Saprotrophic basidiomycete mycelia and their interspecific interactions affect the spatial distribution of extracellular enzymes in soil

Authors


  • Editor: Jim Prosser

Petr Baldrian, Laboratory of Environmental Microbiology, Institute of Microbiology of the ASCR, Vídeňská 1083, 14220 Prague 4, Czech Republic. Tel.: +420 723 770 570; fax:+420 241 062 396; e-mail: baldrian@biomed.cas.cz

Abstract

Saprotrophic cord-forming basidiomycetes are important decomposers of lignocellulosic substrates in soil. The production of extracellular hydrolytic enzymes was studied during the growth of two saprotrophic basidiomycetes, Hypholoma fasciculare and Phanerochaete velutina, across the surface of nonsterile soil microcosms, along with the effects of these basidiomycetes on fungi and bacteria within the soil. Higher activities of α-glucosidase, β-glucosidase, cellobiohydrolase, β-xylosidase, phosphomonoesterase and phosphodiesterase, but not of arylsulphatase, were recorded beneath the mycelia. Despite the fact that H. fasciculare, with exploitative hyphal growth, produced much denser hyphal cover on the soil surface than P. velutina, with explorative growth, both fungi produced similar amounts of extracellular enzymes. In the areas where the mycelia of H. fasciculare and P. velutina interacted, the activities of N-acetylglucosaminidase, α-glucosidase and phosphomonoesterase, the enzymes potentially involved in hyphal cell wall damage, and the utilization of compounds released from damaged hyphae of interacting fungi, were particularly increased. No significant differences in fungal biomass were observed between basidiomycete-colonized and noncolonized soil, but bacterial biomass was reduced in soil with H. fasciculare. The increases in the activities of β-xylosidase, β-glucosidase, phosphomonoesterase and cellobiohydrolase with increasing fungal : bacterial biomass ratio indicate the positive effects of fungal enzymes on nutrient release and bacterial abundance, which is reflected in the positive correlation of bacterial and fungal biomass content.

Introduction

Saprotrophic basidiomycetes are the major agents of the decomposition of lignocellulose in wood and leaf litter. Unlike saprotrophic bacteria and resource-restricted saprotrophic fungi, some saprotrophic basidiomycetes form large mycelia extending over tens of centimetres or metres (Smith et al., 1992; Ferguson et al., 2003; Boddy et al., 2009). This makes these organisms particularly well adapted for terrestrial environments, which are characterized by a largely heterogeneous distribution of resources in both space and time (Fricker et al., 2008a). The size and morphology of mycelia enables saprotrophic basidiomycetes to capture and assimilate such dispersed nutrients efficiently and translocate them to those parts of mycelia where they are required (Boddy, 1993; Lindahl et al., 2001; Harris & Boddy, 2005; Fricker et al., 2008b). In forest soil, mycelia form a network interconnecting resource patches at the soil/litter interface and sometimes penetrate deeper into soil (Boddy, 1993, 1999; Carlile, 1994; Andersson et al., 2001; Šnajdr et al., 2008b). The network architecture of fungal mycelia reflects the distribution of nutrient patches in soil or litter in that it shows preferential biomass production near nutrient patches and strong connections between these patches (Boddy et al., 2006; Howard et al., 2007). Saprotrophic basidiomycetes produce a wide variety of extracellular enzymes (Valášková et al., 2007; Šnajdr et al., 2010), considerably broader than those produced by saprotrophic ascomycetes (Baldrian et al., 2011). When colonizing a soil profile, saprotrophic basidiomycetes increase the activity of decomposition-related enzymes in soil (Šnajdr et al., 2008a). The distribution of extracellular enzymes in forest topsoil is highly spatially variable (Šnajdr et al., 2008b) and the activity of several extracellular enzymes correlates positively with fungal biomass (Baldrian et al., 2010a, b). Because the size of activity hotspots of several enzymes corresponds with the size of individual basidiomycete mycelia, the presence of saprotrophic basidiomycetes is probably a major determinant of the spatial distribution of enzyme activities.

When saprotrophic basidiomycete mycelia meet, they interact aggressively, the overall outcome being a deadlock in which neither mycelium gains headway, or replacement, where one mycelium gains territory from the other (Boddy, 2000). These interspecific mycelial interactions are complex phenomena that can start before physical contact of mycelia through the production of volatile and/or diffusible organic compounds, often resulting in the inhibition or the stimulation of other organisms in the environment, for example the microinvertebrate grazers of fungal mycelia (Hynes et al., 2007; Evans et al., 2008). Interspecific mycelial interactions are characterized by changes in mycelial morphology, increased activities of oxidative enzymes, especially laccases or manganese peroxidases (Lang et al., 1997; Baldrian, 2004; Chi et al., 2007; Hiscox et al., 2010), or the production of reactive oxygen species (Tornberg & Olsson, 2002), both in the immediate region of interaction and elsewhere in the mycelia. Because of the release of organic compounds from damaged hyphae (Wells & Boddy, 2002), zones of mycelial interactions represent distinct patches with altered nutrient availability. Although the increase of activity of hydrolytic enzymes is expected in the interaction zones, their production has not yet been experimentally addressed.

The fact that saprotrophic basidiomycetes decompose recalcitrant organic matter and thus alter the distribution of available nutrients implies that they must affect other microorganisms. Even though bacteria occupy different nutritional niches to fungi (de Boer et al., 2005), they compete (Romani et al., 2006) and the presence of saprotrophic basidiomycetes in wood or soil affects the bacterial community (Folman et al., 2008; Valášková et al., 2009; de Boer et al., 2010). The production of oxidative enzymes and toxic oxygen species impacts negatively on bacteria, but the availability of nutrients released by extracellular digestion and from fungal hyphae may affect soil microbial communities positively. The extent to which these factors affect soil bacteria and fungi has not yet been elucidated.

The objective of this study was to investigate the production of extracellular hydrolytic enzymes involved in decomposition during the growth of two saprotrophic basidiomycetes, Hypholoma fasciculare and Phanerochaete velutina, in nonsterile soil microcosms. While H. fasciculare represents a fungus with an exploitative strategy, forming mycelial fans densely covering soil or litter surfaces, P. velutina produces sparser, explorative mycelium that extends rapidly in the search for patchy lignocellulose nutrients. We hypothesized that (1) the presence of each of these two saprotrophic fungi will increase the activity of decomposition-related enzymes, but (2) the exploitative strategy of H. fasciculare will result in higher enzyme production. Furthermore, (3) increased activity of extracellular enzymes is expected during mycelial interactions due to the presence of nutrients released from the interacting hyphae. The distribution of eight hydrolytic enzymes involved in the C, N, P and S cycle was quantified beneath fungal mycelia and in the zones of mycelial interaction. Because the activity of fungal extracellular enzymes leads to local changes in nutrient availability, the effects of fungal presence and interactions on the soil microbial community can be expected. Bacterial and fungal biomass in soil beneath the two competing basidiomycetes, in noncolonized soil and in the interaction zone, was quantified to evaluate these effects.

Materials and methods

Fungal strains and cultivation

Saprotrophic cord-forming basidiomycetes H. fasciculare DD2 and P. velutina Pv29 were obtained from the culture collection of Cardiff University. Fungi were maintained on malt extract agar (MEA; 20 g L−1 malt extract, 15 g−1 agar). Beech (Fagus sylvatica) wood blocks (2 × 2 × 2 cm) were cut from a freshly felled tree and stored at −18 °C until required. Before use, they were soaked overnight in deionized water to defrost and then autoclaved twice at 121 °C for 45 min in sealed autoclave bags, with a 24-h interval. Blocks were colonized, by placing them onto 30-day cultures of fungi on MEA in 14-cm-diameter Petri dishes, and were incubated at 20 °C in the dark for 15 weeks.

Soil microcosms

Topsoil was collected from <20 cm depth from deciduous woodland in the Coed Beddick Inclosure (Tintern, UK; National Grid ref. SO 528 018). Wood and leaf litter were removed and the soil was sieved (10 mm mesh) and stored in plastic bins. Soil organic matter was 12% and soil pH was between 4.6 and 4.7. Soil was air-dried for 21 days before use, frozen for 1 day to kill soil fauna and then sieved through a 2-mm-metal mesh to remove organic material and stones. Soil was rewetted to a water potential of −0.012 MPa and compacted evenly into 24 × 24 cm lidded bioassay dishes (200 g wet soil per dish, soil depth approximately 4 mm). Dishes were weighed and rewetted to the original weight with a fine mist of water every 7 days. Inoculated wood blocks from culture Petri dishes were scraped free of adhering mycelium and agar and pushed firmly into the centre of each dish for the single-species microcosms. In the microcosms used to study P. velutina × H. fasciculare interactions, wood blocks precolonized with the respective fungi were placed approximately 12 cm from each other along a diagonal (Fig. 1). Inoculated dishes were incubated in the dark at 19 ± 1 °C (Rotheray et al., 2008). There were five replicates per treatment.

Figure 1.

 Mycelial growth of Hypholoma fasciculare and Phanerochaete velutina and the activity of selected extracellular enzymes in single-species and interaction-soil microcosms. Heat maps show the relative enzyme activities in individual soil squares. Thin black lines indicate the extent of the young and old H. fasciculare mycelia in the H. fasciculare microcosms, old P. velutina mycelia in P. velutina microcosm and the interaction zone in the interaction microcosm. Black dots mark the location of fungal inocula. The bar charts show the mean and SE of the mean by soil region. Different letters above bars indicate statistically significant (P≤0.05) differences among regions. I, interaction zone; Ao, old mycelium of H. fasciculare; Bo, old mycelium of P. velutina; A, young mycelium of H. fasciculare; B, young mycelium of P. velutina; C, no colonization by saprotrophic basidiomycetes.

Analysis of soil from microcosms

The fungi growing alone and interacting with each other were left to colonize soil for approximately 1 month before harvesting; this was approximately 10 days after the first contact of H. fasciculare and P. velutina mycelia. Microcosms were photographed and the central parts of microcosms were cut into a square 2.2 × 2.2 cm grid (100 squares per microcosm). Cut soil pieces were classified by a visual inspection of fungal colonization into regions of noncolonized soil (C), soil colonized by young mycelium (<10 days) of each individual fungus and soil colonized by old mycelium. In interaction microcosms, squares containing mycelia of both fungi were designated as belonging to the interaction zone (I). For all treatments (H. fasciculare, P. velutina, interaction), all soil squares (100 samples) from one whole microcosm were analysed separately. In addition, at least five samples per sample type per microcosm were analysed from at least three other microcosms per treatment. Subsamples of soil for ergosterol extraction (0.5 g), DNA extraction (0.3 g) and enzyme assays (0.3 g) were immediately frozen and kept at −18 °C until analysis. The dry mass of soil in each square was assessed by incubating soil at 85 °C until a constant mass.

Image capture and analysis

Digital images of microcosms were captured using an μ780 camera (Olympus, Tokyo, Japan), mounted on a stand at a height of 45 cm, before sampling. Saved images were processed using imagej (National Institute of Health, Bethesda). Microcosm edges were electronically removed from all images, before conversion to greyscale (eight bit), and images were adjusted to 2600 × 2600 pixels. One sampling square was represented by 260 × 260 pixels. Images were then subject to manual thresholding, to reduce the effect of any unevenness in soil colour; any pixels with a grey value less than the threshold were converted to black, to represent soil, and pixels with values higher than the threshold were converted to white, representing fungal mycelium. Hyphal coverage was determined as the percentage of white pixels in the binary image (Wood et al., 2006).

Enzyme assays

For enzyme activity assays, 0.3 g soil was homogenized in 50 mL of 50 mM sodium acetate buffer (pH 5.0) using an UltraTurrax (IKA Labortechnik, Staufen, Germany) for 3 min at 8000 r.p.m. in an ice bath (Štursová & Baldrian, 2011). The activities of hydrolytic enzymes arylsulphatase (E.C. 3.1.6.1), 1,4-α-glucosidase (E.C. 3.2.1.20), cellobiohydrolase (CBH; E.C. 3.2.1.91), 1,4-β-xylosidase (E.C. 3.2.1.37), 1,4-β-glucosidase (E.C. 3.2.1.21), N-acetylglucosaminidase (E.C. 3.2.1.30), phosphodiesterase (E.C. 3.1.4.1) and phosphomonoesterase (E.C. 3.1.3.2) were then measured using 4-methylumbelliferol (MUF)-based substrates as described previously (Baldrian, 2009). Substrates (40 μL in dimethylsulphoxide) at a final concentration of 500 μM were combined with three technical replicates of 200 μL of soil homogenates in a 96-well plate. For the background fluorescence measurement, 200 μL of each soil homogenate was combined with 40 μL of MUF standards to correct the results for fluorescence quenching. The multiwell plates were incubated at 40 °C and fluorescence was recorded from 5 to 125 min using a microplate reader infinite (TECAN, Männedorf, Switzerland), with an excitation wavelength of 355 nm and an emission wavelength of 460 nm. The quantitative enzymatic activities were calculated after blank subtraction based on a standard curve of MUF. One unit of enzyme activity was defined as the amount of enzyme releasing 1 nmol of MUF min−1 and expressed as g−1 soil dry mass. To correct for differences in the mean enzyme activities among individual microcosms (± 20%), enzyme activities were expressed as a percentage of the microcosm mean.

Quantification of microbial biomass

Total ergosterol was extracted and analysed as described previously (Šnajdr et al., 2008a). Briefly, samples (0.5 g) were sonicated with 3 mL 10% KOH in methanol at 70 °C for 90 min. Distilled water (1 mL) was added and the samples were extracted three times with 2 mL cyclohexane, evaporated under nitrogen, redissolved in methanol and analysed isocratically using a Waters Alliance HPLC system (Waters, Milford) with methanol as a mobile phase at a flow rate of 1 mL min−1. Ergosterol was quantified by UV detection at 282 nm.

Quantitative PCR(qPCR) was used to determine the bacterial and fungal biomass. For the relative quantification of H. fasciculare and P. velutina, qPCR with species-specific primers for the rRNA region of genomic DNA was used. DNA from soil samples (0.3 g) was isolated using the modified Miller-SK method (Sagova-Mareckova et al., 2008) and purified using the Geneclean Turbo Kit (MP Biomedicals). Two sets of specific PCR primers were used to quantify the relative amounts of fungal and bacterial DNA, qITS1 (TCCGTAGGTGAACCTGCGG) and qITS2* (TTYGCTGYGTTCTTCATCG) for fungi, and 1108F (ATGGYTGTCGTCAGCTCGTG)/1132R (GGGTTGCGCTCGTTGC) for bacteria as described previously (Vìtrovský et al., 2010). The primer qITS2* was originally designed by Prof. Lee Taylor (University of Alaska, Fairbanks; Wakelin et al., 2007) as a modification of the widely used primer ITS2 (White et al., 1990) to enable the amplification of the widest possible range of fungi, based on additional sequences that revealed several mismatches with the original ITS2. For qPCR purposes, the primer was tested with DNA from isolated strains of soil fungi from hardwood forest soil (Baldrian et al., 2011) and with environmental DNA samples from a wide set of soils (Štursová & Baldrian, 2011). Amplifications were performed on a StepOne Plus cycler (Applied Biosystems, Carlsbad) using optical grade 96-well plates. Each 20 μL reaction mixture contained 10 μL SYBR Green Master Mix (Applied Biosystems), 0.9 μL bovine serum albumin (BSA) (10 mg mL−1), 1.35 μL of each primer, 1.5 μL of template and 6.1 μL of water. The PCR cycling protocol was the same for fungal and bacterial DNA quantification: 56 °C for 2 min, 95 °C for 10 min, 95 °C for 15 s and 60 °C for 1 min (40 cycles). Plasmids with cloned inserts of the targeted regions from Streptomyces lincolnensis DNS40335 and H. fasciculare CCBAS281 were used as PCR standards.

Previously described primers HfF (CACCTTTTGTAGACCTGGATT) and HfR (AGTGCTATAAACGGCAAATAG) (Eyre et al., 2010) for the quantitative PCR detection of H. fasciculare and newly designed primers PvF (AACGCACCTTGCGCTCCCT) and PvR (CTTCACGACCACGGCGCAGA) were optimized and used for P. velutina. Amplifications were performed on a StepOne Plus cycler (Applied Biosystems) using optical grade 96-well plates. Each 20 μL of reaction mixture contained 10 μL of SYBR Green Master Mix (Applied Biosystems), 0.9 μL of BSA (10 mg mL−1), 1.35 μL of each primer, 1.5 μL of template and 6.1 μL of water. The PCR cycling protocol for the HfF/HfR pair was as follows: 56 °C for 2 min, 95 °C for 10 min, 95 °C for 15 s and 60 °C for 1 min (40 cycles). For the PvF/PvR pair, the annealing and the extension temperature was 69 °C. Plasmids with cloned inserts of the targeted regions from H. fasciculare CCBAS281 and P. velutina Pv29 were used as PCR standards.

Plasmids with inserted target sequences and total genomic DNA of H. fasciculare and P. velutina were used as standards; DNA content was expressed in gene abundance g−1 soil dry mass or as a ratio of different qPCR targets. To correct for the difference in rRNA copy number ng−1, genomic DNA of the studied basidiomycetes (2.96 × 106 ng−1 DNA in H. fasciculare and 2.27 × 106 ng−1 DNA in P. velutina), the biomass of these fungi was also quantified as ng of their genomic DNA instead of copy numbers.

Statistical analyses

Statistical analyses were performed using the software package statistica 7 (StatSoft, Tulsa). Interaction microcosms with all regions (including the interaction zones) were used for the comparison of enzyme activities. Similar values of enzyme activities in the noncolonized soil regions and regions covered by mycelium of individual basidiomycetes were found in the single-fungus microcosms, but these did not allow comparison with the interaction zone. Differences between soil regions were tested using one-way anova, followed by the Fisher LSD post hoc test; for qPCR data, which followed a lognormal distribution, data were log-transformed before analysis. Principal component analysis was used to analyse the differences in the relative activities of individual enzymes in different areas of soil microcosms. Correlation analysis was used to determine the relationships between mycelial coverage and enzyme activities. These analyses were only derived from single-fungus microcosms, where mycelia could be reliably assigned to one of the two basidiomycete fungi. In all cases, differences at P≤0.05 were regarded as statistically significant.

Results

Distribution of extracellular enzymes in soil with individual and interacting fungal mycelia

Within 1 month of microcosm set-up, H. fasciculare mycelium in single-species microcosms colonized approximately 25% of the total area of the microcosm, exploring soil within 3–6 cm around the wood inoculum. Phanerochaete velutina mycelium colonized an area with a radius of 7 cm, covering approximately 35% of the microcosm. Fungal growth did not differ substantially among the replicates. Mycelium close to wood inocula (older than 10 days) was always denser in both species than in the growing younger zones, but colony morphology differed between species. Phanerochaete velutina formed mycelia with long, branching cords while H. fasciculare formed thick, dense mycelial fans (Fig. 1).

In the interaction microcosms collected approximately 10 days after the first contact of mycelia, fungi formed mycelia of similar size, showing no inhibition of growth before contact. Phanerochaete velutina colonized approximately 64% of the microcosm area, 22% of the area was colonized by H. fasciculare and 24% was not colonized. About 10% of the soil microcosms were colonized by mycelium of both fungi and represented the interaction zone (Fig. 1). Evidence that P. velutina is the better combatant was provided by lower extension of H. fasciculare in the area of interaction and the presence of P. velutina cords growing over and replacing H. fasciculare mycelia. Within the next 20 days, the mycelium of H. fasciculare stopped growing and P. velutina mycelium colonized the entire microcosm including the inoculum wood block of H. fasciculare, whose mycelium became largely disintegrated (data not shown).

Soil in the microcosms exhibited a high spatial variation in enzyme activities (Fig. 1). Microcosm regions colonized by the saprotrophic basidiomycetes exhibited higher activities of α-glucosidase, β-glucosidase, cellobiohydrolase, β-xylosidase, N-acetylglucosaminidase and phosphodiesterase, and the activity of phosphomonoesterase was significantly (P<0.05) higher only under old mycelium. The activities of α-glucosidase, cellobiohydrolase and N-acetylglucosaminidase were especially low in the control soil and only 6–19% of the microcosm mean. The enzyme activities in soil areas covered by the older fungal mycelium were generally higher than in the soil covered by the young mycelium, but, due to the high level of spatial variation, significant differences (P≤0.05) were only recorded for β-glucosidase and β-xylosidase in H. fasciculare and phosphomonoesterase in P. velutina (Fig. 1, Table 1). The activity of phosphomonoesterase, N-acetylglucosaminidase, and especially α-glucosidase, was significantly (P≤0.05) increased in the interaction zone compared with the single-fungus-colonized soil. As an exception, the activity of arylsulphatase did not exhibit any significant treatment effect.

Table 1.  Activity of extracellular enzymes, mycelial cover and ergosterol content in different regions of soil microcosms colonized by Hypholoma fasciculare, Phanerochaete velutina or by both fungi interacting
Regionα-Glucosidase
(% of mean)
Cellobio-
hydrolase
(% of mean)
β-Xylosidase
(% of mean
Phospho-
monoesterase
(% of mean)
Phospho-
diesterase
(% of mean)
Mycelial
cover (%)
Ergosterol
(μg g−1)
  1. The data represent means and standard error of the mean.

  2. Different letters indicate statistically significant differences among regions at P<0.05.

Control6.4 ± 2.7 c11.3 ± 4.5 c42.7 ± 4.8 c34.3 ± 1.5 c61.4 ± 2.7 c6.88 ± 0.76 d3.43 ± 0.25 ab
H. fasciculare (young)67.2 ± 14.7 b66.9 ± 18.9 ab83.6 ± 8.3 b120.0 ± 15.7 b117.3 ± 8.2 a80.41 ± 2.34 b4.02 ± 0.30 a
H. fasciculare (old)100.5 ± 35.7 b146.6 ± 27.4 ab155.4 ± 13.6 a95.0 ± 11.4 b104.8 ± 4.1 a93.73 ± 0.86 a3.80 ± 0.35 ab
P. velutina (young)105.2 ± 29.7 b52.4 ± 14.8 ab67.2 ± 5.2 b73.5 ± 4.1 bc81.4 ± 4.1 b37.71 ± 2.16 c3.31 ± 0.31 ab
P. velutina (old)69.3 ± 12.8 b102.7 ± 25.7 b130.2 ± 26.1 ab99.0 ± 7.4 b109.6 ± 8.5 a81.01 ± 1.83 b2.99 ± 0.30 b
Interaction zone271.6 ± 66.9 a219.5 ± 91.1 a113.9 ± 16.3 ab184.7 ± 22.5 a130.8 ± 7.5 a92.63 ± 1.50 a3.61 ± 0.37 ab

To reduce the effects of a large variation in enzyme activities in microcosms, the relative composition of the extracellular enzyme pool was used to identify the enzymes preferentially produced in different soil regions. In the biplot of the two principal components that explained 36% and 22% of the total variability, soil samples clustered generally by soil regions, the samples from the control soil and the interaction zone being separated the most (Fig. 2). anova, followed by the Fisher LSD post hoc test showed significant (P≤0.05) separation of samples from the control soil from all the other treatments, and also of samples from the interaction zone from all the other treatments, except the old H. fasciculare mycelium. In the latter case, the samples were significantly (P≤0.05) separated along the second component axis. While the interaction zone exhibited high relative activities of N-acetylglucosaminidase, α-glucosidase and cellobiohydrolase, the control soil was richer in β-glucosidase, β-xylosidase and phosphodiesterase. There was no significant difference (P>0.05) in enzyme production by the two tested basidiomycetes, but, interestingly, samples of soil colonized by old and young H. fasciculare mycelium differed in the enzyme composition.

Figure 2.

 Principal component analysis of the relative activity of eight extracellular enzymes in soil microcosms with interacting mycelia of Hypholoma fasciculare and Phanerochaete velutina. Localization of individual soil samples in the PCA plot (a) and contributions of individual enzymes to sample localization (b).

Fungal and bacterial biomass in soil with interacting fungal mycelia

While the soil in which mycelia of introduced saprotrophic basidiomycetes were absent showed only scarce and low coverage by mycelia of resident soil fungi (<7%), the regions colonized by P. velutina and H. fasciculare and the interaction zones had high mycelial density (Table 1). Mycelial coverage differed between the two species: coverage was much less for P. velutina than for H. fasciculare and coverage in old and young regions was similar in H. fasciculare, but less dense in the older regions of P. velutina than in the younger regions. While image analysis considered only surface fungal mycelia predominantly produced by the introduced basidiomycetes, ergosterol and DNA were extracted from soil samples (which included surface mycelium). Ergosterol analysis indicated that the control soil had similar amounts of fungal biomass as soil covered by H. fasciculare or P. velutina (Table 1).

qPCR analyses of rRNA gene abundance in soil regions demonstrated a high variation in the content of fungal DNA within samples, but confirmed that there was no significant (P>0.05) difference in the fungal rRNA abundance among the control regions and regions whose surface was colonized by basidiomycetes (Fig. 3). The use of species-specific qPCR probes for H. fasciculare and P. velutina enabled approximation of the contribution of DNA of introduced basidiomycetes to total soil fungal DNA. The proportion of H. fasciculare DNA in total fungal DNA decreased with the age of mycelium, while the proportion of P. velutina DNA was higher in regions covered by old mycelia, and most of the fungal DNA in these soil regions belonged to this introduced species. In the interaction zones, most DNA belonged to P. velutina (Fig. 3), which is consistent with overgrowth of H. fasciculare mycelium and its replacement by P. velutina.

Figure 3.

 Quantification of fungal and bacterial DNA and the DNA of the introduced saprotrophic basidiomycetes in soil microcosms with interacting mycelia of Hypholoma fasciculare and Phanerochaete velutina based on qPCR analysis. The bar graphs show the medians (line), interquartile ranges (end of boxes), means (squares) and 90% percentiles (whiskers). Different letters above plots indicate statistically significant (P≤0.05) differences among regions. The data are based on rRNA abundance g−1 soil dry mass, except for the basidiomycetes/fungi panel, where the ratio is based on ng species-specific DNA ng−1 total DNA to correct for the difference in rRNA copy number ng−1 genomic DNA in H. fasciculare and P. velutina. I, interaction zone; Ao, old mycelium of H. fasciculare; Bo, old mycelium of P. velutina; A, young mycelium of H. fasciculare; B, young mycelium of P. velutina; C, no colonization by saprotrophic basidiomycetes.

Individual regions of soil exhibited median values of bacterial rRNA abundance between 25 × 106 and 250 × 106 g−1 soil. Compared with the control soil, the proportion of bacterial DNA in total DNA was significantly (P≤0.05) lower under the old H. fasciculare mycelium and in the zone of mycelial interaction (Fig. 3). The ratio of fungal to bacterial DNA was higher in soil colonized by H. fasciculare than in all other treatments (Fig. 3). The abundance of fungal mycelium in soil did not affect bacterial abundance negatively: bacterial rRNA abundance significantly increased with that of fungi (P≤0.001). Interestingly, no significant (P>0.05) correlations were found among enzyme activities and total fungal biomass expressed as ergosterol content. This may reflect a variation in the ergosterol content with age and functional status of fungal mycelia; the importance of fungi in the production of certain extracellular enzymes was shown by significant increases in the activities of several enzymes with the soil fungal : bacterial (F : B) DNA ratio. This was observed for β-xylosidase (P<0.001), β-glucosidase (P=0.002), phosphomonoesterase (P=0.024) and cellobiohydrolase (P=0.025). In single-species microcosms, where mycelia can be assigned reliably to each saprotrophic basidiomycete, the activity of several enzymes tended to increase with the mycelial density (Fig. 4), but a significant (P≤0.05) correlation was only detected for N-acetylglucosaminidase. Arylsulphatase significantly (P≤0.05) decreased with increasing mycelial cover of P. velutina, suggesting that this fungus might negatively affect the microbial producers or arylsulphatase.

Figure 4.

 Correlations between the activities of N-acetylglucosaminidase (a) and arylsulphatase (b) with mycelial coverage in soil microcosms colonized by Hypholoma fasciculare or Phanerochaete velutina. The analysis only considered samples where the saprotrophic basidiomycetes were present. Linear regression fit, F and P values for one-way anova are included.

Discussion

Saprotrophic cord-forming basidiomycetes are typical inhabitants of forest soil, preferentially colonizing the litter/soil organic horizon interface (Boddy, 1999). They actively contribute to organic matter decomposition due to the ability to produce a large variety of extracellular enzymes (Osono, 2007; Baldrian, 2008), and their presence in soils might therefore be one of the likely causes of locally increased enzyme titres. In the present study, where the growth of saprotrophic cord-forming basidiomycetes on the soil surface was simulated in a microcosm, the presence of both H. fasciculare and P. velutina resulted in the elevated activity of seven of eight hydrolytic enzymes involved in the C, N and P cycles (Lindahl & Finlay, 2006). Higher activities of several enzymes were recorded under the old mycelium rather than in the peripheral zone of the mycelial colony. It should also be noted that the distribution of enzymes under fungal mycelia was not homogeneous, showing patches of high or low activities in certain parts of the colony (Fig. 1). Surprisingly, enzyme activities under H. fasciculare and P. velutina mycelia were similar, even though the explorative mycelium of P. velutina was much less dense than the exploitative mycelial cover of H. fasciculare. Hypothesis 2 was therefore rejected. Moreover, only N-acetylglucosaminidase activity increased significantly with mycelial cover, showing that enzyme activity does not solely reflect fungal biomass. Nevertheless, the results provided evidence for hypothesis 1 that enzyme activity will increase under mycelia.

Interaction between mycelia of different species results in changes in mycelial morphology and physiology in the interaction zone and elsewhere (Boddy, 2000). The production of dissolved (Heilmann-Clausen & Boddy, 2005) and volatile organic compounds (Evans et al. 2008) and increases in the production of reactive hydroxyl radicals (Tornberg & Olsson, 2002) and oxidative enzymes (Hiscox et al. 2010) are all involved in these combative interactions. In laboratory studies, increased activities of laccase and Mn-peroxidase were recorded in the mycelial interaction zones involving basidiomycetes (Baldrian, 2004; Ferreira Gregorio et al., 2006; Chi et al., 2007; Hiscox et al., 2010). The former enzyme was previously shown to be involved in fungal defence by cell wall melanization (Baldrian, 2006). Here, we report for the first time that the interaction zone between basidiomycete mycelia of P. velutina and H. fasciculare, the respective winner and loser of this interspecific combat, is characterized by elevated activities of phosphatase, N-acetylglucosaminidase and α-glucosidase in soil, thereby providing evidence for hypothesis 3. N-acetylglucosaminidase as a chitinolytic enzyme is likely involved in the attack of the cell wall of the interacting fungi. Activities of chitinase have been shown to increase during the secondary colonization of wood blocks precolonized by another fungus compared with noncolonized wood (Lindahl & Finlay, 2006). In addition to the acquisition of N, N-acetylglucosaminidase can thus induce the damage of fungal hyphae, ultimately leading to the leakage of cell content in the interaction zone (Wells & Boddy, 2002). The increased activity of α-glucosidase might be a reflection of the increased utilization of glycogen as an energy storage resource either to support competing mycelium or as a result of its leakage from the damaged hyphae of the defeated fungus. The higher activity of phosphomonoesterase might indicate the increased availability of phosphates in the interaction zone after hyphal damage.

Previous studies of saprotrophic basidiomycetes in wood showed that fungi can selectively affect the composition of microbial communities (Tornberg et al., 2003; Folman et al., 2008; Valášková et al., 2009; de Boer et al., 2010). White-rot fungi produce organic acids, which act as chelating agents in the formation of reactive radicals during lignin decomposition, but also ensure high efficiency of polysaccharide hydrolysis by acidification of the environment. Indeed, wood colonized by H. fasciculare is strongly acidic, with pH ranging from 3.6 to 4.3 (Valášková et al., 2009; de Boer et al., 2010), and the change of pH in soil below fungal mycelia of H. fasciculare and P. velutina was also recorded in this experiment, where pH decreased from 5.5 in the control soil to 4.6–5.0 in the fungus-colonized area and the interaction zone. Although H. fasciculare has the potential to produce several toxic secondary metabolites and reactive oxidative compounds, no evidence was found for their involvement in the reduction of the bacterial counts in fungus-colonized wood (de Boer et al., 2010). It seems that rapid acidification may be the main factor reducing wood-inhabiting bacteria upon colonization by H. fasciculare (de Boer et al., 2010). In addition, a shift to the increased proportion of low pH-tolerant and slow-growing Proteobacteria and Acidobacteria representing more than 90% of the total bacterial biomass was recorded (Folman et al., 2008; Valášková et al., 2009), which might efficiently reduce the bacterial acquisition of available nutrients released by the fungus.

While the slight acidification of soil might affect microbial communities in basidiomycete-colonized soil, due to the extracellular nature of enzymes hydrolysing cell wall biopolymers, their higher activity increases the availability of small molecular mass nutrients for both their producers and other soil microorganisms. This was reflected here by the increased activity of β-xylosidase, β-glucosidase, phosphomonoesterase and cellobiohydrolase with increasing F : B ratio and the positive correlation of bacterial and fungal biomass. The positive effect of nutrient release by fungi was increased in the interaction zone by the release of nutrients from damaged hyphae, but counteracted by the expected production of hydroxyl radicals and the activity of oxidative enzymes. The present study clearly shows that the quantitative effects of basidiomycete colonization and interactions on microbial communities in soils are rather limited. The fungal biomass content was not elevated in the basidiomycete-colonized soil and, while the P. velutina biomass represented a significant proportion of the total fungal biomass in the soil colonized by its mycelium, H. fasciculare represented only a minor fraction of the total fungal community even in the densely colonized soil. Higher F : B biomass ratio and reduction of bacterial biomass were, however, recorded under the mycelium of this fungus, showing that the effects of saprotrophic basidiomycetes on bacteria are likely species specific.

Acknowledgements

This work was supported by the Academy of Sciences of the Czech Republic (M200200901), the Ministry of Education, Youth and Sports of the Czech Republic (LC06066, OC10064) and the Institutional Research Concept of the Institute of Microbiology of the ASCR, v.v.i. (AV0Z50200510).

Ancillary