Plant–fungal symbiosis en gros and en détail


Mycorrhiza: systems research from genes to communities, Monte Verità, Ascona, Switzerland, March 2006

Mutualistic relationships between plants and fungi are an important component of ecological systems worldwide. Most higher land plants rely on a fungal partner for efficient supply of macronutrients, mainly phosphorus. In the three major types of symbiosis – the arbuscular mycorrhizas (AM), the ectomycorrhizas (ECM) and the ericoid mycorrhizas – fungal hyphal structures extend into the surrounding environment and thereby enlarge the nutrient accessible area for plant roots. In AM and ECM, these hyphae can form extensive interconnected networks, and even link roots from different plants. In exchange, the plant host provides the fungus with carbon and it has been estimated that up to 20% of all photosynthates might be delivered to the fungal partner. Arbuscular mycorrhizal fungi, which belong to the phylum Glomeromycota and form the oldest of these mycorrhizas, dating back to the appearance of the first land plants, are completely dependent on the host plant for growth.

Despite their ecological and economical importance, we are still only beginning to understand the processes of mutual recognition, development and establishment of symbiotic structures and nutrient exchange. Also, little is known about the composition and ecology of fungal communities and their impact on the flora. This is partly caused by technical difficulties in identification and handling of mycorrhizal fungi, especially the Glomeromycota AM fungi.

A recent international meeting ‘Mycorrhiza: systems research from genes to communities’ ( assembled some of the world leaders in mycorrhiza research. Well organized by Thomas Boller (University of Basel, Switzerland) and Marcel Bucher (ETH Zürich, Switzerland), it was held at the scenic Centro Stefano Franscini in Monte Verità, overlooking Ascona and Lago Maggiore. The main focus was on AM, with reports also on ECM and pathogenic systems.

‘. . . strigolactones induce an increase in mitochondrial density and respiration, the earliest known hallmarks of fungal response to plant hosts.

Diversity in arbuscular mycorrhiza

Under laboratory conditions, AM fungi appear to be fairly nonspecific in their choice of plant host. Many different fungus–plant combinations form symbiotic structures. This is somewhat in disagreement with field studies in which coexisting plants, even from closely related species, were found to harbour distinct fungal communities. It has also been reported that composition of the AM community has an impact on the plant community, and that AM fungal taxa differ with respect to their carbon-uptake abilities, phosphate-transfer efficiency, and ability to confer protection from pathogenic attack on the host plant. Alastair Fitter (University of York, UK) provided two possible solutions to this apparent paradox: either all AM fungi are nonspecific but adapt to different soil niches; or there are specific AM fungi that adapt to the root niche and are therefore very difficult to isolate and subculture. These specialized species could remain undetected by the classical screening system of trap culture with model host plants.

What, then, is the best way to detect mycorrhizal fungi? For morphological distinction, only the very limited characteristics of the spores can be used, but trap culture might have a bias towards more nonspecific and more vigorous AM fungal species. Furthermore, detection of a certain species in an ecosystem does not necessarily correlate with that species being engaged in productive symbiosis, as Damond Kyllo (Smithsonian Tropical Research Institute, Panama) pointed out. He showed examples where species distribution from spore sampling was different from that found actually associated with plant roots; certainly a more sensitive and reproducible detection system would be helpful for future studies. Real-time PCR, as presented by Hannes Gamper (University of York, UK), might be a promising candidate as it would potentially also allow quantification of AM fungi.

These are not the only difficulties in studying mycorrhizal ecology. Dirk Redecker (University of Basel, Switzerland) estimated by sequence comparison that the roughly 200 species currently described, which are mostly distinguished by spore morphology, might actually constitute more than 2000 different species. In the future, better and more numerous molecular markers will be needed for species characterization. Sequencing of different Glomus intraradices isolates showed that the intraspore variation of nuclear rDNA is relatively high compared with that of mitochondrial rDNA (Raab et al., 2005). This intra-individual sequence heterogeneity may be a general feature of Glomeromycota, and may make it difficult to use nuclear markers for characterization. Mitochondrial markers such as mtLSU and its introns might be better choices. From these data, Redecker also concluded that G. intraradices is not one species, but a very heterogeneous group or several different groups. This was also pointed out by Søren Rosendahl (University of Copenhagen, Denmark), and could explain the finding that G. intraradices is one of the few globally distributed ‘species’, as reported by Maarja Öpik (University of Tartu, Estonia).

Jan Jansa (ETH Zürich, Switzerland) argued that more diverse AMF communities might be more beneficial than single species because of niche and functional partitioning. Fungal communities are definitely not random, but differ between sites and plant species present, and over time, as shown by Rosendahl and Peter Young (University of York, UK). However, John Klironomos (University of Guelph, Canada), at least in his experiments, could not find a correlation between fungal species richness and plant species richness. He concluded that plant pathogens are probably the more important driving force shaping plant communities.

Genetic control of mycorrhization

In recent years, one of the most important advances in AM research has been the identification of common symbiotic (SYM) genes in legumes, mainly Lotus japonicus and Medicago truncatula, that are needed for AM, as well as the nitrogen-fixing root nodule symbiosis. Five of these SYM genes have been identified so far, and Katsuharu Saito (University of Tokyo, Japan) and Martin Parniske (University of Munich, Germany) presented a sixth and a seventh member: the nucleoporin gene NUP85; and CYCLOPS, encoding a nuclear protein. This completed the molecular identification of the seven common SYM genes in L. japonicus so far identified by forward genetics (Kistner et al., 2005), so the current focus is on AM-specific genes. After penetration of wild-type plant roots, fungal hyphae invade and colonize the root cortex to form eventually highly branched, tree-like fungal structures (arbuscules) inside plant cells. The large interaction surface is thought to be the area of nutrient exchange. Sonja Kosuta (The Sainsbury Laboratory, Norwich, UK) presented three new EMS-induced mutants in L. japonicus: red, dis and small. These are normally nodulated by Rhizobium bacteria: arbuscule formation is disturbed in symbiosis with either G. intraradices or Gigaspora rosea. From a genetic screen of transposon-mutagenized petunias, Didier Reinhardt (University of Freiburg, Germany) presented three mutants: con1, being affected during early interaction; and blo1 and arb1, both of which exhibit unusual arbuscule development. These mutants are promising candidates to further our understanding of plant–fungal communication and the subcellular development of the symbiosis.

Vivienne Gianinazzi-Pearson (INRA-CSME, Dijon, France) presented transcriptome profiles of M. trunculata. By comparing expression profiles of mycorrhized plants with those of dmi3 mutants that lack a calcium calmodulin protein kinase essential for successful fungal penetration, several plant and fungal genes were found to be upregulated during symbiosis. These genes may play a role during symbiotic development. The next step will be phenotypic testing of plants in which the respective genes have been knocked down by RNAi. As Arabidoposis thaliana does not symbiose, it cannot be used as a mycorrhiza model system; however, Ari Jumpponen (Kansas State University, Manhattan, KS, USA) called this into question by presenting data on the interaction of A. thaliana with the endophyte Periconia macrospinosa. Although this is no mycorrhiza sensu strictu, the plant host appears to benefit from the endophyte's presence in terms of growth stimulation and decreased susceptibility to the foliar pathogen Botrytis cinerea. Arabidoposis thaliana may therefore have a potential function as a model system for screening genes responsible for both phenotypic effects.

One of the most helpful and long-awaited developments in AM research is the still ongoing sequencing project on G. intraradices. However, as Francis Martin (INRA-UMR, Champenoux, France) remarked in his overview on the now-finished sequencing project for the ectomycorrhiza Laccaria bicolor, this will be only the starting point, and a community effort will be required for correct functional annotation of the genomes. Because of the divergence of the Glomeromycota from other fungi, many novel and specific genes are expected in the G. intraradices genome that can be reliably annotated only with the aid of a sufficiently scaled cDNA sequencing project.

Early signalling in mycorrhiza

The biggest breakthrough in AM research in the past year was the identification of strigolactones as the branching factors (BF) that exude from plant roots and lead to early responses in the approaching hyphae (Akiyama et al., 2005). Interestingly, the same class of molecules act as germination stimulants for seeds of Orobrancheaceae parasitic weeds such as Striga (witchweed), which attack plant roots and deprive them of water and nutrients. Strigolactones are thought to belong to the sesquiterpene lactones, and since they are highly unstable, their purification was a major task, as Kohki Akiyama (Osaka Prefecture University, Osaka, Japan) explained. In a first attempt to obtain the structure, strigolactones were purified from 9920 l of root exudate from a hydroponic L. japonicus culture. Subsequently, a strategy was developed in which BF was enriched by circulation of the exudate through an activated charcoal trap. Application of purified strigolactone, as well as synthetic derivatives, led to hyphal branching in Gigaspora margarita; however, this bioassay does not work with G. intraradices because of spontaneous branching. Guillaume Bécard (University of Toulouse, France) also reported fungal species-dependent responses to strigolactones. But, more importantly, he confirmed their role as signalling molecules in demonstrating that strigolactones induce an increase in mitochondrial density and respiration, the earliest known hallmarks of fungal response to plant hosts (Besserer et al., 2006). Such a general recognition mechanism correlates well with the observed promiscuity of AM interactions.

Thus far, the strigolactone biosynthesis pathway is unclear. But Harro Bouwmeester (Plant Research International, Wageningen, the Netherlands) presented a postulated synthesis scheme based on observations that point to carotenoid as the strigolactone precursor: root exudates from plants treated with the carotenoid biosynthesis inhibitor fluridone, or from maize mutants with disturbed carotenoid biosynthesis, had a significantly reduced germination stimulation effect on Striga seeds (Matusova et al., 2005). The next steps will be elucidation of the biosynthetic steps and generation of respective plant mutants to verify the role of strigolactone in AM during early recognition events, and to investigate possible roles during the later stages of symbiosis.

As production of the postulated Myc factor(s) might be induced in the presence of strigolactones (Kosuta et al., 2003), discovery of BF might facilitate their characterization, as well. Akiyama, Bécard and David Drissner (ETH Zürich, Switzerland) all presented attempts to isolate the Myc factor. They followed slightly different approaches, but with a similar idea: to look for plant reporter-gene activation by fungal exudate fractions. Akiyama could characterize the Myc factor as nonpolar, lipophilic and Nod-factor-like. He found several different active exudate fractions, which could mean there is not one Myc factor, but rather several active compounds. This is in agreement with the fact that the promoter–GUS constructs used by Drissner and Akiyama are induced at different time points during symbiosis.

The arbuscular stage of AM symbiosis

Most, if not all plant species synthesize apocarotenoids at late stages of symbiosis (Fester et al., 2002), but their function is unknown. As Michael Walter (Leibniz Institute of Plant Biochemistry, Halle, Germany) explained, downregulation of apocarotenoid synthesis by RNAi in M. truncatula led to significantly reduced levels of fungal colonization, whereby all symbiotic stages are affected. This might be caused by lowered strigolactone levels, although the use of an inhibitor in the same pathway in a similar experiment did not show an effect on strigolactone-related functions (Matusova et al., 2005). For apocarotenoid, Walter hypothesized a role as a signal for degradation of arbuscular structures at later time points.

In the light of these new data, Thomas Kuyper (Wageningen University, the Netherlands) remarked that, some time ago, a negative effect was found between Striga infection and AM colonization (Lendzemo & Kuyper, 2001). Similarly, Horst Vierheilig (Institute for Plant Protection, BOKU, Vienna, Austria) showed that preinfection of plant roots with AM suppressed subsequent infection by the same fungal species. This phenomenon was also observed between different AM fungal species. A similar, but weaker, negative effect was found between AM and root nodule formation, and even for the formation of AM on Nod-factor treated roots. To explain these effects, Vierheilig postulated an autoregulatory mechanism in the plant host that possibly leads to changes in root exudates (Scheffknecht et al., 2006).

A paradigm shift in our understanding of plant–fungus interactions resulted from the work of Andrea Genre (University of Torino, Italy). It now seems that the plant host dictates the development of symbiosis. Before fungal penetration, plant cells form an intracellular, tunnel-like prepenetration apparatus (PPA; Genre et al., 2005). Within 2 h of appressoria formation, the nucleus is repositioned underneath the fungus, followed by formation of the PPA, which is finally disassembled after successful guidance of the fungal hypha through the plant cell. Interestingly, M. trunculata dmi2 or dmi3 mutants show repositioning of the nucleus in response to approaching hyphae, but no PPA formation.

At the stage of arbuscule formation, morphological changes within the plant cell can also be observed. Thomas Fester (Leibniz Institute of Plant Biochemistry, Halle, Germany) showed that plastids undergo proliferation during arbuscule development. In the beginning, proliferation of lens-shaped plastids and enhanced production of fatty acids and amino acids were observed, but after 7–10 d (the arbuscule life span in his system) more tubular plastids and accumulation of apocarotenoid were seen (Lohse et al., 2005). The latter structures may have a function in the recycling of fungal material.

Physiology of symbiosis

The central part of AM symbiosis is thought to be nutrient exchange at the arbuscular interface – the peri-arbuscular space. The main nutrient flows are presumed to be phosphate from fungus to plant, and hexose in the opposite direction. At the meeting, candidates for a plant phosphate transporter as well as for a fungal hexose transporter were presented. Maria Harrison (Boyce Thompson Institute for Plant Research, Ithaca, NY, USA) showed that RNAi downregulation of the phosphate transporter MtPt4 in M. truncatula led to reduced colonization and aberrant arbuscule development. This phenotype was confirmed in an EMS mutant identified by TILLING (Targeted Induced Local Lesions IN Genomes; McCallum et al., 2000). The phenotype was assessed by a novel method to analyse quantitatively the arbuscule development in wild-type and mutant plants. This was achieved by the use of Glomus versiforme and synchronized infection through pregermination with root exudates. Such studies will be much easier in the future with the employment of synthetic strigolactones.

Using a very different model system, the AM-like endosymbiosis between Geosiphon pyriformis and the cyanobacterium Nostoc, Arthur Schüßler and coworkers (University of Technology, Darmstadt, Germany) were able to isolate the first monosaccharide transporter gene of a glomeromycotan fungus. The transporter was functionally characterized in yeast and is postulated to transport hexoses across the perisymbiotic membrane (a derivative of the plasma membrane) in Geosiphon. Schüßler believes that orthologous proteins from other Glomeromycota are located within the arbuscular plasma membrane. If correct, this type of transporter would be responsible for a large portion of the carbohydrate transport towards the fungus in the AM.

Genome structure of AM fungi

A very interesting and controversial question is whether the multiple nuclei of Glomeromycota fungi are hetero- or homokaryotic. As all stages are multinucleate, and fungi form extensive hyphal networks, Ian Sanders (University of Lausanne, Switzerland) reasoned strongly in favour of heterokaryosis. However, heterokaryosis should lead to allelic segregation and allelic dropout, in contrast to the observed stability of the Glomeromycota genome. This problem could be solved if hyphal networks formed extended networks by anastomosis formation, which might provide the structural basis for the exchange of genetic material. Anastomosis formation was suggested by Manuela Giovannetti (University of Pisa, Italy) to occur very frequently within networks of G. intraradices, but not within Gigaspora or Scutellospora species. She showed the functionality of these structures through cytoplasmic streaming and nuclear migration (Giovannetti et al., 1999). Anastomoses also form between hyphal networks emerging from different plants, but never between different fungal species. Interestingly, no directed recognition process could be observed. Recognition appears to occur only on contact, and can lead to incompatibility reactions.

Sanders supported the concept of nuclear exchange by presenting experimental data on AFLP analyses of daughter spores that resulted from cocultivation of hyphal networks originating from two spores with different AFLP patterns. The daughter spores showed signals from both parental spores, therefore Glomeromycota genomes could have a rather fluid structure. As it allows genetic exchange between different isolates, this new experimental technique might provide the opportunity to perform AM fungal genetics in the future.


This meeting foreshadowed some of the discoveries we will hopefully witness in the near future: characterization of more plant genes involved in symbiosis by direct as well as reverse genetic screens; elucidation of the strigolactone biosynthesis pathway and generation of plant mutants, necessary for assessing the relevance of strigolactones for symbiosis, and potentially useful in ecological competition studies of cost–benefit studies of mycorrhization in natural and agronomic environments; and identification of Myc factor(s) and synthesis of analogues for synchronized plant-response activation. The long-awaited completion of the G. intraradices genome sequence will be extremely helpful for AM research, and might also clarify the heterokaryon/homokaryon debate.

During this meeting it became clear that, especially in AM, there are a number of significant differences between different fungal species and host plants, which will certainly increase further. We still have only very limited basic knowledge about the fungal partner in AM at the cytological, molecular and biological levels. Overcoming the technical difficulties in cultivation and application to molecular studies will be one of the important tasks for the future.