Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Distinctive Feature of Microbial Communities and Bacterial Functional Profiles in Tricholoma matsutake Dominant Soil

Distinctive Feature of Microbial Communities and Bacterial Functional Profiles in Tricholoma matsutake Dominant Soil

  • Seung-Yoon Oh, 
  • Jonathan J. Fong, 
  • Myung Soo Park, 
  • Young Woon Lim
PLOS
x

Abstract

Tricholoma matsutake, the pine mushroom, is a valuable forest product with high economic value in Asia, and plays an important ecological role as an ectomycorrhizal fungus. Around the host tree, T. matsutake hyphae generate a distinctive soil aggregating environment called a fairy ring, where fruiting bodies form. Because T. matsutake hyphae dominate the soil near the fairy ring, this species has the potential to influence the microbial community. To explore the influence of T. matsutake on the microbial communities, we compared the microbial community and predicted bacterial function between two different soil types—T. matsutake dominant and T. matsutake minor. DNA sequence analyses showed that fungal and bacterial diversity were lower in the T. matsutake dominant soil compared to T. matsutake minor soil. Some microbial taxa were significantly more common in the T. matsutake dominant soil across geographic locations, many of which were previously identified as mycophillic or mycorrhiza helper bacteria. Between the two soil types, the predicted bacterial functional profiles (using PICRUSt) had significantly distinct KEGG modules. Modules for amino acid uptake, carbohydrate metabolism, and the type III secretion system were higher in the T. matsutake dominant soil than in the T. matsutake minor soil. Overall, similar microbial diversity, community structure, and bacterial functional profiles of the T. matsutake dominant soil across geographic locations suggest that T. matsutake may generate a dominance effect.

Introduction

The pine mushroom, Tricholoma matsutake (S. Ito and S. Imai) Singer, is an important ectomycorrhizal fungus in Asia due to its value as a food product [1, 2]. Fruiting bodies of T. matsutake grow in a fairy ring (also called a shiro) where hyphae aggregate densely with soil in the shape of ring around ectomycorrhizal host trees (Pinaceae and Fagaceae) [1, 3]. In hopes of successful cultivation, previous research focused on abiotic factors such as climate [46] and nutritional condition [79] for understanding the ecology of T. matsutake. However, studies on the biotic interaction of T. matsutake, especially microbial interaction, are limited.

Microbial interaction may be limited due to the dominance effect of T. matsutake in the fairy ring. Ohara and Hamada [10], using a culture-dependent method, found no bacteria in the fairy ring. Subsequent studies using culture-dependent [11], traditional molecular [12, 13], and next generation sequencing (NGS) based methods [14] corroborated the results of Ohara and Hamada [10], finding lower microbial diversity or abundance within the fairy ring of T. matsutake. In contrast, Kim et al. [15] found no difference in bacterial diversity between the fairy ring and non-fairy ring soil, with both having high diversity.

Together with reduced diversity, the dominance of T. matsutake may shape a distinct microbial community. Although previous studies found specific microbes present in the fairy ring of T. matsutake [1113, 16], some studies used culture-dependent methods, which may have underestimated microbe diversity. Culture-independent and high-throughput technology (e.g. NGS) is a useful tool to fully understand microbial characteristics in T. matsutake dominant soil. Preliminary studies of microbial communities in the fairy ring demonstrated the usefulness of NGS technology [14, 15], and the next step is to extend analyses to include multiple samples across geographic locations to verify trends.

Although NGS has revealed high microbial diversity in ecosystems [17], functional analyses are needed to understand the role of microbes in the ecosystem. While metagenomics, metatranscriptomics, and functional arrays are good approaches for functional analysis [18, 19], these methods are constrained by cost and difficulty [20, 21]. Recently, Langille et al. [22] developed the software Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt), which analyzes microbial function by combining phylogenetic analyses and genome database data to predict functional profiles. To date, PICRUSt has contributed to our understanding of microbial functional profiles in the animal gut [23], human mouth [24], ocean [25], and soil [26].

Our main objective was to explore the microbial characteristics associated with T. matsutake by comparing the diversity and community structure of bacteria and fungi, as well as bacterial functional profiles in T. matsutake dominant (Tm-dominant) soil and minor (Tm-minor) soil near the fairy ring. A dominance effect of T. matsutake would be implied if microbial characteristics were similar across multiple Tm-dominant soils, but different from the Tm-minor soil. We used pyrosequencing because it has proven to be a powerful method for revealing microbial interactions [27, 28], and expanded on the design of past studies by comparing multiple soil types from two geographic locations. Additionally, we used PICRUSt to predict and compare bacterial functional profiles of soil bacteria in the Tm-dominant soil. We expect this study to contribute to our understanding of the ecology of T. matsutake and microbial interactions in the natural environment.

Materials and Methods

Sampling location and strategy

In 2011, sampling was conducted during the fall, when T. matsutake produced fruiting bodies. Samples were collected from two locations in South Korea that are separated by ~220 km: Gyeongju (35° 48' 46"N, 129° 13' 48"E) and Hongcheon (37° 41' 35"N, 127° 58' 51"E). Sampling sites in Gyeongju and Hongcheon are research forests of government institutes (Gyeongsanbuk-do Forest and Environment Research Institute, and National Institute of Forest Science, respectively). Approval for sampling was granted by each institute. Both locations consisted of mixed forest dominated by the Japanese red pine (Pinus densiflora), oak trees (Quercus spp.), and ericaceous shrubs (Rhododendron spp.). For collecting the Tm-dominant soil samples, we conducted sampling near the fairy ring. Generally, abundance of T. matsutake is high in the fairy ring with whitish-gray colored soil and hyphae aggregations. We took three soil samples at 30 cm intervals near the fairy ring, and conducted sampling twice for each geographical location. In total, 12 soil samples taken from Hongcheon and Gyeongju were transported on ice and stored at -4°C for a maximum of one week before DNA extraction.

It has been shown that morphological characterization of the fairy ring may not correspond to soil dominated by T. matsutake if multiple fungal species co-exist [29]. Therefore, we confirmed soil type based on the sequence read abundance. We categorized soil samples into two types based on the proportion of T. matsutake sequences in the NGS analysis: Tm-dominant soil (≥50%) and Tm-minor soil (<50%). We added a status code (Tm-dominant soil: D, Tm-minor soil: m) to end of the sample name to denote its classification after sequence based confirmation (S1 and S2 Tables).

DNA extraction, PCR amplification, and Pyrosequencing

For each sample, genomic DNA was extracted from 500 mg of soil using the FastDNATM SPIN kit for soil (MP Biomedicals, USA) following the manufacturer’s instructions. For fungi, the nuclear ribosomal internal transcribed spacer (ITS) region was amplified using ITS1F and ITS4 primers [30], while for bacteria, the 16S ribosomal DNA (16S) region was amplified using 27F [31] and 518R primers [32]. Amplicon libraries were prepared using primers with a 454 pyrosequencing adaptor and multiple identifier (MID) tag. The PCR program was as follows: 5 min at 94°C; 35 cycles of 30 s at 94°C, 30 s at 55°C, and 40 s at 72°C; and a final extension step of 5 min at 72°C. PCR products were confirmed using gel electrophoresis and purified using the HighPureTM PCR Product Purification Kit (Roche, Germany). Individual PCR products were quantified using a NanoDrop spectrophotometer (Thermo, USA) and separately pooled for bacteria and fungi. Pyrosequencing was performed with a 454 GS Junior platform (Roche, USA) at ChunLab (Seoul, Korea). DNA was sequenced only in the reverse direction, and data were deposited in the NCBI Sequence Read Archive (SRP046049).

Bioinformatics process and statistical analysis

Raw data were processed with QIIME 1.8.0 [33]. Reads having sequences shorter than 200 bp, an average nucleotide quality score under 25, ambiguous nucleotides, or mismatches in the MID tag were excluded from analyses. The remaining reads had MID tags removed and were denoised using Denoiser v1.7.0 [34] implemented in QIIME. Different workflows were used for the fungal and bacterial dataset; the bacterial workflow was chosen so that we could perform a functional prediction analysis. For the fungal dataset, chimeric sequences were removed and fungal operational taxonomic units (fOTUs) were clustered with 97% similarity using the USEARCH v5.2.236 [35] implemented in QIIME. Universal singleton fOTUs in total of samples were discarded from analyses. Taxonomy was assigned to fOTUs using the RDP classifier [36] to query the representative sequence (the most abundant sequence) against the UNITE database v. 7.0 [37]. For the bacterial dataset, we clustered bacterial OTUs (bOTUs) by using the open reference OTU picking process in UCLUST [35] with 97% similarity based on the Greengenes database (version May 2013) [38]. The final identification of bOTUs was assigned using the RDP classifier with an 80% confidence threshold based on the EzTaxon database [39]. The representative sequences of bOTUs were aligned to the Greengenes core alignment dataset using PyNAST [40]. Chimeric sequences were detected and removed using ChimeraSlayer [41]. Columns of data with only gaps were filtered from the alignment and a phylogeny was constructed using FastTree [42]. As for fOTU, universal singleton bOTUs were removed from analysis.

Before further analysis, we normalized sequences using the minimum number of sequences. Alpha diversity indices (Chao1, Shannon, and Shannon’s equitability) and taxonomic composition were calculated in QIIME. Before performing calculations for fungi, we removed sequences of T. matsutake to eliminate the estimation bias from the T. matsutake abundance. In addition to the other diversity indices, phylogenetic diversity index (PD) [43] was calculated for bacteria. Statistical analyses (ANOVA and Wilcoxon rank sum test) and graphical plots (boxplots and principal coordinates analysis [PCoA]) were performed using R v. 3.1.0 [44]. We used ANOVA for comparing diversity indices of microbial communities and a Wilcoxon rank sum test for comparing the OTU richness at the phylum level. PCoA plots were drawn using the phyloseq package with Bray-Curtis distances for fungi and weighted UniFrac distances for bacteria [45]. We tested the influence of two factors: soil type (the Tm-dominant soil and Tm-minor soil) and geography (Hongcheon and Gyeongju). To determine the explanatory coverage of the factors, we performed Constrained Analysis of Principal coordinates (CAP) analyses for CAP models specified in two ways: both factors together (~soil type + geography) and constrained to only one factor. In models constrained to one factor, we eliminated the effect of the other factor by assigning a conditioned factor (~geography+ Condition[soil type] and ~soil type + Condition[geography]). To test for the statistical significance of each CAP model, we performed an ANOVA with 999 permutations. We tested for statistical difference of microbial communities between samples using PERMANOVA (adonis) with 999 permutations implemented in QIIME, and compared the distribution of genera between soil types using a discovery odds ratio test in the metagenomeSeq package [46]. P-values in the discovery odds ratio test were corrected using the false discovery rate (FDR) of Benjamini and Hochberg [47]. To avoid misinterpretation, we considered only significant results for common and major genera (present more than 1% in at least three samples of one of the soil types) that had sufficiently large changes (|log2 Odds ratio| >1).

Functional profiles of bacterial communities were predicted using PICRUSt [22]. Since fungal genome data are currently limited, we performed PICRUSt analyses only for bacterial communities. We followed the tutorial instructions on the PICRUSt website for our workflow. Briefly, we predicted gene family profiles and 16S copy number by constructing a phylogeny with representative bOTU and Greengenes sequences for which genome data were available, followed by an ancestral state reconstruction. Metagenome prediction of bacterial communities was conducted using the calculated dataset after normalizing for 16S copy number. We obtained Nearest Sequenced Taxon Index (NSTI) scores for evaluating the accuracy of predicted metagenome profiles [22], with a low NSTI indicating high accuracy. The functions of predicted metagenomes were categorized with the KEGG pathways database [48]. We compared bacterial functional profiles at KEGG modules levels 4. We used the edgeR package [49] for analyzing differential abundance of modules by soil type. After normalizing module abundance using the trimmed mean of M values (TMM) method in edgeR, we retained modules with more than 100 counts per million (CPM) in the 3 samples at least, and measured fold changes of modules between the Tm-dominant soil and Tm-minor soil. Lastly, an exact test was conducted with multiple test correction using the FDR of Benjamini and Hochberg. We retained statistically significant modules with large fold change (|log2 fold change| >4).

Results

A total of 237,107 fungal and 111,374 bacterial sequence reads were obtained, and after filtering 76,294 and 40,558 reads were retained respectively (S1 and S2 Tables). The proportions of T. matsutake reads in each sample were calculated for classification of soil types. In Hongcheon, T. matsutake reads were dominant (>80%) in H2 and H5, and almost nonexistent (0–1 reads) in the remaining four samples (H1, H3, H4, H6). In Gyeongju, three samples (G1, G2, and G3) were dominated by T. matsutake reads (>80%), but G4 and G6 (<2%), and G5 (34.7%) had a low proportion of T. matsutake. Therefore, we confirmed the samples to be five Tm-dominant soil (G1D, G2D, G3D, H2D, H5D) and seven Tm-minor soil samples (G4m, G5m, G6m, H1m, H3m, H4m, H6m) (S1 and S2 Tables).

Microbial diversity associated with T. matsutake

Fungal diversity.

A total of 76,294 reads and 187 fOTUs were detected. After removing T. matsutake reads, we normalized sequences to 570 reads (secondly smallest number) for further analysis. Because only 179 reads remained after removing T. matsutake reads, G1D sample was excluded the fungal community analysis. All Good's coverage estimates were 0.97–0.99, indicating that sequencing depth was appropriate for representing fungal diversity (S1 Table). The richness (Chao1) was significantly lower in the Tm-dominant soil than in the Tm-minor soil (Corrected P = 0.037), while both of diversity and evenness (Shannon) indices were not significantly different (Shannon: Corrected P = 0.579; Equitability: Corrected P = 0.779) (Fig 1A and S1 Table). The diversity and evenness showed significant differences for geographies (Shannon: Corrected P = 0.042; Equitability: Corrected P = 0. 042), while the richness did not (Corrected P = 0.063). No index showed a significant difference of interaction effects for soil type and geography (Chao1: Corrected P = 0.621; Shannon: Corrected P = 0.525; Equitability: Corrected P = 0. 525). Among fungal phyla, the fOTU richness of Basidiomycota was significantly lower in the Tm-dominant soil than the Tm-minor soil (Corrected P = 0.029) (Fig 2A). However, the fOTU richness in the other phyla did not have significant differences (Fig 2A).

thumbnail
Fig 1.

Diversity indices of Tm-dominant and Tm-minor soil samples for (A) fungi and (B) bacteria. All calculations were performed on normalized data. (*: Corrected P < 0.05, **: Corrected P < 0.01).

https://doi.org/10.1371/journal.pone.0168573.g001

thumbnail
Fig 2.

OTU richness at the phylum level in Tm-dominant and Tm-minor soil samples of (A) fungal and (B) bacterial communities. Significance of differential richness between the Tm-dominant and Tm-minor soil was evaluated using Wilcoxon rank sum test. The phyla that were significantly different between the soil types (Corrected P < 0.05) are represented with an asterisk.

https://doi.org/10.1371/journal.pone.0168573.g002

Bacterial diversity.

A total of 40,558 reads and 1,639 bOTUs were detected. The number of sequences was normalized to 1300 reads for further analysis. All samples had Good's coverage that indicated sufficient sequencing depth for characterizing bacterial diversity (0.94–0.99) (S2 Table). All indices of Chao1, Shannon diversity, Equitability, and phylogenetic diversity (PD) were significantly lower in the Tm-dominant soil than Tm-minor soil samples (Chao1: Corrected P = 0.026; Shannon: Corrected P = 0.005; Equitability: Corrected P = 0.014; PD: Corrected P = 0.005) (Fig 1B and S2 Table). Similarly, all indices were significantly different between geographies (Shannon: Corrected P = 0.017; Equitability: Corrected P = 0.033; PD: Corrected P = 0.019), except for Chao1 richness (Corrected P = 0.086). There were no significant interaction effects for soil type and geography. In comparing bOTU richness at the phyla level between soil types, Firmicutes was significantly higher in the Tm-dominant soil (Corrected P = 0.048), while Acidobacteria, Proteobacteria, and Verrucomicrobia were significantly higher in the Tm-minor soil (Corrected P: 0.048, 0.048, and 0.048, respectively) (Fig 2B).

Microbial community structure associated with T. matsutake

Fungal community.

In the ordination analysis, points clustered based on soil type and geography. CAP analyses found the combination of soil type and geography to have a significant effect (P = 0.004) and explained 42.0% of the clustering pattern (S1A Fig). Geography alone had a significant effect on clustering (P = 0.004, 31.8% explanatory power), while soil type alone did not (P = 0.159, 9.8% explanatory power) (Fig 3A and 3B). PERMANOVA results were in agreement, indicating that the samples were differed significantly according to geography (P = 0.003), while soil type did not (P = 0.271).

thumbnail
Fig 3. Constrained Analysis of Principal coordinates (CAP) plots for microbial communities in the Tm-dominant and Tm-minor soil.

CAP model for (A) fungal and (C) bacterial communities constrained by soil type and conditioned by geographic location. CAP models for (B) fungal and (D) bacterial communities constrained by geographic location and conditioned by soil type. CAP analyses were conducted on Bray-Curtis distances for fungal communities and weighted Unifrac distances for bacterial communities. Significance of CAP models was evaluated using ANOVA with 999 permutations.

https://doi.org/10.1371/journal.pone.0168573.g003

The results from the discovery odds ratio test indicated that some fungal genera had significantly different abundance patterns between soil types across geographies (Table 1 and S2A Fig). Among the major genera, Umbelopsis was significantly more abundant in the Tm-dominant soil than the Tm-minor soil. In contrast, Mortierella was significantly more abundant in the Tm-minor soil.

thumbnail
Table 1. Microbial genera list of significantly enriched in a discovery odds ratio test.

P-values were corrected to false discovery rates using the Benjamini and Hochberg method.

https://doi.org/10.1371/journal.pone.0168573.t001

Bacterial community.

Bacterial communities also clustered according to soil type and geography. In CAP analyses, the combination of soil type and geography showed a significant effect (P = 0.002) and explained 30.9% of the clustering pattern (S1B Fig). When analyzed separately for each factor, only soil type had a significant effect on clustering (P = 0.003, 20.2% explanatory power), while geography was not significant (P = 0.361) (Fig 3C and 3D). PERMANOVA tests supported this result, showing that bacterial communities were significantly different by soil type (P = 0.002), but was not by geography (P = 0.154).

The results of a discovery odds ratio test showed that some bacterial genera were differentially abundant between soil types (Table 1 and S2B Fig). Bacillus, Burkholderia, Mycobacterium, and Paenibacillus were significantly more abundant in the Tm-dominant soil, while Koribacter and Solibacter were significantly more abundant in the Tm-minor soil.

Predicted functional profiles of bacterial communities associated with T. matsutake

We predicted the functional profiles of bacterial communities with PICRUSt. Considering that we were studying novel environmental communities, the statistical relevance of prediction was relatively well supported (NSTI = 0.03–0.17). Functional profiles were significantly different between the soil types. Six level-4 KEGG modules were statistically more abundant in the Tm-dominant soil, while none of significantly abundant modules were detected in the Tm-minor soil (Table 2). For these six modules, three modules were a Type III secretion system (T3SS) and effectors (M00332, M00660, and M00542), two were amino acid uptake (M00225 and M00226), and one involved in carbohydrate metabolism (M00519) (Table 2).

thumbnail
Table 2. List of KEGG modules that were significantly abundant in the Tm-dominant soil.

Fold change is for Tm-dominant soil samples relative to Tm-minor soil samples. Counts per million (CPM) were calculated with normalized abundances using edgeR’s TMM method. P-values were corrected to false discovery rates using the Benjamini and Hochberg method.

https://doi.org/10.1371/journal.pone.0168573.t002

Discussion

Hyphal dominant environments (often named mycosphere, mycorrhizosphere, ectomycorrhizal mat, and fairy ring) influence both abiotic and biotic factors in the soil ecosystems [29, 5054]. Tricholoma matsutake forms its hyphal dominant environment around the fairy ring. Although the hyphal dominant environment can influence microbial communities [55], microbial interaction and functional activity in the Tm-dominant soil remain poorly characterized. In the work reported here, we compared soil communities between soil types (Tm-dominant and Tm-minor soil) across geographic locations. Our results showed an influence of T. matsutake dominance on the microbial diversity and community structure of other soil microbes as well as predicted bacterial functional profiles.

Microbial diversity in the Tm-dominant soil

Consistent with most of previous studies (e.g., Ohara and Hamada [10]), we found that both fungal and bacterial diversity were significantly lower in the Tm-dominant soil than in the Tm-minor soil. However, this is in contrast to Kim et al. [15], who used a similar NGS approach as us, but found equal levels of bacterial diversity between the fairy ring and non-fairy ring zones in Korea. While the differences between our results and those of Kim et al. [15] could reflect a difference between geographic location, we consider this unlikely, given that previous studies in Japan [10, 13] also found lower bacterial diversity in the fairy ring. A more likely explanation is that the results of Kim et al. [15] reflect different sampling methods. Kim et al. [15] combined multiple samples from each of the three zones (inside, beneath, and outside a fairy ring). In our experience, morphological soil classification may be inconsistent with the dominance of T. matsutake; Kim et al. [15] may have combined samples from different soil types, which would have obscured the pattern of lower bacterial richness in the fairy ring.

As noted above, the low fungal richness in the Tm-dominant soil we found is consistent with previous studies [11, 14]. Because we excluded T. matsutake when calculating diversity indices, the dominance effect of low fungal richness is not simply a result of T. matsutake abundance. Although the Shannon’s diversity and evenness of fungal community were not significantly different between the Tm-dominant and Tm-minor soil, the dominance effect of T. matsutake likely reduced fungal richness. At the phylum level, Basidiomycota showed significant differences in fOTU richness between the soil types (Fig 2A). Considering that a large proportion of ectomycorrhizal fungi belong to Basidiomycota [56], the reduction of Basidiomycota richness may reflect exclusion by T. matsutake. Similar to other ectomycorrhizal fungi [57, 58], T. matsutake can secrete antifungal compounds [59] to exclude other fungal species, promoting its own fitness by reducing competitors.

Our results are consistent with previous studies showing that bacterial diversity is reduced in Tm-dominant soil across geographical locations [10, 13]. This result is also consistent with mycosphere studies under the fruiting body of another ectomycorrhizal fungus (Laccaria proxima) [54, 60]. The major difference we found between the Tm-dominant soil and Tm-minor soil was a reduction of the bOTU richness of the phyla Acidobacteria and Proteobacteria (Fig 2B). These phyla are abundant in soil environments [61, 62], but reduced richness in the Tm-dominant soil could be the result of a negative effect of T. matsutake, such as competition for resources or secretion of antibiotics to exclude bacteria [1]. In contrast, the bOTU richness of Firmicutes was higher in Tm-dominant soil. Firmicutes contains the genera Bacillus, Cohnella, and Paenibacillus, which are commonly found in mycorrhiza-associated environments [6365]. Therefore, higher richness of Firmicutes may suggest a conserved trait associated with fungi-bacteria interaction or resistance to antibiotics. The interrelationship between these genera and T. matsutake is discussed below.

As shown in the Tm-dominant soil, low microbial diversity may be a general feature in hyphal dominant environments. In the mycosphere of Laccaria, bacterial diversity was significantly reduced compared to the bulk soil [66]. On the other hand, fungal diversity was also relatively low in the mycorrhizospheres of Amanita and Laccaria [67]. This trend is not limited to ectomycorrhizal fungi. In the mycosphere soil of several saprophytic mushrooms (Lachnocladiaceae, Lepiotaceae, and Marasmiaceae), pyrosequencing detected reduced microbial diversity [68]. However, some fungi showed higher or unchanged diversity in the mycosphere or mycorrhizosphere soil [67]. Therefore, reduction of microbial diversity is likely a species-specific feature among fungi such as T. matsutake. This alteration of microbial diversity between soil types suggests that the hyphae of T. matsutake can shape distinctive environments different from the adjacent soil. In addition, this environment can be harsh (as in those of Basidiomycota, Acidobacteria, Proteobactera, and Verrucomicrobia) or favorable (as in those of Firmicutes), depending on the taxa, which suggest that this is a conserved trait.

Distinct microbial community structure

If the microbial community associated with the Tm-dominant soil is distinct and shared across geographic locations, the dominance effect of T. matsutake may cause spatially distinct microbial communities to become more similar. CAP and PERMANNOVA analyses showed different trends in fungal and bacterial communities. Bacterial communities were significantly different between the Tm-dominant soil and Tm-minor soil, and shared community structure within soil types, while fungal community was not (Fig 3A and 3C). This suggests that dominance of T. matsutake may have stronger influence on bacterial communities than on fungal communities. If T. matsutake has a negative effect on microbial diversity, genera that were enriched in the Tm-dominant soil may have a specific relationship with T. matsutake.

Although the fungal community structure was not different between soil types, Umbelopsis was significantly more abundant in Tm-dominant soil samples across geographic locations (Table 1 and S2A Fig). Umbelopsis was frequently detected from the fruiting body and the fairy ring of T. matsutake [13, 69, 70]. Because Umbelopsis species formerly belonged to Mortierella, Umbelopsis isolated from the fairy ring was identified as Mortierella [71]. However, our result showed that Mortierella and Umbelopsis have opposite distribution trends associated with T. matsutake. Umbelopsis may have positive interactions with T. matsutake. In addition, Umbelopsis is likely a helper fungus that promotes growth of T. matsutake; a previous study showed that metabolites from “Mortierellanana (= Umbelopsis nana) enhanced hyphal growth of T. matsutake [72].

Bacterial communities also showed distinct community structures in the different soil types (Fig 3C). Bacillus, Burkholderia, Mycobacterium, and Paenibacillus were statistically more abundant in the Tm-dominant soil than the Tm-minor soil (Table 1 and S2B Fig). Most of these genera were previously detected in the fairy ring of T. matsutake [13, 15], and are commonly found in the rhizosphere or mycorrhizosphere [6365, 73]. In addition, these taxa may be similar on a functional level. For example, Bacillus, Burkholderia, and Paenibacillus are known to be mycorrhiza helper bacteria (MHB) that promote the growth and colonization of mycorrhizae [7476]. Co-occurrence suggests that T. matsutake may recruit mutualistic microbes as part of its dominance effect. On the other hand, these bacteria actively orientate to Tm-dominant soil for exploiting nutrient from T. matsutake. Diverse nutrients are exuded from hyphae (e.g. carbon hydrates from plant) [54], and hyphae can be a nutrient source itself for the mycophagous bacteria [77]. Therefore, bacteria abundant in the Tm-dominant soil are either recruited by T. matsutake, gather by themselves, or a combination of both.

Distinct microbial communities and enrichment of specific microbes in the mycorrhizosphere and mycosphere have been reported [66, 78]. Among the bacteria enriched in the hyphal dominant environment, Burkholderia, Pseudomonas, and Variovorax associated with various fungi [54, 66, 7981], thus, these bacteria can be classified as fungiphilic bacteria [66]. Enrichment of fungiphiles in the hyphal dominant environment is thought to be connected with specific traits. For example, fungiphiles can use surface of hyphae to move on like a highway [82, 83], and have T3SS mediating interactions with fungi [60]. Moreover, fungiphiles can use the material excreted from the hyphae (e.g. carbohydrates [maltose and trehalose] and amino acids [alaninamine and ornithine]) as an energy source [66]. Fungal communities are also different in the hyphal dominant environment [80, 84]. In addition, Sabella et al. [85] showed the potential of mycorrhizal helper fungus (Arthrinium phaeospermum for Tuber) like MHB. Therefore, mutualistic fungi may co-exist in the hyphal dominant environment, and our results highlighted Umbelopsis as a likely mutualistic fungus.

Functional association of distinct bacterial community

We analyzed whether the bacterial community of the Tm-dominant soil produces a distinct functional profile in the soil. If true, this would link T. matsutake to specific functions of the soil microbiota. The low NSTI value (0.03–0.17) from our PICRUSt analysis indicated a reasonable accuracy of prediction. Our results indicated that Tm-dominant soil and Tm-minor soil were functionally distinct. The six functional modules found to be statistically higher in the Tm-dominant soil than in the Tm-minor soil can be organized into two groups based on their putative functions: nutrient uptake (amino acid and carbohydrate) and type III secretion systems (T3SS). These functions can be associated with microbial interaction and exploitation by increasing bacterial survival and mediating interactions between bacteria and T. matsutake. For the first effect, improved bacterial survival can be facilitated by uptake of amino acids (arginine, histidine, lysine, and ornithine) and carbohydrates (Table 2). As mentioned above, an increase of nutrient metabolism associated modules suggests that these bacteria favor Tm-dominant soil because bacteria in the Tm-dominant soil may more easily acquire nutrients.

The second effect, mediation of T. matsutake-bacteria interactions, could be enhanced by T3SS. While T3SS is a well known system for mediating bacterial interactions with various eukaryotes [86], T3SS is also crucial for fungi-bacteria interactions [87]. Studies of bacteria associated with mycorrhizal fungi found increased numbers of T3SS genes in both the mycorrhizosphere and mycosphere [60, 88]. Additionally, T3SS of Pseudomonas fluorescens has an MHB effect, enhancing growth of Laccaria bicolor [89].

By characterizing functional profiles of bacteria in the Tm-dominant soil, we identified several functional modules that were potentially associated with microbial interactions with T. matsutake. Increased functions associated with nutrient uptake suggest that microbial interaction in the Tm-dominant soil may improve bacterial fitness. However, the benefit for T. matsutake remains unknown. Although there are limits to the interpretation of functional profile predictions, we have identified functions with potential positive effects. Future studies can target these functions to clarify the dynamics between organisms in the Tm-dominant soil.

In contrast to our results, the studies on the functional traits in the hyphal dominant environment showed an increase in traits not only for bacterial fitness, but also for the hyphal host's fitness. This may be explained by previous studies. For examples, Pseudomonas fluorescens promotes hyphal growth [90] and ectomycorrhizal formation of Laccaria bicolor [91], and T3SS (enriched modules in the Tm-dominant soil) is essential component to promoting effect [89]. Moreover, it is possible to help fungi indirectly by improving fungi-plant interaction. For example, transferring nitrogen and phosphate is important in ectomycorrhiza-plant relationships. Bacteria that can carry out mineral weathering can enhance ectomycorrhizal capability [92, 93]. Although culture-dependent and gene targeted studies showed distinct functional traits in the hyphal dominant environment, there are limited studies of functional traits using metagenomic or metatranscriptomic approaches because of cost and complexity of soil environment [20, 21]. As we have shown, predicted functional profiles using PICRUSt can be an intermediate method using targeted amplicon datasets for glancing overall functional profiles in metagenomic samples.

Conclusions

We compared Tm-dominant soil and Tm-minor soil samples across geographic locations in order to investigate possible dominance effects of T. matsutake. The similarity of microbial communities and bacterial functional profiles in the Tm-dominant soil suggests that this fungus may have a dominance effect on soil microorganisms. We identified several dominant fungal (Umbelopsis) and bacteria genera (Bacillus, Burkholderia, Mycobacterium, and Paenibacillus) in the Tm-dominant soil that may exploit or help T. matsutake. Regarding functional analyses, we identified functional modules in the Tm-dominant soil that are associated with nutrient metabolism for bacteria and bacterial-fungal interactions. The Tm-dominant soil is a complex environment, and our work implies multiple symbiotic relationships between microorganisms—Tricholoma matsutake may shape bacterial/fungal diversity, while microorganisms associated with the Tm-dominant soil may enhance T. matsutake growth. As co-occurrence of microorganisms does not necessarily mean interaction, additional study of microbial taxa and functional roles is needed to clarify microbial interactions and functions caused by the dominance effect of T. matsutake. Nevertheless, we believe our study enhances the understanding of T. matsutake in terms of microbial interactions, and the dominance effect of T. matsutake could provide knowledge of fungal effect on the microbial community in soil ecosystems.

Supporting Information

S1 Fig. Constrained Analysis of Principal coordinates (CAP) plots for microbial communities in the Tm-dominant and Tm-minor soil.

CAP model for (A) fungal and (B) bacterial communities, constrained by soil type and geographic location. CAP analyses were conducted on Bray-Curtis distances for fungi and weighted Unifrac distances for bacteria. Significance of CAP models was evaluated using ANOVA with 999 permutations.

https://doi.org/10.1371/journal.pone.0168573.s001

(PDF)

S2 Fig. Differential distribution of major microbial genera between soil types (the Tm-dominant and Tm-minor soil).

Only genera that were present above 1% are shown for (A) fungi and (B) bacteria. Significance of differential abundance between the Tm-dominant and Tm-minor soil was evaluated using a discovery odds ratio test. The genera that were significantly different between soil types (Corrected P < 0.05) are represented with asterisk. Genera are grouped based on phylum membership (Basi: Basidiomycoota, Zygo: Zygomycota, Acid: Acidobacteria, Acti: Actinobacteria, Chlo: Chloroflexi, Firm: Firmicutes, Prot: Proteobacteria, Verr: Verrucomicrobia).

https://doi.org/10.1371/journal.pone.0168573.s002

(PDF)

S1 Table. Information on the fungal communities for soil samples.

Diversity indices and Good’s coverage were calculated after normalization (570 reads) without T. matsutake reads. Because the number of G1D sequences without T. matsutake reads was too small, G1D was excluded for diversity measurement.

https://doi.org/10.1371/journal.pone.0168573.s003

(DOCX)

S2 Table. Information on the bacterial communities for soil samples.

Diversity indices and Good’s coverage were calculated after normalization (1300 reads).

https://doi.org/10.1371/journal.pone.0168573.s004

(DOCX)

Acknowledgments

We would like to appreciate Dr. Gordon Robertson (Canada's Michael Smith Genome Sciences Centre, British Columbia Cancer Agency) and Professor Dr. Colette Breuil (University of British Columbia) for thoughtful review and comments on the manuscript. We would like to thank Dr. John A. Eimes (Seoul National University) for assistance with English editing.

Author Contributions

  1. Conceptualization: S-YO YWL.
  2. Data curation: S-YO JJF.
  3. Formal analysis: S-YO JJF.
  4. Funding acquisition: YWL.
  5. Investigation: YWL MSP.
  6. Project administration: YWL.
  7. Resources: YWL.
  8. Supervision: YWL.
  9. Visualization: S-YO.
  10. Writing – original draft: S-YO JJF YWL.
  11. Writing – review & editing: S-YO JJF MSP YWL.

References

  1. 1. Wang Y, Hall IR, Evans LA. Ectomycorrhizal fungi with edible fruiting bodies 1. Tricholoma matsutake and related fungi. Econ Bot. 1997;51(3):311–27.
  2. 2. Hall IR, Yun W, Amicucci A. Cultivation of edible ectomycorrhizal mushrooms. Trends Biotechnol. 2003;21(10):433–8. pmid:14512229
  3. 3. Matsushita N, Kikuchi K, Sasaki Y, Guerin-Laguette A, Lapeyrie F, Vaario L-M, et al. Genetic relationship of Tricholoma matsutake and T. nauseosum from the Northern Hemisphere based on analyses of ribosomal DNA spacer regions. Mycoscience. 2005;46(2):90–6.
  4. 4. Kang A, Cha D, Kim Y, Park Y, You C. Studies on analyzing meteorological elements related with yield of Tricholoma matsutake (S. Ito et Imai) Singer. Korean J Mycol. 1989;17(2):51–6.
  5. 5. Koo C-D, Kim J-S, Lee S-H, Park J, Ahn K. Spatio-temporal soil water changes in fairy-ring colony of Tricholoma matsutake. J Korean For Soc. 2003;92(6):632–41.
  6. 6. Yang X, Luedeling E, Chen G, Hyde KD, Yang Y, Zhou D, et al. Climate change effects fruiting of the prize matsutake mushroom in China. Fungal Divers. 2012;56(1):189–98.
  7. 7. Guerin-Laguette A, Vaario L-M, Matsushita N, Shindo K, Suzuki K, Lapeyrie F. Growth stimulation of a Shiro-like, mycorrhiza forming, mycelium of Tricholoma matsutake on solid substrates by non-ionic surfactants or vegetable oil. Mycol Prog. 2003;2(1):37–43.
  8. 8. Lee C-Y, Hong O-P, Jung M-J, Han Y-H. Effect of carbon sources and vitamins on mycelial growth of Tricholoma matsutake. Korean J Mycol. 1997;25(3):226–32.
  9. 9. Kim I, Jung G, Han S, Cha J, Sung J. Favorable condition for mycelial growth of Tricholoma matsutake. Korean J Mycol. 2005;33(1):22–9.
  10. 10. Ohara H, Hamada M. Disappearance of bacteria from the zone of active mycorrhizas in Tricholoma matsutake (S. Ito et Imai) Singer. Nature. 1967;213:528–9.
  11. 11. Song H, Min K. Microfungal flora of Tricholoma matsutake producing and nonproducing sites in the forest of Pinus densiflora. Korean J Mycol. 1991;19(2):109–19.
  12. 12. Vaario LM, Fritze H, Spetz P, Heinonsalo J, Hanajik P, Pennanen T. Tricholoma matsutake dominates diverse microbial communities in different forest soils. Appl Environ Microbiol. 2011;77(24):8523–31. pmid:21984247
  13. 13. Kataoka R, Siddiqui ZA, Kikuchi J, Ando M, Sriwati R, Nozaki A, et al. Detecting nonculturable bacteria in the active mycorrhizal zone of the pine mushroom Tricholoma matsutake. J Microbiol. 2012;50(2):199–206. pmid:22538647
  14. 14. Kim M, You YH, Woo JR, Lee GM, Kong WS. Metagenomic Analysis of Fungal Communities Inhabiting the Fairy Ring Zone of Tricholoma matsutake. J Microbiol Biotechn. 2013;23(10):1347–56.
  15. 15. Kim M, Yoon H, Kim YE, Kim YJ, Kong WS, Kim JG. Comparative analysis of bacterial diversity and communities inhabiting the fairy ring of Tricholoma matsutake by barcoded pyrosequencing. J Appl Microbiol. 2014;117(3):699–710. pmid:24922184
  16. 16. Lian C, Narimatsu M, Nara K, Hogetsu T. Tricholoma matsutake in a natural Pinus densiflora forest: correspondence between above- and below-ground genets, association with multiple host trees and alteration of existing ectomycorrhizal communities. New Phytol. 2006;171(4):825–36. pmid:16918553
  17. 17. Buee M, Reich M, Murat C, Morin E, Nilsson RH, Uroz S, et al. 454 Pyrosequencing analyses of forest soils reveal an unexpectedly high fungal diversity. New Phytol. 2009;184(2):449–56. pmid:19703112
  18. 18. Antony‐Babu S, Deveau A, Van Nostrand JD, Zhou J, Le Tacon F, Robin C, et al. Black truffle‐associated bacterial communities during the development and maturation of Tuber melanosporum ascocarps and putative functional roles. Environ Microbiol. 2014;16(9):2831–47. pmid:24118660
  19. 19. Chen L-x, Hu M, Huang L-n, Hua Z-s, Kuang J-l, Li S-j, et al. Comparative metagenomic and metatranscriptomic analyses of microbial communities in acid mine drainage. ISME J. 2014:e245.
  20. 20. Carvalhais LC, Schenk PM. Sample processing and cDNA preparation for microbial metatranscriptomics in complex soil communities. Methods Enzymol. 2013;531:251–67. pmid:24060125
  21. 21. Sharpton TJ. An introduction to the analysis of shotgun metagenomic data. Front Plant Sci. 2014;5:209. pmid:24982662
  22. 22. Langille MG, Zaneveld J, Caporaso JG, McDonald D, Knights D, Reyes JA, et al. Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences. Nat Biotechnol. 2013;31(9):814–21. pmid:23975157
  23. 23. Corrigan A, de Leeuw M, Penaud-Frézet S, Dimova D, Murphy R. Phylogenetic and functional alterations in bacterial community compositions in broiler ceca as a result of mannan oligosaccharide supplementation. Appl Environ Microbiol. 2015;81(10):3460–70. pmid:25769823
  24. 24. Kirst ME, Li EC, Alfant B, Chi Y-Y, Walker C, Magnusson I, et al. Dysbiosis and alterations in predicted functions of the subgingival microbiome in chronic periodontitis. Appl Environ Microbiol. 2015;81(2):783–93. pmid:25398868
  25. 25. de Voogd NJ, Cleary DF, Polónia AR, Gomes NC. Bacterial community composition and predicted functional ecology of sponges, sediment and seawater from the thousand islands reef complex, West Java, Indonesia. FEMS Microbiol Ecol. 2015;91(4):fiv019.
  26. 26. Ma B, Lyu XF, Zha T, Gong J, He Y, Xu JM. Reconstructed metagenomes reveal changes of microbial functional profiling during PAHs degradation along a rice (Oryza sativa) rhizosphere gradient. J Appl Microbiol. 2015;118(4):890–900. pmid:25619404
  27. 27. Barberán A, Bates ST, Casamayor EO, Fierer N. Using network analysis to explore co-occurrence patterns in soil microbial communities. ISME J. 2011;6(2):343–51. pmid:21900968
  28. 28. Bakker MG, Bradeen JM, Kinkel LL. Effects of plant host species and plant community richness on streptomycete community structure. FEMS Microbiol Ecol. 2013;83(3):596–606. pmid:23013423
  29. 29. Trappe MJ, Cromack K, Caldwell BA, Griffiths RP, Trappe JM. Diversity of mat-forming fungi in relation to soil properties, disturbance, and forest ecotype at Crater Lake National Park, Oregon, USA. Diversity. 2012;4(2):196–223.
  30. 30. Gardes M, Bruns TD. ITS primers with enhanced specificity for basidiomycetes‐application to the identification of mycorrhizae and rusts. Mol Ecol. 1993;2(2):113–8. pmid:8180733
  31. 31. Weisburg WG, Barns SM, Pelletier DA, Lane DJ. 16S ribosomal DNA amplification for phylogenetic study. J Bacteriol. 1991;173(2):697–703. pmid:1987160
  32. 32. Muyzer G, Hottenträger S, Teske A, Wawer C. Denaturing gradient gel electrophoresis of PCR-amplified 16S rDNA—a new molecular approach to analyse the genetic diversity of mixed microbial communities. Molecular microbial ecology manual. 1996;3(4):1–23.
  33. 33. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 2010;7(5):335–6. pmid:20383131
  34. 34. Reeder J, Knight R. Rapidly denoising pyrosequencing amplicon reads by exploiting rank-abundance distributions. Nat Methods. 2010;7(9):668–9. pmid:20805793
  35. 35. Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioinformatics. 2010;26(19):2460–1. pmid:20709691
  36. 36. Wang Q, Garrity GM, Tiedje JM, Cole JR. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol. 2007;73(16):5261–7. pmid:17586664
  37. 37. Kõljalg U, Nilsson RH, Abarenkov K, Tedersoo L, Taylor AF, Bahram M, et al. Towards a unified paradigm for sequence‐based identification of fungi. Mol Ecol. 2013;22(21):5271–7. pmid:24112409
  38. 38. DeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL, Keller K, et al. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microbiol. 2006;72(7):5069–72. pmid:16820507
  39. 39. Chun J, Lee J-H, Jung Y, Kim M, Kim S, Kim BK, et al. EzTaxon: a web-based tool for the identification of prokaryotes based on 16S ribosomal RNA gene sequences. Int J Syst Evol Microbiol. 2007;57(10):2259–61.
  40. 40. Caporaso JG, Bittinger K, Bushman FD, DeSantis TZ, Andersen GL, Knight R. PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformatics. 2010;26(2):266–7. pmid:19914921
  41. 41. Haas BJ, Gevers D, Earl AM, Feldgarden M, Ward DV, Giannoukos G, et al. Chimeric 16S rRNA sequence formation and detection in Sanger and 454-pyrosequenced PCR amplicons. Genome Res. 2011;21(3):494–504. pmid:21212162
  42. 42. Price MN, Dehal PS, Arkin AP. FastTree 2–approximately maximum-likelihood trees for large alignments. PLoS ONE. 2010;5(3):e9490. pmid:20224823
  43. 43. Faith DP. Conservation evaluation and phylogenetic diversity. Biol Conserv. 1992;61(1):1–10.
  44. 44. R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2014.
  45. 45. McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS ONE. 2013;8(4):e61217. pmid:23630581
  46. 46. Paulson JN, Stine OC, Bravo HC, Pop M. Differential abundance analysis for microbial marker-gene surveys. Nat Methods. 2013;10(12):1200–2. pmid:24076764
  47. 47. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Series B (Methodological). 1995;57(1):289–300.
  48. 48. Kanehisa M, Goto S, Sato Y, Kawashima M, Furumichi M, Tanabe M. Data, information, knowledge and principle: back to metabolism in KEGG. Nucleic Acids Res. 2014;42(D1):D199–D205.
  49. 49. Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26(1):139–40. pmid:19910308
  50. 50. Kluber LA, Tinnesand KM, Caldwell BA, Dunham SM, Yarwood RR, Bottomley PJ, et al. Ectomycorrhizal mats alter forest soil biogeochemistry. Soil Biol Biochem. 2010;42(9):1607–13.
  51. 51. Griffiths R, Baham J, Caldwell B. Soil solution chemistry of ectomycorrhizal mats in forest soil. Soil Biol Biochem. 1994;26(3):331–7.
  52. 52. Griffiths RP, Castellano MA, Caldwell BA. Hyphal mats formed by two ectomycorrhizal fungi and their association with Douglas-fir seedlings: a case study. Plant Soil. 1991;134(2):255–9.
  53. 53. Cromack K, Fichter B, Moldenke A, Entry J, Ingham E. Interactions between soil animals and ectomycorrhizal fungal mats. Agric Ecosyst Environ. 1988;24(1):161–8.
  54. 54. Boersma F, Otten R, Warmink J, Nazir R, Van Elsas J. Selection of Variovorax paradoxus-like bacteria in the mycosphere and the role of fungal-released compounds. Soil Biol Biochem. 2010;42(12):2137–45.
  55. 55. Zhang M, e Silva MCP, Maryam CDM, van Elsas JD. The mycosphere constitutes an arena for horizontal gene transfer with strong evolutionary implications for bacterial-fungal interactions. FEMS Microbiol Ecol. 2014;89(3):516–26. pmid:24807742
  56. 56. Smith SE, Read DJ. Mycorrhizal symbiosis: Academic press; 2010.
  57. 57. Šašek V, Musilek V. Cultivation and antibiotic activity of mycorrhizal basidiomycetes. Folia microbiologica. 1967;12(6):515–23. pmid:6080528
  58. 58. Kope H, Fortin J. Antifungal activity in culture filtrates of the ectomycorrhizal fungus Pisolithus tinctorius. Can J Bot. 1990;68(6):1254–9.
  59. 59. Takakura Y. Tricholoma matsutake fruit bodies secrete hydrogen peroxide as a potent inhibitor of fungal growth. Can J Microbiol. 2015;61(999):1–4.
  60. 60. Warmink JA, van Elsas JD. Selection of bacterial populations in the mycosphere of Laccaria proxima: is type III secretion involved? ISME J. 2008;2(8):887–900. pmid:18421258
  61. 61. Janssen PH. Identifying the dominant soil bacterial taxa in libraries of 16S rRNA and 16S rRNA genes. Appl Environ Microbiol. 2006;72(3):1719–28. pmid:16517615
  62. 62. Fierer N, Leff JW, Adams BJ, Nielsen UN, Bates ST, Lauber CL, et al. Cross-biome metagenomic analyses of soil microbial communities and their functional attributes. Proceedings of the National Academy of Sciences. 2012;109(52):21390–5.
  63. 63. Bending GD, Poole EJ, Whipps JM, Read DJ. Characterisation of bacteria from Pinus sylvestrisSuillus luteus mycorrhizas and their effects on root–fungus interactions and plant growth. FEMS Microbiol Ecol. 2002;39(3):219–27. pmid:19709201
  64. 64. Timonen S, Hurek T. Characterization of culturable bacterial populations associating with Pinus sylvestris-Suillus bovinus mycorrhizospheres. Can J Microbiol. 2006;52(8):769–78. pmid:16917536
  65. 65. Nguyen NH, Bruns TD. The microbiome of Pinus muricata ectomycorrhizae: Community assemblages, fungal species effects, and Burkholderia as important bacteria in multipartnered symbioses. Microb Ecol. 2015;69(4):914–21. pmid:25687126
  66. 66. Warmink J, Nazir R, Van Elsas J. Universal and species‐specific bacterial ‘fungiphiles’ in the mycospheres of different basidiomycetous fungi. Environ Microbiol. 2009;11(2):300–12. pmid:19196267
  67. 67. Voronina EY. Effect of the mycorrhizosphere on soil micromycete biodiversity and community structure and its relation to the rhizosphere and hyphosphere effects. Microbiology. 2011;80(4):584–90.
  68. 68. Halsey JA, e Silva MdCP, Andreote FD. Bacterial selection by mycospheres of Atlantic Rainforest mushrooms. Antonie van Leeuwenhoek. 2016;109(10):1353–65. pmid:27411813
  69. 69. Tominaga Y. Studies on the life history of Japanese pine mushroom, Armillaria matsutake Ito et Imai. Bull Hiroshima Agric Coll. 1963;2:105–45.
  70. 70. You Y-H, Yoon H-J, Woo J-R, Rim S-O, Lee J-H, Kong W-S, et al. Diversity of Endophytic Fungi Isolated from the Rootlet of Pinus densiflora Colonized by Tricholoma matsutake. Korean J Mycol. 2011;39(3):223–6.
  71. 71. Summerbell RC. From Lamarckian fertilizers to fungal castles: recapturing the pre-1985 literature on endophytic and saprotrophic fungi associated with ectomycorrhizal root systems. Stud Mycol. 2005;53:191–256.
  72. 72. Ogawa M, Kawai M. Studies on the artificial reproduction of Tricholoma matsutake (S. Ito et Imai) Sing., III. Effects of growth promoting natural products on the vegetative growth of T. matsutake. Trans Mycol Soc Japan. 1976;17:492–8.
  73. 73. Andreote FD, Rossetto PB, Mendes R, Avila LA, Labate CA, Pizzirani-Kleiner AA, et al. Bacterial community in the rhizosphere and rhizoplane of wild type and transgenic eucalyptus. World J Microb Biot. 2009;25(6):1065–73.
  74. 74. Frey-Klett P, Garbaye J, Tarkka M. The mycorrhiza helper bacteria revisited. New Phytol. 2007;176(1):22–36. pmid:17803639
  75. 75. Requena N, Jimenez I, Toro M, Barea J. Interactions between plant‐growth‐promoting rhizobacteria (PGPR), arbuscular mycorrhizal fungi and Rhizobium spp. in the rhizosphere of Anthyllis cytisoides, a model legume for revegetation in mediterranean semi‐arid ecosystems. New Phytol. 1997;136(4):667–77.
  76. 76. Hrynkiewicz K, Baum C, Niedojadło J, Dahm H. Promotion of mycorrhiza formation and growth of willows by the bacterial strain Sphingomonas sp. 23L on fly ash. Biol Fert Soils. 2009;45(4):385–94.
  77. 77. Leveau JH, Preston GM. Bacterial mycophagy: definition and diagnosis of a unique bacterial-fungal interaction. New Phytol. 2008;177(4):859–76. pmid:18086226
  78. 78. Vik U, Logares R, Blaalid R, Halvorsen R, Carlsen T, Bakke I, et al. Different bacterial communities in ectomycorrhizae and surrounding soil. Scientific reports. 2013;3:3471. pmid:24326907
  79. 79. Uroz S, Oger P, Morin E, Frey-Klett P. Distinct ectomycorrhizospheres share similar bacterial communities as revealed by pyrosequencing-based analysis of 16S rRNA genes. Appl Environ Microbiol. 2012;78(8):3020–4. pmid:22307291
  80. 80. Kluber LA, Smith JE, Myrold DD. Distinctive fungal and bacterial communities are associated with mats formed by ectomycorrhizal fungi. Soil Biol Biochem. 2011;43(5):1042–50.
  81. 81. Nazir R, Boersma F, Warmink J, Van Elsas J. Lyophyllum sp. strain Karsten alleviates pH pressure in acid soil and enhances the survival of Variovorax paradoxus HB44 and other bacteria in the mycosphere. Soil Biol Biochem. 2010;42(12):2146–52.
  82. 82. Nazir R, Tazetdinova DI, van Elsas JD. Burkholderia terrae BS001 migrates proficiently with diverse fungal hosts through soil and provides protection from antifungal agents. Frontiers in microbiology. 2014;5:598. pmid:25426111
  83. 83. Simon A, Bindschedler S, Job D, Wick LY, Filippidou S, Kooli WM, et al. Exploiting the fungal highway: development of a novel tool for the in situ isolation of bacteria migrating along fungal mycelium. FEMS Microbiol Ecol. 2015;91(11):fiv116.
  84. 84. Izumi H, Finlay RD. Ectomycorrhizal roots select distinctive bacterial and ascomycete communities in Swedish subarctic forests. Environ Microbiol. 2011;13(3):819–30. pmid:21176055
  85. 85. Sabella E, Nutricati E, Aprile A, Miceli A, Sorce C, Lorenzi R, et al. Arthrinium phaeospermum isolated from Tuber borchii ascomata: the first evidence for a “Mycorrhization Helper Fungus”? Mycol Prog. 2015;14(8):1–9.
  86. 86. Preston GM. Metropolitan microbes: type III secretion in multihost symbionts. Cell host & microbe. 2007;2(5):291–4.
  87. 87. Nazir R, Warmink JA, Boersma H, Van Elsas JD. Mechanisms that promote bacterial fitness in fungal-affected soil microhabitats. FEMS Microbiol Ecol. 2010;71(2):169–85. pmid:20002182
  88. 88. Viollet A, Corberand T, Mougel C, Robin A, Lemanceau P, Mazurier S. Fluorescent pseudomonads harboring type III secretion genes are enriched in the mycorrhizosphere of Medicago truncatula. FEMS Microbiol Ecol. 2011;75(3):457–67. pmid:21204867
  89. 89. Cusano AM, Burlinson P, Deveau A, Vion P, Uroz S, Preston GM, et al. Pseudomonas fluorescens BBc6R8 type III secretion mutants no longer promote ectomycorrhizal symbiosis. Environ Microbiol Rep. 2011;3(2):203–10. pmid:23761252
  90. 90. Deveau A, Palin B, Delaruelle C, Peter M, Kohler A, Pierrat J-C, et al. The mycorrhiza helper Pseudomonas fluorescens BBc6R8 has a specific priming effect on the growth, morphology and gene expression of the ectomycorrhizal fungus Laccaria bicolor S238N. New Phytol. 2007;175(4):743–55. pmid:17688589
  91. 91. Frey-Klett P, Pierrat JC, Garbaye J. Location and survival of mycorrhiza helper Pseudomonas fluorescens during establishment of ectomycorrhizal symbiosis between Laccaria bicolor and Douglas fir. Appl Environ Microbiol. 1997;63(1):139–44. pmid:16535478
  92. 92. Li C, Massicote H, Moore L. Nitrogen-fixing Bacillus sp. associated with Douglas-fir tuberculate ectomycorrhizae. Plant Soil. 1992;140(1):35–40.
  93. 93. Frey‐Klett P, Chavatte M, Clausse ML, Courrier S, Roux CL, Raaijmakers J, et al. Ectomycorrhizal symbiosis affects functional diversity of rhizosphere fluorescent pseudomonads. New Phytol. 2005;165(1):317–28. pmid:15720643