Transcription Analysis of Recombinant Trichoderma reesei HJ-48 to Compare the Molecular Basis for Fermentation of Glucose and Xylose

Profiling the transcriptome changes involved in xylose metabolism by the fungus Trichoderma reesei allows for the identification of potential targets for ethanol production processing. In the present study, the transcriptome of T. reesei HJ-48 grown on xylose versus glucose was analyzed using nextgeneration sequencing technology. During xylose fermentation, numerous genes related to central metabolic pathways, including xylose reductase (XR) and xylitol dehydrogenase (XDH), were expressed at higher levels in T. reesei HJ-48. Notably, growth on xylose did not fully repress the genes encoding enzymes of the tricarboxylic acid and respiratory pathways. In addition, increased expression of several sugar transporters was observed during xylose fermentation. This study provides a valuable dataset for further investigation of xylose fermentation and provides a deeper insight into the various genes involved in this process.


Ethanol Fermentation
To study the effects of different carbon sources, specifically glucose and xylose, on ethanol fermentation, fermentation by the recombinant T. reesei strain HJ48 was performed anaerobically in fermentation medium (FM) supplied with 50 g/l glucose (GM) and 50 g/l xylose (XM). As shown in Figs. 1A and 1B, the recombinant strain HJ48 showed different fermentation modes depending on the carbon source of the fermentation medium. When the recombinant strain HJ48 was grown in GM on anaerobic conditions, the glucose was almost consumed by HJ48 after 72 h of cultivation. Meanwhile, an ethanol concentration of 4.8 ± 2 g/l was detected.
When xylose was provided as a sole carbon source, half of the xylose was consumed by HJ48 after 72 h of cultivation, demonstrating that the recombinant strain HJ48 consumes xylose at a slower rate than it does glucose after 72 h of cultivation. Additionally, an ethanol concentration of 1.1 g/l was detected. The maximum ethanol production of the recombinant strain HJ48 achieved in fermentation with XM was lower than that observed with GM.

Transcriptome Analysis of T. reesei Strain HJ48 Growth on Xylose Versus Glucose
The transcriptome of T. reesei growing on glucose and xylose was analyzed using RNA-sequencing after 72 h of fermentation under anaerobic conditions. The total RNA of the recombinant strain HJ48 was extracted, and RNAsequencing was conducted using Illumina HiSeq2000. After removing low quality reads, clean reads (20.56 G) were obtained, and were used as the basis of all downstream studies. The sequence data produced in this study can be accessed [GEO: GSE84393]. Detailed information on the assembly process is shown in Table 2. Clean reads were mapped to the T. reesei genome. The normalized expression values of each annotated gene were calculated   1, G: glucose, X: xylose;"-1","-2" and "-3" represent biological replicates. 2, Q20 and Q30 indicate that the rates of correct base recognition were 99% and 99.9%, respectively.
using FPKM (expected number of Fragments Per Kilobase of transcript sequence per Million base pairs sequenced). In this study, only the genes with FPKM >1 were recognized as potentially expressed. To study significantly upregulated and downregulated genes during growth on either glucose or xylose, only the genes with q-values less than 0.05 were selected for further analysis. Overall, 5,065 genes showed significantly different expression when grown on xylose compared to that on glucose. We observed that 2,584 genes were upregulated, whereas 2,481 genes were downregulated (Fig. 2). These results suggest that there were significant differences in T. reesei gene expression under glucose and xylose growth conditions. We next performed gene ontology (GO) term analysis to classify the functions of the differentially expressed genes. Surprisingly, we did not identify any significant enrichment in GO terms (p < 0.05) for the genes found to be downregulated during growth on glucose or xylose. Table 3 shows the GO terms (p < 0.05) found to be enriched for the genes that were significantly upregulated. Based on common biological properties, the identified genes could be classified into three groups: biological processes, cellular components, and molecular functions (Fig. 3). As shown in Fig The GO terms cytoplasm, structural constituent of ribosome, ribosome, ribonucleoprotein complex, and cellular protein metabolic process, were all found to be enriched within the group of upregulated genes involved in the cellular component (Fig. 4). The group of upregulated genes for cellular component seems to directly correlate to the efficiency of cell growth. The genes     [tre01200], starch and sucrose metabolism [tre00500], and fructose and mannose metabolism [tre00051], although this enrichment was not statistically significant (FDR ≤ 0.05).

qPCR
In order to assess the accuracy of the RNA-Seq results, qPCR was implemented. Thirteen genes involved in different metabolism pathways were randomly selected (S1). The values shown are the mean values of three parallel experiments. As expected, the results of qPCR were consistent with those of RNA-Seq, indicating that the RNA-Seq results were credible (Fig. 5).

Analysis of Genes Involved in Central Carbon Pathways
The fermentation of xylose to ethanol is achieved in T. reesei via central carbon metabolism pathways that consist of the oxidoreductive pathway, the pentose phosphate pathway, glycolysis, and the ethanol fermentation pathway. To gain insight into the genes regulated by xylose, gene expression related to the central carbon metabolism was examined, using glucose as a control. As expected, there were a large number of differentially expressed genes between xylose and glucose ( Table 4, Fig. 6).
Xylose must first be converted into xylulose, which is then phosphorylated by xylulokinase. Xylose reductase  (XR), the first enzyme in the oxidoredutive pathway, is fundamental for xylose utilization [22]. Interestingly, the expression of xylose reductase and xylitol dehydrogenase (XDH) was found to be upregulated in this study. The xylulose kinase (XK) however, was not differentially expressed in T. reesei during xylose fermentation. The pentose phosphate pathway is subdivided into two biochemical branches known as the oxidative and nonoxidative pentose phosphate pathway [23]. In the oxidative component of the pentose phosphate pathway, the enzymatic reactions are considered unidirectional. In order to help replenish NADPH-reducing equivalents, the enzymes glucose-6-phosphate dehydrogenase (G6PDH) and NADPH-producing 6-phosphogluconate dehydrogenase (6PGDH) are significantly upregulated with xylose in S. cerevisiae during the stress response [24]. However, we observed that the enzymes of the oxidative component of the pentose phosphate pathway were expressed at a relatively high level during glucose fermentation. The genes encoding the most important enzymes in the non-oxidative pentose phosphate pathway, the transaldolase (TAL) and transketolase (TKL), were significantly downregulated with xylose. In contrast, other enzymes in the non-oxidative pentose phosphate pathway, such as ribose 5-phosphate isomerase A, ribokinase, and ribose-phosphate pyrophosphokinase (PRPS), were expressed significantly higher with xylose compared to that with glucose.
The expression levels of several genes in the tricarboxylic acid cycles were upregulated with xylose compared to glucose. ATP-citrate synthase subunit 1 (ACL1) and ATP-citrate synthase subunit 2 (ACL2), which catalyze the condensation of acetyl coenzyme A and oxaloacetate to form citrate, were upregulated on xylose. Aconitate hydratase (ACO), which is repressed by ethanol, was expressed at a relatively low level during the xylose fermentation. The expression level of isocitrate dehydrogenase remained unchanged, whereas the expression of isocitrate dehydrogenase subunit 1 (IDH1), a mitochondrial NADP + -dependent enzyme, decreased with xylose. As for the KGD genes that encode 2-oxoglutarate dehydrogenase, the expression level of KGD1 and KGD2 encoding 2-oxoglutarate dehydrogenase E1 and 2-oxoglutarate dehydrogenase E2 components, respectively, were unchanged. Succinyl-CoA synthetase, various subunits of succinate dehydrogenase, as well as fumarate hydratase were all significantly upregulated with xylose. The expression level of cytoplasmic malate dehydrogenase (MDH), which is involved in the gluconeogenesis and glyoxylate cycles, was found to be unchanged. We also observed that genes encoding enzymes of the tricarboxylic acid cycle and respiration were upregulated during xylose metabolism. Moreover, Hap2, a subunit of the transcriptional activator complex Hap2/3/5, was upregulated with xylose under anaerobic conditions. In contrast, Hap3 was found to be upregulated during glucose fermentation. We next analyzed the expression level of genes involved in glycolysis and alcohol fermentation during xylose and glucose fermentation. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of hexokinases, which catalyze the conversion of glucose to glucose-6-phosphate, revealed that three genes were differentially expressed (TRIREDRAFT_73665, TRIREDRAFT_80231, TRIREDRAFT_79677). TRIREDRAFT_73665 and TRIREDRAFT_80231 were more highly expressed during xylose fermentation than during glucose fermentation. In contrast, the gene TRIREDRAFT_79677 was expressed more highly with glucose than with xylose. In S. cerevisiae, the expression of hexokinase1 and glucokinase was upregulated under aerobic and anaerobic conditions during xylose growth [25], and during growth on non-fermentable carbon sources [26]. The enzymes of glucose-6-phosphate isomerase (GPI) and fructose-1,6-bisphosphatase 1 (FBP1) are the major regulatory enzymes in the gluconeogenesis pathway. FPB1 was found to be upregulated in T. reesei during xylose fermentation, while GPI was downregulated with xylose. Pyruvate kinase (PYK), which is also known to be required in gluconeogenesis pathway, was found to be upregulated with xylose.
The expression levels of several enzymes in the lower half of the glycolytic pathway were unchanged, with the exception of enolase (ENO) and fructose-bisphosphate aldolase (FBA), which were downregulated, and pyruvate kinase and phosphoenolpyruvate carboxykinase (PCK), which were upregulated, with xylose compared to that with glucose. In the ethanol fermentation pathway, several genes showed significantly different expression. The isozymes of the pyruvate decarboxylase (PDC) gene TRIREDRAFT_121534 were upregulated with xylose, whereas expression of TRIREDRAFT_59267 was not significantly different. Transcripts of ADH, which encodes alcohol dehydrogenase, decreased with xylose. The result is consistent with our previous finding that only one alcohol dehydrogenase gene (TRIREDRAFT_22633), which catalyzes the reduction of acetaldehyde to ethanol, was upregulated in T. reesei during ethanol fermentation. Nevertheless, ethanol production is catalyzed by multiple alcohol dehydrogenase isozymes in S. cerevisiae [27]. Most of the genes involved in acetyl-CoA formation were upregulated during xylose fermentation, including pyruvate dehydrogenase E2 component (PDC-E2), dihydrolipoamide dehydrogenase (DLD), aldehyde dehydrogenase (ALD), and acetyl-CoA synthetase (ACS). In the pyruvate metabolism pathway, the expression of genes NAD-dependent malate dehydrogenase (MDH), fumarate hydratase (FH), and malate synthease (MS) was upregulated during xylose fermentation. Importantly, NAD-dependent malate dehydrogenase is the key enzyme involved in malate synthesis [28]. The increased expression of genes involved in malate synthesis suggests that the malate concentration is perhaps elevated during xylose fermentation.

Analysis of Genes Involved in Sugar Transport
In the present study, several of the transporter genes in T. reesei were differentially expressed between xylose fermentation and glucose fermentation (Table 5). Among all the differentially expressed sugar transporter genes, 24 genes were more highly expressed during glucose fermentation, whereas 15 genes were more highly expressed during xylose fermentation. SNF3, a glucose receptor that generates a signal for induction of intracellular HXT expression, was downregulated with xylose. The glucose transporter Hgt-1 was upregulated with xylose. Rgt1, a transcription factor that regulates expression of HXT genes during glucose fermentation, was upregulated with glucose in S. cerevisiae.

Discussion
We demonstrated from the genomic level the enormous advantages of using the filamentous fungus, T. reesei, over the popular yeast, S. cerevisiae, in ethanol fermentation from the pentose lignocellulosic sugar, xylose. Our T. reesei strain was able to utilize and ferment about 25 g/l of xylose, producing 1.1 g/l of ethanol under anaerobic condition after 72 h (Fig. 1). We attributed this successful xylose utilization/fermentation to the presence and expression of the essential genes responsible for adequate xylose utilization as sole carbon source in T. reesei (Figs. 5 and 6). Moreso, our T. reesei strain also wonderfully utilized and fermented glucose as carbon source, producing a peak ethanol concentration of 4.8 g/l from an initial 50 g/l glucose by 96 h (Fig. 1). These results therefore indicate the potentials and efficiency of T. reesei to both use hexose and pentose sugars in bioethanol production.
Xylose is converted to xylulose in filamentous fungi via the oxidoreductive pathway, which includes two reactions. First, xylose is reduced to xylitol by a NAD(P)H-dependent xylose reductase. Then, xylitol is oxidized to xylulose by a NADP + -dependent xylitol dehydrogenase [29,30]. S. cerevisiae is unable to use xylose as a sole carbon source due to its lack of metabolic pathways for xylose utilization. This has led to the generation of several genetically modified strains of S. cerevisiae in recent years to increase the efficiency of xylose utilization, including expression of native pentose phosphate pathway enzymes such as xylose reductase and xylitol dehydrogenase [31,32]. Unfortunately, the expression of these enzymes in these modified yeast strains has been generally poor. Yang et al. reported a promising transformation of the industrial S. cerevisiae KF-7 strain to utilizing xylose. Better xylitol production from xylose was recorded when the xylose reductase gene, XYL1, was overexpressed with lower xylose specificity. However, increasing the copy number of XXL1 provided little improvement in xylitol production by the yeast and this was partially attributed to inadequate cofactor regeneration [33].
On the contrary, the genome of T. reesei contains all the genes for the metabolic pathways needed for xylose utilization. It is interesting to note that both xylose reductase and xylitol dehydrogenase were highly expressed in T. reesei during growth on xylose as sole carbon source in this study (Figs. 5 and 6). Similarly, the T. reesei strain created by Hong et al. with an increased copy number of xylose reductase gene was demonstrated to show higher reductase expression, unlike that of S. crevisiae, which consequently resulted in increased xylitol production and the amount of xylose consumed [34]. These results therefore highlight the enormous potentials of adopting T. reesei for consolidated bioprocessing.
The efficiency of the pentose phosphate pathway has been considered a barrier to efficient xylose utilization by S. cerevisiae, with the reactions of the non-oxidative pentose phosphate pathway catalyzed by the enzymes transaldolase and transketolase controlling the flux [35]. In addition, these enzymes regulate the switch between glycolysis and the pentose phosphate pathways [43]. The results presented here are consistent with previous observations, as the expression levels of transaldolase and transketolase were downregulated on xylose compared to those on glucose, which limited the rate of xylose metabolism [36]. Genetic methods to improve the rate of xylose utilization have focused primarily on enhancing the non-oxidative phases of the pentose phosphate pathway [37][38][39].
Kurylenko et al. reported that both transketolase and transaldolase were needed for xylose fermentation in Ogataea polymorpha, though both enzymes were not required for growth on xylose as a sole carbon source. Therefore, overexpression of transketolase and transaldolase elevated ethanol production from xylose in the O. polymorpha [40]. Overexpression of genes involved in non-oxidative pentose phosphate pathway, including TAL1 and TKL1, also improved cell growth and increased the rate of xylose consumption in xylose-utilizing yeast strains [38]. In addition, it was also reported that the expression level of transketolase was upregulated through Msn2/4p-mediated stress responses, rather than in response to xylose as a sole carbon source in S. cerevisiae [11].
Overexpression of the XR/XDH pathway was utilized to engineer S. cerevisiae for xylose fermentation with xylose reductase using NAD(P)H as a cofactor and xylitol dehydrogenase, NAD + . In addition, these cofactors act greatly as limiting factors to xylose metabolism in transformed S. cerevisiae. However, we demonstrated that the expression levels of both xylose reductase and xylitol dehydrogenase were upregulated on xylose carbon source in T. reesei (Table 4, Fig. 6) thereby making this fungus an essential consolidated bioethanol producer from xylose. NAD(P)H is regenerated in S. cerevisiae mainly through the oxidative branch of the pentose phosphate pathway and the isocitrate dehydrogenases. Moreover, we found that the expression levels of glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase in the oxidative branch of the pentose phosphate pathway, and that of isocitrate dehydrogenase, were downregulated on xylose in T. reesei.
Jeppsson et al. had initially reported that ethanol yield in S. cerevisiae was increased by lowering the flux via the NADPH-producing pentose phosphate pathway [41]. This flux was lowered by disrupting both 6phosphogluconate dehydrogenase and glucose-6-phosphate dehydrogenase. Decreasing the enzyme activity of phosphoglucose isomerase was also reported to lower the flux. Consequently, lowering the rate of the oxidative components of the pentose phosphate pathway led to reduced xylose uptake rate, attributed to the fact that limited NADPH-mediated xylose reduction induced the low rate of xylose fermentation [41].
Furthermore, Miskovic et al. also suggested that increasing the activity of phosphoglucose isomerase, which catalyzes the conversion of glucose-6-phosphate to fructose-6-phosphate, results in a reduced rate of xylose consumption [42]. This is because of the decreased reduction of glucose-6-phosphate to fructose-6-phosphate which consequently results in a reduced flux through the oxidative component of pentose phosphate pathway during glucose-xylose fermentation. This assertion is consistent with the experimental study by Jeppsson et al. [41], as lower phosphoglucose isomerase activity reduced the flux through the oxidative component of pentose phosphate pathway in batch fermentations with xylose as the sole carbon source [42]. Results from our study also concur with these two experimental studies regarding the effects of glucose-6-phosphate dehydrogenase, 6phosphogluconate dehydrogenase, and phosphoglucose isomerase on xylose uptake rate in T. reesei. The expression levels of 6-phosphogluconate dehydrogenase, glucose-6-phosphate dehydrogenase, and glucose-6phosphate isomerase were downregulated with xylose, resulting in a lower rate of xylose consumption [43] .
Many genes coding for enzymes of the tricarboxylic acid cycle were not repressed in glucose-rich medium. However, those genes were still expressed at a relatively high level in xylose-rich medium compared to glucose in our study (Figs. 5 and 6). Meanwhile, the genes encoding respiratory enzymes were also upregulated during xylose metabolism compared to glucose. The Hap complex consists of four subunits: Hap2p, Hap3p, Hap5p -each containing a highly conserved core domain required for DNA binding -and Hap4p (yeast) or HapX (filamentous fungi), which is a critical component of the transcriptional activator proteins [44][45][46]. In fungi, the Hap complex has been linked to fundamentally important biological processes, including nutrient acquisition, oxidative stress responses, and asexual reproduction [47]. In S. cerevisiae, the expression level of Hap4p was upregulated during xylose fermentation compared to that during glucose fermentation [11]. Nevertheless, the genes of Hap2 and Hap3 were expressed very differently in T. reesei between xylose and glucose fermentation, while HapX has not been found in T. reesei transcription (Fig. 6). Zeng et al. showed that the expression of tricarboxylic acid cycle enzymes and respiratory enzymes was not repressed by xylose in transformed S. cerevisiae [29]. In that study, the expression level of CIT2 and CIT3, which catalyze the condensation of acetyl coenzyme A and oxaloacetate, respectively, to form citrate, was upregulated during xylose fermentation, especially xylose as a sole carbon source. The transcripts of cytoplasmic malate dehydrogenase, which is involved in the gluconeogenesis and glyoxylate cycles, were also increased with xylose. These results are consistent with our findings and those of Matsushika et al. that growth of both yeast and filamentous fungi does not recognize xylose as a fermentable carbon source [11].
The transport of xylose into the cell is the initial rate-limiting step of xylose utilization [48,49]. The predominant sugar transporters in S. cerevisiae are members of the HXT family [50]. These HXT genes belong to the major facilitator superfamily (MFS) of transporters [51]. Previous reports have demonstrated that MFS transporters display a higher affinity for glucose over xylose, thus contributing to limit pentose utilization during ethanol fermentation. Notably, xylose uptake in S. cerevisiae is very slow, and can be inhibited by glucose in the lignocellulosic sugar media. The use of engineered pentose transporters is therefore a promising approach to improve overall pentose uptake [52]. Nevertheless, very few pentose transporters have been functionally characterized and further research is needed to identify these targets. Filamentous fungi, on the other hand, possess a strong capacity to utilize lignocellulosic sugars. In fact, many pentose-assimilating fungal species possess specific transporters for pentose uptake. Several recombinant S. cerevisiae strains that were generated via introduction of heterologous xylose transporters have been reported to have significantly improved xylose uptake activity.
Jing et al. identified two novel xylose specific transporters, An25 and Xyp29, from Neurospora crassa and Pichia stipites, respectively, and characterized them in S. cerevisiae at the molecular level [53]. Colabardini et al. discovered a high affinity xylose transporter XtrD in Aspergillus nidulans, which encodes an MFS transporter [49]. S. cerevisiae cells engineered to produce XtrD were capable of growth using xylose, glucose, galactose, and mannose as sole carbon sources, indicating that XtrD could transport multiple sugars. It was observed that expression of the XtrD transporter improved xylose uptake in a mutant S. cerevisiae strain.
Furthermore, transporter genes can be engineered to improve xylose uptake activity by introducing directed evolution. Using directed evolution, four novel xylose transporters that remained inhibited by glucose were created. Young et al. aimed to engineer hexose transporters with improved xylose uptake capacity [48]. Based on our transcriptome data, a number of sugar transporters were induced by xylose, indicating that the T. reesei genome may also possess high affinity xylose transporters. As of now, three A. niger (XltA, XltB, and XltC) and three T. reesei (Str1, Str2, and Str3) xylose transporters have been reported by Sloothaak et al. [54]. This result demonstrated that six transporters were able to transport xylose into yeast cells, but the substrate affinity and biochemical properties of these six transporters were significantly different with respect to their uptake of xylose. Amino acid sequence analysis showed the specific residues and motifs of transporters associated with xylose transporters. Specifically, XltA and Str1 were induced by xylose and were dependent on the XlnR/Xyr1 regulators, indicating that both of these transporters play a key role in xylose utilization.
Industrial microorganisms that can produce alcohol utilizing both hexose and pentose sugars simultaneously are essential for reducing the cost of lignocellulose conversion to bioethanol. T. reesei is considered as a candidate microorganism to be used industrially, although several limitations of its use exist currently. The present study provides the first transcriptomic comparison analysis of T. reesei exposed to either xylose or glucose. We revealed significant transcriptomic changes in genes involved in xylose metabolism and their significance. However, upstream regulation of these targets would demand further studies.