T. rubra YIM 77501^T (= DSM 45614^T = CCTCC AA 2011014^T) was cultured on R₂A medium (yeast extract, 0.5 g/l; peptone, 0.5 g/l; glucose 0.5 g/l; soluble starch, 0.5 g/l; casein acid hydrolysates, 0.5 g/l; sodium pyruvate, 0.5 g/l; K₂HPO₄, 0.3 g; MgSO₄, 0.024 g/l) at 50°C. Genomic DNA was purified from 100 ml of mid exponential phase R₂A cultures using a MasterPure Gram-Positive DNA Purification kit (Epicentre MGP04100) following the standard DNA isolation procedure recommended by the manufacturer with modifications (Wu et al., 2009). Genomic DNA was sequenced using Illumina technology (Bennett, 2004) at the Beijing Genomics Institute (BGI Tech Solutions, Shenzhen, China). After the raw sequences were trimmed and their quality filtered (Q > 30), the clean reads with high quality were assembled using de Bruijn graphs in SOAP de novo v.1.05 (http://www.seekbio.com/soft/2754.html) with the K-mer parameter set to 41 (Li et al., 2008) and draft genomes were generated. The SOAPaligner v.2.21 alignment tool (http://soap.genomics.org.cn/soapaligner.html#down2) was used to align these reads against the de novo scaffolds to map reads and account for single nucleotide errors (Gu et al., 2013). Glimmer v3.0 (Chen et al., 2011) was used for gene prediction in assembled sequences of strain T. rubra. The sequence data described here have been deposited in JGI IMG (Submission ID: 105093) and DDBJ/ENA/GenBank (Accession number: MSZZ00000000). Carbohydrate-active enzymes (CAZymes) of T. rubra were determined using the CAZymes Analysis Toolkit (http://mothra.ornl.gov/cgi-bin/cat/cat.cgi; Petit et al., 2015). The results of GHs (Glycoside hydrolase families) were analyzed using the HMMER software based on the Pfam database (http://pfam.xfam.org/; Finn et al., 2011).

T. rubra YIM 77501^T was cultured in a modified form of a previously described R₂A medium containing the following (g/l): yeast extract, 0.5 g; peptone, 0.5 g; casein acid hydrolysates, 0.5 g; sodium pyruvate, 0.5 g; K₂HPO₄, 0.3 g; MgSO₄, 0.024 g; with 2 g/l CMC sodium as the test sample (R₂A-CMC medium) and with 2 g/l glucose as control (R₂A-glucose medium), with the pH adjusted to 7.0 using KOH. The colony of T. rubra was inoculated to R₂A-CMC and R₂A-glucose media. The growth curves under different conditions were determined (Figure S3). After incubation to mid exponential phase (at 30°C 180 rpm for 7 days, 40°C 180 rpm for 3.5 days, and 50°C 180 rpm for 2 days), cells were harvested by centrifugation for 5 min at 12,000 × g at 4°C. Three independent biological replicates were performed for each condition. The cell biomass was washed with phosphate buffered saline, re-suspended in 100 μl TE buffer pH 8 (EMD Chemicals) containing 2 mg/ml lysozyme (Merck, USA), and incubated at 37°C for 40 min. Total RNA was isolated using the RNeasy RNA purification kit (QIAGEN, Germany) according to the manufacturer's instructions. Contaminating DNA was removed with RNase-free DNase I (QIAGEN, Germany). Ribosome RNA in total RNA preparation was removed by using the Ribo-Zero™ Magnetic Kit for Gram-Positive Bacteria (Epicentre, USA). The quality of RNA samples was assessed on the Agilent Bioanalyzer 2100 system. Library construction and Illumina sequencing (Illumina HiSeq™ 2500) were performed at the Beijing Novogene Biological Information Technology Co., Ltd., Beijing, China (http://www.novogene.cn/).

An RNA-seq analysis was performed according to the protocol recommended by the manufacturer (Illumina Inc.). Raw reads of fastq files were first processed through in-house Perl scripts. In this step, clean reads were utilized by removing reads containing adapter, reads containing poly-N and low quality reads from raw data. At the same time, the Q20, Q30, and GC contents of the clean data were calculated. All the following analyses were based on clean data with high quality. The reads from different conditions were mapped to the whole-genome of strain T. rubra using Bowtie 2.2.3 (Langmead and Salzberg, 2012). HTSeq v0.6.1 was used to count the read numbers mapped to each gene. Then, the FPKM of each gene was calculated based on the length of the gene and reads count mapped to this gene (FPKM, expected number of fragments per kilobase of transcript sequence per millions base pairs sequenced, considers the effect of sequencing depth and gene length for the reads count at the same time and is currently the most commonly used method for estimating gene expression levels; Trapnell et al., 2009). Here, FPKM > 1 means gene expression.

Differential expression analysis of two conditions (cultures in CMC media and glucose media) at the same temperature (three biological replicates per condition) were performed using the DESeq R package (1.18.0). DESeq provided statistical methods for determining differential expression in digital gene expression data using a model based on a negative binomial distribution. The resulting p-values were adjusted using the Benjamini and Hochberg approach for controlling false discovery rate. Genes with an adjusted p < 0.05 found by DESeq were assigned as differentially expressed (Wang et al., 2010).

The full-length sequences of cellulase genes were amplified from T. rubra genomic DNA. Chromosomal DNA of T. rubra was isolated using the Ezup Bacteria DNA Kit (Sangon Biotech, China) according to the manufacturer's instructions. Base on genome sequences, primers of cellulase genes (Table S1) were designed by using Primer Premier 5 (http://www.bioprocessonline.com/doc/primer-premier-5-design-program-0001). The complete ORFs of cellulase genes were amplified by PCR using the TransStar FastPfu Fly DNA Polymerase (TransGen Biotech, China). Amplification was performed for 34 cycles of 98°C for 20 s, 65°C for 20 s, and 72°C for 1 min, 72°C for 5 min with initial 2 min denaturation at 98°C. PCR products were ligated into pEASY-Blunt E1 vector (TransGen Biotech, China) and transformed into Escherichia coli DH5α. After sequencing verification, the entire cellulase gene was confirmed; positive recombinant vectors were transformed into E. coli BL21 (DE3) for cellulase gene expression.

Transformants were cultured overnight in LB culture medium with 100 μg/ml ampicillin at 37°C and 220 rpm. Then, 1 ml of the cells was added to 100 ml LB medium at 25°C and 220 rpm. During cultivation, isopropyl β-D-1-thiogalactoside (IPTG) was added to a final concentration of 1 mM at mid-exponential phase (OD600 ≈ 0.6) and followed by further incubation 8 h at 25°C with 220 rpm. Cells were harvested by centrifugation at 4,000 × g and suspended in 20 ml PBS buffer (pH 8.0).

After ultrasonic cell disintegration and centrifugation at 12,000 × g for 30 min at 4°C, cell-free extracts were applied to a Ni-chelating affinity column (GE, USA) because the proteins possess an N-terminal His-tag. Then, the extract was washed with five column volumes of buffer A (20 mM sodium phosphate, 0.3 M NaCl, pH 8.0), followed by 10 column volumes of buffer A with 20 mM imidazole, pH 8.0, and was eluted with buffer A with 200 mM imidazole, pH 8.0. The eluted protein was used for enzyme characterization.

The homogeneity of the purified enzyme was monitored by SDS-PAGE using 10% acrylamide gels. Proteins were visualized by Coomassie brilliant blue R-250 as described by Liu et al. (2010). Protein concentration was determined with a protein assay kit (Sangon Biotech, China) using bovine serum albumin as a standard. Activity against CMC was determined by measuring the release of reducing sugar, with 1% (w/v) CMC as substrate, by the 3,5-dinitrosalicylic acid (DNS) assay (Miller, 1959). One unit (U) of CMCase activity was defined as the amount of enzyme to release 1 μmol glucose-equivalent reducing sugars per minute. β-glucosidase activity was assayed using a 200 μl reaction mixture containing 2.5 mM p-nitrophenyl-β-D-glucopyranoside (pNPGlu) (Sigma, St. Louis, MO, USA). After 5 min of incubation at optimal temperature, the reaction was stopped by adding 0.6 ml of 1 M Na₂CO₃ (Yang et al., 2015). The p-nitrophenol was determined by monitoring the absorbance at 405 nm (Harnpicharnchai et al., 2009). One unit of β-glucosidase activity is equivalent to 1 μmol of p-nitrophenol released from the pNPGlu in 1 min. The effect of temperature was determined at different temperatures from 20 to 90°C in optimal pH. Thermal stability of enzymes was determined by incubating equivalents of purified enzyme solutions for varied length of time intervals at 30, 40, and 50°C. The residual activity was determined by the standard method.

The draft genome sequence of T. rubra YIM 77501^T consisted of 191 scaffolds, with 8,233,369 bp. The GC content was 71.7% and the genome contained 8,114 coding sequences, with an average length of 914 bp. The general genomic features of T. rubra are listed in Table S2. Among the predicted genes, 60% (4867 genes) have been assigned a function, and 40% (3,247 genes) have an unknown function. In addition, genes encoding one rRNA operon were found in proximity to the origin of function, and there were 54 tRNAs (Table S2).

A CAZymes analysis was conducted to identify potential enzymes with plant cell-wall degradation ability. By applying this analysis, a total of 403 glycoside hydrolases (GHs) were distributed into 58 families, 226 carbohydrate-binding modules (CBMs) were distributed into 20 families, 20 polysaccharide lyases (PLs) were distributed into 7 families, 109 carbohydrate esterases (CEs) were distributed into 10 families, and 339 glycosyl transferases (GTs) were distributed into 26 families, are encoded in the genome of T. rubra (Figure 1). After analysis of GHs by using the HMMER software based on the Pfam database, 108 GH genes were distributed into 36 families (Figure 1). Twenty-seven cellulase genes were distributed to 6 GH families. The function of these 6 GH (GH1, GH3, GH5, GH6, GH9, and GH48) families may directly relate to cellulose digestion.

Global gene expression profiles of T. rubra cultured on R₂A-glucose and R₂A-CMC media under different temperatures (30, 40, and 50°C) were examined using transcriptome sequencing. Finally, there are 406.48 million clean reads and 50.81 GB of RNA-seq data in treated and control strains after quality filtering (Error rate = 0.01%, Q20 > 98%, Q30 > 95%). More than 99% of the clean reads were mapped to the T. rubra YIM 77501^T genome. The GC content of clean reads is within the scope of 68.2–69.2% for all samples (Table S3). All of the Pearson's correlations between biological replicates were >0.95 and indicated a high reliability of the experiment and rationality of sample selection (Figure S4). Comparing percentage of reads mapped to the genome regions at different temperatures, more reads were mapped to intergenic region at higher temperatures. Cultures of R₂A-CMC media contained ~8.9, 12.3, and 21% reads mapping to intergenic regions at 30, 40, and 50°C, respectively. Cultures of R₂A-glucose media contained ~10.8, 19.9, and 22.1% reads mapping to intergenic regions at 30, 40, and 50°C, respectively (Figure S5). These results revealed that intergenic regions might play some functions in adapting to higher temperatures for strain T. rubra.

Total 18, 19, and 21 cellulase genes showed expression (FPKM > 1) at 30, 40, and 50°C, respectively. In these genes, seven up-regulated cellulase genes belonging to GH1, GH3, and GH6 were detected, comparing the cellulase genes expressions of CMC cultures with glucose cultures. This means that not all cellulase genes expressed during our culture conditions and the presence of CMC were no need for expression of some cellulase genes. Comparing the cellulase genes expressions of CMC cultures with glucose cultures at same temperature, four cellulase genes (TrBG2, TrBG3, TrBG4, and ThrCel6B) were up-regulated at 30, 40, and 50°C (p < 0.01). The rates of gene expression of TrBG2, TrBG3, TrBG4, and ThrCel6B were relatively higher at 50°C; on the contrary, they were relatively lower at 40°C. One cellulase gene (TrBG1) was especially up-regulated at 30°C (p < 0.01); two cellulase genes (TrBG5 and ThrCel6A) were merely up-regulated at 50°C (p < 0.01; Figure 2). These findings suggested that culture temperature can affect cellulase gene expressions for T. rubra.

After data analysis, 7,084 and 7,132 genes expressed (FPKM > 1) when T. rubra was cultured at 30°C in R₂A-CMC and R₂A-glucose media, respectively; 7,009 and 6,839 genes were expressed (FPKM > 1) at 40°C in R₂A-CMC and R₂A-glucose media, respectively; 7,345 and 7,228 genes were expressed (FPKM > 1) at 50°C in R₂A-CMC and R₂A-glucose media, respectively, (Figure 3, Table S4). The expression of different genes in T. rubra showed high fluctuation when cultured at different temperatures on R2A-CMC and R2A-glucose media. High expression genes was observed (FPKM > 60), when T. rubra was cultured at 30 and 50°C as compared at 40°C. The genes 2,725 and 2,750, 2,231 and 2,543, 2,835 and 2,852 showed high expression (FPKM > 60) when T. rubra was cultured at 30, 40, and 50°C on R2A-CMC and R2A-glucose media, respectively (Table S4).

The genes expression profiles displayed remarkable differences when T. rubra was cultured at different temperatures in the same media (R₂A-CMC or R₂A-glucose). Comparing 30°C with 40°C, 4,473 genes (2,253 genes up-regulated and 2,220 genes down-regulated) were significant differentially expressed in R₂A-CMC media and 3,894 genes (2,093 genes up-regulated and 1,801 genes down-regulated) were significant differentially expressed in R₂A-glucose media. Comparing 30°C with 50°C, 4,918 genes (2,489 genes up-regulated and 2,429 genes down-regulated) were significant differentially expressed in R₂A-CMC medium and 4,988 genes (2,510 genes up-regulated and 2,478 genes down-regulated) were significant differentially expressed in R₂A-glucose medium. Comparing 40°C with 50°C, 4,293 genes (2,219 genes up-regulated and 2,074 genes down-regulated) were significant differentially expressed in R₂A-CMC medium and 3,760 genes (1,805 genes up-regulated and 1,955 genes down-regulated) were significant differentially expressed in R₂A-glucose medium (Figure S6A). The functions of these differentially expressed genes at different temperatures are mainly reflected in the biological process, cellular component, and molecular function (Figure S6B).

Comparing genes expression profiles of samples from R₂A-CMC and R₂A-glucose media at different temperatures, 3,398 genes (1,706 up-regulated genes and 1,692 down-regulated genes) were found at 30°C, 3,560 genes (1,929 up-regulated genes and down-regulated 1,631 genes) were found at 40°C, and 2,869 genes (1,412 up-regulated genes and 1,457 down-regulated genes) were found at 50°C. Among these differentially expressed genes at different temperatures, 1,278 genes were up- and down- regulated at all temperatures, and 732, 504, and 577 genes were specifically expressed at 30, 40, and 50°C, respectively (Figure 4, Figure S7). This also showed more genes were specific expression when T. rubra is cultured at 30 and 50°C than at 40°C.

TrBG2, TrBG3, TrBG4, ThrCel6B, TrBG1, TrBG5, and ThrCel6A were cloned and expressed in E. coli BL21. TrBG1, TrBG2, TrBG3, TrBG1, TrBG5, and ThrCel6A were purified (Figure S8) and their enzyme activities were tested (Figure S9). The kinetic parameters of these enzymes are shown in Table S5. The enzyme activities of TrBG4 was tested using crude enzymes. Based on experimental results, TrBG1, TrBG2, TrBG3, TrBG4, and TrBG5 showed β-glucosidase enzyme activity; ThrCel6A showed CMCase enzyme activity. The optimal temperatures of TrBG1, TrBG2, TrBG3, TrBG4, TrBG5, and ThrCel6A were 40, 60, 50, 60–70, 50–60, and 70°C, respectively. TrBG1 showed high relative enzyme activity at 30°C (~40%). TrBG2, TrBG3, and TrBG4 showed more than 20% relative enzyme activity at 30–50°C ThrCel6B cannot be expressed in E. coli (by expression vector). TrBG5 and ThrCel6A showed high relative enzyme activity at 50°C (~100 and 78%, respectively; Figure 5).

Comparing effects of temperatures on enzyme stability, TrBG1, TrBG2, TrBG3, TrBG4, TrBG5, and ThrCel6A kept more than 70% relative enzyme activity after incubating at 30 and 40°C for 6 h. TrBG2 and TrBG3 lost more than 50% relative enzyme activity after incubating at 50°C for 6 h, while TrBG1 and TrBG5 lost most enzyme activity (more than 80%) after incubating at 50°C for 3 h. These results revealed TrBG1 and TrBG5 may be unstable at 50°C in vitro. TrBG4 and ThrCel6A were stable (kept more than 90% relative enzyme activity) after incubating at 50°C for 6 h (Figure 6).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.