Author's personal copy Appl Microbiol Biotechnol (2015) 99:1131–1144 DOI 10.1007/s00253-014-6098-4
BIOTECHNOLOGICAL PRODUCTS AND PROCESS ENGINEERING
Studies on the mechanism of synthesis of ethyl acetate in Kluyveromyces marxianus DSM 5422 Christian Löser & Thanet Urit & Peter Keil & Thomas Bley
Received: 10 July 2014 / Revised: 10 September 2014 / Accepted: 15 September 2014 / Published online: 9 December 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Kluyveromyces marxianus converts whey-borne sugar into ethyl acetate, an environmentally friendly solvent with many applications. K. marxianus DSM 5422 presumably synthesizes ethyl acetate from acetyl-SCoA. Iron limitation as a trigger for this synthesis is explained by a diminished aconitase and succinate dehydrogenase activity (both enzymes depend on iron) causing diversion of acetyl-SCoA from the tricarboxic acid cycle to ester synthesis. Copper limitation as another trigger for ester synthesis in this yeast refers to involvement of the electron transport chain (all ETC complexes depend on iron and complex IV requires copper). This hypothesis was checked by using several ETC inhibitors. Malonate was ineffective but carboxin partially inhibited complex II and initiated ester synthesis. Antimycin A and cyanide as complexes III and IV inhibitors initiated ester synthesis only at moderate levels while higher concentrations disrupted all respiration and caused ethanol formation. A restricted supply of oxygen (the terminal electron acceptor) also initiated some ester synthesis but primarily forced ethanol production. A switch from aerobic to anaerobic conditions nearly stopped ester synthesis and induced ethanol formation. Ironlimited ester formation was compared with anaerobic ethanol production; the ester yield was lower than the ethanol yield but a higher market price, a reduced number of process stages, a faster process, and decreased expenses for product recovery by stripping favor biotechnological ester production.
C. Löser (*) : T. Urit : P. Keil : T. Bley Institute of Food Technology and Bioprocess Engineering, TU Dresden, 01062 Dresden, Germany e-mail:
[email protected] T. Urit Department of Biology and Biotechnology, Faculty of Science and Technology, Nakhon Sawan Rajabhat University, Nakhon Sawan 60000, Thailand
Keywords Kluyveromyces marxianus . Ethyl acetate . Oxygen . Carboxin . Antimycin A . Cyanide
Introduction Ethyl acetate is a microbially degradable and thus environmentally friendly solvent with an annual world production of 1.7 million tons (Posada et al. 2013). Today, ethyl acetate is exclusively produced from fossil fuels in chemical processes (Löser et al. 2014). Microbial synthesis from renewable raw materials could be an alternative. Many yeasts synthesize ethyl acetate (Westall 1998), but the most potent producers are Candida utilis (Thomas and Dawson 1978; Armstrong et al. 1984a, b; Armstrong and Yamazaki 1984; Corzo et al. 1995), Pichia anomala (Davies et al. 1951; Bol et al. 1987; Rojas et al. 2001; Fredlund et al. 2004), and Kluyveromyces marxianus (Willetts 1989; Kallel-Mhiri et al. 1993; KallelMhiri and Miclo 1993; Löser et al. 2012, 2013, 2014; Urit et al. 2011, 2012, 2013a, b). K. marxianus is of special interest due to its fast growth and the ability to consume whey-borne lactose (Urit et al. 2013b; Löser et al. 2014). K. marxianus is a Crabtree-negative yeast (Fonseca et al. 2007, 2008) which allows to control its metabolism by the amount of supplied oxygen. In yeasts, three pathways for synthesis of ethyl acetate have been described (Park et al. 2009; Löser et al. 2014): esterification of acetate with ethanol, the hemiacetal reaction (acetaldehyde and ethanol spontaneously form a hemiacetal which is enzymatically oxidized), and alcoholysis of acetyl-SCoA with ethanol. Synthesis of ethyl acetate as a bulk product (also referred to as bulky synthesis) via esterification is very unlikely since esterhydrolyzing esterase activities dominate over the estersynthesizing activities (Löser et al. 2014). The hemiacetal reaction as found in C. utilis (Kusano et al. 1999)
Author's personal copy 1132
is characterized by ethanol as a substrate (directly applied or formed from sugar (Löser et al. 2014)) and by acetaldehyde as a side-product (Armstrong et al. 1984b). K. marxianus DSM 5422 formed ethyl acetate from lactose but hardly from ethanol, and acetaldehyde was never detected as a by-product (Urit et al. 2012; Löser et al. 2013), i.e., this yeast strain presumably produces ethyl acetate via alcoholysis. Alcohol acetyltransferase as the catalyzing enzyme is associated with intracellular lipid particles (Verstrepen et al. 2004) which are located in the cytosol (Zweytick et al. 2000). The level of iron is a crucial factor for synthesis of ethyl acetate; much ester was formed at iron limitation in C. utilis and K. marxianus while supplementing the medium with iron always repressed ester formation (Thomas and Dawson 1978; Armstrong and Yamazaki 1984; Armstrong et al. 1984a, b; Willetts 1989; Kallel-Mhiri et al. 1993). A deficit of iron as a trigger for ester synthesis has also been confirmed for K. marxianus DSM 5422 (Urit et al. 2011, 2012; Löser et al. 2012, 2013). Thomas and Dawson (1978) speculated for C. utilis that iron limitation reduces the flux of acetyl-SCoA into the tricarboxic acid (TCA) cycle due to a diminished aconitase and succinate dehydrogenase activity (both enzymes depend on iron) and that accumulated acetyl-SCoA lastly reacts with ethanol to form ethyl acetate. At that time, it was still unknown that C. utilis forms ethyl acetate mainly from acetaldehyde rather than acetyl-SCoA (Kusano et al. 1999); i.e., ester synthesis in this yeast occurs by conversion of pyruvate to acetaldehyde due to a reduced flux of pyruvate into the mitochondria. Metabolite-profiling studies confirmed the postulated reduction of the metabolic flux through the TCA cycle at iron limitation for K. marxianus DSM 5422 (Löser et al. 2012). Copper limitation as another trigger for this route of ester synthesis refers to involvement of the electron transport chain (ETC) in this process (Urit et al. 2012). Both iron and copper are important for the function of the electron-transferring proteins (Levi and Rovida 2009; Cuillel 2009). A lack of these metals should make the ETC less efficient and slow down oxidation of reduced nicotinamide adenine dinucleotide (NADH) to oxidized nicotinamide adenine dinucleotide (NAD+); the declining level of mitochondrial NAD+ could lastly limit the flux of acetyl-SCoA into the TCA cycle and divert this compound to ester synthesis (Fig. 1). Thomas and Dawson (1978) already referred to involvement of the ETC in the event of ester synthesis since iron limitation reduced the available energy in C. utilis. The effect of glycerol, iron, or copper limitation on the oxygen-related growth yield of C. utilis (YX/O2 =1.49, 0.94, or 0.49 g/g) and on the mitochondrial P:O ratio (P:O=3, 2, or 1) under these conditions demonstrate the impact of lacking iron or copper on the respiratory efficiency (Alexander and Jeffries 1990). Bulky synthesis of ethyl acetate from sugar is, in fact, an aerobic process as becoming visible from an overall reaction
Appl Microbiol Biotechnol (2015) 99:1131–1144
equation for a metabolism solely aimed at conversion of sugar into ethyl acetate: C6H12O6 + O2 → CH3-CO-O-CH2-CH3 + 2 CO2 + 2 H2O (Löser et al. 2014). Much ester was formed at aerobic conditions while ester production failed without oxygen (Davies et al. 1951; Tabachnick and Joslyn 1953; Yong et al. 1981; Willetts 1989; Kallel-Mhiri and Miclo 1993). Oxygen as the terminal electron acceptor is expected to influence the formation of ethyl acetate (Fig. 1). The effect of oxygen on synthesis of ethyl acetate has been repeatedly studied (Gray 1949; Armstrong et al. 1984a; Armstrong and Yamazaki 1984; Bol et al. 1987; Kallel-Mhiri et al. 1993; Rojas et al. 2001; Fredlund et al. 2004), but the published data are inconsistent and in part unreliable (for details, see (Löser et al. 2014)). The effect of oxygen on synthesis of ethyl acetate therefore requires re-examination under well-defined conditions. The electron transport is also affected by specific ETC inhibitors (Alexander and Jeffries 1990; Wallace and Starkov 2000) which thus represent valuable tools for studying the interrelation between ester synthesis and ETC activity. Most research on the ETC in yeasts has been done on Saccharomyces cerevisiae, some is known for Kluyveromyces lactis (Overkamp et al. 2002; Tarrío et al. 2006; Saliola et al. 2008), but the knowledge for K. marxianus in this relation is poor (Lertwattanasakul et al. 2009). Figure 1 depicts the proposed mechanism of ester synthesis in K. marxianus DSM 5422 including sources and sinks of reduction equivalents. Lactose is taken up and hydrolyzed to form glucose and galactose with subsequent epimerization of the latter. Glucose is catabolized to pyruvate which passes the mitochondrial membrane and reacts to acetyl-SCoA which then enters the TCA cycle. Three TCAcycle reactions generate NADH being fed to the ETC. The ETC consists of a sequence of at most four multienzyme complexes, denoted as complex I to IV. Complex I is absent in K. marxianus (Büschges et al. 1994; Veiga et al. 2003), but external and internal NADH dehydrogenases (Nde and Ndi) exist (Lertwattanasakul et al. 2009) which have mainly been studied in S. cerevisiae (Kerscher 2000; Bakker et al. 2001; Murray et al. 2011) and K. lactis (Overkamp et al. 2002; Tarrío et al. 2006). Ndi assimilates mitochondrial NADH while Nde uses cytosolic NADH. Both dehydrogenases oxidize NADH, transfer electrons to ubiquinone (Q), and form ubiquinol (QH2). The Q/QH2 pool in the hydrophobic center of the inner mitochondrial membrane works as a redox buffer (Fig. 1; Lenaz et al. 2007). Complex II represents succinate dehydrogenase (Sdh), oxidizes succinate to fumarate, and supplies electrons to ubiquinone (Lancaster 2002). Sdh is inhibited by malonate and carboxin. Complex III transfers electrons from QH2 to cytochrome c (Cyt c) while complex IV transmits electrons from Cyt c to oxygen (Overkamp et al. 2002; Lertwattanasakul et al. 2009). Complex III is inhibited by antimycin A and complex IV is blocked by cyanide. An alternative oxidase which mediates
Author's personal copy Appl Microbiol Biotechnol (2015) 99:1131–1144
1133
Cultivation medium NADH
2 NADH
H2O
Lactose Hydrolysis
Glucose Glycolysis
Pyruvate
Acetaldehyde
Pdc CO2
Galacose Epimerization
Aldh
Acetate
Cytosol
Acetyl-SCoA H+
II Sdh
NADH
H+
IV e– Ha CuA Ha3 CuB
H2O
Antimycin A
NAD+
Fumarate
+ ½ O2 H
Cyanide
H+
Pyruvate
Malate
Pdh NADH
Malonate
Ethyl acetate
Cyt c
Cyt b
H+
Aat
ATP synthase
Shuttle systems
H+
Carboxin
Ndi
NADH
Acs
NADH NAD+
Nde e– Q/QH e– III e– 2 Cyt c1 ee–
Ethanol
Adh
Succinate
Oxaloacetate
CO2
Acetyl-SCoA
H2O
NADH
Citrate
Succinyl-SCoA
Aco
CO2
Isocitrate
2-Oxoglutarate CO2
Mitochondrial matrix Inner mitochondrial membrane Intermitochondrial space Outer mitochondrial membrane Cell wall
Fig. 1 Proposed mechanism for synthesis of ethyl acetate as a bulk product from lactose by K. marxianus DSM 5422 at aerobic conditions including sources and sinks of reduction equivalents, here typified by NADH (NAD+ and HSCoA not pictured); an important source of NADH is the TCA cycle, and most NADH is oxidized in the electron transport chain (organization of the ETC adapted from Lertwattanasakul et al.
(2009)). Aat alcohol acetyltransferase, Aco aconitase, Acs acetyl-SCoA synthase, Adh alcohol dehydrogenase, Aldh aldehyde dehydrogenase, Cyt c cytochrome c, Nde external NADH dehydrogenase, Ndi internal NADH dehydrogenase, Pdc pyruvate decarboxylase, Pdh pyruvate dehydrogenase, Q/QH2 ubiquinone/ubiquinol pool, Sdh succinate dehydrogenase
cyanide-insensitive respiration as known for many yeasts is absent in K. marxianus (Veiga et al. 2003). Complexes I (not existing in K. marxianus), III, and IV pump protons from the mitochondrial matrix to the intramembrane space; the generated proton gradient across the inner membrane is utilized by adenosine triphosphate (ATP) synthase for producing ATP, accompanied by remigration of protons (Fig. 1; Wallace and Starkov 2000; Lancaster 2002). Neither Ndi and Nde nor Shd translocate protons in this way and, hence, do not directly contribute to ATP production (Kerscher 2000; Lancaster 2002). Shuttle systems transfer reduction equivalents across the mitochondrial membrane for maintaining the redox balance between cytosol and mitochondrial matrix. Several shuttles have been described for S. cerevisiae (Bakker et al. 2001). In K. marxianus, the ethanol-acetaldehyde shuttle is of importance (Lertwattanasakul et al. 2009).
A deficit of iron or copper could lower the ETC efficiency, reduce the NADH oxidation, and thus decrease the flux of acetyl-SCoA in the TCA cycle; excess acetyl-SCoA is then transported to the cytosol (e.g., via the citrate-malate shuttle) where it forms ethyl acetate (Löser et al. 2012). Cytosolic acetate could also contribute some to ester synthesis. The objective of this study was to analyze the effect of oxygen as the terminal electron acceptor on ester synthesis by K. marxianus DSM 5422 under well-defined conditions; a limited availability of oxygen is expected to reduce NADH oxidation and to lower the mitochondrial NAD+ level which could divert acetyl-SCoA from the TCA cycle to ester synthesis. The supply of oxygen was restricted or was completely disrupted. Absent oxygen favors ethanol synthesis as another product of microbial whey utilization. The obtained data were used for some comparative economic considerations on wheybased ester or ethanol production. These studies were
Author's personal copy 1134
complemented by a second series of experiments for exploring the influence of ETC activity on synthesis of ethyl acetate using specific ETC inhibitors.
Materials and methods Microorganism and culture media K. marxianus DSM 5422 from the Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (Braunschweig, Germany) was maintained with the Cryoinstant preservation system (Lomb Scientific Pty. Ltd., Austria), cultivated on yeast-glucose-chloramphenicol agar (Roth GmbH, Germany) for 2 days at 32 °C, and then used as an inoculum. DW basic medium was prepared from concentrated and partially demineralized sweet whey permeate (processed by Sachsenmilch Leppersdorf GmbH, Germany) in 1-L batches as formerly described (Urit et al. 2012). This medium holds 10 g/L (NH4)2SO4 to compensate a deficit of nitrogen and sulfur. DW basic medium contains 78 g/L sugar, various inorganic ions (Urit et al. 2011), and trace metals (Urit et al. 2012). DW basic medium was in part supplemented with normative trace-element solution (preparation described in (Urit et al. 2012)). Media for shake-flask experiments were based on highly diluted and buffered DW basic medium as already used in other bottle tests (Löser et al. 2012; Urit et al. 2013a, b). This medium is received by mixing 50 mL DW basic medium, 500 mL 50 mM K-phosphate buffer pH 6, 450 mL water, and 0.1 mL normative trace-element solution (all of them autoclaved separately) resulting in pH 6, 3.9 g/L sugars, and trace metals in excess. Shake-flask cultivation These experiments were conducted in sealed bottles for precise quantification of the highly volatile ester. A formerly developed cultivation system (Löser et al. 2011) was not used since pure oxygen as headspace gas proved to be inhibitory for the studied strain (Urit et al. 2013a). The cultivation occurred in 1000-mL Schott bottles fitted with perforated screw cap and septum (3 mm silicone, inside coated with PTFE for preventing losses of volatiles). This construction allows sampling of the gas and liquid phase; it is tight for ethyl acetate, and the headspace delivers enough oxygen for the microbial process; the applicability of this cultivation system has already been discussed thoroughly elsewhere (Urit et al. 2013a). Pre-cultures were prepared as follows: A bottle was filled with 50 mL highly diluted and buffered DW basic medium and then inoculated with 50 μL cell suspension obtained by suspending one loop biomass from a plate culture in 1 mL PBS. A 0.9×50-mm injection needle with a syringe filter
Appl Microbiol Biotechnol (2015) 99:1131–1144
(cellulose acetate membrane, 0.45-μm pore size) was pierced through the septum for pressure compensation. The bottle was cultivated overnight for 14 h on a shaker at 32 °C and 250 rpm, and used for inoculating main cultures. Main cultures were prepared in the same manner with the modification that fresh medium was mixed with pre-culture in such a way that the resulting volume was 50 mL and the OD600 amounted to 0.25. The main cultures contained a specific ETC inhibitor which was added in various ways. Malonate and KCN were introduced with fresh medium (malonate-containing stock medium required pH correction to 6 by 2 M KOH), while carboxin (5,6-dihydro-2-methyl-1,4oxathiin-3-carboxanilide; C12H13NO2S; 235.3 g/mol; solubility in water 0.147 g/L at 20 °C (www.capl.sci.eg)) and antimycin A (C28H40O9N2; 548.6 g/mol; nearly insoluble in water (Strong 1958)) were supplied as iso-propanolic stock solutions. Every experiment comprised several bottles with various inhibitor concentrations. All bottles were prepared in duplicate; one was used for liquid-phase sampling and the other for gas-phase sampling. This doubling avoided that liquid sampling interferes with ester analysis. The bottles were shaken for 24 h at 250 rpm and 32 °C. The sampling occurred hourly over a period of at least 10 h and then again at the end of cultivation. Gas and liquid samples (0.4 and 1.5 mL) were taken by syringes pierced through the septum. The OD was measured at 600 nm with a Beckman DU 520 photometer (after pre-dilution to OD
