Journal of Bioscience and Bioengineering VOL. xx No. xx, 1e9, 2015 www.elsevier.com/locate/jbiosc

Metabolomic analysis of acid stress response in Saccharomyces cerevisiae Riyanto Heru Nugroho, Katsunori Yoshikawa, and Hiroshi Shimizu* Department of Bioinformatic Engineering, Graduate School of Information Science and Technology, Osaka University, 1-5 Yamadaoka, Suita, Osaka 565-0871, Japan Received 6 November 2014; accepted 19 February 2015 Available online xxx

Acid stress has been reported to inhibit cell growth and decrease productivity during bio-production processes. In this study, a metabolomics approach was conducted to understand the effect of lactic acid induced stress on metabolite pools in Saccharomyces cerevisiae. Cells were cultured with lactic acid as the acidulant, with or without initial pH control, i.e., at pH 6 or pH 2.5, respectively. Under conditions of low pH, lactic acid led to a decrease in the intracellular pH and specific growth rate; however, these parameters remained unaltered in the cultures with pH control. Capillary electrophoresis-mass spectrometry followed by a statistical principal component analysis was used to identify the metabolites and measure the increased concentrations of ATP, glutathione and proline during severe acid stress. Addition of proline to the acidified cultures improved the specific growth rates. We hypothesized that addition of proline protected the cells from acid stress by combating acid-induced oxidative stress. Lactic acid diffusion into the cell resulted in intracellular acidification, which elicited an oxidative stress response and resulted in increased glutathione levels. Ó 2015, The Society for Biotechnology, Japan. All rights reserved. [Key words: Metabolome; Acid stress; Saccharomyces cerevisiae; Proline; Lactic acid]

The use of bio-based substances has gained wide interest in industrial applications such as biodiesel, bio-gasoline, and bioplastic production (1,2). Yeast strains such as Saccharomyces cerevisiae, are commonly used for bioproduction owing to their natural tolerance to stressors like ethanol (3,4), salinity and osmotic pressure (5), heat (6,7), and weak acids (8,9). During bioproduction process, yeast cells have been exposed to various stress conditions such as acid stress (10,11), varying osmotic pressures or ethanol (12), exposure to furan aldehydes or aliphatic acids (13,14), and freezing (15). These stress conditions decreased cell growth and productivity. Acid stress is one of the critical problems encountered during yeast mediated bioproduction processes, e.g., lignocelluloic (or derivative) ethanol production (16e19) and lactic acid production (20,21). Acid stress may occur when the extracellular pH is lower than the pKa of acid (logarithmic constant of acid dissociation). If the pH is higher than the pKa, the dissociated form of acid outside the cells is abundant and this is relatively harmless compared to its undissociated form. When the pH is lower than the pKa, undissociated form of acid is existing in large amounts, which can diffuse more freely into the cell than dissociated acids, which needs specific transport due to selective permeability of the cell membrane. Within the cell, dissociation of acid results in accumulation of protons and can cause intracellular acidification. This acidification may result in intracellular pH drop (11,22), inhibition of cell

* Corresponding author: Tel./fax: þ81 6 6879 7446. E-mail addresses: [email protected] (R.H. Nugroho), yoshikawa@ ist.osaka-u.ac.jp (K. Yoshikawa), [email protected] (H. Shimizu).

growth and product formation (11,18) and thereby affect cellular metabolism. Moreover, acid stress also leads to the accumulation of anions within the cells (23), and this can lead to the degradation of vacuoles (24,25). In practice, the productivity of ethanol production was decreased at a pH equal to or lower than the pKa of acetic acid (11). Several studies have been performed for a better understanding of the acid stress mechanism and improvement of acid stress tolerance of cells. Hasunuma et al. showed that ethanol production using xylose was improved under acetic acid or formic acid stress conditions by overexpression of TAL1 in yeast (26). Abbott et al. (20) found that overexpression of CTT1 encoding cytosolic catalase increased specific growth rate under acetic acid treatment at pH 3 in yeast.Furthermore, adaptive response to propionic acid study in S. cerevisiae suggested that RIM101 expression is necessary in counteracting propionic acidinduced cytosolic acidification (27). Genome wide-screening studies have reported gene sensitivity towards acid stress (9,25,28). However, the knowledge needed for further improvement of acid stress tolerance is limited and warrants further investigations. Metabolomic analysis is a powerful tool to observe comprehensive intracellular metabolite concentrations. Metabolite concentrations are the results of complicated cellular mechanisms including gene regulations and translational regulations; therefore, metabolome data reflect the metabolic state of the cell better than transcriptomic or proteomic data. Metabolomic analysis has been used to identify the key metabolites and metabolic reactions required for stress tolerance, and to improve the stress tolerance (16,26,29e32).

1389-1723/$ e see front matter Ó 2015, The Society for Biotechnology, Japan. All rights reserved. http://dx.doi.org/10.1016/j.jbiosc.2015.02.011

Please cite this article in press as: Nugroho, R. H., et al., Metabolomic analysis of acid stress response in Saccharomyces cerevisiae, J. Biosci. Bioeng., (2015), http://dx.doi.org/10.1016/j.jbiosc.2015.02.011

2

NUGROHO ET AL.

Here, we studied the metabolic responses of S. cerevisiae to acid stress in an effort to improve its acid stress tolerance. Lactic acid was used as an acidulant mainly because lactic acid stress constitutes one of the major stress conditions during its production in yeast (20,21). In addition, lactic acid is a prospective monomer for polylactic acid (PLA) which has growing interest and demand in bioplastic bioindustries (33,34) for replacing plastic produced from fossil source in order to reduce global environmental waste problem. During lactic acid production, a decrease in the pH of yeast cells has been reported to cause stress and thereby lower productivity (21). Here, metabolome analysis was performed using capillary electrophoresis time-of-flight mass spectrometry (CE-TOFMS) in the presence of varying concentrations of lactic acid, with and without pH control. Several metabolites responding to acid stress were identified and the effect of these on acid stress tolerance was studied. MATERIALS AND METHODS Strain and growth conditions S. cerevisiae strain BY4739 (MATa leu2D0 lys2D0 ura3D0, Open Biosystems, Huntsville, AL, USA) was used in this study. Cells were cultured at 30  C in synthetically defined (SD) minimal media containing 2% glucose, 0.67% yeast nitrogen base without amino acids (Difco Lab., Detroit, MI, USA) and amino acids; 0.0076% L-leucine, 0.038% L-lysine monohydrochloride and 0.0076% uracil (Wako Pure Chemical Industries, Ltd., Osaka, Japan). Inoculum was grown overnight in 5 mL SD media on a bioshaker (BR-21FH, Taitec Corp., Saitama, Japan) at 150 strokes per min and inoculated into Sakaguchi flask containing 100 mL SD media at initial optical density of 660 nm (OD660) z 0.1 at 120 strokes per min (MM10, Taitec Corp., Saitama, Japan). L-Lactic acid (either 10 or 14 g/L, Wako Pure Chemical Industries, Ltd.) were added in experiments used to monitor acid stress. Two pH conditions were set for those cultures where lactic acid was added viz. uncontrolled pH and pH 6 initially controlled using 2 M sodium hydroxide. 0.6 g/L of proline (Nacalai Tesque, Kyoto, Japan) was added to the SD medium for investigation of the effect of proline. The values of OD660 were converted to dry-cell weight (DCW) of yeast cells using the following relation: OD660 of 1: 0.22 g DCW. Samples used for metabolite profiling were harvested at the exponential phase at OD660 z 0.5. Sample analyses Cell concentration was measured by OD660 using a UV-VIS spectrophotometer (UVmini-1240, Shimadzu Corp., Kyoto, Japan). Glucose and Llactate concentrations were measured using an enzyme electrode sensor (Biosensor BF-5, OJI Scientific Instruments, Hyogo, Japan). Ethanol was analyzed by gas chromatography (7890 GC System, Agilent Technologies, Palo Alto, CA, USA) equipped with Stabilwax column dimension of 0.32 mm ID, 60 m length, 1 mm thickness (Restek Co., Bellefonte, PA, USA) using helium gas as the carrier and 3-methyl-1-butanol (0.1 %v/v) as the internal standard. Glycerol was analyzed by F-kit (Roche Diagnostics, Mannheim, Germany). Concentration of proline in the culture broth was analyzed using ultra performance liquid chromatography (UPLC, Acquity UPLC System, Waters, Massachusetts, USA) and Waters Empower chromatography software. The UPLC components consisted of ACQUITY Column Heater (equipped by ACQ-Tag Ultra C18 1.7 mm, 2.1  100 mm column and guard column of Acquity UPLC HSS T3 1.8 mm, 201  5 mm Van-guard pre-column), UPLC Auto Sampler Manager and UPLC Binary Solvent Manager. The mobile phase used acetonitrile solutions (Eluent A and Eluent B, Waters, MA, USA). Amino acid standard was purchased from Thermo Fisher Scientific (IL, USA) and 250 mM norvaline was used as the internal standard. Derivatized solution was prepared using Waters AccQ-Tag Ultra Derivatization Kit. Reagent in 2A (diluent, 100% acetonitrile) was poured into the 2B encoded bottle containing reagent powder and was vortex-mixed for 10 s. The solution mixture was heated for 10 min at 55  C. The norvaline derivatizating solutions and Waters AccQ-Tag Ultra Borate Buffer were mixed and the solution was heated for 10 min at 55 C and measured by UPLC. Metabolome analysis Intracellular metabolite concentrations were measured using CE-TOFMS. Detailed methods are described in a previous report (35). Samples were harvested at OD660 z 0.5 and were filtered using a 0.4 mm filter (Millipore Isopore, MA, USA). Ultra-sonication was performed after the membrane was immersed in an internal standard containing methanol solution (H3304-1002, Human Metabolome Technologies, Yamagata, Japan). Chloroform and water were added to the solution after removal of the membrane and the aqueous portion was collected after extraction using a vortex mixer which was dried using a SpeedVac (Thermo Fisher Scientific) before the addition of the second internal standard (H3304-1004, Human Metabolome Technologies). Samples were analyzed using CE-TOFMS (Agilent 7100 CE system with Agilent 6224 TOF-MS, Agilent Technologies). Intracellular pH measurement Intracellular pH was measured using the dye labeling method as previously reported (36e39). The dye i.e., fluorescence-agent used in this measurement was 5-(and-6-)-carboxyfluorescein diacetate

J. BIOSCI. BIOENG., succinimidyl ester (CFDA-SE; CFSE mixed isomers, Life Technologies Corp., NY, USA) dissolved in anhydrous dimethyl sulfoxide (DMSO, Sigma Aldrich Corp., MO, USA). The dye is permeable into yeast cells and upon permeation into the cell, the CFDASE is expected to react with cellular esterase and form CFSE, which possesses low membrane permeability. CFSE is reactive toward amino acids and retains dye form (40). Intracellular pH calibration curve was made by growing cells on SD medium until OD660 z 0.5 was reached. Cells were centrifuged and phosphate buffered saline (PBS) solutions (Gibco, Life Technologies Corp.) of varying pH viz., 2.76, 3.27, 3.65, 3.97, 4.10, 5.10, 6.06, 7.20, 8.33, and 9.70 were added to each sample followed by loading with a dye of around 10 mM (final concentration). This was followed by addition of valinomycin (1 mM of final concentration, Sigma Aldrich Corp.) and nigericin (1 mM of final concentration, sodium salt form, Sigma Aldrich Corp.). Samples were mixed by vortexing for 1 min and washed twice, using PBS possessing the same pH. After this, the contents of the sample vials were moved into 96 well plates (costar, Corning Incorporation, NY, USA) and excitation spectra of 485 and 450 at emission wavelength of 535 nm were measured using a multilabel counter (1420 Arvo MX, Perkin Elmer, MA, USA). For intracellular pH measurement of the samples, the dye was loaded into the samples, which were then quick-washed twice using PBS (pH 7.2) to remove non-diffused dye and the excitations were measured at the above mentioned wavelengths. pH values were obtained by extrapolating the measured excitation ratios of 485 and 450 nm from the calibration curve. Statistical test Significant difference in parameters of growth rates and metabolite concentrations between the two conditions was calculated by two-tailed t-test using MS Excel 2010. Metabolome data were normalized and analyzed by principal component analysis (PCA) using R function “prcomp”.

RESULTS Effects of acid stress on yeast cultures Acid-stress conditions were created by addition of either 10 or 14 g/L of lactic acid to the culture medium because these concentrations clearly decreased the yeast growth (data not shown). Because pH plays an important role in the dissociation of weak acids, two pH conditions were selected for each concentration of lactic acid. The first condition had no a pH control (i.e., approximately pH 2.5). Because the pKa of lactic acid is 3.86, lactic acid is retained predominantly in the un-dissociated form in the culture condition. The pH was initially controlled approximately at 6 in the second condition, where lactic acid is mostly in the dissociated form. Five culture experiments were performed as outlined in Table 1. Cultures to which lactic acid was not added, and in which pH was not controlled, were used as controls. Culture conditions to which 10 or 14 g/L of lactic acid was added, but pH was not controlled are indicated as La10 and La14, respectively. Those conditions where 10 or 14 g/L of lactic acid was added and pH were initially controlled (pH 6) using NaOH are indicated as La10 (pH 6) and La14 (pH 6), respectively. These culture conditions were classified according to the initial pH of the culture as: (i) below the pKa of lactic acid (La10 and La14), and (ii) above pKa of lactic acid [control, La10 (pH 6) and La14 (pH 6)] (Table 1). The culture profiles are summarized in Fig. 1 and Table 1. In cultures with pH control [i.e., La10 (pH 6) and La14 (pH 6)], the cell growth, glucose consumption, and rates of glycerol and ethanol formation were close to those in control cultures (Table 1). Under low pH conditions (i.e., La10 and La14), an increase in lactic acid concentrations led to a decrease in specific cell growth rate from 0.260/h in control to 0.160/h in La10, and further to 0.099/h in La14. Other parameters such as the specific glucose consumption rate, specific ethanol production rate, and specific glycerol production rate were not significantly affected (Table 1). Lactic acid concentrations were not altered from the corresponding concentrations during culture experiments, suggesting that lactic acid was not consumed by yeast cells (data not shown). Both extracellular and intracellular pH at mid-log phase (OD660 z 0.5) were measured (Table 1). Extracellular pH in all conditions was similar to the initial observed pH. Extracellular pH of La10 and La14 were 2.50 and 2.42, respectively, which was lower than the pKa of lactic acid. Similar to the extracellular pH, La10 and La14 showed low intracellular pH of 3.86 and 3.48, respectively,

Please cite this article in press as: Nugroho, R. H., et al., Metabolomic analysis of acid stress response in Saccharomyces cerevisiae, J. Biosci. Bioeng., (2015), http://dx.doi.org/10.1016/j.jbiosc.2015.02.011

VOL. xx, 2015

METABOLOMICS STUDY OF ACID STRESS IN YEAST

3

TABLE 1. Summary of the cultures used in this studya. Culture code

Initial culture pH

Culture pH at OD660 z 0.5

Intracellular pH at OD660 z 0.5

Specific growth rate (/h)

Specific glucose uptake rate (g/g DCW-h)

Specific ethanol formation rate (g/g DCW-h)

Specific glycerol formation rate (g/g DCW-h)

Description

Control

4.80  0.20

4.04  0.02

4.28  0.20

0.260  0.001

0.046  0.006

0.019  0.003

0.005  0.000

La10

2.54  0.03

2.50  0.06

3.86  0.57

0.160  0.031

0.044  0.005

0.017  0.005

0.006  0.002

La14

2.42  0.05

2.42  0.01

3.48  0.24

0.099  0.004

0.042  0.008

0.018  0.003

0.007  0.001

La10 (pH 6)

5.87  0.62

5.53  0.48

6.38  0.16

0.263  0.008

0.051  0.000

0.018  0.002

0.005  0.000

La14 (pH 6)

5.62  0.36

5.55  0.43

6.28  0.16

0.266  0.012

0.053  0.004

0.024  0.004

0.005  0.000

PLa0

5.27  0.03

3.81  0.05

5.82  0.55

0.267  0.015

0.039  0.004

0.019  0.000

0.003  0.000

PLa10

2.54  0.01

2.54  0.02

5.77  0.49

0.187  0.004

0.038  0.001

0.018  0.003

0.004  0.000

PLa14

2.45  0.01

2.44  0.00

4.70  0.48

0.121  0.009

0.029  0.005

0.020  0.003

0.004  0.001

No lactic acid addition No pH control 10 g/L lactic acid addition No pH control 14 g/L lactic acid addition No pH control 10 g/L lactic acid addition Initial pH control at 6 14 g/L lactic acid addition Initial pH control at 6 No lactic acid addition No pH control Proline addition of 0.6 g/L 10 g/L lactic acid addition No pH control Proline addition of 0.6 g/L 14 g/L lactic acid addition No pH control Proline addition of 0.6 g/L

a Evaluation times for specific rates were during logarithmic growth state; approximately between 0 and 12 h for control, PLa0, La10 (pH 6) and La14 (pH 6); 12e36 h for La10 and PLa10; 24e48 h for La14 and PLa14. Errors indicate the standard deviation of biological triplicate experiments.

0

0 20 40 60 Time (h)

4

5

2

0

0

0 20 40 60 Time (h)

D) La10 (pH 6) Glucose (g/L)

20

10 8

15

6

10

4

5

2

0

0 20 40 60 Time (h)

0

10 8

15

6

10

4

5

2

0

0 20 40 60 Time (h)

E) La14 (pH 6) 20

0

10 8

15

6

10

4

5

2

0

0 20 40 60 Time (h)

0

OD660, Ethanol (g/L), Glycerol (g/L)

2

0

6

10

20

OD660, Ethanol (g/L), Glycerol (g/L)

4

5

8

15

C) La14 Glucose (g/L)

6

10

20

10

Glucose (g/L)

8

15

B) La10 Glucose (g/L)

Glucose (g/L)

20

10

OD660, Ethanol (g/L), Glycerol (g/L)

A) Control

OD660, Ethanol (g/L), Glycerol (g/L)

Intracellular metabolite profile Metabolome analysis was performed at mid-log phase (OD660 z 0.5) using CE-TOFMS to investigate the metabolic state of cultures under conditions of acid stress. All metabolome data are summarized in Supplementary Table S1. The metabolome data was categorized based on metabolic pathways (Fig. 2). Principal component analysis (PCA) was performed to identify metabolites whose concentrations changed upon exposure to acid stress. The proportion of variance of principal component 1

(PC1) and PC2 were 47% and 20%, respectively (Fig. 3). PCA indicated three different culture pH regions: pH z 6 [La10 (pH 6) and La14 (pH 6)], pH z 4 (control), and pH z 2.5 (La14 and La10). It can be observed from Fig. 3, that PC1 represents the effect of culture pH with a high variance proportion (47%). A positive PC1 score indicates high culture pH conditions while a negative score indicates low culture pH conditions. PC2 represents culture conditions with and without external supplementation of lactic acid. To identify metabolites responsive to acid stress under these conditions, metabolites that contributed to PC1 were extracted based on factor loading scores (Table 2). Intracellular concentrations of metabolites with high and low factor loadings decreased and increased respectively under low pH conditions.

OD660, Ethanol (g/L), Glycerol (g/L)

while those of La10 (pH 6) and La14 (pH 6) were at 6.38 and 6.28, respectively. These results clearly indicate that the un-dissociated form of lactic acid in low extracellular pH conditions leads to decreased intracellular pH and inhibited cell growth.

FIG. 1. Culture profiles of OD660 (closed circles), glucose (closed triangles), glycerol (open diamonds) and ethanol (open squares). Errors indicate the standard deviation of biological triplicate experiments.

Please cite this article in press as: Nugroho, R. H., et al., Metabolomic analysis of acid stress response in Saccharomyces cerevisiae, J. Biosci. Bioeng., (2015), http://dx.doi.org/10.1016/j.jbiosc.2015.02.011

4

NUGROHO ET AL.

J. BIOSCI. BIOENG.,

FIG. 2. Metabolite concentrations obtained by CE-TOFMS analysis. (A) Central metabolic pathway metabolites; (B) energy-related metabolites; (C) glutathione; (D) amino acids and related metabolites.

Please cite this article in press as: Nugroho, R. H., et al., Metabolomic analysis of acid stress response in Saccharomyces cerevisiae, J. Biosci. Bioeng., (2015), http://dx.doi.org/10.1016/j.jbiosc.2015.02.011

VOL. xx, 2015

METABOLOMICS STUDY OF ACID STRESS IN YEAST

12 (2)

8

(3)

4 PC2 (20%)

Oxidative stress is also reported to cause rapid consumption of S7P to generate NADPH for the anti-oxidative stress defense system, which corresponds to increased DHAP accumulation (45). These results are in agreement with our data. G3P concentrations in La10 (pH 6) and La14 (pH 6) were significantly increased from control (Fig. 2A). The accumulation of the glycerol biosynthesis precursor (G3P) could be explained by the osmotic stress response, since lactic acid is present in its sodium salt form in the condition. The high-osmolarity glycerol (HOG) pathway is responsible for the high osmotic stress tolerance by increasing intracellular glycerol concentrations (46,47).

pH 4

(1) (1)

0

(1)

(2) (3)

-8

-8

-4

(2) (3)

pH 6

pH 2.5

-12 -12

(1)

(2) (1)

(2)

(3)

-4

(3)

0 4 PC1 (47%)

8

12

FIG. 3. PCA result of metabolome data. Open diamonds, control; closed squares, La10; closed triangles, La14; open squares, La10 (pH 6); open triangles, La14 (pH 6). Numbers indicated in blankets on the x- and y-axis indicate the proportion of variance for PC1 and PC2, respectively. Numbers in the blanket inside PCA graph indicate three independent biological data. Three pH groups are corresponding to culture pH at OD660 z 0.5.

Central metabolism pathway Pool sizes of many metabolites were decreased in low pH conditions (Fig. 2A), such as fructose-1-6-diphosphate (F1,6 dP), 3-phosphoglycerate (3 PG), 2phosphoglycerate (2 PG), glycerol-3-phosphate (G3P), phosphoenolpyruvate (PEP), sedoheptulose-7-phosphate (S7P), malate, and fumarate. On the other hand, DHAP showed high pools especially in severe acid stress (La14) compared to control and pH z 6 cultures (p values