Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
• The interaction between the mycorrhiza helper bacteria Streptomyces nov. sp. 505 (AcH 505) and Streptomyces annulatus 1003 (AcH 1003) with fly agaric (Amanita muscaria) and spruce (Picea abies) was investigated.
• The effects of both bacteria on the mycelial growth of different ectomycorrhizal fungi, on ectomycorrhiza formation, and on fungal gene expression in dual culture with AcH 505 were determined.
• The fungus specificities of the streptomycetes were similar. Both bacterial species showed the strongest effect on the growth of mycelia at 9 wk of dual culture. The effect of AcH 505 on gene expression of A. muscaria was examined using the suppressive subtractive hybridization approach. The responsive fungal genes included those involved in signalling pathways, metabolism, cell structure, and the cell growth response.
• These results suggest that AcH 505 and AcH 1003 enhance mycorrhiza formation mainly as a result of promotion of fungal growth, leading to changes in fungal gene expression. Differential A. muscaria transcript accumulation in dual culture may result from a direct response to bacterial substances.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The majority of plants form symbiotic associations, termed mycorrhizas, with filamentous fungi. Mycorrhizal symbiosis improves plant nutrient and water uptake, while the fungal partner gains carbohydrates from its host plant (Smith & Read, 1997). Bacteria have been visualized inside mycorrhizas and as colonies on soil-colonizing fungal hyphae (Foster & Marks, 1966; Nurmiaho-Lassila et al., 1997), and the plant partner selects for bacterial strains beneficial for the symbiosis and for the plant (Frey-Klett et al., 2005). Specifically, the growth of ectomycorrhizal fungi and mycorrhiza formation are promoted by some of the mycorrhizosphere bacteria, termed ‘mycorrhiza helper bacteria’ (MHB; Garbaye, 1994). MHB have predominantly been reported from fluorescent pseudomonads (Garbaye & Bowen, 1989; Founoune et al., 2002) and from Burkholderia and Rhodococcus (Poole et al., 2001).
Actinomycetes are known for their capacity to control plant diseases. A number of investigations have reported antagonism of soil actinomycetes to root pathogenic fungi (Rose et al., 1980; Crawford et al., 1993), and that streptomycetes in particular provide a rich source of antifungal compounds (Mousavi & Robson, 2004). When 12 actinomycete isolates were tested for their effects on mycelial growth of ectomycorrhizal fungi (Richter et al., 1989), the bacterial isolates inhibited, promoted or showed no significant effects on hyphal extension in dual culture (DC). Three fungal species were tested in DC, Laccaria bicolor, Laccaria laccata and Thelephora terrestris, of which the slowest growing fungus, T. terrestris, was most sensitive to both growth-promoting and antagonistic actinomycetes (Richter et al., 1989). Some antagonistic actinomycetes are also producers of plant growth-promoting substances, helping plants to withstand adverse conditions and attacks by pathogens (Igarashi et al., 2002). Thus, the opportunity exists to select for actinomycetes stimulatory to certain symbiotic fungi and plant growth but antagonistic to pathogenic organisms.
We have isolated a collection of actinomycetes from the hyphosphere of a spruce (Picea abies) stand (Maier, 2003). One of these isolates, Streptomyces sp. nov. 505 (AcH 505), significantly increases the mycelial growth and mycorrhiza formation rate of Amanita muscaria but suppresses the mycelial extension of the plant pathogens Armillaria obscura and Heterobasidion annosum (Hampp & Maier, 2004; Maier et al., 2004). In contrast to AcH 505, the second MHB isolated, Streptomyces annulatus 1003 (AcH 1003), does not affect the growth of the tested plant pathogenic fungi.
The aim of the experiments presented in this study was to compare the fungus specificity of AcH 505 to that of AcH 1003, and to test the MHB function of these bacteria in a seminatural perlite-moss culture system. Given the present lack of information regarding fungal genes expressed in response to the interaction with MHB, we have used suppression subtractive hybridization (SSH; Diatchenko et al., 1996) to identify genes of A. muscaria that are preferentially expressed in DC with AcH 505. For subtraction we used the first 1 cm from the actively growing hyphal front of 9-wk-old DC, as AcH 505 most strongly promotes mycelial growth at this stage. Here we report several A. muscaria genes that are differentially transcribed in DC with the bacterium AcH 505.
Materials and Methods
Influence of streptomycetes on fungal growth in axenic conditions
The conditions for bacterial and fungal single and dual cultures, and the screening of bacterial effects on fungal growth, were essentially as described in Maier et al. (2004). Amanita muscaria growth-promoting Streptomyces sp. nov. 505 (AcH 505) and Streptomyces anulatus (Beijerinck) Waksman 1003 (AcH 1003), and the control strain Streptomyces argenteolus (Tresner et al., 1961) 504 (AcH 504) exerting no effect on hyphal growth in DC (Maier, 2003), isolated from the hyphosphere of Picea abies (L.) Karst (Norway spruce) in Haigerloch, Germany (Maier et al., 2004), were cultivated on ISP-2 agar (Shirling & Gottlieb, 1966) in the dark at 20°C.
Amanita muscaria (Fr.) Hooker strains 6 and MEII, isolated from fruiting bodies collected from the Schönbuch forest near Tübingen, Germany, Amanita muscaria strain CS83 (Schaeffer, 1995), Hebeloma cylindrosporum Romagnesi strain H1-H7 (Debaud & Gay, 1987), Paxillus involutus (Batsch: Fr.) Fr. strain ATCC 200175, originally isolated from a fruiting body associated with 15–30-yr-old Betula pendula trees growing on coal waste in Midlothian, UK, and Suillus bovinus (L.: Fr.) O. Kuntze strain K3 (Sen, 1990) were cultivated in the dark at 20°C on MMN (modified Melin–Norkrans medium) agar (Molina & Palmer, 1982) with 10 g l−1 glucose. Cultures for growth experiments were started with 4-mm-diameter inoculates, obtained from the actively growing rims of single fungal colonies placed in the centre of a sterile cellophane sheet (exclusion limit 10 kDa; Folia, Wendelstein, Germany) on MMN agar (Maier et al., 2004).
The bacterial effect on fungal growth was tested in DC without direct contact between fungus and bacterium on MMN medium [see Maier et al. (2004) for details]. A single thin streak (4 cm) of bacteria, obtained from 1-wk-old culture, was applied to the edge of a Petri dish (outside the cellophane sheet) with a sterile inoculation loop. Ten replicates were used. Culture periods were between 6 and 11 wk, depending on the mycelial growth rate of the fungal species in question. At harvest, the first 1 cm of mycelium was rapidly scraped off the cellophane layer, and the material was immediately frozen in liquid nitrogen and stored at −80°C.
Suspension culture experiments
Fungal suspension cultures were started from actively growing hyphae which were finely cut, transferred to MMN medium, homogenized and incubated at 20°C on a rotary shaker at 80 rpm. The MMN medium was exchanged once a week. Carbon-based induction of A. muscaria acetoacyl coenzyme A synthetase (AmAacs) expression was investigated according to Nehls et al. (1999). For A. muscaria cyclophilin 40 kDa (AmCyp40) induction experiments, suspension cultures of A. muscaria, grown for 5 d after MMN medium exchange, were used. Streptomycete suspension cultures were started by transferring an appropriate amount of 1-wk-old bacterial culture to MMN medium and grown for 7 d.
For DC, 10 ml of a filtered (0.22-m mesh) cell-free 7-d-old streptomycete culture supernatant, taken at an optical density of 600 nm of 0.2, was added to 15 ml of a fungal suspension culture and grown for 3 h. An untreated fungal suspension culture served as a control. All suspension cultures were kept at 20°C on a rotary shaker at 80 rpm. At harvest, suspension cultures were filtered through 100-m mesh nylon membranes. The material was immediately frozen in liquid nitrogen and stored at −80°C.
Synthesis of ectomycorrhizas and analysis of seedlings
Seeds of Norway spruce (P. abies) were obtained from Staatsklenge Nagold (Nagold, Germany) and those of Scots pine (Pinus sylvestris L.) from Forelia (Häkkilä, Finland). Seedlings were cultivated until 4 wk old according to Schaeffer et al. (1996). Ectomycorrhizas were synthesized using a Petri dish method modified from Poole et al. (2001). Moss peat (Botanical Garden, Tübingen, Germany) and perlite (Knauf Perlite GmbH, Dortmund, Germany) were mixed in a 1 : 1 ratio, crushed, moistened with double-distilled water (10 g water : 15 g mixture) and autoclaved. A slit was cut into the side of the Petri dishes and their cover using a hot scalpel. One part MMN medium (without glucose) was mixed with 4 parts peat–perlite mixture and evenly spread onto the Petri dishes. Plants and fungal inocula (A. muscaria 6 or S. bovinus K3) were positioned according to Poole et al. (2001). Briefly, plants were inserted into the slit, with the root systems placed on the peat–perlite mixture and the shoots exposed to the atmosphere. Three fungal inocula, consisting of 5-mm-diameter discs cut from the growing margin of an MMN agar culture, were placed upside-down along each side of the root at 15-mm intervals, at a distance of 5 mm from the root. For bacterial inoculations, 1-wk-old AcH 505 or AcH 1003 culture in MMN medium was centrifuged (1500 g for 15 min) and the pellets were resuspended at OD 0.6 in MMN medium without glucose. Two 500-l aliquots of bacteria were pipetted between the roots and the fungal plugs. Four-month-old seedlings were harvested for analysis. For all inoculations, 10 replicates were used. Significance was analysed with Student's t-test.
Nucleic acid extractions and construction of a subtracted cDNA library
The first 1 cm of the actively growing rims of 9-wk-old A. muscaria 83 mycelia inoculated with AcH 505 was used as starting material for cDNA subtraction. Total RNA from A. muscaria was extracted according to Nehls et al. (1998). cDNA was synthesized from 0.2 g of total RNA (Superscript II; Invitrogen, Groningen, the Netherlands) using a Smart PCR cDNA synthesis kit (Clontech, Palo Alto, CA, USA) and then PCR-amplified for 17–18 cycles using conditions recommended by the manufacturer (Clontech). Suppression subtractive hybridization (SSH) was performed between A. muscaria–streptomycete 505 DC and A. muscaria pure culture using the PCR-Select cDNA Subtraction Kit (Clontech). Subtracted gene fragments were cloned into the TOPO TA cloning vector and transformed into TOPO cells (Invitrogen).
Dot blot analysis of the subtracted library
Samples of bacterial cultures were used for PCR amplification of cDNA inserts from the subtracted library with the primers sense, 5′-TCGAGCGGCCGCCCGGGCAGGT-3′ and antisense, 3′-AGCGTGGTCGCGGCCGAGGT-5′. The conditions for the PCR reaction were as follows: 98°C for 3 min, and 30 cycles of 94°C for 30 s, 68°C for 30 s and 72°C for 2 min. cDNA inserts (200) from randomly selected clones were arrayed on duplicate positively charged nylon membranes (Roche, Mannheim, Germany). For probe synthesis, digoxigenin-UTP (DIG) labelled total cDNA was amplified from 100 ng total cDNA from A. muscaria–AcH 505 DC or A. muscaria pure culture using the Smart PCR cDNA synthesis kit (Clontech). The strength of the DIG label in the cDNA populations was equalized by dilution series to allow semiquantitative comparisons of hybridization intensities between filters. The dots were quantified using Quantity One software (Bio-Rad, Munich, Germany). Hybridizations were performed at 68°C for 16 h, and membranes were treated according to the manufacturer's instructions (Roche). Array analyses were replicated two times using first-strand cDNA from two independent cultures. The Thermo Sequence fluorescent-labelled primer cycle sequencing kit with 7-deaza-dGTP (Amersham Pharmacia Biotech, Freiburg, Germany) was used to obtain sequence information from differentially hybridized inserts. Software for the sequencing was Base Imagir 4.0 (Li-Cor, MWG-Biotech, Ebersberg, Germany). DNA sequences we compared with already known entries using the BLASTX algorithm (Altschul et al., 1990) of NCBI (http://www.ncbi.nlm.nih.gov/BLAST).
Rapid amplification of cDNA ends (RACE) PCR
The SMART RACE cDNA Amplification Kit (Clontech) was used according to the manufacturer's instructions. The Aacs cDNA was completed using the gene-specific primer 5′-TGCCGCTGGCCCGATAAGATAGC-3′ for 3′-RACE. Primers for Cyp40 RACE reactions were 5′-CCTCTATGGTATAAGGGAAG-3′ for 3′-RACE and 5′-CCTCGAATGATGAACAGGGAAACC-3′ for 5′-RACE. All RACE reactions were conducted using the respective primers for 3′- and 5′-RACE provided by the SMART RACE cDNA Amplification Kit and under standard PCR conditions.
Fragments obtained were cloned into pCR2.1 according to instructions given for the TOPO TA Cloning Kit (Invitrogen).
Nucleic acid hybridizations
For northern blot analyses, 2–3 g of total RNA was separated on a 1% agarose gel in the presence of formamide and formaldehyde in 20 mm MOPS [3-(4-Morpholino) propane sulphonic acid] buffer, and transferred to a nylon membrane. To confirm dot blot results, membranes were hybridized with DIG-labelled cDNA inserts from the subtracted library. For gene-specific hybridizations, DIG-labelled PCR fragments were obtained from the noncoding 3′ regions, using the forward primer 5′-AACTGCCGGAAAATTCTC-3′ and the reverse primer 5′-TTTCCCATGTCTTGTTCC-3′ for AmCyp40, and the forward primer 5′-CATCCTGTTCGTCAAGC-3′ and the reverse primer 5′-GCTGGGGCAGCTTAC-3′ for AmAacs, with the following PCR programme: 94°C for 3 min; 32 cycles of 94°C for 30 s, 46°C for 30 s and 72°C for 2 min; 72°C for 5 min. PCR-based labelling with DIG and hybridization procedures under high-stringency conditions were performed according to instructions provided in the PCR DIG Probe Synthesis Kit (Roche).
Amino acid sequence analysis
Web-based algorithms provided by ExPASY Molecular Biology Server (PFSCAN; http://www.isrec.isb-sib.ch/software/PFSCAN_form.html) and NCBI (conserved domain search; Marchler-Bauer et al., 2003) were used to analyse the protein sequences. The predicted full-length protein sequences of AmCyp40 and AmAacs were used for phylogenetic analyses. For the alignments, homologous sequences from other fungi detected by BLASTX were selected. In addition, the genomic DNA sequences of Coprinus cinereus (http://www.broad.mit.edu/annotation/fungi/coprinus_cinereus/) and Phanerochaete chrysosporium (Martinez et al., 2004) were used to search for homologues using the tBLASTN (AmCyp40 protein vs genomic nucleotide sequence) search. The detected amino acid sequences and their properties in comparison with the predicted A. muscaria homologues are shown in Table 1.
Table 1. Cyclophilin (Cyp) and acetoacyl coenzyme A synthetase (Aacs) homologues from the genomic DNA sequences of Phanerochaete chrysosporium and Coprinus cinereus
Theoretical isoelectric point
Theoretical MW (kDa)
% identity to Amanita muscaria protein
The properties of the predicted amino acid sequences are shown in comparison to the deduced Amanita muscaria proteins. Cyp40 and Cyp18 indicate cyclophilin proteins with the predicted theoretical molecular weights (MWs) of 40 and 18 kDa, respectively.
The peptidyl-prolyl isomerase domains of cyclophilins and full-length Aacs sequences were used for phylogenetic analyses. Sequences were aligned in ClustalX (Jeanmougin et al., 1998). To estimate the phylogenetic relationships, alignments were analysed using a Bayesian approach based on Markov chain Monte Carlo (MCMC; Larget & Simon, 1999) as implemented in the program MrBayes 3.0b4 (Huelsenbeck & Ronquist, 2001). Four incrementally heated simultaneous Monte Carlo Markov chains were run over 2 × 106 generations using random starting trees and assuming a percentage of invariable alignment sites with gamma-distributed substitution rates of remaining sites. The amino acid substitution model was sampled from the substitution models implemented in MrBayes. Trees were sampled every 100 generations, resulting in an overall sampling of 20 000 trees. The stationarity of the process was controlled using Tracer software, version 1.0 (Rambaut & Drummond, 2003). The first 8000 trees were discarded (‘burn in’) and the remaining 12 000 were used to compute a 50% majority rule consensus tree to obtain estimates of posterior probabilities. Branch lengths were averaged over the sampled trees and scaled in terms of expected numbers of amino acid substitutions per site. To test the independence of the results, the Bayesian MCMC phylogenetic analyses were performed twice for each data set. Resulting MCMC trees were rooted with Streptomyces coelicolor Aacs as outgroup for the Aacs tree and with Streptomyces antibioticus Cyp18 as outgroup for the cyclophilin tree.
GenBank/EMBL/Swissprot accession numbers for the cyclophilin sequences are as follows. PcCyp40: P. chrysosporium (not annotated); AmCyp40: A. muscaria (AY773481); CcCyp40: C. cinereus (not annotated); ScCyp40: Saccharomyces cerevisiae (NP_013317); AtCyp40: Arabidopsis thaliana (NP_565381); NcCyp40: Neurospora crassa (Q9P3 × 9); SpCyp40: Schizosaccharomyces pombe (Q11004); MmCyp40: Mus musculus (Q9CR16); HsCyp40: Homo sapiens (Q08752); HsCyp18: H. sapiens (P62937); MmCyp18: M. musculus (P17742); DlCyp18: Digitalis lantana (T50769); PcCyp18: P. chrysosporium (not annotated); CcCyp18: C. cinereus (not annotated); ScCyp18: S. cerevisiae (NP_010439); PoCyp18: Pleurotus ostreatus (CAD10797); PaCyp18: P. abies (CAC81066); StraCyp18: Streptomyces antibioticus (P83221). Aacs sequences derived from public databases were as follows. AmAacs: A. muscaria (DQ089688); CnAacs: Cryptococcus neoformans (CNBE4890); UmAacs: Ustilago maydis (UM05131.1); RnAacs: Rattus norvegicus (AB026291); AnAacs: Aspergillus nidulans (CAE47912); StrcAacs: S. coelicolor (AL939108.1); CcAacs: C. ciniereus (not annotated); PcAacs: P. chrysosporium (not annotated).
Effects of bacteria on fungal growth and mycorrhiza formation
The two streptomycete isolates, AcH 505 and AcH 1003, showed similar fungus specificities in axenic culture (Table 2). The growth of all three tested A. muscaria isolates was promoted, but significant changes in mycelial growth rates were only detected for the slowly growing (6 and 83) strains of this fungus. The extent of response to either AcH 505 or AcH 1003 varied significantly in S. bovinus and H. cylindrosporum. Growth of the S. bovinus K strain was promoted, yet both bacteria suppressed H. cylindrosporum. The margins of the fungal colonies remained isodiametric in all culture experiments. A mycorrhiza formation test in a peat–perlite culture system was conducted to determine whether AcH 505 and AcH 1003 increase the mycorrhiza formation rate (Table 3). Consistent with the results on mycelial growth promotion in DC, AcH 1003 promoted mycorrhiza formation more strongly than AcH 505. The number of mycorrhizal second-order lateral roots in Norway spruce was increased with AcH 505 (20%) and with AcH 1003 (39%) in comparison with the control treatment. In pine seedlings the number of mycorrhizal first-order lateral roots was increased by 27% with AcH 505 and by 39% with AcH 1003, and the number of second-order lateral roots was increased by 50% with AcH 505 and by 72% with AcH 1003. Furthermore, AcH 505 induced a significant increase in secondary lateral root formation in pine, and a slight increase in formation of these roots in spruce seedlings. No other significant effects on seedling growth of either plant were recorded.
Table 2. The effect of streptomycete isolates AcH 505 and AcH 1003 on the mycelial growth of a selection of ectomycorrhizal fungi
Colony area (cm2)
Values followed by a sign are significantly higher (+) or lower (–) than the value corresponding to the control treatment in the same line of the table using Student's t-test (P < 0.001). Dual culture results (AcH 505 vs AcH 1003) with differing (P < 0.1) values are marked with asterisks.
Amanita muscaria 6
Amanita muscaria 83
Amanita muscaria MEII
Suillus bovinus K3
Paxillus involutus 031
Hebeloma cylindrosporum H1–H7
Table 3. Effects of AcH 505 and AcH 1003 on lateral root and mycorrhiza formation in individual seedlings of Norway spruce (Picea abies) and Scots pine (Pinus sylvestris)
No. of first-order uninfected lateral roots
No. of second-order uninfected lateral roots
No. of first-order mycorrhizal lateral roots
No. of second-order mycorrhizal lateral roots
Total no. of roots
First-order root mycorrhizal rate (%)
Second-order root mycorrhizal rate (%)
Total Mycorrhizal rate (%)
Picea abies was grown with Amanita muscaria for 12 wk and Pinus sylvestris with Suillus bovinus K3 for 16 wk. Values significantly higher than the control treatment (P < 0.01) are indicated with (+). Mean values of 10 replicates are given, with standard deviations in parentheses underneath.
Picea abies-Amanita muscaria 6
Pinus sylvestris–Suillus bovinus K3
The use of the cDNA subtraction technique to select preferentially expressed genes
As hyphal growth promotion was found to be an important factor for the MHB effect, we performed an SSH analysis with the aim of isolating fungal genes preferentially expressed in DC, i.e. the genes that are related to the induction of growth of A. muscaria hyphae through AcH 505. First, the time course of A. muscaria growth promotion was analysed to determine the appropriate stage for mRNA isolations (Fig. 1). The strongest induction with AcH 505 and AcH 1003 was detected between 8 and 10 wk of growth. Nine-wk-old AcH 505-induced A. muscaria 83 mycelia were used for gene expression analysis. Differentially expressed sequences were cloned to obtain a subtracted Am-505 cDNA library. The average size of DNA inserts in the subtracted library was 450 bp.
Altered fungal gene expression in DC
Two hundred randomly picked clones from the subtracted library were analysed by dot blot analysis (data not shown). Fifty-eight clones (28%) showed at least 2-fold increase or decrease in hybridization intensities, 45 of them being up-regulated and 13 down-regulated. Twenty-one of these differentially regulated genes showed statistically significant matches against database sequences (expected value < e-10) (Table 4). Differential expression of nine genes was validated by northern analysis. These genes were related to important aspects of the fungus–bacterium interaction, or were unknown genes with significantly altered hybridization intensities. The observed changes in expression rates were comparable to those detected in dot blot analysis (data not shown). For genes of known function, the highest differential expression (6-fold up-regulated) was observed for cyclophilin 40, and the lowest (3-fold down-regulated) for the gamma-aminobutyric acid/polyamine permease. Among the genes of unknown function, clones with strongly increased (6-fold up-regulated; CV632830) and decreased (5-fold down-regulated; DN779411) levels of transcription were found (Fig. 2).
The GenBank accession numbers of the nucleotide sequences are given. BLASTX searches gave differing degrees of identity to genes in other organisms (expected values). The first clone has the highest and the last clone the lowest apparent Am–AcH 505/Am transcript ratio, detected by dot blot analysis. Clones marked with an asterisk are less strongly expressed in the interacting fungal hyphae.
Am, Amanita muscaria; PAK, p21-activated kinase; Tpk3, Protein kinase A catalytic subunit 3.
The majority of the genes with known function were related to signal transduction pathways (seven genes; protein kinases and transcription factors) and metabolism (five genes; carbon-metabolism: acetoacyl coenzyme A synthetase; sulphur-metabolism: adenylyl sulphate kinase and sulphite reductase; nitrogen metabolism: glutamine synthetase and protease). In addition, genes coding for proteins involved in cell growth or cell stress (five genes; cyclophilin 40, gamma-aminobutyric acid/polyamine transporter, ribosomal protein genes and histone H4) and cell structure [two genes; beta-glucosidase and p21-activated kinase (PAK)] were identified.
Two up-regulated genes were selected for further analysis: Aacs, to analyse whether the increased Aacs expression was related to growth activation of A. muscaria mycelia, and Cyp40, to investigate whether Cyp40 expression increased as a result of the addition of bacterial culture supernatants.
Isolation and expression analysis of Aacs
Aacs catalyses the activation of acetoacetate, which is needed for efficient sterol biosynthesis (Bergstrom et al., 1984). Increased expression of a functionally related gene, acetyl coA acetyl transferase (Acat), was observed in L. bicolor hyphae grown in contact with pine root exudates (Podila et al., 2002). We investigated whether the AmAacs transcription level was increased as a result of activation of fungal mycelium, as was suggested for L. bicolor Acat by Podila et al. (2002).
As no previous information about fungal Aacs proteins existed, we isolated the full-length AmAacs cDNA by RACE. The predicted 702 amino acid AmAacs protein sequence includes two conserved and overlapping regions according to conserved domain search (CD-Search; Marchler-Bauer et al., 2003), namely the AMP-binding enzyme region (amino acids 124–577) and the acyl coenzyme A synthetase/AMP (fatty) acid ligase region (amino acids 81–667). Using the predicted AmAacs amino acid sequence, we detected homologue sequences in the published genomes of the homobasidiomycetes C. cinereus and P. chrysosporium (66% and 66% identity at the amino acid level, respectively), which are both close relatives of A. muscaria. Several AmAacs homologues were found by BLASTX, such as the hypothetical protein CNBE4890 from the human opportunistic pathogen C. neoformans (48% identity), the hypothetical protein UM05131.1 from the maize pathogen Ustilago maydis (45%), and the rat Aacs (41%; Iwahori et al., 2000). By contrast, no AmAacs homologue was found in the budding yeast genome database. The predicted amino acid sequences of the AmAacs homologues had comparable molecular weight, pI, AMP-binding and acyl coenzyme A synthetase domain sequences (data not shown). As expected, the homobasidiomycete (A. muscaria, C. cinereus and P. chrysosporium) Aacs proteins grouped together in phylogenetic analysis (Fig. 3a).
We used submerged A. muscaria culture to analyse whether the changed AmAacs expression in DC is a result of the activation of dormant hyphae (Podila et al., 2002). The effect of addition of glucose on Aacs expression was determined in A. muscaria suspension culture. The glucose-supported (100 mm) fungal culture showed a 3-fold increase in Aacs expression in comparison with glucose-starved hyphae (Fig. 3b). A 2-fold up-regulation of Aacs expression was observed in DC at 9 wk of culture (Fig. 3c).
Isolation and characterization of Cyp40 cDNA
The Cyp40 proteins are prolyl isomerases, which are necessary for efficient cell growth and endurance of cell stress in S. cerevisiae (Duina et al., 1996). Of all subtracted genes of known function, the level of A. muscaria Cyp40 transcription is most significantly increased in DC hyphae, and we decided to investigate whether comparable differential expression occurs already in a short-term culture.
Using RACE PCR, the complete cDNA sequence of the first cyclophilin of A. muscaria was determined and denoted AmCyp40 for the A. muscaria cyclophilin 40-kDa protein. Homology searches using genome sequences or the BLASTX algorithm revealed related proteins belonging to the cyclophilin family, most notably from C. cinereus and P. chrysosporium genomes (70% and 63% identity at the amino acid level, respectively).
A motif scan of the predicted AmCyp40 protein sequence by PFSCAN detected a cyclophilin-type peptidyl-prolyl cis-trans isomerase domain of 166 amino acids. In addition, the profile scan also detected the presence of two putative tetratricopeptide repeat (TPR) motifs for protein–protein interactions and a putative calmodulin-binding site. An analysis using the ‘PESTfind’ algorithm revealed the presence of two putative PEST sequences. A study of 12 other fungal and mammalian 40-kDa cyclophilins obtained from the GenBank/EMBL database showed that only the fungal cyclophilins contain PEST sequences. TPR motifs are only absent from the 40-kDa cyclophilin wis2 of S. pombe, and all investigated proteins contain the putative calmodulin-binding site. A phylogenetic analysis of AmCyp40 based on the conserved peptidyl-prolyl cis-trans isomerase region (166 amino acids) of the deduced protein sequence was used for the determination of the phylogenetic position of AmCyp40 among the eukaryotic cyclophilins. Cyclophilins clustered (i) according to the size of the entire protein and (ii) according to the relatedness of organisms. AmCyp40 grouped with the other homobasidiomycete 40-kDa cyclophilins (Fig. 4a).
AmCyp40 is highly expressed in the DC hyphae with both AcH 505 and AcH 1003 (data not shown). In order to determine whether the increase in AmCyp40 transcript is a growth effect, the expression was analysed in short-term experiments in suspension culture. A. muscaria suspension cultures were incubated with a cell-free supernatant of the streptomycete strains AcH 505 and AcH 1003 and cultivated for 6 h. Incubation with supernatant of the strain AcH 504, which exerts no effects on A. muscaria growth (Maier et al., 2004), served as a control. Levels of AmCyp40 transcription were strongly increased in the presence of AcH 505 and AcH 1003 supernatants after 6 h, while there was no effect with control strain AcH 504 (Fig. 4b).
The two streptomycete strains investigated in this contribution have similar specificities for ectomycorrhizal fungi. This is surprising, as Maier et al. (2004) showed previously that AcH 505 suppresses the growth of H. annosum and Armillaria abietina, whereas AcH 1003 has no effect on these plant pathogens. This indicates an important application for ectomycorrhiza helper bacteria such as AcH 505: the simultaneous growth promotion of mycorrhizal and growth suppression of pathogenic fungi.
Resistance to antifungal substances can be an important characteristic for fungal species colonizing the rhizosphere (Whipps, 2001). H. cylindrosporum is inhibited in its growth and mycorrhiza formation by MHB (Bacillus, Pseudomonas) associated with L. laccata (Garbaye & Duponnois, 1992). Accordingly, the streptomycetes of this study suppressed H. cylindrosporum growth in DC experiments, indicating that H. cylindrosporum is sensitive to antifungal substances produced by very different bacterial species.
Fungi and bacteria were spatially separated in DC, although a continuum of nutrient solution existed in the Petri dish cultures. The shape of fungal colonies remained isodiametric in all cultures, indicating that the substances that promote (A. muscaria and S. bovinus) mycelial growth must readily diffuse in the culture medium, more rapidly than the mycelial growth rate of responsive fungi. Such an effect has been previously shown for the ECM fungus Pisolithus tinctorius, as even a local application of the small fungal growth-promoting plant secondary substance rutin did not result in asymmetric colony extension of this fungus (Lagrange et al., 2001). Although streptomycetes are known to produce a variety of volatile compounds (Schöller et al., 2002), volatiles from AcH 505 and AcH 1003 did not show any effects on the mycelial extension rates of the six tested fungal isolates. To investigate the differential responses of the fungal species in detail, we are currently analysing mycelial growth-stimulating and -suppressing substances from the AcH 505 culture supernatant.
Differential gene expression in DC
Subtractive hybridization is a well-accepted method to detect differentially regulated genes in ectomycorrhizal fungi. It has been used to select fungal genes expressed preferentially in the presymbiotic phase (Podila et al., 2002), in ectomycorrhizal tissues (Voiblet et al., 2001; Peter et al., 2003) and in extramatrical hyphae (Morel et al., 2005). We present the first detailed report on differential fungal gene expression during DC with mycorrhiza helper bacteria. Becker et al. (1999) showed modulation of in vitro translation products after inoculation of L. bicolor with culture filtrates from a growth-promoting and a growth-suppressing streptomycete strain, but no information concerning the identity of the regulated mRNAs was given in their contribution. The presented data thus identify the first fungal genes with altered expression rates in DC with ectomycorrhiza helper bacteria, as well as providing the first clues to the mechanisms that lead to growth promotion of fungal mycelia. Three aspects of mycorrhiza helper bacteria were chosen for a more detailed analysis: enhancement of the fungal growth rate, the stress response and signalling between bacteria and fungi.
Accelerated fungal growth and cell stress The presented data show that A. muscaria is able to change its gene expression in response to bacterial stimuli. The largest number of induced genes (five) are related to fungal metabolism. These include AmAacs, the first acetoacetyl-coA synthetase cDNA isolated from fungi. Aacs activity (Bergstrom et al., 1984) and expression profiles (Ohgami et al., 2003) suggest that this enzyme is involved in cholesterol biosynthesis. We suggest that part of the response of A. muscaria to AcH 505 is the induction of increased expression of AmAacs, leading to higher production of ergosterol, which is needed for promoted mycelial growth. As carbon, especially glucose, is a key regulator of fungal gene expression in ECM symbiosis (Nehls et al., 1999; Nehls, 2004), the levels of AmAacs expression were analysed in carbon-depleted (dormant) and in well-supported (activated) A. muscaria suspension cultures. We found an up-regulation of Aacs expression in the glucose-supplemented culture, comparable to that seen in DC experiments at 9 wk of growth. After 9 wk of DC, the fungal mycelium nearly covered the Petri dish, and it had probably consumed most of the available glucose. As no agar-degrading activity that might lead to an additional carbon supply has been observed in bacterial cultures, the observation suggests that (i) the fungus is able to change its gene expression in response to bacterial stimuli, and (ii) an effective bacterial compound acting in a similar manner to glucose could be produced by the bacterium.
The transport of polyamines and gamma-aminobutyric acid are mediated by the vacuolar Uga4 permease in budding yeast S. cerevisiae (Uemura et al., 2004). The inhibitor of polyamine biosynthesis difluoromethylornithine decreases the mycelial growth rate of P. involutus (Zarb & Walters, 1994). The down-regulation of Uga4 in DC may thus reflect changes in individual polyamine concentrations in the DC hyphae, necessary for the faster growth of fungal mycelia.
Peptidyl-prolyl cis-trans isomerases (PPIases) are enzymes that bind to mature proteins to control their activity, subunit assembly and localization (Dolinski & Heitman, 1997). Cyp40 proteins are involved in the regulation of cell proliferation and differentiation and in the control of cell stress. The closest characterized homologues of AmCyp40 are the yeast Cyp40 family members encoded by cpr6 and cpr7 genes. Yeast strains with disrupted cpr7 show a significant impairment of the rate of cell division, leading to the slow growth phenotype (Duina et al., 1996). The elevation of AmCyp40 transcription in coculture could thus play a role in the enhanced hyphal growth rate observed in DC. As the streptomycete culture supernatants were added to submerged fungal cultures, an increased level of AmCyp40 expression was observed with both AcH 505 and AcH 1003 supernatants. By contrast, the use of the culture supernatant of S. argenteolus 504, a streptomycete strain exerting no effects on A. muscaria growth, resulted in basal levels of AmCyp40 expression. This indicates that additional factors, perhaps the presence of stimulatory compounds in the supernatants of AcH 505 and AcH 1003, lead to increased AmCyp40 expression.
ECM fungi are known to release extracellular enzymes, such as mono-oxidases, capable of oxidizing aromatic (Gramss et al., 1998) and nonaromatic (Takakura & Kuwata, 2003) substances. A. muscaria mono-oxidase expression was observed to be increased in DC hyphae, suggesting an activation of specific steps in either secondary metabolism or catabolism in A. muscaria.
Signalling between bacteria and fungi Changes in gene expression levels occur in parallel with modulated transcription of signalling-related genes such as PAK and cAMP-dependent protein kinase (PKA). PAK is a central regulator of polarized growth in eukaryotes (Leveleki et al., 2004), and the induced level of PAK expression suggests modulated organization of the actin cytoskeleton in DC hyphae. PKA regulates the cell cycle and metabolism in yeast and mammalian cells. The increased expression level of the gene related to the catalytic subunit of PKA, Tpk3, in A. muscaria indicates altered growth-related signal transduction in DC.
The differentially expressed genes detected in this in vitro approach could serve as a basis for the analysis of A. muscaria gene expression in more natural culture systems such as nonsterile microcosms or pot cultures. Analysis of comparable fungal genes in the growth-promoted cultures of S. bovinus, and in the growth-inhibited cultures of H. cylindrosporum and H. annosum, would provide information about the specificity of the fungal response to AcH 505.
Mycorrhiza formation tests
Our findings support the view of Poole et al. (2001) that MHB have a ubiquitous distribution, and that the main helper mechanism is fungal growth promotion. This widens the scope of MHB investigations, as previous studies mainly focused on Bacillus and Pseudomonas species (Garbaye, 1994; Founoune et al., 2002), and resulted in the isolation of MHB from Burkholderia and Rhodococcus (Poole et al., 2001).
As the ability of the bacteria to stimulate mycelial growth correlated with their effect on mycorrhiza formation, we suggest that the effective mechanism for MHB is predominantly hyphal growth promotion. When Bacillus and Pseudomonas MHB were investigated, the behaviour of fungal mycelium in DC correlated with mycorrhiza formation tests (Garbaye & Duponnois, 1992; Frey-Klett et al., 1997; Founoune et al., 2002), and Poole et al. (2001) suggested that promoted mycelial growth of Lactarius rufus is the probable mechanism for the MHB outcome in the Scots pine model system. According to the observations of Richter et al. (1989) with T. terrestris, the A. muscaria strain with the slowest observed growth rate (strain 6) was the most sensitive to the streptomycete isolates in in vitro tests. This suggests that MHB may compensate for slow mycelial growth rates in receptive fungal isolates.
For A. muscaria, however, we have observed that the fungal mycelial density decreases in DC with both streptomycete isolates (Maier, 2003; M.T., unpublished data). This suggests either that the observed MHB effect is caused by a faster spreading of fungal mycelia on the peat–perlite culture system (but not by the higher mycelial biomass) or that additional, plant-related factors are involved. As originally suggested by Garbaye (1994), the latter could include increased receptivity of the roots to the fungal hyphae, or improved fungal colonization of the fine roots produced by plant cell wall modifying bacterial substances. According to our results, and a previous study on mycorrhiza formation in Norway spruce on a MMN agar medium conducted by Maier (2003), AcH 505, but not AcH 1003, promotes root branching. Plant growth promoters with auxin-like activity have previously been isolated from Streptomyces hygroscopicus (Igarashi et al., 2002).
To conclude, the present data confirm that the streptomycetes AcH 505 and AcH 1003 are MHB, exerting their positive action mainly through the activation of fungal metabolism. As a result of the changed gene expression levels, fly agaric mycelia activated in DC should have a modulated physiology. The question now arises as to which bacteria-generated substances give rise to the improved growth of selected mycorrhizal fungi and inhibit preferentially the growth of phytopathogenic fungi.
We are indebted to Dr Andreas Maier for advice and assistance in setting up the dual cultures, Dr Uwe Nehls for helpful suggestions and for critical reading of the manuscript, Dr Ingrid Kottke for advice on the implementation of peat–perlite cultures and Dr Michael Weiss for advice on phylogenetic analysis. We thank Dr Robin Sen, Dr Anders Tunlid and Dr Daniel Wipf for the strains of ectomycorrhizal fungi. This work was supported by the Strukturfonds of the University of Tübingen (M.T.) and the German Research Foundation (D.F.G.; graduate school ‘Infection Biology’; S.D.S.).