Skip to main content
  • Original article
  • Open access
  • Published:

Complementary function of two transketolase isoforms from Moniliella megachiliensis in relation to stress response


Two transketolase isogenes, MmTKL1 and MmTKL2, isolated from Moniliella megachiliensis were investigated for their roles in stress response and erythritol biosynthesis. The encoded proteins were highly homologous in amino acid sequence and domain structure. Two stress response elements (STREs) were found upstream of MmTKL1, while no STRE was found upstream of MmTKL2. In contrast, two Ap-1 elements were present upstream of MmTKL2, but none were detected upstream of MmTKL1. MmTKL2 partially complemented the aromatic amino acid auxotrophy of a Saccharomyces cerevisiae tkl1 deletion mutant, suggesting that at least one of the MmTKLs functioned as a transketolase in vivo. In response to short-term osmotic stress (20% glucose or 1.2 M NaCl) in Moniliella cells, MmTKL1 expression increased rapidly through the first 40 min before subsequently decreasing gradually, while MmTKL2 expression showed no significant change. In contrast, short-term oxidative stress (0.15 mM menadione) induced considerable increases in MmTKL2, while MmTKL1 expression remained low under the same conditions. Long-term osmotic stress (20% glucose) yielded increased expression of both genes starting at 12 h and continuing through 72 h. During either osmotic or oxidative stress, intracellular erythritol accumulation could clearly be correlated with the pattern of expression of either MmTKL1 or MmTKL2. These results strongly suggested that MmTKL1 is responsible primarily for the response to osmotic stress, while MmTKL2 is responsible primarily for the response to oxidative stress. Thus, we postulate that the two transketolase isoforms of M. megachiliensis play distinct and complementary roles in coordinating erythritol production in response to distinct environmental stresses.


Moniliella megachiliensis SN-124A, a yeast-like fungus isolated from dry fruit, is a highly osmotolerant microorganism. This strain grows even in a 60% glucose solution, and produces a significant amount of erythritol [with maximum yields exceeding 40% (w/v)] as an osmoregulatory compatible solute when cultivated in high-glucose medium (de Hoog et al. 2011; Ishizuka et al. 1989). Erythritol, the sugar alcohol of tetrose, is distributed widely in nature, and is present at low levels in fruits, mushrooms, and fermented foods (Shindo et al. 1988; Yoshida et al. 1984). Erythritol is currently used as a low calorie sweetener, having a cool and plain sweetness (70% that of sugar) and low energy value (0.4 kcal/g) due to being non-metabolizable in the human body (Noda et al. 1994). Furthermore, erythritol has been reported to have antioxidant properties: the compound is an efficient hydroxyl radical scavenger and may help protect against hyperglycemia-induced vascular damage (den Hartog et al. 2010). Hence, this polyol could serve as an antioxidant sweetener for use by diabetics. Moreover, erythritol recently has been proposed for use as a feedstock (in place of petrochemicals) for the synthesis of bio-plastics (Amada et al. 2012). Owing to these appealing characteristics, demand for erythritol continues to expand year by year.

In the prokaryotic bacteria Oenococcus oeni (formerly Leuconostoc oenos), erythritol is synthesized from erythritol-4-phosphate, which is itself generated from fructose-6-phosphate by the action of phosphoketolase (Veiga-da-Cunha et al. 1992, 1993; Richter et al. 2001). In contrast, in eukaryotes like Moniliella or Candida, erythritol is produced via the pentose phosphate pathway (PPP), whereby the precursor erythrose, formed through dephosphorylation of erythrose-4-phosphate, is enzymatically reduced to erythritol by an NADP+-dependent erythrose reductase (ER) (Lee et al. 2003; Ookura et al. 2005; Kim et al. 2013). Previously, we noted that ER, a member of the aldoketo reductase family, is not found in Saccharomyces cerevisiae, and that transgenic expression of ER (an ER-encoding gene) in S. cerevisiae does not provide ER activity, despite the accumulation of ER protein (Kobayashi et al. 2013). Nonetheless, the PPP is considered to play an important role in erythritol biosynthesis (Kobayashi et al. 2013; Sawada et al. 2009). Among the enzymes involved in PPP, transketolase (TKL) is a key enzyme that transfers a ketol group from xylulose-5-phosphate to ribose-5-phosphate, yielding glyceraldehyde-3-phosphate and sedoheptulose-7-phosphate. In addition, TKL can produce fructose-6-phosphate and glyceraldehyde-3-phosphate via ketol transfer from xylulose-5-phosphate to erythrose-4-phosphate (Lindqvist et al. 1992; Schenk et al. 1998; Nikkola et al. 1994). Notably, fructose-6-phosphate and glyceraldehyde-3-phosphate also are intermediates that participate in glycolysis. Hence, TKL is a multi-functional enzyme that modulates both the pentose phosphate and glycolytic pathways in the cell.

Escherichia coli harboring a knock-out of the TKL-encoding gene require aromatic amino acids (AAAs) for normal growth; this deficiency reflects the requirement for erythrose-4-phosphate as a precursor for the synthesis of AAAs such as phenylalanine, tyrosine, and tryptophan (Josephson and Fraenkel 1969). S. cerevisiae harboring a deletion in the TKL1 (one of two yeast TKL-encoding paralogs) was still able to grow in synthetic complete medium lacking AAAs (Sundström et al. 1993). This observation suggested that S. cerevisiae possesses two isogenes of TKL (TKL1 and TKL2), and that TKL2 could complement the AAAs synthesis function of TKL1 (Schaaff-Gerstenschläger et al. 1993). However, the detailed mechanisms of regulation and roles of two paralogs remained unclear.

In the present study, we identified two TKL isogenes from M. megachiliensis (MmTKL1 and MmTKL2). To investigate the role of these two genes in cell metabolism under conditions of stress, we analyzed (1) the structure of the genomic regions flanking the MmTKL1 and MmTKL2 coding regions and (2) the expression profiles and functions of each gene in relation to stress response and erythritol biosynthesis. This work provides new insights into the physiological significance of TKL isoforms and of erythritol biosynthesis in the cell protection systems of microorganisms. In addition, the results are expected to facilitate improved yields of erythritol, a promising biomaterial, via metabolic engineering.

Materials and methods

Strains and growth condition

The microbial strains and plasmids used in this study are listed in Table 1. E. coli DH5α (TOYOBO, Osaka, Japan), used as a host cell for molecular constructs, was cultured overnight in LB medium (2.0% tryptone, 1.0% yeast extract, 2.0% NaCl) at 37 °C. M. megachiliensis SN-124A (National Food Research Institute, Microbial Gene Bank, Tsukuba, Japan) was pre-cultured in 100 mL GY medium (2% glucose and 0.5% yeast extract) or YPD medium (1.0% yeast extract, 2.0% peptone, 2.0% glucose) in a 500-mL flask at 30 °C and 200 rpm until the OD600 reached 1.0. The pre-culture then was used as a 1% inoculum (i.e., 10 mL into 1 L; in a 3-L flask) in GY or YPD medium, and the resulting culture was further cultivated at 30 °C and 160 rpm until the OD600 reached 1.0.

Table 1 Fungal strains and plasmids used in this study

For short-term stress loading, the 1-L GY or YPD cultures then were divided into 100-mL aliquots, and individual aliquots were supplemented by the addition of a stressing agent (glucose, NaCl, or menadione). Specifically, stressing agents were added to yield final concentrations as follows: 20% (w/v) glucose in GY medium, 1.2 M NaCl in YPD medium, or 0.15 mM menadione in YPD medium. The resulting 100-mL cultures then were incubated at 30 °C and 200 rpm for up to 120 min.

For long-term stress loading, pre-culture prepared as above was used as 1% inoculum in 100 mL of GY medium containing 20% glucose. The resulting culture was incubated at 30 °C and 200 rpm for 72 h, and aliquots were harvested for analysis every 12 h.

Cloning of TKL genes from M. megachiliensis

The partial DNA sequence of a gene encoding a microbial TKL homolog was obtained from the cDNA library of M. megachiliensis SN G-42 (mutant strain of M. megachiliensis SN-124A). Primers specific for that sequences were designed and used to screen (via direct phage plaque PCR) a M. megachiliensis SN-124A genomic phage library for that TKL-encoding clone. The nucleotide sequence of the recovered clone was determined and used to design additional primers for recovering the full-length gene (designated MmTKL1) from the genomic phage library. Homology searches of the M. megachiliensis SN-124A draft genome sequence with MmTKL1 identified a second gene encoding a TKL homolog. The corresponding DNA sequence was used to design gene-specific primers, which were in turn used to PCR amplify the second gene (designated MmTKL2) using M. megachiliensis SN-124A cDNA as the template. The nucleotide sequence of the recovered clones were determined and used to clone genomic sequences corresponding to the region upstream of MmTKL2. MmTKL1 and MmTKL2 then were separately cloned into the pGEM® T-easy vector (Promega, Co., Madison, WI). The primers used in this experiment are listed in Table 2. Similarly, the amplified TKL1 gene fragment was obtained from cDNA of S. cerevisiae BY4741 using PCR with degenerate primers, ScTKL1-Met-forward and ScTKL1-Stop-reverse (Table 2). The resulting fragment was cloned into pGEM®-T Easy vector (Promega Corporation, Madison, WI, USA).

Table 2 Primers used in this study

Nucleotide sequence analysis

The nucleotide sequences of DNAs were analyzed using the BigDye® Terminator v3.1/1.1 Cycle Sequencing Kit (Applied Biosystems, Tokyo, Japan) with an ABI automatic sequencer (PerkinElmer Japan, Tokyo, Japan). The nucleotide and amino acid sequence data were processed using GENETYX-Mac NETWORK software, version 15 (GENETYX CORPORATION, Tokyo, Japan). For both MmTKL1 and MmTKL2, the 1000 bp upstream of each gene were searched using the TFSEARCH database for potential transcription factor binding motifs.

Yeast transformation

A yeast centromere plasmid vector was constructed to evaluate whether MmTKL1 and MmTKL2 can complement the S. cerevisiae BY4741 Δtkl1 mutant and rescue the phenotypes associated with this mutation, the growth deficiency in the SC plate lacking AAAs. Specifically, the multi-cloning site of pBluescript II SK (+) was inserted into a yeast centromere plasmid vector (pDEST32; Invitrogen, Carlsbad, CA, USA) that had been digested with HindIII and SacI; the resulting episome was designated pDB05 (Yoshida et al. 2013). NotI-ended fragments harboring MmTKL1, MmTKL2, or ScTKL1 (obtained from the corresponding clones in the pGEM® T-easy vector) were ligated into NotI-digested pDB05 to generate pDB05-MmTKL1, pDB05-MmTKL2, or pDB05-ScTKL1 respectively. Plasmids with and without inserts were separately transformed into the S. cerevisiae Δtkl1 mutant using the lithium chloride method (Finlayson et al. 1991).

Growth test of transformants

Leu+ transformants were selected on SC medium (synthetic complete medium) lacking leucine (SC-Leu). To evaluate the function of the MmTKL genes in vivo, the transformants were grown to logarithmic growth phase in SC medium. Following the cell pellets were resuspended in sterile distilled water at densities of 2 × 107 cells/mL. Suspensions were subjected to six steps of tenfold serial dilutions, and the dilutions were spotted (5 µL/spot) on parallel plates of SC agar medium with and without added AAAs (tryptophan, phenylalanine, and tyrosine).

Total RNA preparation and cDNA synthesis

Total RNA was extracted from M. megachiliensis using the previously described lithium chloride method (Iwata et al. 2015). An aliquot of 50 µg of total RNA then was treated with recombinant DNase I (Takara Bio, Shiga, Japan), extracted with PCI (phenol: chloroform: isoamylalcohol = 25:24:1), and purified by ethanol precipitation. cDNA was synthesized from 5 µg of the resulting RNA preparation using PrimeScript™ Reverse Transcriptase (Takara Bio, Shiga, Japan) according to the manufacturer’s protocol.

Semi-quantitative PCR and real-time PCR

Semi-quantitative PCR amplification was performed using Go Taq® polymerase (Promega, Madison, WI). The reaction mixtures were prepared using the manufacturer’s protocol. One round of PCR consisted of denaturation at 95 °C for 30 s, annealing at 58 °C for 30 s, and extension at 72 °C for 30 s. The amplified DNA fragments were detected by agarose-gel electrophoresis. Real-time quantitative PCR was performed using reaction mixtures incorporating Thunderbird SYBR qPCR Mix (TOYOBO, Osaka, Japan) according to the previously reported method (Iwata et al. 2015). The amplification of each cDNA template was performed in triplicate independently using a CFX96 real-time PCR system (Bio-Rad Laboratories, Inc., CA, USA). The primer sets used for PCR amplification are listed in Table 2.

Determination of intracellular polyol content

The amount of intracellular polyol was determined as previously described (Kobayashi et al. 2013). Briefly, cell suspensions of M. megachiliensis were immersed in boiling water for 10 min, mixed with 100% (w/v) trichloroacetic acid, and vigorously stirred with a vortex mixer for 20 min at room temperature. After centrifugation at 13,400×g for 10 min, polyol in the supernatant was amperometrically assayed using a 4 mm × 250 mm CarboPac MA-1 column with an ICS-3000 chromatographic system (Dionex, Osaka, Japan) at flow rate of 0.3 mL of 500 mM NaOH as solvent per minutes. For dry cell weight determinations, cell suspensions (1 mL each) were dried by incubating for 5 h in a drying oven maintained at 80 °C. The polyol and dry cell weight examination were performed in triplicate independently.


Cloning and sequencing of MmTKL genes from M. megachiliensis

Two transketolase isogenes were isolated from M. megachiliensis; these genes, designated MmTKL1 (Accession Number LC163538) and MmTKL2 (Accession Number LC163539), were identified from the genomic phage library and draft genome sequence data, respectively. MmTKL1 consists of a 1956-bp ORF predicted to encode a 652-amino-acid protein with an estimated molecular mass of 70.8 kDa. This amino acid sequence shows 61% identity to putative or confirmed transketolases from Cryptococcus bacillisporus WM276 (Accession Number ADV_20663.1), C. neoformans var. neoformans JEC21 (Accession Number XP_570357. 1), Puccinia granminis f. sp. tritici CRL 75-36-700-3(Accession Number EFP74700. 1), and Coprinopsis cinerea okayama7#130 (Accession Number XP_001836588. 2), and 28% identity to the S. cerevisiae transketolase 1 (Accession Number NP_015399). MmTKL2 also consists of a 1956-bp ORF encoding a 652-amino-acid protein with an estimated molecular mass of 70.8 kDa. This amino acid sequence exhibits 66% identity to the Cryptococcus gatti WM276 transketolase (Accession Number XP_003192450), 65% identity to the C. neoformans var. neoformans JFC21 transketolase (Accession Number XP_570357.1), and 26% identity to the S. cerevisiae transketolase 1 (Accession Number NP_015399). MmTKL1 and MmTKL2 show 74% identity with each other.

BLAST search analysis revealed that both MmTKL1 and MmTKL2 are composed of domains typically shared among known TKLs (Fig. 1). From the N- to C-termini, these domains correspond to the TPP-binding module of the TKL subfamily of the thiamine pyrophosphate (TPP) family; the pyrimidine (PYR) -binding domain of 1-deoxy-D-xylulose-5-phosphate synthase (DXS); and the C-terminal domain of TKLs. The Ile189, Ile416, and Ala449 residues present in the TPP-binding site of S. cerevisiae TKL1 appear to correspond to respectively Val205, Val377, and Val413 in MmTKL1, and to respectively Val206, Ile414, and Ala449 in MmTKL2.

Fig. 1
figure 1

Alignment of amino acid sequence and domain structure of transketolases. MmTKL1, MmTKL2, and ScTKL1 indicate respectively the transketolase 1 and transketolase 2 proteins from Moniliella megachiliensis and the transketolase 1 from Saccharomyces cerevisiae. Light gray zone thiamine pyrophosphate (TPP) family, transketolase subfamily, TPP-binding module; gray zone thiamine pyrophosphate (TPP) family, pyrimidine (PYR)-binding module; dark gray zone transketolase, C-terminal module

Two putative STREs (stress response element: AGGGG or CCCCT; binding sites involved in osmotic stress response) were detected in the sequences upstream of the MmTKL1 ORF (at −556 to −561 bp and −220 to −225 bp with respect to the start codon).In contrast, two putative AP-1 s (activator protein 1 response elements: TGACTCA or TGAGTCA; binding sites involved in oxidative stress response) were detected in the sequences upstream of the MmTKL2 ORF (at −891 to −881 bp and −689 to −669 bp with respect to the start codon) (Fig. 2).

Fig. 2
figure 2

Putative stress response element (STREs) and Ap-1 response elements (AP-1 s) detected upstream of the MmTKL1 and MmTKL2 ORFs. Analysis was limited to the 1000 bp upstream of each ORF. Black bow tie STRE, gray bow tie AP-1 element, numerals represent base position with respect to ORF start codon

In vivo complementation by MmTKL1 and MmTKL2

To investigate whether MmTKL1 and MmTKL2 can function in vivo as TKLs, we evaluated the ability of each gene to complement a S. cerevisiae tkl1 mutant (An equivalent mutant of M. megachiliensis has not yet been constructed). Complementation was evaluated based on cell growth on SC agar medium lacking AAAs (SC-AAA), since TKLs are involved in AAA biosynthesis via effects on the supply of erythrose-4-phosphate. Both the parent strain (S. cerevisiae BY4741) and a complemented mutant [S. cerevisiae tkl1 transformed with a plasmid containing the S. cerevisiae gene (ScTKL1)] grew well on SC-AAA plates. The yeast mutant transformed with the MmTKL2 plasmid showed significant growth on the SC-AAA plate, although this growth was less vigorous than that observed in the positive controls. In contrast, growth of the yeast mutant harboring the MmTKL1 plasmid was indistinguishable from that of the uncomplemented mutant or from the mutant transformed with empty plasmid (pDB05) (Fig. 3).

Fig. 3
figure 3

Complementation of Saccharomyces cerevisiae tkl1 mutant by MmTKL1 and MmTKL2. This assay tested complementation of the aromatic amino acid (AAA) auxotrophy of the S. cerevisiae BY4741 Δtkl1 mutant by plasmids harboring MmTKL1, MmTKL2, or ScTKL1 (the endogenous S. cerevisiae gene) or with no insert (pDB05). Also tested were the untransformed mutant and the wild-type parent (BY4741). Growth was tested by spotting serial dilutions at an estimated 105–100 cells/spot, left to right on parallel synthetic complete (SC) agar medium with or without added AAAs (tryptophan, tyrosine, phenylalanine) and incubating the plates at 30 °C for 3 days

Expression of MmTKL genes and erythritol production under short-term osmotic stress

Gene expression in M. megachiliensis was analyzed using semi-quantitative PCR during short-term (120-min) growth under various hyperosmotic conditions. In the presence of 20% glucose, the expression level of MmTKL2 did not change with time, while that of MmTKL1 gradually increased up to 40 min before subsequently falling (Fig. 4a). In the presence of 1.2 M NaCl, MmTKL1 transcript accumulated with time, reaching a peak at 50 min 11-fold higher than baseline at 0 min (Fig. 5). However, hypersalinity did not result in apparent changes in MmTKL2 expression levels through 120 min.

Fig. 4
figure 4

Gene expression and intracellular erythritol accumulation under short-term osmotic stress with 20% glucose in Moniliella megachiliensis. a Levels of gene expression were analyzed by semi-quantitative PCR analysis; data provided are representative ultraviolet-illuminated bands following electrophoresis in ethidium bromide-stained agarose gels. b Intracellular erythritol content is presented as µg/mg dry cells (mean ± SD from three independent determinations). When OD600 reached approximately 1.0, glucose was added to the cell culture to a final concentration of 20%, and the culture was incubated for 120 min at 30 °C. Gene expression was analyzed by semi-quantitative PCR using GAPDH as a standard. Intracellular erythritol content was chromatographically determined with an ICS-3000 system, and is represented as µg/mg dry cells

Fig. 5
figure 5

Gene expression and intracellular erythritol accumulation under short-term osmotic stress with 1.2 M NaCl in Moniliella megachiliensis. When OD600 reached approximately 1.0, NaCl was added to the cell culture to a final concentration of 1.2 M, and the culture was incubated for 120 min at 30 °C. Gene expression was analyzed by real-time PCR, and normalized using GAPDH expression level as a standard. Intracellular erythritol content was determined chromatographically with an ICS-3000 system, and is presented as µg/mg dry cells

In short-term osmotic stress in 20% glucose medium, intracellular erythritol accumulation appeared to track with that of the MmTKL1 transcript, with erythritol levels peaking at fivefold above baseline at 40 min before subsequently falling gradually (Fig. 4b).

Expression of MmTKL genes and erythritol production under short-term oxidative stress

Under oxidative stress (induced by supplementation with 0.15 mM menadione), the expression profiles of MmTKL1 and MmTKL2 differed considerably. MmTKL2 showed marked induction of expression during the 120-min time course, with transcript levels after 90 min peaking at 12-fold over those at baseline (0 min) (Fig. 6). In contrast, levels of MmTKL1 remained low throughout the 120 min of oxidative stress (Fig. 6). Notably, the patterns of expression of the two genes under oxidative stress were complementary to the patterns seen under osmotic stress with glucose or NaCl. Intracellular erythritol levels under oxidative stress increased with time through 120 min.

Fig. 6
figure 6

Gene expression and intracellular erythritol accumulation under short-term oxidative stress with 0.15 mM menadione in Moniliella megachiliensis. When OD600 reached approximately 1.0, menodione was added to the cell culture to a final concentration of 0.15 mM, and the culture was incubated for 120 min at 30 °C. Gene expression was analyzed by real-time PCR, and normalized using GAPDH expression level as a standard. Intracellular erythritol content was determined chromatographically with an ICS-3000 system, and is presented as µg/mg dry cells

Expression of MmTKL genes and erythritol production during long-term growth

Gene expression analysis also was performed during long-term (72-h) growth under various stress conditions. In medium containing 20% glucose, expression of both MmTKL1 and MmTKL2 began increasing rapidly starting at around 12 h (early exponential growth) (Fig. 7). The levels of MmTKL1 transcript peaked at 48 h before subsequently decreasing slightly; in contrast, MmTKL2 transcript continued to accumulate through 72 h under these conditions.

Fig. 7
figure 7

Gene expression and intracellular erythritol accumulation during long-term growth. M. megachiliensis was cultivated at 30 °C for 72 h in GY medium containing 20% glucose. Every 12 h, samples were removed for analysis. Gene expression was analyzed by real-time PCR, and normalized using GAPDH expression level. Intracellular erythritol content was determined chromatographically with an ICS-3000 system, and is presented as µg/mg dry cells

In long-term osmotic stress in 20% glucose medium, intracellular erythritol levels increased rapidly starting at around 24–36 h after inoculation, with the increase tracking with those of both MmTKL1 and MmTKL2 expression. Peak erythritol levels were 40-fold higher than those at 12 h (Fig. 7).


Two TKL-encoding isogenes (MmTKL1 and MmTKL2) were cloned and sequenced from M. megachiliensis, a hyper-osmotolerant basidiomycetous yeast-like fungus. The amino acid sequences of the MmTKL1 and MmTKL2 proteins exhibited lower levels of identity to that of S. cerevisiae TKL1 (28 and 26%, respectively) than to those of other basidiomycetous fungi. Strikingly, several amino acid residues known to be involved in TPP binding were substituted in the MmTKL proteins compared to the corresponding residues of ScTKL1. However, all these substitutions were conservative, representing amino acids of the same family with hydrophobic side chains. Thus, despite sequence divergence, MmTKL1 and MmTKL2 were expected to possess functions similar to those of the TKLs of S. cerevisiae.

The MmTKL1 and MmTKL2 proteins were found to exhibit strongest homologies (61–66% identity) to TKL proteins of Cryptococcus or Puccinia species. Pathogenic Cryptococcus (Wong et al. 1990) and Aspergillus (Wong et al. 1989) have been reported to accumulate high levels of mannitol in response to hyper-osmotic environments, a strategy that enhances survival when infecting host cells. S. cerevisiae or Candida glycerinogenes also are known to accumulate glycerol in response to conditions of hyper-osmosis, with glycerol serving as an osmo-regulatory-compatible solute (O’ Rourke et al. 2002; Chen et al. 2008). Under hyper-osmotic conditions, the S. cerevisiae Hog1 protein (the downstream-most protein kinase of the HOG (high osmolarity glycerol) pathway is activated via phosphorylation and rapidly translocates to the nucleus (Edmunds and Mahadevan 2004). Upon translocation to the nucleus, phosphorylated Hog1 (in cooperation with other transcription factors) stimulates transcription of the genes encoding GPD1 (glycerol-3-phosphate dehydrogenase 1) and GPP2 (glycerol-3 phosphate phosphatase 2) via STREs located upstream of the corresponding genes, resulting in glycerol biosynthesis (Alepuz et al. 2001; Ansell et al. 1997). Hence, STREs are believed to play an essential role in the osmotic stress response in yeast cells. We previously reported that M. megachiliensis possesses three erythrose reductase isogenes (ER1, ER2, ER3) and two transaldolase isogenes (TAL1, TAL2); the products of these genes are known to be involved in erythritol biosynthesis (Kobayashi et al. 2013, Iwata et al. 2015). Among these genes, ER3 and TAL2 harbor putative STREs within 1000 bp upstream of the respective ORFs. In the present work, we noted that the MmTKL1 ORF is preceded (at −556 and −220 bp) by two upstream putative STREs; in contrast, the MmTKL2 ORF appears to lack STREs within 1000 bp upstream of the initiation codon. Hence, we postulate that MmTKL1 is involved in regulation of the osmotic-stress response via the PPP. On the other hand, two putative AP-1 elements were found upstream of the MmTKL2 ORF, but not proximal to MmTKL1. Since AP-1 has been shown in other organisms to mediate responses to oxidative stress (Toone and Jones 1999), we hypothesize that MmTKL2 is involved in the oxidative stress response. Similar results have been obtained for MmTAL1 (Iwata et al. 2015).

In addition to stress response, the function of MmTKL1 and MmTKL2 was evaluated based on another criterion: nutritional requirement. The erythrose-4-phosphate generated by TKLs can be converted (via the PPP; in microorganisms, fungi, and plants) to AAAs by way of the shikimic acid pathway (Hermann and Weaver 1999). Notably, a S. cerevisiae tkl1 deletion mutant is not able to grow in synthetic complete medium lacking AAAs. Our transformation tests demonstrated that MmTKL2 (but not MmTKL1) can partially complement the AAA auxotrophy of a S. cerevisiae tkl1 mutant. We have confirmed this result, including demonstration that the MmTKL1 plasmid is indeed present in the S. cerevisiae transformant (data not shown). The reason for this failure to complement is unknown, but may reflect the absence of heterologous expression of MmTKL1 in the yeast background under the plate assay conditions. Meanwhile, we are not convinced that MmTKL1 and MmTKL2 are orthlogs of TKL1 and TKL2 of S. cerevisiae, respectively, because number of MmTKL is not strictly defined. In fact, we have found three TKL homologues in M. megachiliensis draft genome sequence decoded, and obtained two of them, MmTKL1 and MmTKL2, as shown in this study. It is unclear that putative third TKL gene may complement S. cerevisiae TKL function. Analyses to determine MmTKL1 expression in the transformant and further, putative third gene will be needed.

In the presence of 20% glucose in M. megachiliensis, endogenous MmTKL1 expression peaked at 40 min after osmotic stress loading before subsequently gradually decreasing through 120 min. In contrast, MmTKL2 showed an approximately constant expression level during this osmotic stress interval. Similar results were obtained for gene expression profiles under NaCl-induced osmotic stress. In contrast, distinct results were obtained under conditions of short-term (120-min) oxidative stress, with the level of MmTKL1 expression remaining relatively low while MmTKL2 transcript exhibited marked accumulation. These results implicate MmTKL1 as a major mediator of the response to hyper-osmotic stress; in contrast, MmTKL2 is inferred to be a major mediator in the response to oxidative stress. The oxidative stressor used here (menadione) is metabolized to semiquinone by the oxido-reductase system of the cell, and then subsequently converted to a quinone that generates reactive oxygen species (ROS) (Yamashoji et al. 1991). ROS often induce oxidative damage and impair cell survival (Yashiki and Yamashoji 1996). S. cerevisiae TKL1 reportedly is induced by oxidative agents like hydrogen peroxide or acetoaldehyde (Jamieson 1998). Furthermore, TKL mediated by Yap1p and Skn7p in S. cerevisiae has been reported to contribute to the regulation of glutathione and NADPH for cell redox homeostasis (Carter et al. 2005; Slekar et al. 1996). We postulate that the ROS generated by menadione degradation similarly induces MmTKL2 expression in M. megachiliensis, such that this isoform of TKL contribute to the regulation of glutathione and NADPH for eliminating ROS in this yeast-like fungus.

We used 72-h growth in medium containing 20% glucose to model long-term exposure to hyper-osmotic conditions; these conditions parallel those used in industrial fermentation for production of erythritol. Under these conditions, MmTKL1 expression rapidly increased starting at 12 h and reached a maximum at 48 h, while MmTKL2 expression increased throughout the 72-h experiment. A clear correlation was observed between MmTKL1 expression and erythritol accumulation under conditions of hyper-osmotic glucose stress; no such correlation was observed between MmTKL2 expression and erythritol production. As postulated for short-term stress, MmTKL1 expression appears to be associated with the long-term response to osmotic stress in this organism. The expression of MmTKL2 increased with time during long-term osmotic stress, and this isogene may contribute to elimination of ROS that accumulate during long-term stress in high-glucose culture, which is more or less similar to oxidative stress caused by menadione. Regarding stress responses, compensatory behavior of isogenes is known to apply to S. cerevisiae GPD1 and GPD2, which encode isoforms of a key enzyme of glycerol biosynthesis (Ansell et al. 1997).

Based on the results obtained here, we consider that MmTKL1 is involved in the M. megachiliensis response to osmotic stress. In contrast, MmTKL2 appears to be involved in the response to oxidative stress, while also contributing to the AAA supply that is essential for growth in synthetic and minimal media. Intriguingly, S. cerevisiae also has been reported to possess two TKL isogenes, TKL1 and TKL2. Based on mutant phenotype, TKL1 is presumed to contribute to the supply of AAAs; the function of the ScTKL2 isogene presumed complementary of TKL1 remains unclear.

In summary, our results suggest that MmTKL1 and MmTKL2 may play distinct and complementary roles in M. megachiliensis defense against environmental stress, mediated by induction of erythritol production. To our knowledge, the results obtained in our study are the first instance of complementary function of TKL isogenes in association with stress response. We are now going to analyze the detailed mechanism of erythritol biosynthesis involved in ROS elimination in stress response of M. megachiliensis.


Download references

Authors’ contributions

Experiment planning and design: HI, TW, JO, and KT; cloning of TKLs and gene expression analysis: HI, YK, and TK; Erythritol production and analysis: HI, and DM; HPLC analysis of erythritol: HI, YK, and DM; heterologous expression and analysis: WI, and TK; Paper construction: HI, TW, JO and TK. All authors read and approved the final manuscript.


We thank Mr. Yosuke Tanaka of Nihon University for participating in critical genetic research and useful discussions.

Competing interests

The authors declare that they have no competing interests.

Availability of data and materials

All data and materials supporting conclusions of the manuscript are included within the article. Translated amino acid sequences obtained in this study are available in the DNA Data Bank of Japan.

Ethical approval and consent to participate

This article does not contain any studies with human participants or animals performed by any of the authors.


A part of this study was supported by a grant from the Advanced Low Carbon Technology Research and Development Program (ALCA), a project of the Japan Science and Technology Agency (JST), Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Takafumi Kasumi.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Iwata, H., Kobayashi, Y., Mizushima, D. et al. Complementary function of two transketolase isoforms from Moniliella megachiliensis in relation to stress response. AMB Expr 7, 45 (2017).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: