For a long time, the mycorrhizal symbiosis has been considered as a bipartite relationship between plant roots and mycorrhizal fungi. However, in natural conditions, mycorrhizas are surrounded by complex bacterial and fungal communities, which interact with the mycorrhiza–plant symbiosis at physical, metabolic and functional levels. That is why it is more relevant today to qualify mycorrhizal roots and associated microbial communities as a multitrophic mycorrhizal complex (Fig. 1; Frey-Klett et al., 2005). Although it is quite clear that the mycorrhizal complexes play a major role in gross production and nutrient cycling, the structure and the functioning of these complexes, and more particularly the importance of the interactions between bacteria and the mycorrhizal symbiosis, have been so far very poorly documented. Seminal investigations of Bowen & Theodorou (1979) and then Garbaye & Bowen (1989) demonstrated that the rhizosphere microflora could have a positive or negative impact on the mycorrhizal symbiosis, depending on the bacterial isolates. Since that time, several studies have been conducted, on either endomycorrhizal or ectomycorrhizal symbiosis, to identify bacterial isolates promoting the mycorrhizal symbiosis, so-called ‘mycorrhiza helper bacteria’ (Garbaye, 1994). These helper bacteria belong to many bacterial groups and genera, such as Proteobacteria (Pseudomonas: Duponnois & Garbaye, 1991 and Founoune et al., 2002; Burkholderia: Poole et al., 2001; Bradyrhizobium: Xie et al., 1995), Firmicutes (Bacillus: von Alten et al., 1993 and Dunstan et al., 1998; Paenibacillus: Budi et al., 1999 and Poole et al., 2001) and Actinomycetes (Rhodococcus: Poole et al., 2001; Streptomyces: Schrey et al., this issue, pp. 205–216).
‘The challenge now is to monitor the kinetics of the expression of Amanita genes in the presence of different fungal-associated bacterial isolates, including mycorrhiza helper bacteria, not only in vitro but also in more natural conditions – in other words, in the presence of plant roots.’
Fungus specificity and practical relevance of mycorrhiza helper bacteria
Very few papers published to date have addressed the question of the fungus-specificity of mycorrhiza helper bacteria – the paper of Schrey et al. in this issue is one of them. The specificity of the helper effect appears to vary with the bacterial strains. Indeed, whereas Garbaye & Duponnois (1992) and Dunstan et al. (1998) demonstrated a fungus specificity in the interactions between the ectomycorrhizal fungus Laccaria and different mycorrhiza helper bacterial isolates, Duponnois & Plenchette (2003) described that the same mycorrhiza helper Pseudomonas monteilli isolate was able to promote the symbiosis between Acacia holosericea and the ectomycorrhizal fungi Pisolithus or Scleroderma, as well as the endomycorrhizal fungus Glomus intraradices. Schrey et al. focused on the effect of two Streptomyces strains, isolated from the mycorrhizosphere of spruces extensively mycorrhizal with Amanita muscaria, on the in vitro growth of different ectomycorrhizal fungi. These Streptomyces strains promoted the growth of Amanita muscaria and Suillus bovinus, whereas they inhibited Hebeloma cylindrosporum growth and had no significant effect on Paxillus involutus. Schrey et al. rightly underlined the practical potential application of one of these two Streptomyces isolates, which had previously also showed an inhibitory effect towards two phytopathogenic fungi, Heterobasidion annosum and Armillaria obscura (Maier et al., 2004). Such a mycorrhiza helper bacterial strain could be used in forest nurseries to promote controlled mycorrhization of tree seedlings and prevent phytopathogen attacks. Of course, the results of Maier et al. (2004) and Schrey et al. should be confirmed by field experiments because the significance of in vitro tests is often controversial (Whipps, 1987). Nevertheless, the work of Schrey et al. is in keeping with the small number of previous studies that have addressed the practical use of mycorrhiza helper bacteria in agriculture and forestry (Duponnois et al., 1993; Barea et al., 1998; Becker et al., 1999; Budi et al., 1999). By their positive and even sometimes selective effect on mycorrhizal fungi and their negative effect on pathogenic ones, mycorrhiza helper bacterial inoculants represent a promising environmentally safe process to be used in sustainable agriculture and forestry.
Mechanisms underlying the mycorrhiza helper effect: a focus on fungal gene regulation by helper bacteria
Garbaye (1994) reviewed the possible mechanisms underlying the mycorrhiza helper effect (Fig. 2). A direct effect of the helper bacteria on the root receptivity to mycorrhizal fungi has been frequently evoked in the different papers that deal with the mechanisms of the mycorrhiza helper effect (e.g. Schrey et al.). However, the main mechanism favoured so far in all these studies is the direct effect of helper bacteria on the presymbiotic survival and growth of the mycorrhizal fungi in the soil (Bruléet al., 2001; Founoune et al., 2002, Schrey et al.). At the molecular level, this mechanism likely relies on the modification of the fungal nutrient use efficiency and/or on the regulation of the fungal cell cycle (i.e. hyphal proliferation) by the helper bacteria. To date, little is known about the signal molecules produced by the helper bacteria, the fungal factors that recognize the bacterial signal molecules as well as the fungal gene networks underlying the fungal–bacterial interactions. Xie et al. (1995) demonstrated the involvement of bacterial Nod factors in the helper effect of Bradyrhizobium japonicum on the Glomus mosseae–soybean endomycorrhizal symbiosis. Moreover, by differential RNA display, Requena et al. (1999) were able to identify a cDNA fragment coding a mRNA that was five-fold down-regulated when sporocarps from the endomycorrhizal fungus G. mosseae where inoculated with a rhizobacterial strain of Bacillus subtilis that was proved to promote the hyphal growth of G. mosseae. This transcript corresponded to a highly conserved gene encoding a multifunctional protein of the peroxisomal β-oxidation that might play a role in the catabolism of long-chain fatty acids during the fungal presymbiotic growth. Even less is known about the interactions between mycorrhiza helper bacteria and ectomycorrhizal fungi: for these fungi, no differentially regulated genes have been identified, although changes in gene profiles and protein patterns have been reported in the presence of helper bacteria. Indeed, Becker et al. (1999) described the effect of two ectomycorrhiza-associated streptomycete isolates on global gene expression and protein synthesis of the ectomycorrhizal fungus Laccaria bicolor in in vitro assays where liquid cultures of the fungus where incubated over 1–3 d with bacterial culture supernatants. Interestingly, the two bacterial isolates significantly, but differently, altered the fungal gene expression and protein patterns.
Schrey et al. also analyzed the promoting effect of a second Streptomyces isolate, on the in vitro Amanita hyphal growth and the in vitro Spruce–Amanita mycorrhiza formation. A comparative analysis of the global effect of these two helper Streptomyces isolates, as well as of the effect of other fungus-associated bacterial isolates, on the regulation of the Amanita gene expression is now required. This would complete, at the molecular scale, the analysis that Schrey et al. performed, at the phenotypic level, about the specificity of the mycorrhiza helper effect. This question of specificity was partly addressed by Schrey et al. by studying the regulation of the gene AmCyp40, which encodes a cyclophilin, a protein involved in transduction pathways that might be involved also in the regulation of cell proliferation during the bacterial enhancement of the fungal growth in vitro. The culture supernatants of the two mycorrhiza helper Streptomyces isolates induced the AmCyp40 gene expression in short-term liquid culture assay, whereas no induction was observed in the presence of a control Streptomyces isolate that exerted no effect on A. muscaria growth in vitro (Maier et al., 2004). As a consequence, it is possible that Schrey et al. have identified a key candidate gene involved in the interactions between an ectomycorrhizal fungus and a mycorrhiza helper bacterium. The challenge now is to monitor the kinetics of the expression of Amanita genes in the presence of different fungal-associated bacterial isolates, including mycorrhiza helper bacteria, not only in vitro but also in more natural conditions – in other words, in the presence of plant roots. This would confirm the assumption by Schrey et al. about the mechanisms of the mycorrhiza helper effect of their Streptomyces isolates, which were mostly based on a bacterial promotion of the Amanita hyphal growth. This would also open the way for detailed analysis of the bacterial signal molecules, the fungal major regulators and downstream signalling molecules, which are likely associated with fungal–bacterial interactions, especially in the case of mycorrhiza helper bacteria. To overcome these bottlenecks, the mining and exploitation of the data obtained from genomics and the related research areas of genome-wide transcriptomics, proteomics and metabolomics will bring an undeniable support (Martin et al., 2004).
Wider and applied prospects
As mentioned above, because of the potential practical application of mycorrhiza helper bacteria in agriculture and forestry, there is a need now for the identification of new screening criteria allowing a quick and efficient selection of performing bacterial isolates. The screening strategies used so far are too time-consuming. Therefore, any molecular approach leading to identify fungal marker genes, such as master regulators specific for the mycorrhiza helper effect, will have crucial practical outputs, especially for the improvement of tree yields in poor forest soils. Requena et al. (1999), Becker et al. (1999) and Schrey et al. can undisputably be considered as pioneers in the analysis of the molecular mechanisms underlying the interactions between helper bacteria and mycorrhizal fungi. However, many complementary studies on different mycorrhizal fungi are still required in order to identify and validate the preliminary marker genes that have been identified so far. In parallel to these studies, one should pay attention to the research on fungal–bacterial interactions presently developing in other fields, such as plant protection and medicine. Indeed, Schoonbeek et al. (2002) on the Botrytis cinerea–Pseudomonas spp. model system and Hogan et al. (2004) on Candida albicans–Pseudomonas aeruginosa have already identified different fungal genes regulated by bacteria. Comparing the mycorrhizal fungus genes differentially expressed in the presence of helper bacteria to the ones identified in these other model systems would allow us to overcome, more efficiently, the bottlenecks in studying the mechanisms of the mycorrhiza helper effect. Conversely, any breakthrough in the understanding of the mechanisms underlying the interactions between helper bacteria and mycorrhizal fungi will undoubtedly benefit the other research areas where fungal–bacterial interactions play a major practical and economical role.
We would like to thank F. Martin (INRA Nancy, France) for useful discussions and comments.