Analysis of ethanol fermentation mechanism of ethanol producing white-rot fungus Phlebia sp. MG-60 by RNA-seq

Background The white-rot fungus Phlebia sp. MG-60 shows valuable properties such as high ethanol yield from several lignocellulosic materials, although white-rot fungi commonly degrade woody components to CO2 and H2O. In order to identify genes involved in ethanol production by Phlebia sp. MG-60, we compared genes differentially expressed by the ethanol producing fungus Phlebia sp. MG-60 and the model white-rot fungus Phanerochaete chrysosporium under ethanol fermenting and non-fermenting conditions using next-generation sequencing technologies. Results mRNAs from mycelia of Phlebia sp. MG-60 and P. chrysosporium under fermenting and non-fermenting conditions were sequenced using the MiSeq system. To detect differentially expressed genes, expression levels were measured in fragments per kilobase of exon per million mapped reads (FPKM). Differentially expressed genes were annotated using BLAST searches, Gene Ontology classifications, and KEGG pathway analysis. Functional analyses of differentially expressed genes revealed that genes involved in glucose uptake, glycolysis, and ethanol synthesis were widely upregulated in Phlebia sp. MG-60 under fermenting conditions. Conclusions In this study, we provided novel transcriptomic information on Phlebia sp. MG-60, and these RNA-seq data were useful in targeting genes involved in ethanol production for future genetic engineering. Electronic supplementary material The online version of this article (doi:10.1186/s12864-016-2977-7) contains supplementary material, which is available to authorized users.


Background
Most bioethanol production in the world is currently from food crops, which leads to competition with food and feed uses. Advanced second-generation bioethanol is a renewable transportation fuel made from lignocellulosic biomass that does not compete with food or feed [1]. Lignocellulosic biomass is mainly composed of cellulose, hemicellulose, and lignin and is the most abundant material for bioethanol production. Wood-rot basidiomycetes play key roles in the carbon cycle in forest ecosystems through multi-enzyme systems that degrade lignocelluloses [2]. Wood-rot basidiomycetes are considered the most efficient degraders of lignocellulose in nature. Biological delignification by white-rot basidiomycetes, which is a useful pretreatment for enzymatic saccharification of lignocellulosic biomass, is summarized by Moreno et al. [3]. Basidiomycetes are expected to be used for pretreatment in bioethanol production from lignocellulosic materials. Therefore, some researchers have reported use of white-rot fungi in pretreatment for enzymatic saccharification of lignocellulosic biomass [4][5][6]. Furthermore, several reports have been published about genome projects on Basidiomycota, Agaricomycetes and Polyporales lignocellulose-degrading fungi, and comparative genomic studies recently [7][8][9][10].
Several white-rot fungi have been reported to ferment oligosaccharide materials to ethanol. The white-rot fungi Peniophora cinerea and Trametes suaveolens efficiently convert hexoses to ethanol [11], and Trametes hirsuta shows efficient fermentation of starch, wheat bran and rice straw to ethanol without acid or enzymatic hydrolysis [12]. Recently, Okamoto et al. documented that the white-rot fungus Trametes versicolor KT9427 can produce ethanol from starch, cellulose, xylan, wheat bran and rice straw [13].
The white-rot fungus Phlebia sp. MG-60 was selected as a hypersaline-tolerant lignin-degrading fungus from 28 mushrooms and samples of driftwood based on decolorization and delignification abilities after collection from mangrove stands in Okinawa, Japan [14]. Phlebia sp. MG-60 was capable of converting lignocellulose to ethanol directly with high yield [6,15,16]. When this fungus was cultured with 20 g L −1 of unbleached hardwood kraft pulp for 168 h, 71.8 % of the theoretical maximum yield of ethanol was observed, and when it was cultured with waste newspaper, 51.5 % of the theoretical maximum yield was observed [16]. The delignification, anaerobic saccharification, and fermentation of oak wood using only Phlebia sp. MG-60, without addition of chemicals or enzymes, has been also reported [16]. Additionally, alkaline pretreated sugarcane bagasse was fermented well directly, without addition of cellulases, by Phlebia sp. MG-60 [6]. Thus, Phlebia sp. MG-60 possesses not only wood degrading ability but also ethanol fermentability. However, the detailed mechanism of fermentation by this fungus remains unknown.
The goal of this study was to characterize the specific genes for ethanol production, and to predict the mechanism behind the high yield of ethanol by Phlebia sp. MG-60. In the present study, we analyzed differential gene expression of the ethanol producing white-rot fungus Phlebia sp. MG-60 and the model white-rot fungus P. chrysosporium, used as the control, under fermenting and non-fermenting conditions by next-generation sequencing. This is the first report of transcriptomic data of the ethanol producing white-rot fungus Phlebia sp. MG-60.

Production of ethanol from glucose
In order to equalize the growth of mycelia, Phlebia sp. MG-60 and P. chrysosporium were incubated on PDA plates at 30°C for 5 and 3 days, respectively. 10-mm diameter disks were then punched out from the growing edge of the mycelia using a sterile cork borer. Two mycelial disks for each strain were placed into a 100-mL Erlenmeyer flask containing 10 mL basal liquid medium (20 g L −1 glucose, 10 g L −1 yeast extract, 10 g L −1 KH 2 PO 4 , 2 g L −1 (NH 4 ) 2 SO 4 , and 0.5 g L −1 MgSO 4 -7H 2 O, pH 4.5). After sealing the flask with a silicone plug stopper (to ensure semi-aerobic conditions), the culture was statically incubated at 30°C for 10 days. Cultures were filtered through a 0.2-μm membrane filter, and the filtrate was then separated by high-performance liquid chromatography (HPLC) using a Shodex SH1821 column (8.0 mm × 300 mm, Showa Denko K.K., Tokyo, Japan) at 75°C with 0.5 mM H 2 SO 4 as the mobile phase at a flow rate of 0.6 mL min −1 , and ethanol and glucose concentrations in the cultures were measured using an online refractive index detector. The pH of the culture was also measured by a glass electrode (D-51S, Horiba Ltd., Kyoto, Japan).

Mycelial dry weight
For monitoring the growth of Phlebia sp. MG-60 and P. chrysosporium, mycelial dry weights obtained from liquid culture which described as above were measured. Cultures were filtered through 0.2-μm membrane filter, mycelium and filter were then dried. The mycelial dry weight calculates by (weight of filter + dried residue) -(weight of filter paper).

cDNA library preparation for DNA sequencing
To construct RNA-seq libraries, total RNA was isolated from the mycelia of Phlebia sp. MG-60 after 2 and 9 days of incubation and from mycelia of P. chrysosporium after 3 and 9 days of incubation. Total RNA was first purified from cultured mycelia with three biological replicates obtained from separate cultures using a Qiagen RNeasy Mini Kit (Hilden, Germany). The quality and quantity of each RNA sample were assessed as described previously [17]. Agarose gel electrophoresis and the OD260/OD280 ratio were used for assessing quality of total RNA.
Each extracted total RNA sample was treated with DNase I and repurified using an RNeasy Mini Kit (Qiagen) following the manufacturer's protocol. Purified RNA (1 μg) was used for first-strand cDNA synthesis using an oligo-dT primer and PrimeScript reverse transcriptase (Takara). An equal quantity of 50 ng total RNA was used for PCR. Primer sequences and the expected product sizes are shown in Additional file 1: Table S1. RT-PCR was performed using 1 μg of total RNA with the PrimeScript RT-PCR Kit and the RT-PCR amplified fragments were analyzed by agarose gel electrophoresis. The libraries for strand-specific RNA sequencing were constructed using a SureSelect Strand-Specific RNA Library prep kit (Agilent Technologies) according to the manufacturer's protocol. We generated libraries of each biological replicates, and then libraries derived from same species were pooled together. The two individual sequencing runs for Phlebia sp. MG-60 and P. chrysosporium were performed, respectively. Transcriptome sequencing of paired-end reads (75 bp) was performed by a MiSeq system (Illumina).

De novo assembly and differential expression analysis
The raw reads were processed using cutadapt version 1.8.1 to remove adapter sequences [18], low-quality bases (quality scores <30) and reads shorter than 50 nt. The last 76 bases were trimmed by FASTX-Toolkit [19]. After quality trimming, the high-quality reads were assembled into unigenes by Trinity (version 2.0.6) [20] Oases (version 0.2.08) [21], Trans-ABySS (version 1.5.4) [22] and SOAPdenovo-Trans (version 1.03) [23]. The resulting unigenes was further analyzed using DETONATE (de novo transcriptome RNA-seq assembly with or without the truth evaluation) [24]. In de novo transcriptome assembly by Trinity program, we selected the Jaccard-clip option to reduce the generation of chimeric transcripts. rRNA were excluded from the unigenes by removing sequences matching entries in the SILVA rRNA database by the Megablast program [25]. The genes derived from mitochondria sequences were manually removed from the unigenes using the result of local BLASTX and BLASTN search against nr and nt database. High-quality short reads were mapped to the rRNA-removed unigenes as a reference using Bowtie [26], and then transcript abundance was estimated using RSEM software [27]. To identify differentially expressed genes (DEGs), P-values and fold changes were computed using the edgeR package [28].

Functional annotation
The unigenes were searched against the Swiss-Prot database by a local BLASTX algorithm (E-value cut-off was set at 10e-5) to predict the biological functions [29]. Open reading frames (ORFs) and their protein sequences were predicted from unigenes using TransDecoder, which is included in the Trinity package. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were assigned to unigenes using the BlastKOALA server [19,30]. Gene ontology (GO) annotation of the transcriptome was performed using InterproScan software version with the "goterms" option [31]. Orthologous protein pairs between Phlebia sp. MG-60 and P. chrysosporium were identified by the FastOrtho program (http://enews.patricbrc.org/fastortho/) [32] as described previously [33], which is a reimplementation of the OrthoMCL program [34]. The longest peptides translated by TransDecoder were used for orthologous protein analysis as representatives of peptides encoded by the Trinity unigenes. Proteins showing one-to-one correspondence between Phlebia sp. MG-60 and P. chrysosporium were regarded as orthologous proteins.

Semi-quantitative RT-PCR and quantitative RT-PCR (qRT-PCR)
To analyse gene expression in Phlebia sp. MG-60, samples were prepared as described in the section Production of ethanol from glucose. The semi-quantitative RT-PCR method was as described in our previous study [35]. All the selected unigenes used in RT-PCR and qRT-PCR evidenced high similarity with known genes with very low E-values (< E-20), and these transcripts were manually confirmed mapping rate of sequencing reads from each sample and sequence errors using the visualized software IGV tools version 2.0 (http://software.broad institute.org/software/igv/home). PCR was performed for 28 cycles, with template denaturation at 95°C for 30 s, primer annealing at 58°C for 30 s, and DNA extension at 72°C for 1 min using Ex Taq DNA Polymerase (TaKaRa Bio).
For the quantitative real-time RT-PCR assays, the Roche LightCycler 480 system was used. Each reaction contained 50 ng of first-strand cDNAs, 2 μL PCR primers, 7 μL water, and 10 μL master mix (Roche). Cycling conditions were set as follows: pre-incubation, 95°C for 10 min; amplification, 45 cycles of 95°C for 10 s, 60°C for 10 s, and 7°C for 10 s. The reference genes actin and hydroxymethylbilane synthase (HMBS) were used to test for sample-to-sample variation. Relative quantitation using the comparative Ct method was calculated as ΔΔCt = (ΔCt target -ΔCt control ) fermenting conditions − (ΔCt target -ΔCt control ) non-fermenting conditions ..

Production of ethanol from glucose
The time courses of ethanol production from glucose and glucose consumption by the ethanol producing white-rot fungus Phlebia sp. MG-60 and the control white-rot fungus P. chrysosporium are shown in Fig. 1. Phlebia sp. MG-60 produced the maximum ethanol concentration of 9.5 g L −1 from glucose after 3 days of incubation, and the concentration then gradually diminished. On the other hand, maximum ethanol production by P. chrysosporium was 2.4 g L −1 after 8 days of incubation ( Fig. 1). The glucose concentration of Phlebia sp. MG-60 and P. chrysosporium were 12.8 g L −1 and 18.2 g L −1 at 1 day, and then almost decreased to 0 g L −1 until 2 days and 5 days after incubation, respectively. We also measured the mycelial growth and pH changes in liquid cultures of both species (Additional file 2: Figure S1). Although the growth rate of P. chrysosporium was higher than that of Phlebia sp. MG-60, ethanol productivity of Phlebia sp. MG-60 was much higher than that of P. chrysosporium. The pH in liquid cultures of both species was between 4.5 and 3.8 during the experiment. All the results indicated that there exists a correlation between consumption of glucose and production of ethanol in the mycelia culture of Phlebia sp. MG-60 at 2 days after incubation. The consumed glucose was mainly used for its ethanol production in Phlebia sp. MG-60. The maximum ethanol productivity was at 2 days in Phlebia sp.
MG-60 and 3 days of incubation in P. chrysosporium, and that ethanol production stopped at 9 days of incubation in both fungi. Therefore, 2-day-incubated mycelia of Phlebia sp. MG-60 and 3-day-incubated mycelia of P. chrysosporium were used as the fermenting samples, and 9-day-incubated mycelia of Phlebia sp. MG-60 and 9-day-incubated mycelia of P. chrysosporium were used as the non-fermenting samples for further experiments.

Sequencing and de novo assembly
Total RNA was extracted from Phlebia sp. MG-60 and P. chrysosporium for transcriptome analysis. Samples included biological triplicates of cDNA libraries from Phlebia sp. MG-60 after 2 and 9 days of incubation and P. chrysosporium after 3 and 9 days of incubation (termed MG2D, MG9D, PC3D, and PC9D, respectively). Total mRNA was purified, and then cDNA libraries were constructed and sequenced on the MiSeq platform.  Table S2). After removal of low-quality regions of the reads (quality values <30), each of the high-quality reads from Phlebia sp. MG-60 and P. chrysosporium were assembled into unigenes using the Trinity, Oases, Trans-ABySS and SOAPdenovo-Trans program, and further analyzed using DETONATE. The resulting RSEM-EVAL score from DET-ONATE analysis indicated that Trinity produced the most accurate assemblies in all data sets. Next, rRNA and genes derived from mitochondria sequences were excluded from Trinity unigenes by removing those matching entries in the SILVA 111 rRNA [36], nr and nt database. Ultimately, we obtained 34,731 (N50 length = 1,422 bp) unigenes in Phlebia sp. MG-60 and 27,956 (N50 length = 1,533 bp) in P. chrysosporium (Table 1).

Functional annotations of differentially expressed genes
All unigenes were searched against the Swiss-Prot database using local BLASTX (Additional file 4: Table S3, Additional file 5: Table S4), and further annotated with GO terms using InterProScan. To identify DEGs, we compared the unigenes that were over 2-fold up-or downregulated with false discovery rate (FDR) < 0.05 between the fermenting and non-fermenting conditions. In Phlebia sp. MG-60, 2,944 genes were upregulated, and 3,910 genes were downregulated (Additional file 4: Table  S3). There were 1,689 upregulated genes and 1,901 downregulated genes in P. chrysosporium (Additional file 5: Table S4). We performed GO enrichment analysis of Phlebia sp. MG-60 and P. chrysosporium, which revealed that 27 GO terms were upregulated and 5 were downregulated in Phlebia sp. MG-60. The most enriched GO terms were "ATP binding" (GO: 0005524) in the molecular functions category, "fatty acid biosynthetic process" (GO: 0006633) in biological processes, and "integral component of membrane" (GO: 0016021) in cellular components under fermenting conditions (Table 2). In contrast, only "catalytic activity" (GO: 0003824) in molecular functions was significantly enriched in P. chrysosporium (data not shown).
To evaluate the high ethanol productivity in Phlebia sp. MG-60, we mapped all genes in Phlebia sp. MG-60 and P. chrysosporium to KEGG metabolic pathways  using BlastKOALA, and focused on genes involved in ethanol fermentation. Metabolic pathways and biosynthesis of secondary metabolites were the most frequently represented pathways, including glycolysis and pyruvate oxidation, which are involved in ethanol production pathways. Based on the KEGG pathway assignments, we compared the genes involved in the glycolysis/gluconeogenesis pathway in Phlebia sp. MG-60 and P. chrysosporium (Additional file 6: Figure S2). Lists of transcripts related to the glycolysis/gluconeogenesis pathway are shown in Additional file 7: Table S5 and Additional file 8: Table S6. In all, 40 genes were mapped to the glycolysis/gluconeogenesis pathway in Phlebia sp. MG-60, but only 18 in P. chrysosporium.  (Fig. 2b).

Semi-quantitative RT-PCR and qRT-PCR of Phlebia sp. MG-60
To validate the reliability of the expression profiles obtained by RNA-seq, expression levels of five highly expressed genes related to ethanol fermentation in Phlebia sp. MG-60, including one sugar transporter (TR10028|c0_g1), three genes mapped to the glycolysis/gluconeogenesis pathway (TR8916|c1_g1, TR11270|c0_g1, TR9324|c0_g1), and one ADH (TR11797|c0_g1), were compared by semiquantitative RT-PCR and qRT-PCR. As shown in Fig. 3a, semi-quantitative RT-PCR indicated that all of  these genes showed higher expression at 2 days of incubation. Using qRT-PCR, we examined the expression of these genes at 2 and 9 days of incubation and normalized expression to the actin and HMBS genes (Fig. 3b). These results all indicated that expression levels measured by semi-quantitative RT-PCR and qRT-PCR correlated with the RNA-seq analysis.

Discussion
In this study, we analyzed differential gene expression of the ethanol producing white-rot fungus Phlebia sp. MG-60 under fermenting and non-fermenting conditions using next-generation sequencing techniques, and identified the genes involved in its high ethanol production.
Gene expression of ligninolytic enzymes in Phlebia sp.

MG-60
Biological delignification is currently attracting much attention as an alternative technology to traditional physicochemical methods for the saccharification of lignocellulosic biomass [3]. White-rot fungi have a unique ability to degrade lignin via extracellular ligninolytic enzymes such as lignin peroxidase, manganese peroxidase (MnP), and laccase [37]. MnP oxidizes Mn 2+ to Mn 3+ , and Mn 3+ acts on monomeric phenol, phenolic lignin dimers, and synthetic lignin [38,39]. Laccases are a group of multi-copper oxidases, which have the ability to oxidize both phenolic and non-phenolic lignin units [40,41]. Following BLASTx searches of Phlebia sp. MG-60 against amino acid sequences in the Swiss-Prot database, thirteen MnP genes and eleven laccase genes were detected (Additional file 4: Table S3). Recently, MnP gene transformants of Phlebia sp. MG-60 showed higher MnP activity, and overexpression of the MnP gene improved delignification ability of Phlebia sp. MG-60 [42]. The present study was done using glucose as carbon source and therefore not adapted to discuss about ligninolytic enzymes involved in lignocellulose conversion. However, Phlebia sp. MG-60 is a candidate for an integrated fungal fermentation process due to its efficient delignification.
Expression of glucose transporter gene in Phlebia sp.

MG-60
GO functional enrichment analysis of Phlebia sp. MG-60 indicated that two groups, "integral component of membrane" (GO: 0016021) and "membrane" (GO: 0016020), were significantly enriched in the cellular components ( Table 2). The first step of sugar metabolism is its transport across the cell membrane [43]. Efficient sugar uptake through the expression of hexose transporter genes can improve fermentation of lignocellulosic biomass to ethanol [44]. Yeast hexose transporters have been most extensively researched, and Saccharomyces cerevisiae has at least eight hexose transporters that mediate the uptake of glucose [45,46]. In the present study, nine glucose transporter genes were expressed in Phlebia sp. MG-60. In particular, TR10228|c0_g1, TR10117|c0_g1, TR2714|c0_g1 were characterized as high-affinity glucose transporters, and their expression increased 47.8-, 12.9-, and 4.9-fold under fermenting conditions (Additional file 4: Table  S3). Ali et al. reported that the overexpression of a high-affinity glucose transporter gene in the fungus Fusarium oxysporum directly affected the glucose and xylose transport capacity and ethanol yield [47]. These results confirm that high-affinity glucose transporter genes expressed at a high level in Phlebia sp. MG-60 play an important role in the initiation of ethanol fermentation from glucose.

Expression of genes involved in glycolysis
DEGs involved in the glycolysis pathway and ethanol fermentation in Phlebia sp. MG-60 are summarized in Fig. 4 (Fig. 4, Additional file 7: Table S5). KEGG annotations indicated that all genes involved in the production of ethanol from glucose were expressed in Phlebia sp. MG-60 (Additional file 6: Figure S2). In contrast, only nine of these genes were expressed in P. chrysosporium (Additional file 6: Figure S2). In current commercial ethanol production, the yeast S. cerevisiae is mostly researched, which shows high ethanol yield from glucose, and its prime metabolic pathway of ethanol fermentation is glycolysis [48]. Our results clarified that the ethanol producing white-rot fungus Phlebia sp. MG-60 produces ethanol through the same glycolytic pathway, and that it produces ethanol efficiently by upregulating all genes involved in glycolysis.  Fig. 4, were consistently upregulated in both Phlebia sp. MG-60 and P. chrysosporium under fermenting condition (Fig. 2a) [49].

Orthologous analysis between
In the three-step ethanol fermentation pathway, pyruvate is oxidatively decarboxylated to acetyl-CoA by pyruvate ferredoxin oxidoreductase and pyruvate formate lyase. Acetyl-CoA is then converted to acetaldehyde by a CoAdependent-acetylating acetaldehyde dehydrogenase. Finally, ADH converts acetaldehyde to ethanol [50,51]. The three-step pathway is widespread in bacteria, but not in white-rot fungi. In the present study, pyruvate ferredoxin oxidoreductase, pyruvate formate lyase, and CoA-dependent-acetylating acetaldehyde dehydrogenase were not detected in the DEGs of Phlebia sp. MG-60 under fermenting conditions. These observations indicated that Phlebia sp. MG-60 mainly uses the two-step pathway of ethanol production from pyruvate, and that the PDC and ADH genes play the major roles in ethanol production. Recently, we reported that highly expressing transformants of the PDC gene in the white-rot fungus P. sordida YK-624 showed improved ethanol production [52]. As shown in Additional file 7: Table S5 and Additional file 8: Table S6, two genes of PDC in Phlebia sp. MG-60 increased PDC expression by 5.0-and 6.3-fold, and three ADH genes increased ADH expression by 11.6-, 4.4-, and 4.1-fold under fermenting conditions. Although six ADH genes were identified in this study, three were downregulated under fermenting conditions. These three upregulated ADH genes may have been the main ones used for ethanol production in the fermentation conditions of this study. In contrast, PDC expression was 2.9-and 3.0fold increased, and ADH expression was 2.9-fold decreased in P. chrysosporium based on RNA-seq data. Additionally, eleven genes for aldehyde dehydrogenase [EC: 1.2.1.3], which is responsible for the subsequent oxidation of acetaldehyde into acetate, were identified in Phlebia sp. MG-60. However, all aldehyde dehydrogenase genes of Phlebia sp. MG-60 showed low expression in fermentation conditions (Additional file 4: Table S3), which suggested that low expression of aldehyde dehydrogenase in Phlebia sp. MG-60 might be a cause of high ethanol productivity. During ethanol fermentation, the ethanol produced inhibits the growth and viability of the microorganism [53,54]. Thus, a high level of ethanol tolerance is considered to be important for a high yield of ethanol. It has been reported that alteration in ethanol tolerance could be affected by many factors, such as fatty acid composition and activity of plasma membrane H + -ATPase [55][56][57][58]. In this study, "fatty acid biosynthetic process" (GO: 0006633) in the biological processes category and "ATP binding" (GO: 0005524) in the molecular functions were the most abundant terms in GO enrichment analysis of Phlebia sp. MG-60 under fermenting conditions ( Table 2). These observations indicated that a high resistance of Phlebia sp. MG-60 to ethanol stress might have an effect on its high ethanol productivity.

Conclusions
In this study, we provided novel transcriptomic information on the ethanol producing white-rot fungus Phlebia sp. MG-60, and investigated the genes involved in its high ethanol yield by comparing them to the model