Background: Saccharomyces cerevisiae is a host for the industrial production of S-adenosyl-l -methionine (SAM), which has been widely used in pharmaceutical and nutritional supplement industries. It has been reported that the intracellular SAM content in S. cerevisiae can be improved by the addition of ethanol during cultivation. However, the 13 13 metabolic state in ethanol-assimilating S. cerevisiae remains unclear. In this study, C-metabolic flux analysis ( C-MFA) was conducted to investigate the metabolic regulation responsible for the high SAM production from ethanol. Results: The comparison between the metabolic flux distributions of central carbon metabolism showed that the metabolic flux levels of the tricarboxylic acid cycle and glyoxylate shunt in the ethanol culture were significantly higher than that of glucose. Estimates of the ATP balance from the C-MFA data suggested that larger amounts of excess ATP was produced from ethanol via increased oxidative phosphorylation. The finding was confirmed by the intracellular ATP level under ethanol-assimilating condition being similarly higher than glucose. Conclusions: These results suggest that the enhanced ATP regeneration due to ethanol assimilation was critical for the high SAM accumulation. Keywords: Saccharomyces cerevisiae, C-based metabolic flux analysis, Ethanol metabolism, Central carbon metabolism, Redox balance Background regeneration in Kyokai strains contribute to high SAM Saccharomyces cerevisiae has been used in several indus- production . trial processes such as for the production of S-adenosyl- It has been reported that the SAM production was l -methionine (SAM) [1, 2]. SAM is synthesized from improved by feeding with ethanol. For instance, the l -methionine and ATP in S. cerevisiae, and acts as a Kyokai strains produce 10.8 g/L of SAM in a 10-L fer- biological methyl group donor involved in many meta- menter under ethanol feed conditions . While the bolic reactions such as transmethylation of proteins. The transcriptome analysis of S. cerevisiae grown on ethanol metabolite has been widely used in pharmaceutical and indicated the transcripts related to gluconeogenesis, the nutritional supplement industries . For SAM produc- glyoxylate shunt, and the tricarboxylic acid (TCA) cycle tion, S. cerevisiae strains used for Japanese sake brew- were upregulated compared with those grown on glucose ing (Kyokai strains) are suitable owing to these higher , the in vivo activity of the metabolic pathway involved intracellular contents among microorganisms . The in the SAM accumulation under ethanol-assimilating metabolic analysis showed that enhanced energy (ATP) condition is still far from clear. To further understand the mechanism in response to SAM production from ethanol, we performed a meta *Correspondence: email@example.com bolic flux analysis for central carbon metabolism since Department of Bioinformatic Engineering, Graduate School SAM is biosynthesized from precursors to key interme- of Information Science and Technology, Osaka University, 1-5 Yamadaoka, Suita, Osaka 565-0871, Japan diates including ribose 5-phosphate, ATP, oxaloacetate, Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), 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. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/ publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 2 of 13 and l-methionine using reductive power (NADPH) and contained 3.8 g/L ethanol, 5.0 g/L (NH ) SO , 0.50 g/L 4 2 4 chemical motive force (ATP). While a metabolic flux K HPO , and 3.4 g/L of yeast nitrogen base without 2 4 distribution in the ethanol-assimilating S. cerevisiae amino acids and ammonium sulfate (Difco Laboratories, was estimated through metabolic flux balancing using a Franklin Lakes, NJ, USA). Vitamins and minerals were compartmented stoichiometric model in silico , the added to the medium, with the final composition as fol - experimental determination by C-metabolic flux anal - lows: 2.0 g/L K H PO , 1.5 g/L MgSO , 2.5 mg/L ZnSO , 2 4 4 4 ysis ( C-MFA) is required to comprehensively quantify 2.4 mg/L MnSO , 0.27 mg/L CuSO , 0.20 g/L CaCl , 4 4 2 central carbon metabolism on ethanol in SAM-produc-4.0 mg/L FeCl , 0.40 mg/L N aMoO ∙2H O, 1.0 mg/L 3 4 2 ing S. cerevisiae . To our knowledge, there have been H BO , 0.20 mg/L KI, 0.20 g/L NaCl, 34 µg/L biotin, 3 3 no prior C-MFA studies on ethanol metabolism in S. 1.6 mg/L Ca-pantothenate, 13 mg/L inositol, 7.8 mg/L cerevisiae. thiamine-HCl, 2.3 mg/L pyridoxine-HCl, 0.40 mg/L In the present study, C-MFA was conducted on a para-aminobenzoic acid, 0.40 mg/L riboflavin, 0.80 mg/L 13 13 high-SAM-producing S. cerevisiae strain with C-labeled niacin, and 4.0 µg/L folic acid. For C-MFA, the carbon 13 13 ethanol as the sole carbon source. C-MFA is based source in the medium was replaced with [2- C]etha- on the cultivation of the cells in medium containing nol. Samples taken from the reactor were centrifuged at C-labeled carbon sources. A metabolic flux distribution 18,800×g for 5 min at 4 °C. The supernatant was used for was estimated from the C-labeling patterns of intracel- analysis of extracellular metabolites. The cell pellet was lular metabolites using mass spectrometry. The selection used for C-labeling analysis by gas chromatography- of a suitable C-labeled carbon source is critical in the mass spectrometry (GC–MS). design of a C-MFA experiment since the accuracy of the flux estimation depends on the labeling patterns of the carbon source. However, a suitable experimental design Offline measurement remains unclear for the C-MFA of the central carbon The analysis of SAM, cell dry weight (CDW), glucose, metabolism in S. cerevisiae cultured with ethanol as the ethanol, glycerol and organic acids such as acetate were sole carbon source. Thus the design of the C-MFA performed as described by Hayakawa et al. . SAM experiment was optimized in this study by a computer concentration was measured by high-performance liq- simulation and found that 100% [2- C] ethanol was the uid chromatography (HPLC) LC2010A-HT (Shimadzu, best carbon source for the precise measurements of met- Kyoto, Japan) after 10% HClO extraction. CDW was abolic fluxes. Using the optimized experimental design, estimated using OD = 1 corresponding to 0.21 g /L. 600 CDW the intracellular metabolic flux distribution of central Glucose and ethanol concentrations were measured carbon metabolism was successfully determined for the enzymatically by using an analyzer (BF-7, Oji Scientific high SAM-producing S. cerevisiae strain (Kyokai no. 6) Instruments, Hyogo, Japan). An enzymatic kit was used in the ethanol limited-chemostat culture. The results for the quantification of glycerol (F-kit Glycerol, R-Biop - showed that the metabolic flux levels through the gly - harm, Washington, MO, USA) in the supernatant. The oxylate shunt and the later TCA cycle were upregulated concentrations of acetate, pyruvate, succinate, fumarate, during growth on ethanol, and the resultant activa- malate, citrate, and lactate in the supernatant were meas- tion of the oxidative phosphorylation should contribute ured by HPLC (LC2010A-HT, Shimadzu). the high SAM accumulation in the ethanol-assimilating The intracellular ATP, l-methionine, and SAM concen - conditions. trations were measured using capillary electrophoresis time-of-flight mass spectrometry (CE-TOFMS) accord - ing to previously described methods . Cultivated cells Methods were harvested by filtration and washed with Milli-Q Strain and growth conditions water. After the membrane was soaked in methanol The S. cerevisiae strain used for Japanese sake brew- solution containing an internal standard (H3304-1002, ing, Kyokai no. 6 (NBRC2346), was purchased from the Human Metabolome Technologies, Yamagata, Japan), National Biological Resource Center (NBRC, Chiba, ultra-sonication was performed. After removal of the Japan). This S. cerevisiae strain was cultivated in an membrane, chloroform and water were added to the aerobic carbon-limited chemostat culture with a work- solution. The aqueous portion collected after mixing with ing volume of 100 mL in a 250-mL fermentor (ABLE a vortex mixer was filtered by Ultrafree MC-PLHCC 250 Co., Tokyo, Japan). The pH was maintained at 5.5 by the (Human Metabolome Technologies) and dried. For CE- automatic addition of 1.0 N NaOH. The temperature was TOFMS, pellets were suspended in the second internal maintained at 30 °C. The stirring speed was 1000 rpm. standard (H3304-1004, Human Metabolome Technolo- The aeration rate was 200 mL/min. The dilution rate −1 gies) solution. Samples were analyzed using the Agilent was 0.06 h . The synthetic medium used for cultivation Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 3 of 13 7100 CE system with Agilent 6224 TOF–MS (Agilent the stoichiometry of anabolic metabolism, the demands Technologies, Santa Clara, CA, USA). of precursor metabolites for cell growth were calculated (Additional file 1: Table S7) and used as constraint condi- 13 13 Optimization of the C tracer for C‑MFA tions in C-MFA (Additional file 1: Table S8). by computational simulation A computer simulation was conducted for the design GC–MS analysis of proteinogenic amino acids of the C-MFA experiment as previously described . Cell pellets were hydrolyzed in 6 N HCl at 105 °C for The intracellular flux distribution data was determined 18 h. After removal of the debris by filtration, the hydro - using the modified literature data through metabolic lysate was dried and dissolved in acetonitrile. For GC– flux balancing . Data for the consumption and pro - MS, the hydrolysate was mixed with an equal volume of duction rates were obtained in this study. A composi- N-(tert-butyldimethylsilyl)-N-methyl-trifluoroacetamide tion of [ C]ethanol was predetermined. The simulated (MTBSTFA), and the mixture was incubated at 95 °C C-enrichments were calculated for the 21 fragments of for 1 h. Next, 1 μL of the sample was injected into the the following amino acids: (M−57) fragments for Ala, GC–MS system (7890A GC and 5975C GC/MSD, Agi- Asp, Glu, Gly, Phe, and Thr; (M−85) fragments for Ala, lent Technologies, Santa Clara, CA, USA), according to Asp, Glu, Gly, Ile, Leu, Pro, Thr, and Val; (M−159) frag- a previously described method [17, 18]. The GC–MS ments for Glu, Ile, Leu, Pro, and Val; and (M302) frag- data were corrected considering the natural abundance ment for Asp; and then the Gaussian noise at 1% levels of C, H, N, O, and Si isotopes for C-MFA . The was added to produce hypothetical measurements for the C-enrichments of the amino acid fragments at an iso- C-enrichment data. A metabolic flux distribution was topic steady state [χ(∞)] can be calculated from the fol- estimated by minimizing the residual between the hypo- lowing equation: thetical measurement and simulated C-enrichments of −μt χ(t) − e · χ(0) amino acid fragments. The 95% confidence intervals were χ ∞ = . (2) ( ) −μt 1 − e determined using the grid search method [10, 11]. Preci- sion scoring was calculated by the following equation to evaluate the range of the 95% confidence intervals: In Eq. (2), t represents the time that has elapsed since feeding C-labeled ethanol-containing medium, χ(t) and S = χ(0) are the C-enrichments of amino acid at t and start (1) [10:0:0:0],i 13 time since feeding the C-labeled ethanol-containing medium, respectively, and, µ is the dilution rate . where r , r , and S are the range of the 95% confi - i [10:0:0:0], i i dence intervals, the range of 95% confidence intervals in C‑metabolic flux analysis 13 13 C-labeled ethanol composition of 10:0:0:0 (100% non- The computational procedure for C-MFA was per- labeled ethanol), and precision score for the ith flux. The formed using a Python version of OpenMebius maximum net flux was 500, which was normalized to an implemented in Python 2.7.9 [9, 21], by which 22 inde- ethanol consumption rate of 100. The metabolic network pendent fluxes were iteratively tuned by minimizing shown in the following was used. the residual between the experimental and simulated C-enrichments of proteinogenic amino acid fragments. Metabolic network Furthermore, the C-labeling patterns of CO were inde- The metabolic model of S. cerevisiae for the C-MFA pendently optimized in the flux estimation. Amino acid was based on previously published models, including fragments used for the flux optimization were shown glycolysis, the pentose phosphate (PP) pathway, the ana- above. Nonlinear optimization was performed using the plerotic pathways, the TCA cycle, gluconeogenesis, and SLSQP (sequential least squares programming) func- the transport reactions between the cytosol and mito- tion implemented in PyOpt 1.2 . The 90% confidence chondria (Additional file 1: Tables S1 and S2) [4, 12, 13]. intervals were determined using the grid search method In the metabolic model, the reaction from the mitochon- [10, 11]. drial oxaloacetate and acetyl-CoA to isocitrate via citrate was assumed to be reversible. The amino acid biosyn - NADPH and ATP demands for cell growth and SAM thesis pathways, including two pathways for glycine and production alanine biosynthesis, were employed (Additional file 1: The NADPH demands for cell growth in assimilat - Table S2) . The composition of the S. cerevisiae bio - ing glucose and ethanol were calculated by using Addi- mass was determined on basis of the literature (Addi- tional file 1: Tables S3–S6 and the KEGG database tional file 1: Tables S3–S6) [15, 16]. Using these data and (http://www.genom e.jp/kegg/) (9.42 and 11.95 mmol/ Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 4 of 13 g , respectively). The literature data were used for glucose as with previous research [25, 26]. In contrast, CDW the ATP demands for cell growth on glucose and etha- the acetate yield in assimilating ethanol was higher than nol (39.78 and 105.56 mmol/g , respectively) . A that of glucose. This might be related to the enhanced CDW non-growth-associated maintenance ATP requirement expression of acetaldehyde dehydrogenase (ALD4 and of 1.0 mmol ATP/g /h was also considered . The ALD6) under assimilating ethanol conditions . CDW total ATP cost for SAM biosynthesis was found to be 13.0 mol ATP per mol SAM . 13 13 Design of C‑MFA using C‑labeled ethanol As mentioned in the background, C-MFA is based on Results 13 the cultivation of cells in medium containing C-labeled Physiological parameters of ethanol‑assimilating S. 13 carbon sources. Selection of a suitable C-labeled car- cerevisiae 13 bon source is critical in the design of a C-MFA experi- The S. cerevisiae strain used for Japanese sake brewing ment since the accuracy of the flux estimation depends (Kyokai no. 6, NBRC2346) was cultivated in an ethanol- on labeling patterns of the carbon source. In this study, limited chemostat culture under aerobic condition at a 13 a suitable experiment was designed for the C-MFA of −1 dilution rate of 0.06 h . The physiological parameters the central carbon metabolism in S. cerevisiae cultured in the cultivation are shown in Table 1. A comparison on ethanol as sole carbon source. A relationship between with the previous glucose-limited chemostat culture 13 13 the composition of C-labeled ethanol including [1- C],  showed that the SAM specific production rate and 13 13 [2- C], and [U- C]ethanol, and 95% confidence inter - intracellular content in assimilating ethanol (0.27 μmol/ vals of the estimated flux levels were investigated by a g /h and 1.8 mg/g , respectively) were greater than 13 CDW CDW computer simulation of C-MFA experiment. that of a glucose-limited culture under the same dilution The metabolic model of S. cerevisiae describing the and carbon atom supply rate conditions (0.089 μmol/ stoichiometry equations and carbon atom transitions of g /h and 0.58 mg/g , respectively). The biomass CDW CDW each metabolic reaction was constructed from previously yield in assimilating ethanol was lower than that of published models with modifications (Additional file 1: Tables S1 and S2) [4, 12, 13]. The computer simulation was conducted by the following procedure : using a Table 1 Fermentation profiles of the high‑SAM ‑producing flux distribution of ethanol-assimilating S. cerevisiae esti - strain (Kyokai no. 6) under aerobic carbon‑limited mated by an in silico calculation method , a series of −1 chemostat cultures (dilution rate of 0.06 h ) artificial mass spectra datasets of amino acid fragments were generated for various compositions of [ C]ethanol Carbon source 13 13 13 (all patterns of non-labeled, [1- C], [2- C], and [U- C] Glucose Ethanol ethanol with a 50% step size) by adding Gaussian noise at Carbon source consumption rate (mmol/g /h) 0.65 2.5 1% levels. The metabolic flux distribution and 95% confi - CDW SAM content (mg/g ) 0.58 1.8 dence intervals were determined using the artificial mass CDW Yield on carbon source spectra datasets by the C-MFA procedure (Additional Cell (g /mol C) 15 12 file 1: Table S9). A precision score S was determined for CDW CO (mol C/mol C) 0.41 0.54 each metabolic reaction i from the data of the 95% con- Specific production rate (μmol/g /h) fidence intervals. Useful compositions of [ C]ethanol CDW SAM 0.089 0.27 were investigated by comparing the precision score S of Malate 3.6 5.1 estimated flux distributions. Acetate 4.9 29 For example, the 95% confidence intervals of isoci - Lactate 5.9 2.0 trate dehydrogenase (IDH) flux (isocitrate→α- Pyruvate 0 0.27 ketoglutarate + CO ) with mixtures of non-labeled, 13 13 13 Citrate 1.3 0.51 [1- C], [2- C], [U- C]ethanol at 10:0:0:0, 5:5:0:0, and Succinate 4.2 4.3 0:0:10:0 were estimated to be 2.5–45.1, 19.0–42.3, and Glycerol 0.35 0.44 27.4–41.6, from which S were estimated to be 1.0, 0.55, IDH and 0.33, respectively (Additional file 1: Table S9). Since a Ethanol 0 – smaller S score shows a more precise flux estimation, IDH All data were obtained from a single experiment. CDW cell dry weight a the results indicated that the mixture of non-labeled, Data from previous study  13 13 13 [1- C], [2- C], [U- C]ethanol at 0:0:10:0 was a suitable Estimated by the carbon recoveries for biomass synthesis calculated from the specific growth rate and the reported elemental composition of carbon carbon sources for the flux estimation of this reaction (glucose: 0.455 g/g , ethanol: 0.467 g/g ) , and by the specific rates for CDW CDW in the three mixtures. Figure 1 shows a heatmap in the carbon source consumption and for products including SAM productions. The total carbon recovery was considered to be 1.0 precision score S based on the range of 95% confidence i Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 5 of 13 13 13 13 Fig. 1 Heatmap of the precision score S levels estimated by the computer simulation of C-MFA using 10 mixtures of non-labeled, [1- C], [2- C], and [U- C]ethanol as the carbon sources. The magenta and green colors in the boxes represent narrower (better precision) and wider (poorer precision) 95% confidence interval levels of the estimated metabolic fluxes on each reaction, respectively interval for the estimated flux in each composition of depended greatly on the composition of C-labeled etha- 13 13 C-labeled ethanol. The magenta and green colors rep - nol. The mixture ratio at 0:0:10:0 (100% [2- C]ethanol) resent narrower (better precision) and wider (poorer pre- showed the smallest sum of the precision score Si in the cision) 95% confidence intervals, respectively. This result TCA cycle and PP pathway among all C-labeling con- indicates that the range of the 95% confidence interval ditions (Fig. 1). Based on the results, [2- C]ethanol was Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 6 of 13 (See figure on next page.) −1 Fig. 2 Metabolic flux distributions (mmol/g /h) for S. cerevisiae grown on glucose (a)  and ethanol (b) at a dilution rate of 0.06 h . Flux CDW values in the central carbon metabolism are shown with 90% confidence intervals. The width of each arrow indicates best-fit value (Additional file 1: Table S11). The anabolic reactions from metabolic intermediates to biomass are represented by small arrows. The product synthesis reactions are represented by dotted arrows. These flux values are shown in Additional file 1: Table S8. G6P glucose-6-phosphate, F6P fructose-6-phosphate, DHAP dihydroxyacetone phosphate, GAP 3-phosphoglyceraldehyde, PGA 3-phosphoglycerate, PEP phosphoenolpyruvate, Pyr_c cytosolic pyruvate, Pyr_m mitochondrial pyruvate, AcCOA_c cytosolic acetyl-CoA, AcCOA_m mitochondrial acetyl-CoA, AcAl acetaldehyde, Ac acetate, IsoCit isocitrate, aKG α-ketoglutarate, Suc succinate, Mal malate, Oxa_c cytosolic oxaloacetate, Oxa_m mitochondrial oxaloacetate, 6PG 6-phosphogluconate, Ru5P ribulose-5-phosphate, R5P ribose-5-phosphate, Xu5P xylulose-5-phosphate, S7P sedoheptulose-7-phosphate, E4P erythrose-4-phosphate, Glxy glyoxylate employed for the C-MFA of the central carbon metabo- the assimilated ethanol flowed into the EMP pathway lism in S. cerevisiae cultured on ethanol as the sole car- via the glyoxylate shunt and then converted to glucose- bon source. 6-phosphate. The metabolic flux in the glyoxylate shunt (IsoCit→Glxy + Suc, AcCOA_c + Glxy→Oxa_c) for the C‑Metabolic flux analysis glucose and ethanol assimilations were 0.007–0.03 and The Kyokai no. 6 strain was cultivated for C-MFA by an 0.58–0.68 mmol/g /h, respectively. This was in agree - CDW ethanol-limited chemostat culture under aerobic condi- ment with the gene transcription levels of ICL1, ICL2, −1 tions at a dilution rate 0.06 h . At 98 h after the start MLS1, and MLS2 in the glyoxylate shunt that were upreg- of cultivation when the cell growth reached steady-state, ulated in a transcriptome analysis of S. cerevisiae grown the medium containing 100% [2- C]ethanol was fed to on ethanol described previously . Since the expres- the chemostat cultures. The biomass samples were har - sion of malate synthase (AcCOA_c + Glxy→Oxa_c) was vested at 1.3, 2.3, 2.7, 3.7, 4.1, and 5.1 residential times essential for growth on ethanol , the result confirmed after feeding C-labeled ethanol. Following acid hydroly- that the pathway was active during the assimilation of sis and derivatization, C-enrichment of proteinogenic ethanol. amino acids was measured using GC–MS analysis (Addi- To investigate the cofactor balance under the ethanol- tional file 1: Fig. S1). The intracellular metabolism might assimilating conditions, metabolic flux levels respon - reach an isotopic steady-state after an infinite amount of sible for the NADPH and ATP regeneration were time. In this study, the C-enrichment data indicated by compared (Fig. 3a, b). In S. cerevisiae, NADPH was Eq. (2) were used to calculate metabolic flux distributions mainly regenerated via the PP pathway (glucose-6-phos- (Additional file 1: Table S10 and Fig. S1). phate dehydrogenase and phosphogluconate dehydro- A metabolic flux distribution was estimated by mini - genase), malic enzyme, and isocitrate dehydrogenase. mizing the difference between the computationally sim - The results showed that the TCA cycle functioned as ulated C-enrichments of proteinogenic amino acids the main provider of NADPH on ethanol since the flux and the experimentally obtained data by GC–MS. The level of isocitrate dehydrogenase (IsoCit + NADP → differences (residual sum of squares) between experi - aKG + NADPH + CO ) in the TCA cycle was higher mental and simulated C-enrichment data for the best- on ethanol (0.91–1.1 mmol/g /h) compared to glu- CDW fitted metabolic flux distributions that passed the χ -test cose (0.34–0.42 mmol/g /h) (Figs. 2 and 3a). On the CDW (α = 0.05) were small (69.5) , indicating that the esti- other hand, the glucose-6-phosphate dehydrogenase mated flux distributions could explain the experimentally and phosphogluconate dehydrogenase flux (G6P + 2 13 + obtained C-enrichments (Additional file 1: Table S10). NADP →Ru5P + 2 NADPH + CO ) in the oxidative PP Based on the best-estimate metabolic flux distribution, pathway was 0–0.25 mmol/g /h on ethanol, which CDW 90% confidence intervals of the metabolic flux levels were was lower compared to the 0.36–0.58 mmol/g /h on CDW determined using the grid search method [10, 11]. glucose, indicating that the carbon fluxes of the central The 90% confidence intervals of the flux levels on glu - carbon metabolism on ethanol were redirected into the cose and ethanol in the Kyokai no. 6 strain at a dilution TCA cycle (Fig. 2). The malic enzyme flux (Mal + NADP −1 + rate 0.06 h are shown in Fig. 2. The results for glucose →Pyr_m + NADPH + CO ) was similar on glucose and have been previously published . The results revealed ethanol (0.04–0.08 and 0–0.12 mmol/g /h, respec- CDW that the metabolic flux distribution was totally rewired tively), which showed that this reaction did not con- in the ethanol culture from that under the glucose cul- tribute greatly in the regeneration of NADPH on both ture. The Embden–Meyerhof–Parnas (EMP) pathway carbon sources (Fig. 2). was reversed from the glycolytic direction in the glucose Figure 3b shows the ATP balance in the glucose- and culture to the direction for gluconeogenesis because ethanol-assimilating conditions. The data revealed that Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 7 of 13 oxidative phosphorylation using the electron transport are from the EMP pathway, TCA cycle, glyoxylate shunt, system during respiration was a major source of ATP alcohol dehydrogenase, and aldehyde dehydrogenase. regeneration. The metabolic flux distribution showed The flux levels of these reactions were also largely differ - that the NADH supply for the electron transport system ent between glucose and ethanol. For instance, the malate Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 8 of 13 detected enzyme activity under the ethanol-grown cul- ture . Redox balance As mentioned above, C-MFA showed that the flux lev - els related to NADPH and ATP regeneration were largely different during growth on glucose and ethanol, while the flux of cytosolic oxaloacetate to SAM biosynthesis corre - sponded to 0.01% of the ethanol consumption flux value. Hence, the relationship between cofactor (NADPH and ATP) balances and SAM biosynthesis was investigated. To analyze the NADPH balance, the NADPH regenera- tion rates estimated from the metabolic flux data were compared with the NADPH consumption rates for cell growth. Since the composition of S. cerevisiae biomass differs among the carbon source used for the culture, NADPH demand levels were determined using Addi- tional file 1: Tables S3–S6 and the KEGG database (see Materials and methods). As shown in Fig. 3a, NADPH regeneration rates were larger than the consumption rates for cell growth in both conditions. In assimilating glucose and ethanol, the differences between NADPH regenera - Fig. 3 Redox and energy balances. NADPH (a) and ATP (b) balances tion and consumption rates were 0.75 and 0.42 mmol/ in assimilating glucose and ethanol were determined at a dilution −1 g /h, respectively, indicating that the difference on rate of 0.06 h . The error bars at each bar represent the 90% CDW confidence intervals of the flux estimations glucose was greater than that of ethanol (Fig. 3a). Excess NADPH was assumed to be used for ATP production by the oxidation of NADPH as previously shown . dehydrogenase flux (Mal + NAD →Oxa_m + NADH) in the TCA cycle on ethanol corresponded to 59–65% of ATP balance ethanol consumption rate (1.5–1.6 mmol/g /h), while Since the high levels of SAM production by the Kyokai CDW the flux level on glucose corresponded to 35–48% of glu - no. 6 strain could be attributed to enhanced ATP regen- cose consumption rate (0.23–0.31 mmol/g /h) (Fig. 2). eration , ATP balances for regeneration and consump- CDW In contrast, the pyruvate dehydrogenase (Pyr_m + NAD tion were investigated (Fig. 3b). Total ATP regeneration →AcCOA_m + CO + NADH) in the EMP pathway was rates were calculated based on the metabolic flux of inactive on ethanol because of a low flux (0–0.06 mmol/ reactions responsible for the respiration and substrate g /h), indicating that the TCA cycle also functioned level phosphorylation. The biosynthetic ATP demand CDW as the provider of NADH on ethanol (Fig. 2). Therefore, was estimated from the literature data (see “Methods”). this result suggests that the high TCA activity in ethanol The ATP consumption rate for cytosolic acetyl-CoA assimilation caused the lower cell yield on carbon source synthesis was calculated from acetyl-CoA synthase flux compared with glucose assimilation, since two CO mol- (Ac→AcCOA_c). The difference between ATP regenera - ecules were released from the TCA cycle (Table 1). tion and consumption on ethanol was 5.1 mmol/g /h, CDW Next, the carbon flows around cytosolic oxaloacetate, indicating that the difference on ethanol was 1.9-fold which is the precursor for SAM, were compared. The lev - greater than that of glucose (2.8 mmol/g /h) (Fig. 3b). CDW els of SAM biosynthesis flux from cytosolic oxaloacetate on glucose- and ethanol-assimilating (0.09 and 0.3 µmol/ Intracellular ATP, l ‑methionine, and SAM concentration g /h, respectively) were only 0.01% or less of the glu- The C-metabolic flux analysis revealed that SAM CDW cose- and ethanol-consumption flux values (0.65 and accumulation on ethanol should be derived from 2.5 mmol/g /h, respectively) (Fig. 2). In turn, the flux enhanced ATP regeneration. It was also expected that CDW levels of pyruvate carboxylase (Pyr_c + CO →Oxa_c) on the intracellular ATP concentration on ethanol could glucose and ethanol were 0.14–0.18 and 0.42–1.4 mmol/ be higher than that of glucose since the SAM content −1 g /h, respectively, indicating that pyruvate carboxy- in assimilating ethanol at a dilution rate of 0.06 h CDW lase were used to supply cytosolic oxaloacetate on both (1.8 mg/g ) (Table 1) was greater than that of glu- CDW −1 carbon sources (Fig. 2). This result was confirmed by the cose at a dilution rate of 0.1 h (1.0 mg/g ) . CDW Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 9 of 13 Then, intracellular ATP concentrations were deter - grown on ethanol (24.8 ± 1.6 μmol/g ) was 1.4-fold CDW mined by CE-TOFMS analysis to further investigate higher than that of glucose (17.5 ± 2.6 μmol/g ) CDW the effects of ATP levels on SAM productivity. For the (Fig. 4b). This result confirmed that the S. cerevisiae comparison of ATP levels, glucose- and ethanol-limited cells grown on ethanol had the potential for high SAM chemostat cultures were conducted at dilution rates of content due to high intracellular ATP levels as shown −1 0.1 and 0.06 h , respectively. After the batch phase (at in the metabolic flux distribution estimated using C- 11 and 44 h after the initiation of cultivation for glucose MFA. Similarly, the intracellular l -methionine and and ethanol, respectively), the carbon-limited chemo- SAM levels on ethanol (0.10 ± 0.0019 μmol/g and CDW stat cultures were started. After reaching steady-state 1.2 ± 0.067 μmol/g ) were 1.7- and 1.2-fold higher CDW cell growth, the S. cerevisiae cells were repeatedly sam- than those of glucose (0.060 ± 0.0024 μmol/g and CDW pled (at 75, 82, and 99 h on glucose; and at 112, 119, 1.0 ± 0.12 μmol/g ), respectively, indicating that CDW and 136 h on ethanol; respectively) (Fig. 4a). In this improvement of intracellular ATP and l -methionine study, three samples were considered to be experimen- levels enhanced SAM biosynthesis (Fig. 4c, d). tal triplicate. The intracellular ATP levels in the cells Fig. 4 Carbon-limited chemostat cultures of S. cerevisiae. a Cell dry weight (CDW; g /L); b intracellular ATP concentration (µmol/g ); c CDW CDW intracellular l -methionine concentration (µmol/g ); d intracellular SAM concentration (µmol/g ). The bold arrows indicate the time points CDW CDW at the start of the carbon-limited chemostat cultures. The dash arrows represent the time points of sampling for intracellular metabolite analysis. Values of intracellular ATP concentration represent the average of five measurements with error bars calculated as the standard deviations of the means. Values of intracellular l -methionine and SAM concentration represent the average of three measurements with error bars calculated as the standard deviations of the means Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 10 of 13 Discussion under the ethanol-assimilating conditions (Fig. 5). Fur- In the present study, C-MFA was conducted to investi- thermore, estimation of the ATP balance revealed that a gate a metabolic regulation responsible for the high intra- higher amount of excess ATP was produced on ethanol cellular SAM content in ethanol-assimilating S. cerevisiae Kyokai no. 6 strain [5, 28]. The design of the C-MFA experiment was optimized by computer simulations, which showed that 100% [2- C]ethanol was the best carbon source for the precise estimation of metabolic fluxes through the TCA cycle and PP pathway (Fig. 1 and Additional file 1: Table S9). An advantage of having 100% 13 13 [2- C]ethanol in C-MFA was that the second carbon of ethanol contributes to the generation of various meta- bolic intermediates with various C-labeling pattern, which in turn increases the sensitivity to fluxes. On the other hand, the first and fourth carbons of mitochondrial oxaloacetate derived from the first carbon of ethanol are released as CO in the TCA cycle . The C-MFA in Escherichia coli cultured on acetate was performed using the mixtures of non-labeled, [2- C]acetate at a ratio of 8:2 because of the same reason . A comparison of the metabolic flux distribution on glu - cose and ethanol revealed that the metabolic flux through the EMP pathway was reversed by the activation of the glyoxylate shunt and gluconeogenesis during growth on ethanol (Fig. 2). The metabolic redirection coincided with the regulation in mRNA and protein expression levels as reported in previous studies. It has been reported that S. cerevisiae grown on ethanol increased the transcript levels of genes involved in gluconeogenesis and the gly- oxylate shunt compared to cells grown on glucose . Furthermore, comparable results have been observed in a proteome analysis under chemostat cultures limited for glucose and ethanol . The previous transcriptome and proteome studies during the diauxic shift in S. cerevisiae [32–34] also showed that the glyoxylate shunt, gluconeo- genesis, and oxidative phosphorylation were activated after starting the assimilation of the produced ethanol. On the other hand, the downregulation of the pyruvate dehydrogenase flux level (Pyr_m→AcCOA_m + CO ) at the entry point of the TCA cycle disagreed with a previ- ous transcriptome analysis because the expression levels of PDA1, PDB1, PDX1, and LPD1 genes in the ethanol- assimilating conditions increased from that of glucose (Fig. 2) . This may have been due to pyruvate dehy - drogenase being allosterically regulated by NADH . These results suggested that the fluxes through the cen - tral carbon metabolism in S. cerevisiae were regulated at the transcript, translational, and post-translational levels [6, 31, 36]. Fig. 5 Hypothesized mechanism for the high SAM accumulation. It has been reported that the ethanol-assimilating S. a Glucose; b ethanol (carbon source for higher SAM accumulation). cerevisiae cells showed a higher oxygen uptake rate . 13 The dotted gray arrows represent respiration. The C-MFA results The results of C-MFA confirmed that ATP regeneration revealed that high SAM accumulation can be explained by enhanced ATP regeneration with high respiration activity via the oxidative phosphorylation significantly increased Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 11 of 13 (Fig. 3b), since upregulation of ATP regeneration via the uncovered by a systems-level analysis of the metabolism oxidative phosphorylation was greater than the increase in ethanol-assimilating S. cerevisiae using C-MFA [46, in ATP consumption for gluconeogenesis and cyto- 47]. solic acetyl-CoA synthesis during ethanol assimilation . These results suggested that the excess ATP should activate SAM biosynthesis. However, the excess ATP Conclusions regeneration flux (5.1 mmol/g /h) was clearly greater In this study, the design of the C-MFA experiment was CDW than the ATP consumption flux for SAM production optimized by computer simulations, which showed that (3.5 µmol/g /h) (Table 1 and Fig. 3b). The excess ATP 100% [2- C]ethanol was the best carbon source for the CDW may be utilized for the response of stress caused by acet- precise estimation of metabolic fluxes through the TCA aldehyde and acetate [37, 38]. cycle and PP pathway. The C-MFA result revealed that The analysis of the intracellular metabolites (Fig. 4) the metabolic flux distribution was totally rewired in revealed that the ATP level in the cells grown on ethanol the ethanol culture from that of glucose. Estimates of was higher than that of glucose (Fig. 4b). It was reported the ATP balance from the C-MFA data suggested that that the increase of intracellular ATP level enhanced the larger amounts of excess ATP was produced from etha- SAM production in S. cerevisiae. The addition of sodium nol via increased oxidative phosphorylation. The find - citrate improved the isocitric acid dehydrogenase activ- ing was confirmed by the intracellular ATP level under ity and ATP level in the cell, which promoted the conver- ethanol-assimilating condition being similarly higher sion of methionine into SAM . Optimization of the than glucose. These results suggest that the excess ATP culture medium also revealed that the restriction of cell regeneration via the activation of oxidative phosphoryla- growth, and the enhancement of the intracellular ATP tion was a mechanism responsible for the SAM overpro- level and SAM production were achieved by reducing the duction under the ethanol-assimilating conditions, which supplemented yeast extract . Moreover, the result of also provided general knowledge for the development of this study that the intracellular l-methionine level in eth - S. cerevisiae cell factories for the cytosolic acetyl-CoA- anol assimilating was higher than that of glucose (Fig. 4c). derived products. Since l-methionine biosynthesis consumes cytosolic Additional file acetyl-CoA in S. cerevisiae , the upregulation might be concerned with the higher flux level of acetyl-CoA Additional file 1. Additional tables and figures. synthase (Ac→AcCOA_c) on ethanol (2.5–2.6 mmol/ g /h) compared to glucose (0.13–0.43 mmol/g /h) CDW CDW (Fig. 2). These results indicated that both ATP and Authors’ contributions l-methionine levels in cells had a positive effect on SAM KH designed and carried out all experiments, and drafted the manuscript. FM revised the manuscript and further strengthened the interpretation of the biosynthesis (Table 1 and Fig. 4d). data. HS supervised and coordinated the research, and revised the manuscript. In S. cerevisiae, cytosolic acetyl-CoA was synthesized All authors read and approved the final manuscript. from acetate and CoA by acetyl-CoA synthase with con- Author details suming ATP . The increased ATP regeneration in the Department of Bioinformatic Engineering, Graduate School of Informa- ethanol assimilation conditions would also enhance the tion Science and Technology, Osaka University, 1-5 Yamadaoka, Suita, production of other useful chemicals, since the cytosolic Osaka 565-0871, Japan. KANEKA Fundamental Technology Research Alliance Laboratories, Graduate School of Engineering, Osaka University, 2-8 acetyl-CoA is a common precursor for the biosynthesis Yamadaoka, Suita, Osaka 565-0871, Japan. Biotechnology Development of fatty acid ethyl ester (biodiesel), geraniol (flavor), and Laboratories, Health Care Solutions Research Institute, Kaneka Corporation, 1-8 amorpha-4, 11-diene (precursor to artemisinin) [42–44]. Miyamae-cho, Takasago-cho, Takasago, Hyogo 676-8688, Japan. The results of this study also suggested that the saving on Acknowledgements the ATP consumption for cytosolic acetyl-CoA would We thank Dr. Yoshihiro Toya for his helpful comments regarding this manu- further increase SAM production. For instance, it has script, Dr. Nobuyuki Okahashi for his support to perform GC–MS analysis, Dr. Shuichi Kajihata and Mr. Kousuke Maeda for their support to perform C-MFA, been reported that introducing the metabolic reaction and Dr. Yoshifumi Fukui and Dr. Shinji Ozawa for their helpful suggestions. that produces acetyl-CoA from acetaldehyde using an ATP-independent enzyme could increase cell yield . Competing interests The authors declare that they have no competing interests. The C-MFA conduced in this study revealed that excess ATP regeneration via the activation of oxidative Availability of data and materials phosphorylation was a mechanism responsible for the The datasets used in the current study are available from the corresponding author on reasonable request. SAM overproduction under the ethanol-assimilating conditions. A detailed mechanism for the metabolic Consent for publication redirection in the central carbon metabolism could be Not applicable. Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 12 of 13 Ethics approval and consent to participate 17. Mori E, Furusawa C, Kajihata S, Shirai T, Shimizu H. Evaluating C enrich- Not applicable. ment data of free amino acids for precise metabolic flux analysis. Biotech- nol J. 2011;6(11):1377–87. Funding 18. Toya Y, Hirasawa T, Morimoto T, Masuda K, Kageyama Y, Ozaki K, Ogasa- This work was supported in-part by Grants-in-Aid for Scientific Research (grant wara N, Shimizu H. C-Metabolic flux analysis in heterologous cellulase no. 17H06303). production by Bacillus subtilis genome-reduced strain. J Biotechnol. 2014;179:42–9. 19. van Winden WA, Wittmann C, Heinzle E, Heijnen JJ. Correcting mass isoto- Publisher’s Note pomer distributions for naturally occurring isotopes. Biotechnol Bioeng. Springer Nature remains neutral with regard to jurisdictional claims in pub- 2002;80(4):477–9. lished maps and institutional affiliations. 20. van Winden W, Schipper D, Verheijen P, Heijnen J. Innovations in genera- 13 1 tion and analysis of 2D [ C, H] COSY NMR spectra for metabolic flux Received: 9 February 2018 Accepted: 22 May 2018 analysis purposes. Metab Eng. 2001;3(4):322–43. 21. Kajihata S, Furusawa C, Matsuda F, Shimizu H. OpenMebius: an open source software for isotopically nonstationary C-based metabolic flux analysis. Biomed Res Int. 2014;2014:627014. 22. Perez RE, Jansen PW, Martins JRRA. pyOpt: a Python-based object- References oriented framework for nonlinear constrained optimization. Struct 1. Kanai M, Mizunuma M, Fujii T, Iefuji H. A genetic method to enhance the Multidiscip Optim. 2012;45:101–18. accumulation of S-adenosylmethionine in yeast. Appl Microbiol Biotech- 23. Verduyn C. Physiology of yeasts in relation to biomass yields. Antonie Van nol. 2017;101(4):1351–7. Leeuwenhoek. 1991;60(3–4):325–53. 2. Shiozaki S, Shimizu S, Yamada H. Unusual intracellular accumulation 24. Verduyn C, Postma E, Scheffers WA, van Dijken JP. Energetics of Saccharo - of S-Adenosyl-l -methionine by microorganisms. Agric Biol Chem. myces cerevisiae in anaerobic glucose-limited chemostat cultures. J Gen 1984;48(9):2293–300. Microbiol. 1990;136(3):405–12. 3. Shobayashi M, Mukai N, Iwashita K, Hiraga Y, Iefuji H. A new method 25. Verduyn C, Stouthamer AH, Scheffers WA, van Dijken JP. A theoreti- for isolation of S-adenosylmethionine (SAM)-accumulating yeast. Appl cal evaluation of growth yields of yeasts. Antonie Van Leeuwenhoek. Microbiol Biotechnol. 2006;69(6):704–10. 1991;59:49–63. 4. Hayakawa K, Kajihata S, Matsuda F, Shimizu H. C-metabolic flux analysis 26. de Jong-Gubbels P, Vanrolleghem P, Heijnen S, van Dijken JP, Pronk JT. in S-adenosyl-l -methionine production by Saccharomyces cerevisiae. J Regulation of carbon metabolism in chemostat cultures of Saccha- Biosci Bioeng. 2015;120:532–8. romyces cerevisiae grown on mixtures of glucose and ethanol. Yeast. 5. Shiozaki S, Shimizu S, Yamada H. Production of S-adenosyl-l -methionine 1995;11(5):407–18. by Saccharomyces sake. J Biotechnol. 1986;4(6):345–54. 27. Chen Y, Siewers V, Nielsen J. Profiling of cytosolic and peroxisomal acetyl- 6. Daran-Lapujade P, Jansen ML, Daran JM, van Pronk W, de Winde JH, Gulik CoA metabolism in Saccharomyces cerevisiae. PLoS ONE. 2012;7(8):e42475. JT. Role of transcriptional regulation in controlling fluxes in central carbon 28. Shiozaki S, Shimizu S, Yamada H. S-Adenosyl-l -methionine production metabolism of Saccharomyces cerevisiae. A chemostat culture study. J Biol by Saccharomyces sake: optimization of the culture conditions for the Chem. 2004;279(10):9125–38. production of cells with a high S-Adenosyl-l -methionine content. Agric 7. Toya Y, Shimizu H. Flux analysis and metabolomics for systematic meta- Biol Chem. 1989;53:3269–74. bolic engineering of microorganisms. Biotech Adv. 2013;31(6):818–26. 29. Matsuoka Y, Shimizu K. The relationships between the metabolic fluxes 8. Hayakawa K, Matsuda F, Shimizu H. Metabolome analysis of Saccharo- and C-labeled isotopomer distribution for the flux analysis of the main myces cerevisiae and optimization of culture medium for S-adenosyl- metabolic pathways. Biochem Eng J. 2010;49(3):326–36. l -methionine production. AMB Express. 2016;6(1):38. 30. Zhao J, Shimizu K. Metabolic flux analysis of Escherichia coli K12 grown on 9. Maeda K, Okahashi N, Toya Y, Matsuda F, Shimizu H. Investigation of C-labeled acetate and glucose using GC-MS and powerful flux calcula- useful carbon tracers for C-metabolic flux analysis of Escherichia coli by tion method. J Biotechnol. 2003;101(2):101–17. considering five experimentally determined flux distributions. Metab Eng 31. Kolkman A, Olsthoorn MM, Heeremans CE, Heck AJ, Slijper M. Com- Commun. 2016;3:187–95. parative proteome analysis of Saccharomyces cerevisiae grown in 10. van Winden WA, van Dam JC, Ras C, Kleijn RJ, Vinke JL, van Gulik WM, chemostat cultures limited for glucose or ethanol. Mol Cell Proteomics. Heijnen JJ. Metabolic-flux analysis of Saccharomyces cerevisiae CEN.PK113- 2005;4(1):1–11. 7D based on mass isotopomer measurements of C-labeled primary 32. DeRisi JL, Iyer VR, Brown PO. Exploring the metabolic and genetic control metabolites. FEMS Yeast Res. 2005;5(6–7):559–68. of gene expression on a genomic scale. Science. 1997;278(5338):680–6. 11. Antoniewicz MR, Kelleher JK, Stephanopoulos G. Determination of 33. Murphy JP, Stepanova E, Everley RA, Paulo JA, Gygi SP. Comprehensive confidence intervals of metabolic fluxes estimated from stable isotope temporal protein dynamics during the diauxic shift in Saccharomyces measurements. Metab Eng. 2006;8(4):324–37. cerevisiae. Mol Cell Proteomics. 2015;14(9):2454–65. 12. Gombert AK, Moreira dos Santos M, Christensen B, Nielsen J. Network 34. Zampar GG, Kümmel A, Ewald J, Jol S, Niebel B, Picotti P, Aebersold R, identification and flux quantification in the central metabolism of Sac- Sauer U, Zamboni N, Heinemann M. Temporal system-level organization charomyces cerevisiae under different conditions of glucose repression. J of the switch from glycolytic to gluconeogenic operation in yeast. Mol Bacteriol. 2001;183(4):1441–51. Syst Biol. 2013;9:651. 13. Frick O, Wittmann C. Characterization of the metabolic shift between 35. Kim Y, Ingram LO, Shanmugam KT. Dihydrolipoamide dehydrogenase oxidative and fermentative growth in Saccharomyces cerevisiae by com- mutation alters the NADH sensitivity of pyruvate dehydrogenase com- parative C flux analysis. Microb Cell Fact. 2005;4:30. plex of Escherichia coli K-12. J Bacteriol. 2008;190(11):3851–8. 14. Monschau N, Stahmann KP, Sahm H, McNeil JB, Bognar AL. Identification 36. Oliveira AP, Ludwig C, Picotti P, Kogadeeva M, Aebersold R, Sauer U. Regu- of Saccharomyces cerevisiae GLY1 as a threonine aldolase: a key enzyme in lation of yeast central metabolism by enzyme phosphorylation. Mol Syst glycine biosynthesis. FEMS Microbiol Lett. 1997;150(1):55–60. Biol. 2012;8:623. 15. Stückrath I, Lange HC, Kötter P, van Gulik WM, Entian KD, Heijnen JJ. Char- 37. Tuma DJ, Casey CA. Dangerous byproducts of alcohol breakdown–focus acterization of null mutants of the glyoxylate cycle and gluconeogenic on adducts. Alcohol Res Health. 2003;27(4):285–90. enzymes in S. cerevisiae through metabolic network modeling verified by 38. Caspeta L, Castillo T, Nielsen J. Modifying yeast tolerance to inhibitory chemostat cultivation. Biotechnol Bioeng. 2002;77(1):61–72. conditions of ethanol production processes. Front Bioeng Biotechnol. 16. Förster J, Famili I, Fu P, Palsson BØ, Nielsen J. Genome-scale reconstruc- 2015;3:184. tion of the Saccharomyces cerevisiae metabolic network. Genome Res. 39. Chen H, Wang Z, Wang Z, Dou J, Zhou C. Improving methionine and ATP 2003;13(2):244–53. availability by MET6 and SAM2 co-expression combined with sodium citrate feeding enhanced SAM accumulation in Saccharomyces cerevisiae. World J Microbiol Biotechnol. 2016;32(4):56. Hayakawa et al. Microb Cell Fact (2018) 17:82 Page 13 of 13 40. Thomas D, Surdin-Kerjan Y. Metabolism of sulfur amino acids in Saccharo- 44. Jiang GZ, Yao MD, Wang Y, Zhou L, Song TQ, Liu H, Xiao WH, Yuan YJ. myces cerevisiae. Microbiol Mol Biol. 1997;61(4):503–32. Manipulation of GES and ERG20 for geraniol overproduction in Saccharo- 41. de Kok S, Kozak BU, Pronk JT, van Maris AJ. Energy coupling in Saccharo- myces cerevisiae. Metab Eng. 2017;41:57–66. myces cerevisiae: selected opportunities for metabolic engineering. FEMS 45. Kozak BU, van Rossum HM, Niemeijer MS, van Dijk M, Benjamin K, Wu L, Yeast Res. 2012;12(4):387–97. Daran JM, Pronk JT, van Maris AJ. Replacement of the initial steps of etha- 42. Westfall PJ, Pitera DJ, Lenihan JR, Eng D, Woolard FX, Regentin R, Horning nol metabolism in Saccharomyces cerevisiae by ATP-independent acety- T, Tsuruta H, Melis DJ, Owens A, Fickes S, Diola D, Benjamin KR, Keasling lating acetaldehyde dehydrogenase. FEMS Yeast Res. 2016;16(2):fow006. JD, Leavell MD, McPhee DJ, Renninger NS, Newman JD, Paddon CJ. 46. Costenoble R, Picotti P, Reiter L, Stallmach R, Heinemann M, Sauer U, Production of amorphadiene in yeast, and its conversion to dihydroar- Aebersold R. Comprehensive quantitative analysis of central carbon temisinic acid, precursor to the antimalarial agent artemisinin. Proc Natl and amino-acid metabolism in Saccharomyces cerevisiae under multiple Acad Sci USA. 2012;109(3):E111–8. conditions by targeted proteomics. Mol Syst Biol. 2011;7:464. 43. de Jong BW, Shi S, Siewers V, Nielsen J. Improved production of fatty 47. Hackett SR, Zanotelli VR, Xu W, Goya J, Park JO, Perlman DH, Gibney PA, acid ethyl esters in Saccharomyces cerevisiae through up-regulation of Botstein D, Storey JD, Rabinowitz JD. Systems-level analysis of mecha- the ethanol degradation pathway and expression of the heterologous nisms regulating yeast metabolic flux. Science. 2016;354(6311):aaf2786. phosphoketolase pathway. Microb Cell Fact. 2014;13(1):39. Ready to submit your research ? Choose BMC and benefit from: fast, convenient online submission thorough peer review by experienced researchers in your ﬁeld rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Microbial Cell Factories – Springer Journals
Published: May 31, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera