Analysis of Bacterial Diversity and Communities Associated with Tricholoma matsutake Fruiting Bodies by Barcoded Pyrosequencing in Sichuan Province, Southwest China

Abstract

Endophytes play an important role in the growth and development of the host. However, the study of endophytes is mostly focused on plants, and reports on bacteria associated with fungi are relatively rare. We studied the bacteria associated with fruiting bodies of Tricholoma matsutake picked from seven main T. matsutake-producing areas in Sichuan, China, by barcoded pyrosequencing. About 8,272 reads were obtained per sample, representing 40 phyla, 103 classes, and 495 genera of bacteria and archaea, and 361-797 operational taxonomic units were observed at a 97% similarity level. The bacterial community was always both more abundant and more diverse than the archaeal community. UniFrac analysis showed there were some difference of bacterial communities among the samples sites. Three bacterial phyla, Proteobacteria, Bacteroidetes, and Firmicutes, were dominant in all samples. Correlation analysis showed there was a significant correlation between some soil properties and bacterial community associated with T. matsutake. This study demonstrated that the bacteria associated with T. matsutake fruiting bodies were diversified. Among these bacteria, we may find some strains that can promote the growth of T. matsutake.

Introduction

Tricholoma matsutake (S. Ito et Imai) is an ectomycorrhizal basidiomycete associated with Pinaceae and Fagaceae trees in China, Korea, and elsewhere in the Northern Hemisphere [43,49]. Its fruiting body, the pine mushroom, is commercially important as a valuable food because of its medicinal effects and attractive flavor [14,23]. Polysaccharides and terpenoids extracted from its fruiting body have antitumor and antioxidant properties [19,38,48,51]. The growth of wild T. matsutake is extremely slow, and it is selective to the environmental conditions, growing in virgin forests without pollution and human intervention [31,42,47]. Artificial cultivation of T. matsutake has not been successful [21,46].

It is still unclear how ecological factors such as host plant, soil properties, and the associated microbial communities influence the development of the T. matsutake fruiting body [2,23,28]. The soil surrounding T. matsutake contains diverse microbial communities that may affect the growth of T. matsutake mycelia and the formation of mycorrhiza [25]. The soil microbial communities may also play a role in the material exchange between mushroom mycelia and plant host. Furthermore, the soil bacteria may live extracellularly inside fungal tissue [15]. Their role in fungal development is still unclear. Therefore, understanding the bacteria affiliated with T. matsutake and the microbial community structure underneath the mushroom has important implications for the artificial domestication and cultivation of T. matsutake. The microbial community in soil-mycelia aggregates and in the T. matsutake fairy ring zone has been studied [26,41]. Park et al. [32] also found a new species associated with the pine mushroom. This suggests that the surroundings of T. matsutake contains rich microbial resources for us to understand. The microbial communities associated with T. matsutake fruiting bodies have been analyzed using denaturing gradient gel electrophoresis (DGGE) [29]. However, studies on the bacterial diversity associated with T. matsutake fruiting bodies have rarely used barcoded pyrosequencing.

A majority (over 99%) of the microbes living in natural environments have not been cultured. Despite continuing development of culture techniques for the isolation and identification of microbes, it is still difficult to assess the true diversity in microbial communities using the currently available culture techniques owing to their limitations [1,34]. Metagenomics is the study of genetic material recovered directly from environmental samples. Because of its ability to reveal the previously hidden diversity of microscopic life, metagenomics offers a powerful lens for viewing the whole microbial world in a sample, including unculturable microbes [18,34,35].

In this paper, we studied the bacteria associated with fruiting bodies of T. matsutake picked from seven main T. matsutake-producing areas in Sichuan, China. This paper provides knowledge on the fundamental aspects of T. matsutake, such as T. matsutake-associated bacterial diversity, that may be of particular significance to the artificial domestication and cultivation of T. matsutake.

 

Materials and Methods

Mushroom Sampling Strategy and Soil Analyses

T. matsutake fruiting bodies were picked from Xiaojin, Yajiang, Muli, Yanyuan, Yanbian, Huidong, and Mianning county at its mature stage in Sichuan, China (Fig. 1, Supplemental Fig. S1), and kept in sterile sealed bags on ice. We collected three fruiting bodies in each producing area, based on altitude difference, and 500 g of soil beneath the fruiting bodies with a soil sampler. DNA was extracted from the three fruiting bodies and mixed as one sample to analyze the bacterial diversity by barcoded pyrosequencing. Soil samples from the same site were mixed separately to analyze their properties. Soil particle size distribution was determined using the pipette method [22]. pH was measured in soil water extracted by dissolving air-dried soil in distilled water at a ratio of 1:5. Organic matter content was estimated using the Tyurin method [33]. Total nitrogen was determined by the Kjeldahl method [7]. Molybdenum antimony ascorbic acid spectrophotography was used to estimate total phosphorus. Total potassium was analyzed by flame photometry. Effective nitrogen was measured by the alkali solution diffusion method. Available phosphorus was determined by the baking soda leaching - molybdenum antimony colorimetric method. Available potassium was determined by ammonium acetate extraction - flame photometry [30]. Soil Mg, Ca, Cu, Mn, and Zn were determined by inductively coupled plasma optical emission spectroscopy (Optima 2000 DV; PerkinElmer, USA), with yttrium as the internal standard [8].

Fig. 1.Location of the sampling sites of Tricholoma matsutake fruiting bodies and soil beneath the fruiting bodies.

DNA Extraction and MiSeq Sequencing of 16S rRNA Gene Amplicons

DNA extraction was conducted by using the E.Z.N.A. Fungal DNA Kit (Omega Bio-Tek, USA). The DNA concentration and quality were checked using a NanoDrop Spectrophotometer. Extracted DNA was diluted to 10 ng/μl and stored at -40℃. Universal primers 515F (5’-GTGCCAGCMGCCGCGGTAA-3’) and 806R (5’-GGACTACHVGGGTWTCTAAT-3’) [45] with 10nt barcodes were used to amplify the V4 hypervariable regions of 16S rRNA genes for pyrosequencing using the MiSeq sequencer [11,12]. The PCR mixture (25 μl) contained 1 μl PCR buffer, 1.5 mM MgCl2, each deoxynucleoside triphosphate at 0.4 mM, each primer at 1.0 mM, 0.5 U of TransStart Fast Pfu DNA Polymerase (TransGen, China), and 10 ng of soil genomic DNA. The PCR amplification program included initial denaturation at 94℃ for 3 min, followed by 30 cycles of 94℃ for 40 sec, 56℃ for 60 sec, and 72℃ for 60 sec, and a final extension at 72℃ for 10 min. Two PCRs per sample were combined together after PCR amplification. PCR products were subjected to electrophoresis using 1.0% agarose gel. The band with a correct size was excised and purified using the Gel Extraction Kit (Omega Bio-tek, USA) and quantified with Nanodrop. All samples were pooled together, with an equal molar amount from each sample. A TruSeq DNA kit was used to prepare the sequencing samples. The purified library was diluted, denatured, re-diluted, and then mixed with PhiX (about 30% of final DNA amount) as described in the Illumina library preparation protocols, and then the samples were applied to an Illumina MiSeq system for sequencing with the Reagent Kit ver. 2 2×250 bp according to the manufacturer’s manual.

Pyrosequence Data Analysis

The sequence data were processed using QIIME Pipeline ver. 1.7.0 (http://qiime.org/tutorials/tutorial.html) [10]. Multiple steps were required to trim the sequences, such as trimming the barcoded fusion primers, and filtering low-quality sequences (read length <150 bp or average quality value <25, ambiguous base calls⩾2) out. Sequences were clustered into operational taxonomic units (OTUs) at a 97% identity threshold, and the cut-off values used for taxonomic assignments were as follows (x = similarity): genus (97% > x ⩾ 94%), family (94% > x ⩾ 90%), order (90% > x ⩾ 85%), class (85% > x ⩾ 80%), and phylum (80 > x ⩾ 75%). If the similarity was lower than the specific cut-off value, the sequence was assigned as “unclassified” [27]. The aligned 16S rRNA gene sequences were used for chimera check using the Uchime algorithm [17]. All the samples were randomly resampled to 9,700 reads. We conducted alpha-diversity (phylogenetic distance whole tree, Chao1 estimator of richness, observed species, and Shannon’s diversity index) and beta-diversity (PCoA, UniFrac) analyses [9], for which the rarefaction curves were generated from the observed species. Taxonomy was assigned using the Ribosomal Database Project classifier [44].

 

Results

Diversity Indices

About 8,272 reads were obtained per sample, representing 40 phyla, 103 classes, and 495 genera of bacteria and archaea, and 361–797 OTUs were observed at a 97% similarity level (Table 1). The numbers of observed OTUs were significantly different between the sampling sites (Table 1). The rarefaction curves (Fig. 2) calculated with QIIME pipeline at 97% similarity also showed a different OTU richness pattern for all sites. Other OTU richness estimations, such as Chao1, indicated that samples picked from Mianning contained the lowest number of bacteria. According to the comprehensive OTU richness results (Table 1 and Fig. 2), the bacterial community of fruiting bodies from Yanyuan was the most diverse (Simpson index, 0.969; Shannon index, 6.38), whereas the lowest bacterial OTU diversity (Simpson index, 0.932; Shannon index, 4.90) was in the samples from Huidong.

Table 1.Chao1, estimator of richness. Good’s coverage is proportional to the nonsingleton phylotypes in all sequences.

Fig. 2.Rarefaction curves for bacterial operational taxonomic units (OTUs) in each sampling site (cut-off value at 97% similarity).

Characteristics of Soil Beneath Tricholoma matsutake Fruiting Bodies

Soil samples, down to a depth of 15 cm, were collected from inside the T. matsutake fairy ring at seven different sites. The physicochemical properties of the soils in the sampling sites differed slightly (Table 2). The pHs of the soil samples obtained from the fairy rings varied between 5.08 and 6.48. The soil collected from Yanyuan had the highest content of organic matter (189.6 g/kg), total nitrogen (4.39 g/kg), total phosphorus (0.42 g/kg), and available phosphorus (10.12 mg/kg), and it also contained the lowest total potassium (17.3 g/kg). Soil samples from Mianning had the most acidic environment (pH 5.08) and the lowest available phosphorus (1.21 mg/kg). In terms of soil texture, the clay composition was lower (1.73%) and the sand composition was higher (88.5%) in Mianning than in the other sites. The mineral element contents were slightly different at different sites (Table 2).

Table 2.OM, organic matter; TN, total nitrogen; TP, total phosphorus; TK, total potassium; AN, effective nitrogen; AP, available phosphorus; AK, available potassium; TMn, total manganese; TCu, total copper; TZn, total zinc; TCa, total calcium; TMg, total magnesium.

Taxonomic Analyses of Bacterial Communities

Each bacterial 16S rRNA gene sequence was taxonomically assigned from the phylum level to the species level using the Ribosomal Database Project classifier. The proportion of unclassified bacteria was no more than 0.1%. A total of 40 phyla were identified, and 17 of the 40 phyla were identified in all seven samples (Fig. 3A). The bacterial community was always both more abundant and more diverse than the archaeal community. Two archaeal phyla, Crenarchaeota and Euryarchaeota, accounted for less than 1% of all bacteria associated with T. matsutake fruiting bodies. The most abundant archaeal community was Euryarchaeota, about 0.3% of all bacteria detected. More than 37 bacterial phyla were detected in the samples. Three bacterial phyla, Proteobacteria, Bacteroidetes, and Firmicutes, were dominant in all samples. The relative abundance of Proteobacteria was more than 47% of all bacteria detected. Overall, Firmicutes and Bacteroidetes were less abundant. However, in samples picked from Yajiang, Firmicutes presented 41.0% of the OTUs, and in Yanyuan samples, Bacteroidetes presented 44.1% of the OTUs.

Fig. 3.Taxonomic composition analysis of bacterial communities (from phylum to family).

A total of 103 classes were identified, and 26 of the 103 classes were identified in all seven sites (Fig. 3B). The class Gammaproteobacteria was dominant in all samples (the average relative abundance was 52.2%). In particular, Bacilli (39.3%) and Betaproteobacteria (36.9%) were the dominant classes in Yajiang and Muli samples, respectively. In general, the classes Gammaproteobacteria, Betaproteobacteria, Bacilli, and Sphingobacteria occupied a dominant position in T. matsutake fruiting bodies from different sites.

A total of 183 orders were observed, and 38 of these were detected in all samples (Fig. 3C). The most abundant observed orders at all sites were Pseudomonadales (average relative abundance 35.8%), Burkholderiales (average relative abundance 15.8%), and Enterobacteriales (average relative abundance 11.3%). In particular, Burkholderiales (35.4%), Enterobacteriales (28.3%), and Bacillales (37.1%) were the dominant classes in the Muli, Yanbian, and Yajiang samples, respectively.

A total of 310 families were observed and 65 were identified in all seven sites (Fig. 3D). The relative abundance of some families, including Pseudomonadaceae (average of 35.3%), Enterobacteriaceae (11.3%), Oxalobacteraceae (8.3%), and Sphingobacteriaceae (7.3%), were dominant in all samples, but differed significantly between the sites. Pseudomonadaceae was the biggest family in most samples except samples from Yanyuan and Yanbian, in which Sphingobacteriaceae and Enterobacteriaceae occupied a dominant position, respectively.

The relative abundance of the genera Pseudomonas (in the family Pseudomonadaceae), unclassified genus (in the family Enterobacteriaceae), unclassified genus (in the family Pseudomonadaceae) and Janthinobacterium (in the family Oxalobacteraceae) was dominant in all samples (Table 3). Among them, Pseudomonas was the most abundant (average abundance 26.3%) and occupied the dominant position in samples from Xiaojin, Yanyuan, Huidong, and Mianning. Janthinobacterium was the most abundant genus in Muli sample (25.4%). Unclassified genus (in the family Enterobacteriaceae) was dominant in Yanbian sample (28.2%). In particular, unclassified genus (in the family Planococcaceae) occupied a dominant position in Yajiang sample (22.1%).

Table 3.List of bacterial genera (>1% in at least one experimental site) associated with the Tricholoma matsutake fruiting bodies of different sites.

Several species showed average abundances greater than 1% (Table 4). The communities from the seven different sites shared the same dominant species, Pseudomonas sp., Janthinobacterium sp., Janthinobacterium lividum, Pseudomonas viridiflava, Chryseobacterium sp., and other unclassified species (in the family Enterobacteriaceae, Comamonadaceae, Xanthomonadaceae, Pseudomonadaceae, and so on). The predominant species were two Pseudomonas sp., belonging to the family Pseudomonadaceae.

Table 4.List of bacterial species (average abundance >1%) associated with the Tricholoma matsutake fruiting bodies of different sites.

UniFrac Analysis

The differences in bacterial communities between the samples were estimated using UniFrac analysis (Fig. 4). The bacteria community structure in Yajiang and Yanyuan were significantly different from the other sites. The communities from the geographically close and similar environments like Huidong and Mianning were highly similar, implying that the environmental conditions can affect the T. matsutake-associated bacterial community structure.

Fig. 4.Principal coordinate analysis using weighted UniFrac.

Correlation Analysis Between Bacterial Community Associated with the Fruiting Bodies of Tricholoma matsutake and Soil Properties

There was a significant correlation between some soil properties and bacterial community associated with T. matsutake fruiting bodies (Table 5). Chao1 and the numbers of observed OTUs were correlated with clay (positively) content and sand (negatively) content. The Simpson index and Shannon index were positively correlated with total nitrogen and available phosphorus. Chao1 was also positively correlated with total nitrogen and total zinc, and negatively correlated with total magnesium. The relative abundance of the phylum Proteobacteria, the predominant phylum in this study, was positively correlated with the presence of total potassium. In contrast, the relative abundance of the phylum Bacteroidetes was negatively correlated with total potassium but positively correlated with organic matter. The abundance of Actinobacteria was negatively correlated with available phosphorus and total zinc. Total zinc was also negatively correlated with the abundance of the phylum Nitrospirae and Euryarchaeota. The abundance of Acidobacteria was positively correlated with total nitrogen and total copper. There were no clear correlations between soil pH and phyla abundance.

Table 5.OM, organic matter; TN, total nitrogen; TP, total phosphorus; TK, total potassium; AN, effective nitrogen; AP, available phosphorus; AK, available potassium; TMn, total manganese; TCu, total copper; TZn, total zinc; TCa, total calcium; TMg, total magnesium. *Significant at p < 0.05; **Significant at p < 0.01. Chao1, estimator of richness.

 

Discussion

Host-associated microbes play an important role in the growth and development of the host [16,39]. They can participate in the metabolic processes of the host, produce biological macromolecules with growth-promoting or antibacterial activities, and affect the yield and quality of the host [3,4,50]. However, the study of host-associated microbes is mostly focused on plants, and reports on bacteria associated with mushrooms are relatively rare [6,15,20,33]. Many large mushrooms have a complex life history and demanding environmental requirements. Studying the community structure of bacteria in mushrooms may be significant for their cultivation.

Most microbes in nature cannot be obtained in pure culture because of the difficulty of simulating the conditions required for their growth and reproduction. Earlier, we characterized the T. matsutake fruiting body-associated bacteria using DGGE [29]. For a more comprehensive picture of the bacterial diversity, we applied barcoded pyrosequencing for culture-independent bacterial community analysis. The results indicated that the bacteria associated with T. matsutake fruiting bodies were relatively abundant. In agreement with the earlier results [29], the bacterial communities were also varied in different samples, possibly due to the different ecological environments, as the communities from similar environments were similar. In the fairy ring zone of T. matsutake, Proteobacteria and Acidobacteria were the dominant phyla, and the relative abundance of the Proteobacteria was approximately twice that of the Acidobacteria [27]. The bacterial communities associated with the fruiting bodies were different, as Proteobacteria, Bacteroidetes, and Firmicutes were distributed in all samples, and Proteobacteria were over 50 times more abundant than Acidobacteria. This indicated that microbial taxons associated with the T. matsutake fruiting bodies were selectively enriched or reduced compared with microorganisms in the fairy ring zone. It remains to be seen if this change in the composition of microbial populations is related to the growth of T. matsutake. Pseudomonas spp. that were abundant in T. matsutake fruiting bodies have shown the ability to promote the growth of plants [13,40]. Therefore, they might be made into microbial fertilizer applied in the artificial cultivation of T. matsutake. Janthinobacterium sp. and Pedobacter sp. have shown antimicrobial activity against pathogenic organisms [5,36]. Whether these bacteria are active against pathogens inside fruiting bodies needs to be tested.

Previous studies showed that soil properties significantly affect the microbial community structure beneath T. matsutake fruiting bodies [27]. Similarly, in our study, there was a significant correlation between some soil properties and bacteria associated with T. matsutake. From the previous studies, we knew that the clay content in the fairy ring of T. matsutake positively affected the OTU diversity and evenness. In addition, the relative abundance of the phylum Actinobacteria was similarly positively correlated with clay and negatively correlated with sand content [27]. Interestingly, in this study, the clay content and sand content were also correlated with the numbers of observed OTUs and Chao1 in the same pattern, yet there was no significant correlation with the abundance of the phylum Actinobacteria. In addition, we found that some mineral elements in soil also correlated significantly with some bacterial taxa. For example, total zinc was negatively correlated with the abundance of Actinobacteria, Acidobacteria, and Euryarchaeota, and total copper positively affected the abundance of Acidobacteria, possibly related to the demand of these mineral elements by the bacteria. Moreover, there were no clear correlations between phyla abundance and soil pH in the fairy ring zone or fruiting bodies.

Altogether, the diverse T. matsutake-associated bacteria showed good prospects for the cultivation of T. matsutake, yet the questions to be answered are many. Further study of the growth-promoting bacteria should be tested, and their role in the material exchange between host plant and mycorrhizal fungi and in pest resistance should be assessed. In conclusion, this study provides important knowledge about the bacterial community inhabiting the fruiting bodies of T. matsutake and will lay a good foundation for the cultivation of T. matsutake.