Simulating Extracellular Glucose Signals Enhances Xylose Metabolism in Recombinant Saccharomyces cerevisiae

Abstract

Efficient utilization of both glucose and xylose from lignocellulosic biomass would be economically beneficial for biofuel production. Recombinant Saccharomyces cerevisiae strains with essential genes and metabolic networks for xylose metabolism can ferment xylose; however, the efficiency of xylose fermentation is much lower than that of glucose, the preferred carbon source of yeast. Implications from our previous work suggest that activation of the glucose sensing system may benefit xylose metabolism. Here, we show that deleting cAMP phosphodiesterase genes PDE1 and PDE2 increased PKA activity of strains, and consequently, increased xylose utilization. Compared to the wild type strain, the specific xylose consumption rate (r_xylose) of the pde1Δ pde2Δ mutant strains increased by 50%; the specific ethanol-producing rate (r_ethanol) of the strain increased by 70%. We also show that HXT1 and HXT2 transcription levels slightly increased when xylose was present. We also show that HXT1 and HXT2 transcription levels slightly increased when xylose was present. Deletion of either RGT2 or SNF3 reduced expression of HXT1 in strains cultured in 1 g L⁻¹ xylose, which suggests that xylose can bind both Snf3 and Rgt2 and slightly alter their conformations. Deletion of SNF3 significantly weakened the expression of HXT2 in the yeast cultured in 40 g L⁻¹ xylose, while deletion of RGT2 did not weaken expression of HXT2, suggesting that S. cerevisiae mainly depends on Snf3 to sense a high concentration of xylose (40 g L⁻¹). Finally, we show that deletion of Rgt1, increased r_xylose by 24% from that of the control. Our findings indicate how S. cerevisiae may respond to xylose and this study provides novel targets for further engineering of xylose-fermenting strains.

1. Introduction

Lignocellulosic biomass is considered as an abundant, renewable, and environment-friendly material for biofuels and chemicals production. Converting the two most abundant sugars in hydrolysates of lignocellulosic materials, glucose and xylose, can increase both the environmental and economic benefits of lignocellulose as a source for biofuels [1,2]. Saccharomyces cerevisiae has been considered a highly competitive cell factory for conversion of lignocellulosic materials to biofuels and chemicals because it is a generally recognized as safe (GRAS) microorganism by the U.S. Food and Drug Administration, has strong glucose-metabolizing capacity, and has been well-studied. However, S. cerevisiae cannot utilize xylose because of its inability to process xylose in its metabolic pathways [1,3].

In recent decades, continuous efforts have been made to construct xylose-utilizing S. cerevisiae and improve the xylose metabolic capacity of these recombinant strains. Xylose reductase (XR) and xylitol dehydrogenase (XDH) of Scheffersomyces stipitis or xylose isomerases (XI) of bacteria and fungi have been introduced into S. cerevisiae to build pathways for xylose metabolism. In recombinant S. cerevisiae strains, xylose is transported by hexose transporters and metabolized in sequence through the XR-XDH or XI, pentose phosphate, and glycolysis pathways to produce pyruvate, which is then converted to ethanol and other products [3,4,5]. Therefore, the genes of xylulokinase and the non-oxidative part of the pentose phosphate pathway (PPP) were overexpressed to enhance the downstream flux of xylose metabolism [6,7,8].

Unfortunately, the attempts of metabolic engineering mentioned above are far from enough, and the adaptive evolution in the medium with xylose as the sole carbon source is necessary to produce a S. cerevisiae strain that can efficiently utilize xylose [8,9,10,11,12]. Many researchers have endeavored to reveal the differences in the “omics” between the evolved strains (with high xylose utilization capacity) and their parents (with low xylose utilization capacity) [11,13,14,15], as well as the differences in the “omics” between the strains cultured in xylose and in glucose [16,17]. The results suggested that an important reason that limited the xylose fermentation rate is the fact that S. cerevisiae lacks a signaling pathway to recognize xylose as a carbon source and regulate the cells to convert to a state that promotes xylose utilization. For another, our previous work suggests that extracellular glucose signals can promote xylose utilization. In a strain that could transport xylose but not glucose intracellularly, we observed that xylose metabolism was enhanced by the presence of extracellular glucose [18] (Figure A1). Extensive studies on glucose signaling pathways and their controls on glucose metabolism showed that efficient hexose transporters and glycolysis, which are the factors for efficient xylose metabolism, depends on activation of glucose signaling pathways. Unfortunately, how these signaling pathways may respond to xylose is not clear.

There are two signaling pathways that respond to extracellular glucose [19]. The first is the cAMP-PKA pathway (Figure 1A) where the transmembrane protein, Gpr1, undergoes an allosteric effect when extracellular glucose or sucrose binds to it. Then, the allosteric Gpr1 stimulates the transition of the small G protein Gpa2 from an inactive state (binding with GDP) to an active state (binding with GTP). The GTP-bound Gpa2 activates adenylate cyclase Cyr1 which catalyzes the conversion of ATP to cAMP and subsequently increases the intracellular levels of cAMP. cAMP binds to the regulatory subunit Bcy1 of PKA and exposes the active site of Tpk1/Tpk2/Tpk3, thus activating PKA. Meanwhile, the cAMP-PKA pathway is also regulated by an intracellular protein, Ras. Intracellular glucose and its metabolites stimulate the GDP-bound Ras (inactive) to convert to GTP-bound Ras (active). The active Ras can also increase Cyr1 activity and consequently, PKA activity. Additional proteins that affect the pathway are two cAMP phosphodiesterases, Pde1 and Pde2, that catalyze cAMP to AMP to prevent the overactivation of PKA [19]. The active PKA up-regulates glycolysis and down-regulates gluconeogenesis at both transcriptional and translational levels [20,21]. Increased activity of PKA benefits xylose utilization because xylose metabolism occurs through the PPP and glycolysis pathway.

In the Rgt2/Snf3-Rgt1 pathway, which also responds to extracellular glucose (Figure 1B), high and low levels of extracellular glucose are sensed by the membrane-spanning proteins Rgt2 and Snf3, respectively [22]. The signal is then relayed into the cell, leading to the phosphorylation-dependent degradation of Mth1 and Std1, which are co-repressors necessary for the DNA-binding of Rgt1. The elimination of Mth1 and Std1 exposes Rgt1 to phosphorylation, which releases it from gene promoters and derepresses the expression of target genes, including the ones encoding hexose transporters [19,23]. Lacking a xylose transporter, the transport of xylose in S. cerevisiae depends on these hexose transporters (HXT) [24]. Thus, the repression of Rgt1 is expected to negatively affect hexose transporter genes and thus xylose utilization. In a recent work, researchers were able to show that extracellular xylose can induce the fluorescent signal of the green fluorescent protein gene (yEGFP3) expressed under the control of the HXT1, 2, and 4 promoters. Their data suggest that the Rgt2/Snf3-Rgt1 pathway responds to extracellular xylose, although the mechanistic details are still not clear [25].

The aims of the present work were to determine how signals from extracellular glucose can enhance xylose utilization and the mechanism of how S. cerevisiae can sense extracellular xylose. Here, we demonstrated that the PKA level in cells cultured on xylose was much lower than that of cells cultured on glucose. Additionally, expression of the Gpa2^G132V allele or knock outs of both PDE1 and PDE2 increased the PKA activity, regulated expression of carbohydrate metabolism genes, and increased the xylose consumption rate of the S. cerevisiae strain. We also explored how S. cerevisiae may sense extracellular xylose by examining the transcriptional levels of HXT1 and HXT2 in rgt2Δ and snf3Δ strains cultured in media with different levels of glucose and xylose. The results demonstrate that Rgt2 and Snf3 can only weakly respond to high concentrations of xylose and slightly increase the expression of transporter genes. Then, we demonstrated that deleting Rgt1 improved the expression of transporter genes and xylose utilization. Our work will improve our understanding of the effect of xylose on the major glucose signaling pathways in S. cerevisiae and demonstrates promising strategies for delivering simulated glucose signals to induce S. cerevisiae cells to modify their gene-expression state to a state fit for xylose utilization.

2. Materials and Methods

2.1. Strains and Plasmids

All S. cerevisiae strains and plasmids are listed in

Table 1. Strains and plasmids used in this work.

S. cerevisiae Strains and Plasmids	Description	Sources
Plasmids
pUG6	E. coli plasmid with segment LoxP-KanMX44-LoxP	[28]
pJFE1	YCplac33; CEN, Ampr, URA3, TEF1p-PGK1t	[26]
pJX7	ΥΕplac195; 2μ, Ampr, URA3, TEF1p-Ru-xylA-PGK1t	[26]
YEp-CH	YEp; containing hygromycin B resistant gene and Cre gene under GAL2 regulative regulation promoter	[26]
S. cerevisiae Strains
CEN.PK 113-5D	MATa; ura3-53, belongs to CEN.PK strain family	[27]
BSPC039	CEN.PK113-5D derivative; XKS1(-194,-1)::loxP-PTEF1, gre3(-241, +338)::PTPI1-RKI1-TRKI1-PPGK1-TAL1-TTAL1-PFBA1-TKL1-TTKL1-PADH1-RPE1-TRPE1-loxP	[12]
BSL01	BSPC039 derivative; pJX7	This work
BSL06	BSL01 derivative; GPA2G132V	This work
BSL08	BSL01 derivative; RAS2G19V	This work
BSL10	BSL01 derivative; pde1Δ	This work
BSL16	BSL01 derivative; pde1Δ; pde2Δ	This work
BSL20	BSL01 derivative; rgt1Δ	This work
BSWW1	Derivative of CEN.PK 113-5D: snf3Δ	This work
BSWW2	Derivative of CEN.PK 113-5D: rgt2Δ	This work

. The primers are listed in

Table A1. The primers used in this study.

Primers	Sequence 5′→3′
For gene expression
BamHI-GPA2-1up	GCGCGGATCCATGGGTCTCTGCGCATCTTC
GPA2-1dn	ACTTACCACTTTCAACGGCACCCAG
GPA2-2up	CTGCTGGGTGCCGTTGAAAGTGGTAAG
SbfI-GPA2-2dn	CGCGCCTGCAGGTCATTGTAACACTCCAGAGTCTTTC
GRE3-ty1up	CAGATGACGAGAAGAAAGGTCACATC
GRE3-ty1dn	CCGAAGGAGGAGTAAGCAACTAC
KanMX-up0505	ATATCCAAGTAGTTGCTTACTCCTCCTTCGGTAGGTCTAGAGATCTGTTTAGCTTGCC
KanMX-dn0505	CATTTTGAAGCTATGGTGTGTGGGCCACTAGTGGATCTGATATCACC
GPA2-up0505	GTGATATCAGATCCACTAGTGGCCCACACACCATAGCTTCAAAATGTTTC
GPA2-dn0505	GCCCATCTAAGCAATACTTGGGAAGCATGATTACGCCAAGCTTTAAC
GRE3-ty2up	CGTAATCATGCTTCCCAAGTATTGCTTAGATGGGC
GRE3-ty2dn	TCAGGCAAAAGTGGGGAATTTACC
G-int-up	CCATTTGAAGAGAAATACCCTCCAG
RAS-ty1up	GAAAACGGGAAACAAGGTTCACATCAG
RAS-ty1dn	CGACTTGTTCAAAGGCATTTTTTTTTCTGTATATCTCCTTTCAATTCGAAAACGG
Rasup1	AAGAGAGTACAAGCTAGTCGTCGTTGGTGGTGTTGGTGTTGGTAAATCTGC
Rasup2	ATGCCTTTGAACAAGTCGAACATAAGAGAGTACAAGCTAGTCGTCG
RAS-up	GAATTGAAAGGAGATATACAGAAAAAAAAATGCCTTTGAACAAGTCGAAC
RAS-dn	AGCGTACGAAGCTTCAGCTTTAACTTATAATACAACAGCCACCCGATC
K-RAS-up	AGTTAAAGCTGAAGCTTCGTACGCTG
K-RAS-dn	GCGTTTCTACAACTATTTCCTTTTTAGCATAGGCCACTAGTGGATCTG
RAS-ty2up	CCACTAGTGGCCTATGCTAAAAAGGAAATAGTTGTAGAAACGCTAAGACG
RAS-ty2dn	CTCTGGAACGTCCTCATATTCACC
seq-rasR	GTATTCTGGGCCTCCATGTC
int-Y	CAGAAGGACCTTTTCATTCACC
RuXI-up	GGATCCATGGCAAAAGAATATTTTCCGTTTAC
RuXI-dn	CCTGCAGGTTATTTGCAGTGGAGGGCG
For gene knockout
PDE1-ty1up	GAATGTACCAGCTACGGGAGATG
PDE1-ty1dn	CGTACGAAGCTTCAGCTCAGCCCACTAATATGGTCCAGATG
K-PDE1-up	CATCTGGACCATATTAGTGGGCTGAGCTGAAGCTTCGTACGCTG
K-PDE1-dn	CTAATATACCCTTGAGATGCACAAGCGCATAGGCCACTAGTGGATCTG
PDE1-ty2up	GATCCACTAGTGGCCTATGCGCTTGTGCATCTCAAGGGTATATTAG
PDE1-ty2dn	CTCAAGTCTCCTAAGTTCCTCTCC
PDE2-ty1up	ATGTCCACCCTTTTTCTGATTGG
PDE2-ty1dn	AGCGTACGAAGCTTCAGCTCTCTCGTGTATGATCTATCGCAG
K-PDE2-up	TCTGCGATAGATCATACACGAGAGAGCTGAAGCTTCGTACGCTG
K-PDE2-dn	CTATCCTCATACTGAGAATGCAATGCGCATAGGCCACTAGTGGATCTG
PDE2-ty2up	GATCCACTAGTGGCCTATGCGCATTGCATTCTCAGTATGAGGATAG
PDE2-ty2dn	CGTTTGGAATATGCGGATGGTC
KanMXhomoup-F	TGTGACGTGGCTTATGAGCACCAG
KanMXhomoup-R	AGTACAAGGGAGAGGCGATTTCCAACG
KanMXhomodn-F	ATGCTCCCTTCTCCTGTAGGTCAGG
KanMXhomodn-R	TGGCAATGGCAATAGTGATGGCACC
KanMX-F	GGAAATCGCCTCTCCCTTGTACTGACATGGAGGCCCAGAATAC
KanMX-R	CTGACCTACAGGAGAAGGGAGCATCAGTATAGCGACCAGC
SNF3-1	AGTGGCGTATTGGCCTTGTT
SNF3-2	GCCGAGGATAGGACTATTGTCTAG
SNF3-3	CAATAGTCCTATCCTCGGCACGCGGCCGCCAGCTGAAGCTTCG
SNF3-4	GTAACCCAGAATCATGGACTGGACCGGCAGATCCGCGGCCGCA
SNF3-5	CCAGTCCATGATTCTGGGTTACTG
SNF3-6	CAACAACCAGCACCTTACGTCTAC
RGT2-1	CTGTTCCTGCATCGTCCATCGGC
RGT2-2	CGGCAGGTAGTCACCGTTGAGTC
RGT2-3	TCAACGGTGACTACCTGCCGACGCGGCCGCCAGCTGAAGCTTCG
RGT2-4	CTTGGATCTTCGATCGGGAGGCCACCGGCAGATCCGCGGCCGCATAG
RGT2-5	GGCCTCCCGATCGAAGATCCAAG
RGT2-6	ACTGAAGGCGACGTTGACGGC
For qPCR
Actin-F	CAAACCGCTGCTCAATCTTC
Actin-R	AGTTTGGTCAATACCGGCAG
HXT1-qPCR-F	GTCGGTATGGTCTGCTGTTATG
HXT1-qPCR-R	ACAGTTACCAGCACCCTTTG
HXT2-qPCR-F	GGCTCTCAACAAACTTCTATCCAC
HXT2-qPCR-R	GGGAGTTCAGCGTTAGTGTATTC

.

Strain BSL01 (BSPC039 [12] derivative; pJX7 [26]) was used as the parent strain and control. The strain BSL06(GPA2^G132V) was obtained by integrating GPA2^G132V at the GRE3 site of BSL01. The fragment GRE3-ty1-loxP-KanMX-loxP-TEF1p-GPA2^G132V-PGKt-GRE3-ty2 was used for the integration. The homologous arms GRE3-ty1 and GRE3-ty2 were cloned from the CEN.PK113-5D [27] genome. The selection marker loxP-KanMX-loxP was cloned from plasmid pUG6 [28]. The GPA2^G132V ORF fragment was obtained by fusion PCR, using the CEN.PK113-5D genome as a template, and then treated with BamHI and SbfI and ligated between TEF1p and PGKt of plasmid pJFE1 [26]. Then, the expression cassettes TEF1p-GPA2^G132V-PGKt, GRE3-ty1, GRE3-ty2, and loxP-KanMX-loxP were ligated by fusion PCR to obtain the full fragment GRE3-ty1-loxP-KanMX-loxP-TEF1p-GPA2^G132V-PGKt-GRE3-ty2. The KanMX marker was discarded by transferring plasmid YEp-CH [26] into the strains and inducing the expression of the Cre recombinase. Thus, the KanMX4 cassette between the two loxP sequences was removed by the recombination between the two loxP sites. The construction process of BSL08(RAS2^G19V) was similar to that of BSL06(Gpa2^G132V).

The deletions of genes PDE1, PDE2, RGT1, SNF3, and RGT2 were all performed by replacing the target gene with the KanMX4 cassette via homologous recombination, and then the KanMX4 was removed as mentioned above. The integration arms of genes were all cloned from the CEN.PK113-5D genome.

2.2. Cultivation Conditions and Batch Fermentation

BSL01 and its derivatives were cultured in SD-URA medium, which contained 1.7 g L⁻¹ yeast nitrogen base (YNB, Sangon, Shanghai, China), 5 g L⁻¹ ammonium sulfate (Sangon, Shanghai, China), 0.77 g L⁻¹ CSM-URA (MP Biomedicals, Solon, OH, USA), and 20 g L⁻¹ glucose or xylose as the carbon source. BSWW1 (snf3Δ) and BSWW2 (rgt2Δ) were cultured in YPD medium. For selection of transformants, 800 mg L⁻¹ G418 or 400 mg L⁻¹ hygromycin B was added into the medium when necessary. The fermentations were performed in SD-URA medium with 20 g L⁻¹ xylose and/or 20 g L⁻¹ glucose at 30 °C. A single colony from a plate of SD-URA medium was incubated in the liquid medium and collected in the middle of exponential growth to inoculate 100-mL flasks containing 40 mL fermentation medium. The initial biomass was 0.23 g DCW L⁻¹ and pH was 4.5. The fermentation was performed at 200 rpm, 30 °C.

2.3. Determination of cAMP

Preparation of samples was performed as instructed in the manual of the cAMP Direct Immunoassay Kit (BioVision, Milpitas, CA, USA). Cells in the middle of exponential growth were collected by centrifugation at 6000 rpm for 5 min at 4 °C. The 80 mg (DCW) cells were cooled in liquid nitrogen and resuspended with 500 μL HCl (0.2 M pre-cooled to 4 °C) and 0.6 g acid-washed glass beads (φ = 0.5 mm). Then cells were lysed using a FastPrep cell homogenizer (Thermo Savant, Carlsbad, CA, USA) at a speed of 5000 rpm for 30 s and repeated twice. The homogenate was centrifuged at 13,000 rpm for 5 min at 4 °C, and the cAMP levels in this supernatant were determined by the cAMP Direct Immunoassay Kit (BioVision, Milpitas, CA, USA). The t-tests were applied to evaluate the differences between means.

2.4. Assay of Trehalase Activity

The preparation of crude enzyme samples and the assay of trehalase activity were performed based on a previously reported method [29]. Cells were collected by centrifugation at 6000 rpm for 5 min at 4 °C. The 40 mg (DCW) cells were cooled in liquid nitrogen and resuspended with 500 μL HEPES (pH 7.0) and 0.8 g acid-washed glass beads (φ = 0.5 mm). These cells were lysed using a FastPrep cell homogenizer (Thermo Savant, Carlsbad, CA, USA) at a speed of 5000 rpm for 30 s and repeated five times. The homogenate was centrifuged at 13,000 rpm for 5 min at 4 °C, and the supernatant was used as the crude enzyme sample. The total cellular protein concentration was measured using a BCA protein assay reagent kit (Sangon Biotech Co., Ltd., Shanghai, China). Trehalase activities were determined at 30 °C by measuring the glucose produced by hydrolysis of trehalose. Reactions contained 10 μL extract and 140 μL water in 100 μL substrate buffer (250 mM trehalose in 50 mM HEPES buffer, pH 7, with 125 μM CaC1₂). They were incubated at 30 °C for 30 min and then immediately halted by boiling for 3 min at 100 °C. The amount of glucose produced was determined by a d-Glucose (GOPOD Format) Assay Kit (Megazyme, Dublin, Ireland). Specific activity of trehalase was expressed as nmol glucose liberated min⁻¹ (mg protein)⁻¹. The t-tests were applied to evaluate the differences between means.

2.5. Analysis of Metabolites

The concentrations of glucose, xylose, and ethanol were determined using the Prominence LC-20A HPLC (Shimadzu, Kyoto, Japan) equipped with an Aminex HPX-87H ion-exchange column (Bio-Rad, Hercules, CA, USA) and a refractive index detector RID-10A. Samples were eluted from the column at 45 °C with 5 mmol L⁻¹ H₂SO₄ at a flow rate of 0.6 mL min⁻¹ [16].

2.6. Calculation of Physiological Parameters

The cell density (OD₆₀₀) was determined with a UV–visible spectrophotometer (Eppendorf, Hamburg, Germany). The biomass was estimated according to the correlation of measured OD₆₀₀ and dry weight [16]. One unit of OD₆₀₀ equaled 0.208 g DCW L⁻¹. The specific growth rate (μ) was the regression coefficient of the log-linear regression of the OD₆₀₀ versus time during the exponential growth phase. The specific xylose consumption rate (r_xylose), specific glucose consumption rate (r_glucose), and specific ethanol production rate (r_ethanol) were calculated using the following Equation (1) as previously described [12]:

$$\mathrm{r}=\frac{A_n-A_m}{\frac{1}{2}{{\displaystyle \sum }}_{i=m+1}^n\left(B_i+B_{i-1}\right)\times \left(t_i-t_{i-1}\right)},$$

(1)

where r is the specific utilization or production rate during the phase from sampling point m to sampling point n; and A, B, and t are the metabolite concentration, biomass concentration, and time, respectively, at sampling points n, i, and m.

The t-tests were applied to evaluate the differences between means.

2.7. Quantitative PCR

qPCR data were analyzed according to the 2^−ΔΔCt method [30]. RNA was extracted from exponentially growing cells using a UNlQ-10 Column Trizol Total RNA Isolation Kit (Sangon Biotech Co., Ltd., Shanghai, China). The cDNA was obtained using a PrimeScript^™ RT reagent Kit (TaKaRa, Shiga, Japan). The gene transcription levels were determined using the equation N = 2^{Ct (reference gene)}/2^{Ct (target gene)}. ACT1 was used as the reference gene, and t-tests were applied to evaluate the differences between means [16].

3. Results

3.1. Expression of the GPA2^G132V Allele and Deletion of Both PDE1 and PDE2 Increased PKA Activity in S. cerevisiae

To investigate differences in cAMP-PKA responses in S. cerevisiae to glucose and xylose, intracellular cAMP concentration and trehalase activity level of strains cultured in media using glucose or xylose as the sole carbon source were determined, since trehalase activity is tightly connected to PKA activity [31]. The results showed that intracellular cAMP concentrations of the strain BSL01 in glucose and xylose fermentation were similar (Figure 2). However, intracellular trehalase activity of yeast cells in xylose fermentation was only 35% of that in glucose fermentation (Figure 2), implying that the PKA activity of cells in xylose fermentation was lower than that in glucose fermentation.

Three genetic strategies were used to increase PKA activity. The first strategy promoted expression of the GPA2^G132V allele, which can constitutively activate the cAMP-PKA pathway whether or not Gpr1 senses the extracellular glucose signals [21,32]. The second promoted expression of the RAS2^G19V allele, which may induce high levels of cAMP and PKA activity [21,33]. In the third method, we deleted the genes encoding two phosphodiesterase (both Pde1 and Pde2 or only the primary one, Pde1), which functions in the feedback control by PKA in degrading cAMP [19]. Thus, we constructed strains BSL06(GPA2^G132V), BSL08(RAS2^G19V), BSL10(pde1Δ), and BSL16(pde1Δ pde2Δ), which were derived from the xylose-utilizing strain BSL01. The BSL01 strain consisted of a xylose isomerase gene cloned from a metagenomics library of the bovine rumen microbiome. Moreover, the genes encoding xylulokinase and the enzymes in the non-oxidative part of the PPP were overexpressed in BSL01.

The results show that expressing GPA2^G132V and deleting both PDE1 and PDE2 increased the intracellular cAMP concentration by respectively 15% and 30% in glucose fermentation (Figure 2A). Consistent with this, trehalase activity increased by respectively 76% and 82% (Figure 2B), which implied the increase of PKA activity. In xylose fermentation, cAMP levels of the five strains were similar. However, expressing GPA2^G132V and deleting both PDE1 and PDE2 increased trehalase activity by respectively 73% and 114%. Thus, the trehalase activity of these two strains in xylose fermentation respectively reached 61% and 70% of that of BSL01 in glucose fermentation (Figure 2B). In contrast, expressing RAS2^G19V and only deleting PDE1 had no effect on cAMP level and trehalase activity in both glucose and xylose fermentations (Figure 2).

3.2. Expression of GPA2^G132V Allele and Deletion of Both PDE1 and PDE2 Affected Glucose and Xylose Metabolism

The fermentation dynamics of strains BSL06(GPA2^G132V), BSL16(pde1Δpde2Δ), and the control BSL01 were compared. The results revealed that the specific growth rates (μ) of these strains with 20 g L⁻¹ xylose as the sole carbon source were the same (Figure 3). The specific xylose consumption rates (r_xylose, xylose consumed per unit of cell mass) of BSL06(GPA2^G132V) and BSL16(pde1Δpde2Δ) were 19% and 51% higher than that of BSL01, respectively. The specific ethanol production rates (r_ethanol) of BSL06(GPA2^G132V) and BSL16(pde1Δpde2Δ) were greater by 74% and 72%, respectively, compared to that of BSL01 (Figure 3,

Table 2. The xylose fermentation and glucose–xylose co-fermentation characteristics of strains.

Strains	Xylose Fermentation		Glucose–Xylose Co-Fermentation
	Specific Xylose Consumption Rates (g L−1 h−1 g−1 DCW)	Specific Ethanol Production Rates (g L−1 h−1 g−1 DCW)	Specific Glucose Consumption Rates (g L−1 h−1 g−1 DCW)	Specific Xylose Consumption Rates (g L−1 h−1 g−1 DCW)	Specific Ethanol Production Rates (g L−1 h−1 g−1 DCW)
BSL01(Control)	0.103 ± 0.003	0.047 ± 0.002	1.218 ± 0.004	0.046 ± 0.002	0.428 ± 0.005
BSL06(GPA2G132V)	0.123 ± 0.003	0.082 ± 0.001 ***	1.354 ± 0.019 **	0.0481 ± 0.002	0.489 ± 0.002 ***
BSL16(pde1Δ pde2Δ)	0.156 ± 0.001 **	0.081 ± 0.003 **	1.260 ± 0.008 *	0.068 ± 0.004 *	0.523 ± 0.005 ***

The fermentations were performed in the 100 mL shake flasks with 40 mL medium. For the xylose fermentation, the medium was SC-URA with 20 g L⁻¹ xylose as carbon source; for the glucose–xylose co-fermentation, the medium was SC-URA with 20 g L⁻¹ glucose and 20 g L⁻¹ xylose medium. The cells were cultured at 30 °C, 200 rpm, with initial OD₆₀₀ of 2.5 (≈0.5 g L⁻¹ biomass). The experiments were performed in triplicate. The values are given as the averages ± standard deviations of three measurements. DCW, dry cell weight. * p value < 0.05; ** p value < 0.01; *** p value < 0.005.

).

In the co-fermentation of 20 g L⁻¹ glucose and 20 g L⁻¹ xylose as carbon sources, the μ of strains were also similar (Figure 4). However, BSL06(GPA2^G132V) and BSL16(pde1Δpde2Δ) shifted out of the exponential growth phase before BSL01′s shift. Correspondingly, biomass yields of BSL06(GPA2^G132V) and BSL16(pde1Δpde2Δ) were lower than biomass of BSL01. Interestingly, the lower biomass yields did not result in lower consumption rates of sugars; in fact, sugar metabolism rates of the mutant strains were greater than the rate of the wild type. Compared to BSL01, the specific glucose consumption rate (r_glucose) of BSL06(GPA2^G132V) increased by 11.1%; r_xylose of BSL16(pde1Δpde2Δ) increased by 48%; and r_ethanol of BSL06(GPA2^G132V) and BSL16(pde1Δpde2Δ) increased by 14% and 22%, respectively (Figure 4, Table 2).

3.3. Weak Responses of Rgt2 and Snf3 to High Concentrations of Extracellular Xylose

To investigate if the Rgt2/Snf3-Rgt1 pathway, another well-known glucose signaling pathway, can respond to extracellular xylose and induce expression of transporter genes, the transcriptional levels of genes HXT1 and HXT2 in yeast cells cultured on different carbon sources were measured by quantitative PCR.

A previous study reported that a high concentration of glucose (40 g L⁻¹) induced expression of HXT1, and the main sensor of the high-level glucose in S. cerevisiae was Rgt2 [22]. In addition, a low concentration of glucose (1 g L⁻¹) induced expression of HXT2, and the main sensor of the low-level glucose was Snf3 [22]. Our results supported these findings (Figure 5). HXT1 and HXT2 transcriptional levels in the wild type strain CEN.PK 113-5D grown in the medium containing 40 g L⁻¹ glucose were approximately 400 and 25 times greater than those observed in cells cultured in a medium containing glycerol as the sole carbon source (hereafter referred to as glycerol medium), respectively. Furthermore, in comparison to HXT1 expression in the wild type exposed to the high concentration of glucose, expression in rgt2Δ mutants was greater by 58% and unaffected by the deletion of SNF3. In a similar comparison, HXT2 expression induced by 1 g L⁻¹ glucose was lower by 82% in snf3Δ mutants and unaffected by the deletion of RGT2.

The results illustrate additional relationships of the four genes and the different carbon sources (Figure 5). The presence of 1 g L⁻¹ glucose, 40 g L⁻¹ xylose, or 1 g L⁻¹ xylose also increased the transcriptional levels of HXT1 in CEN.PK 113-5D, although the levels were much lower than the level in which 40 g L⁻¹ glucose was present. Thus, all four carbohydrate conditions can induce the expression of HXT1. However, deletion of RGT2 or SNF3 reduced expression of HXT1 in strains cultured in 1 g L⁻¹ xylose, the weak response of the strains to the low xylose concentration (1 g L⁻¹) may be a cumulative effect that involves these two sensors. Exposure to 40 g L⁻¹ glucose and 40 g L⁻¹ xylose also raised transcriptional levels of HXT2 in CEN.PK 113-5D, although the levels were much lower than the exposure to 1 g L⁻¹ glucose. By contrast, the signal from 1 g L⁻¹ xylose was likely too weak to induce expression of HXT2. Furthermore, deletion of SNF3 significantly weakened the expression of HXT2 in the yeast cultured in 1 g L⁻¹ glucose or 40 g L⁻¹ xylose, while deletion of RGT2 did not weaken expression of HXT2. These results suggest that S. cerevisiae mainly depends on Snf3 to sense a high concentration of xylose (40 g L⁻¹), similar to its Snf3-dependent response to the low level of glucose, though the response was clearly much weaker than the response to glucose.

3.4. Deleting RGT1 Enhanced Xylose Utilization in S. cerevisiae by Upregulating the Expression of Hexose Transporter Genes

Rgt1, a transcription repressor of HXTs, is phosphorylated when the glucose signal is transmitted into the cell through Rgt2 and Snf3, and then released from the promoters of HXTs; therefore, it derepresses the expression of hexose transporters. Thus, to simulate the condition that glucose signals have occurred, rgt1 was deleted in strain BSL01. The qPCR results indicate that deletion of RGT1 increased transcriptional levels of HXT1 and HXT2 by 17.32 ± 2.23- and 3.75 ± 0.38-fold, respectively, in cells cultured with 20 g L⁻¹ xylose, but RGT1 had no notable effect on cells cultured with 20 g L⁻¹ glucose. Furthermore, the results from fermentation experiments (Figure 6) showed that the specific xylose consumption rate (r_xylose) of BSL20(rgt1Δ) in xylose fermentation was 0.133 ± 0.002 g L⁻¹ h⁻¹ g⁻¹ DCW, which was 24% higher than the r_xylose of BSL01 (0.107 ± 0.006 g L⁻¹ h⁻¹ g⁻¹ DCW). Consistent with qPCR results, the RGT1 deletion did not affect glucose utilization. However, the r_xylose of BSL20(rgt1Δ) in the glucose–xylose co-fermentation during the xylose consumption phase (20–72 h) was 0.053 ± 0.003 g L⁻¹ h⁻¹ g⁻¹ DCW and 23% higher than that of BSL01 (0.043 ± 0.003 g L⁻¹ h⁻¹ g⁻¹ DCW).

4. Discussion

Xylose is an important and abundant carbon source second to glucose in industrial fermentation processes to produce fuels and chemicals. However, S. cerevisiae is only able to efficiently metabolize glucose. Despite the many attempts to genetically modify the carbohydrate metabolism of yeast to obtain more efficient xylose fermentation rates, large gaps in our knowledge and understanding of the carbohydrate metabolic pathways and their mechanisms hamper our pursuit in developing recombinant strains that can exploit the potential energy from xylose as well as they do from glucose.

Hexose transporters and glycolysis, which are controlled by glucose signaling pathways, are important to xylose metabolic efficiency, and our previous work suggests that extracellular glucose signals can promote xylose utilization [18] (Figure A1). Despite extensive studies on glucose signaling pathways and their controls on glucose metabolism, less is known about how these signaling pathways may respond to xylose. Here, we demonstrated that S. cerevisiae cells can only weakly respond to xylose, which may limit its potential to metabolize xylose. Furthermore, we also demonstrated that simulating the occurrence of glucose signals in yeast strains with a mutation (GPA2^G132V) or deletions (PDE1 and PDE2, and RGT1) of specific genes found in two different glucose signaling pathways (cAMP-PKA and Rgt2/Snf3-Rgt1 pathways) are feasible strategies to enhance xylose utilization and/or ethanol production in yeast.

Brink et al. [25] developed a successful method by applying a green fluorescence protein gene (yEGFP3) to monitor effects of eight endogenous yeast promoters, which are found in the three major sugar metabolism pathways Snf3/Rgt2, Snf1/Mig1, and cAMP/PKA, on glucose and xylose signaling responses in yeast. They determined that only two of these promoters respond to xylose; the two successful promoters were the HXT2p/4p sensors of the Snf3/Rgt2 signaling pathway. We also measured the responses of the two promoters, HXT1 and HXT2, but in rgt2Δ and snf3Δ mutants, as well as a strain with the wildtype Snf3/Rgt2 pathway. Our results revealed that the expression of HXT1 responded to both high and low concentrations of xylose (50 and 1 g L⁻¹, respectively), while the expression of HXT2 only responded to the high xylose concentration. Our results confirmed that the Snf3/Rgt2 pathway can minimally respond to xylose. Furthermore, we found that S. cerevisiae’s response to the high xylose level is similar to its response to the low glucose level, which corroborates the findings of Osiro et al. [34].

Özcan et al. (1996) revealed that a dominant mutation in RGT2 and SNF3 causes constitutive induction of HXT gene expression, even in the absence of the inducer glucose, which suggests this mutation converts the glucose sensors into their glucose-bound form [22]. Furthermore, their study on the rgt2Δsnf3Δ strain, as well as the rgt2Δ and snf3Δ strains, revealed that HXT1 and HXT2 took even lower response to the glucose in the double mutant than in the single mutants, which confirmed that Snf3 and Rgt2 have separate but overlapping functions [35]. In the present work, deletion of either RGT2 or SNF3 reduced expression of HXT1 in strains cultured in 1 g L⁻¹ xylose, which suggest that the weak response of the strains to the low xylose concentration (1 g L⁻¹) may be a cumulative effect that involves these two sensors. Moreover, xylose can bind both Snf3 and Rgt2 and slightly alter their conformations. In contrast, deletion of SNF3 significantly weakened the expression of HXT2 in the yeast cultured in 40 g L⁻¹ xylose, while deletion of RGT2 did not weaken expression of HXT2, suggesting that S. cerevisiae mainly depends on Snf3 to sense a high concentration of xylose (40 g L⁻¹). Our results also showed that the response of cells to xylose was clearly much weaker than the response to glucose. This may be due to either a low affinity of xylose to Snf3/Rgt2 is low or the conformations of xylose-bound Snf3/Rgt2 are not close to the glucose-bound Snf3/Rgt2. To determine the cause, will require substantially greater research efforts in the future.

Here, we used a recombinant strain with xylose metabolic capacity. This strain was derived from CEN.PK-113-5D, which belongs to the family of the CEN.PK strain [27]. Reportedly, the CYR1 allele in CEN.PK strains encodes a mutant adenylate cyclase (Cyr1^K1876M), and the constitutive activation of the cAMP pathway by Ras2^G19V and Gpa2^G132V is suppressed by this mutant [27]. Supporting this report, our result showed that the cAMP level and PKA activity of strain BSL08(RAS2^G19V) were not higher than those of the control [36]. However, although the cAMP level of the strain BSL06(GPA2^G132V) showed an insignificant increase in xylose, the increased PKA activity in BSL06(GPA2^G132V) cells cultured on glucose and xylose were observed. This may be due to Gpa2 inhibiting Gpb1 and 2, which inhibit PKA activity [37]. Moreover, deletion of both PDE1 and PDE2 increased PKA activity and xylose utilization. Recent works have reported that deletion of IRA2 inhibits PKA activity by negatively regulating RAS or BCY1 (a gene encoding a PKA inhibitor) and thus increases the specific xylose consumption rates (xylose consumed per unit of cell mass) of strains [14,15,38]. Our results directly support their viewpoint that active PKA benefits xylose metabolism in a certain way.

Moreover, we found that increasing PKA activity did not affect the specific growth rate of the yeast strain; however, it shortened the exponential growth phase and decreased biomass yield in glucose fermentation. High PKA activity promotes aging and early cell death and leads to a short lifespan [39,40], and thus the high PKA activity may have been the cause of this observation. On the other hand, in xylose fermentation, the increased PKA activity did not affect the exponential growth phase and biomass yield of yeast. This may be due to that the PKA activity in strains (control and mutations) cultured in xylose are lower than that in the control strain cultured in glucose. However, the short lifespan should be avoided in ethanol production using materials containing both glucose and xylose. Therefore, increase PKA activity only when glucose is running out in glucose–xylose co-fermentation may be a way to enhance xylose utilization without shorten the lifespan of yeast cells.