Background

Lignocellulose, a widely distributed, renewable and enormous biomass resource, is one of the most important raw materials for bioethanol production [1]. Bamboo, a lignocellulosic feedstock, is a regenerated biomass material with abundant resources, short growth cycle, high yield and similar chemical composition as wood, and it is considered as a potentially excellent raw material [2, 3]. Many studies have evaluated bamboo lignocellulose degradation and bioethanol production, with a focus on using physical and chemical pre-treatment [4, 5]. However, studies reporting the biodegradation of bamboo lignocellulose using microbes such as bacteria and fungi are scarce.

Lignocellulose hydrolysis, especially cellulose degradation, remains a considerable challenge in lignocellulosic bioethanol production [6]. In nature, numerous examples for lignocellulose degradation are present; of these, phytophagous insects are considered the most notable. In these insects, intestinal symbiotic microbes played important roles in lignocellulose degradation [7]. Therefore, the intestines of phytophagous insects were considered as important locations for isolating lignocellulolytic microbes [8].

Microbial degradation of lignocellulose is a green biological refining method with advantages over physical and chemical methods [9]. The bacterial genus Bacillus is an excellent degrader that exhibits various abilities for degrading lignocellulose biomass, including cellulose, hemicellulose and lignin [10, 11]. Furthermore, genome sequencing, considered an efficient method for investigation of function, has been utilized in lignocellulose degradation research. However, the lignocellulose degradation of Bacillus is still unclear [12]. Dunlap et al. [13] reported that B. oryzicola and B. methylotrophicus, were classified into the B. velezensis group. Recently, the complete genome and genes associated with lignocellulose degradation of several B. velezensis strains were sequenced and are enriched in the genome [14,15,16]. However, its potential application in converting lignocellulose into bioethanol has received little attention.

In the present study, we isolated an endophytic bacteria from the gut of Cyrtotrachelus buqueti that showed a bamboo lignocellulose-degrading ability [17] and sequenced the whole genome of the bacteria B. velezensis LC1, determined the cellulase activities and analysed the ethanol production of bamboo shoot. CAZy genes involved in degradation of lignocellulose were identified through genomic analysis. The chemical changes of the cell wall components were investigated, as well as the hydrolytic and ethanol-fermenting properties of bamboo shoots.

Results and discussion

Identification and cellulose-degrading potential of Bacillus velezensis LC1

Five cellulolytic strains, including PX9, PX10, PX11, PX12 and PX13, which produced clear zones around the colonies after Congo red staining, were isolated from the intestine of C. buqueti on CMC agar. Among the five strains, PX12 exhibited the highest cellulose hydrolysis capacity, with a higher hydrolysis capacity ratio (HCR: 4.71) than PX9 (HCR: 2.41), PX10 (HCR: 1.92), PX11 (HCR: 2.12) or PX13 (HCR: 2.56), as determined using the cellulose hydrolysis assay (Fig. 1a, b; Additional file 1: Figure S1). Based on the HCR ratio, many potent cellulolytic bacteria were previously screened from various regions, such as Geobacillus sp. from a hot spring and Paenibacillus lautus BHU3 from a landfill site [18, 19]. Similarly, PX12 was considered as a good cellulolytic bacterium and was used for further study.

Fig. 1
figure 1

Isolation, identification and enzyme activities of Bacillus velezensis LC1. a Cellulolytic activities of PX12 cultured on carboxymethyl cellulose agar with Congo red stain. b Cellulose hydrolysis capacity ratio of the isolates PX9, PX10, PX11, PX12 and PX13. c Phylogenetic analysis of 16S rRNA sequences among B. velezensis LC1 and the closest Bacillus BLAST hits. d Phylogenetic analysis of house-keeping gene sequences among B. velezensis LC1 and the closest Bacillus BLAST hits. e Endoglucanase activity during the culture for 1st day, 3rd day and 6th day. f Exoglucanase activity during the culture for 1st day, 3rd day and 6th day. g β-glucosidase activity during the culture for 1st day, 3rd day and 6th day. The different normal letters indicate a significant difference in gene expression at different time points with p value at 0.05 level (n = 3)

PX12 was identified and confirmed by 16S rRNA sequencing. Phylogenetic analysis of the obtained PX12 sequence revealed a 99% resemblance with B. velezensis strain JTYP2 (NZ_CP020375.1) (Fig. 1c). Moreover, a house-keeping gene was used for phylogenetic analysis; it showed that the sequence of PX12 had a 98.6% similarity with B. velezensis strain S3-1 (NZ_CP016371.1) (Fig. 1d). Overall, the PX12 was identified as B. velezensis and was named as B. velezensis LC1.

Several B. velezensis strains have been noted for their lignocellulose-degrading abilities [14,15,16]. To investigate the cellulose-degrading ability, B. velezensis LC1 was cultured on CMC agar to determine cellulase activities by the dinitrosalicylic acid spectrophotometric (DNS) method for 6 days (Fig. 1e–g) [20]. The cellulase activities of strain LC1 were then determined. The endoglucanase activity was 0.689 ± 0.011 U/ml at day 1 and increased to 0.752 ± 0.013 U/ml at day 6, which was in accordance with the exoglucanase activity (from 0.359 ± 0.016 U/ml to 0.385 ± 0.022 U/ml), whereas the β-glucosidase activity decreased from day 6 to day 1. Previous studies have reported the cellulase activities of other lignocellulolytic Bacillus strains. For example, Bacillus sp. 275, Bacillus sp. R2, B. velezensis 157 and B. velezensis ZY-1-1 [10, 11, 21, 22] showed similar results as those achieved in our study. This indicated that the strain played a potential role in cellulose degradation.

Genome sequencing and assembly of Bacillus velezensis LC1

The genome contributes to a clear understanding of bacterial decomposition mechanisms of cellulose; thus, the genome of B. velezensis LC1 was analysed to decipher the genetic code involved in cellulose degradation. The complete genome sequence of B. velezensis LC1 was assembled into a ring chromosome with 3,929,782 bp and had a GC content of 46.5% (Fig. 2). A length of 3,502,596-bp genes was found based on gene prediction, and the ratio of gene length/genome was 89.13%. The intergene region/whole genome ratio was 7.19%, and the GC contents of the gene and intergene region were 47.29% and 40.04%, respectively. Furthermore, 4018 CDSs were contained in the genome, and all were assigned functions. CDSs were further annotated in NR, Swiss-Prot, COGs, KEGGs, GO and Pfam, and their numbers were 4018, 3520, 2996, 2186, 2718 and 3315, respectively (Table 1).

Fig. 2
figure 2

The whole genome of Bacillus velezensis LC1. The genome map is composed of seven circles. From the outer circle to inner circle, each circle displays information regarding the genome of (1) forward CDS, (2) reverse CDS, (3) forward COG function classification, (4) reverse COG function classification, (5) nomenclature and locations of predictive secondary metabolite clusters, (6) G+C content and (7) GC skew

Table 1 Genome features of Bacillus velezensis LC1

COGs involved in carbohydrate metabolism

In total, 3046 genes were classified into 2996 COGs, of which carbohydrate transport and metabolism, amino acid transport and metabolism and transcription were the most enriched COGs, which represented 9.46%, 7.62% and 7.29%, respectively (Fig. 3a). To elucidate the function of B. velezensis LC1 in cellulose degradation at the genetic level, specific COGs involved in carbohydrate metabolism were analysed. A total of 222 genes were annotated into carbohydrate metabolism, including 130 COGs, of which the most abundant COGs were COG0366 (alpha-amylase), COG0524 (pfkb domain protein), COG2814 (Major facilitator), COG0726 (4-amino-4-deoxy-alpha-l-arabinopyranosyl undecaprenyl phosphate biosynthetic process), COG1263 (PTS System), COG0477 (major facilitator superfamily), COG1455 (pts system), COG1940 (ROK family) and COG2723 (beta-glucosidase) (Additional file 2: Table S1). COG366 encodes an alpha-amylase that acts on a bond between starch and glycogen, hydrolysing polysaccharides into glucose and maltose [23]. COG2814 is involved in cellular transport of some complexes, such as carbohydrates and amino acids. COG0477, a secondary active transporter, helps to catalyse the transport of various substrates [24, 25]. Moreover, other important COGs in carbohydrate metabolism were annotated, e.g. COG0395 was reported to participate in carbohydrate uptake [26] and COG1109 catalysed the conversion of glucosamine-6-phosphate [27]. The high diversity of function annotations indicated that B. velezensis LC1 had a potent capability in lignocellulose degradation.

Fig. 3
figure 3

Functional categories of Bacillus velezensis LC1. a Clusters of Gene Ontology (GO) annotation. b Clusters of Orthologous Groups of proteins (COGs) annotation. c Clusters of KEGG annotation. d Gene count distributions of carbohydrate-active enzyme families. GH glycoside hydrolases, GT glycosyl transferases, PL polysaccharide lyases, CE carbohydrate esterases, CBM carbohydrate-binding modules, AA auxiliary activities

GO terms annotations

To explain the relevance of the genome of B. velezensis LC1, GO analysis was used to categorize genes into three categories according to matches with known sequences. In three categories, molecular function contained most numerous GO terms and gene number (3730), followed by biological process (Gene number: 3025) and cellular component (Gene number: 1637) (Fig. 3b). In molecular function, the most five pathway was ATP binding (GO:0005524; 314 genes), DNA binding (GO:0003677; 254 genes), transcription factor activity (GO:0003700; 115 genes), metal ion binding (GO:0046872; 114 genes) and hydrolase activity (GO:0016787; 84 genes). Oxidation–reduction process (GO:0055114) and regulation of transcription (GO:0006355) were most pathways in the biological process, and integral component of membrane (GO:0016021), cytoplasm (GO:0005737) and plasma membrane (GO:0005886) pathways in cellular component.

Furthermore, we analysed the GOs that are associated with carbohydrate metabolism. We identified 114 GO items associated with carbohydrate metabolism, including GO:0004553 (hydrolase activity that hydrolyses O-glycosyl compounds), GO:0005975 (carbohydrate metabolic processes) and GO:0016787 (hydrolase activity) (Additional file 3: Table S2).

KEGG annotations

The CDSs of B. velezensis LC1 were submitted to KAAS and KEGG pathways to identify metabolism pathways (Additional file 4: Table S3). As shown in Fig. 3c, of the six classification of KEGG pathways, metabolism contained the most numbers of genes, followed by environmental information processing. In KEGG metabolism annotations of B. velezensis LC1, carbohydrate metabolism and amino acid metabolism, which are considered its main functions, contained 392 and 285 genes, respectively. For these metabolisms, some pathways were dominant, such as sucrose and starch metabolism (ko00500), glycolysis/gluconeogenesis (ko00010), and amino and nucleotide sugar metabolism (ko00520). Forty-one genes were related to ko00500, and common enzyme endoglucanase (EC.3.2.1.4), present in ko00500, was involved in cellulose degradation (Fig. 3c). Starch and sucrose metabolic pathways occurred in B. velezensis LC1, indicating that cellulose could be hydrolysed into cellobiose and ultimately, β-d-glucose. In the genome, 39 genes were found in ko00010, in which d-glucose was phosphorylated into d-glucose-6-phosphate. In addition, ko00010 was linked with other pathways. For example, d-glucose-6-phosphate could be converted to pyruvate, which can be oxidized to acetyl-CoA, having an ability to enter the citrate cycle. Furthermore, ko00520 indicated that glucose from the ko00010 finally entered other pathways under various catalytic reactions. α-d-galactose can be transferred and isomerized in ko00520 and then entered into the ascorbate and aldarate metabolism pathways. Additionally, fructose, 1,4-β-d-xylan, and extracellular mannose were metabolized in ko00520.

The annotation involved in the degradation of lignin or aromatic compounds has also been identified. We identified various enzymes associated with lignin degradation, including oxidoreductase, reductases, dehydrogenases, esterases, thioesterases, transferases and hydrolases. Moreover, 13 monooxygenases, 12 dioxygenases, 2 peroxidases (including one DyP-type peroxidase) and 1 laccase were been identified (Additional file 5: Table S4).

Carbohydrate-active enzyme (CAZyme) annotation

The bCAN carbohydrate-active enzymes (CAZy) annotation algorithm was used to analyse CAZy annotations to identify genes involved in lignocellulose degradation. The results showed that 136 genes were identified from CAZy families and distributed into five subfamilies. In B. velezensis LC1, glycoside hydrolases (GHs), which play key roles in carbohydrates degradation, contained 44 members [28]. Additionally, 38 glycosyl transferases (GTs), 30 carbohydrate esterases (CEs), 3 polysaccharide lyases (PLs), 6 enzymes with auxiliary activities (AAs) and 15 carbohydrate-binding modules (CBMs) were identified (Fig. 3d; Additional file 6: Table S5).

The GH family contained various hydrolases that acted on the glycosidic bond. For example, endoglucanases (EC 3.2.1.4) from Bacillus that have cellulose degradation function usually belonged to the GH5 families [29] (Table 2). In the genome, six GH13s were obliged to hydrolyze starch, such as α-amylase, α-glucosidase and α-glycosidase [30]. Four GH4s, three GH1s and one GH16 exhibited potential cellulose degradation because of enzyme function. Three GH32s were found in the genomic annotations, which contained some hydrolases and levanases, thereby revealing their ability to hydrolyse sucrose [31]. Additionally, Four GH43s, two GH51s and one GH30, responsible for xylan degradation, were considered as other important members for hemicellulose degradation (Table 2). GH43 is an important component of xylan degradation system [32]. One GH5, GH30 and GH1 each were annotated as potential β-glucosidases to utilize cellobiose. One GH53, which hydrolyses (1 → 4)-β-d-galactosidic linkages, was reported as an endo-1,4-β-galactosidase [33]. β-mannosidase was classified into GH26, hydrolysing parts of mannan polysaccharides [34]. Additionally, maltose phosphorylase, which belongs to GH65, was reportedly involved in trehalose degradation [35].

Table 2 Annotated genes encoding lignocellulose-degrading enzymes of B. velezensis LC1, with a focus on enzymes degrading cellulose, hemicellulose and lignin

CEs contributing to the decomposition of xylans were also identified in the genome, including two CE3s, one CE7s, three CE10s, and seven CE4s. CE3 as a potential acetyl xylan esterase enhanced xylan solubilization [36]. The acetylxylan esterase CE7 was considered as a capable xylan-degrading enzyme [37]. CE10 previously exhibited carboxylesterase and xylanase activities involved in hemicellulose degradation [38]. Polysaccharide deacetylases, which play a role in degrading polysaccharides and are classified as a CE4, were also identified. CE4 contained not only highly specific acetylxylan esterases, but also peptidoglycan N-deacetylates involved in chitin degradation [39].

Moreover, two PL1s and one PL9 were annotated to degrade pectin. Pectate lyase (EC 4.2.2.2) usually has (1 → 4)-α-d-galacturonan cleavage function, causing oligosaccharides present at the end [40] (Table 2). AA4, AA6, AA7 and AA10 were included in the genome. AA4 included vanillyl-alcohol oxidases, which could transform some phenols [41]. Additionally, AA7 enzymes were involved in biotransformation or detoxification of lignocelluloses [42]. Finally, we proposed a hypothetical cellulose-degrading and ethanol-producing pathway for B. velezensis LC1 (Table 3).

Table 3 Enzymes in cellulose-degrading and ethanol-producing pathways of B. velezensis LC1

Comparative genomic analysis of CAZymes with other B. velezensis strains

The assembled genome of B. velezensis LC1 was compared to the genomes of other 10 B. velezensis strains, including, B. velezensis S3-1, B. velezensis LS69, B. velezensis JTYP2, B. velezensis DR-08, B. velezensis FZB42, B. velezensis LPL-K103, B. velezensis TB1501, B. velezensis UCMB5036, B. velezensis LB002 and B. velezensis 157. The result showed that GH, GT and PL family numbers were the same in theses strains, while the strain LC1 contained more CE and AA family members and less CBM members (Table 4). The coexistence of these genes suggests that they play important roles in the enzymatic degradation of cellulose and hemicellulose. We consider these degradation enzymes in B. velezensis to have potential use for bioethanol production.

Table 4 Comparative genomic analysis of CAZymes with other B. velezensis strains

Gene expression analysis for B. velezensis LC1 cultured in bamboo shoot powder or glucose medium

Expression profiles of selected B. velezensis LC1 lignocellulolytic enzymes genes were monitored in cultures grown on bamboo shoot powder compared to cultures utilizing glucose. Quantitative real-time PCR (qRT-PCR) investigation results of genes encoding endoglucanase (gene1950), beta-glucanase (gene3931), 6-phospho-beta-galactosidase (gene1280), glucan endo-1,6-β-glucosidase (gene2084), alpha-glucosidase (gene2945), acetyl xylan esterase (gene0351), xylan 1,4-beta-xylosidase (gene1870), arabinoxylan arabinofuranohydrolase (gene1955), alpha-N-arabinofuranosidase (gene2785), and arabinan endo-1,5-alpha-l-arabinosidase (gene2795) are shown in Fig. 4. Transcript levels of all the genes of interest were significantly up-regulated (at least P < 0.05) in BSP cultures compared to glucose cultures. It indicated that these genes involved in BSP degradation by B. velezensis LC1.

Fig. 4
figure 4

Lignocellulolytic enzyme genes relative expression levels of B. velezensis LC1 in the presence of different substrates. G glucose substrate, BSP bamboo shoot powder substrate. gene1950: endoglucanase, gene3931: beta-glucanase, gene1280: 6-phospho-beta-galactosidase, gene2084: glucan endo-1,6-β-glucosidase, gene2945: alpha-glucosidase, gene0351: acetyl xylan esterase, gene1870: xylan 1,4-beta-xylosidase, gene1955: arabinoxylan arabinofuranohydrolase, gene2785: alpha-N-arabinofuranosidase, gene2795: arabinan endo-1,5-alpha-l-arabinosidase. The values represent the means of the three replicates with the standard deviation (SD). Asterisks represent significant differences from the glucose-containing medium (statistical significance: ** P < 0.01)

Cellulose degradation efficiency and fermentation efficiency of bamboo shoots by B. velezensis LC1

Several previously reported genomes of B. velezensis strains contained genes encoding enzymes having lignocellulose-degrading potential [14,15,16]. However, lignocellulose-degrading abilities have not been completely verified in B. velezensis [12]. B. velezensis LC1, isolated from the intestine of C. buqueti, which was reported to be a microflora with lignocellulose-degrading ability, was prepared in a bamboo shoot powder (BSP) degradation assay [17]. We first determined the cellulose degradation efficiency to be 39.32% (Fig. 5a). Furthermore, we determined the glucose and xylose content in the degradation products to be 55.30 ± 1.40 mg/L and 488.81 ± 45.06 mg/L, respectively (Fig. 5b). The reducing sugars in the culture medium was mainly derived from the hydrolysis of cellulose and hemicellulose in BSPs, the reducing sugar content was determined to reflect the degree of conversion of lignocellulose. It indicated that the cellulose and hemicellulose of BSP were degraded with incubation with B. velezensis LC1.

Fig. 5
figure 5

Cellulose-degrading efficiency and fermentation performance of BSP by Bacillus velezensis LC1. a Cellulose degradation efficiency. b Glucose and xylose contents of the hydrolysate. c Ethanol production at 48 h, 60 h, 72 h, 84 h and 96 h

Shimokawa et al. [43] reported that bamboo shoot was an excellent biomass stock for ethanol production due to its high saccharification efficiency. To determine the ethanol production capacity of B. velezensis LC1, a 6-day-treated hydrolysate was prepared for ethanol fermentation using glucose-fermenting S. cerevisiae and xylose-fermenting E. coli KO11. Ethanol production from BSP hydrolysate using B. velezensis LC1 is illustrated in Fig. 5c. The result showed that the ethanol yield continually increased during 48–96 h, and reached 7.21 ± 0.24 g/L at 96 h, while the reducing sugar continually decreased after incubation 48 h. This indicated that B. velezensis LC1 had a potent ability to convert lignocelluloses in bamboo shoot to ethanol.

Conclusions

The genome of a gut symbiotic bacteria, B. velezensis LC1, was sequenced and analysed. The genome comprises a ring chromosome of 3,929,782 bp and has a GC content of 46.5%. A total of 136 CAZyme genes involved in lignocellulose degradation were annotated in the genome, and cellulose-degrading and ethanol-producing pathways were proposed. Moreover, the expression level of some CAZyme genes involved in cellulose and hemicellulose degradation, were up-regulated in bamboo shoot powder substrate. Moreover, a transcriptomic study would perform to identify the role of CAZyme genes in lignocellulose degradation as a future study. The cellulose degradation of bamboo shoot by the strain was 39.32%, and the hydrolysate was subjected to ethanol fermentation; the ethanol yield was 7.2 g/L at 96 h. This study suggests that B. velezensis LC1 could be used for bamboo lignocellulose degradation and bioconversion of lignocelluloses to ethanol.

Methods

Insect sample, isolation and identification of cellulolytic bacteria

Cyrtotrachelus buqueti specimens were sampled from the Muchuan County (E 103° 98′, N 28° 96′), China. Gut was extracted from individual insects and stored at 4 °C for isolation of bacteria. The gut was blended, homogenized and serially diluted (10−1 to 10−9), and inoculums of 10−7 to 10−9 dilution were plated on carboxymethyl cellulose (CMC) agar [44] for cellulolytic bacteria screening. Congo red dye was used to screen the cellulose-degrading bacteria as described by Teather and Wood [45]. Hydrolysis zone = clearance zone/colony diameter.

Molecular characterization of bacterial isolate

The 16sRNA V3–V4 region was considered for amplification for bacterial identification and amplified using bacterial primers (27F 5′-AGAGTTTGATCMTGGCTCAG-3′ and 1492R 5′-TACGGYTACCTTGTACGACTT-3′); moreover, F 5′-GCCCATATTTCCATTTCTCC-3′ and R 5′-GTGGTCGTTATGGAAATAAAGG-3′ were selected for amplification of the house-keeping gene rpoB. Thermocycling conditions were as follows: initial denaturation at 94 °C (2 min), followed by 30 cycles of denaturation at 94 °C (30 s), annealing at 55 °C (30 s) and extension at 72 °C (100 s), ultimately extending at 72 °C (2 min). The amplicons were checked by electrophoresis on a 1% agarose gel. MEGA5 was used to establish phylogenetic relationships among the obtained sequence and reference genes that were retrieved in NCBI GenBank through the neighbour-joining method.

Genomic DNA extraction and genome sequencing

Genomic DNA of B. velezensis LC1 was extracted according to the CTAB method based on the protocol described by Lin [46], and it was purified using the Wizard Genomic DNA Purification Kit (Vazyme Biotech Co., Ltd, Nanjing, China). Whole-genome sequencing, performed by Shanghai Majorbio Bio-pharm Technology Co., Ltd (Shanghai, China), was fulfilled through PacBio systems with an average genome coverage of 100 × for the raw data [47]. HGAP 2.0 was used to filter and assemble reads to the scaffold.

Function annotation of B. velezensis LC1

Glimmer 3.02 was used to predict coding DNA sequences (CDSs). A BLAST search was then conducted for CDSs in some widely used databases: NCBI non-redundant (NR) database, Swiss-Prot, COGs, Gene Ontology (GO) and KEGGs [48, 49]. GO, an important bioinformatics tool, unified expressions of gene and genetic products in all species [50]. KEGGs, a database resource to understand high-level functions, also could analyse metabolic pathways. Additionally, the CAZymes were identified, classified and annotated using CAZymes database (CAZyDB: http://www.cazy.org/).

Quantitative real‑time PCR

The primers used for qRT-PCR in this study were performed in Additional file 7: Table S6. PCR was performed under following conditions: 10 min initial denaturation at 95 °C, 45 cycles of 5 s denaturation at 95 °C, 50–65 °C anneal for 30 s, and 30 s extension at 55 °C, finally 10 s extension at 95 °C. All experiments were performed three times and analysed by 2− ΔΔCT Method. 16S rRNA was used as reference gene.

Bamboo shoots degradation by B. velezensis LC1

To obtain fermentable sugar from bamboo shoots, B. velezensis LC1 was used to degrade the bamboo shoot powder (BSP), which was prepared as described by Luo et al. [17]. B. velezensis LC1 was cultured in liquid medium at a pH of 7.2, temperature of 37 °C and 200 rpm for 6 days. The culture medium comprised BSP 10 g/L, (NH4)2SO4 2 g/L, K2HPO4 1 g/L, KH2PO4 1 g/L, MgSO4 0.2 g/L, CaCl2 0.1 g/L, FeSO4·7H2O 0.05 g/L and MnSO4·H2O 0.02 g/L. The reaction mixture was incubated at 100 °C for 30 min to terminate the reaction and centrifuged at 13,000 rpm for 10 min, and the hydrolysate-containing supernatant and deposit were collected separately. The obtained deposit was dried and weighed to determine the cellulose levels using the Van Soest method [51]. The hydrolysate-containing supernatant was used to determine the glucose and xylose contents according to the NREL methods [52]. The BSP hydrolysate supernatant was sterilized and stored at − 20 °C for ethanol fermentation.

Fermentation

Saccharomyces cerevisiae, a glucose-fermenting strain, was pre-cultured in YPD at 30 °C for 24 h, and E. coli KO11, a xylose-fermenting strain, was cultured in LB at 37 °C for 24 h. S. cerevisiae (50 g/L) and E. coli KO11 (100 g/L) were prepared after centrifugation of pre-cultured cells. The initial cell concentrations were 0.33 g/L (S. cerevisiae) and 1.0 g/L (E. coli KO11) at the beginning of fermentation. 100 ml hydrolysate was used for ethanol fermentation in 250-ml serum bottles under anaerobic conditions. The fermentation was performed at 37 °C and 200 rpm for 96 h. After 48 h fermentation, ethanol production was monitored every 12 h. The ethanol concentration was determined via HPLC. All reactions were repeated three times.