Construction of an Efficient Mutant Strain of Trichosporonoides oedocephalis with HOG1 Gene Deletion for Production of Erythritol

Abstract

The mitogen-activated protein kinase HOG1 (high-osmolarity glycerol response pathway) plays a crucial role in the response of yeast to hyperosmotic shock. Trichosporonoides oedocephalis produces large amounts of polyols (e.g., erythritol and glycerol) in a culture medium. However, the effects of HOG1 gene knockout and environmental stress on the production of these polyols have not yet been studied. In this study, a To-HOG1 null mutation was constructed in T. oedocephalis using the loxP-Kan-loxP/Cre system as replacement of the targeted genes, and the resultant mutants showed much smaller colonies than the wild-type controls. Interestingly, compared with the wild-type strains, the results of shake-flask culture showed that To-HOG1 null mutation increased erythritol production by 1.44-fold while decreasing glycerol production by 71.23%. In addition, this study investigated the effects of citric acid stress on the T. oedocephalis HOG1 null mutants and the wild-type strain. When the supplementation of citric acid in the fermentation medium was controlled at 0.3% (w/v), the concentration of erythritol produced from the wild-type and To-HOG1 knockout mutant strains improved by 18.21% and 21.65%, respectively.

Introduction

Erythritol is a four-carbon sugar alcohol that has been approved for use as a natural sweetener and a food additive in the United States [5]. Thus far, several studies have reported that erythritol possesses a much lower energy value (0.2 kcal/g) than sucrose (4 kcal/g) and has 60-70% of the sweetness of sucrose in a 10% (w/v) solution [23,25]. This makes erythritol a preferred non-nutritive or low-calorie sweetener for the dietary management of diabetes mellitus. In addition, erythritol is used as a flavor enhancer, humectant, stabilizer, formulation aid, and thickener [14]. In particular, its toxic activity that helps inhibit the fruit fly Drosophila melanogaster has been reported [1]. Overall, there is considerable interest in exploring genetically engineered erythritol because of its characteristic natural sweetness and low calorific value.

Erythritol is usually manufactured by microbial methods, using osmophilic yeasts or certain bacteria, by a simple process involving cheap starting materials [26,29,34]. It is efficiently produced by several high-yield strains such as Trichosporonoides megachiliensis SNG-42 or Candida magnoliae JH110 [13,16]. Moreover, T. oedocephalis, which can tolerate up to 60% (w/v) of glucose, is an ideal organism that can be used for the industrial production of erythritol [27]. Unfortunately, the large amount of glycerol generated as a byproduct during the production of erythritol by T. oedocephalis hinders further separation and purification [6]. Therefore, limiting the quantity of the glycerol byproduct is particularly important for increasing the yield of erythritol. In our previous study, we observed that the production of glycerol could be dramatically reduced by optimizing the culture conditions, such as controlling the dissolved oxygen and stirring in the process of ribitol production by T. oedocephalis ATCC 16958 [19]. By adopting this strategy, glycerol becomes the main product in the early stage of the fermentation process while erythritol and ribitol are mainly synthesized in the stationary phase. However, the molecular mechanism underlying glycerol regulation and erythritol production during the culture of T. oedocephalis is still unclear.

It is well known that HOG1 (high-osmolarity glycerol response pathway), a member of the mitogen-activated protein kinases (MAPKs) family, plays an important role in the accumulation of the compatible solute glycerol. The three-tiered cascade of protein kinase HOG1, known as MAPK, is the molecular device for eliciting these responses [4]. In cells, partially redundant MAPK kinases including Pbs2p can stimulate the phosphorylation and nuclear translocation of HOG1p, leading to the expression of a number of genes and production of glycerol so as to prevent dehydration [3,21]. In particular, in the response of yeast to hyperosmotic stress, the activation of the HOG pathway by the increase in extracellular osmolarity elicits a wide range of responses, and transcription of the glycerol-3-phosphate dehydrogenase gene (GPDH) dramatically increases, resulting in the accumulation of the compatible solute glycerol [31]. However, whether HOG1 also regulates glycerol and erythritol production in T. oedocephalis is still unclear.

The knockout procedure is a key strategy for the molecular dissection of gene function in many microbes. As an efficient gene disruption technology, the Cre/loxP cassette system is widely applied to yeast in combination with the heterologous dominant KanR. The short yeast DNA sequences on either side of a marker gene are integrated into the yeast genome by homologous recombination [2]. KanR, with resistance to the antibiotic G418, eliminates the need for yeast strains that are auxotrophic for the markers normally used for gene disruption [30]. In addition, the KanR without homology to the yeast genome is driven by the Ashbya gossypii TEF2 promoter and terminated by the A. gossypii TEF2 terminator, which leads to minimal recombination between the yeast genome and the internal parts of the KanMX disruption cassette [11]. In view of the widespread use of the Cre/loxP system in yeast, it is surprising that previous studies have rarely reported systematic approaches to the gene deletion of T. oedocephalis.

In this study, we initially attempted to adopt the Cre/loxP system containing the KanMX gene disruption cassette for HOG1 gene knockout in T. oedocephalis. Then, the T. oedocephalis HOG1 knockout mutants were constructed and evaluated for their ability to produce glycerol and erythritol in shake flasks. Meanwhile, the effects of citric acid on the production of glycerol and erythritol by mutant and wild-type T. oedocephalis were investigated in detail. This research could help enhance the application of genetically engineered bacteria for deletion HOG1 site in erythritol biosynthesis.

 

Materials and Methods

Strains and Plasmids

The wild-type strain T. oedocephalis ATCC16958 used for this study was purchased from American Type Culture Collection (ATCC, USA). The plasmids used in this study, pug6 and psh65, were obtained from Addgene, USA. E. coli JM109 was selected as the host for plasmid construction and was propagated at 37℃ in LB medium (10 g/l tryptone, 5 g/l yeast extract, and 10 g/l NaCl). T. oedocephalis was cultivated in yeast extract peptone dextrose (YPD) medium (600 g/l potato, 20 g/l glucose, and 5 g/l yeast extract).

Analysis of Conserved Sequence and Amplification of HOG1

From the gene bank, four DNA sequences of HOG1 that belonged to Saccharomyces cerevisiae S288c, Candida albicans SC5314, Candida dubliniensis CD36, and T. megachiliensis were analyzed to design the primers for amplification of the full-length sequence of HOG1 by DNAMAN Primer 5.0. Primers F-cqHOG1 and R-cqHOG1 were synthesized according to the conserved sequence of HOG1 in primer 5.0, adhering to the general rules of PCR primer design. The primer sequences are given in Table 1.

Table 1.Oligonucleotide primers used in the study.

Genomic DNA and RNA were extracted from T. oedocephalis, which were harvested using the Dr. GenTLE (from Yeast) High Recovery Kit and Yeast RNAiso Kit (TaKaRa, Tokyo, Japan). For amplification of the HOG1 of T. oedocephalis, 20 ng of total DNA was used as the template in a 50 μl reaction mixture containing 5 μl of PCR buffer (10 mM Tris/HCl; pH: 9.0, 50 mM KCl, and 1.5 mM MgCl2), 2 μl each of forward (cqHOG1-F) and reverse (cqHOG1-R) primers, and 2.5 units of Taq DNA polymerase. The PCR cycling profiles were as follows: one cycle at 95℃ for 3 min; 39 cycles at 95℃ for 30 sec, 58℃ for 30 sec, and 72℃ for 2 min; followed by a final extension step at 72℃ for 10 min.

Construction of Gene Knockout Subassembly of HOG1

The gene disruption segment was constructed by PCR, and the pug6 sequence was used as the template in combination with forward (knockout HOG1-F) and reverse (knockout HOG1-R) primers, which are listed in Table 1. The PCR conditions were similar to those described previously, except for the annealing temperature of 68℃. Oligonucleotides (knockout subassembly) carry the same 20 bp 3’-nucleotides, while the 60 bp 5’-nucleotides must be homologous to the sequences to the left or right of the gene to be deleted.

Transformation of T. oedocephalis in Combination with Zymolyase-100T

T. oedocephalis, a strain of osmotolerant yeast, was grown in 50 ml of YPD medium in 250 ml flasks at 30℃ with a rotary shaker at 200 rpm. After culturing for 24 h, the transformation of knockout subassembly was performed by the LiAc/SS carrier DNA/PEG method in combination with Zymolyase-100T (a kind of enzyme to digest yeast cell wall) [10]. After detailed analysis and preliminary experiments, 30 μl of Zymolyase-100T was added to the reaction system before transformation. Transformants were selected from the solid medium containing 10 g/l of glucose and 5 g/l of yeast extract. Then, 400 μg/ml of G418 sulfate (Gibco BRL Germany; batch 11811-031, activity may be batch dependent) was dissolved in water and added to autoclaved YPD cooled to 40℃ with a final active concentration of 200 μg/ml, and the positive clones were picked out by G418. PCR was performed, as previously described, to identify and exclude the false positives. Twenty nanogram of total DNA was used as the template in a 50 μl reaction mixture containing 5 μl of PCR buffer, 2 μl each of forward and reverse primers, and 2.5 units of Taq DNA polymerase. The PCR cycling conditions were as follows: 1 cycle at 95℃ for 3 min; 39 cycles at 95℃ for 30 sec, 58℃ for 30 sec, and 72℃ for 30 sec; followed by a final extension step at 72℃ for 10 min.

The Cre expression plasmid psh65 was transformed in the same manner. Zeocin-resistant transformants were selected from the YPD plates with 40 μg/ml of Zeocin (PHLEL0100; Cayla, Toulouse, France) and confirmed by PCR using the primers SC-F and SC-R, as described in Table 1.

Knockout HOG1 of Wild-Type T. oedocephalis

To knock out HOG1 of T. oedocephalis, the loxP-Kan-loxP/Cre system was adopted in this osmotolerant yeast. The loxP marker gene loxP/Cre system was constructed to carry out the deletion of HOG1, and the transformation of the yeast was as previously described. The Cre expression plasmid psh65 was constructed to carry the Kan’ gene from the bacterial transposon, which confers resistance to the antibiotic Zeocin. Finally, repeated genes were disrupted by the construction of loxP marker gene loxP cassettes and plasmid psh65. The engineered strains were cultured and subcultured for six generations to maintain their genetic stability. The To-HOG1 knockout mutant and wild-type strain were cultivated on agar plates consisting of 2% glucose, 1% yeast extract, 2% agar, and 2% peptone at 30℃ and pH 7.0.

Knockout of Gene Expression by Quantitative Real-Time PCR

For quantitative real-time PCR (qRT-PCR), T. oedocephalis strains were grown to late exponential phase. Total DNA was extracted using the DNA extraction kit, following the manufacturer’s protocol. Subsequently, the quantitative analysis of HOG1 was performed with a SYBR Green PCR Master Mix Kit on Step One (Applied Biosystems, USA), using time-F and time-R as forward and reverse primers for RT-PCR; the original DNA was diluted 0-, 10-, 100-, 1,000-, and 10,000-fold as templates. The conditions for qRT-PCR cycles were as follows: 95℃ for 6 min, followed by 40 cycles of 95℃ for 20 sec and 60℃ for 1 min. The standard curve method was applied to analyze the qRT-PCR data.

The deletion levels were determined with the qRT-PCR experiment using the DNA from the original and engineered strains. The reaction mixture was prepared with the SYBR Green PCR Master Mix (Applied Biosystems). Primers time-F and time-R (Table 1) were used to amplify HOG1. Approximately 5 ng of template cDNA was used in the reaction with the thermal profile of 2 min at 68℃, 3 min at 95℃, 25 cycles of 30 sec at 95℃ and 1 min at 68℃, followed by a dissociation curve. All reactions were repeated three times for each sample, and the ΔΔCt-method was analyzed by the formula 2 − (ΔCt target − ΔCt reference sample) [24].

Cultivation Test of Wild Type and To-HOG1 Knockout Mutant

The TO-HOG1 knockout mutant strains were inoculated in 250-ml shake flasks containing 50 ml of seed medium for 48 h at 30℃ with an agitation speed of 200 rpm, and the wild-type T. oedocephalis was studied as the control. The seed medium contained 10 g/l of glucose and 5 g/l of yeast extract. The seed culture was then transferred into the fermentation medium with a 5% inoculum level (v/v). The cultivation experiments were performed in 250 ml shake flasks at 30℃, 200 rpm, and pH 5.5. The fermented medium contained 200 g/l of glucose and 5 g/l of yeast extract. Moreover, in order to study the effect of citric acid on the production of polyols by the TO-HOG1 knockout mutant and wild-type strain, citric acid was added to the medium from a stock solution of 3.5 M, and the pH was adjusted to 5.5 using 6 M NaOH.

HPLC Analysis

High-performance liquid chromatography (HPLC) analysis of substrates and end-products showed consistent results with our previous report [13]. The samples of fermentation broth were filtered through 0.20 μm membranes and assayed for glycerol and erythritol by HPLC using a Rezex RCM-Monosaccharide column (300 mm × 7.8 mm; Phenomenex, Torrance, CA, USA) and detected using the Waters 2410 Refractive Index (RI) Detector. The mobile phase was ultrapure water. For sample analysis, the column was eluted at 60℃ with ultrapure water at a flow rate of 0.6 ml/min. This method enabled the quantification of glycerol and erythritol. All experiments were performed in triplicates.

GPDH Activity Analysis

In the present study, 1 ml of the fermentation broth along with 3 ml of methanol (−20℃) were homogenized in an ultrasonic ice bath for 10 min and then centrifuged at 12,000 ×g for 10 min at 4℃. Next, the supernatants were assayed for GPDH activity using GPDH activity assay kits (TaKara Bio Inc., Tokyo, Japan) following the methods described in previous studies [7,36]. The specific GPDH activity was determined as units per liter of the sample (U/l), where 1 unit of GPDH activity was defined as 1 ml of sample consuming 1 μmol of NADH in 1 min. The experiments were performed in triplicates and the average of three observations was taken.

 

Results and Discussion

Cloning of HOG1 and Construction of Knockout Subassembly

The hyperosmolar signaling pathway (HOG1) in fungi involves mitogen-activated protein kinases and regulates environmental signals by activating gene transcription involving the glycerol synthesis enzyme in order to obtain high concentrations of glycerol and maintain the turgor pressure of cells [12,22]. In 2005, Kayingo and Wong [15] tentatively researched MAPK HOG1 signaling pathways in Candida and observed that the production of D-arabinitol partly depends on the HOG1 signaling pathway under oxygen stress by changing the culture of stress to control the amount of glycerin and D-arabinitol. In 2013, Yoshida et al. [37] reported that the function of MAPK and polyhydric alcohol production were affected by the length of the C-terminal domain in HOG1, by comparing the gene encoding function for MAPK between Saccharomyces cerevisiae and T. megachiliensis SN-124A (Tm-HOG1). Thus, the HOG1 signaling pathway is involved in polyol production at the initial stages. However, the effect of HOG1 on fermentation in T. oedocephalis was not well understood.

In the present study, we first cloned HOG1 in T. oedocephalis. The agarose gel electrophoresis of HOG1 in T. oedocephalis (To-HOG1) is shown in Fig. 1A. The HOG1 gene of T. oedocephalis was 691 bp in length, which was shorter than that of the other yeasts mentioned above. To-HOG1 was compared with the Tm-HOG1 available from the NCBI database, using the BLASTP program. To-HOG1 showed a significant sequence homology with Tm-HOG1 (87%), and the maximum score was 73.4 with e-values lower than 6.00E-17. These findings indicated that the To-HOG1 obtained was credible, which would provide a solid basis for the further knockout of To-HOG1.

Fig. 1.Gel electrophoresis of PCR products. (A) Agarose gel electrophoresis induced by T. oedocephalis HOG1. Lane 1 represents the PCR marker. Lanes 2 and 3 represent PCR products of T. oedocephalis HOG1. The loading volume of sample is 5 μl. (B) Agarose gel electrophoresis induced by the gene interference segment (knockout subassembly). Lane 1 represents the PCR marker; Lanes 2 and 3 represent PCR products of the gene interference segment. The loading volume of sample is 5 μl.

To investigate the effects of To-HOG1 on polyol production, a gene disruption strategy was employed in the present study. For this purpose, a gene interference segment was constructed by PCR, and the quality of the PCR-amplified fragments was detected by agarose gel electrophoresis. As shown in Fig. 1B, the length of the PCR-amplified fragments ranged 1,500–2,000 bp, which was consistent with the result of the desired knockout subassembly. The knockout subassembly constructed in this study comprised the loxP-Kan-loxP sequence (1,653 bp) and 60 bp upstream sequences from To-HOG1 along with 60 bp downstream sequences. Therefore, this amplified fragment could be identified as the knockout subassembly. The transformants were selected randomly for further studies.

Transformation and qRT-PCR

At this stage, the cells of T. oedocephalis were transformed with a knockout subassembly that contained the loxP-Kan-loxP/Cre system and grown in YPD selection plates to identify the transformed cells. The transformation was performed by the LiAc/SS Carrier DNA/PEG method in combination with Zymolyase-100T, and the transformation rate was maintained at 30% [10].

The successful adoption of the loxP-Kan-loxP/Cre system for use in yeast is well known [2]. To knock out the HOG1 of T. oedocephalis, we constructed a loxP marker gene loxP/Cre system to carry out the deletion of T. oedocephalis HOG1. The scheme for deleting the T. oedocephalis HOG1 is presented in Fig. 2; this scheme was validated by agarose gel electrophoresis analysis. As shown in Fig. 3A, lane 1 is the marker and lanes 3 and 4 show the products of colony PCR ranging 1,500–2,000 bp, indicating that the sequence of the knockout subassembly was obtained again from the strains that were transformed once, called the “positive clones TO1.” The Cre expression plasmid psh65 was constructed to carry the Kan’ gene from the bacterial transposon, which confers resistance to the antibiotic Zeocin. As observed in Fig. 3B, lane 1 is the marker and lanes 3 and 5 show products of colony PCR with positive clones. Moreover, two bands are observed: the first band indicates that KanR was removed and only one loxP site was left; the second band indicates that KanR was retained; only the strains that lost KanR were named TO2. Finally, repeated genes were disrupted by the construction of loxP marker gene loxP cassettes and plasmid psh65. Furthermore, the engineered strains were cultured and subcultured for six generations to keep them genetically stable. Of the products of colony PCR through the template of the engineered strains shown in Fig. 3C, four measured 140 bp. Thus, it was obvious that both the HOG1 gene and KanR gene were deleted with only the loxP sequence left. The strains that successfully deleted HOG1 were named as TO3. Finally, in order to eliminate the error of colony PCR, we again extracted the DNA of the To-HOG1 knockout mutant strain. The results are shown in Fig. 3D. Lane 1 shows the PCR marker and lane 2 shows the PCR products of the To-HOG1 knockout mutant strain. There was no HOG1 gene, and only the loxP site was left.

Fig. 2.Construction of the HOG1 gene knockout subassembly. This knockout subassembly consists of 60 bp upstream and 60 bp downstream sequences of HOG1, and a 20 bp pug6 sequence. The HOG1 gene was substituted by knockout subassembly in either of the two chains, and this strain was named TO1. Deletion of the Kan’ site and one loxP site in strain TO1 gave strain TO2. Deletion of the HOG1 gene by knockout subassembly in another chain, which was the repetition of the (A) and (B) experimental procedures. HOG1 genes were deleted by knockout subassembly in two chains, and the engineered microorganism was named TO3.

Fig. 3.Agarose gel electrophoresis of the colony PCR products. All the loading volume of samples for agarose gel electrophoresis are 5 μl. (A) Lane 1 represents the PCR marker. Lanes 3 and 4 represent colony PCR products of strain TO1. (B) Lane 1 represents the PCR marker. Lanes 3 and 5 represent colony PCR products of strain TO2. (C) Lane 1 represents the PCR marker. Lanes 3 and 4 represent colony PCR products of strain TO3. (D) Lane 1 represents the PCR marker. Lane 2 represents the PCR product of the To-HOG1 knockout mutant strain.

qRT-PCR experiments were also performed to confirm HOG1 expression in the wild-type T. oedocephalis and To-HOG1 knockout mutants. Analysis of relative gene expression data using real-time quantitative PCR and the 2-ΔΔCt-method showed that the concentration of HOG1 was decreased to 0.204% in the To-HOG1 knockout mutants, compared with the original strain. Values <5% were considered statistically significant. The qRT-PCR data showed that in the To-HOG1 knockout mutant strain, HOG1 expression was significantly lower than in the wild type, which was in agreement with the results of the abovementioned transformation experiment. Altogether, these data further confirmed successful production of the To-HOG1 knockout mutant strain of T. oedocephalis.

In addition, in the solid YPD plate culture, the colonies of To-HOG1 knockout mutants grew more slowly than those of the wild-type strain. The colonies of the wild-type strain and the To-HOG1 knockout mutant are shown in Fig. 4. Distinct variation in the colonial morphology was observed in the To-HOG1 knockout mutant. In particular, the wild-type strain showed bigger colonies with a sebaceous surface. Although the To-HOG1 knockout mutant colonies also showed the sebaceous surface, they were smaller in size. As is already known, the HOG pathway plays an important role in multiple biological and physiological processes in Aspergillus fumigatus and Saccharomyces cerevisiae cells. Most of the genes of this pathway are involved not only in the adaptation to diverse stresses but also in the regulation of morphogenesis [20,28]. There seems to be a relationship between the HOG1 gene knockout in T. oedocephalis and the change in colony morphology.

Fig. 4.Disruption of HOG1 influences the growth of Trichosporonoides oedocephalis. (A) Colony growth of wild-type Trichosporonoides oedocephalis. (B) Colony growth of the To-HOG1 knockout mutant strain.

Comparison of Polyol Production between Wild-Type and Mutant Strains

Previous studies indicated that the decrease in the glycerol yield in the process of erythritol production is extremely essential and has potential industrial prospects. The HOG1 gene encodes a MAPK that plays an important role in osmoadaptation. In particular, HOG1 regulates the transcription of the gene encoding GPDH via the stress response elements. In this study, to investigate whether glycerol and erythritol are excreted by the parental or the To-HOG1 knockout mutant strain, T. oedocephalis ATCC 16958 (i.e., the parent strain) and its To-HOG1 knockout mutant strain (TO3) were cultivated using shake flasks in YPD medium containing 200 g/l glucose. The biomass, glycerol, and erythritol production from glucose by T. oedocephalis ATCC16958 and its mutant TO3 in shake flask are shown in Table 2. Cell dry weight data indicated that the stationary phase of cell growth was reached at 48 h of fermentation. The biomass of the mutant strain was slightly lower than that of the wild-type strain. As observed in Table 2, the glycerol concentration in the wild-type strain ranged 6.69–36.22 g/l, whereas the erythritol concentration ranged 6.03–39.37 g/l. In the mutant strain, the glycerol concentration increased from 2.17 to 10.42 g/l along with the increase in erythritol content from 19.89 to 56.82 g/l. This indicated that a nutritional resource was used in cell growth at the early stage, and the massive increase in the polyols was observed 72 h later. The amount of total erythritol plus glycerol was 1.12-fold higher in the parental strain T. oedocephalis ATCC16958 than in the To-HOG1 knockout mutant strain. In the To-HOG1 knockout mutant strain, however, deletion of the HOG1 gene improved the amount of erythritol by 1.44-fold and concomitantly decreased the glycerol production by 71.23%. In other words, knockout of HOG1 in T. oedocephalis resulted in a significant increase in erythritol biosynthesis. Several studies have focused on increasing erythritol production and reducing glycerol production using mutation technology. For example, the strain Candida magnoliae 12-2 was obtained by ultraviolet and chemical mutagenesis, and this mutant led to a 2.4-fold increase in erythritol (20.32 g/l) and a 5.5-fold decrease in glycerol production compared with the wild-type strain [9]. Moreover, mutants of Yarrowia lipolytica were generated by UV mutation for enhancing erythritol production and removing other byproducts (particularly glycerol) [8]. It must be emphasized that our results may provide a novel idea for increasing erythritol production. The results in Table 2 also showed that the value of Yp/s increased considerably in the To-HOG1 knockout mutant strain. The values of Yp/s after 120 h of fermentation by wild-type T. oedocephalis and its To-HOG1 knockout mutant strain were 0.197 and 0.284, respectively. This indicates that the efficiency of erythritol production from glucose increased greatly during the fermentation of the To-HOG1 knockout mutant strain, since HOG1 is known to regulate glycerol production in response to osmotic stress in some microbial species [32]. In this process, the transcription of the GPDH was regulated by the HOG1 pathway [31]. Furthermore, at fixed intervals of time, a similar amount of fermentation broth was sampled and the total GPDH activity was determined using the commercial detection kits. Compared with the To-HOG1 knockout mutant strain, the wild-type strain showed a 3.24-fold increase in GPDH activity after 120 h (Fig. 5). Thus, glycerol is a major osmoprotectant in T. oedocephalis and its synthesis is controlled by the HOG1 gene. With regard to S. cerevisiae, it has been reported that glycerol accumulates rapidly via the glycolytic pathway and the action of GPDH on dihydroxyacetone phosphate as a template. For glycerol production, the translocation and accumulation of HOG1 in the nucleus as well as HOG1 activation via phosphorylation are critical [28]. In this study, the disruption of HOG1 in T. oedocephalis resulted in decreased glycerol accumulation in the fermentation broths. Moreover, the metabolic shift strongly affected the flux partitioning between glycolysis and the pentose phosphate pathway (PPP). As is well known, erythritol is produced from the PPP, with a significant amount of carbon flux being channeled into the PPP in the To-HOG1 knockout mutant strain. As a result, erythritol production was greatly increased in the fermentation process. To our best knowledge, it has been reported that T. megachiliensis switches the compatible solute biosynthesis from glycerol to erythritol throughout the growth stage [17]. Our results suggest that erythritol was successfully accumulated by HOG1 gene knockout in T. oedocephalis. Moreover, the present study demonstrates the feasibility of engineering T. oedocephalis for the production of erythritol from D-glucose. Further studies, however, are needed to optimize the fermentation conditions in the bioreactor before the industrial application of erythritol production from D-glucose by the To-HOG1 knockout mutant strain.

Table 2.CDW: Cell dry weight; Yp/s: Erythritol yield from glucose.

Fig. 5.GPDH activity at different culture times in fermentation broths of wild-type T. oedocephalis and its To-HOG1 knockout mutant strain.

Effects of Citric Acid Stress on Polyol Production in Wild-Type and Mutant Strains

Citric acid is an intermediary metabolite of the tricarboxylic acid cycle and is a key component of normal respiratory metabolism in yeast cells. It activates the HOG1p MAPK pathway, which subsequently regulates the expression of a number of proteins [18]. Therefore, the effects of citric acid on polyol production by parental and To-HOG1 knockout mutant strains in shake-flask culture were also evaluated in the present study. The cultures were grown in a medium containing 200 g/l of glucose and 0.1-0.5% (w/v) citric acid. After 120 h of fermentation at 200 rpm and 30℃, the variations in biomass, and the values of Yp/s, erythritol concentration, and glycerol concentration could be clearly observed, as shown in Table 3. A citric acid concentration of 0.4% or more negatively affected strain growth and erythritol synthesis. With regard to wild-type and To-HOG1 knockout mutant strains with citric acid concentrations ranging 0.1–0.3%, the increase in the citric acid concentration contributed to an increase in the amount of erythritol and a decrease in glycerol concentration. For instance, with an increase in the citric acid concentration from 0.1% to 0.3%, the erythritol concentration increased from 56.82 to 69.12 g/l in the fermentation process of the To-HOG1 knockout mutant strain. This could be explained by the fact that citric acid promotes the absorption of calcium ions and inhibits the activity of pyruvate kinase, benefiting the PPP and enhancing the erythritol yield [35]. When the citric acid concentration was increased from 0.3% to 0.4%, irrespective of being produced by the wild-type strain or the To-HOG1 knockout mutant strain, the yield of erythritol and the value of Yp/s were obviously reduced. This might be attributed to an osmotic effect on the cells, or the relatively higher concentrations of citric acid that would inhibit cell growth and influence glucose-metabolizing enzymes [21]. In this study, the optimum concentration of citric acid supplement for erythritol production was therefore determined to be 0.3%. In comparison with the blank control, when the citric acid concentration was 0.3%, the production of erythritol by the wild-type and To-HOG1 knockout mutant strains improved by 18.21% and 21.65%, respectively.

Table 3.CDW: Cell dry weight; Yp/s: Erythritol yield from glucose.

Similar to Saccharomyces cerevisiae, relative to the wild-type control, when the concentration of citric acid was increased to 0.1% during cultivation of the To-HOG1 knockout mutant strain, the glycerol production varied proportionally, with its concentration increasing from 10.42 to 12.11 g/l (Table 3). Moreover, by increasing the citric acid content from 0 to 0.10%, the GPDH activity increased from 4.20 to 5.80. Glycerol production in the To-HOG1 knockout mutant strain was almost the same as that in Saccharomyces cerevisiae. This could be attributed to the upregulation of the genes and proteins involved in glycerol biosynthesis [33]. The results further indicated that the HOG1 of T. oedocephalis is involved in glycerol synthesis by regulating the expression of the glycerol synthesis gene. In contrast, when the concentration of citric acid was increased from 0.1% to 0.2% or 0.3%, glycerol production decreased slightly, but the concentration of erythritol increased markedly. At this point, erythritol is a major solute that accumulates in the To-HOG1 knockout mutant strain under conditions of citric acid stress. When the organisms were stressed under 0.3% citric acid, the concentration of erythritol increased rapidly in the To-HOG1 knockout mutant strain until it was 1.49-fold higher than the level of wild-type T. oedocephalis. These findings suggest that MAPK HOG1 participates in regulating glycerol and erythritol production during citric acid stress.