OUP user menu

Altered expression and activities of enzymes involved in thiamine diphosphate biosynthesis in Saccharomyces cerevisiae under oxidative and osmotic stress

Ewa Kowalska, Marta Kujda, Natalia Wolak, Andrzej Kozik
DOI: http://dx.doi.org/10.1111/j.1567-1364.2012.00804.x 534-546 First published online: 1 August 2012


Thiamine diphosphate (TDP) serves as a cofactor for enzymes engaged in pivotal carbohydrate metabolic pathways, which are known to be modulated under stress conditions to ensure the cell survival. Recent reports have proven a protective role of thiamine (vitamin B1) in the response of plants to abiotic stress. This work aimed at verifying a hypothesis that also baker's yeast, which can synthesize thiamine de novo similarly to plants and bacteria, adjust thiamine metabolism to adverse environmental conditions. Our analyses on the gene expression and enzymatic activity levels generally showed an increased production of thiamine biosynthesis enzymes (THI4 and THI6/THI6), a TDP synthesizing enzyme (THI80/THI80) and a TDP-requiring enzyme, transketolase (TKL1/TKL) by yeast subjected to oxidative (1 mM hydrogen peroxide) and osmotic (1 M sorbitol) stress. However, these effects differed in magnitude, depending on yeast growth phase and presence of thiamine in growth medium. A mutant thi4Δ with increased sensitivity to oxidative stress exhibited enhanced TDP biosynthesis as compared with the wild-type strain. Similar tendencies were observed in mutants yap1Δ and hog1Δ defective in the signaling pathways of the defense against oxidative and osmotic stress, respectively, suggesting that thiamine metabolism can partly compensate damages of yeast general defense systems.

  • oxidative stress
  • osmotic stress
  • thiamine phosphate synthase
  • thiamine pyrophosphokinase
  • transketolase
  • vitamin B1


Prokaryotes as well as some eukaryotic organisms (fungi and green plants) can synthesize thiamine (vitamin B1) de novo from simple precursors, whereas animals are dependent on its nutritional uptake (Friedrich, 1987). The active form of vitamin B1 is thiamine diphosphate (TDP), a universal cofactor involved in pivotal cellular pathways. Thiamine and TDP biosynthesis (Fig. 1) includes the separate formation of two aromatic components, 4-amino-5-hydroxymethyl-2-methylpyrimidine diphosphate (HMP-PP) and 5-(2-hydroxyethyl)-4-methylthiazole phosphate (HET-P), which are then condensed to form thiamine monophosphate (TMP). TMP is subsequently dephosphorylated to free thiamine, which can be transformed to TDP by thiamine pyrophosphokinase (Kowalska & Kozik, 2008). Albeit in general known and similar to those in bacteria and plants, the thiamine biosynthetic pathways in yeast are structurally and mechanistically incompletely understood. The rate of this process is controlled by the intracellular TDP level and in the regulatory mechanisms in yeast engage a set of proteinaceous transcription factors (Kawasaki et al., 1990; Nishimura et al., 1992).


Thiamine (vitamin B1) biosynthesis in yeast. The scheme shows the synthetic pathways of late thiamine precursors, thiamine activation (i.e., TDP formation) and TDP distribution according to the subcellular localization of major TDP-dependent enzymes. The symbols of the genes involved are specified. Dashed arrows indicate the pathways for salvage of the two aromatic moieties of thiamine. Double strand arrows symbolize the transport of thiamine and TDP. Dotted arrows show the incorporation of TDP into cytosolic and mitochondrial TDP-dependent enzymes as well as the nucleus-localized regulation of thiamine biosynthesis level. Other abbreviations are HET, 5-(2-hydroxyethyl)-4-methylthiazole; HET-P, HET phosphate; HMP, 4-amino-5-hydroxymethyl-2-methylpyrimidine; HMP-P, HMP phosphate; HMP-PP, HMP diphosphate; TKL, transketolase.

Thiamine has recently been proven to play important roles in the responses of green plants to stress factors such as a pathogen attack or oxidative and osmotic stress chemical inducers (Ahn et al., 2005; Rapala-Kozik et al., 2008, 2012). In studies on bacteria, an accumulation of thiamine compounds under amino acid starvation and other energy-stress conditions was reported (Gigliobianco et al., 2010).

Only a few reports have suggested that a relationship between thiamine and the cellular responses to stress can also exist in yeast. Machado et al. (1997) proposed a role of thiazole synthase (THI4, encoded by ortholog of plant THI1 gene with confirmed dual role in thiamine biosynthesis and stress response) in the Saccharomyces cerevisiae resistance to DNA damaging agents. THI4 mutants were more susceptible to DNA damage and reactive oxygen species (ROS) attack and showed increased frequencies of colonies with mitochondrial disorders (pétites), and upon heat treatment, an altered THI4 gene expression was also observed (Medina-Silva et al., 2006).

As oxidative stress conditions seem to be most harmful to the cell (Jamieson, 1998), yeasts possess a group of defense systems that provides oxidative protection, including, among others, the yap1- and SKN7-signaling pathways (Beckhouse et al., 2008), antioxidant enzymes (e.g. superoxide dismutase, SOD) and small-molecular-weight antioxidants such as glutathione. The yeast cell protection against adverse environmental conditions includes the overexpression of over 200 different genes (Gasch et al., 2000), and the number of genes for which the specific roles have been resolved have been progressively increasing during the last decade (Lelandais & Devaux, 2010; Lushchak et al., 2010).

It is well known that several serious human diseases are caused by thiamine deficiency or its disturbed transport (Sriram et al., 2012). They are often associated with imbalanced cellular redox state (Jhala & Hazell, 2011). A more detailed knowledge of thiamine activation and utilization by TDP-dependent enzymes under stress conditions in the model eukaryotic organism, S. cerevisiae, can shed a light on the mechanisms of those pathological states, in spite of the inability of human cells to synthesize thiamine de novo.

This work aimed at detecting changes of the activities of thiamine biosynthesis (the expression of THI4,THI6 and THI80 genes and the activities of the encoded enzymes) as well as of a TDP-dependent metabolic route (the activity of transketolase involved in NADPH-generating pentose phosphate pathway) in baker's yeast subjected to oxidative and osmotic stress treatments. To support a hypothesis on a protective role of thiamine in yeast response to harmful factors, mutants defective in thiamine biosynthesis (thi4Δ,thi6Δ) or general stress signaling system (yap1Δ,hog1Δ) were also tested.

Materials and methods


YPD medium was obtained from Difco and Edinburgh Minimal Medium (EMM2) from US Biological. Amino acid supplement, vitamin-free casein hydrolysate, glucose, thiamine, TMP, TDP, NADPH, NADH, nitroblue tetrazolium, HPLC-grade reagents, d-ribose-5-phosphate, α-glycerophosphate dehydrogenase-triosephosphate isomerase and G418 disulfate salt were from Sigma and HPLC-grade acetonitrile from Merck. ATP and PMSF were purchased from Fluka. Reagents used for genetic investigation were obtained from Invitrogen (TRIzol Reagent, dT18 primers and SuperScript III Reverse Transcriptase), Ambion (PureLink RNA Mini Kit and PureLink On-Column DNase Set) and Sigma (SYBR Green® JumpStart™ Taq ReadyMix™).

Yeast strains and growth conditions

Saccharomyces cerevisiae mutant strains lacking different biosynthetic/regulatory genes used in this study are listed in Table 1. All yeast strains were obtained from Euroscarf (Germany). Saccharomyces cerevisiae were grown in YPD medium (10 g L−1 yeast extract, 20 g L−1 peptone, 20 g L−1 dextrose) or in a defined, free of thiamine, EMM2 medium (10.2 g L−1 base, 5 g L−1 glucose, 0.5 g L−1 mineral stock) supplemented with vitamin-free casein hydrolysate (20 g L−1), amino acid mix (containing 20 mg L−1 of each: tryptophan, methionine, leucine and histidine), uracil (20 mg L−1) and vitamin mix (containing 400 μg L−1 of each: pyridoxine, niacin and pantothenic acid, 200 μg L−1 of riboflavin and 2 μg L−1 of biotin). Yeasts were grown at 30 °C with 200 RPM shaking till appropriate growth phase defined for the purpose of this study as early logarithmic (OD600 nm ≈ 0.3), mid-logarithmic (OD600 nm ≈ 0.7), late logarithmic (OD600 nm ≈ 1.2), or early stationary phase (OD600 nm ≥ 2.0) [growth curves tested and compared with (Kennedy et al., 2005)] and then subjected to treatment with stress factors. Turbidity measurements were carried with a UV mini 1240 spectrophotometer (Shimadzu, Kyoto, Japan). All stress treatments lasted 30 min for genetic analyses or one hour for enzyme activity assays. Stress treatment was carried mainly in early logarithmic phase [wild-type (WT) strain and all mutated strains] or in late logarithmic phase (only for WT strain). Oxidative and osmotic stresses were induced every time by treatment with 1 mM hydrogen peroxide and 1 M sorbitol, respectively.

View this table:

Saccharomyces cerevisiae strains used in this study

BY4741 (WT)MATa; his3D1; leu2D0; met15D0; ura3D0
thi4ΔMATa; his3D1; leu2D0; met15D0; ura3D0; YGR144w::kanMX4
thi6ΔMATa; his3D1; leu2D0; met15D0; ura3D0; YPL214c::kanMX4
yapMATa; his3D1; leu2D0; met15D0; ura3D0; YML007w::kanMX4
hog1ΔMATa; his3D1; leu2D0; met15D0; ura3D0; YLR113w::kanMX4

Extract preparation for the enzyme activity determination

Cells from experimental cultures were harvested by centrifugation (5 min, 5000 g ). Pellets were suspended in 0.1 M Tris–HCl buffer pH 8.0, containing protease inhibitor cocktail. Extracts were prepared by Fast-Prep machine using glass beads (425–600 μm; Sigma) with time 45 s and speed 6.0. The precipitated cellular debris was centrifuged for 15 min at 10 000 g at 4 °C. The fresh supernatants were used for the antioxidant enzyme activity determination, while THI6, THI80 and TKL activities were determined in the supernatants after dialysis for 24 h at 4 °C against 0.1 M Tris–HCl buffer pH 8.0 with 1 mM PMSF and 1 mM EDTA with shaking.

RNA isolation and quantitative RT-PCR

Total RNA was isolated from yeast cells grown in 20 mL of broth medium using TRIzol Reagent with PureLink RNA Mini Kit and PureLink On-Column DNase Set and its quality was assessed by separation in agarose gel in denaturing conditions. First strand cDNA was synthesized using 1.5 μg of total RNA, dT18 primers, and SuperScript III Reverse Transcriptase. cDNA was diluted 2× and 2 μL was added to each real-time PCR reaction in a final volume of 15 μL. Amplification and detection were performed with Corbett Rotor-Gene 6000 PCR system using SYBR Green® JumpStart Taq ReadyMix for fluorescent labeling. The applied pairs of gene-specific primers are listed in Table 2. The reaction conditions were 95 °C for 10 min, followed by 40 cycles of 94 °C for 15 s, 59 °C for 15 s, and 72 °C for 20 s. The RDN18 gene was used as a reference in all experiments because it showed the most stable expression under various stress conditions. Appropriate negative controls with RNA or water instead of cDNA were also used. Relative fold changes in expression levels were calculated using the 2−ΔΔCT method (Livak & Schmittgen, 2001). PCR for each sample was performed at least in duplicate.

View this table:

List of primers used in this study


Quantification of thiamine and thiamine phosphate esters

After centrifugation, the yeast pellets were treated with 12% trichloroacetic acid (TCA) and subjected to cell wall disruptions with glass beads using Fast-Prep machine at 4 °C. The precipitated proteins were centrifuged for 15 min at 10 000 g . TCA was removed from the supernatant by diethyl ether extraction, and thiamine compounds levels were quantified by reverse-phase high-pressure liquid chromatography (RP-HPLC) separation, with a post-column derivatization and fluorometric detection (Lee et al., 1991).

Enzyme activity assays

The activity of SOD was determined according to Giannopolitis & Ries (1977). One unit of SOD (U) was defined as the activity able to reduce NBT oxidation by 50%. Activities of THI6 and THI80 were assessed through the determinations of the products formed (TMP and TDP, respectively) by the RP-HPLC method described earlier. For THI6 assay, the reaction mixture contained in 100 μL total volume: 10 mM MgCl2, 20 μM HET-P, 25 μM HMP-PP, dialyzed yeast extract, and 50 mM buffer Tris–HCl pH 7.5. The samples were incubated at 37 °C for 45 min, and the reaction was stopped by mixing for 5 min with 15% metaphosphoric acid and centrifuged. For THI80 assay, the reaction mixture (100 μL volume) contained 10 mM MgCl2, 25 μM thiamine, 20 mM ATP, dialyzed yeast extract, and 50 mM borate-phosphate buffer pH 9.0. The samples were incubated for 60 min and treated with metaphosphoric acid as earlier. The activity of transketolase (TKL) was determined with a modified spectrophotometric method that monitored the decrease in NADH absorbance at 340 nm (Chamberlain et al., 1996).

Protein concentration determination

Protein concentration was measured by the Bradford method (Bradford, 1976).

Statistical analysis

All experiments were repeated 3–5 times to ensure proper analysis of statistical significance (t-test, P < 0.05).


Reference characteristics of thiamine metabolism in S. cerevisiae, depending on the growth phase and the presence of thiamine in the medium

The parameters chosen to characterize thiamine metabolism under stress conditions were first determined in the unstressed yeast cells to be used later as reference values. The variability of the growth conditions included the presence of external thiamine, repeatedly reported to repress thiamine synthesis in all organisms with this biosynthetic ability and the growth phase.

The results obtained in our experiments confirmed strong repression of thiamine metabolism in the presence of thiamine (data not shown), especially in the case of thiamine biosynthetic genes: THI4 (> 1000-fold decrease) and THI6 (35-fold decrease). The results for the latter gene were in a good correlation with the results of TMP synthase activity determinations (50-fold decrease in the thiamine-containing medium). The impact of thiamine on the expression of THI80 and TKL1 genes was barely noticeable while corresponding enzymes showed in both cases the sixfold increase in activity in thiamine-free medium.

The growth phase dependence was studied on yeast grown in thiamine-free medium. We determined the expression of thiamine biosynthetic genes and, where appropriate, the activity of encoded enzymes, as well as the levels of all major thiamine compounds, including free thiamine, TMP and TDP. An example of such the analysis is presented in Fig. 2. The expression levels of THI6 and TKL1 genes were relatively constant over the early- to mid-logarithmic phase but started to decrease at late logarithmic phase to drop to < 10% of initial values at the stationary phase. In contrast, the expression level of THI80 gene was apparently independent from the growth phase. The enzymatic activities of THI80 and TKL, highest at the early log phase, regularly decreased toward the stationary phase while thiamine synthase activity was constant till the late log phase to sharply drop thereafter. The total level of thiamine compounds in yeast cells decreased with culture aging, mainly on the expense of a regular decrease in free thiamine content and, but only at the stationary phase, a drop of TMP level. In contrast, TDP level slightly increased toward the stationary phase.


The gene expression (a) and the activity (b) of TMP synthase (THI6, THI6), thiamine pyrophosphokinase (THI80, THI80) and transketolase (TKL1, TKL) and the contents of thiamine compounds (c) in baker's yeast grown in thiamine-free medium, depending on the growth phase. The specified OD determines the growth phase as follows: 0.3, early logarithmic; 0.7, mid-logarithmic; 1.2, late logarithmic, 2.0, stationary. In each panel, the mean values from two independent triplicate determinations (n = 6) with the standard error bars were plotted. The asterisk indicates a statistically significant difference when compared with the samples from early logarithmic phase – OD600 nm ≈ 0.3 (t-test, P < 0.05). RNA expression was evaluated according to RDN18 gene, which was used as a reference in all experiments.

Taken together, all these results indicated that maximal thiamine biosynthesis occurs at early logarithmic phase. Hence, the analysis of the impact of the stress conditions on thiamine biosynthesis was performed mainly in that growth stage.

Thiamine metabolism under stress conditions

Oxidative stress was induced in yeast cells by treatment with 1 mM hydrogen peroxide and the osmotic one with 1 M sorbitol, after literature suggestions (Gasch et al., 2000) and individual optimization (data not shown). To confirm that the intracellular stress response was actually induced under all these conditions, the gene expression of appropriate stress markers (glutaredoxin, GRX2, and NAD-dependent glycerol-3-phosphate dehydrogenase, GPD1) as well as the activity of an antioxidant enzyme (SOD) were analyzed (Fig. 3). The performed tests confirmed not only appropriate stress state occurrence but also induction of oxidative stress as a consequence of the osmotic one. This part of experiments showed also significantly higher activity of SOD in thiamine (−) medium than that in thiamine (+) medium.


The expression of glutaredoxin (GRX2) and NAD-dependent glycerol-3-phosphate dehydrogenase (GPD1) (a) and the activity of SOD (b) in yeast subjected to oxidative (OX, 1 mM H2O2) and osmotic (OSM, 1 M sorbitol) stress in both thiamine (+) and thiamine (−) media. CON – the unstressed yeast culture as a control. The mean values from two independent triplicate determinations (n = 6) with the standard error bars were plotted. The asterisk indicates a statistically significant difference when compared with the control (nonstressed) samples (t-test, P < 0.05). RNA expression was evaluated relatively to RDN18 gene, which was used as a reference in all experiments.

A further analysis of thiamine metabolism under stress conditions was performed on yeast cultures at the early and late logarithmic phases in the presence [thiamine (+)] or absence [thiamine (−)] of thiamine in the medium. A stress-dependent modulation of the expression of thiamine biosynthetic, activating and utilizing genes at the early logarithmic phase, is illustrated in Fig. 4a,b. Interestingly, more diversified results, especially for THI4 gene, were obtained for thiamine-containing medium (Fig. 4b), where, theoretically, thiamine biosynthesis de novo is not necessary. In thiamine (+) medium, THI4 expression level increased threefold and sevenfold under oxidative and osmotic stress conditions, respectively, while in thiamine (−) medium, the increase was only 1.2-fold at best. The expression of THI6 gene showed a significant (1.5-fold) up-regulation under oxidative stress in thiamine (−) medium and a twofold increase in osmotic stress in thiamine (+) medium. The expression of THI80 was elevated in thiamine (−) medium in both types of stress while in thiamine (+) medium slightly decreased. TKL1 gene was up-regulated only in thiamine (−) medium under oxidative stress conditions (1.4-fold increase). These tendencies were upheld by the enzyme activities (Fig. 4c,d), which showed spectacular increases in the oxidative stress, especially in thiamine (−) medium while, in contrast, under osmotic stress conditions, the thiamine biosynthesis was strongly modulated in thiamine (+) medium.


The expression of genes involved in thiamine biosynthesis and utilization (graphs a and b) and activity of corresponding encoded proteins (c and d) under stress conditions. The yeast were cultured in thiamine-free (left panel – a and c) and thiamine-rich (right panel – b and d) and harvested for the analyses in the early logarithmic phase (OD600 nm = 0.3). Results are presented as a percentage of control values (enzymatic activity) obtained for the corresponding cultures of unstressed yeast and relative mRNA expression, where proper control values are equal 1. All control values were close to those given in Fig. 2. The mean values from two independent triplicate determinations (n = 6) with the standard error bars were plotted. The asterisk indicates a statistically significant difference when compared with the nonstressed samples (t-test, P < 0.05).

At the late logarithmic phase of yeast growth (Fig. 5), a significant stress-dependent up-regulation of thiamine metabolism in thiamine (+) medium was found, especially under osmotic stress conditions (twofold, 3.4–fold, and twofold for THI6,THI80 and TKL, respectively) while in thiamine (−) medium, no significant activation could be observed.


Activities of TMP synthase, thiamine pyrophosphokinase and transketolase in yeast subjected to oxidative (OX) and osmotic (OSM) stress in the late logarithmic phase of growth in thiamine (−) (a) or thiamine (+) medium (b). The results are presented as percentage the values obtained for control cultures of unstressed yeast. The mean values from two independent triplicate determinations (n = 6) with the standard error bars were plotted. The asterisk indicates a statistically significant difference when compared with nonstressed samples (t-test, P < 0.05).

Thiamine metabolism in mutant strains under stress conditions

To further elucidate the effects of stress on thiamine/TDP biosynthesis and TDP-dependent metabolism, mutant strains defective in thiamine biosynthesis (thi4∆,thi6∆) and in the genes involved in the global control of stress responses in yeast (yap1Δ,hog1∆) were used as models. All of them were exposed to the stress factors at early-logarithmic phase, to obtain results comparable with WT strain. A comparison of the activities of thiamine/TDP synthesizing enzymes, THI6 and THI80 between the WT strains and thi6∆ and thi4∆ mutants, could only be tested in thiamine (+) medium where these strains could survive, taking up the external thiamine. As shown in Fig. 6, both biosynthetic mutants showed a significant increase in SOD activity as compared to WT strains while in regulatory mutants, a decreased activity of this antioxidant enzyme was rather observed. The THI6 activity seemed to be elevated in the regulatory mutants but ceased in the thi4∆ mutant. All tested mutants but thi6∆ presented a significant (1.5-fold) increase in THI80 activity. The regulatory mutants under control conditions (no stress factors applied) showed a nearly half reduced free thiamine level while TDP content increased, especially in hog1∆ (1.6-fold, data not shown).


A comparison of the activities of SOD, THI6 and THI80 in the WT yeast strains and (a) mutants defective in thiamine biosynthesis (thi4Δ and thi6Δ) grown in thiamine (+) medium, and (b) yeast strains with mutations in the genes of global control of stress reaction in yeast (yap1Δ and hog1Δ) grown in thiamine (−) medium measured in early logarithmic phase (OD600 nm = 0.3). The mean values from two independent triplicate determinations (n = 6) with the standard error bars were plotted. The asterisk indicates a statistically significant difference when compared with the control samples of wild-type strain (t-test, P < 0.05).

The results from similar analyses of thiamine/TDP biosynthesis for mutants under stress conditions are presented in Fig. 7. In the thi4Δ mutant, TMP synthase (THI6) activity was twofold elevated under both oxidative and osmotic stresses and THI80 activity also increased threefold and twofold, respectively.


Activities of thiamine biosynthetic enzymes in THI4 and THI6 mutated strains cultured in thiamine (+) medium under oxidative and osmotic stress conditions introduced in early logarithmic phase (OD600 nm = 0.3). The mean values from two independent triplicate determinations (n = 6) with the standard error bars were plotted (t-test, P < 0.05). The asterisk indicates a statistically significant difference when compared with the nonstressed samples of each strain (t-test, P < 0.05).

In cellular defense-defective mutants, the yap1- and HOG1- disrupted strains, thiamine biosynthesis decreased as a result of harmful treatments, especially strongly in the latter mutant under osmotic stress conditions (Fig. 8). TDP synthesis level was elevated under oxidative stress in yap1Δ (> 2.5-fold) and under osmotic stress in hog1Δ (c. threefold).


Changes of the activities of thiamine/TDP biosynthetic enzymes (a) and the intracellular levels of thiamine compounds (b) in the cellular defense regulation mutants yap1 and HOG1 under oxidative (OX) and osmotic (OSM) stress conditions studied for early logarithmic phase (OD600 nm = 0.3). The yeasts were cultured in thiamine (−) medium. TMP level remained at a very low and did not change during stress treatment; it was thus not presented on graph b. Results are presented as % of control sample value, which are presented on Fig. 6. The mean values from two independent triplicate determinations (n = 6) with the standard error bars were plotted (t-test, P < 0.05). The asterisk indicates a statistically significant difference when compared with the nonstressed samples of each strain (t-test, P < 0.05).

The intracellular levels of thiamine compounds also changed in regulatory mutants under the influence of stress factors. In the case of yap1Δ, the results were comparable in both types of stress applied and were characterized by a slightly increased thiamine level (1.5-fold) and almost constant TDP content. In contrast, the hog1Δ mutant presented more variable response, such as an almost unchanged thiamine/TDP levels under oxidative stress conditions and a significant elevation of total thiamine pool, especially of free thiamine, under osmotic stress.


It is well known that all living cells adapt their growth and division rate to environmental conditions. Only recently has a hypothesis of an essential role of thiamine in the stress responses started to emerge, mainly from studies on plants and bacteria. In this work, for the first time, we provide data on the modulation of thiamine biosynthesis and TDP-dependent pathways in baker's yeast subjected to the oxidative and osmotic stress treatments.

Reference characteristics of thiamine biosynthesis in baker's yeast

At the beginning, we determined reference values of parameters intended to be used in the stress study and their variability with the phase of yeast culture growth as well as a presence of external thiamine. We confirmed that THI6 gene/TMP synthase were expressed/active in thiamine (+) medium indicating that reported repression of thiamine biosynthesis by exogenous thiamine (Nosaka et al., 2005) is by no means absolute.

We took under consideration also the growth phase. The THI6 expression and corresponding enzyme activity were strongly correlated with maintain the TMP pool on almost constant level during yeast culture growth until the stationary phase when they suddenly dropped. A clear correlation between the gene expression level and the activity of encoded protein product was also observed in the case of transketolase, whose level decreased to a very low value at the latest growth phase studied. It was previously reported in the literature that a decrease in activities of TDP-dependent enzymes such as pyruvate dehydrogenase, α-ketoglutarate dehydrogenase, and just transketolase caused a deceleration of growth limiting biosynthetic processes in Candida lipolytica (Ermakova et al., 1979) and Salmonella enterica serovar Typhimurium (Frodyma et al., 2000). We hereby confirmed this phenomenon in S. cerevisiae, which can be explained by taking into consideration the carbohydrate metabolism (Galdieri et al., 2010). Saccharomyces cerevisiae is a facultative anaerobe that executes fermentation in 0.5% or more glucose content, and after glucose depletion switches to the use of ethanol as a energy source, entering the same stationary phase (Marks et al., 2008). TDP is crucial for both pathways following the glycolysis because pyruvate decarboxylase (anaerobic) as well as pyruvate dehydrogenase (aerobic) is TDP-dependent enzymes. Therefore, we suspected that the activity of thiamine pyrophosphokinase would be modulated during yeast cultivation in spite of a known constitutive expression of its gene (Voskoboev et al., 1978). In our experiments, we confirmed an almost constant THI80 gene expression, but the activity of encoded enzyme, similarly to that of a TDP-dependent enzyme, transketolase, gradually felt with yeast culture aging. This finding strongly suggested that some so far unknown post-transcriptional mechanisms must regulate the activity of this enzyme in yeast. It is quite possible that it might be controlled through a regulatory mechanism somehow similar to that of pyruvate decarboxylase (Nosaka et al., 2005), as thiamine significantly influenced THI80 expression.

Thiamine metabolism in S. cerevisiae under stress conditions

Baker's yeast strains differ in sensitivity to numerous oxidants; hence, we chose hydrogen peroxide because it seems to be most universal oxidative stress inducer, and moreover, it is well known to activate the yap1-dependent signaling pathway (Lushchak, 2010). We tested H2O2 concentration from 0.5 to 5.0 mM (data not shown) to choose 1 mM finally. As an osmotic stress factor, we used a solute commonly used in similar studies, sorbitol at high 1 M concentration (Westfall et al., 2008). To confirm the actual intracellular stress state, we analyzed the expression of commonly accepted genetic markers of oxidative and osmotic stress such as GRX2 and GPD1, respectively. We also checked the expected up-regulation of a representative antioxidant enzyme, SOD which is universally essential to oxidative stress protection, because the sod∆ mutants are hypersensitive to oxygen action (Lushchak et al., 2005). The biological role of SOD is much broader, including, among others, responsibility for internal homeostasis, protective function in relation to catalase and participation in hyperosmosis protection (Garay-Arroyo et al., 2003; Lushchak et al., 2005). Our tests repeatedly showed significant increases of SOD activity in yeast under the influence of hydrogen peroxide, but also under osmotic stress treatment, clearly confirming that the oxidative stress co-occurs with the osmotic one, a phenomenon so far supported by a limited number of reports for S. cerevisiae (Pastor et al., 2009). The most interesting finding for SOD was that its activity in thiamine (−) medium was significantly (almost twofold) higher than in thiamine (+) medium. The hypothesis that a lack of thiamine biosynthesis might cause mobilization of antioxidant defense is also supported by a report of an interaction between thiamine biosynthetic regulatory gene THI2 and SOD1 gene (Zheng et al., 2010) and a postulated direct cell-protecting antioxidant property of thiamine molecule based on its oxidation to thiochrome (Lukienko et al., 2000; Jung & Kim, 2003; Zakrzewska et al., 2011).

Of four genes analyzed for the expression level, three could also be tested for enzymatic activities. We conducted our research on yeast cultures grown in thiamine (−) as well as in thiamine (+) medium not to overlook the changes that could occur when the de novo thiamine biosynthesis was not necessary. As it is known that under adverse environmental conditions, yeast cell economizes its highly energy-consuming processes, thiamine biosynthesis activation in thiamine (+) medium could provide evidence that the biological role of thiamine might be extended on the cell protection. In thiamine (+) medium, the expression of THI4 gene, previously postulated to be involved in some stress responses in yeast (Machado et al., 1997; Medina-Silva et al., 2006), increased under stress treatment, especially under osmotic stress conditions when also the THI6 gene expression was up-regulated. A relationship between thiamine content and oxidative and osmotic stress was confirmed by the determinations of all tested enzyme activities that showed up to eightfold activation of TKL, while in the case of oxidative stress in both medium types, there were equally strong increases (4–5-fold) in thiamine metabolic enzyme activities. This strong correlation of thiamine biosynthesis with osmotic stressor treatment might provide an interpretation for a previously reported interaction between THI2 regulatory gene and SAT4 (HAL4) gene of Ser/Thr protein kinase involved in the salt/osmotic tolerance (Bandyopadhyay et al., 2010). In our studies, we also looked for some specific effects of 1 M NaCl but did not note essential differences with those observed during the sorbitol treatment, in agreement with the literature (Hirasawa et al., 2006); therefore, our corresponding results are not presented here. The explanation of why in thiamine (+) medium the activation of thiamine biosynthesis was so evident might be similar to the results obtained for sunflower where the level of soluble sugars synthesized on TDP-dependent carbohydrate metabolic pathways was significantly higher in thiamine-supplemented plants, strengthening their resistance to osmotic stress conditions (Sayed & Gadallah, 2002).

In thiamine (−) medium, the most significant results were obtained for the early logarithmic phase, especially under oxidative stress conditions. This might be interpreted as indicating a need for intensification of cellular metabolic processes including TDP-dependent pathways and possibly taking advantage of protective role of thiamine to survive and maintain growth processes that are considerably impaired under oxidative stress (Zakrzewska et al., 2011). Such a lack of visible regulation of studied parameters in thiamine (−) medium as compared with thiamine (+) medium may result from the elevated activity of antioxidant defense mechanisms in the absence of thiamine (confirmed by SOD activity level). These findings suggest that thiamine biosynthesis in thiamine (−) medium is from its early stages connected with the environmental stress response.

The aim of this part of the study was also to compare thiamine metabolism level at initial dynamic growth and when yeast entered the stationary phase, to investigate the real thiamine role in the cell protection, as it is also known that cells modulate their antioxidant system functioning depending on the growth phase (Drakulic et al., 2005). At the late logarithmic phase of yeast growth, the activation of thiamine biosynthesis was also recorded in thiamine (+) medium, confirming that under stress conditions, it is essential for the cell to maintain a sufficient pool of thiamine and its derivatives despite the depletion of nutrients in the culture medium.

A strong acceleration of thiamine biosynthesis and activation, and in some cases, a dramatic increase in transketolase activity is consistent with the involvement of this TDP-dependent enzyme in the yeast stress response via pentose phosphate pathway (PPP). It was previously reported that under stress conditions, yeasts redirect carbohydrate flux from glycolysis to PPP to alter the redox equilibrium (Ralser et al., 2007). There are also reports of an overlapping role of SOD and PPP under oxidative stress conditions (Slekar et al., 1996) not restricted to glucose-6-phosphate dehydrogenase, which catalyzes irreversible oxidative branch of that pathway, but also dependent on enzymes engaged in nonoxidative branch of this cycle (transketolase and transaldolase), which regenerate substrates for NADPH-producing reaction. It was also confirmed by a report that TKL1 mutants are hypersensitive to H2O2 treatment (Juhnke et al., 1996). The above-mentioned examples showed mainly TKL1 gene importance because of unfavorable conditions while our results confirm an important metabolic role of transketolase protein connected with the elevated TDP request.

The main way for cells to cope with osmotic stress factor treatment is an accumulation of compatible solutes. An interaction between TKL2 and trehalose synthase genes (Szappanos et al., 2011) was reported while it has been well known that this sugar plays an important role in the cell protection against the osmotic stress (Hounsa et al., 1998). However, the main role in the osmoprotection is assigned to glycerol that is synthesized from dihydroxyacetone. Sorbitol-induced stress state activates sorbitol dehydrogenase to produce d-fructose that enters glycolysis pathway and is a source of dihydroxyacetone. Accelerated glycolysis is also conditioned by elevated PPP rate while products of TKL-catalyzed reaction are also fructose-6-phosphate and glyceraldehyd-3-phosphate, the latter undergoing an isomerization to dihydroxyacetone. Moreover, glycerol inhibits fermentation (Converti et al., 1995), forcing PPP to produce more NADPH that is required for aerobic growth. It was also reported that ROS inhibit fermentation and promote respiration (Allen et al., 2010) and that the fermenting cells exhibit a lower viability than aerobically respiring cells under adverse environmental conditions (Cabiscol et al., 2000) whereas our studies of SOD activity under both types of stress conditions confirmed that ROS level was indeed elevated.

The above-mentioned explanation involving the switch from fermentation to respiration by yeast treated with stress factors for sure does not cover all needs for THI80/THI80 activation because there are also some indirect interactions found for THI80 gene with genes engaged in oxidative and osmotic stress signaling pathways (Fiedler et al., 2009; Szappanos et al., 2011). These data may suggest an existence of a very complicated and based on extremely subtle interactions cellular defense system.

Thiamine biosynthetic mutants under stress conditions

We chose two strains with thiamine biosynthesis disruption: thi4Δ that was previously reported to have an altered stress protection, and thi6Δ, with knock-out in the major biosynthesis gene encoding the bifunctional enzyme TMP synthase/HET kinase. We focused on the enzyme activity determinations to see the final effects of mutations on the de novo thiamine/TDP biosynthesis. Both mutants showed increase in SOD activity under control conditions, even though they were cultured in thiamine-containing medium, where these mutations were not expected to exert a big impact on cellular metabolism but in fact they did, affecting also growth rate of thi6Δ (decreased, data not shown). It looks like this pivotal enzyme responsible for thiamine-characteristic two-ring structure formation is also important for general cellular functioning.

The analysis of the stress effect in the thi4Δ mutant confirmed data from thiamine (+) stress experiments on the WT yeast where THI4 gene expression was strongly induced. Mutated strain exhibited a great activation of thiamine metabolism under oxidative stress. Osmotic stress conditions also activated thiamine biosynthesis significantly. In the case of thi6Δ mutant, the activation of thiamine/TDP biosynthesis was barely visible and in the osmotic stress even decreased, confirming that this strain is metabolically inefficient under stress conditions, partly because the decrease in TDP level results in a decrease in PPP rate and hence lower supply of NADPH required by respiratory growth, whereas stress conditions block fermentative pathway. Results obtained with THI4 mutant agreed with the literature reports of their elevated susceptibility to stress treatment (Machado et al., 1997; Medina-Silva et al., 2006). For THI6 mutants, there was no reference study of their behavior under unfavorable conditions, but putative interactions of THI6 gene with processes such as the ubiquitination or transcription modulation (Collins et al., 2007; Szappanos et al., 2011) can be taken into account in the data interpretation.

Defense response mutants under stress conditions

For this part of our studies, we chose two mutated strains, yap1Δ and hog1Δ, which are extremely susceptible to oxidative and osmotic stress disruptions, respectively. We first tested the basal thiamine metabolism in the yap1Δ and hog1Δ mutants and found that the activities of TMP synthase and thiamine pyrophosphokinase were significantly elevated. In contrast, the control determination of SOD activity in hog1Δ did not differ from that in the WT yeast and in yap1Δ was clearly decreased. It seems that thiamine biosynthesis activation may be the way of compensation of the stress response disruption. Such a mechanism, based on thiamine biosynthesis stimulation, was previously reported for a bacterial system with inactivated DNA repair genes (Fukui et al., 2011).

Thiamine compound levels also differed between the WT strain and the yap1/HOG1 mutants, especially the latter, having a decreased thiamine pool and accumulating TDP. This finding can be explained by taking into consideration the regulatory role of TDP for thiamine/TDP biosynthesis in yeast. These regulatory mechanisms are not well understood in S. cerevisiae but are unlikely to involve TDP-dependent riboswitches, abundant in plants and bacteria but not yet detected in yeast. However, a lack of thiamine induces SNO and SNZ genes that physically interact with THI5 family genes (Rodriguez-Navarro et al., 2002), and the latter show a marked similarity to bacterial thiY gene, coding for a HMP-binding protein regulated by riboswitches (Bale et al., 2010). Moreover, all four members of the THI5 family (THI5/11/12/13) showed differed response to nutrient limitation (Wightman & Meacock, 2003).

Data on the stress effects in the yap1 and HOG1 mutants confirmed our assumption that thiamine metabolism should be elevated in the yeast strains with disrupted cellular defense system. Moreover, we observed a close correlation between type of mutation and the response to stress treatment. Thus, yap1Δ activated THI80 mainly under H2O2 treatment while hog1Δ under sorbitol/NaCl conditions. We also showed that under osmotic stress, yeast accumulate free thiamine, in agreement with the above-described hypothesis of protective role of thiamine.

There is also some relationship between the tested stress response in mutants and PPP. We demonstrated a concerted elevation of THI80 and TKL activities, suggesting PPP stimulation. This hypothesis has an additional support from literature where some interaction between TKL1 and HOG1 genes was reported (Costanzo et al., 2010).


Data obtained in this work strongly suggest that thiamine biosynthesis ability and defense responses in yeast markedly overlap as judged from the observation that thiamine biosynthetic mutants have significantly elevated SOD activity while defense regulatory mutants enhance thiamine synthesis and activation. A connection between thiamine biosynthesis, pentose phosphate pathway, and the resistance to different disadvantageous conditions was hereby confirmed. Several of observed effects reflect a switch of yeast metabolism from fermentation to respiratory growth under the stress conditions. Further investigations are necessary to fully appreciate a global role of thiamine, small multifunctional molecule that is able to affect not only protein but even nucleic acid structure to control its own biosynthesis and also other important metabolic processes.


This work was supported in parts by the Ministry of Science and Higher Education/National Science Center, Poland (grant No. N N303 809540) and the Jagiellonian University (statutory funds DS/37/2011/WBBiB) (both grants to E.K.).


  • Editor: Jens Nielsen


View Abstract