. (2013). Th e SIL- for including humus forms in the World Reference VA ribosomal RNA gene database project: improved Base for Soil Resources (WRB-FAO). Geoderma data processing and web-based tools. Nucleic Acids 192:286-294. Res 41:590-596. Jacob SM, Bhagwat AM, Kelkar-Mane V. (2015). R Core Team. (2013). R: A language and environ-

57 ment for statistical computing. R Foundation for Strickland MS, Rousk J. (2010). Considering fungal:- Statistical Computing, Vienna, Austria. URL http:// bacterial dominance in soils – Methods, controls, www.R-project.org/. and ecosystem implications. Soil Biol Biochem Rappe MS, Giovannoni SJ. (2003). Th e uncultured 42:1385-1395. microbial majority. Annual Rev Microbiol 57:369- Tedersoo L, Bahram M, Põlme S et al. (2014). Global 394. diversity and geography of soil fungi. Science 346. Reichenbach T, Mobilia M, Frey E. (2007). Mobility Torres MJ, Brandan CP, Petroselli G, Erra-Balsells R, promotes and jeopardizes biodiversity in rock-pa- Audisio MC. (2016). Antagonistic eff ects of Bacillus per-scissors games. Nature 448:1046-1049. subtilis subsp. subtilis and B. amyloliquefaciens against Rice EW, Baird RB, Eaton AD, Clesceri LS. (2012). Macrophomina phaseolina: SEM study of fungal Standard methods for the examination of water and changes and UV-MALDI-TOF MS analysis of their wastewater, 22nd Edition. American Public Health bioactive compounds. Microbiol Res 182:31-39. Association, American Water Works Association, Veech JA. (2013). A probabilistic model for analy- Water Pollution Control Federation: Washington. sing species co‐occurrence. Global Ecol Biogeogr Ritz K, Young IM. (2004). Interactions between soil 22:252-260. structure and fungi. Mycologist 18:52-59. Wang Q, Garrity G, Tiedje J, Cole J. (2007). Naive Sato Y, Narisawa K, Tsuruta K et al. (2010). Detec- Bayesian classifi er for rapid assignment of rRNA se- tion of Betaproteobacteria inside the Mycelium of quences into the new bacterial taxonomy. Appl En- the Fungus Mortierella elongata. Microbes Environ viron Microbiol 73:5261-5267. 25:321-324. Waring BG, Averill C, Hawkes CV. (2013). Diff erences Scheublin TR, Sanders IR, Keel C, van der Meer JR. in fungal and bacterial physiology alter soil carbon (2010). Characterisation of microbial communities and nitrogen cycling: insights from meta-analysis colonising the hyphal surfaces of arbuscular mycor- and theoretical models. Ecol Lett 16:887-894. rhizal fungi. ISME J 4:752-763. Wells JM, Boddy L, Evans R. (1995). Carbon translo- Schloss P, Gevers D, Westcott S. (2011). Reducing the cation in mycelial cord systems of Phanerochaete velu- eff ects of PCR amplifi cation and sequencing arti- tina (DC: Pers.) Parmasto. New Phytol 129(3):467- facts on 16S rRNA-based studies. PLOS One 6. 476. Schloss P, Westcott S, Ryabin T et al. (2009). Introdu- Wells JM, Boddy L. (1995). Phosphorus translocation cing mothur: open-source, platform-independent, by saprotrophic basidiomycete mycelial cord systems community-supported software for describing and on the fl oor of a mixed deciduous woodland. Mycol comparing microbial communities. Appl Environ Res 99(8):977-980. Microbiol 75:7537-7541. WRB IWG. (2014). World Reference Base for Soil Schütte KH. (1956). Translocation in the fungi. New Resources 2014. Vol. 106. World Soil Resources Re- Phytol 55:164-182. ports: Rome. Simon A, Bindschedler S, Job D et al. (2015). Exploi- Yadav DR, Kim SW, Adhikari M et al. (2015) Th ree ting the fungal highway: development of a novel tool new records of Mortierella species isolated from crop for the in situ isolation of bacteria migrating along fi eld soil in Korea. Mycobiology 43:203-209. fungal mycelium. FEMS Microbiol Ecol 91.

58 Chapter 2 - Culture-independent approach

Supplementary information

Figure S1. UPGMA clustering of the soil bacterial (top) Figure S3. Rarefaction curves for fungal and bacterial and fungal (bottom) communities based on Jaccard dis- communities in the three soils. tances showing a higher similarity of the communities within a specifi c soil type, independently of the composition of the medium in the columns.

Figure S2. Venn diagrams representing the number of specifi c and shared OTUs among the three compartments (soil, middle of the columns, top of the columns) for each soil and for both bacterial and fungal communities.

59 00) 1 a( ll e nt e m o T 00) g 1 aceae( r o h ep l e fTh similarity). es l a r o h ep l e Th es t yce m co ri ga A a t yco m o i mycota Incertae_sedis Mortierellales f__Mortierellaceae(100) g__Mortierella(100) d i as iB g n ungiungiungi Ascomycotaungi Basidiomycotaungi Basidiomycotaungi Basidiomycotaungi Sordariomycetes Agaricomycetes Zygomycotaungi Microbotryomycetes Basidiomycotaungi Agaricomycetes Zygomycota Zygomycota Incertae_sedis Hypocreales Sporidiobolales Agaricomycetes Ascomycota Russulales Incertae_sedis Russulales Incertae_sedis Dothideomycetes Thelephorales Mortierellales Unclassified Incertae_sedis Mortierellales f__Russulaceae(100) Mortierellales Incertae_sedis f__Russulaceae(100) f__Thelephoraceae(100) f__Mortierellaceae(100) f__Mortierellaceae(100) f__Myxotrichaceae(100) f__Mortierellaceae(100) g__Russula(100) Unclassified g__Ilyonectria(100) g__Tomentella(100) Unclassified g__Mortierella(100) g__Mortierella(100) g__Oidiodendron(100) g__Mortierella(100) ungiungiungi Ascomycota Basidiomycota Zygomycota Leotiomycetes Agaricomycetes Incertae_sedis Russulales Helotiales Mortierellales f__Russulaceae(100) Incertae_sedis f__Mortierellaceae(100) g__Lactarius(100) g__Mortierella(100) g__Tetracladium(100) ungi Zygomycota Incertae_sedis Mortierellales f__Mortierellaceae(100) g__Mortierella(100) ungiungi Zygomycotaungi Ascomycotaungiungi Ascomycotaungi Incertae_sedis Rozellomycotaungi Dothideomycetes Ascomycotaungi Ascomycotaungi Sordariomycetes Unclassified Zygomycotaungi Ascomycotaungi Mortierellales Capnodiales Leotiomycetes Ascomycotaungi Sordariomycetes Ascomycotaungi Incertae_sedis Hypocreales Basidiomycotaungi Sordariomycetes Ascomycotaungi Unclassified Leotiomycetes Zygomycota Helotiales Sordariomycetes Hypocreales Tremellomycetes Basidiomycotaungi f__Mortierellaceae(100) f__Mycosphaerellaceae(100) Ascomycotaungi Mortierellales Eurotiomycetes Coniochaetalesungi Incertae_sedis Agaricomycetes f__Nectriaceae(100) Zygomycotaungi Helotiales Hypocreales Cystofilobasidiales Basidiomycotaungi Leotiomycetes g__Mycosphaerella(100) Unclassified Basidiomycotaungi Incertae_sedis Zygomycota Eurotiales g__Mortierella(100) ungi Incertae_sedis Agaricomycetes f__Helotiaceae(100) f__Coniochaetaceae(100) Ascomycota Athelialesungi Mortierellales f__Mortierellaceae(100) Tremellomycetes f__Cystofilobasidiaceae(100) Ascomycota Rozellomycotaungi Helotiales Incertae_sedis f__Nectriaceae(100) g__Volutella(100) Zygomycotaungi Leotiomycetes f__Helotiaceae(100) Agaricalesungi Mortierellales Sordariomycetes Unclassified Filobasidiales g__Guehomyces(100) g__Coniochaeta(100) Unclassifiedungi Basidiomycota g__Mortierella(100) f__Trichocomaceae(100)ungi f__Mortierellaceae(100) Incertae_sedis g__Tricladium(100) Zygomycotaungi f__Atheliaceae(100) Mortierellales Unclassified g__Ilyonectria(100) Ascomycotaungi Helotiales Unclassified Hypocreales Tremellomycetes Ascomycotaungi f__Vibrisseaceae(100) g__Neonectria(100) Ascomycotaungi Unclassified f__Mortierellaceae(100) Incertae_sedis f__Filobasidiaceae(100) Unclassified Basidiomycota f__Clavariaceae(100)ungi Mortierellales Eurotiomycetes g__Penicillium(100) Ascomycota g__Mortierella(100) ungi Pezizomycetes Filobasidiales Ascomycotaungi f__Mortierellaceae(100) Sordariomycetes Tremellomycetes Unclassified Unclassifiedungi g__Piloderma(100) Incertae_sedis Ascomycotaungi Mortierellales Sordariomycetes Incertae_sedis g__Phialocephala(100) Basidiomycota Chaetothyriales g__Mortierella(100) ungi g__Cryptococcus(100) Sordariomycetes Unclassified Ascomycota Pezizalesungi f__Mortierellaceae(100) Unclassified Xylariales Trichosporonales Unclassified Ascomycotaungi Eurotiomycetes f__Filobasidiaceae(100) Agaricomycetes Ascomycota g__Mortierella(100) ungi Hypocreales Zygomycotaungi Leotiomycetes Coniochaetales Unclassified Ascomycota f__Herpotrichiellaceae(100)ungi f__Mortierellaceae(100) Sordariomycetes Ascomycotaungi f__Trichosporonaceae(100) Eurotiomycetes Zygomycota Chaetothyriales Agaricales g__Mortierella(100) Unclassifiedungi Incertae_sedis g__Acremonium(100) Ascomycotaungi g__Cryptococcus(100) Leotiomycetes f__Sarcosomataceae(99) g__Leptodontidium(100) f__Hyponectriaceae(100) Ascomycotaungi Helotiales Leotiomycetes Hypocreales g__Exophiala(100) Unclassified f__Hypocreaceae(100) Basidiomycotaungi Incertae_sedis Unclassified g__Trichosporon(100) Zygomycota Eurotiales g__Mortierella(100) ungi Saccharomycetes Ascomycota Unclassified Mortierellales Eurotiomycetes Agaricomycetes Basidiomycota Helotiales Unclassified g__Strumella(99) Unclassified f__Clavariaceae(100) Unclassified Helotiales Incertae_sedis Unclassified Saccharomycetales Mortierellales Dothideomycetes Tremellomycetes Incertae_sedis g__Trichoderma(100) Incertae_sedis Eurotiales Thelephorales Unclassified f__Trichocomaceae(100) f__Mortierellaceae(100) Unclassified Mortierellales Incertae_sedis Cystofilobasidiales Unclassified Incertae_sedis g__Ramariopsis(100) f__Mortierellaceae(100) f__Dermateaceae(100) Unclassified f__Thelephoraceae(100) Unclassified Unclassified g__Paecilomyces(100) Incertae_sedis g__Mortierella(100) f__Trichocomaceae(100) g__Cephalosporium(100) f__Myxotrichaceae(100) f__Mortierellaceae(100) g__Helicodendron(100) g__Mortierella(100) Unclassified Unclassified Unclassified g__Tetracladium(100) g__Penicillium(100) Unclassified g__Oidiodendron(100) g__Mortierella(100) g__Mrakiella(100) Unclassified ungiungiungi Ascomycotaungi Ascomycotaungi Ascomycotaungi Basidiomycotaungi Eurotiomycetes Basidiomycotaungi Leotiomycetes Basidiomycotaungi Eurotiomycetes Agaricomycetes Basidiomycotaungi Agaricomycetes Ascomycotaungi Agaricomycetes Ascomycota Chaetothyrialesungi Agaricomycetes Unclassifiedungi Helotiales Basidiomycota Chaetothyriales Auricularialesungi Eurotiomycetes Ascomycota Russulalesungi Dothideomycetes Ascomycota Russulalesungi Unclassified Agaricomycetes Ascomycota Agaricales f__Herpotrichiellaceae(100)u Ascomycota Sordariomycetes Ascomycota Eurotiales f__Herpotrichiellaceae(100) Pleosporales Unclassified Unclassified Leotiomycetes f__Helotiaceae(97) Cantharellales Sordariomycetes g__Phialophora(100) f__Auriscalpiaceae(100) Unclassified Sordariomycetes Hypocreales f__Russulaceae(100) g__Rhinocladiella(100) f__Psathyrellaceae(100) Unclassified Helotiales Incertae_sedis Hypocreales f__Trichocomaceae(100) Hypocreales f__Ceratobasidiaceae(100) g__Clavicorona(100) g__Articulospora(97) Unclassified f__Nectriaceae(100) g__Russula(100) g__Coprinellus(100) Unclassified Unclassified g__Penicillium(100) Unclassified f__Hypocreaceae(100) Unclassified f__Nectriaceae(100) g__Didymella(100) g__Flagellospora(100) g__Trichoderma(100) Unclassified g__Fusicolla(98) Unclassified Unclassified ungi Ascomycota Leotiomycetes Helotiales Incertae_sedis g__Tetracladium(100) ungi Rozellomycota Unclassified Unclassified Unclassified Unclassified ungiungi Zygomycota Zygomycotaungi Incertae_sedisungi Incertae_sedis Ascomycota Basidiomycota Mortierellales Eurotiomycetes Agaricomycetes Mortierellales Eurotiales Russulales f__Mortierellaceae(100) f__Mortierellaceae(100) g__Mortierella(100) f__Russulaceae(100) f__Elaphomycetaceae(100) g__Mortierella(100) g__Elaphomyces(100) g__Russula(100) ungi Zygomycota Incertae_sedis Mortierellales f__Mortierellaceae(100) g__Mortierella(100) ungi Ascomycota Leotiomycetes Helotiales Unclassified Unclassified ungi Ascomycota Eurotiomycetes Eurotiales f__Elaphomycetaceae(100) g__Elaphomyces(100) Fungi Ascomycota Leotiomycetes Helotiales Unclassified Unclassified F R2A Kingdom Phylum Class Order Family Genus Histosol soil R2A Histosol top Histosol middle R2A 9111 3 0 1355 Fungi Zygomycota Incertae_sedis Mortierellales f__Mortierellaceae(100) g__Mortierella(100) LMA 1665 4 0 32 Fungi Zygomycota Incertae_sedis Mortierellales f__Mortierellaceae(100) g__Mortierella(100) Histosol soil 00001F 01000F 02000F 00000F 02000F 00000F 00000 3747 448 2 0 70 Fungi Basidiomycota Microbotryomycetes Leucosporidiales f__Leucosporidiaceae(100) g__Leucosporidium(100) Histosol middle LMA 7 R2A Fluvisol soil R2A Fluvisol top Fluvisol middle R2A LMA 1211 7 0 760 1863 1 0 216 1297 9 0 1042 1457 0 0 1475 3864 4666 1 3569 2985 17 2 492 Fluvisol soil LMA Fluvisol top 00000000000F 2802400230931609965FungiZygo Fluvisol middle LMA soil R2A Cambisol top R2A Cambisol Cambisol middle R2A soil LMA Cambisol top LMA Cambisol . Summary of the sequencing results for fungi, and taxonomic assignment using the UNITE database (assignment at a level of 80 % assignment using the UNITE database (assignment at a level for fungi, and taxonomic of the sequencing results . Summary 000000005280043300000F 0000000035003 Cambisol middle LMA 7 u009 t OTUs Otu0020 Otu002100100000000007800784F Otu0022001930016700700006000421F Otu00230000000000000700794F Otu0024001230027800000093360048F Otu0025000000004500032600000F Otu0026000000006390011000000F Otu00275840024001009000040F Otu0028004230013500000200000F Otu00180020900127300000001000F Otu001900000060881094700000F Otu001700400800 Otu002900100000000075060024F O Otu003000000000451008100000F Otu0031Otu0032000000002510022400000F Otu0033003630010600000000000F Otu00340000000043100700000F 0Otu0035000005002530016800000F Otu003600000000341033300000F Otu0037000000001510020700000F Otu00380000000000000800346F Otu00390021001100232208800000F 0Otu0040200000000000582704020F Otu004100231009100000000000F Otu004200000000860023100000F Otu00430000000000000264000F Otu00440000000000000000255F 85Otu0045Otu0046000000142022007700000F Otu004700000000000001100225F Otu004800500500100018432007F 0 0Otu004910000010000016267000F Otu00500000000000000700222F Otu005100200000172004900000F Otu005200150006900000000000F Otu0053006515013200000000000F 0 0Otu0054008300270000000460051 Otu005500000000820010700000F Otu005600000000000011170006F Otu0057100141202400000000000F Otu005800134003400200500000F 64 21Otu00590029004500500016080025F Otu00600000000092007100000F Otu006120490000052005600000F Otu00620000020074007400000F 0 0Otu006300000000100004500000F Otu00640000000073006200000F Otu00650013200000000000000F Otu00660000000000000200127F Otu00670000000096003300000F 0 0Otu006800000000111001100000F Otu0069004300220040001500000F Otu00700000000011900000000F Otu00710000000053006500000F 136Otu00720072004400000000000F 0Otu00730020020091002100000F Otu00740000000000000470066F Otu00750073003600000000000F Otu00760000000058004800000F 0Otu007750027200800000000000F 0Otu0078003000500160020000033F Otu00790000000022207700000F Otu008000000000160050210059F Otu00810054004500000000000F 0 0 162 65 0 0 12 0 0 165 0 0 18 0 Fungi Ascomycota 0 Sordariomycetes 0 Hypocreales 0 Fungi Unclassified Ascomycota Sordariomycetes Hypocreales Unclassified Incertae_sedis g__Stilbella(100) t08070018 Otu00820000000000000210077F Otu0083009600000000000000F Otu00840078001700000000000F Otu00850000000061003200000F Otu0086009200000000000000F Otu00870000000055003500000F Otu008800720 Otu00890040000051003400000F Otu0090000000008600000000F Otu00910000000024006000000F Otu00920000000054002800000F Otu00930000000047003400000F Otu00940000000016006300000F Otu00950000000030004400000F Otu00960040017004500600000F Otu0001 4390 3 960Otu00152054147072000100269 Otu001600000000 21556 34 8238 Otu00022025986021360020013441 Otu001400000000 Otu0013001300120000000184002222F Otu00100030300010900000000000F Otu0005Otu000610001010690 Otu0007500000100000 Otu0008 576Otu00090000000000000242003360F Otu00110027530034500000001002F 0Otu001200100000 0 0 147 3041 45 5 0 32 0 12 2 2 4 836 0 87 167 8 532 745 0 16 467 16 761 0 1500 0 2 0 0 Fungi 204 Zygomycota Fungi Incertae_sedis Zygomycota Incertae_sedis Mortierellales Mortierellales f__Mortierellaceae(100) f__Mortierellaceae(100) g__Mortierella(100) g__Mortierella(100) Otu000347052238041214034087931672712000F Otu00041046003910 Table S1 Table

60 ungiungiungi Basidiomycotaungi Unclassifiedungi Ascomycotaungi Wallemiomycetes Zygomycotaungi Basidiomycotaungi Unclassified Ascomycotaungi Sordariomycetes Ascomycotaungi Incertae_sedis Geminibasidiales Agaricomycetes Ascomycotaungi Ascomycotaungi Eurotiomycetes Ascomycotaungi Eurotiomycetes Xylariales Unclassified Unclassifiedungi Leotiomycetes Zygomycota Thelephorales f__Geminibasidiaceae(100)ungi Mortierellales Dothideomycetes Ascomycotaungi Sordariomycetes Ascomycota Chaetothyrialesungi Unclassified Zygomycota Chaetothyrialesungi Incertae_sedis Zygomycotaungi Rhytismatales Pleosporales Sordariomycetes g__Geminibasidium(100) Zygomycotaungi Leotiomycetes Hypocreales Unclassified f__Thelephoraceae(100) f__Amphisphaeriaceae(100) Basidiomycotaungi f__Mortierellaceae(100) Incertae_sedis Basidiomycota Unclassifiedungi Incertae_sedis Ascomycota Unclassified Unclassifiedungi Mortierellales Incertae_sedis Sordariales Agaricomycetes Basidiomycotaungi Tremellomycetes Ascomycota g__Seimatosporium(100) f__Rhytismataceae(100)ungi Helotiales Incertae_sedis Ascomycotaungi Unclassified Mortierellales Dothideomycetes Agaricomycetes Incertae_sedis Unclassified g__Mortierella(100) ungi Mortierellales Unclassified Thelephoralesungi Mortierellales Sordariomycetes Tremellales Basidiomycotaungi f__Mortierellaceae(100) Sordariomycetes Unclassified Ascomycota f__Chaetomiaceae(100)ungi Capnodiales Unclassified Unclassified g__Hypohelion(100) Ascomycota Agaricalesungi Unclassified Tremellomycetes Ascomycota Unclassified ungi f__Mortierellaceae(100) Sordariales Incertae_sedis Basidiomycota Unclassified ungi f__Mortierellaceae(100) Lecanoromycetes Hypocreales f__Thelephoraceae(100) Ascomycota f__Mortierellaceae(100) Sordariomycetes g__Pyrenochaeta(100) Zygomycota g__Mortierella(100) ungi Incertae_sedis Dothideomycetes Microbotryomycetes Trichosporonales g__Ilyonectria(100) g__Humicola(100) Ascomycota Unclassifiedungi f__Mycosphaerellaceae(100) Unclassified Lecanoralesungi Eurotiomycetes Ascomycota f__Psathyrellaceae(100) g__Mortierella(100) ungi Incertae_sedis Sordariales Leucosporidiales Ascomycota g__Mortierella(100) f__Lasiosphaeriaceae(100)ungi Unclassified Venturiales Sordariomycetes Unclassified f__Hypocreaceae(100) Unclassified g__Mortierella(100) f__Trichosporonaceae(100) g__Mycosphaerella(100) Basidiomycotaungi Eurotiomycetes g__Cadophora(100) Unclassified Eurotialesungi Leotiomycetes Unclassifiedungi Mortierellales f__Leucosporidiaceae(100) Unclassified Hypocreales Tremellomycetes f__Parmeliaceae(100) g__Psathyrella(100) Unclassified g__Podospora(81) Unclassified g__Cryptococcus(100) ungi g__Trichosporon(100) Ascomycota f__Lasiosphaeriaceae(100)ungi Unclassified Basidiomycota Eurotiales g__Trichoderma(100) f__Venturiaceae(100)ungi Zygomycotaungi Helotiales Unclassified Cystofilobasidiales Unclassified g__Leucosporidiella(100) ungi Leotiomycetes Tremellomycetes Basidiomycota f__Trichocomaceae(100) Unclassifiedungi f__Mortierellaceae(100) f__Bionectriaceae(100) g__Zopfiella(100) Zygomycota g__Parmelia(100) ungi Incertae_sedis Ascomycota Unclassifiedungi Unclassified Microbotryomycetes f__Cystofilobasidiaceae(100) Ascomycotaungi Unclassified Cystofilobasidiales Unclassified Zygomycota f__Trichocomaceae(100) Unclassifiedungi Incertae_sedis Incertae_sedis Unclassified Basidiomycotaungi Orbiliomycetes Heterogastridiales Incertae_sedis g__Penicillium(100) Ascomycota g__Mortierella(100) ungi Mortierellales Dothideomycetes Unclassified g__Mrakia(100) g__Bionectria(100) Unclassifiedungi Incertae_sedis Tremellomycetes Incertae_sedis Basidiomycota Unclassifiedungi Unclassified Unclassifiedungi Mortierellales Sordariomycetes g__Penicillium(100) Incertae_sedis f__Heterogastridiaceae(100) Basidiomycota Orbilialesungi Capnodiales Unclassified Agaricomycetes Unclassified Basidiomycotaungi Filobasidiales Unclassifiedungi Mortierellales f__Mortierellaceae(100) Unclassified Agaricomycetes Unclassifiedungi Hypocreales Microbotryomycetes Ascomycota g__Heterogastridium(100) ungi g__Infundichalara(100) Unclassified Unclassified Agaricalesungi f__Mortierellaceae(100) Unclassified Ascomycota Unclassifiedungi Unclassified Unclassified Sporidiobolales f__Mycosphaerellaceae(100) g__Syzygospora(100) Zygomycota Cantharellalesungi Leotiomycetes f__Filobasidiaceae(100) Unclassified Unclassified g__Mortierella(100) Unclassifiedungi f__Mortierellaceae(100) Unclassified Unclassified g__Meliniomyces(100) Unclassifiedungi Unclassified Incertae_sedis Ascomycotaungi Incertae_sedis Unclassified g__Mycosphaerella(100) Ascomycota g__Mortierella(100) Unclassifiedungi Unclassified Incertae_sedis Basidiomycota f__Clavariaceae(100) Unclassifiedungi Helotiales Unclassified Unclassified f__Ceratobasidiaceae(100) Rozellomycotaungi g__Cryptococcus(100) Dothideomycetes Zygomycota g__Mortierella(100) Unclassifiedungi Dothideomycetes Unclassified Tremellomycetes Unclassified Ascomycota Unclassifiedungi Mortierellales Fungi Unclassified Unclassified Unclassified Unclassified Unclassified 00000000F 0000000016001000000F Otu0110 0 0 30 0 0 16 0 0 11 Otu01610000000014001700000F Otu0111000000009004800000F Otu01120000000000000150039F Otu0113000000000000044009F Otu01140000000027002600000F Otu0115005200000000000000F Otu0116004800300000000000F Otu0117005000000000000000F Otu01180000000038001200000F Otu01190000000028002100000F Otu01200000000000000380010F Otu01211500320000000000000F Otu01220030010025001700000F Otu01230000000026001900000F Otu0124000004400000000000F Otu01250010000012003100000F Otu0126000000000004400000F Otu0127000000000000020042F Otu0128000000004100300000F Otu01290000000017002600000F Otu0130000000009003300000F Otu0131000000003400800000F Otu01320016002100000500000F Otu01330000000018002300000F Otu01340000000021002000000F Otu01350021000000000317000F Otu01360018002100000002000F Otu01370000000013002700000F Otu0138000000000000020038F Otu0139000000000000000039F Otu01400025001400000000000F Otu0141000000000003900000F Otu01420016002300000000000F Otu0143000000003300400000 Otu0144002800800000001000F Otu01450020001700000000000F Otu0146003000600000000000F Otu01470000000000023013000F Otu0148000000000003600000F Otu0149Otu0150000000008002700000F Otu0151000000000000070027F Otu0152009002500000000000F 0Otu0153000000000000034000F Otu0154000000003400000000F Otu0155000000004002900000F Otu0156000000002900400000F Otu0157003200000000000000F 0Otu0158000000000000090023F Otu01590000000000000120020F Otu0160000000000000818005F Otu0162000000003100000000F 11Otu0163000000001002900000F Otu01640017001200000000000F Otu0165000000002000900000F Otu0166000000008002100000F 0Otu0167002200600000000000F Otu0168000000000002800000F Otu0169000000002800000000F Otu0170002300400000000000F Otu01710010001700000000000F 0Otu0172000000000000010026F Otu0173001800900000000000F Otu01740000000017001000000F Otu0175000000002600000000F Otu01760000000013001300000F 4Otu0177 Otu0178000000000000060020F Otu0179000000002300300000F Otu0180006001600300100000F Otu01810000000016001000000F 0Otu0182000000000000000026F Otu01830013000000000012000F Otu0184000000000000020023F Otu01850012001300000000000F 0Otu018600900700000001008F Otu0187004005005001000000F Otu0188000000001700700000F Otu0189000000002400000000F Otu01900000000013001100000F 5Otu0191000000009001500000F Otu0192002200100000000000F Otu0193000000002300000000F Otu0194000000000000022001F Otu0195000000000002300000F 0Otu0196000000000002300000F Otu0197000000007001600000F Otu0198007001600000000000F Otu0199000000000000040019F Otu0200002300000000000000F 0Otu0201002000000000002000F Otu0202001800400000000000F Otu0203000000001500600000F Otu02040011001000000000000F 15Otu0205000000000002100000F Otu0206001500600000000000F Otu0207001300100000700000F Otu0208000000002100000000F Otu0209001700400000000000F 0 0 0 0 0 Fungi Basidiomycota Tremellomycetes Filobasidiales f__Filobasidiaceae(100) g__Cryptococcus(100)

61 ) 100 ( ocrea yp H )g 100 ( ocreaceae yp ocreales f H yp cetes H y cota Sordariom y i Ascom g ungiungiungi Ascomycotaungi Ascomycotaungi Basidiomycotaungi Ascomycotaungi Pezizomycetes Ascomycotaungi Sordariomycetes Ustilaginomycetes Ascomycotaungi Unclassifiedungi Dothideomycetes Ascomycotaungi Eurotiomycetes Basidiomycota Pezizales Urocystidalesungi Unclassified Hypocreales Ascomycotaungi Unclassified Unclassifiedungi Capnodiales Sordariomycetes Microbotryomycetes Basidiomycotaungi Basidiomycota Eurotialesungi Leotiomycetes Ascomycotaungi Unclassified Sporidiobolales Agaricomycetes Ascomycota Unclassified f__Urocystidaceae(100)ungi Hypocreales Agaricomycetes f__Nectriaceae(100) Zygomycota Unclassifiedungi f__Tuberaceae(100) Ascomycotaungi Sordariomycetes Unclassified Basidiomycotaungi Helotiales Eurotiomycetes Ascomycota Agaricalesungi Incertae_sedis Incertae_sedis Unclassified Boletales f__Trichocomaceae(100) Unclassifiedungi Sordariomycetes Agaricomycetes Ascomycota g__Urocystis(100) ungi Hypocreales Unclassified f__Nectriaceae(100) Ascomycotaungi Sordariomycetes Unclassified Unclassified Ascomycota Onygenalesungi Unclassified g__Tuber(100) Unclassifiedungi Mucorales Sordariomycetes Hypocreales Ascomycota Agaricalesungi Sordariomycetes Unclassified g__Eupenicillium(100) Ascomycota f__Entolomataceae(100)ungi Sordariomycetes Xylariales Unclassified Unclassifiedungi Unclassified f__Paxillaceae(100) Unclassified Unclassified Zygomycotaungi Leotiomycetes Hypocreales g__Neonectria(100) g__Sporobolomyces(89) Ascomycota Unclassifiedungi Leotiomycetes Incertae_sedis Hypocreales Ascomycotaungi Unclassified Hypocreales Incertae_sedis Unclassified f__Umbelopsidaceae(100) Incertae_sedis Unclassified g__Entoloma(100) Ascomycota f__Inocybaceae(100)ungi Sordariomycetes Unclassified Ascomycota Unclassifiedungi Incertae_sedis Leotiomycetes f__Hyponectriaceae(100)ungi Helotiales Unclassified g__Gyrodon(100) f__Bionectriaceae(100) Ascomycotaungi Leotiomycetes Unclassified Unclassified f__Ophiocordycipitaceae(100) Ascomycota Unclassifiedungi Mortierellales g__Umbelopsis(100) Leotiomycetes Hypocreales Unclassified f__Nectriaceae(100) Unclassifiedungi Unclassified Basidiomycotaungi Leotiales Sordariomycetes g__Arthropsis(100) g__Inocybe(100) Incertae_sedis Basidiomycota Unclassified ungi Eurotiomycetes g__Elaphocordyceps(100) Unclassified g__Myrothecium(100) Basidiomycota Unclassifiedungi Helotiales Unclassified Agaricomycetes Unclassified Ascomycotaungi Helotiales Agaricomycetes f__Hyaloscyphaceae(100) Basidiomycotaungi f__Mortierellaceae(100) Hypocreales Agaricomycetes Unclassified f__Hypocreaceae(100) Basidiomycotaungi g__Cosmospora(88) Ascomycota Eurotialesungi Pezizomycetes Agaricomycetes Unclassified Athelialesungi Unclassified Tremellomycetes f__Leotiaceae(100) Ascomycota Sebacinales Unclassifiedungi g__Lasiobelonium(100) Unclassified Ascomycota Agaricalesungi Sordariomycetes f__Helotiaceae(100) g__Collophora(100) Basidiomycota g__Mortierella(100) ungi Unclassified Incertae_sedis f__Nectriaceae(100) Ascomycota Sebacinales g__Sepedonium(100) Pezizalesungi Incertae_sedis Unclassified Filobasidiales Zygomycotaungi Dothideomycetes Agaricomycetes Unclassified f__Trichocomaceae(100)ungi Xylariales Ascomycotaungi f__Atheliaceae(100) Dothideomycetes f__Sebacinaceae(100) Unclassified Unclassified Basidiomycota Unclassified ungi Incertae_sedis Unclassified f__Strophariaceae(100) Unclassifiedungi Incertae_sedis Pleosporales Unclassified Unclassified Ascomycota Boletalesungi Sordariomycetes f__Sebacinaceae(100) f__Filobasidiaceae(100) Agaricomycetes g__Fusarium(86) Ascomycotaungi Pleosporales f__Pyronemataceae(100) Unclassified g__Aspergillus(100) Ascomycotaungi Unclassified g__Leptodontidium(100) Ascomycotaungi Mortierellales Sordariomycetes g__Sebacina(100) f__Xylariaceae(100) Unclassifiedungi g__Amphinema(100) Sordariomycetes Hypocreales Incertae_sedis g__Hypholoma(100) Ascomycota Agaricales Unclassifiedungi Sordariomycetes f__Venturiaceae(100) Unclassified Zygomycotaungi g__Cryptococcus(100) Leotiomycetes g__Sebacina(100) g__Tarzetta(100) Unclassifiedungi Unclassified f__Paxillaceae(100) f__Pleosporaceae(100) Hypocreales Unclassified Ascomycota Unclassifiedungi Sordariomycetes Hypocreales Ascomycotaungi f__Mortierellaceae(100) Incertae_sedis Coniochaetales Unclassifiedungi Unclassified Incertae_sedis g__Nemania(100) Ascomycotaungi Helotiales Sordariomycetes Unclassified Ascomycota f__Cortinariaceae(100) Unclassified ungi Eurotiomycetes Hypocreales Ascomycota Unclassifiedungi Unclassified g__Epicoccum(100) f__Clavicipitaceae(100) g__Amplistroma(100) Ascomycotaungi Mortierellales Unclassified g__Melanogaster(100) Unclassified f__Coniochaetaceae(100) f__Nectriaceae(100) Ascomycota g__Mortierella(100) ungi Eurotiomycetes Xylariales Unclassified Ascomycota Unclassifiedungi Leotiomycetes Basidiomycota Eurotialesungi Sordariomycetes Ascomycota g__Cortinarius(100) ungi Sordariomycetes Unclassified f__Ophiocordycipitaceae(100) Ascomycota Unclassifiedungi Leotiomycetes Tremellomy Fungi Basidiomycota Agaricomycetes Sebacinales f__Sebacinaceae(100) g__Sebacina(100) 00600000500000000F 000000000000011008F Otu0222 Otu0272000000000000020011F Otu0223000001002001500000F Otu0224000000000001800000F Otu0225001800000000000000F Otu0226001300500000000000F Otu0227000000000000080010F Otu0228001100200500000000F Otu0229001300500000000000F Otu0230001000000000000008F Otu0231001700100000000000F Otu0232001700000000000000F Otu0233000000001000700000F Otu023400000000800900000F Otu0235001100600000000000F Otu0236004001300000000000F Otu023700800900000000000F Otu0238000000001400300000F Otu0239000000001700000000F Otu0240000000001600000000F Otu0241000000005001100000F Otu0242000000001100500000F Otu024300000000000007009F Otu0244002000001100300000F Otu0245000000001600000000F Otu0246000000000000020014F Otu0247000000001600000000F Otu0248001600000000000000F Otu0249000000000000016000F Otu025000000000800700000F Otu0251001500000000000000F Otu0252001500000000000000F Otu0253000000001500000000F Otu025400000000900200004F Otu0255000000001500000000 Otu0256000000000001400000F Otu0257001300100000000000F Otu025800000000500900000F Otu025900000000800006000F Otu0260000000000001400000F Otu0261000000001400000000F Otu0262000000001400000000F Otu0263000000002001100000F Otu0264000000000000130000F Otu0265000000000001300000F Otu0266000000002001100000F Otu026700900400000000000F Otu026800000000000004009F Otu026900000000600700000F Otu0270001100110000000000F Otu0271000130000000000000F Otu0273000000000001300000F Otu0274000000001300000000F Otu0275000000000000000013F Otu0276000000001300000000F Otu0277001100100000000000F Otu0278000000000000010011F Otu027900500700000000000F Otu0280000000000000000012F Otu0281001200000000000000F Otu0282000001200000000000F Otu0283000000000001200000F Otu0284000000000001200000F Otu0285001000200000000000F Otu028600000000700500000F Otu028700000000600600000F Otu0288000000001200000000F Otu0289 Otu0290000000001100000000F Otu029100000000300800000F Otu0292000000001100000000F Otu0293000000001100000000F Otu0294000000001100000000F Otu0295000000000001100000F Otu0296000001100000000000F Otu0297000000000000011000F Otu0298000000001100000000F Otu029900200000900000000F Otu030000000000200900000F Otu030100000000000008003F Otu030200000000000009002F Otu0303000000001100000000F Otu0304000000000000110000F Otu030500000000400700000F Otu0306000000001100000000F Otu0307000000000001100000F Otu0308000000000001100000F Otu0309000000001100000000F Otu0310000000001100000000F Otu031100000000600400000F Otu0312000001000000000000F Otu0313001000000000000000F Otu031400600400000000000F Otu0315000000001000000000F Otu031600000000700300000F Otu0317001000000000000000F Otu0318000000000001000000F Otu0319000000000001000000F Otu0320000000000001000000F Otu032100200400200200000Fun

62 ) 100 ( Trichoderma )g 100 ( ocreaceae yp ocreales f H yp cetes H y cota Sordariom y i Ascom g ungiungiungi Ascomycotaungi Zygomycotaungi Ascomycotaungi Unclassifiedungi Dothideomycetes Ascomycotaungi Incertae_sedis Unclassifiedungi Unclassified Basidiomycotaungi Unclassified Basidiomycotaungi Incertae_sedis Orbiliomycetes Ascomycotaungi Unclassified Agaricomycetes Chytridiomycotaungi Mortierellales Unclassified Basidiomycotaungi Zygomycota Unclassified Unclassifiedungi Leotiomycetes Basidiomycota Unclassified Unclassifiedungi Agaricomycetes f__Myxotrichaceae(100) Zygomycota Athelialesungi Unclassified Unclassifiedungi Incertae_sedis Agaricomycetes Ascomycotaungi f__Mortierellaceae(100) Unclassified Ascomycotaungi Helotiales Incertae_sedis Unclassified Unclassified Agaricalesungi Unclassified Unclassified Ascomycotaungi Sordariomycetes g__Oidiodendron(100) Unclassified Unclassified Ascomycota Athelialesungi Mortierellales Sordariomycetes Ascomycotaungi f__Atheliaceae(100) Unclassified Unclassified Ascomycota g__Mortierella(100) ungi Mortierellales Sordariomycetes Unclassified Ascomycotaungi Dothideomycetes Hypocreales Basidiomycota Unclassifiedungi Unclassified Eurotiomycetes Xylariales Incertae_sedis Unclassified f__Entolomataceae(100)ungi Eurotiomycetes Ascomycotaungi f__Mortierellaceae(100) Dothideomycetes Hypocreales Agaricomycetes Unclassified Unclassifiedungi f__Atheliaceae(100) Incertae_sedis Ascomycotaungi f__Mortierellaceae(100) g__Athelopsis(100) Unclassified Unclassified Zygomycota Chaetothyrialesungi Dothideomycetes Unclassified Unclassified f__Clavicipitaceae(100) Ascomycota Chaetothyrialesungi Hysteriales Unclassified g__Entoloma(100) Unclassified Ascomycota Thelephorales Sordariomycetes Unclassified f__Amphisphaeriaceae(100) Unclassified g__Mortierella(100) ungi Incertae_sedis Unclassified Incertae_sedis f__Pseudeurotiaceae(100) Unclassifiedungi Hysteriales Dothideomycetes g__Scleropezicula(100) Unclassified Unclassified g__Mortierella(100) Unclassified f__Herpotrichiellaceae(100)ungi g__Amphinema(100) Dothideomycetes Ascomycota f__Herpotrichiellaceae(100)ungi Unclassified Hypocreales g__Metarhizium(100) Glomeromycota Unclassified Unclassifiedungi Unclassified f__Thelephoraceae(100) Zygomycota Unclassifiedungi Mortierellales Botryosphaeriales g__Leuconeurospora(100) Unclassified Glomeromycetesungi Incertae_sedis Pezizomycetes g__Cladophialophora(100) Unclassifiedungi g__Rhinocladiella(100) Unclassified f__Gloniaceae(100)ungi Incertae_sedis Unclassified Unclassified Ascomycota Unclassifiedungi Unclassified Unclassified g__Stachybotrys(100) f__Ophiocordycipitaceae(100) Ascomycota Unclassified Glomeralesungi Unclassified Unclassified Unclassified Zygomycota Pezizalesungi f__Mortierellaceae(100) Unclassified Unclassified f__Myxotrichaceae(100) Ascomycotaungi Saccharomycetes Rozellomycotaungi Mortierellales Sordariomycetes g__Elaphocordyceps(100) Zygomycotaungi Incertae_sedis Ascomycota Unclassified Unclassifiedungi g__Cenococcum(100) Leotiomycetes Unclassified Unclassified Basidiomycota Saccharomycetales Unclassifiedungi Unclassified f__Glomeraceae(100) Zygomycota g__Mortierella(100) Unclassifiedungi Incertae_sedis Hypocreales g__Oidiodendron(100) Ascomycotaungi Sordariomycetes f__Pezizaceae(100) Unclassified Agaricomycetes Ascomycotaungi Mortierellales f__Mortierellaceae(100) Unclassified Unclassified Incertae_sedisungi Thelebolales Incertae_sedis Unclassified Zygomycotaungi Unclassified Leotiomycetes Unclassified Unclassifiedungi Mortierellales Saccharomycetes Sordariales Unclassified Zygomycota Agaricalesungi g__Glomus(100) Unclassified Unclassified f__Cordycipitaceae(100) Ascomycotaungi Incertae_sedis Ascomycota g__Mortierella(100) ungi Mortierellales Unclassified Unclassified Unclassified g__Pachyphloeus(100) Rozellomycota Unclassifiedungi Helotiales Incertae_sedis Unclassified f__Thelebolaceae(100) Basidiomycotaungi Leotiomycetes Unclassified Basidiomycotaungi f__Mortierellaceae(100) Incertae_sedis Unclassified Basidiomycota g__Peterozyma(100) Unclassified Unclassifiedungi Basidiobolales g__Cordyceps(100) Agaricomycetes Basidiomycota f__Inocybaceae(100)ungi Microbotryomycetes Basidiomycota Unclassifiedungi Mortierellales f__Mortierellaceae(100) Unclassified Tremellomycetes Unclassifiedungi Helotiales Unclassified Microbotryomycetes Unclassified Unclassified Unclassifiedungi Incertae_sedis Unclassified Sporidiobolales Agaricomycetes f__Dermateaceae(100) Ascomycota Polyporales g__Mortierella(100) ungi Unclassified Unclassified Zygomycota Unclassified ungi Unclassified Unclassified Filobasidiales Unclassified Ascomycotaungi Unclassified g__Inocybe(100) Ascomycota g__Mortierella(100) ungi f__Mortierellaceae(100) Eurotiomycetes Unclassified Unclassified Ascomycota Agaricalesungi Incertae_sedis Incertae_sedis Incertae_sedis Ascomycota Unclassified ungi Eurotiomycetes f__Dermateaceae(100) g__Marssonina(100) Unclassifiedungi Eurotiomycetes f__Meruliaceae(100) Unclassified Basidiomycota Unclassifiedungi Leotiomycetes f__Filobasidiac Fungi Basidiomycota Microbotryomycetes Leucosporidiales f__Leucosporidiaceae(100) g__Mastigobasidium(100) 00200200000000000F 00000500000000000F Otu0446 Otu049600000000000004000F Otu044700050000000000000F Otu044800000000300200000F Otu044900500000000000000F Otu045000500000000000000F Otu045100000500000000000F Otu045200300200000000000F Otu045300500000000000000F Otu045400100000000001003F Otu045500000000000500000F Otu045600500000000000000F Otu045700000000000000005F Otu045800000000500000000F Otu045900000000000014000F Otu046000000000500000000F Otu046100000000500000000F Otu046200000000000500000F Otu046300500000000000000F Otu046400000000500000000F Otu046500000000500000000F Otu046600000000000000005F Otu046700100400000000000F Otu046800500000000000000F Otu046900000000000000005F Otu047000500000000000000F Otu047100000000000005000F Otu047200000000500000000F Otu047300400100000000000F Otu047400000000500000000F Otu047500000000400000000F Otu047600000000000003001F Otu047700000000400000000F Otu047804000000000000000F Otu047900000000000040000 Otu048000400000000000000F Otu048100000000400000000F Otu048200000000400000000F Otu048300000000000400000F Otu048400000000400000000F Otu048500000000400000000F Otu048600000000000004000F Otu048700000000000400000F Otu048800000400000000000F Otu048900000000400000000F Otu049000300100000000000F Otu049100000000400000000F Otu049200400000000000000F Otu049300000000000400000F Otu049440000000000000000F Otu049500000000000400000F Otu049700000000400000000F Otu049800000000000001003F Otu049900000000200200000F Otu050000000000200200000F Otu050100400000000000000F Otu050200400000000000000F Otu050300000000000400000F Otu050400000400000000000F Otu050500000000000400000F Otu050600000000000040000F Otu050700000000000001003F Otu050800000000000400000F Otu050900000000000000004F Otu051000000000000400000F Otu051100400000000000000F Otu051200000000400000000F Otu0513 Otu051400000400000000000F Otu051500400000000000000F Otu051600000000000400000F Otu051700000000000000003F Otu051800000000000003000F Otu051900000000300000000F Otu052000000000000003000F Otu052100000000000300000F Otu052200000000000300000F Otu052300000000000001002F Otu052400000000300000000F Otu052500000000300000000F Otu052600000000300000000F Otu052700000000000000003F Otu052800000000200100000F Otu052900300000000000000F Otu053000000000000000003F Otu053100000000000300000F Otu053200000000000003000F Otu053300000300000000000F Otu053400000000300000000F Otu053500000000300000000F Otu053600000000000003000F Otu053700000000000000003F Otu053800000000300000000F Otu053900000000000003000F Otu054000000000300000000F Otu054100000000000000003F Otu054200000000000300000F Otu054300300000000000000F Otu054400000000300000000F Otu054500000000300000000Fun

63 ungiungiungi Ascomycotaungi Zygomycotaungi Ascomycotaungi Unclassifiedungi Dothideomycetes Ascomycotaungi Incertae_sedis Unclassifiedungi Unclassified Basidiomycotaungi Unclassified Basidiomycotaungi Incertae_sedis Orbiliomycetes Ascomycotaungi Unclassified Agaricomycetes Chytridiomycotaungi Mortierellales Unclassified Basidiomycotaungi Zygomycota Unclassified Unclassifiedungi Leotiomycetes Basidiomycota Unclassified Unclassifiedungi Agaricomycetes f__Myxotrichaceae(100) Zygomycota Athelialesungi Unclassified Unclassifiedungi Incertae_sedis Agaricomycetes Ascomycotaungi f__Mortierellaceae(100) Unclassified Ascomycotaungi Helotiales Incertae_sedis Unclassified Unclassified Agaricalesungi Unclassified Unclassified Ascomycotaungi Sordariomycetes g__Oidiodendron(100) Unclassified Unclassified Ascomycota Athelialesungi Mortierellales Sordariomycetes Ascomycotaungi f__Atheliaceae(100) Unclassified Unclassified Ascomycota g__Mortierella(100) ungi Mortierellales Sordariomycetes Unclassified Ascomycotaungi Dothideomycetes Hypocreales Basidiomycota Unclassifiedungi Unclassified Eurotiomycetes Xylariales Incertae_sedis Unclassified f__Entolomataceae(100)ungi Eurotiomycetes Ascomycotaungi f__Mortierellaceae(100) Dothideomycetes Hypocreales Agaricomycetes Unclassified Unclassifiedungi f__Atheliaceae(100) Incertae_sedis Ascomycotaungi f__Mortierellaceae(100) g__Athelopsis(100) Unclassified Unclassified Zygomycota Chaetothyrialesungi Dothideomycetes Unclassified Unclassified f__Clavicipitaceae(100) Ascomycota Chaetothyrialesungi Hysteriales Unclassified g__Entoloma(100) Unclassified Ascomycota Thelephorales Sordariomycetes Unclassified f__Amphisphaeriaceae(100) Unclassified g__Mortierella(100) ungi Incertae_sedis Unclassified Incertae_sedis f__Pseudeurotiaceae(100) Unclassifiedungi Hysteriales Dothideomycetes g__Scleropezicula(100) Unclassified Unclassified g__Mortierella(100) Unclassified f__Herpotrichiellaceae(100)ungi g__Amphinema(100) Dothideomycetes Ascomycota f__Herpotrichiellaceae(100)ungi Unclassified Hypocreales g__Metarhizium(100) Glomeromycota Unclassified Unclassifiedungi Unclassified f__Thelephoraceae(100) Zygomycota Unclassifiedungi Mortierellales Botryosphaeriales g__Leuconeurospora(100) Unclassified Glomeromycetesungi Incertae_sedis Pezizomycetes g__Cladophialophora(100) Unclassifiedungi g__Rhinocladiella(100) Unclassified f__Gloniaceae(100)ungi Incertae_sedis Unclassified Unclassified Ascomycota Unclassifiedungi Unclassified Unclassified g__Stachybotrys(100) f__Ophiocordycipitaceae(100) Ascomycota Unclassified Glomeralesungi Unclassified Unclassified Unclassified Zygomycota Pezizalesungi f__Mortierellaceae(100) Unclassified Unclassified f__Myxotrichaceae(100) Ascomycotaungi Saccharomycetes Rozellomycotaungi Mortierellales Sordariomycetes g__Elaphocordyceps(100) Zygomycotaungi Incertae_sedis Ascomycota Unclassified Unclassifiedungi g__Cenococcum(100) Leotiomycetes Unclassified Unclassified Basidiomycota Saccharomycetales Unclassifiedungi Unclassified f__Glomeraceae(100) Zygomycota g__Mortierella(100) Unclassifiedungi Incertae_sedis Hypocreales g__Oidiodendron(100) Ascomycotaungi Sordariomycetes f__Pezizaceae(100) Unclassified Agaricomycetes Ascomycotaungi Mortierellales f__Mortierellaceae(100) Unclassified Unclassified Incertae_sedisungi Thelebolales Incertae_sedis Unclassified Zygomycotaungi Unclassified Leotiomycetes Unclassified Unclassifiedungi Mortierellales Saccharomycetes Sordariales Unclassified Zygomycota Agaricalesungi g__Glomus(100) Unclassified Unclassified f__Cordycipitaceae(100) Ascomycotaungi Incertae_sedis Ascomycota g__Mortierella(100) ungi Mortierellales Unclassified Unclassified Unclassified g__Pachyphloeus(100) Rozellomycota Unclassifiedungi Helotiales Incertae_sedis Unclassified f__Thelebolaceae(100) Basidiomycotaungi Leotiomycetes Unclassified Basidiomycotaungi f__Mortierellaceae(100) Incertae_sedis Unclassified Basidiomycota g__Peterozyma(100) Unclassified Unclassifiedungi Basidiobolales g__Cordyceps(100) Agaricomycetes Basidiomycota f__Inocybaceae(100)ungi Microbotryomycetes Basidiomycota Unclassifiedungi Mortierellales f__Mortierellaceae(100) Unclassified Tremellomycetes Unclassifiedungi Helotiales Unclassified Microbotryomycetes Unclassified Unclassified Unclassifiedungi Incertae_sedis Unclassified Sporidiobolales Agaricomycetes f__Dermateaceae(100) Ascomycota Polyporales g__Mortierella(100) ungi Unclassified Unclassified Zygomycota Unclassified ungi Unclassified Unclassified Filobasidiales Unclassified Ascomycotaungi Unclassified g__Inocybe(100) Ascomycota g__Mortierella(100) ungi f__Mortierellaceae(100) Eurotiomycetes Unclassified Unclassified Ascomycota Agaricalesungi Incertae_sedis Incertae_sedis Incertae_sedis Ascomycota Unclassified ungi Eurotiomycetes f__Dermateaceae(100) g__Marssonina(100) Unclassifiedungi Eurotiomycetes f__Meruliaceae(100) Unclassified Basidiomycota Unclassifiedungi Leotiomycetes f__Filobasidiaceae(100) G Fungi Basidiomycota Microbotryomycetes Leucosporidiales f__Leucosporidiaceae(100) g__Mastigobasidium(100) 00200200000000000F 00000500000000000F Otu0446 Otu049600000000000004000F Otu044700050000000000000F Otu044800000000300200000F Otu044900500000000000000F Otu045000500000000000000F Otu045100000500000000000F Otu045200300200000000000F Otu045300500000000000000F Otu045400100000000001003F Otu045500000000000500000F Otu045600500000000000000F Otu045700000000000000005F Otu045800000000500000000F Otu045900000000000014000F Otu046000000000500000000F Otu046100000000500000000F Otu046200000000000500000F Otu046300500000000000000F Otu046400000000500000000F Otu046500000000500000000F Otu046600000000000000005F Otu046700100400000000000F Otu046800500000000000000F Otu046900000000000000005F Otu047000500000000000000F Otu047100000000000005000F Otu047200000000500000000F Otu047300400100000000000F Otu047400000000500000000F Otu047500000000400000000F Otu047600000000000003001F Otu047700000000400000000F Otu047804000000000000000F Otu047900000000000040000 Otu048000400000000000000F Otu048100000000400000000F Otu048200000000400000000F Otu048300000000000400000F Otu048400000000400000000F Otu048500000000400000000F Otu048600000000000004000F Otu048700000000000400000F Otu048800000400000000000F Otu048900000000400000000F Otu049000300100000000000F Otu049100000000400000000F Otu049200400000000000000F Otu049300000000000400000F Otu049440000000000000000F Otu049500000000000400000F Otu049700000000400000000F Otu049800000000000001003F Otu049900000000200200000F Otu050000000000200200000F Otu050100400000000000000F Otu050200400000000000000F Otu050300000000000400000F Otu050400000400000000000F Otu050500000000000400000F Otu050600000000000040000F Otu050700000000000001003F Otu050800000000000400000F Otu050900000000000000004F Otu051000000000000400000F Otu051100400000000000000F Otu051200000000400000000F Otu0513 Otu051400000400000000000F Otu051500400000000000000F Otu051600000000000400000F Otu051700000000000000003F Otu051800000000000003000F Otu051900000000300000000F Otu052000000000000003000F Otu052100000000000300000F Otu052200000000000300000F Otu052300000000000001002F Otu052400000000300000000F Otu052500000000300000000F Otu052600000000300000000F Otu052700000000000000003F Otu052800000000200100000F Otu052900300000000000000F Otu053000000000000000003F Otu053100000000000300000F Otu053200000000000003000F Otu053300000300000000000F Otu053400000000300000000F Otu053500000000300000000F Otu053600000000000003000F Otu053700000000000000003F Otu053800000000300000000F Otu053900000000000003000F Otu054000000000300000000F Otu054100000000000000003F Otu054200000000000300000F Otu054300300000000000000F Otu054400000000300000000F Otu054500000000300000000F

64 ungiungiungi Ascomycotaungi Basidiomycotaungi Basidiomycotaungi Ascomycotaungi Dothideomycetes Microbotryomycetes Ascomycotaungi Microbotryomycetes Unclassifiedungi Unclassifiedungi Sordariomycetes Heterogastridiales Unclassifiedungi Pleosporales Sordariomycetes Leucosporidiales Ascomycotaungi Unclassified Unclassifiedungi Unclassified Ascomycotaungi Unclassified Hypocreales f__Heterogastridiaceae(100) Unclassifiedungi Dothideomycetes Sordariales Unclassifiedungi f__Leucosporidiaceae(100) Unclassified Unclassifiedungi Archaeorhizomycetes f__Sporormiaceae(100) Unclassified Unclassifiedungi Unclassified Basidiomycota g__Heterogastridium(100) Unclassifiedungi Capnodiales Unclassified Archaeorhizomycetales Unclassified Unclassifiedungi Unclassified f__Ophiocordycipitaceae(100) Basidiomycota g__Leucosporidium(100) ungi Unclassified Microbotryomycetes Basidiomycota Unclassified f__Lasiosphaeriaceae(100)ungi f__Archaeorhizomycetaceae(100) Ascomycotaungi g__Preussia(100) Unclassified Agaricomycetes Ascomycota Unclassifiedungi g__Tolypocladium(100) Sporidiobolales Agaricomycetes Unclassified g__Archaeorhizomyces(100) Basidiomycota Unclassifiedungi Unclassified f__Mycosphaerellaceae(100) Unclassified Unclassifiedungi Dothideomycetes Unclassified g__Lasiosphaeria(100) Ascomycota Unclassifiedungi Sordariomycetes Agaricomycetes Unclassified Agaricalesungi Unclassified Unclassified Polyporales Unclassifiedungi Unclassified g__Mycosphaerella(100) Incertae_sedis Ascomycotaungi Capnodiales Dothideomycetes Unclassified Zygomycotaungi Unclassified Sordariales Unclassified Unclassified Thelephoralesungi Unclassified Unclassified Unclassifiedungi Leotiomycetes Unclassified Ascomycota Pleosporales Incertae_sedis Unclassified Basidiomycota f__Tricholomataceae(100) Unclassifiedungi Unclassified Unclassified f__Meruliaceae(100) Unclassified Chytridiomycotaungi Unclassified Unclassified f__Mycosphaerellaceae(100) Unclassifiedungi Sordariomycetes Agaricomycetes f__Thelephoraceae(100) Basidiomycota Chytridiomycetes Unclassified f__Lasiosphaeriaceae(100)ungi Helotiales Unclassified g__Rhodotorula(100) Zygomycotaungi Mortierellales Unclassified g__Fayodia(100) ungi f__Lophiostomataceae(100) Unclassified Agaricomycetes g__Mycosphaerella(100) Ascomycota Unclassifiedungi Hypocreales Unclassified Unclassified Ascomycota Polyporales Rhizophydiales Unclassifiedungi Incertae_sedis g__Hyphoderma(100) Unclassified g__Cercophora(100) Ascomycotaungi g__Tomentella(100) Unclassified Unclassified Unclassified Chytridiomycotaungi Sordariomycetes Unclassified g__Lophiostoma(100) Unclassified Athelialesungi f__Mortierellaceae(100) Sordariomycetes Unclassified Incertae_sedis Zygomycota Unclassifiedungi Eurotiomycetes Zygomycotaungi Mortierellales f__Rhizophydiaceae(100) Unclassified f__Clavicipitaceae(100) Basidiomycotaungi Unclassified Hypocreales f__Xenasmataceae(100) Unclassified Zygomycota Unclassifiedungi Incertae_sedis Hypocreales Unclassifiedungi Incertae_sedis Agaricomycetes Ascomycota Eurotiales g__Mortierella(100) ungi Unclassified Unclassified Ascomycotaungi f__Atheliaceae(100) Incertae_sedis g__Rhizophydium(100) Ascomycotaungi Unclassified Unclassified Unclassified g__Pochonia(100) Unclassified g__Xenasmatella(100) Unclassifiedungi Mortierellales Dothideomycetes Unclassified g__Scleropezicula(100) f__Nectriaceae(100) Chytridiomycota Agaricalesungi Mortierellales Eurotiomycetes Unclassified f__Cordycipitaceae(100) Basidiomycotaungi Sordariomycetes Unclassified Chytridiomycetesungi Mortierellales Unclassified Unclassified Ascomycota f__Trichocomaceae(100)ungi Unclassified Capnodiales Unclassified Agaricomycetes Ascomycota Unclassifiedungi g__Piloderma(100) Zygomycota Eurotialesungi f__Mortierellaceae(100) Unclassified Sordariales Unclassified Unclassified Rhizophydialesungi f__Mortierellaceae(100) Sordariomycetes g__Beauveria(100) g__Cosmospora(100) Ascomycota f__Bolbitiaceae(100)ungi Eurotiomycetes Ascomycota Agaricales Unclassified Unclassifiedungi f__Mortierellaceae(100) Incertae_sedis g__Sagenomella(100) Ascomycotaungi Unclassified f__Mycosphaerellaceae(100) Ascomycotaungi Pezizomycetes Chaetosphaeriales Unclassified Unclassified Basidiomycota g__Mortierella(100) Unclassifiedungi Eurotiomycetes f__Rhizophydiaceae(100) Zygomycota Chaetothyriales g__Mortierella(100) Unclassified f__Lasiosphaeriaceae(100)ungi Sordariomycetes Ascomycota Unclassified ungi Mortierellales Dothideomycetes Agaricomycetes g__Mycosphaerella(100) g__Conocybe(100) Ascomycota g__Mortierella(100) ungi Unclassified f__Chaetosphaeriaceae(100) Ascomycota f__Psathyrellaceae(100) Pezizales Unclassifiedungi Incertae_sedis Unclassified Zygomycota Eurotialesungi Sordariomycetes Incertae_sedis g__Rhizophydium(100) g__Podospora(100) Basidiomycota f__Herpotrichiellaceae(100)ungi Venturiales Sordariomycetes Unclassified Ascomycota Athelialesungi Leotiomycetes g__Cylindrotrichum(100) Unclassifiedungi f__Mortierellaceae(100) Fungi Unclassified Unclassified Unclassified Unclassified Unclassified 00000000200000000F 00000000300000000F Otu0558 Otu060800100100000000000F Otu055900000000000000003F Otu056000000000000030000F Otu056100300000000000000F Otu056200000000000000003F Otu056300000000000300000F Otu056400000000300000000F Otu056500300000000000000F Otu056600000000000300000F Otu056700300000000000000F Otu056800000000000000003F Otu056900000000000300000F Otu057000000000000000003F Otu057100000000000300000F Otu057200300000000000000F Otu057300000000000300000F Otu057400300000000000000F Otu057500000000000002001F Otu057600000000000003000F Otu057700000000300000000F Otu057800000000300000000F Otu057900100200000000000F Otu058000300000000000000F Otu058100300000000000000F Otu058200000000000000003F Otu058300000000000002001F Otu058400200000000000000F Otu058520000000000000000F Otu058600200000000000000F Otu058700100100000000000F Otu058800000000000001001F Otu058900000200000000000F Otu059000000000000200000F Otu059100000000000200000 Otu059200000000000002000F Otu059300100000000001000F Otu059400200000000000000F Otu059500000000000200000F Otu059600000000200000000F Otu059700200000000000000F Otu059800000000200000000F Otu059900000000200000000F Otu060000000000020000000F Otu060120000000000000000F Otu060200000000200000000F Otu060300000000020000000F Otu060400000000000002000F Otu060500000200000000000F Otu060600000000200000000F Otu060700200000000000000F Otu060900000000000000002F Otu061000000200000000000F Otu061100000000000000002F Otu061200000000200000000F Otu061300200000000000000F Otu061400000000200000000F Otu061500200000000000000F Otu061600000000000200000F Otu061700000000000200000F Otu061800000000200000000F Otu061900200000000000000F Otu062000200000000000000F Otu062100000000020000000F Otu062200000000000000002F Otu062300200000000000000F Otu062400000000000002000F Otu0625 Otu062600000000200000000F Otu062700000000000002000F Otu062800200000000000000F Otu062900000000000000002F Otu063000000000000002000F Otu063100000000000020000F Otu063200000000000002000F Otu063300000000000000002F Otu063400000000200000000F Otu063500000200000000000F Otu063600200000000000000F Otu063700000000200000000F Otu063800000000200000000F Otu063900020000000000000F Otu064000000000200000000F Otu064100200000000000000F Otu064200000000000002000F Otu064300000000000002000F Otu064400000000200000000F Otu064500000000000200000F Otu064600000200000000000F Otu064700000000200000000F Otu064800000200000000000F Otu064900000000000020000F Otu065000200000000000000F Otu065100000000200000000F Otu065200200000000000000F Otu065300200000000000000F Otu065400200000000000000F Otu065500000000000002000F Otu065600000000020000000F Otu065700200000000000000F

65 ungiungiungi Unclassifiedungi Ascomycotaungi Ascomycotaungi Chytridiomycotaungi Unclassified Ascomycotaungi Dothideomycetes Chytridiomycota Chytridiomycetesungi Dothideomycetes Ascomycotaungi Ascomycota Chytridiomycetesungi Pezizomycetes Basidiomycotaungi Capnodiales Unclassified Unclassified Unclassifiedungi Pleosporales Sordariomycetes Basidiomycotaungi Sordariomycetes Agaricomycetes Basidiomycota Rhizophydialesungi Ascomycota Pezizalesungi Unclassified Unclassified Unclassifiedungi Hypocreales Agaricomycetes Unclassifiedungi Incertae_sedis f__Teratosphaeriaceae(100) Zygomycota Agaricalesungi Archaeorhizomycetes f__Venturiaceae(100) Unclassified Unclassified Unclassifiedungi f__Rhizophydiaceae(100) Unclassified Unclassifiedungi Unclassified Archaeorhizomycetales Ascomycota Thelephorales Unclassifiedungi Unclassified Incertae_sedis g__Teratosphaeria(100) Ascomycotaungi Unclassified f__Pyronemataceae(100) f__Hypocreaceae(100) Chytridiomycota Incertae_sedisungi Unclassified f__Archaeorhizomycetaceae(100) Unclassifiedungi Dothideomycetes g__Rhizophydium(100) Unclassified f__Pterulaceae(100) Chytridiomycetes g__Cylindrosympodium(100) Unclassifiedungi Leotiomycetes g__Archaeorhizomyces(100) Unclassified Unclassifiedungi Mortierellales f__Thelephoraceae(100) Zygomycotaungi Unclassified Unclassified g__Genea(100) Unclassified Unclassified Unclassifiedungi Pleosporales Unclassified Unclassified Unclassified Unclassified g__Trichoderma(100) Rhizophydiales Unclassifiedungi Unclassified Zygomycotaungi Helotiales Incertae_sedis Basidiomycotaungi Unclassified Unclassified Unclassified g__Globulicium(100) g__Myrmecridium(100) ungi g__Tomentella(100) f__Mortierellaceae(100) Unclassified Unclassified Unclassified Unclassified Incertae_sedis Microbotryomycetes Basidiomycota Unclassifiedungi Incertae_sedis f__Didymosphaeriaceae(100) Unclassified Unclassified Unclassifiedungi Mortierellales Unclassified Unclassifiedungi Unclassified Sporidiobolales Agaricomycetes Unclassified Unclassifiedungi Unclassified Unclassified Unclassified Ascomycota g__Mortierella(100) Unclassifiedungi Mortierellales g__Didymosphaeria(100) Unclassified Unclassifiedungi Unclassified Basidiomycotaungi Unclassified Unclassified Unclassified Agaricales Unclassifiedungi f__Mortierellaceae(100) Leotiomycetes Unclassified Unclassified Incertae_sedis Zygomycota Unclassifiedungi Unclassified Unclassified Agaricomycetes Glomeromycotaungi g__Operculomyces(100) Unclassified Ascomycota Unclassifiedungi f__Mortierellaceae(100) Unclassified Unclassified Unclassified Unclassified Glomeromycetesungi Incertae_sedis Unclassified Basidiomycota Unclassifiedungi Helotiales Unclassified Trechisporales g__Mortierella(100) ungi Dothideomycetes Unclassified Unclassified Ascomycota f__Hygrophoraceae(100) Unclassifiedungi Unclassified Agaricomycetes Unclassified Ascomycota Glomerales Unclassified Zygomycota g__Mortierella(100) Unclassified Mortierellales Unclassified Unclassified Unclassified g__Rhodotorula(100) Basidiomycota Capnodiales Sordariomycetes Unclassified Sordariomycetes f__Hydnodontaceae(100) Unclassified Cantharellales Incertae_sedis g__Hygrophorus(100) Unclassified Agaricomycetes Unclassified Unclassified Unclassified Unclassified Xylariales f__Glomeraceae(100) Unclassified f__Mortierellaceae(100) Incertae_sedis Unclassified Unclassified f__Mycosphaerellaceae(100) g__Trechispora(100) Polyporales Mortierellales Unclassified f__Hydnaceae(100) Unclassified Unclassified g__Mycosphaerella(100) f__Annulatascaceae(100) g__Mortierella(100) g__Glomus(100) Unclassified Unclassified f__Xylariaceae(100) Unclassified f__Mortierellaceae(100) Unclassified f__Ganodermataceae(100) g__Sistotrema(100) Unclassified g__Conlarium(100) g__Ganoderma(100) g__Lopadostoma(100) g__Mortierella(100) Unclassified Unclassified Fungi Chytridiomycota Chytridiomycetes Rhizophydiales f__Rhizophydiaceae(100) g__Rhizophydium(100) 00000000000000002F Otu0670 Otu067100000200000000000F Otu067200000200000000000F Otu067300000000200000000F Otu067400000200000000000F Otu067500000000200000000F Otu067600200000000000000F Otu067700200000000000000F Otu067800000000000000002F Otu067900000000000000002F Otu068000000000000000002F Otu068100200000000000000F Otu068200000000000000002F Otu068300000000000000002F Otu068400000000000002000F Otu068520000000000000000F Otu068600200000000000000F Otu068700000000000000002F Otu068800000000200000000F Otu068900200000000000000F Otu069000000000200000000F Otu069100000000000200000F Otu069200000000000200000F Otu069300000200000000000F Otu069410010000000000000F Otu069500000000000200000F Otu069600000000200000000F Otu069700000000200000000F Otu069800000000000000002F Otu069900000200000000000F Otu070000000000000000002F Otu070100000000000000002F Otu070200200000000000000F Otu070300000000000200000 Otu070400000000200000000F Otu070500200000000000000F Otu070600200000000000000F Otu070700200000000000000F Otu070800000000000200000F Otu070920000000000000000F Otu071000000000000200000F Otu071100200000000000000F Otu071200200000000000000F Otu071300000200000000000F Otu071400000000100100000F Otu071500000000000200000F Otu071600200000000000000F Otu071700020000000000000F Otu071800000000000200000F

66 f 80 % similarity). Summary of the sequencing results for bacteria, and taxonomic assignment using the SILVA database v119 (assignment at a level o database v119 (assignment at a level assignment using the SILVA for bacteria, and taxonomic of the sequencing results Summary Table S2. Table

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 Table S3. Bacterial OTUs present only at the top of the columns. Last column indicates if the same genus is present in the soils.

91 Table S4. List of fungal and bacterial OTUs presenting a positive co-occurrence among all the samples (p-value ≤ 0.001). Th e co-occurrence probabilities between the bacterial and fungal OTUs were evaluated with the Veech probabilist model of species co-occurrence (Veech, 2013), using the cooccur package in R.

92 Chapter 2 - Culture-independent approach

Additional experiments

Bacterial population analysis with denaturing gel gradient electrophoresis (DGGE)

In order to have a fi rst insight into bacterial diversity in the diff erent soil types, and within the diff erent column sections, we performed a DGGE on DNA extracts from 2 columns per soil type (one column with LMA medium, and one with R2A medium) before performing high-throughput sequencing.

Material and methods Because of the reduced DNA yield on the DNA extracts from the target culture media, we performed fi rst a PCR amplifi cation of a partial fragment of the 16S rRNA gene with primers EUB 9-27f (5’-AGAAAGGAGGTGATCCAGCC-3’) and EUB 1542r (5’-AGAAAGGAGGTGATCCAGCC-3’; Liesack et al., 1991), followed, after purifi cation, by 3 PCR amplifi cations of the V3 region of the 16S rRNA gene, with primers 338f (3’-AC TCC TAC GGG AGG CAG CAG-5’ + 40 bp GC; Lane et al., 1991) and 518r (5’- ATT ACC GCG GCT GCT GG-3’; Muyzer et al., 1993). On the DNA extracts of the attracting culture media and the soil, we only performed PCR amplifi cations of the V3 region of the 16S rRNA gene (without the preliminary amplifi cation of a larger fragment of the 16S rRNA gene). For the PCR with primers EUB 9-27f and EUB 1542r, the master mix contained (in 25 μl of fi nal . -1 volume): 5 μl buff er (with 1.5 mM MgCl2), 0.2 mM dNTPs mix, 0.2 mM of each primer, 1 mg ml BSA and 0.5 U DNA Polymerase (Kapa Biosystems, Inc., Wilmington, USA). 1 μl of DNA templat was added (≤ 2 ng.μl-1 of DNA). PCR was carried out in an Arktik thermocycler (Th ermo Fisher Scien- tifi c, Waltham, USA), with an initial denaturation at 95 °C for 2 min, followed by 10 cycles consisting of denaturation at 95 °C for 30 s, annealing at 60 °C (-0.5 °C per cycle) for 20 s, and elongation at 72°C for 1 min, then 25 cycles consisting of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s and elongation at 72 °C for 1 min. Final extension was performed at 72 °C for 1 min. For PCR of the V3 region of the 16S rRNA gene, the master mix contained (in 25 μl of fi nal volume): 5 μl buff er (with 1.5 mM MgCl2), 0.2 mM dNTPs mix, 0.2 mM of each primer, 1 mg.ml-1 BSA, 5 μl Enhancer (Kapa Biosystems, Inc., Wilmington, USA) and 1 U DNA Polymerase (Kapa Biosystems, Inc., Wilmington, USA). 1 μl of DNA template was added (2 ng.μl-1 of DNA). PCR was carried out in an Arktik thermocycler, with an initial denaturation at 95 °C for 5 min, followed by 10 cycles consisting of denaturation at 95 °C for 30 s, annealing at 65 °C (-1 °C per cycle) for 30 s, and elongation at 72 °C for 1 min, then 25 cycles consisting of denaturation at 95 °C for 1 min, annealing at 55 °C for 30 s and elongation at 72 °C for 1 min. Final extension was performed at 72 °C for 10 min. Th e fi nal PCR products were diluted in order to obtain solutions containing 500 ng of DNA. DGGE was performed using a DCode system (Bio-Rad, A.G. München, Germany). Th e purifi ed PCR products (500 ng in 15 μl) were loaded directly onto the gel with 5 μl of loading buff er. Separa-

93 Figure A1. Bacterial fi ngerprinting obtained with DGGE analysis. Clusterization obtained with Jaccard clustering and UPGMA distance. Red: DNA from Cambisol. Blue: DNA from Fluvisol. Green: DNA from Histosol.

tion was carried out in 7.5% polyacrylamide gels with a gradient of 30–60% of denaturants (100% denaturant solution with 420 g.L-1 urea and 400 ml.l-1 deionized formamide in 0.5X TAE). Gels were runnnig during 5 h at 150 V at 60 °C. Th e gel was stained with 0.01% SYBR-Gold (BioTium Inc., CA) at 4 °C in the dark for 30 min. Image was acquired with a Transilluminator (VWR International, Radnor, USA). Th e normalization, clustering, and band selection were carried out with the software GELCOMPAR II (Sint-Martens-Latem, Belgium).

Results According to the DGGE gels, the bacterial community composition of the diff erent soils and columns varied (Figure A1), and there was no reduction of the amount of bands in the top of the columns compared with the soils. Th is is why, in order to obtain additional insights in the analysis of the taxonomic composition of the diff erent communities colonizing the columns of the three soils, we performed 454 pyrosequencing on fungal and bacterial amplicons.

References Lane DJ (1991) 16S/23S rRNA sequencing. In: Stackebrandt E & Goodfellow M, eds. Nucleic acid techniques in bacterial systematics. Wiley, West Sussex, UK. Liesack W, Weyland H & Stackebrandt E (1991) Potential risks of gene amplifi cation by PCR as determined by 16S rDNA analysis of a mixed-culture of strict barophilic bacteria. Microb Ecol 21: 191-198. Muyzer G, Waal Ed & Uitterlinden A (1993) Profi ling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplifi ed genes coding for 16S rRNA. Appl Environ Microbiol 59: 695-700.

94 Chapter 2 - Culture-independent approach

Main fi ndings and perspectives

Main fi ndings

Only fungi belonging to the genus Mortierella (with the exception of one unidentifi ed OTU) were identifi ed in the columns. Th ese fungi are r-strategists, and our fi ndings suggest that the importance of the activity of these fungi in soils might be underestimated.

In the three soils, bacterial phyla Acidobacteria and Actinobacteria showed a high richness, but they represented a very small fraction of the OTUs inside the columns. A higher proportion of OTUs belonging to phyla Firmicutes, Nitrospirae, Planctomycetes and Actinobacteria were observed in the columns. Th ese phyla might preferentially interact with fungi (and in particular Mortierella spp.), while Acidobacteria and Actinobacteria are less susceptible to interact with these fungi. Th us, bacterial dispersal along fungal hyphae can be considered as a structuring factor of microbial communities in soils.

Some bacterial genera were present only in the columns, and not in the soils. Most of these genera belong to bacterial groups including intrahyphal bacteria. As fungi Mortierella spp. are known for hosting intrahyphal bacteria, we hypothesized that the recovered bacteria might be intrahyphal. As Mortierella spp. are coenocytic fungi, we hypothesized that these bacteria were dispersed inside the fungi the same way organelles do.

Th e proportional bacterial richness recovered inside the columns was similar for the three soils, independently from the soil physico-chemical properties. Transport of bacteria along fungi, whether intra- or extraphyphal might be more dependent on the fungal physiological state than on the soil properties.

Perspectives

We would like to fi nd a way for isolating slower-growing fungi, and assess their specifi c role in bacterial dispersal.

Chapter 3

Element cycling and metabolism

Chapter 3 - Element cycling and metabolism

Foreword

In Chapter 1 of this thesis, we presented fungi and associated cultivable bacteria isolated from a Moroc- can soil with fungal highway columns. We observed that the migration of these bacteria along hyphae of their associated fungus was varying according to the culture medium. Our hypothesis was that fungi might be able to actively promote or restrain bacterial migration along their hyphae. Th us, we wanted to assess the diff erent metabolic abilities of these organisms, in order to understand if the associated fungi and bacteria showed complementary features. As well, we wanted to observe migration of bacteria in other culture media, in order to better understand which elements could trigger this diff erentiated bacterial migrationmigration..

Scanning electron microscopy picture of calcium oxalate crystals.

Th is chapter is subdivided in three sections. Chapter 3a - Fungal oxalotrophy is related to an observation I made while conducting culturing tests for metabolic activities: all my fungal strains from the genus Fusarium were able to solubilize calcium oxalate. After bibliographic researches and discussions with colleagues, I realized that no one ever isolated fungi from the oxalate-carbonate pathway that were able to dissolve calcium oxalate after only a few weeks. Nitrogen fi xation was also part of the tested abilities. Even if it is commonly accepted that fungi are not able to fi x nitrogen, I decided to conduct these tests the same way than the other. Hence I inoculated each fungal strain in a classical nitrogen-free culture medium. Surprisingly, all fungi were well-growing in this culture medium. After having questioned colleagues about this phenomenon, I thought that the best explanation would be that fungi still grow thank to their nitrogen stocks, and that successive inoculations would fi nally starve them. After 4 successive generations, I could still not starve them, and we decided to conduct further experiments in order to fi nd an explanation to this phenomenon. Th is research in presented in Chapter 3b - Fungi and nitrogen fi xation. Th e other metabolic abilities I tested are presented in Chapter 3c - Migration and metabolism. As well, this section presents the observations we made on the ability of each bacterium to migrate on its associated fungus in diff erent culture media.

Chapter 3a - Fungal oxalotrophy

Manuscript

Th is section presents the following manuscript, in preparation:

Simon A, Braissant O, Verrecchia E, Junier P. An unexpected role of fungi in the oxalate-carbonate pathway: oxalotrophy and dispersal of non-oxalotrophic bacteria.

101

An unexpected role of fungi in the oxalate-carbonate pathway: oxalotrophy and dispersal of non-oxalotrophic bacteria

Anaele Simon1, Olivier Braissant2, Eric Verrecchia3, Pilar Junier1 1 Laboratory of Microbiology, Institute of Biology, University of Neuchâtel, Neuchâtel, Switzerland 2 Center of Biomechanics & Biocalorimetry, Biozentrum/Pharmazentrum University of Basel, Basel, Switzer- land 3 Biogeosciences laboratory, Institute of Earth Surface Dynamics, University of Lausanne, 1015 Lausanne, Switzerland

Abstract

Th e oxalate-carbonate pathway involves the interaction of oxalogenic plants, oxalotrophic bacteria, and fungi. While the role of plants as calcium oxalate producers and the role of bacteria as oxalate consumers are generally accepted, the role of fungi has been proposed to be more indirect either as contributing to the pool of oxalic acid released in soil or as “highways” for the migration of oxalotrophic bacteria. In this study, we analyzed the role of selected fungi and their associated bacteria in the turnover of calcium oxalate. Th ese fungi have been isolated as potential dispersal routes for bacteria but their metabolic contribution to oxalate cycling was unknown. Contrary to what has been reported so far, our selected fungi were able to oxidize calcium oxalate, which redefi nes their role in the oxalate-carbonate pathway. Associated bacteria were not able to oxidize calcium oxalate, but they better migrated towards a medium containing calcium oxalate as sole carbon source than towards a rich culture medium. Our fi ndings propose that fungi must be considered as calcium oxalate degraders in the oxalate-carbonate pathway, and fungal highways might not only be a mean for bacteria to reach otherwise inaccessible nutrients, as previously thought.

Introduction the subsequent oxidation of calcium oxalate by oxalotrophic bacteria increases the pH, and might lead to se- condary precipitation of calcium carbonate (Braissant Oxalic acid (H2C2O4) and oxalate minerals, especial- et al., 2004, Verrecchia et al., 2006). Th e coupling of ly calcium oxalate (CaC2O4), are produced by many plants as biosynthetic byproducts related to photosyn- calcium and carbon cycles during these biogeochemi- thesis (Franceschi & Horner, 1980, Nakata, 2003). cal reactions is called the oxalate-carbonate pathway, Th ese compounds are deposited in soils through root and it is gaining increasing interest as a potential long- exudation and litter decomposition, and can be used term terrestrial CO2 sink (Cailleau et al., 2011, Gadd as a carbon and energy source by oxalotrophic bacteria et al., 2014, Marin et al., 2014). Currently, the ability (Sahin, 2003). Various fungal species are able to secrete of oxalotrophic bacteria to oxidize calcium oxalate is oxalic acid (Gadd et al., 2014), which can lead to the considered as crucial in the oxalate-carbonate pathway. precipitation of metal oxalates such as calcium oxalate However, recent studies have demonstrated that under (Guggiari et al., 2011) or copper oxalate (Joseph et al., soil-mimicking conditions, oxalate oxidation by bacte- 2012) on the surface of hyphae or near them. In soils, ria only occurs in the presence of fungi (Martin et al.,

103 2012). While the role of fungi in this interaction is not here to evaluate the metabolic activity of each indi- clearly established, it has been assumed that they in- vidual microorganism and fungal-bacterial co-cultures directly favor bacterial metabolic activity. Th e hyphal under diff erent nutrient conditions. In addition, we network may allow oxalotrophic bacteria to disperse assessed how nutrient conditions could infl uence the actively in the soil, taking the so-called fungal highways migration of bacteria along fungal hyphae. Finally, the (Kohlmeier et al., 2005, Banitz et al., 2013). As bacte- combined analysis of metabolic activity and migration rial motility is otherwise very limited in water-unsatu- tests was used to evaluate whether the association of rated soils, the presence of fungi might be essential for specifi c fungal-bacterial couples could have a positive oxalotrophic bacteria to reach their nutrient source, in or negative eff ect on the turnover of oxalate on an oxa- this case calcium oxalate. Indeed, it has been demons- late-carbonate system. trated that many oxalotrophic bacteria are able to actively move along fungal hyphae (Bravo et al., 2013). However, little is known about the metabolic activity Materials and Methods of fungi, their contribution to the oxalate-carbonate pathway, and the active selection they could exert on Fungal and bacterial strains bacteria using their mycelia as dispersal paths. Couples of fungi and migrating bacteria were pre- Unlike in a previous experiment, in which isolation viously isolated from a soil profi le located under an was conducted using calcium oxalate as sole carbon oxalogenic plant, Opuntia fi cus-indica, in a semi-arid and energy source (Bravo et al., 2013), the fungal and area of Morocco (Simon et al., 2015). Th e isolation bacterial strains used in the current study were isolated was performed with fungal highway columns, a device with a specifi c device, the fungal highway columns (Si- allowing for targeted isolation and separation of soil mon et al. 2015), in which other substrates (i.e. malt fungi and associated migrating bacteria. Further de- extract) rather than calcium oxalate could sustain fun- tails about the sampling location, isolated organisms, gal and bacterial growth. In this way, the selection of as well as composition and validation of the fungal fungi and bacteria was not directly restricted by the highway columns are presented in Simon et al. (2015). metabolism of oxalate, but by their ability to co-colo- Amongst the isolated organisms, we selected the fol- nize a medium that is separated from the initial soil by lowing fungal-bacterial associates for this study: Fusa- an unsaturated patch forcing bacteria to associate to rium nygamai (A) with Achromobacter mucicolens (A1); fungi for dispersal. Th e eff ect of the fungal and bac- Fusarium equiseti (D) with Achromobacter spanius (D1) terial strains isolated in this way on calcium oxalate and Stenotrophomonas humi (D3); Fusarium oxysporum turnover is unknown. (F) with Ochrobactrum sp. (F2) and Stenotrophomonas Th e aim of this study was to assess the metabolic maltophilia (F3), Fusarium chlamydosporum (G) with properties of fungi and bacteria with regards to cal- Ochrobactrum pecoris (G1) and Pseudomonas frederiks- cium oxalate. Th e ability to degrade calcium oxalate bergensis (G2); Fusarium chlamydosporum (H) with was tested for both fungi and associated bacteria. For Acinetobacter calcoaceticus (H3) and Stenotrophomonas this purpose, a culture-based method in which cal- rhizophila (H5). cium oxalate oxidation is directly visualized by the formation of a degradation halo around the colonies Calcium oxalate degradation (Braissant et al. 2002), was combined with isother- Th e ability of each organism to degrade calcium oxa- mal microcalorimetry (IMC), an approach that al- late was tested in a modifi ed DSM81 medium (Brais- lows for the monitoring of microbial activity through sant et al., 2002), hereafter called Ca-Oxalate medium, measurements of heat production rates (Kong et al., containing calcium oxalate as sole carbon source. Th is 2009, Braissant et al., 2013, Wu et al., 2014). IMC medium consisted of two solid layers successively has been used in the past to monitor bacterial oxalo- poured into Petri dishes. trophic activity (Bravo et al., 2011) and thus it can be Th e fi rst layer (ca. 20 mL) consisted of a 2:1 (v/v) considered as a complementary proxy to measure in a mixture of a solution A and an agar solution. Solu- non-destructive way the metabolic activity of microor- tion A was composed of (per L of milliQ water) 9 g . ganisms cultured in calcium oxalate. IMC was used Na2HPO4 12 H2O, 1.5 g KH2PO4, 1 g NH4Cl, 0.2 g

104 Figure 1. Experimental setup. a) Fungi and bacteria were cultured in Ca-Oxalate medium for visual detection of calcium oxalate solubilization. b) Th e ability of each bacterium to migrate along its associated fungus was assessed with separated Petri dishes. A part of the culture medium at the opposite part of the separation was sub-cultured in NA medium and growth of bacterial colonies was visually assessed. c) Activity of pure cultures of fungi and bacteria, as well as of co- cultures, was assessed in two culture media (MA and Ca-Oxalate) using isothermal microcalorimetry.

. MgSO4 7 H2O, 1 mL trace elements solution DSM27 during 2 weeks on malt agar (MA) medium, composed . . -1 (composition per L of milliQ water: 10 mg ZnSO4 of 12 g L malt extract (Sios Homebrew Shop, Wald, . . -1 7 H2O, 3 mg MnCl2 4 H2O, 30 mg H3BO3, 20 mg Switzerland) and 15 g L agar (fi nal pH = 5.5), and a . . . 3 CoCl2 6 H2O, 1 mg CuCl2 2 H2O, 2 mg NiCl2 6 cube of 3 mm was cut out from the pre-culture and . H2O, 3 mg Na2MoO4 2 H2O, pH=3.5), and adjusted placed on the Ca-Oxalate medium. Th e presence of a to pH = 7.2. Th e agar solution was composed of 15 g degradation halo was visually assessed after 20 days for purifi ed agar (Biolife Italiana, Milano, Italy) per L of both bacteria and fungi. Bacterium Variovorax soli was milliQ water. used a positive control. Assessment of the dissolution Th e second layer (ca. 5 mL) consisted of a 2:1 (v/v) of calcium oxalate was only conducted in pure cultures mixture of solution A supplemented with 4 g.L-1 Ca- and not with co-cultures (Figure 1a). . C2O4 H2O, and the agar solution. Th is second mixture was supplemented with 2 ml.L-1 of a solution B. Diff erential bacterial migration Solution B was composed of (per L of milliQ water) Th e ability of each isolated bacterium to migrate along . . 0.5 g Fe(NH4)(SO4)2 12 H2O and 1 g CaCl2 H2O. its associated fungus was determined using separated Bacteria were pre-cultured during 24 to 48 hours Petri dishes. Th ey consisted of two culture media in a on nutrient agar (NA) medium, composed of 8 g.L-1 Petri dish, separated from each other by a 5 mm gap nutrient broth (Biolife Italiana, Milano, Italy) and 15 without culture medium (Figure 1b). Two culture me- g.L-1 agar, and a loop of the pre-culture was inoculated dia associations were used. Th e fi rst type of separated on the Ca-Oxalate medium. Fungi were pre-cultured Petri dish contained MA medium on both sides of the

105 Figure 2. Fungal strains growing on Ca-Oxalate medium after 20 days of incubation (diameter = 7 cm). a) Fusarium nygamai; b) Fusarium equiseti; c) Fusarium oxysporum; d) Fusarium chlamydosporum; e) Fusarium chlamydosporum. Sterile Ca-Oxalate medium was used as negative control. Bacterium Variovorax soli was used as positive control. Degradation zones (darker areas) are visible on all plates, except for the negative control. For better visibility, all pictures (included the negative control) were modifi ed as following: contrast +80 %, brightness -20 %, sharpness +50 %.

separation. Th e second type of separated Petri dish Th e fi rst culture medium was MA medium, and the contained MA medium on one side of the separation, second was Ca-Oxalate medium, containing calcium and Ca-Oxalate medium on the other side. oxalate as sole carbon source. Isothermal microcalori- Bacteria were pre-cultured during 48 h in NA me- metry (IMC) measurements were performed on pure dium. Th e content of one Petri dish was taken with a fungal or bacterial cultures, as well as on fungal-bacte- loop and transferred into 1 ml of sterile physiological rial co-cultures. water. Th e fi nal OD550 of this solution varied from 1 For the analysis, 4 mL vials were fi lled with 2 mL of to 10. Ten microliters of this solution were inoculated slanted culture medium (MA, resp. Ca-Oxalate). All in the MA medium of a separated Petri dish, on a 2 bacterial strains were pre-cultured during 3 days in nu- cm line, 1.5 cm away from the separation. Fungi were trient broth (NB) medium (8 g.L-1 nutrient broth), and pre-cultured in MA medium during two weeks. A 5 4 μl of the pre-culture was inoculated at the bottom mm diameter circle was cut out and inoculated in the of the vials. All fungal strains were pre-cultured du- separated Petri dish, 2 cm away from the separation, ring 13 days on MA medium, and a cylinder of 1 mm behind the bacterial inoculum. Both microorganisms in diameter was cut out of the pre-culture and placed were inoculated simultaneously (Figure 1b). After fun- 5 mm above the bottom of the vials. Each microor- gal colonization, the culture medium at the opposite ganism was inoculated alone in MA and Ca-Oxalate part of the inoculum was scrubbed with a loop and media. In addition, co-cultures of the specifi c couples subcultured in NA medium to assess the presence of isolated from the same column were inoculated in bacteria (dispersed on fungal hyphae). Th e formation both media, with bacteria at the bottom of the vials of bacterial colonies was visually checked after 48 h and fungi 5 mm above (Figure 1c). (Figure 1b). Th e vials were incubated at 30 °C in an isothermal heat conduction microcalorimeter (TAM III, Waters/ Microcalorimetric measurement of metabolic activity TA Instruments, New Castle, USA), and the produced In order to assess the metabolic activity of the orga- heat fl ow was measured every 5 min during 160 hours. nisms, two diff erent culture media were compared. Th e baseline of heat measurement was obtained from

106 vials containing sterile medium only. Th e heat fl ow After 20 days of incubation on Ca-Oxalate medium, data (actual calorimetric measurements) were used to containing calcium oxalate as sole carbon source, all determine the maximal activity peak of each organism bacteria were able to form small colonies. However, no or co-culture, as well as the time needed to reach this visible degradation of calcium oxalate was observed, point. Th e integration of these data was used to deter- except for the positive control (bacterium Variovorax mine the cumulated heat. soli). To the contrary, all fungal strains were able to As the purpose of these microcalorimetric measure- dissolve calcium oxalate after 20 days of incubation ments was to determine if fungi alone, resp. bacteria (Figure 2). alone, showed a diff erent metabolic pattern than fungal-bacterial co-cultures (regardless of the taxonomic Diff erential bacterial migration assignment), we performed a Kruskal-Wallis rank We observed the ability of bacteria to migrate along sum test (command kruskal.test) in association with the fungal mycelium in separated Petri dishes using a pairwise test for multiple comparisons of mean rank MA medium and Ca-Oxalate medium. After 13 days sums (command posthoc.kruskal.nemenyi.test, dist= of incubation, all fungi had visibly colonized the area “Tukey”, with package PMCMR; Pohlert, 2015) in R beyond the separation. Sub-culturing in NA medium (R Core Team, 2015) for the analysis of the variance using the separated compartment was used to assess between the total heat produced by all fungi alone, all bacterial migration. Colonies of the bacterial species bacteria alone, and all fungal-bacterial co-cultures. Th is S. maltophilia and P. frederiksbergensis were visible for was performed for each medium type. In addition, a the migration tests in MA and Ca-oxalate media. Co- Wilcox test (command wilcox.test) was performed in lonies of A. mucicolens, A. spanius, Ochrobactrum sp. order to compare the total heat fl ow produced by each and S. rhizophila were only visible when sub-cultured co-culture with the cumulated heat fl ow produced by from separated Petri dishes containing Ca-Oxalate. the corresponding single fungal and bacterial cultures. Colonies of S. humi, O. pecoris and A. calcoaceticus, were not visible on the sub-cultures for any of the two Results conditions (Table 1).

Calcium oxalate degradation Microcalorimetric measurement of metabolic activity

Table 1. Migration of the selected bacteria along their associated fungal hyphae in separated Petri dishes, containing MA and MA, resp. MA and Ca-Oxalate (Ca-Ox) media.

Migration (13 days) Migration (8 days) Fungi (code) Bacteria (code) MA/MA MA/Ca-Ox MA/MA SEM* Fusarium nygamai Achromobacter - + - (A) mucicolens (A1) Achromobacter - + - Fusarium equiseti spanius (D1) (D) Stenotrophomonas - - - humi (D3) Stenotrophomonas + + - Fusarium oxysporum maltophilia (F3) (F) - + - Ochrobactrum sp. (F2)

Ochrobactrum - - + Fusarium pecoris (G1) chlamydosporum Pseudomonas (G) + + - frederiksbergensis (G2) Acinetobacter - - - Fusarium calcoaceticus (H3) chlamydosporum (H) Stenotrophomonas - + - rhizophila (H5) *Data from Simon et al. 2015, observation with scanning electron microscopy

107 Figure 3. Calorimetric measurements of heat fl ow (scale on the left) and total heat produced (bars, scale on the right) by fungi, bacteria, and co-cultures. a) In MA medium. b) In Ca-Oxalate medium, with values for mean consumption rates.

Th e average heat fl ow produced by fungi, bacteria, and 9.1 J, the total heat produced by fungi alone varied fungal-bacterial co-cultures in both Ca-Oxalate and between 11.2 and 15.9 J, and the total heat produced MA media are shown in Figure 3. Using as threshold a by fungal-bacterial co-cultures varied between 8.9 and 10 μW baseline to separate the heat produced without 16.4 J. metabolic activity compared to the catabolic activity When we compared the metabolic activity for fun- on a substrate (Trampuz et al., 2007), we considered gal, bacterial or fungal-bacterial co-cultures in each that bacteria did not show any activity in Ca-Oxalate type of medium, Kruskall-wallis and post-hoc test de- medium, while in the case of fungi and the co-cultures, monstrated that values for fungi and co-cultures were the heat fl ow overpassed the threshold slightly during comparable in both culture media (p-value = 0.9 in the fi rst 48 h of monitoring. Fungal consumption rates MA and p-value = 0.4 in Ca-Oxalate medium), while in Ca-Oxalate medium varied between 0.08 to 0.11 values for bacteria were signifi cantly lower (p-value = μmol.h-1, and bacterial consumption rates in Ca-Oxa- 0.0022** in MA and p-value = 0.00051*** in Ca-Oxa- late medium varied between 0.04 to 0.09 μmol.h-1 late medium). (Table S1). In MA medium, the average heat fl ow of In order to assess the eff ect of co-culturing on the fungal, bacterial and fungal-bacterial co-cultures was metabolic activity on the two substrates, we compared considerably higher than the 10 μW baseline during the total heat produced by fungal-bacterial co-cultures the fi rst 72 h of monitoring (Figure 3). In MA me- with the values obtained when the total heat pro- dium, consumption rates were not calculated due to its duced by each individual organism was summed up. complexity (mix of potential carbon sources). For example in MA medium, the fungus F. nygamai, After 160 h, the total heat produced was signifi cant- produced a total heat of 14.6 J, while the associated ly lower in Ca-Oxalate medium than in MA medium bacterium, A. mucicolens, produced a total heat of 2.6 (p-value = 0.000003****). In Ca-Oxalate medium, J. Th e co-culture of both F. nygamai and A. mucicolens the total heat produced varied between 0.8 and 1.4 produced a total heat of 14.2 J. In this case, the co- J for bacteria alone, between 3.7 and 4.5 J for fungi culture produced less total heat (14.2 J) than the addi- alone, and between 3.7 and 5.2 J for fungal-bacterial tion of both individual cultures (17.2 J). When calcu- co-cultures. In contrast, in MA medium, the total heat lated for all fungal-bacterial couples tested (Figure 4), produced by bacteria alone varied between 2.6 and the total heat produced by co-cultures was signifi cantly

108 Figure 4. Comparison between the heat fl ow produced by co-cultures, and the cumulated heat fl ow produced by the corresponding pure fungal and bacterial cultures (addition). a) In MA medium. b) In Ca-Oxalate medium. c) Total heat produced in MA and Ca-Oxalate media.

lower than the addition of individual cultures (p-va- after 6.9 to 37.6 h in Ca-Oxalate medium (Table S1). lue = 0.0015**) in MA medium. In contrast, these va- Values for fungi and co-cultures were comparable in lues were not signifi cantly diff erent (p-value = 0.7) in both culture media (p-value = 0.7 in MA and p-value = Ca-Oxalate medium. 0.9 in Ca-Oxalate medium), while values for bacteria We also analyzed the time needed to reach the maxi- were signifi cantly lower (p-value = 0.01** in MA and mal heat fl ow. For bacteria, the maximal heat fl ow was p-value = 0.02* in Ca-Oxalate medium). produced after 9.0 to 34.5 h in MA medium and after In order to take into consideration the results of the 1.5 to 16.8 h in Ca-Oxalate medium. For fungi, the migration tests for the metabolic measurements, the maximal heat fl ow was produced after 31.2 to 53.3 h diff erences between the results of IMC for migrating in MA medium, and after 24.5 to 43.8 h in Ca-Oxa- and non-migrating fungal-bacterial co-cultures were late medium. For co-cultures, the maximal heat fl ow considered. In MA medium, for migrating couples was produced after 9.2 to 51.0 h in MA medium, and (M+) and non-migrating couples (M-), total heat pro-

109 duced was the same between fungi and co-cultures though the method used for assessing calcium oxalate (p-value = 1 for M+, and p-value = 0.8 for M-), but dissolution does not represent all the possible comit was signifi cantly lower for non-migrating bacteria binations of environmental conditions found in soils (p-value = 0.008**) and for migrating bacteria (p-value and aff ecting calcium oxalate dissolution (e.g. diff erent = 0.049*). In Ca-Oxalate medium, the total heat pro- nitrogen sources), the association of oxalate-degrading duced was not signifi cantly diff erent between bacteria fungi with non-oxalotrophic bacteria as the result of and fungi, and between each one and the co-cultures the experimental conditions used here, opens up the (p-value = 0.07 between bacteria and fungi for M+, possibility to study the role of fungi in the oxalate-car- p-value = 0.4 between fungi and co-cultures for M+, bonate pathway under an experimental set-up that has p-value = 0.2 between bacteria and fungi for M-, p-va- not been considered in the past. lue = 0.7 between fungi and co-cultures for M-). While there are many reports of the production of In MA medium, for M+, the time needed to reach oxalate-degrading enzymes like oxalate oxidase and the maximal activity was equivalent between bacteria oxalate decarboxylase in fungi (Aguilar et al., 1999, and co-cultures (p-value = 0.4), while for M-, the time Graz et al., 2009, Makela et al., 2009, Cassland et al., was signifi cantly higher for co-cultures than for bac- 2010, Liang et al., 2015), most of these reports are lin- teria (p-value = 0.009**). For M-, the time to reach ked to fungal pathogenicity or detoxifi cation of oxalate maximal activity was signifi cantly higher for fungi crystals. But to the date, there is no reference to the po- than for bacteria (p-value = 0.046*), but it was equiva- tential role of these fungal enzymes in the oxalate-car- lent for M+ (p-value = 0.4). bonate pathway. Th us, our fi ndings propose that the In Ca-Oxalate medium, the time needed to reach role of fungi in the oxalate-carbonate pathway must be the maximal activity was equivalent between bacteria reconsidered, and that fungi should be considered not and co-cultures for M+ and M- (p-value = 0.1 in both only as releasers of oxalic acid, but also as degraders cases). For M+, the time to reach maximal activity was of calcium oxalate, thus potentially having the same signifi cantly higher for fungi than for bacteria (p-va- infl uence on the system than oxalotrophic bacteria. lue = 0.017*), but it was equivalent for M- (p-value = Furthermore, the microcalorimetric measurements 0.07). obtained suggest a metabolic activity in Ca-Oxalate Th e diff erence between the heat produced by co- culture medium (with an average maximal heat of 13.5 culture and addition of individual cultures was not μW), which is above a baseline of 10 μW that has been signifi cantly diff erent in MA medium (p-value = 0.2) suggested to separate the heat produced without me- and in Ca-Oxalate medium (p-value = 0.7). tabolic activity compared to the catabolic activity on a substrate (Trampuz et al., 2007). It is worth indicating that the results of IMC are always diffi cult to analyze Discussion because of the contribution of catabolism and anabo- lism to heat release (Russel & Cook, 1995), a reason Calcium oxalate degradation for us to perform this experiments with a high density In this study, we investigated the metabolic capabilities of biomass and for a short time frame (here 106 h for of a series of potential couples of fungi and bacteria IMC versus 20 days for the culture-based approach). associated by fungal highways, which were initially en- Taking this into consideration, it is clear that the heat riched and isolated from an oxalate-carbonate ecosys- fl ow in Ca-Oxalate medium is much lower than the tem (Simon et al., 2015). A previous study showed that values obtained in MA medium (average maximal heat oxalotrophic bacteria can migrate along fungal hyphae of 76.7 μW), but they are not negligible considering in another oxalate-carbonate ecosystem (Bravo et al., other reports using IMC to monitor fungal metabolic 2013), and thus we expected that most of the bacteria activity. For example, a study on the fungus Candida isolated in association with our fungi would be oxa- albicans, has shown that it produces a typical maximal lotrophic. However, none of the bacteria were able to heat of 35 μW in LB medium (Kong et al., 2009). It dissolve calcium oxalate. In contrast, all selected fungi is also usually assumed that one bacterial cell produces were able to solubilize calcium oxalate in a medium a heat of 2 pW (Bonkat et al., 2012, Braissant et al., containing calcium oxalate as sole carbon source. Al- 2015). Th us, the value we considered as a threshold to

110 assess microbial activity (10 μW; Trampuz et al., 2007) important as a mean for bacteria to reach an otherwise corresponds to the activity of about 106 bacteria. inaccessible carbon source, as originally thought (Mar- Th erefore, our results point to the ability of fungi tin, et al., 2012). to use calcium oxalate as a carbon source. It has been For another couple used as a model for fungal shown that in fungi, the enzyme oxalate decarboxy- highways in the laboratory (the bacterium Pseudomo- lase might play two physiological roles. Th e fi rst and nas putida and fungus Morchella crassipes), we have better known is the involvement of this enzyme in the observed that bacteria that are initially transported control of intracellular levels of oxalate, and the sub- along fungal hyphae could be used later by the fungus sequent maintenance of stable pH levels and oxalate as a carbon source, under nutrient-limiting conditions anions outside fungal hyphae (Micales, 1997, Makela (Pion et al., 2013). In that case, bacterial growth was et al., 2010). Th e second role is related to the sequen- partially sustained by the consumption of fungal exu- tial action with the intracellular enzyme formate de- dates, which can be also the case for non-oxalotrophic hydrogenase, which decomposes formate (itself a pro- bacteria dispersed towards Ca-Oxalate medium. In the duct of oxalate oxidation) into CO2 and NADH. Th is case of the fungal and bacterial species studied here, we NADH could be used for ATP synthesis during fungal could not detect a diff erence in the metabolic activity growth (Watanabe et al., 2005). Th e main reaction is measured by IMC between migrating and non-mi- summarized hereunder: grating couples, in both MA and Ca-Oxalate media. However, in MA medium, the activity of the individual organisms was higher than when the co-cultures Oxalate decarboxylase Oxalate Formate + CO2 were prepared, especially for non-migrating couples. Although a larger number of samples and media need Formate dehydrogenase Formate CO2 + NADH to be considered, our data might provide a fi rst hint on the role of the nutrient conditions on the favoring of bacterial dispersal on fungal networks. It is probable Th ere are reports of the presence of oxalate decar- that these fungal-bacterial interactions are very com- boxylase and formate dehydrogenase in several strains plex, and not interpretable only with measures of heat of Fusarium oxysporum (broadinstitute.org, Uchimura production rates. et al., 2002). Th e same can be the case of our fungal strains, all belonging to the genus Fusarium. Th us, it is possible that the dissolution of calcium oxalate in Conclusion culture media, and the metabolic activity measured when calcium oxalate was provided as sole carbon In this study, we present a fi rst report of the ability of source corresponded to an energy-producing reaction fungi to solubilize calcium oxalate in the oxalate-car- based on calcium oxalate. bonate pathway, which adds an essential role of fungi in this pathway. Our fi ndings point to the fact that Diff erential bacterial migration fungi might not only be able to solubilize calcium oxa- Th e migration of oxalotrophic bacteria on fungal late, but that they might also be able to use this re- highways has been used as the framework to explain source as carbon source. Bacteria potentially associated the role of fungal-bacterial interactions in the oxa- to these fungi were not able to dissolve calcium oxalate-carbonate pathway (Bravo et al., 2013). However, late, but dispersed preferentially along fungal hyphae none of the bacteria studied here were able to dis- towards a medium containing calcium oxalate as sole solve calcium oxalate. While these bacteria grew easily carbon source. Th is was not the case for a rich culture in MA medium (Simon et al. 2015) and developed medium. However, the hypothesis of migrating bacte- poorly in Ca-Oxalate medium, migration was more ria feeding on fungal exudates could not be assessed. frequently observed in separated Petri dishes containing Ca-Oxalate medium (all strains, if only migrating bacteria are considered), compared to MA (2 out of Supplementary information 6 strains). Th us, fungal highways might not only be

111 Supplementary information is available in the digital crobial oxalotrophic activity. FEMS Microbiol Ecol version of this study. 78: 266-274. Bravo D, Cailleau G, Bindschedler S, Simon A, Job D, Verrecchia E & Junier P, 2013. Isolation of oxalotro- Acknowledgements phic bacteria able to disperse on fungal mycelium. FEMS Microbiology Letters. We would like to thank Dr. Vincent Hervé (Univer- Cailleau G, Braissant O & Verrecchia EP, 2011. Tur- sity of Neuchâtel, Switzerland) for his advice on the ning sunlight into stone: the oxalate-carbonate statistical analysis. Th is research was supported by the pathway in a tropical tree ecosystem. Biogeosciences Swiss National Science Foundation through Grants 8: 1755-1767. FN CR22I2-137994/1 and FN CR3212-149853/1. Cassland P, Sjode A, Winestrand S, Jonsson LJ & Nil- vebrant NO, 2010. Evaluation of oxalate decarboxylase and oxalate oxidase for industrial applications. References Appl Biochem Biotechnol 161: 255-263. de Boer W, Folman LB, Summerbell RC & Boddy L, Aguilar C, Urzua U, Koenig C & Vicuna R, 1999. 2005. Living in a fungal world: impact of fungi on Oxalate Oxidase from Ceriporiopsis subvermispora: soil bacterial niche development. FEMS Microbiol Biochemical and Cytochemical Studies. Archives of Rev 29: 795-811. Biochemistry and Biophysics 366: 275-282. Franceschi VR & Horner HTJ (1980) Oxalate crystals Banitz T, Johst K, Wick LY, Schamfuss S, Harms H & in plants. Botanical Review 46: 361-427. Frank K, 2013. Highways versus pipelines: contribu- Gadd GM, Bahri-Esfahani J, Li Q, Rhee YJ, Wei Z, tions of two fungal transport mechanisms to effi cient Fomina M & Liang X, 2014. Oxalate production by bioremediation. Environ Microbiol Rep 5: 211-218. fungi: signifi cance in geomycology, biodeterioration Bonkat G, Bachmann A, Solokhina A, Widmer AF, and bioremediation. Fungal Biology Reviews 28: 36- Frei R, Gasser TC & Braissant O, 2012. Growth 55. of mycobacteria in urine determined by isothermal Fusarium Comparative Sequencing Project, Broad Ins- microcalorimetry: implications for urogenital tuber- titute of Harvard and MIT. http://www.broadinsti- culosis and other mycobacterial infections. Urology tute.org/ (14 December 2015, date last accessed). 80: 1169-1112. Graz M, Jarosz-Wilkolazka A & Pawlikowska-Pawlega Braissant O, Verrecchia EP & Aragno M, 2002. Is the B, 2009. Abortiporus biennis tolerance to insoluble contribution of bacteria to terrestrial carbon budget metal oxides: oxalate secretion, oxalate oxidase acti- greatly underestimated? Naturwissenschaften 89: vity, and mycelial morphology. Biometals 22: 401- 366-370. 410. Braissant O, Cailleau G, Aragno M & Verrecchia EP, Guggiari M, Bloque R, Aragno M, Verrecchia E, Job 2004. Biologically induced mineralization in the D & Junier P, 2011. Experimental calcium-oxalate tree Milicia excelsa (Moraceae): its causes and conse- crystal production and dissolution by selected wood- quences to the environment. Geobiology 2: 59-66. rot fungi. International Biodeterioration & Biode- Braissant O, Bonkat G, Wirz D & Bachmann A, 2013. gradation 65: 803-809. Microbial growth and isothermal microcalorimetry: Joseph E, Simon A, Mazzeo R, Job D & Wörle M, Growth models and their application to microcalori- 2012. Spectroscopic characterization of an innova- metric data. Th ermochimica Acta 555: 64-71. tive biological treatment for corroded metal artefacts. Braissant O, Bachmann A & Bonkat G, 2015. Mi- Journal of Raman Spectroscopy 43: 1612-1616. crocalorimetric assays for measuring cell growth and Kohlmeier S, Smits TH, Ford R, Keel C, Harms H & metabolic activity: methodology and applications. Wick LY, 2005. Taking the Fungal Highway: Mobi- Methods 76: 27-34. lization of Pollutant-Degrading Bacteria by Fungi. Bravo D, Braissant O, Solokhina A, Clerc M, Daniels Environ. Sci. Technol. 39: 4640-4646. AU, Verrecchia E & Junier P, 2011. Use of an iso- Kong WJ, Zhao YL, Xiao XH, Li ZL, Jin C & Li HB, thermal microcalorimetry assay to characterize mi- 2009. Investigation of the anti-fungal activity of

112 coptisine on Candida albicans growth by microcalo- kage=PMCMR (15 November 2015, date last ac- rimetry combined with principal component analy- cessed). sis. J Appl Microbiol 107: 1072-1080. R Core Team. R: a language and environment for sta- Liang X, Moomaw EW & Rollins JA, 2015. Fungal tistical computing. Vienna, Austria: R Foundation oxalate decarboxylase activity contributes to Sclero- for Statistical Computing, 2015. http://www.R-pro- tinia sclerotiorum early infection by aff ecting both ject.org/ (15 November 2015, date last accessed). compound appressoria development and function. Russel BR & Cook GM, 1995. Energetics of bacterial Mol Plant Pathol 16: 825-836. growth: balance of anabolic and catabolic reactions. Makela MR, Hilden K & Lundell TK, 2010. Oxalate Microbiological reviews 59 (1): 48-62. decarboxylase: biotechnological update and preva- Sahin N, 2003. Oxalotrophic bacteria. Research in lence of the enzyme in fi lamentous fungi. Appl Mi- Microbiology 154: 399-407. crobiol Biotechnol 87: 801-814. Simon A, Bindschedler S, Job D et al., 2015. Exploi- Makela MR, Hilden K, Hatakka A & Lundell TK, ting the fungal highway: development of a novel tool 2009. Oxalate decarboxylase of the white-rot fun- for the in situ isolation of bacteria migrating along gus Dichomitus squalens demonstrates a novel en- fungal mycelium. FEMS Microbiol Ecol 91 (11). zyme primary structure and non-induced expression Trampuz A, Steinhuber A, Wittwer M & Leib SL, on wood and in liquid cultures. Microbiology 155: 2007. Rapid diagnosis of experimental meningitis 2726-2738. by bacterial heat production in cerebrospinal fl uid. Marin F, Roy NL, Marie B, Ramos-Silva P, Bundeleva BMC Infect Dis 7: 116. I, Guichard N & Immel F, 2014. Metazoan calcium Uchimura H, Enjoji H, Seki T, Taguchi A, Takaya N carbonate biomineralizations: macroevolutionary & Shoun H, 2002. Nitrate reductase-Formate de- trends - challenges for the coming decade. Bulletin hydrogenase couple involved in the fungal denitri- de la société géologique de France 185: 217-232. fi cation by Fusarium oxysporum. J. Biochem 131: Martin G, Guggiari M, Bravo D et al., 2012. Fungi, 579-586. bacteria and soil pH: the oxalate-carbonate pathway Verrecchia EP, Braissant O & Cailleau G, 2006. Th e as a model for metabolic interaction. Environ Mi- oxalate-carbonate pathway in soil carbone storage: crobiol 14: 2960-2970. the role of fungi and oxalotrophic bacteria. Fungi in Micales JA, 1997. Localization and induction of oxa- Biochemical Cycles 12: 289-310. late decarboxylase in the brown-rot wood decay fun- Watanabe T, Hattori T, Tengku S & Shimada M, gus Postia placenta. International Biodeterioration 2005. Purifi cation and characterization of NAD-de- & Biodegradation 39: 125-132. pendent formate dehydrogenase from the white-rot Nakata PA, 2003. Advances in our understanding of fungus Ceriporiopsis subvermispora and a possible calcium oxalate crystal formation and function in role of the enzyme in oxalate metabolism. Enzyme plants. Plant Science 164: 901-909. and Microbial Technology 37: 68-75. Pion M, Spangenberg JE, Simon A, et al., 2013. Bacte- Wu H, Chen W, Rong X, Cai P, Dai K & Huang Q, rial farming by the fungus Morchella crassipes. Proc 2014. Soil colloids and minerals modulate metabolic Biol Sci 280. activity of Pseudomonas putida measured using mi- Pohlert T. PMCMR: Calculate Pairwise Multiple crocalorimetry. Geomicrobiology Journal 31: 590- Comparisons of Mean Rank Sums. R package ver- 596. sion 1.1., 2015. http://CRAN.R-project.org/pac-

113

Chapter 3b - Fungi and nitrogen fi xation

Manuscript

Th is section presents the following manuscript, in preparation:

Simon A, Zopfi J, Verrecchia E, Junier P. To fi x or not to fi x? Th e struggle of starving fungi of nitrogen.

115

To fi x or not to fi x? Th e struggle of starving fungi of nitrogen

Anaele Simon1, Jakob Zopfi 2, Eric Verrecchia3, Pilar Junier1 1 Laboratory of Microbiology, Institute of Biology, University of Neuchâtel, Neuchâtel, Switzerland 2 Aquatic and Stable Isotope Biogeochemistry, Department of Environmental Sciences, University of Basel, 4056 Basel, Switzerland 3 Biogeosciences laboratory, Institute of Earth Surface Dynamics, University of Lausanne, 1015 Lausanne, Switzerland

Abstract

It is commonly accepted that biological nitrogen fi xation is performed only by prokaryotes, and culture in nitrogen-free media, as well as acetylene reduction tests, are performed in routine to assess the ability of organisms to fi x atmospheric nitrogen. In this study, we tested the ability of various fungal strains of the genus Fusarium and one strain of Chaetomium globosum to grow in nitrogen depleted media and to reduce acetylene. Th e selected fungal strains were able to grow after 9 successive transfers in nitrogen-free media, and to produce ethylene in the acetylene reduction test. Th ey were also able to grow in a nitrogen-free atmosphere, and we suggest as explanation that technical, purifi ed and ultrapure agars and agaroses can be a source of nitrogen for fungi, and possibly also for bacteria. Th us, we highlight the limitation of cultural methods, as well of the acetylene reduction technique, for the assessment of nitrogen-fi xing abilities. In conclusion, we were not able to disprove the ability of fungi to fi x nitrogen.

Introduction - + N2 + 16 Mg ATP + 8 e + 8 H Nitrogen is a very abundant element in the atmosphere Nitrogenase (about 78 % of its volume; Atkins, 1992), but under 2 NH + H +16 Mg ADP + 16 Pi the form of dinitrogen (N ), which is not available for 3 2 2 most of the living organisms. Th us, despite of its abun- Biological nitrogen fi xation is very important for dance, nitrogen is often the limiting factor in terres- agricultural systems, as several crops, and especially trial ecosystem productivity (Levy-Booth et al., 2014). legumes, are naturally associated with nitrogen-fi xing In order to become bioavailable, dinitrogen needs to bacteria (Herridge et al., 2008). be fi xed, meaning that it has to be reduced into am- Diff erent methods are commonly used for the as- monia (NH ; Boyd & Peters, 2013). It is commonly 3 sessment of nitrogen fi xation by bacteria, such as accepted that biological nitrogen fi xation is exclusively observation of growth in nitrogen-free culture media performed by prokaryotic organisms (bacteria and ar- (Döbereiner, 1980, Krieg & Gerhardt, 1994, Park et chaea; Leigh, 2000), and mediated by the enzyme ni- al., 2005, Stella & Suhaimi, 2010), measurement of trogenase, with the following general reaction (Cheng, acetylene reduction (Burris, 1975, Stella & Suhaimi, 2008): 2010, Boyd & Peters, 2013), or specifi c amplifi cation of genes coding for the enzyme nitrogenase (Rosch et al., 2002, Cheng, 2008, Levy-Booth et al., 2014).

117 Researches from the late 19th Century until the Materials and Methods middle of the 20th Century have proposed that fungi could also be able to fi x nitrogen (Lipman, 1911), Fungal strains and this statement has been rejected and re-approved We selected 8 fungal strains previously isolated from a many times (Allison et al., 1934). We were surprised semi-arid Moroccan soil by Simon et al. (2015): one to fi nd that one single study, entitled “Nitrogen fi xa- strain of Chaetomium globosum, two strains of Fusa- tion in Moulds and Yeasts – a reappraisal” (Millbank, rium oxysporum, three strains of Fusarium chlamydos- 1969) seems to be the basis of the nowadays general porum, one strain of Fusarium equiseti, and one strain acceptance of the inexistence of nitrogen fi xation in of Fusarium nygamai. fungi (Yamada & Sakaguchi, 1980, Postgate, 1982, van Kuijk et al., 2015), and many publications focu- Growth in nitrogen-free media sing on nitrogen fi xation in bacteria do not mention Th e classical nitrogen-free (NF) medium used was any source linked to the claim that only prokaryotes based on the one presented by Döbereiner (1980). It can fi x nitrogen (for example, Raymond et al., 2004, consisted of (for 1 l of ultrapure water): 20 g sucrose, Santos et al., 2012). With a closer look to Millbank’s . 0.05 g K2HPO4, 0.15 g KH2PO4, 0.01 g CaCl2 2 study, we realized that the tests to disprove nitrogen . . H2O, 0.2 g MgSO4 7 H2O, 0.002 g Na2MoO4 2 H2O, fi xation was performed only with yeasts, namely Bul- 0.01 g FeCl3, 1 g CaCO3 (all products from Sigma-Al- lera sp., Torulopsis sp., Rhodotorula sp., and Pullularia drich, St-Louis, USA), and 15 g purifi ed agar (Merck, sp. (= Aureobasidium pullulans; Brown & Lindberg, Darmstadt, Germany; lot number VM388714 204). 1967). Th e authors claimed to use yeasts and moulds, Th e exact elemental composition of this lot number of certainly because they considered Pullularia sp. as a purifi ed agar was preliminarily assessed, as well as the mould, which is not correct, as it is a yeast able to grow composition of technical agar (Pion et al., 2013). in both single-cell and fi lamentous forms (Brown & Lindberg, 1967, Zalar et al., 2008). Th e tests perfor- Unless otherwise stated, fungal inoculum consisted med in this study were cultivation on “nitrogen-free” of a circle (Ø 5 mm) cut out from a solid medium and medium (which contained 4.5 mg.l-1 of nitrogen) and placed in the center of a Petri dish. Incubation was acetylene reduction tests. However, it is interesting to performed at room temperature. notice that later publications still refer to the ability of fungi to fi x dinitrogen (Rangsawami et al., 1975, Gin- In order to assess the ability of the fungi to grow terova & Gallon, 1979). Th us, it is important to men- in NF medium, each fungal strain was inoculated in tion that nitrogen fi xation in fungi remains unproven NF medium and incubated during 7 days. As these (Wainwright, 1988, Laere, 1995). fi rst inoculums came from a nitrogen-containing me- Th e present study follows an unexpected observa- dium, 3 successive inoculations from NF medium to tion we made in the laboratory: several fungal strains NF medium were performed, each after 7 days of incu- could grow in a classical nitrogen-free medium. Th us, bation, in order to starve the fungi of nitrogen (Figure we wanted to know more about the relation of these 1). Growth of fungi was visually assessed after 7 days fungal strains with nitrogen. For this, we cultivated se- of incubation, for each generation. veral generations of each fungal strain on nitrogen-free Th e growth rate of each fungus was observed by medium, and made acetylene reduction tests. In addi- measuring the mycelial diameter on three successive tion, we cultivated the fungi in modifi ed atmospheres generations in NF medium, 47 to 128 h after inocula- completely free of nitrogen. tion, in triplicates (Figure 2, Table S1). Our aim was to assess whether these fungi were able to fi x dinitrogen, and if it was not the case, to fi nd ano- An attempt of measuring the biomass and proteins ther explanation to this ability to grow in nitrogen-free produced on several generations of fungi in NF me- medium, and thus to initiate a refl ection on current dium was performed as following. Whatman® polycar- theories concerning the study of nitrogen fi xing proka- bonate circular membranes (ref. 7060-4702, pore size ryotes. 0.2 μm, Ø 47 mm) were autoclaved and placed in a Petri dish containing NF medium. Each fungus was

118 Figure 1. General experimental setup and main results. A) Fungi were cultivated successively in NF medium and in NF medium with fosfomycin. Growth was visible in all plates. B) Acetylene reduction tests were performed. All fungi showed higher values of ethylene production than the negative control. C) Fungi were grown in NF medium with controlled atmosphere. Fungi did grow in atmosphere containing only argon and oxygen. D) Fungi were cultivated successively in NF and UNF media. Growth was visible in all plates. Growth rate was the same in all plates. E) Following D, fungi were cultivated successively in SC media. Growth was visible in the fi rst plate, but it was too hard to ensure presence of fungi in the subsequent inoculum.

inoculated on a membrane, in triplicates. Th e idea was visible, so we did not pursue the experiment. to separate the fungi from the fi lters after 9 days of incubation (the same way it was done in Simon et al. In order to prevent bacterial contamination, that (2015)), to weight each fungal pellet, and to perform might explain the ease of these fungi to grow in NF a protein extraction. Unfortunately, unlike the expe- medium, we prepared NF-F medium, consisting of riments of Simon et al. (2015), in which it was very NF medium added with 128 ml.l-1 of fosfomycin, a ni- easy to detach the fungi from nitrocellulose and cellu- trogen-free large spectrum antibiotic (Lu et al., 2011). lose acetate membranes, in this case it was impossible Th ree fungal strains out of the 8 (Fusarium oxysporum, to properly detach the fungi from the polycarbonate Chaetomium globosum, and Fusarium chlamydosporum) membranes, preventing us from pursuing the experi- were inoculated (from a pre-culture in NF medium, 3rd ment. generation in NF medium) in NF-F medium, in tri- Another attempt of measuring biomass was perfor- plicates. In order to assess that the NF-F medium was med by growing the fungi in NF medium and subse- able to prevent the growth of bacteria, nitrogen-fi xing quently liquefy the medium with micro-waves in order bacterium Azotobacter chroococcus NEU 1159 was used to remove the fungi alone, but we were not able to as a control and inoculated simultaneously in NF and properly separate the medium from the fungi. NF-F media, in triplicates. Growth of fungi and bacte- A third attempt of measuring biomass and protein ria was visually observed after 48h, and the diameter of production was performed by growing the fungi in li- the fungi was compared to the diameter of fungi after quid NF medium (without any agar). All fungi were 48h in NF medium. pre-grown in solid NF medium, inoculated in 10 ml of liquid NF medium, and incubated in an orbital shaker As purifi ed agar was contaminated with nitrogen at 145 rpm. After 3 days of incubation, no growth was (see results section for exact values), two fungal strains

119 (Chaetomium globosum and Fusarium oxysporum) were 2 ml of atmosphere inside the vials were removed and inoculated in NF medium in which the purifi ed agar replaced with 2 ml of acetylene (Carbagas, Gümlin- was replaced by Pulse Field certifi ed Agarose (BioRad, gen, Switzerland). Th e production of ethylene was Hercules, USA), in fi ve successive generations (UNF measured by injecting 100 μl of the headscape into a medium). Th e inoculum was taken from a 3rd genera- SRI Instruments 8610 C gas chromatograph equipped tion in NF medium. Growth of fungi was visually ob- with a HayeSep T column (at 70 °C) and a fl ame io- served, and colony diameters compared to the diame- nization detector (FID at 150 °C). Molecular nitrogen ters in NF media, after 47 to 100 h (Figure 2, Table at a fl ow rate of 29 ml.min-1 was used as carrier gas. S1). Measurements were performed after ca. 8 h, 33 h, and For the growth rate, regression was applied (R2 ≥ 178 h. 0.9995) in order to obtain continuous data, and sta- Azotobacter chroococcus NEU 1159 was used as a po- tistical analyses were performed with R (R Core Team, sitive control, and sterile vials as negative controls. 2013). A Kruskal-Wallis rank sum test was used in association with a pairwise test for multiple comparisons Growth in modifi ed atmosphere of mean rank sums (Nemenyi-tests; Pohlert, 2015) for In order to test whether atmospheric nitrogen was es- the analysis of the variance between the diameter of sential for the growth of our fungal strains, we perfor- the colonies, the generations and the culture medium. med cultures in modifi ed atmosphere. Sterile 250 ml vials were fi lled with 13 ml of hot NF In order to try another way to avoid the presence of medium, and placed on a hotplate in order to keep the nitrogen in the culture medium, we inoculated fungi medium boiling. Th e vials were sealed, and vacuum Chaetomium globosum and Fusarium chlamydosporum, was created inside the bottles, in order to avoid the from the 4th generation in UNF medium (Figure 1), presence of residual atmospheric gases inside the me- in a solid medium containing silicic acid and calcium dium after solidifi cation. Th e bottles were then fi lled as gelifying agents (SC medium; Krieg & Gerhardt, with gas, at a pressure of 0.8 bars: 9 bottles were fi lled 1994). Th e preparation of this medium was as follows: with argon, and 9 bottles were fi lled with dinitrogen

2 l of a 3 % sodium silicate 38 °Bé (Schneiter, Neu- (N2). Th is ensured the absence of any other gas in châtel, Switzerland) in ultrapure water were fi ltrated the media after solidifi cation. Th e bottles were cooled through a cationic resin column (Dowex 50 W x 8, 50- down, opened, and inoculated with the following fun- 100 mesh, charged in H+). Twenty milliliters of this so- gi (pre-cultured in NF medium), in triplicates: Chaeto- lution were poured in each Petri dish, and mixed with mium globosum, Fusarium chlamydosporum, and Fusa- . -1 2 ml of CaCO3 solution (1 g l ). After gelifi cation, 1 rium oxysporum. Each bottle was sealed again, vacuum ml of sterile sucrose solution (6 g.l-1) was spread on was performed, and the atmosphere was replaced as each plate. following: in the bottles preliminarily fi lled with ar- Fungi were incubated during 3 days, and then reino- gon, the atmosphere was replaced by a mixture of 80 culated in another SC medium. Growth of fungi was % Ar and 20 % O2; in the bottles preliminarly fi lled visually assessed. with N2, the atmosphere was replaced with a mixture

of 80 % N2 and 20 % O2. Growth of fungi was visually Acetylene reduction tests assessed after 10 days. In order to test for the ability of fungi to fi x atmospheric dinitrogen, we performed an acetylene reduction test. Th e principle of this test is to add acetylene Results

(C2H2) in the atmosphere, and measure the rates of ethylene (C2H4) produced, as the requirements for Growth in nitrogen-free media

C2H2 reduction are the same as for N2 reduction (Bur- Th e control of the exact composition of purifi ed and ris, 1975). technical agar showed a contamination in nitrogen for All fungal strains were pre-cultured in NF medium both (0.109 % (wt) N in the purifi ed agar, and 0.074 during 7 days, and inoculated in 20 ml vials contai- % (wt) N for the technical agar; Pion et al., 2013). ning 8 ml of NF medium. After 6 days of incubation, Th is means that the NF culture medium contained

120 Figure 2. Regression of the diameter of fungal cultures in function of time (R2 ≥ 0.9995). C 1-3: Fungus Chaetomium globosum on NF plates. C 4,7,8: Fungus C. globosum on UNF plates. F 1-3: Fungus Fusarium chlamydosporum on NF plates. F 4,7,8: Fungus F. chlamydosporum on UNF plates. Growth rate between the two media was signifi cantly diff erent (p-value = 0.00002****).

around 16 mg.l-1 of nitrogen. mycelium from one plate to another, and if we eff ec- All successive generations of fungi showed visible tively faced an absence of growth or an absence of fun- growth in NF, resp. UNF media. gal mycelium. Th e growth rate was signifi cantly diff erent between the two culture media, NF and UNF (p-value = As already mentioned in the material and methods 0.00002****; Figure 2). When we compared the section, several attempts of measuring the biomass and growth rate between generations, there was a signifi - the production of proteins were performed. When cant diff erence between generations 2 and 7 (p-value fungi were grown on fi lters, we were not able to en- = 0.036*) and between generations 2 and 8 (p-value tirely detach them from the fi lters. When fungi were = 0.028*). inoculated in liquid NF medium, no growth was vi- Fungal growth was visible in all NF-F plates with sible. When fungi were grown in solid NF medium, fosfomycin. After 47 h, mean diameter was of 20 ± 0.8 we were not able to properly detach the fungus from mm for F. oxysporum, 16 ± 0 mm for C. globosum, and the medium after its liquefaction. 23 ± 0.8 for F. chlamydosporum. In comparison, at the same time, mean diameter in NF medium was of 21.9 Acetylene reduction test ± 1.6 mm for F. oxysporum, 18.2 ± 1.2 for C. globosum, Acetylene reduction test was performed in order to and 25.3 ± 1.1 for F. chlamydosporum. Growth of bac- assess whether fungi were able to reduce acetylene terium Azotobacter chroococcus was not visible in NF-F into ethylene, as a proxy for nitrogen fi xation. After medium, but it was visible in NF medium. approx. 8 h, the mean values of ethylene production were of 1293 units for the positive controls (Azoto- Fungal growth was visible in all plates of SC me- bacter chroococcus NEU 1159), 36 units for fungi, and dium, in the fi rst generation. In the second generation, 21 for the negative controls. After approx. 33 h, the fungal growth was visible in two of the six plates. But mean values of ethylene production were of 10 995 it appeared that it was very diffi cult to remove some of for the positive controls, 129 for the fungi, and 45 for this culture medium in order to subculture the fungi. the negative controls. After approx. 178 h, the mean Th us we are not certain that we eff ectively transferred values of ethylene production were of 12 772 for the

121 Figure 3. Acetylene reduction test. Left: ethylene production by fungi (blue), negative controls (red) and positive controls (green; Azotobacter chroococcus NEU 1159), independently from time; p-value ≤ 0.03. Right: ethylene production in function of time, same color code.

positive controls, 942 for the fungi and 46 for the ne- nitrogen. However, most references about cultures of gative controls. Ethylene production by fungi was hi- nitrogen-fi xing prokaryotes in nitrogen-free media do gher than ethylene production by the negative control not make any particular statement about the conta- (p-value = 0.025*), but lower than ethylene produc- mination of nitrogen in the agar, and use “classical” tion by the positive control (p-value = 0.007**; Figure bacteriological grade agar (Jensen & Holm, 1975, Dö- 3 and Table S2). bereiner, 1980, Park et al., 2005, Stella & Suhaimi, 2010, Baldani et al., 2014). Th e fact that our fungi Growth in modifi ed atmosphere were able to grow in NF, UNF, and SC media could After 10 days of incubation in an atmosphere of 80 % be explained as follows. Fungi have the ability to selec-

Ar and 20 % O2, all fungi were able to grow, and no tively auto-digest their own hyphae in nutrient-deple- visible diff erence was noticed between these cultures ted areas (Lahoz et al., 1976). As they are also able to and the ones in an atmosphere containing 80 % N2 translocate nutrients and cell wall components inside and 20 % O2. their hyphae (Markham, 1995, Prosser, 1995), they could send the parts they digested to the active apical areas, and continue to grow in nitrogen-depleted areas. Discussion If the mycelial network is less dense in nutrient-poor areas, growth rate could be the same in diff erent me- Even if the chemical structure of agar polysaccha- dia, even if the biomass would be diff erent. But the rides can vary, none of this population of molecules fact that the tested fungi were able to grow successively is supposed to contain nitrogen (Lahaye & Rochas, in 9 generations of “nitrogen-free” media suggests that 1991). In technical agar, we can assume that the they still fi nd an appropriate source of nitrogen in their contamination in nitrogen was due to residual impu- environment. rities after agar extraction from algal cultures (Krieg & Th e source of nitrogen could originate from the at- Gerhardt, 1994). However, it was surprising to fi nd mosphere, and this would imply that fungi are able to that purifi ed agar was more contaminated with ni- fi x dinitrogen, directly or indirectly, the latter implying trogen than technical agar. Unfortunately, we do not presence of endosymbiotic nitrogen-fi xing bacteria in- know in which form the nitrogen is present in agar, side their hyphae. Th e acetylene reduction test showed and the packages inserts we read for commercial agars that fungi were able to produce a certain amount of and agaroses, whether technical, purifi ed, or ultrapure, ethylene, which was signifi cantly higher than a nega- did not mention any measurement of the content in tive control, but signifi cantly lower than a common

122 diazotrophic bacterium, Azotobacter chroococcus (Fi- Acknowledgements gure 3). Th ese results can be interpreted in three ways: fungi are able to fi x nitrogen, or fungi host endosym- We would like to thank Dr. Frank Schreiber (De- biotic bacteria that are able to fi x nitrogen, or fungi partment of Environmental Microbiology, EAWAG produce ethylene constitutively. We found only one Aquatic Research, Switzerland) for his help with the reference of a research work testing Fusarium species generation of modifi ed atmospheres. Th is research for their ability to produce ethylene (Ilag & Curtis, was supported by the Swiss National Science Founda- 1968), but the results section did not contain any data tion through Grants FN CR22I2-137994/1 and FN about ethylene production by Fusarium spp., so we de- CR3212-149853/1. duced without any certitude that the tested Fusarium spp. were negative for ethylene production. We did not fi nd any reference for the ability of Chaetomium References species to produce ethylene. Th e fact that fungi were able to grow in NF medium containing fosfomycin gi- Allison FE, Hoover SR, Morris HJ (1934) Nitrogen ves no indication for the presence of endosymbiotic fi xation studies with Fungi and Actinomyces. J Agric bacteria, but it excludes the presence of free bacteria Res 49: 1115-1123. contaminating the fungal cultures. Atkins P (1992) Chimie générale. InterEditions, Bo- However, if fungi were able to fi x dinitrogen, they logna, Italy. should not be able to grow in NF medium in an at- Baldani JI, Reis VM, Videira SS, Boddey LH, Balda- mosphere that does not contain any nitrogen. Th e fact ni VLD (2014) Th e art of isolating nitrogen-fi xing that our fungi were able to grow in these conditions bacteria from non-leguminous plants using N-free gives us an important information: there is something semi-solid media: a practical guide for microbiolo- in the NF medium that allows fungi to develop, and gists. Plant Soil. it is not possible that this compound is atmospheric Boyd ES, Peters JW (2013) New insights into the ammonium (NH4) or nitrate (NO3) imprisoned inside evolutionary history of biological nitrogen fi xation. the solid medium when it was poured, as vacuum was Front Microbiol 4: 201. made and atmosphere replaced when the medium was Brown RG, Lindberg B (1967) Polysaccharids from liquid. Th is could mean that fungi are able to feed on cell walls of Aureobasidium (Pullularia) pullulans. nitrogen contaminated agars. However, this statement Acta Chemica Scandinavica 21: 2379-2382. would imply that ultrapure grade agarose used for Burris RH (1975) Th e acetylene-reduction technique. UNF medium was enough contaminated in nitrogen In Nitrogen fi xation by free-living micro-organisms. to allow for fungal growth. Stewart WDP (ed.). Cambridge University Press, Cambridge, UK. Cheng Q (2008) Perspectives in biological nitrogen Conclusion fi xation research. J Integr Plant Biol 50: 786-798. Döbereiner J (1980) Forage grasses and grain crops. Th e fact that our fungal strains were able to grow Methods for Evaluating Biological Nitrogen. Ber- in classical nitrogen-free media in nitrogen-free at- gersen FJ (ed.). John Wiley, Sons Ltd, New York. mosphere indicates that they are able to use the ni- Ginterova A, Gallon JR (1979) Pleurotus ostreatus: a trogen that contaminates purifi ed agar and agarose nitrogen-fi xing fungus? Biochem Soc Trans 7: 1293- for their growth. Th is observation, and the fact that 1295. ethylene production was positive for each fungal strain Herridge DF, Peoples MB, Boddey RM (2008) Global in acetylene reduction tests, shows the limitation of inputs of biological nitrogen fi xation in agricultural cultural methods and acetylene reduction tests for the systems. Plant Soil 311: 1-18. assessment of nitrogen-fi xing abilities by fungi or bac- Ilag L, Curtis RW (1968) Production of ethylene by teria. fungi. Science 159: 1357-1358. Jensen W, Holm E (1975) Associative growth of nitrogen-fi xing bacteria with other micro-organisms.

123 In Steward WDP, ed. Nitrogen fi xation by free-li- kage version 1.1. http://CRAN.R-project.org/pac- ving micro-organisms. Cambridge University Press, kage=PMCMR. Cambridge, UK. Postgate JR (1982) Th e fundamentals of nitrogen fi xa- Krieg NR, Gerhardt P (1994) Solid, liquid/solid, and tion. Cambridge University Press, New York, USA. semisolid cultures. In Gerhardt P, Murray RGE, Prosser JI (1995) Kinetics of fi lamentous growth Wood WA, Krieg NR, eds. Methods for general and and branching. In Gow NA, Gadd GM , eds. Th e molecular bacteriology. American Society for Micro- growing fungus. Chapman, Hall, London, UK. biology, Washington DC, USA. R Core Team (2013) R: A language and environment Laere AV (1995) Intermediary metabolism. In Gow for statistical computing. R Foundation for Statisti- NAR, Gadd GM, eds.Th e growing fungus. Chap- cal Computing, Austria. man, Hall, London. Rangsawami G, Kandaswami T, Ramasamy K (1975) Lahaye M, Rochas C (1991) Chemical structure and Pleurotus sajor-caju (Fr) Singer - Protein rich ni- physico-chemical properties of agar. Hydrobiologia trogen-fi xing mushroom fungus. Curr Sci 44: 403- 221: 137-148. 404. Lahoz R, Reyes F, Leblic MIP (1976) Lytic enzymes in Raymond J, Siefert JL, Staples CR, Blankenship RE the autolysis of fi lamentous fungi. Mycopathologia (2004) Th e natural history of nitrogen fi xation. Mol 60: 45-49. Biol Evol 21: 541-554. Leigh JA (2000) Nitrogen fxation in methanogens: the Rosch C, Mergel A, Bothe H (2002) Biodiversity of archaeal perspective. Curr Issues Mol Biol 2: 125- denitrifying and dinitrogen-fi xing bacteria in an acid 131. forest soil. Appl Environ Microbiol 68: 3818-3829. Levy-Booth DJ, Prescott CE, Grayston SJ (2014) Mi- Santos PCD, Fang Z, Mason SW, Setubal JC, Dixon crobial functional genes involved in nitrogen fi xa- R (2012) Distribution of nitrogen fi xation and nition, nitrifi cation and denitrifi cation in forest eco- trogenase-like sequences amongst microbial ge- systems. Soil Biol Biochem75: 11-25. nomes. BMC Genomics 13. Lipman CB (1911) Nitrogen fi xation by yeasts and Simon A, Bindschedler S, Job D, et al. (2015) Exploi- other fungi. J Biol Chem 10: 169-182. ting the fungal highway: development of a novel tool Lu CL, Liu CY, Huang YT, Liao CH, Teng LJ, Tur- for the in situ isolation of bacteria migrating along nidge JD, Hsueh PR (2011) Antimicrobial suscep- fungal mycelium. FEMS Microbiol Ecol 91. tibilities of commonly encountered bacterial isolates Stella M, Suhaimi M (2010) Selection of suitable to fosfomycin determined by agar dilution and disk growth medium for free-living diazotrophs isolated diff usion methods. Antimicrob Agents Chemother from compost. J Trop Agric and Fd Sc 38: 211-219. 55: 4295-4301. van Kuijk SJA, Sonnenberg ASM, Baars JJP, Hendriks Markham P (1995) Organelles of fi lamentous fungi. WH, Cone JW (2015) Fungal treatment of lignocel- In Gow NA, Gadd GM, eds. Th e growing fungus. lulosic biomass: Importance of fungal species, colo- Chapman, Hall, London, UK. nization and time on chemical composition and in Millbank JW (1969) Nitrogen fi xation in moulds and vitro rumen degradability. Anim Feed Sc Tech 209: Yeasts - a reappraisal. Arch. Mikrobiol. 68: 32-39. 40-50. Park M, Kim C, Yang J, Lee H, Shin W, Kim S, Sa Wainwright M (1988) Metabolic diversity of fungi in T (2005) Isolation and characterization of diazotro- relation to growth and mineral cycling in soil - a re- phic growth promoting bacteria from rhizosphere view. Trans Br Mycol Soc 90: 159-170. of agricultural crops of Korea. Microbiol Res 160: Yamada T, Sakaguchi K (1980) Nitrogen fi xation as- 127-133. sociated with a hotspring green alga. Arch Microbiol Pion M, Spangenberg JE, Simon A et al. (2013) Bac- 124: 161-167. terial farming by the fungus Morchella crassipes. Proc Zalar P, Gostincar C, de Hoog GS, Ursic V, Sudhadham Biol Sci 280: 20132242. M, Gunde-Cimerman N (2008) Redefi nition of Pohlert T (2015) PMCMR: Calculate Pairwise Mul- Aureobasidium pullulans and its varieties. Stud My- tiple Comparisons of Mean Rank Sums. R pac- col 61: 21-38.

124 Chapter 3b - Fungi and nitrogen fi xation

Supplementary information

Table S1. Diameter of the fungal mycelium in the successives generations of culture in nitrogen-free media.

Table S2. Measures of ethylene production in the acetylene reduction test.

125

Chapter 3c - Migration and metabolism

Manuscript

Th is section presents the following mauscript, in preparation:

Simon A, Verrecchia E, Junier P. Is bacterial migration along fungal highways related to metabolic activity?

127

Is bacterial migration along fungal highways related to metabolic activity?

Anaele Simon1, Eric Verrecchia2, Pilar Junier1 1 Laboratory of Microbiology, Institute of Biology, University of Neuchâtel, Neuchâtel, Switzerland 2Biogeosciences laboratory, Institute of Earth Surface Dynamics, University of Lausanne, 1015 Lausanne, Swit- zerland

Abstract

In soils, fungi and bacteria play a crucial role in the bioavailability of many nutrients such as nitrogen, phosphorus and iron. Fungal highways, an interaction in which bacteria actively move along fungal hyphae, have been suggested as increasing bacterial activity and impacting nutrient cycling. In this study, we selected several fungal and bacterial strains known as interacting through fungal highways. We tested their respective metabolic abilities, as well as the ability of each bacterium to migrate along its associated fungal partner under diff erent nutritive conditions. Our fi ndings show that bacterial migration is not directly related to metabolic activity. A non-motile bacterial strain, Acinetobacter calcoaceticus, was able to migrate along its fungal partner. We hypothesize that this bacterium might be intrahyphal, and that it follows a translocation mechanism similar to mitochondria.

Introduction teria (Herridge et al., 2008). Phosphorus is often present in soils in association In soils, plants provide most of the organic carbon with metallic ions, which makes it biologically una- through litter deposition and root exudates (Ekschmitt vailable. Only a small fraction of phosphorus in soil et al., 2008). However, plant development, as well as is present as phosphates, which can be directly ab- overall ecosystem productivity, is often limited by ni- sorbed by plants or microbial cells (Ziadi et al., 2013). trogen (Levy-Booth et al., 2014), phosphorus (Ziadi However, some fungi and bacteria are able to solubilize et al., 2013), and iron (Ahmed & Holmstrom, 2014). inorganic phosphorus, thus increasing the amount of Fungi and bacteria play a major role in the bioavailabi- bioavailable phosphorus in soils (Kucey, 1987, Nau- lity of these nutrients. tiyal, 1999). Nitrogen is a very abundant element in the atmos- Iron oxides are also not bioavailable, because of their insolubility. Many fungi and bacteria are able to phere, but in form of dinitrogen (N2), which is an inactive, not bioavailable form. In order to become produce siderophores, which are low-molecular mass bioavailable, dinitrogen needs to be fi xed, i.e. reduced compounds able to chelate metals, thus making them bioavailable (Crowley, 2006, Ahmed & Holmstrom, into ammonia (NH3; Boyd & Peters, 2013). Th e vast majority of nitrogen fi xation is performed biological- 2014). ly (Fowler et al., 2013), and it is generally accepted While fi lamentous fungi are well-suited for the co- that only prokaryotic organisms are able to fi x nitrogen lonization of soils, because their hyphae are able to (Leigh, 2000). Nitrogen-fi xing bacteria play a major bridge the air-fi lled voids, thus colonizing water un- role in agricultural systems, as several crops, and espe- saturated soil patches (Wösten, 2001, Ritz & Young, cially legumes, are naturally associated with these bac- 2004), this is not the case for bacteria, whose disper-

129 sal is usually considered as very limited in soils (Or bacter calcoaceticus (H3) and Stenotrophomonas rhizo- et al., 2007, Ekschmitt et al., 2008). However, it has phila (H5). been demonstrated that some bacteria are able to actively move along fungal hyphae (taking the so-called Metabolic properties fungal highway), thus increasing their motility in soils Th e following properties of fungi and associated bac- (Kohlmeier et al., 2005). It has been hypothesized that teria were tested: calcium oxalate degradation, solu- this phenomenon could increase bacterial activity, and bilization of inorganic phosphorous, growth in ni- thus could have an impact on elemental cycling (Wick trogen-free medium, and production of siderophores. et al., 2007, Martin et al., 2012, Banitz et al., 2013). Th e organisms were grown in diff erent culture media In this study, we selected fungal and bacterial strains in order to assess these properties. For bacteria, a loop previously isolated and known as interacting through from a 24 to 48 h pre-culture in nutrient agar (NA) fungal highways (Simon et al., 2015). We assessed their medium was inoculated in each culture medium. For ability to solubilize calcium oxalate, to fi x nitrogen, to fungi, a 2 mm square from a 2 weeks pre-culture on solubilize inorganic phosphorus and to produce side- malt agar (MA) medium was placed in each culture rophores. We also assessed the ability of each bacte- medium. rium to migrate along its associated fungus towards NA medium was composed of (for 1 l of deionized diff erent culture media. Our aim was to understand if water) 8 g nutrient broth (containing 3 g beef extract these fungal-bacterial interactions are related to com- and 5 g peptone; Biolife Italiana, Milano, Italy) and 15 plementary metabolic abilities between the fungal and g agar (Biolife Italiana, Milano, Italy). Beef extract has the bacterial associate, and if the carbon source and the an average C:N ratio of 0.01 (Ivusic & Santek, 2015) C:N ratio of these media could have an infl uence on and peptone has an average C:N ratio of 8 (Mwangi et bacterial migration. al., 2012). C:N ratio of nutrient broth was calculated as 5. MA medium was composed of (for 1 l of deionized Materials and methods water) 12 g malt (Sios Homebrew Shop, Wald, Swit- zerland) and 15 g agar. C:N ratio of malt extract is Microbial samples ranging between 14 and 28 (Cooke, 1968), and we Couples of fungi and migrating bacteria were pre- considered an intermediate value of 21. viously isolated from a soil profi le located under an Th e ability to degrade calcium oxalate was tested in oxalogenic plant, Opuntia fi cus-indica, in a semi-arid CaOx medium (modifi ed DSM81 medium; Braissant area of Morocco (Simon et al., 2015). Th e isolation et al., 2002), containing calcium oxalate as sole carbon was performed with fungal highway columns, a device source. Th e medium was poured in two solid layers in allowing for targeted isolation of soil fungi and asso- the Petri dishes. Th e fi rst layer (ca.20 ml) consisted of ciated migrating bacteria. Further details about the a 2:1 (v/v) mixture of solution A and an agar solution. sampling location, isolated organisms, as well as com- Th e second layer consisted of a 2:1 (v/v) mixture of so- . -1 . position and validation of the fungal highway columns lution A, added with 4 g l CaC2O4 H2O, and the agar are presented in Simon et al. (2015). solution. Th is second mixture was added with 2 ml.l-1 Th e following associated organisms were selected of solution B. Solution A was composed of (per l of . for this study. Fusarium nygamai (A) with bacterium milliQ water) 9 g Na2HPO4 12 H2O, 1.5 g KH2PO4, . Achromobacter mucicolens (A1), Fusarium chlamydos- 1 g NH4Cl, 0.2 g MgSO4 7 H2O, 1 ml trace elements porum (C) with bacterium Acinetobacter calcoaceticus solution DSM27 (composition (per l of milliQ water): . . (C1), Fusarium equiseti (D) with bacteria Achromobac- 10 mg ZnSO4 7 H2O, 3 mg MnCl2 4 H2O, 30 mg . . ter spanius (D1) and Stenotrophomonas humi (D2), Fu- H3BO3, 20 mg CoCl2 6 H2O, 1 mg CuCl2 2 H2O, 2 . . sarium oxysporum (F) with bacteria Ochrobactrum sp. mg NiCl2 6 H2O, 3 mg Na2MoO4 2 H2O, pH=3.5). (F2) and Stenotrophomonas maltophilia (F3), Fusarium Th e agar solution was composed of 15 g purifi ed agar chlamydosporum (G) with bacteria Ochrobactrum peco- per l of milliQ water. Solution B was composed of (per . ris (G1) and Pseudomonas frederiksbergensis (G2), and l of milliQ water) 0.5 g Fe(NH4)(SO4)2 12 H2O and . Fusarium chlamydosporum (H) with bacteria Acineto- 1 g CaCl2 H2O. Th e presence of a degradation halo

130 was visually assessed after 20 days. Th e bacterium Va- (CAS, C23H13Cl2Na3O9S) solution 0.12 % (w/v), 10 . riovorax soli was used as positive control. C:N ratio was ml FeCl3 6 H2O solution 0.0027 % (w/v) and 40 ml calculated as 3. HDTMA solution 0.18 % (w/v). Th e bacterium Serra- Th e ability of each organism to solubilize inorganic tia ureilytica NEU 1313 was used for positive control. phosphorous was tested in NBRIP medium (National Change of color from blue to red/yellow was visually Botanical Research Institute’s phosphate growth me- assessed after 10 days. C:N ratio was calculated as 17. dium; Nautiyal, 1999), consisting of (per l of milliQ . water) 10 g glucose, 5 g Ca3(PO4)2, 5 g MgCl2 6 H2O, Diff erential bacterial migration . 0.25 g MgSO4 7 H2O, 0.2 g KCl, 0.1 g (NH4)2SO4 Th e ability of each bacterium to migrate along its as- and 15 g purifi ed agar. pH was adjusted to 7.0 before sociated fungus was determined using separated Pe- autoclaving. Th e bacterium Enterobacter cloacae NEU tri dishes consisting of two culture media in a Petri 1027 was used as a positive control. Th e presence of dish, separated from each other by a 5 mm gap wit- a degradation halo was visually assessed after 13 days hout culture medium. One of the media of the sepa- (bacteria), resp. 15 days (fungi). C:N ratio was calcu- rated Petri dishes consisted of NA medium; the other lated as 667. part consisted of MA, solid CAS (with 15 g.l-1 agar), Th e ability of each bacterial strain to fi x nitrogen was NBRIP, CaOx, or NF media. tested in NF medium (nitrogen-free medium; Döbe- Bacteria were pre-cultured during 48 h in NA me- reiner, 1980), consisting of (per l of milliQ water) 15 g dium. Th e content of half a Petri dish was transferred sucrose, 0.05 g K2HPO4, 0.15 g KH2PO4, 0.01 g CaCl2 in 1 ml of sterile physiological water (fi nal OD550 ran- . . . 2 H2O, 0.2 g MgSO4 7 H2O, 0.002 g Na2MoO4 2 ging from 1 to 10), of which 10 μl were inoculated in a

H2O, 0.01 g FeCl3, 1 g CaCO3 and 15 g purifi ed agar. 2.5 cm line in the NA medium-side of a separated Pe- pH was adjusted to 6.8 before autoclaving. Th e bac- tri dish, 1.5 cm away from the separation. Fungi were terium Azotobacter chroococcum NEU 1159 was used pre-cultured on MA medium. A 5 mm diameter circle as a positive control and the bacterium Escherichia coli was cut out and inoculated in the NA medium-side of NEU 1006 as a negative control. Th ree successive ino- the separated Petri dish, 2 cm away from the separa- culations were performed in order to properly deplete tion. Each fungal-bacterial couple was inoculated in the organisms from remaining nitrogen. Presence of triplicate in each type of separated Petri dishes. After colonies was visually assessed on the last inoculated 13 days of incubation at room temperature, the surface media after 48h. C:N ratio was calculated as infi nite, of the opposite culture medium in each separated Petri thus not applicable. dishes was scrubbed with a loop and re-plated in NA Th e ability of each organism to produce siderophores medium. After 24 and 48 h, growth of bacteria was was tested in vials with 10 ml of CAS medium (chrom visually assessed. azurol S medium; Louden et al., 2011, Supanekar et al., 2013), composed of (for 1 l of medium) 850 ml M9-PIPES solution, 3 ml casamino acid solution 10 Results %, 10 ml glucose solution 20 %, 1 ml vitamin solution 10 %, and 100 ml blue dye solution. Th e so- Metabolic properties lutions were mixed together after autoclaving. M9- After 20 days, no degradation halo was visible for bac- PIPES solution consisted of 100 ml M9 salt solution terial strains in CaOx medium, but all fungi produced . (composition in 1 l of milliQ water: 82 g Na2HPO4 a degradation halo (see chapter 3 of the present thesis

2 H2O, 30 g KH2PO4, 5 g NaCl, 10 g NH4Cl), 750 for further information). ml milliQ H2O and 30.24 g PIPES (C8H18N2O6S2). After 13 days of incubation, a degradation halo was Vitamin solution consisted of (per l of milliQ water) clearly visible in the NBRIP medium (inorganic phos- 0.002 g biotin, 0.02 g nicotinic acid, 0.01 g thiamin, phorous solubilization) for 3 out of the 10 bacterial 0.01 g p-aminobenzoate, 0.005 g pantothenic acid, strains, while no halo was visible for any fungus after 0.05 g pyridoxine, 0.01 g cyanocobalamin, 0.01 g 15 days of incubation. folic acid, 0.05 g ribofl avin). Blue dye consisted of a After 24 h of incubation in the third successive ino- mixture of (for 100 ml of dye) 50 ml chrome azurol s culated NF medium, growth was clearly visible for 8

131 out of the 10 bacterial strains, and no growth was vi- in NF medium, for 8 bacterial strains. sible for the two other bacterial strains. After 10 days of incubation, a change of color of the CAS medium, indicating the production of side- Discussion rophores, was observed for 4 out of the 10 bacterial strains, while a change of color was detected for all Metabolic activities fungal strains (Table 1). Regarding the metabolic processes, we observed that all Fusarium strains showed the same pattern on each Diff erential bacterial migration culture medium. Furthermore, while it has been re- Table 1 presents the results of the visual observation ported that fungi are able to solubilize phosphorus made for the migration of each bacterial strain along (Kucey, 1987), this was not the case for our Fusarium hyphae of the associated fungal partner in each condi- strains. It has to be noticed that we also tested the tion, as well as the observations previously made for ability of fungus Penicillium pinophilum to solubi- the same couples with scanning electron microscopy lize inorganic phosphorus (data not shown), which is (SEM) in MA and NA media (Simon et al., 2015). In consistent with the data for another Penicillium specie MA medium, the results obtained by SEM observations reported by Kucey (1987). and those by culturing were not entirely consistent. By As all fungi were able to dissolve calcium oxalate, SEM, only the bacterium O. pecoris (G1) was detected while none of the bacteria could, we expected that the beyond the separation, while in our observations by capabilities would be complementary between the fun- culturing, S. maltophilia (F3) and P. frederiksbergensis gal and the bacterial partner. However, while most of (G2) were detected beyond the separation. the bacteria were able to grow on NF medium, their Th e bacterium Acinetobacter calcoaceticus (H3) was metabolisms were various regarding solubilization of detected in none of the migration tests. None of the inorganic phosphorus and production of siderophores. bacteria was positive for all migration tests. In CaOx medium, bacterial colonies were observed Diff erential bacterial migration in the opposite culture medium of the separated Petri Th e diff erences between the results of migration with dishes in 6 out of the 10 couples tested. In NBRIP me- SEM and with visual detection of bacterial colonies dium, this was the case for 3 bacterial strains. In CAS can be explained in two ways. First of all, for the SEM medium, this was the case for 6 bacterial strains, and observation of the bacterium O. pecoris, we detected

Table 1. Metabolic activity of all selected fungi and bacteria: growth in NA medium, growth in MA medium, dissolution of calcium oxalate, dissolution of inorganic phosphorus, production of siderophores, and growth in nitrogen-free medium. Migration of each bacterial strain along its fungal partner, observations by SEM and cultivable method.

Calcium Inorganic NA MA Siderophores Nitrogen oxalate phosphorus Code Fungi Bacteria Code Growth Migr. Growth Migration Diss. Migr. Diss. Migr. Prod. Migr. Growth Migr. F B F B SEM* Cult F B Cult F B Cult F B Cult B Cult SEM* Fusarium Achromobacter A + + + + + - - + - + - - - + - + - + A1 nygamai mucicolens

Fusarium Acinetobacter C + + - + + - - + - - - + + + - + + + C1 chlamydosporum calcoaceticus Achromobacter + + + - - - + - + + + - + D1 Fusarium spanius D + + + - + equiseti Stenotrophomonas + + + ------+ - D2 humi + + + - - - + + - + - + + Ochrobactrum sp. F2 Fusarium F + + + - + Stenotrophomonas oxysporum + + + - + - + - + + + + + F3 maltophilia Ochrobactrum + + + + - - - +- + - + - G1 Fusarium pecoris G + + + - + chlamydosporum Pseudomonas + + + - + - + + - - + + +/- G2 frederiksbergensis Acinetobacter + - + - - - - + - - - + - H3 Fusarium calcoaceticus H + + + - + chlamydosporum Stenotrophomonas + + + - - - + - - - + + +/- H5 rhizophila * Data from Simon et al. 2015, observation with scanning electron microscopy

132 Table 2. Migration of each bacterial strain along its associated fungus, with emphasis on C :N ratio and main source of carbon in the culture media.

Medium CaOx NA CAS MA NBRIP NF

C:N ratio 3 5 17 21 668 NA

Main source of C CaC2O4 Peptone Glucose Starch Glucose Sucrose

Migration A1 + + + - - + C1 - - + - + + D1 + + + - + + D2 - + - - - - F2 + + - - - + F3 + + + + + + G1 - + - - - - G2 + + + + - + H3 ------H5 + + + - - +

only few bacterial cells in the sample (see additional hyphae. It is interesting to notice that movement of data at the end of this chapter), thus they might not mitochondria in hyphae follows a diff erent mechanism form visible colonies. Furthermore, the samples for than movement of other organelles (Markham, 1995), SEM were fi xed after 8 days, while the samples in the and maybe intrahyphal bacteria, even if non-motile present study were subcultured after 13 days. It is pos- (such as strain A. calcoaceticus (C1)), could be dis- sible that bacteria such as S. maltophilia and P. frede- persed the same way than mitochondria. riksbergensis would have been observed with SEM if the samples were fi xed after 13 days. Whether intrahyphal or extrahyphal, the factors However, bacterium Acinetobacter calcoaceticus (H3) that infl uence bacterial transport remain unclear. In was never observed in the opposite part of the sepa- Chapter 3 of the present thesis, we hypothesized that rated Petri dishes, neither by SEM observations, nor bacterial growth could be partially sustained by the by culture-based observation in diff erent media. It consumption of fungal exudates, and they could be was fi rst hypothesized that this strain could be a hit- transported and used by the fungus as carbon source chhiker on the fungal highway (Simon et al. 2015). under limiting conditions, as observed in a previous Yet, it is interesting to notice that the two strains of experiment (Pion et al., 2013). Even if we could not A. calcoaceticus (C1 and H3) were isolated in associa- validate this hypothesis, factors such as the carbon tion with the fungus Fusarium chlamydosporum. Th e source, or C:N ratio could infl uence bacterial mi- second strain of A. calcoaceticus (C1), while described gration. Indeed, it has been observed that limited C as non-motile (Simon et al., 2015), was able to migrate aff ects bacterial growth (Demoling et al., 2007), and along its associated fungus in media NBRIP, CAS and limited N aff ects fungal growth (Boyle, 1998), and NF. Furthermore, the same bacterial specie was identi- that C:N ratio aff ects the fungal:bacterial ratio in soils fi ed in association with fungus Mortierella sp. in fungal (Waring et al., 2013). Regarding C:N ratio, we did not highway columns (see Chapter 2 of the present thesis), see a clear shift in bacterial migration towards lower or and it was hypothesized that A. calcoaceticus could be higher C:N ratios (Table 2). If we take into conside- an intrahyphal bacterium. ration the carbon source, the results are not more en- It has been previously reported that fungi from the lightening. For example in the four culture media with genus Fusarium host endobacteria (Li et al., 2010, Ting a glucose-based carbon source (CAS (glucose), NBRIP et al., 2010), thus strongly supporting this hypothesis. (glucose), NF (sucrose), MA (starch)), we observed Th e fact that fungi are septate does not prevent them that NF, with a high C:N ratio and an easy-to-use car- to translocate organelles (Markham, 1995). Th us, even bon source, attracted a lot of migrating bacteria (8 out if Fusarium spp. are septate fungi (Samson et al., 2002), of 10), while NBRIP, as well containing a high C:N it is possible that they translocate bacteria inside their ratio and an easy-to-use carbon source, only attracted

133 3 out of 10 migrating bacteria. Boyd ES & Peters JW (2013) New insights into the Th us, it is clear that bacterial migration along fungal evolutionary history of biological nitrogen fi xation. hyphae does not follow a mono-parametric rule, and Front Microbiol 4: 201. it is probable that many parameters infl uence these Boyle D (1998) Nutritional factors limiting the growth fungal-bacterial interactions, and even that each fun- of Lentinula Edodes and other white-rot fungi in gal-bacterial interaction is related to specifi c parame- wood. Soil BiolBiochem 30: 817-823. ters. Braissant O, Verrecchia EP & Aragno M (2002) Is the contribution of bacteria to terrestrial carbon budget greatly underestimated? Naturwissenschaften 89: Conclusion 366-370. Cooke WB (1968) Carbon/nitrogen relathionships of In this study, we observed that fungi and bacteria in- fungus culture media. Mycopathologia et mycologia teracting through fungal highways show various meta- applicata 34: 305-316. bolic properties. Th e migration of bacteria along fun- Crowley DA (2006) Microbial siderophores in the gal hyphae was not directly related to the metabolic plant rhizosphere. In Barton LL & Abadía J, eds. abilities of each partner. We could not fi nd a relation Iron nutrition in plants and rhizospheric microorga- between C:N ratio or carbon source and bacterial mi- nisms. Springer, Dordrecht, Th e Netherlands. gration. We also hypothesize that two bacterial strains Demoling F, Figueroa D & Bååth E (2007) Compari- belonging to the same species (Acinetobacter calcoace- son of factors limiting bacterial growth in diff erent ticus) could be intrahyphal symbionts of fungus Fu- soils. Soil Biol Biochem 39: 2485-2495. sarium chlamydosporum, which would explain the fact Döbereiner J (1980) Forage grasses and grain crops. than one non-motile A. calcoaceticus strain migrates In Methods for evaluating biological nitrogen. Ber- along its associated fungus in specifi c culture media. gersen FJ (ed.). John Wiley & Sons Ltd, New York, In order to better understand these fungal-bacterial USA. associations, it would be interesting to assess which Ekschmitt K, Kandeler E, Poll C, et al. (2008) Soil-car- bacteria are intrahyphal associates, and which are ex- bon preservation through habitat constraints and trahyphal, in order to understand if the parameters ru- biological limitations on decomposer activity. J Plant ling the migrating abilities would be the same between Nutr Soil Sci 171: 27-35. these bacterial groups. Fowler D, Coyle M, Skiba U, et al. (2013) Th e global nitrogen cycle in the twenty-fi rst century. Philos Trans R Soc Lond B Biol Sci 368: 20130164. Acknowledgements Herridge DF, Peoples MB & Boddey RM (2008) Glo- bal inputs of biological nitrogen fi xation in agricul- Th is research was supported by the Swiss Na- tural systems. Plant Soil 311: 1-18. tional Science Foundation through Grants FN Ivusic F & Santek B (2015) Optimization of complex CR22I2-137994/1 and FN CR3212-149853/1. medium composition for heterotrophic cultivation of Euglena gracilis and paramylon production. Bio- process Biosyst Eng 38: 1103-1112. References Kohlmeier S, Smits TH, Ford R, Keel C, Harms H & Wick LY (2005) Taking the fungal highway: mo- Ahmed E & Holmstrom SJ (2014) Siderophores in bilization of pollutant-degrading bacteria by fungi. environmental research: roles and applications. Mi- Environ Sci Technol 39: 4640-4646. crob Biotechnol 7: 196-208. Kucey RMN (1987) Increased phosphorus uptake by Banitz T, Johst K, Wick LY, Schamfuss S, Harms H & wheat and fi eld beans inoculated with a phospho- Frank K (2013) Highways versus pipelines: contri- rus-solubilizing Penicillium bilaji strain and with ve- butions of two fungal transport mechanisms to ef- sicular-arbuscular mycorrhizal fungi. Appl Environ fi cient bioremediation. Environ Microbiol Rep 5: Microbiol 53: 2699-2703. 211-218. Leigh JA (2000) Nitrogen fi xation in methanogens:

134 the archaeal perspective. Curr Issues Mol Biol 2: Biol Sci 280. 125-131. Ritz K & Young IM (2004) Interactions between soil Levy-Booth DJ, Prescott CE & Grayston SJ (2014) structure and fungi. Mycologist 18: 52-59. Microbial functional genes involved in nitrogen fi xa- Samson RA, Hoekstra ES, Frisvad JC & Filtenborg tion, nitrifi cation and denitrifi cation in forest eco- O (2002) Introduction to food- and airborne fun- systems. Soil Biol Biochem 75: 11-25. gi, 6th edition. Ponsen & Looyen, Wageningen, Th e Li XS, Sato T, Ooiwa Y, Kusumi A, Gu J-D & Kataya- Netherlands. ma Y (2010) Oxidation of Elemental Sulfur by Fu- Simon A, Bindschedler S, Job D et al. (2015) Exploi- sarium solani Strain THIF01 Harboring Endobacte- ting the fungal highway: development of a novel tool rium Bradyrhizobium sp. Microb Ecol 60: 96-104. for the in situ isolation of bacteria migrating along Louden BC, Haarmann D & Lynne AM (2011) Use fungal mycelium. FEMS Microbiol Ecol 91. of Blue Agar CAS Assay for Siderophore Detection. Supanekar SV, Sorty AM & Raut AA (2013) Catechol J Microbiol Biol Educ 12. siderophore produced by Klebsiella pneumoniae iso- Markham P (1995) Organelles of fi lamentous fungi. lated from rhizosphere of Saccharum offi cinarum L. In Gow NA & Gadd GM, eds. Th e growing fungus. International Journal of Scientifi c Research 2. Chapman & Hall, London, UK. Ting ASY, Mah SW & Tee CS (2010) Detection of Martin G, Guggiari M, Bravo D et al. (2012) Fungi, potential volatile inhibitory compounds produced bacteria and soil pH: the oxalate-carbonate pathway by endobacteria with biocontrol properties towards as a model for metabolic interaction. Environ Mi- Fusarium oxysporum f. sp. cubense race 4. World J crobiol 14: 2960-2970. Micro Biot 27: 229-235. Mwangi ESK, Gatebe EG & Ndung’u MW (2012) Waring BG, Averill C & Hawkes CV (2013) Diff e- Impact of nutritional (C:N Ratio and source) on rences in fungal and bacterial physiology alter soil growth, oxalate accumulation, and culture pH by carbon and nitrogen cycling: insights from meta-ana- Sclerotinia Sclerotiorum. J Biol Agric Healthc 2. lysis and theoretical models. Ecol Lett 16: 887-894. Nautiyal CS (1999) An effi cient microbiological Wick LY, Remer R, Würz B, Reichenbach J, Braun S, growth medium for screening phosphate solubili- Schäfer F & Harms H (2007) Eff ect of fungal hy- zing microorganisms. FEMS Microbiol Lett 170: phae on the access of bacteria to phenanthrene in 265-270. soil. Environ Sci Technol 41: 500-505. Or D, Smets BF, Wraith JM, Dechesne A & Friedman Wösten HAB (2001) Hydrophobins: multipurpose SP (2007) Physical constraints aff ecting bacterial ha- proteins. Annu Rev Microbiol 55: 625-646. bitats and activity in unsaturated porous media – a Ziadi N, Whalen JK, Messiga AJ & Morel C (2013) review. Adv Water Res 30: 1505-1527. Assessment and modeling of soil available phospho- Pion M, Spangenberg JE, Simon A, et al. (2013) Bac- rus in sustainable cropping systems. Adv Agron 122: terial farming by the fungus Morchella crassipes. Proc 85-126.

135

Chapter 3 - Element cycling and metabolism

Additional experiments

Bacterial migration in malt agar and nutrient agar media

As presented in the manuscript of Chapter 1, bacteria do not migrate the same way in malt agar (MA) medium and in nutrient agar (NA) medium. Here, we present scanning electron microscopy images for all fungal-bacterial couples, in MA and NA medium (Figure A1). Th e samples were taken from separated Petri dishes, at the opposite part of the fungal and bacterial inoculum, after 8 days of incubation at room temperature. Details about the preparation of the samples are presented in the manuscript of Chapter 1.

Results Most of the bacteria were only visible in NA medium. For couple C1, in MA medium, hyphae were imprisoned in a mucilaginous layer, which was not present in NA medium. In NA medium, we observed a decaying fungal cell. Inside this cell, we observed structures that look like organelles, of the same size than bacteria (Figure A1, arrow 1). For couple D3, in MA medium, we observed nanospherolitic structures. Energy-dispersive X-ray spectroscopy (EDS) did not show presence of metals (data not shown), which suggests that they are organic structures. Th ese nanospherolitic structures were also observed in MA medium for couples E4, F3, G3, H2, H3 and H5 (Figure A1, arrows 2). Th ey were absent or less abundant in NA medium. For couple E6, fungus did not reach the opposite culture medium, thus it could be a false negative. For couple G1, only a few single bacteria were visible on the sample in MA medium. In NA medium, some bacteria were visible on the broken part of a hypha (Figure A1, arrow 3). For couple H2, in MA medium fungal hyphae were imprisoned in a mucilaginous layer, and bacteria were also present under this layer (Figure A1, arrow 4). For couple H5, in NA medium, bacteria were present on the hyphae, and we could observe a structure resembling a decaying hypha on which several bacteria were present (Figure A1, arrow 5).

Detection of intrahyphal bacteria

In order to detect intrahyphal bacteria, we performed two diff erent DNA extractions on pure fungal cultures, followed by amplifi cation of a fragment of the bacterial 16S rRNA gene.

Selected fungi Th e fungi we selected were all isolated from a Moroccan soil with fungal highway columns (see Chap- ter 1 of the present thesis for further details about sampling location and isolation processes). Codes in Figure A2 correspond to following fungal strains: Fusarium nygamai (A), Fusarium oxysporum (B),

137 138 139 140 Figure A1. Scanning electron microscopy observations of the medium at the opposite part of fungal-bacterial inoculum in separated Petri dishes. Left: in MA medium. Right: in NA medium.

Fusarium chlamydosporum (C), Fusarium equiseti (D), Chaetomium globosum (E), Fusarium oxysporum (F), Fusarium chlamydosporum (G), Fusarium chlamydosporum (H), Fusarium chlamydosporum (I), Peni- cillium pinophilum (J), Fusarium oxysporum (K).

DNA extraction and PCR amplifi cation We performed two diff erent DNA extractions on fungal pure cultures. Th e fi rst with a guanidin thio- cyanate method, and the second according to the instructions of PowerSoil DNA Isolation Kit (MoBio, Carlsbad, CA, USA) with a bead-beating step of 5 min at 50 beats·s−1 (Qiagen, Hilden, Germany). PCR amplifi cation was performed of a partial fragment of the 16S rRNA gene with the same primers than in Chapter 1 of the present thesis (EUB 9-27f and EUB 1542r). PCR master mix and protocol were exactly the same than in Chapter 1 of the present thesis. Presence of amplifi ed DNA fragments was assessed by a 30 min gel electrophoresis at 100 V on a 1.5 % agarose gel (Bioconcept, Allschwil, Switzerland). DNA was stained for duration of 30 min in a GelRed bath (Biotium, Hayward, CA, USA) before observation with a Transilluminator (VWR Inter- national). Samples presenting amplifi ed DNA were sent for Sanger sequencing to GATC Biotech AG (Konstanz, Germany). Search for similarity against sequences of the 16S rRNA gene was performed using BLAST (Basic Local Alignment Search Tool; Altschul et al., 1990).

Results DNA fragments on gel electrophoresis are presented in Figure A2. PCR performed on DNA extracted with guanidine showed less unspecifi c bands than PCR performed on DNA extracted with the Power-

141 Figure A2. Gel electrophoresis performed on the products of PCR amplifi cation of a partial fragment of the 16S rRNA gene on pure fungal cultures. Samples in green were sent for Sanger sequencing. a) DNA extraction performed with guanidine; b) DNA extraction performed with PowerSoil kit.

Soil kit. A clear band was present in both electrophoresis gels for fungi D (Fusarium equiseti) and F (Fusarium oxysporum). Sanger sequencing was bad for all samples except for fungus D. Th e amplifi ed fragments could not be identifi ed, they were assigned either to Limnohabitus sp. (95 % id.) or Curvi- bacter sp. (95 % id.).

Detection of the nifH gene

In Chapter 4a, we focused on fungi and their relationship to nitrogen, and in Chapter 4b we showed that most of the bacteria isolated in association with these fungi were able to grow in NF medium. However, we didn’t mention any molecular test for the detection of one of the nif genes encoding for nitrogenase enzymes in fungal or bacterial DNA extracts, as a proxy for nitrogen fi xation. Actually, we were looking for the nifH gene in both fungi and bacteria, and we will show our main results in this section.

Microbial strains

Figure A3. Growth of bacteria E6, G2 and H3 after 8 days in nitrogen-free medium (after three successive subcultures in nitrogen-free medium). E6 forms a slimy colony invading all the Petri dish.

142 Figure A4. Gel electrophoresis performed on the products of PCR amplifi cation of a partial fragment of the nifH gene on bacterial cultures. Samples in green were positive for growht in nitrogen-free medium. Th e bacterium Azotobacter chroococcus was used as positive control.

We selected three fungal strains and fi ve bacterial strains, isolated from a Moroccan soil with fungal highway columns (see Chapter 1 of the present thesis for further details about sampling location and isolation processes). Selected fungal strains were Chaetomium globosum (E), Fusarium chlamydosporum (H), and Fusa- rium oxysporum (B). Selected bacterial strains were Olivibacter soli (E6), Pseudomonas frederiksbergensis (G2), Acinetobacter calcoaceticus (H3), Variovorax soli (E2) and Achromobacter mucicolans (A1).

Materials and methods Th e ability of each bacterial strain to grow in nitrogen-free medium was tested (same protocol than in the manuscript of Chapter 4b of the present thesis). We used the same DNA extracts than the ones presented in the Chapter 1 of the present thesis. Th e bacterium Azotobacter chroococcum NEU 1159 was used as positive control. PCR amplifi cation was performed of a fragment of the nifH gene with primers NifHf (5’-AAA GGY GGW ATC GGY AAR TCC ACC AC-3’) and NifHr (5’-TTG TTS GCS GCR TAC ATS GCC ATC AT-3’; Rosch et al., 2002). For PCR, the master mix contained (in 25 μl of fi nal volume): 5 μl buff er (with 1.5 mM MgCl2), 0.2 mM dNTPs mix, 0.2 mM of each primer and 0.5 U DNA Polyme- rase (Kapa Biosystems, Inc., Wilmington, USA). One microliter of DNA template was added. PCR was carried out in an Arktik thermocycler, with an initial denaturation at 95 °C for 4 min, followed by 1 cycle consisting of denaturation at 95 °C for 1 min, annealing at 65 °C for 30 s, and elongation at 72 °C for 30s, then 2 cycles consisting of denaturation at 95 °C for 1 min, annealing at 62 °C for 30 s, and elongation at 72 °C for 30s, then 3 cycles of denaturation at 95 °C for 1 min, annealing at 59 °C for 30 s, and elongation at 72 °C for 30s, then 4 cycles of denaturation at 95 °C for 1 min, annealing at 56 °C for 30 s, and elongation at 72 °C for 30s, followed by 5 cycles of denaturation at 95 °C for 1 min, annealing at 53 °C for 30 s, and elongation at 72 °C for 30s, followed by 25 cycles of denaturation at 95 °C for 1 min, annealing at 50 °C for 30 s, and elongation at 72 °C for 30s. Final extension was performed at 72 °C for 10 min. Presence of amplifi ed DNA fragments was assessed by a 30 min gel electrophoresis at 100 V on a 1.5 % agarose gel. DNA was stained for duration of 30 min in a GelRed bath before observation with a Transilluminator.

Results Bacteria E6, G2 and H3 formed large and slimy colonies in nitrogen-free medium (Figure A3), while

143 bacteria E2 and A1 were not able to grow in this medium. Except for the positive control (Azotobacter chroococcus), none of the bacteria and none of the fungi showed a clear band of 400 bp in the electrophoresis gel. Figure A4 presents the electrophoresis gel for bacteria.

References Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. J Mol Biol 215: 403-410. Rosch C, Mergel A, Bothe H (2002) Biodiversity of denitrifying and dinitrogen-fi xing bacteria in an acid forest soil. Appl Environ Microbiol 68: 3818-3829.

144 Chapter 3 - Element cycling and metabolism

Main fi ndings and perspectives

Main fi ndings

Fungi isolated from a soil containing calcium oxalate were able to solubilize this mineral. Our results redefi ne the role of fungi in the oxalate-carbonate pathway. In addition to the previous statement considering fungi as releasers of oxalic acid, they must also be considered as solubilizers of calcium oxalate.

Bacteria isolated in association with these fungi (fungal highways) were not able to solubilize calcium oxalate, and thus show complementary activity with their associated fungus. However, migration of these bacteria along hyphae of their fungal partner was more effi cient towards a medium containing calcium oxalate as sole carbon source than towards a rich culture medium. We hypothesized that the bacteria would feed on fungal exudates, but we were not able to demonstrate it with calorimetric measurements. Our results show that bacteria migrating along fungal hyphae do not necessarily use the fungal highway in order to reach an otherwise inaccessible carbon source.

Calorimetric measurements showed a fungal activity in a medium containing calcium oxalate as carbon source, and thus point towards an oxalotrophic activity of fungi.

Th e fungi we tested were able to grow successively in 9 generations of nitrogen-free solid media and to produce ethylene in the acetylene reduction test. Th ese fungi were also able to grow in a classical nitrogen-free medium, in a modifi ed atmosphere containing argon and oxygen as sole elements. Our fi ndings propose that the fungi we tested are able to feed on agar, and that ultrapure grades of agar might also be contaminated with nitrogen. Furthermore, the acetylene reduction test might not be suffi cient for the proof of nitrogen fi xation.

As the disproof of nitrogen fi xation in fungi is mainly based on one publication testing only yeasts, that showed no acetylene reduction abilities, we conclude that no statement should be done about nitrogen fi xation in fungi.

Th e nifH gene, coding for the enzyme nitrogenase, was neither detected in the fungi, nor in the bacteria.

All selected Fusarium strains showed the same metabolic activity: they dissolved calcium oxalate, produced siderophores and did not solubilize inorganic phosphorus. While none of the associated bacteria was able to dissolve calcium oxalate, they showed various metabolic activities regarding production of siderophores, solubilization of inorganic phosphorus and growth in nitrogen-free medium.

145 Th ere was no direct correlation between bacterial metabolic activities and migration along fungal hyphae, and we could not see a direct link between C:N ratio or type of carbon source in the culture media and bacterial migration. Th us migration of bacteria along fungal hyphae might be related to complex and multi-parametric factors.

One non-motile bacterial strain was able to migrate along its associated fungus, and we hypothesized that this bacterium might be intrahyphal and migrate inside fungal hyphae the same way mitochondria do.

Perspectives

We would like to know if fungal highway associations are complementary for other metabolic activities, and if the use of fungal highways by bacteria is related to other nutrient sources. Th is is tested in Chapter 4b of the present thesis.

We would like to test whether genes coding for alternative nitrogenases (vnf or anf) are present in the bacteria and in the fungi.

Bacterial migration along fungal hyphae should be observed with mono-parametric changes in the culture media, such as carbon source, or C:N ratio, or various amounts of a specifi c nutrient.

146

General discussion

About the methodology

Th e main goal of the present thesis was to better understand fungal highway interactions, to identify and characterize the organisms involved, and to understand their role in the oxalate-carbonate pathway.

Th e concept of fungal highways was presented a few years before the beginning of this thesis, but this association had been only studied in vitro, with artifi cial fungal (or Oomycetes)-bacterial couples that were not originating from the same environment (Kohlmeier et al., 2005, Wick et al., 2007, War- mink et al., 2009, Furuno et al., 2012). Th e study of model organisms presents many advantages and allows for understanding potential physiological abilities of the associates. Usually, several mutants are also available, which allows for a better understanding of the role of a gene or a physiological trait in the interaction. However, this approach underestimates the impact of the sharing of a habitat on the co-existing organisms. It is for example known that horizontal gene transfer occurs between fungi and from fungi to bacteria (Fitzpatrick, 2012). Th us, one fungal specie isolated from a soil containing a certain fungal and bacterial diversity might present a diff erent phenotype than the same specie isolated from a diff erent environment. Yet, the aim of the present thesis was to isolate organisms co-existing in the same soil and potentially interacting through fungal highways. For this purpose, we developed and validated the fungal highway columns. Fungal highway columns present some disadvantages, such as their size (approx. length of 5 cm) and the fact that fungi have to cross a barrier of glass beads in order to reach the target culture medium. Th us, they could be favoring fast-growing fungi, and bias the fungal population reaching the target. However, a slower-growing fungus was also isolated from a column (Penicillum pinophilum), demonstrating that it was not impossible to recover these types of fungi with the columns. Th e use of fungal highway columns allowed for demonstrating that soil fungi and bacteria are able to interact through fungal highways, and that bacterial dispersion along fungal hyphae must be taken into consideration for the study of colonization of novel habitats.

Redefi ning the oxalate-carbonate pathway

Most of the fungi we isolated from the Moroccan soil were able to dissolve calcium oxalate, while the associated culturable bacteria were not. Our results could have been biased by the culturable approach (i.e. we have no proof that unculturable oxalotrophic bacteria were not present). But there was calcium oxalate in the columns we used, and we isolated the oxalotrophic bacterium Variovorax soli in association with a fungus that did not

151 Figure 1. A redefi ned scheme of the oxalat-carbonate pathway. In addition to the already known roles of each organism in the pathway (a), two novel contributions have been observed (b): the solubilization of calcium oxalate by fungi, and the migration of non-oxalotrophic bacteria along hyphae towards calcium oxalate.

dissolve calcium oxalate, Chaetomium globosum. Furthermore, several culturable oxalotrophic bacteria were isolated in association with a fungus in another study (Bravo et al., 2013), which makes it unlikely that we massively missed oxalotrophic bacteria in our approach. While it is known that the fungus Fusarium oxysporum is able to produce oxalic acid (Swain & Ray, 2009), we did not fi nd any reference about the ability of Chaetomium globosum to release oxalic acid. Th us, in the oxalate-carbonate pathway, fungi might be releasers of oxalic acid, or degraders of oxalic acid, or both. In parallel, their mycelium might be used as routes for the migration of oxalotrophic or non-oxalotrophic bacteria (Figure 1). It is interesting to notice that in our study, as well as in the study of Bravo et al. (2013), the migrating oxalotrophic bacteria were isolated in association with a non-oxalogenic fungus.

As non-oxalotrophic bacteria were preferentially migrating towards a medium containing calcium oxalate as sole carbon source than towards a rich culture medium, it is not excluded that these bacteria might be benefi cial for the oxalate-degrading fungi in the presence of calcium oxalate. Th e fact that fungi are able to easily degrade calcium oxalate raises also the question of the competition that could occur with oxalotrophic bacteria. However, oxalotrophy in bacteria can be facultative and little is known about the relative importance of obligate oxalotrophic bacteria (Hervé et al., 2016). Th us, it is not sure that oxalotrophic bacteria would be impaired by the presence of oxalate-degrading fungi. Th e preferential migration of non-oxalotrophic bacteria towards calcium oxalate questions also the reasons of bacterial migration. While these bacteria clearly do not migrate in order to reach a carbon source in the environment, it would be interesting to understand if oxalotrophic bacteria are also not attracted by the carbon source (and thus reach calcium oxalate in soils “by chance”). Our results showed

152 that bacterial migration was not related to other metabolic properties such as inorganic phosphorus solubilization, nitrogen fi xation or production of siderophores. If we could fi gure out that a specifi c element triggers bacterial movement along fungal hyphae, we could better understand the process, and possibly be able to improve the effi ciency of bacterial activity in soils, whether oxalotrophy in the case of the oxalate-carbonate pathway, or degradation of pollutants in contaminated soils.

Fungal highways as structuring factors of soil bacterial communities

Th is work demonstrates that bacterial dispersal should no longer be considered as very limited in soils, unlike previously stated (de Boer et al., 2005, Or et al., 2007, Reichenbach et al., 2007, Ekschmitt et al., 2008), and thus, studies about soil colonization by bacteria should take into consideration the dispersal along fungal hyphae. Of course, the presence of a second type of organism involves many possible emerging interactions and makes the study more complex. An easy way to minimize this complexity would be at least to add structures such as glass fi bers in the models, as bacteria are also able to migrate along this abiotic network (Pion et al., 2013). However, our results demonstrated that fungi do not act as inert paths, as dispersal along fungal hyphae did not correlate with the ability of the bacteria to grow in a certain culture medium. We could not observe a direct relationship between the carbon source or C:N ratio in the environment and the migration of bacteria. Furthermore, when we compared soil communities and migrating communities in diff erent soils, we observed that bacterial communities linked to fungal highways are more homoge- nous than the soil communities. It appeared also that bacterial phyla such as Acidobacteria and Actino- bacteria were barely involved in fungal highway interactions. Many soil Actinobacteria are fi lamentous (de Boer et al., 2005), which excludes highway-like interactions with fungi. In this respect, nothing is known about these bacteria acting as highways for other bacteria. We didn’t investigate this point, but this interaction seems rather unlikely because of the respective size of the partners (i.e. the potential diffi culty for a motile bacterium to move on such a small surface). Th e bacterial phylum Acidobacteria is known for being one of the most important in soils, but it hosts mostly unculturable bacteria and little is known about their function in soils (Kielak et al., 2009, Nunes da Rocha et al., 2013). However, these bacteria are very diverse metabolically and genetically (Quaiser et al., 2003), and thus it is diffi cult to understand the reason why they are apparently not involved in fungal highway interactions. Our results suggest that bacterial phyla Firmicutes, Nitrospirae, Planctomycetes and Proteobacteria (alpha, bêta and gamma) are more amenable to interact through fungal highways, and thus their diversity and distribution in soils might be at least partially related to fungal networks. Our results with the cultivable couples suggest that fungi might have an important impact on the migration or not of bacteria. For example, non-oxalotrophic bacteria preferentially migrate towards oxalate than towards a carbon source on which they could easily feed, which does not make sense from the bacterial perspective. Th us, fungal physiology might play an important role in the structuring and diversity of soil bacterial communities.

Intrahyphal bacterial migration?

It is interesting to notice that bacteria from the genus Acinetobacter (gammaprotebacteria) were involved in the most curious observations we made with the culturable and the non-culturable approach.

153 In the culturable approach, they were isolated with three diff erent fungal strains, and yet were negative for the fi rst migration tests we performed. One could argue that these bacteria are simply contaminants of the columns, but this is very unlikely, for several reasons. First of all, tests for the contamination of the columns were performed, and almost all the columns did not present any contamination. Th e only contamination we observed was not an Acinetobacter sp. Furthermore, we isolated Acinetobacter spp. from several batches of columns, produced at a diff erent time, in diff erent locations and by diff erent persons. Th us we had to fi nd another explanation to the presence of this bacterial genus in the columns. Further motility tests showed that a non-motile Acinetobacter calcoaceticus strain was able to migrate along its associated fungus in some conditions. Furthermore, the genus Acinetobacter was identifi ed in the columns in the non-culturable tests, but not detected in the soils in contact with the columns. Acinetobacter spp. can live in ubiquitous habitats. While many researches have focused on the fact that they can be opportunistic pathogens in humans, they are actually widespread in soils and aquatic environments (Krizova et al., 2014). In fact, several bacterial genera detected in the columns but not in the soils were aquatic bacteria, such as Rhodobacter (Dang & Lovell, 2002), Aquabacterium (Kalmbach, 1999) and Defl uviimonas (Foesel et al., 2011). In the culture-independent study, many bacterial OTUs assigned to the genus Burkholderia were present in the columns and not in the corresponding soils. As the associated fungus belonged to the genus Mortierella, known for hosting endosymbiotic bacteria, and especially Burkholderiales (Sato et al., 2010), we hypothesized that these bacterial OTUs corresponded to intrahyphal bacteria. As the other bacterial OTUs only present in the columns belonged to bacterial groups known for inhabiting fungal hyphae (Hoff man & Arnold, 2010), we hypothesized that they were also intrahyphal bacteria. Th is hypothesis is supported by the fact that many of these bacteria have been detected in aquatic environments. Indeed, it has been proposed that the paralog polymerase dnaE2 (responsible for the increase of GC content) has played a crucial role in the colonization of land by bacteria, and that aquatic bacteria, as well as symbiotic or parasitic bacteria, would not possess this polymerase. For example, the bacterium Candidatus Glomeribacter gigasporarum, endosymbiot of mycorrhizal fungi, possesses a small genome (~1815 genes), but a high GC content (54.8 %). Th e reduced size of its genome was explained by the loss of dnaE2, because relatives of these bacteria living in soils do possess dnaE2 polymerase (Wu et al., 2014). If this is true, how it is possible that Acinetobacter sp. was not detected by PCR amplifi cation of a fragment of the 16S rRNA gene in pure fungal cultures? Actually, it has been demonstrated that amplifi cation of fragments of the 16S rRNA gene in eukaryotes is not always consistent, because it can lead to the amplifi cation of eukaryotic DNA fragments as well (Huys et al., 2008, Prosdocimi et al., 2013). But if Acinetobacter spp. are intrahyphal bacteria, how does it come that they are sometimes intrahyphal, and sometimes extrahyphal, forming colonies in culture media? Actually, it has been demonstrated that free-living close relatives to the obligate intrahyphal bacterium Candidatus Glomeribacter gigasporarum were able to penetrate fungal spores (van Overbeek & Saikkonen, 2016). Other observations performed in our laboratory showed that intrahyphal bacteria were able to be released from fungal hyphae in certain conditions (e.g. in presence of antifungal antibiotics; unpublished data). Th e transport mechanisms of these intrahyphal bacteria are still unknown. However, it is likely that these bacteria could be transported in both directions (towards the apex or backwards), the same way than mitochondria (Steinberg, 1998).

While we could not fi nd a direct link between nutritive conditions and bacterial transport, whether

154 extra- or intrahyphal, it was recently observed that the intrahyphal bacterium Candidatus Glomeribac- ter gigasporarum produces vitamin B12, an essential compound for fungi, and that the symbiosis would therefore be mutualistic (van Overbeek & Saikkonen, 2016). Th ese results are encouraging for further researches about the factors triggering bacterial migration along fungal hyphae.

Importance of fungal-driven bacterial dispersal in biogeochemical cycling

In both culture-dependent and -independent approaches we performed, mainly fast-growing r-strategists fungi were isolated with the fungal highway columns (Fusarium spp. and Chaetomium sp. from the Moroccan soil, and Mortierella spp. from the Swiss soils). As these columns represent a novel habitat to colonize, the fact that these fungi arrive with their associated migrating bacteria gives both groups a serious advantage in the colonization of novel environments. A recent study observing early colonization of novel environments by bacteria revealed that only a few bacterial genera were able to disperse more than 3 cm away from an inoculum after 48 h (genera Undibacterium, Massilia, Pseudo- monas and Pantoea; Wolf et al., 2015). Th e bacterial diversity found within our columns (i.e. more than 3 cm away from the soil pool) was much higher. While it is supposed that microbial weathering and nutrient cycling plays an important role in early soil development, the exact mechanisms behind these microbe-mineral interactions are poorly known (Ahmed & Holmström, 2015). Our results suggest that r-strategists fungi and their associated migrating bacteria, whether intrahyphal or extrahyphal, might be the fi rst colonizers of novel environments and they might have a great impact on early stages of pedogenesis. It has been demonstrated that microorganisms attached to mineral surfaces are more effi cient for the weathering processes (Ahmed & Holmström, 2015). Th us, if fungi are the fi rst colonizers of novel mineral environments, their associated bacteria will subsequently be in contact with the minerals, and thus the dissolution of minerals might be enhanced. Furthermore, our results showed that the bacterial population able to migrate along fungal hyphae shows various metabolic abilities, suggesting the establishment of a complex nutrient cycling in early stages of pedogenesis.

Conclusion and perspectives

Before the beginning of this thesis, it has been shown that the presence of fungal mycelium allowed for a better degradation of various carbon sources by bacteria able to take the fungal highway. It was therefore hypothesized that fungal highways would be used by bacteria as a mean to reach otherwise inaccessible carbon sources. All the tests were performed with fungal (or Oomycetes) and bacterial strains that were not isolated from the same environment.

In this thesis, we designed the fungal highway columns and isolated for the fi rst time natural fungal-bacterial couples (i.e. organisms actually co-existing in the environment). With a culture-independent approach, we could also have a fi rst insight into the diversity of fungi and bacteria able to interact through fungal highways in soils.

We demonstrated that bacterial migration along fungal hyphae is dependent on the culture medium, and that it is not necessarily linked to the presence of a carbon source for the bacteria inside

155 the media. We could not relate bacterial migration to their ability of using a certain carbon source, of fi xing nitrogen, of producing siderophores, or of solubilizing inorganic phosphorus. Non-swimming and non-swarming bacteria were able to migrate along fungal hyphae.

Our fi ndings suggested that migration of bacteria along fungal hyphae might be either extrahyphal (fungal highways) or intrahyphal (fungal subways), and the diversity of bacteria taking the fungal subway might be underestimated.

Our results demonstrated that members of certain bacterial phyla were preferentially migrating along fungal hyphae in soils. Th ese fungal-bacterial interactions should be considered as crucial for the colonization of novel environments by bacteria, and as an important factor for the structuring of bacterial communities in soil.

We demonstrated that fungi are able to dissolve calcium oxalate in the oxalate-carbonate pathway, as well as to disperse non-oxalotrophic bacteria, thus redefi ning the role of the microbial components in this pathway. We also observed properties of fungi resembling to nitrogen fi xation, but we could neither prove nor disprove this ability.

Our results suggest a novel approach for understanding microbial weathering in early-stage soil development, with fungi and their associated migratory bacteria as pioneer organisms.

In order to better understand fungal-bacterial interactions in soils, it would be crucial to know the factors triggering bacterial movement along fungal hyphae. For example, by multiplying migration tests with well-controlled and various nutrient sources, or by investigating the bacterial production of essential vitamins or other compounds for the fungus. Th is knowledge would be a major advance for the improvement of bioremediation by bacteria, as well as for the effi ciency of the carbon sink resulting from the oxalate-carbonate pathway.

References

Ahmed E, Holmström SJM (2015) Microbe–mineral interactions: Th e impact of surface attachment on mineral weathering and element selectivity by microorganisms. Chem Geol 403: 13–23 Bravo D, Cailleau G, Bindschedler S, Simon A, Job D, Verrecchia E & Junier P (2013) Isolation of oxalotrophic bacteria able to disperse on fungal mycelium. FEMS Microbiol Lett. Dang H & Lovell CR (2002) Numerical dominance and phylotype diversity of marine Rhodobacter species during early colonization of submerged surfaces in coastal marine waters as determined by 16S Ribosomal DNA sequence analysis and fl uorescence in situ hybridization. Appl Environ Mi- crobiol 68: 496-504. de Boer W, Folman LB, Summerbell RC & Boddy L (2005) Living in a fungal world: impact of fungi on soil bacterial niche development. FEMS Microbiol Rev 29: 795-811. Ekschmitt K, Kandeler E, Poll C et al. (2008) Soil-carbon preservation through habitat constraints and biological limitations on decomposer activity. J Plant Nutr Soil Sci 171: 27-35. Fitzpatrick DA (2012) Horizontal gene transfer in fungi. FEMS Microbiol Lett 329: 1-8. Foesel U, Drake HL & Schramm A (2011) Defl uviimonas denitrifi cans gen. nov., sp nov., and Para-

156 rhodobacter aggregans gen. nov., sp nov., non-phototrophic Rhodobacteraceae from the biofi lter of a marine aquaculture. Syst Appl Microbiol 34: 498-502. Furuno S, Remer R, Chatzinotas A, Harms H & Wick LY (2012) Use of mycelia as paths for the isolation of contaminant-degrading bacteria from soil. Microb Biotechnol 5: 142-148. Hervé V, Junier T, Bindschedler S, Verrecchia E & Junier P (2016) Diversity and ecology of oxalotrophic bacteria. World J Microb Biot 32. Hoff man MT & Arnold AE (2010) Diverse bacteria inhabit living hyphae of phylogenetically diverse fungal endophytes. Appl Environ Microbiol 76: 4063-4075. Huys G, Vanhoutte T, Joossens M, et al. (2008) Coamplifi cation of eukaryotic DNA with 16S rRNA gene-based PCR primers: possible consequences for population fi ngerprinting of complex microbial communities. Curr Microbiol 56: 553-557. Kalmbach (1999) Aquabacterium gen. nov., with description of Aquabacterium citratiphilum sp. nov., Aquabacterium parvum sp. nov. and Aquabacterium commune sp. nov., three in situ dominant bacterial species from the Berlin drinking water system. Int J Syst Bacteriol 49: 769-777. Kielak A, Pijl AS, van Veen JA & Kowalchuk GA (2009) Phylogenetic diversity of Acidobacteria in a former agricultural soil. ISME J 3: 378-382. Kohlmeier S, Smits TH, Ford R, Keel C, Harms H & Wick LY (2005) Taking the fungal highway: Mobilization of pollutant-degrading bacteria by fungi. Environ Sci Technol 39: 4640-4646. Krizova L, Maixnerova M, Sedo O & Nemec A (2014) Acinetobacter bohemicus sp. nov. widespread in natural soil and water ecosystems in the Czech Republic. Syst Appl Microbiol 37: 467-473. Nunes da Rocha U, Plugge CM, George I, van Elsas JD & van Overbeek LS (2013) Th e rhizosphere selects for particular groups of acidobacteria and verrucomicrobia. PLoS One 8: e82443. Or D, Smets BF, Wraith JM, Dechesne A & Friedman SP (2007) Physical constraints aff ecting bacterial habitats and activity in unsaturated porous media – a review. Ad Water Resour 30: 1505-1527. Pion M, Bshary R, Bindschedler S, et al. (2013) Gains of bacterial fl agellar motility in a fungal world. Appl Environ Microbiol 79: 6862-6867. Prosdocimi EM, Novati S, Bruno R et al. (2013) Errors in ribosomal sequence datasets generated using PCR-coupled ‘panbacterial’ pyrosequencing, and the establishment of an improved approach. Mol Cell Probes 27: 65-67. Quaiser A, Ochsenreiter T, Lanz C, Schuster SC, Treusch AH, Eck J & Schleper C (2003) Acidobac- teria form a coherent but highly diverse group within the bacterial domain: evidence from environmental genomics. Mol Microbiol 50: 563-575. Reichenbach T, Mobilia M & Frey E (2007) Mobility promotes and jeopardizes biodiversity in rock-pa- per-scissors games. Nature 448: 1046-1049. Sato Y, Narisawa K, Tsuruta K et al. (2010) Detection of Betaproteobacteria inside the mycelium of the fungus Mortierella elongata. Microb Environ 25: 321-324. Steinberg G (1998) Organelle transport and molecular motors in fungi. Fungal Genet Biol 24: 161- 177. Swain MR & Ray RC (2009) Oxalic acid production by Fusarium oxysporum Schlecht and Botryodi- plodia theobromae Pat., post-harvest fungal pathogens of yams (Dioscorea rotundata L.) and detoxifi - cation by Bacillus subtilis CM1 isolated from culturable cowdung microfl ora. Arch Phytopathology Plant Protect 42: 666-675. van Overbeek LS & Saikkonen K (2016) Impact of bacterial-fungal interactions on the colonization of the endosphere. Trends Plant Sci 21: 230-242. Warmink JA, Nazir R & van Elsas JD (2009) Universal and species-specifi c bacterial ‘fungiphiles’ in the

157 mycospheres of diff erent basidiomycetous fungi. Environ Microbiol 11: 300-312. Wick LY, Remer R, Würz B, Reichenbach J, Braun S, Schäfer F & Harms H (2007) Eff ect of fungal hyphae on the access of bacteria to phenanthrene in soil. Environ Sci Technol 41: 500-505. Wolf AB, Rudnick MB, de Boer W & Kowalchuk GA (2015) Early colonizers of unoccupied habitats represent a minority of the soil bacterial community. FEMS Microbiol Ecol 91. Wu H, Fang Y, Yu J & Zhang Z (2014) Th e quest for a unifi ed view of bacterial land colonization. ISME J 8: 1358-1369.

158

Collaborations, presentations and Curriculum vitae

Collaborations

In addition to the chapters presented in this thesis, I collaborated to following publications during the time of my thesis. Th is list does only include publications directly related to the topic of this thesis. Other collaborations are presented in the Curriculum vitae included hereafter.

Hervé V, Simon A, Verrecchia E, Junier P. Functional diversity of the fungi associated with Tama- rindus indica litter. In prep.

Pion M, Spangenberg JE, Simon A, Bindschedler S, Flury C, Chatelain A, Bshary R, Job D, Junier P (2013) Bacterial farming by the fungus Morchella crassipes. Proceedings of the Royal Society B. 208.

Bravo D, Cailleau G, Bindschedler S, Simon A, Job D, Verrecchia E, Junier P (2013) Isolation of oxalotrophic bacteria able to disperse on fungal mycelium. FEMS Microbiology Letters.

163

Presentations

During this thesis, I had the chance to present my work in several sientifi c meetings. Th is list does only include presentation I made as fi rst author.

Oral presentations

FEMS 2015 - 6th Congress of European Microbiologists, June 7-11, Maastricht (Th e Netherlands). Th e ‘fungal highway’ toll: metabolisms and genes involved in fungal-driven bacterial dispersal in natural ecosystems. ISME 2014 - 15th International Symposium on Microbial Ecology, August 24-29, Seoul (South Korea). Do ‘fungal highways’ exist in nature? Design of a new tool to assess the presence of fungal-driven bacterial dispersal in natural ecosystems. SSM 2014 - 72nd Annual Assembly of the Swiss Society for Microbiology, June 19-20, Fribourg (Swit- zerland). Do ‘fungal highways’ exist in nature? Design of a new tool to assess the presence of fungal-driven bacterial dispersal in natural ecosystems.

SME 2013 - 5th Swiss Microbial Ecology Meeting, February 4-6, Murten (Switzerland). Abundance, diversity and activity of fungal highways in natural ecosystems – a new approach.

Posters

SSM 2015 - 73rd Annual Assembly of the Swiss Society for Microbiology, May 28-29, Lugano (Swit- zerland). Taking the ‘fungal highway’: metabolisms and genes involved in fungal-driven bacterial dispersal in natural ecosystems. Annual PhD Students Meeting 2014 - Interuniversity Doctoral Program in Organismal Biology, May 8, Neuchâtel (Switzerland). Do ‘fungal highways’ exist in nature? Design of a new tool to assess the presence of fungal-driven bacterial dispersal in natural ecosystems. Best poster award. Conférences Jacques-Monod 2013 - Bacterial-fungal interactions: a federative fi eld for fundamental and applied microbiology, December 7-11, Roscoff (France). Applying the ‘fungal highways’ concept as an approach to identify fungal‐bacterial associations in natural ecosystems. FEMS 2013 - 5th Congress of European Microbiologists, July 21-25, Leipzig (Germany). Abundance, diversity and activity of fungal highways in natural ecosystems – a new approach. Best poster award.

165

Anaële Simon Phone +41 (0)797922069 E-mail [email protected] Ph.D. in Sciences Date of birth 05.10.1982 MicrobiologistM Nationalities Swiss and French

Highly motivated microbiologist, specialist for mycology and environmental bacteriology. Competences for culture-dependent and culture-independent analyses for fungal and bacterial samples. Experience in industrial microbiology, familiar with Pharmacopoias, ISO standards, GMP and GLP. Experience in scientifi c research, many collaborations with diff erent Swiss and European laboratories. Strong skills in transfer of knowledge and popularization of science, associated with enjoyment of producing graphic supports and logos. Very independent and organized, enjoys meeting new people and maintaining nice working atmosphere. Always enthu- siastic about acquiring new skills and taking up challenges.

Education Professional experience 2012-16 Ph.D. in Sciences (SNSF). 2012-16 Ph.D. in Sciences. Highways and subways - a story of fungi and bac- Teaching (courses and practicals) of mycology, teria in soils. bacteriology, environmental and food microbio- University of Neuchâtel and University of Lausanne, logy. Switzerland. Management of lab technician trainees and Supervisors: Prof. Pilar Junier and Prof. Eric Ver- bachelor students. recchia. In charge of informatics in the laboratory, and responsible for the website. 2005-07 M.Sc. in Biogeosciences. Magna cum laude. In charge of the SEM (Scanning Electron Micro- Study and optimization of fungal copper oxalate copy) for samples preparation and assistant for formation on verdigris for the protection of mo- observation, including EDS (Energy Dispersive numents. X-ray Spectroscopy). University of Neuchâtel, Switzerland. University of Neuchâtel, Switzerland.

2001-05 B.Sc. in Biology. 2011-12 Scientifi c collaborator. University of Neuchâtel, Switzerland. BAHAMAS project (Biological patina for archaeological and artistic metal artefacts). University of Neuchâtel, Switzerland. Informatics 2010-12 Teacher (high school). ++ ++ ++ + + Languages R , Matlab , HTML , CSS , Python Biology, chemistry and physics. Gymnase Cantonal de Chamblandes, Pully, Switzer- +++ +++ +++ Graphics Blender , Adobe Illustrator , Photoshop land.

+++ +++ Offi ce Adobe InDesign , Microsoft Offi ce Word , 2008-10 Application specialist for +++ +++ Excel , PowerPoint Industrial microbiology. Scientifi c support, marketing, teaching. ++ ++ + + Other AliView , BioEdit , Linux , Qiime , ARB (Sil- BioMérieux (Suisse) SA, Genève, Switzerland. va)+, Jahia+++ +++Excellent ++Intermediate +Basic 2007-08 Teacher (high school). Biology and chemistry. Lycée Jean-Piaget, Neuchâtel, Switzerland. Languages 2006-07 Scientifi c collaborator. French++++, Swiss German++++, English+++, German+++, Ita- Expertises in mycology. lian++, Arabic+ University of Neuchâtel, Switzerland. ++++Native +++Fluent ++Basic +Few notions (in progress)

167 Awards and achievements Publications 2015 Editor’s choice article. For: Simon A, Bindschedler Simon A, Bindschedler S, Job D, Wick L, Filippidou S, S, Job D, et al. 2015. Exploiting the fungal highway Kooli W, Verrecchia E, Junier P. 2015. Exploiting the fungal [...]. FEMS Microbiology Ecology 91. highway: Development of a novel tool for the in situ isolation of bacteria migrating along fungal mycelium. FEMS 2014 Congress travel grant. For ISME15, Seoul, South Microbiol Ecol 91. Korea. Provided by the Interuniversity Doctoral Pro- gram in Organismal Biology, Neuchâtel (CH). Pion M, Spangenberg JE, Simon A, Bindschedler S, Flury C, Chatelain A, Bshary R, Job D, Junier P. 2013. Bacterial 2014 Best Poster Award. Interuniversity Progam for Or- farming by the fungus Morchella crassipes. Proc R Soc B 208. ganismal Biology - Annual PhD students meeting, Neuchâtel (CH). Bravo D, Cailleau G, Bindschedler S, Simon A, Job D, Ver- recchia E, Juner P. 2013. Isolation of oxalotrophic bacteria 2013 Best Poster Award. FEMS - 5th Congress of Euro- able to disperse on fungal mycelium. FEMS Microbiol Lett. pean Microbiologists, Leipzig (D). Joseph E, Letardi P, Comensoli L, Simon A, Junier P, Job D, 2012 Grant for fi eld research in Morocco. Provided by the Woerle M. Assessment of a biological approach for the pro- Commission Egalité des chances, Neuchâtel (CH). tection of copper alloys artefacts. In Conference Proceedings of Metal 2013, Interim Meeting of the ICOM-CC Metal WG. Edinburgh, Scotland. Conferences (fi rst author) Joseph E, Simon A, Mazzeo R, Job D, Wörle M. 2012. th 2015 Oral presentation. FEMS - 6 Congress of European Spectroscopic characterization of an innovative biological Microbiologists, Maastricht (NL). treatment for corroded metal artefacts. J Raman Spectrosc 43: 1612-1616. Poster. SSM - 73rd Annual Assembly of the Swis So- ciety for Microbiology, Lugano (CH). Joseph E, Cario S, Simon A, Wörle M, Mazzeo R, Junier P, Job D. 2012. Protection of metal artefacts with the for- th 2014 Oral presentation. ISME - 15 International Sym- mation of metal-oxalates complexes by Beauveria bassiana. posium on Microbial Ecology, Seoul (KR). Front Microbiol 2: 1-8.

Oral presentation. SSM - 72nd Annual Assembly of Joseph E, Simon A, Prati S, Wörle M, Job D, Mazzeo R. the Swis Society for Microbiology, Fribourg (CH). 2011. Development of an analytical procedure for evaluation of the protective behaviour of innovative fungal patinas Poster. Interuniversity Progam for Organismal Biolo- on archaeological and artistic metal artefacts. Anal Bioanal gy - Annual PhD students meeting, Neuchâtel (CH). Chem 399 (9): 2899-2907.

2013 Poster. Conférences Jacques Monod - Fungal bacte- Mazzeo R, Bittner S, Job D, Farron G, Fontinha R, Joseph rial interactions, Roscoff (F). E, Letardi P, Mach M, Prati S, Salta M, Simon A. Deve- lopment and Evaluation of New Treatments for Outdoor th Poster. FEMS - 5 Congress of European Microbio- Bronze Monuments. In Conservation Science 2007. J. logists, Leipzig (D). Townsend, L. Toniolo, F. Cappitelli (Eds.), Archetype publications: London, 2008. Oral presentation. SME - 5th Swiss Microbial Eco- logy Meeting, Murten (CH). Submitted

Simon A, Hervé V, Al-Dourobi A, Verrecchia E, Junier P. Other interests Highways and subways: identifi cation of bacteria dispersing along fungal hyphae in soils. Submitted. Horseriding, natural horsemanship, work on personal lea- dership and self-confi dence with horses. Blandenier Q, Seppey C, Singer D, Vlimant M, Simon A, Graphics, production of several logos and illustrations for Duckert C, Lara E. Picostelium satyrum nov. gen., nov. sp., the various partners. First Cultured Member of the Environmental Dermamoe- NonViolent communication. bidae Clade LKM74 and its Unusual Life Cycle. Submitted. Reading, painting, hiking.

168