Synergistic inhibition effect of 2-phenylethanol and ethanol on bioproduction of natural 2-phenylethanol by Saccharomyces cerevisiae and process enhancement

Journal of Bioscience and Bioengineering VOL. 112 No. 1, 26 – 31, 2011 www.elsevier.com/locate/jbiosc

Synergistic inhibition effect of 2-phenylethanol and ethanol on bioproduction of natural 2-phenylethanol by Saccharomyces cerevisiae and process enhancement Hang Wang, Qingfeng Dong, Ang Guan, Chun Meng, Xian'ai Shi, and Yanghao Guo⁎ College of Biological Science and Technology, Fuzhou University, Fuzhou, 350108, China Received 15 December 2010; accepted 11 March 2011 Available online 2 April 2011

Natural 2-phenylethanol (PEA) could be produced on a large scale by way of bioconversion with yeast from In this work the synergistic inhibition effect of the target product PEA and the byproduct ethanol on the bioconversion rate by Saccharomyces cerevisiae R-UV3 was systematically studied and a new kinetic model with an item representing the synergistic effect was proposed. Optimization strategies to repress the inhibition effect of PEA and ethanol were carried out in the mode of fed-batch culture with ISPR. The glucose concentration was regulated at the level of 0.2 ± 0.1 g/L by controlling a suitable respiratory quotient on line, which could limit the accumulation of the ethanol lower than 10 g/L. In the presence of resin FD0816 with a weight of 10% of the medium, PEA was removed from the broth and the overall PEA concentration and the space-time yield reached 13.7 g/L and 0.39 g L− 1 h− 1 respectively. The semi-continuous process with ISPR was performed, in which the replacement of the resin was operated repeatedly when the aqueous PEA was over 2.7 g/L and bioconversion continued until the bioactivity of the yeast cells declined, consequently achieving a final overall PEA concentration of 32.5 g/L and a space-time yield of 0.45 g L− 1 h− 1. © 2011, The Society for Biotechnology, Japan. All rights reserved.

L-phenylalanine.

[Key words: Natural 2-phenylethanol; Bioconversion; Synergistic inhibition; In situ product removal; Resin; Saccharomyces cerevisiae]

2-phenylethanol (PEA) is a higher aromatic alcohol with a roselike odour and becomes one of the most useful fragrance chemicals in perfume and cosmetic industry. PEA synthesized from petrochemicals is cheap but some of byproducts which are difficult to be separated impart green-gassy or metallic-chlorine off-odours in the final product. Accordingly, the natural PEA is preferred especially when it is used as food additive. The price of natural PEA is two orders of magnitude higher than its chemically produced counterpart (1). The demand for natural PEA would increase quickly with more and more attention to health and life quality. It is a prospective approach to convert natural L-phenylalanine (L-Phe) to natural PEA through Ehrlich pathway by yeasts such as Saccharomyces cerevisiae, Hansenula anomala and Kluyveromyces (1). There are obvious advantages for the bioconversion as follows: (i) the products could be marked with “natural” by European and US food agencies (2); (ii) the raw materials are cost-effective compared with the extract from roses; (iii) the production cycle is short; and (iv) it is easy to realize automatic control and industrial production. However, there is a notable disadvantage in the bioconversion that the final product concentration is too low to conduct a highly effective production due to serious product inhibition. PEA, an aromatic alcohol, can increase the fluidity of cell membrane (3,4) and then results in leakage of ions and a reduced uptake of amino acids and

⁎ Corresponding author. Tel./fax: + 86 591 22866379. E-mail address: [email protected] (Y. Guo).

glucose (5). Ethanol, as a byproduct of the fermentation by yeast, can produce synergistic inhibitory effect with PEA. It is necessary to study the product inhibition in order to enhance the bioconversion efficiency. Stark et al. (6) studied the relationship of specific cell growth rate with the PEA concentration and ethanol concentration respectively, based on the consideration that the PEA production rate could be associated with the specific growth rate of S. cerevisiae Giv 2009. However, to our knowledge, no experimental data about the synergistic inhibition effect of PEA and ethanol on the PEA production rate have been published so far. There are two ways to overcome the product inhibition in the bioconversion process: one is to mutate the biocatalyst to increase its tolerance to the product; the other is to remove the inhibitory product from the bioreaction system as soon as it is formed. Although considerable progress has been achieved by the former measure, enhancements of the production system still need to be applied (7). More and more attentions are paid to the technology of in situ product removal (ISPR). Many separation processes were integrated into the bioproduction of PEA, such as organic solvent extraction (8,9), ionic liquid extraction (10), solid–liquid sorption (11,12), supercritical CO2 extraction (13), immobilized solvent extraction (14,15) and pervaporation (16). The aqueous/organic two-phase bioconversion system succeeded in obtaining high overall PEA concentration. The reported highest overall PEA concentration was achieved in the presence of 48% oleic acid (8). However, organic solvents, e.g. oleic acid (8), might exhibit toxicity to cells, and emulsion might also be developed in a liquid–liquid system

1389-1723/$ - see front matter © 2011, The Society for Biotechnology, Japan. All rights reserved. doi:10.1016/j.jbiosc.2011.03.006

ENHANCED NATURAL 2-PHENYLETHANOL BIOPRODUCTION

when agitating and aerating (9). These drawbacks could be overcome by the solid–liquid system. A 3 L working volume bioreactor generated an overall PEA concentration of 13.7 g/L and a space-time yield of 0.38 g L− 1 h− 1 by using 500 g Hytrel® 8206 polymer as the sequestering phase (11). While using macroporous resin D101 in a solid–liquid system the overall PEA concentration and the space-time yield just reached 6.17 g/L and 0.26 g L− 1 h− 1 respectively in a shake flask (12). Obviously it is necessary to further improve the efficiency of the bioconversion system with ISPR. In this work, the influence of PEA and ethanol on the bioconversion of L-Phe to PEA by S. cerevisiae was systematically studied and then optimization strategies to enhance the PEA production rate and the bioconversion efficiency were developed. MATERIALS AND METHODS Chemicals and resin PEA of 99% purity was purchased from Acros Organics Co., Ltd. (Geel, Belgium). L-Phe of 99% purity as a standard for analysis was supplied by Alfa Aesar Co., Ltd. (USA), while L-Phe of 99% purity for the bioconversion was sourced from Maidan Biology Group Co., Ltd. of Fujian (Fujian, China). The resin FD0816 which surface area is 750–850 m2/g was supplied by Shanghai Huazhen Sci. and Tech. Co., Ltd. (Shanghai, China). The other biologicals and chemicals (HPLC or analytical grade) were obtained from the local suppliers. Strain and medium S. cerevisiae sp. strain R-UV3 is a preserved culture in our institute. The seed culture was inoculated into a 25 mL of pre-culturing medium that included (g/L): glucose 60, urea 2 and yeast extract 5 and incubated at 30°C and 200 rpm for 24 h. Then a 3 mL of cell suspension was transferred into a 250-mL shake flask containing 30 mL of Medium A which consisted of (g/L): glucose 60, KH2PO4 1, urea 1, yeast extract 1, MgSO4 0.3, and CaCl2 0.2. After incubating at 30°C and 200 rpm for 12 h, the cells were harvested by centrifugation at 6000g for 10 min and then were used as a inoculation for the bioconversion. Medium B for the bioconversion comprised (g/L): glucose 10, KH2PO4 1, urea 1, L-Phe 10, MgSO4 0.3, and CaCl2 0.2. Batch bioconversion 3 L Medium B with additional 50 g/L glucose was added into a 5-L fermenter (B. Braun B5, Germany) at 500 rpm. And then the fermenter was inoculated with wet cells until the cell density reached 6 gdcw/L. By controlling suitable air flow rate the dissolved oxygen concentration was maintained above 20% measured by a pO2 probe which was calibrated to 100% in the air beforehand. The pH level was controlled at 5.0 by addition of 2 mol/L NaOH solution and the temperature was set at 30°C. Fed-batch bioconversion The initial medium was 3 L Medium B with additional 2.5 g/L L-Phe in a 5-L fermenter. Carbon dioxide and oxygen in the exhaust gas were monitored online and then respiratory quotient (RQ) was calculated by a gas analyzer (Lokas LKM2000A, South Korea). The RQ value was controlled within the set range by regulating suitable feed rate of 50% (w/w) glucose solution. Other conditions were the same as batch bioconversion. Fed-batch bioconversion with ISPR 300 g of hydrated resin FD0816 was charged in a filter-cloth bag (200 meshes) fixed between baffles and stirring paddles in the 5-L fermenter. A fed-batch bioconversion mentioned above was performed. L-Phe was fed duly to keep its concentration above 17 g/L. PEA elution from the resin PEA adsorbed in the resin was eluted by 95% ethanol solution at flow rate of 1 bed volume per hour until PEA in the eluate was undetectable. The total PEA produced during the bioconversion with ISPR was equal to the sum of PEA in the resin and broth. Analytical procedures PEA and L-Phe were analyzed by HPLC equipped with a Waters 1525 Binary Pump (Waters, USA) and a Waters 2996 Photodiode Array Detector (Waters) and the analytical conditions were as follows: column, Hypersil ODS2 5 μm, 4.6 mm × 250 mm (Elite, China); column temperature, 35°C; solvent for elution, 50% (V/V) methanol; flow rate, 1 mL/min; detection wavelength, 210 nm. Ethanol was detected by GC (Shimadzu GC-2010, Japan) equipped by a SGE AC20 capillary column (SGE, Australia) and a FID detector and the analytical conditions were as follows: injector temperature, 190°C; column temperature, 120°C; detector temperature 210°C; internal standard, 1-butanol. Glucose was measured by an enzymatic kit (Beijing BHKT Clinical Reagent Co., Ltd., China). Dry weight of biomass was determined in duplicate samples of 2 mL each by centrifugation at 10,000 × g for 5 min. The wet pellets were washed twice and then dried for 12 h at 105°C followed by weighing.

27

decreased quickly although there were still sufficient carbon and nitrogen source (data not shown). The synergistic inhibition effect of PEA and ethanol with different concentrations on the bioconversion rate of PEA by strain R-UV3 was investigated systematically in this work. Pre-culture was performed until final PEA concentration reached 2.2 g/L. Then the cells were harvested and used as inoculum for the following bioconversion experiments on Medium B with various original concentrations of PEA and ethanol. Glucose solution (0.25 g, 50% w/w) was supplied every hour to inhibit ethanol production. Samples were withdrawn at hour 1 and hour 3 respectively to measure the L-Phe concentrations and cell concentrations for obtaining the average specific consumption rate of −1 −1 L-Phe, qPhe (g g h ). In view of sensitivity of measurement, qPhe was chosen as the index of bioconversion capability instead of specific PEA production rate. As shown in Fig. 1, higher concentration of ethanol evidently inhibited the bioconversion rate. When the ethanol concentration exceeded 10 g/L, the inhibitory effect on qPhe increased with the increasing ethanol concentration. It was noted that when the ethanol concentration was lower than 10 g/L, the inhibitory effect caused by ethanol could be neglected. High concentration of the target product PEA possessed more severe inhibition effect on qPhe than the byproduct ethanol. When the PEA concentration was beyond 1.6 g/L, qPhe decreased significantly under the condition of the same concentration of ethanol. The remaining bioconversion activity at the PEA concentration of 3.0 g/L was only 20–25% of the activity when the PEA concentration was below 0.8 g/L. The synergistic effect of PEA and ethanol made the product inhibition more serious. When PEA and ethanol concentrations were beyond 1.6 g/L and 20 g/L separately, notable synergistic inhibition to the bioconversion could be observed. For example, if 2.4 g/L PEA or 30 g/L ethanol existed solely in the medium, qPhe declined to 68.5% or 76.0% respectively compared with no inhibition. However, when they both existed, qPhe declined to 34.0% which was only 65.2% of the predicted value based on the hypothesis of no synergism. Seward et al. (5) also observed serious synergistic inhibition of PEA and ethanol to yeast growth. Ethanol could reduce the tolerance of the cells to PEA and thus reduce the final PEA concentration (14). Stark et al. (6) used a linear function to describe the relationship between the specific cell growth rate and the PEA or ethanol concentration. Obviously, the complicated synergistic effect shown in Fig. 1 might require a special model to describe. A new co-inhibition model was proposed in Table 1. In this model, a synergistic item representing the synergistic effect of PEA and ethanol was introduced,

0.1 0.09 0.08

q Phe (g g-1 h-1)

VOL. 112, 2011

0.07 0.06 0.05 0.04 0.03 0.02

RESULTS AND DISCUSSION

0.01 0

Synergistic inhibition effect of PEA and ethanol on the bioconversion Severe inhibition effects of PEA and ethanol on the bioconversion of L-Phe to PEA by S. cerevisiae R-UV3 were observed in the batch cultures. When the PEA and ethanol concentrations were respectively higher than 2.5 g/L and 20.0 g/L in the aqueous broth both PEA production rate and cell growth rate

0

0.5

1

1.5

2

2.5

3

PEA concentration (g/L) FIG. 1. The average specific consumption rates of L-Phe, qPhe, during 1–3 h in shake flasks at various initial exogenous PEA and ethanol concentrations. The co-inhibition model of Table 1 was represented by line. Exogenous ethanol concentrations were 0 g/L (diamonds), 10 g/L (squares), 20 g/L (triangles), and 30 g/L (crosses), respectively.

qPhe

WANG ET AL.

J. BIOSCI. BIOENG.,

TABLE 1. The co-inhibition model of PEA and ethanol. !n2 # 
 " h  n3 i cPEA n1 c cethanol = qm 1− ccPEA 1− ethanol m ⋅ cm m Phe 1− cm PEA ethanol cethanol PEA

−1 −1 qm h Phe = 0.0818 g g

cm PEA = 3.17 g/L n1 = 3.81 cm ethanol = 52.8 g/L n2 = 2.52 n3 = 1.37

and the parameters were estimated using MATLAB programs based on the least square method (Table 1). The average relative error of fitting was 4.48% (Fig. 1). Both the power exponent n1 and n2 were greater than 1, indicating that qPhe declined faster and faster with the increase of PEA or ethanol concentration. The power exponent n1 was greater m than n2 and cm PEA was less than cethanol, suggesting that the cells were more sensitive to PEA than ethanol. The power exponent n3 was greater than 1, which meant that the synergistic effect of PEA and ethanol would become more and more notable with the increasing concentration of PEA or ethanol. Therefore the production of byproduct ethanol had to be prevented as strictly as possible to enhance the bioconversion process. Optimization strategies for the bioconversion The inhibition of PEA and ethanol was the bottle neck for the bioconversion of L-Phe to PEA by S. cerevisiae R-UV3. Limiting ethanol accumulation and removing PEA in situ were the main optimization strategies to enhance the bioconversion. Strategy to limit ethanol accumulation Ethanol production could be restricted by making glucose concentration in the broth as low as possible. However, sufficient supply of glucose was important to achieve high PEA production rate since glucose was used as carbon and energy source. Thus a suitable feed rate of glucose had to be controlled. It is expensive to measure the glucose concentration on line as a feedback signal for controlling the feed rate of glucose, while there is time delay for control if the glucose concentration is measured off-line. Fortunately the respiratory quotient (RQ) could clearly indicate whether any glucose is transferred to ethanol. So a feedback control strategy based on RQ measured on line was proposed in this work. Three experiments were carried out respectively at various RQs to study the influence of RQ on PEA and ethanol productions. As can be seen in Fig. 2, there were significant differences in ethanol production at different RQ values controlled. The higher RQ was set, the more ethanol could be produced and the more glucose would be consumed.

At the highest RQ of 1.5 ± 0.1, a high ethanol production rate was maintained during the entire reaction process and the final ethanol concentration exceeded 10 g/L. However, at the lowest RQ of 0.7 ± 0.1, the ethanol concentration increased slowly to 3 g/L in the first 3 h and then declined gradually, since ethanol was consumed as carbon source due to lack of glucose. When RQ was controlled at 1.1 ± 0.1, the average ethanol production rate from hour 3 to hour 14 was only 0.19 g L− 1 h− 1 and the final ethanol concentration was below 6 g/L. Therefore the ideal RQ value was 1.1 ± 0.1, which was chosen in the following experiments. When RQ was controlled at 1.1 ± 0.1 the glucose concentration in the broth could maintain at 0.2 ± 0.1 g/L. It was suggested that RQ could be used as a sensitive index to control the glucose concentration in the broth and furthermore to restrict ethanol accumulation and eliminate the inhibitory effect of ethanol on yeast. Fed-batch cultures were performed by regulating RQ at 1.1 ± 0.1 to control suitable feed rate of 50% (w/w) glucose solution. The final PEA concentration and the PEA production rate achieved 4.38 g/L and 0.31 g L− 1 h− 1 respectively, which were 56.4% and 158% higher than those in the batch process (Table 2). Stark et al. (6) reported that both the final PEA concentration and the PEA production rate were greatly improved by regulating glucose feed rate based on the ethanol signal of the tail gas. It was suggested that the ethanol concentration being controlled below the threshold could not only improve the PEA production rate but also ameliorate the tolerance of the cells to PEA and accordingly increase the final PEA concentration. Strategies to remove PEA in situ It is important for ISPR technology to select a suitable separation phase which has high partition coefficient, good biocompatibility and quick productremoving rate. In our previous experiments, eight kinds of resins including ion exchange resins and polar and non-polar macroporous resins were screened (data not shown). The non-polar macroporous resin FD0816 was chosen for PEA removal in situ. At 30°C and pH 5.0, its partition coefficients toward PEA and L-Phe were respectively 30– 35 and 5.0–5.6 at the equilibrium concentrations of 3.0–4.0 g/L. A typical fed-batch process with ISPR was shown in Fig. 3. The final overall PEA concentration based on the total volume, including liquid phase and solid phase, was 13.7 g/L and the average space-time yield achieved 0.39 g L− 1 h− 1 in 35 h. The overall PEA concentration and the space-time yield with addition of resin were respectively 213% and 24.6% higher than those without resin, indicating that the bioproduction of PEA by S. cerevisiae R-UV3 was greatly enhanced when integrated with the process of separating PEA from the broth in situ by resin. Quick and continuous removal of PEA by resin effectively hindered the PEA accumulation in the aqueous phase and consequently decreased the inhibitory effect of PEA on cell bioactivity.

12

5

PEA concentration (g/L)

4.5 10

4 3.5

8

3 2.5

6

2 4

1.5 1

2

0.5 0

0

2

4

6

8

10

12

14

ethanol concentration (g/L)

28

0

Time (h) FIG. 2. The fed-batch bioconversions of L-Phe to PEA under various RQ values. Symbols: diamonds, RQ = 0.7 ± 0.1; squares, RQ = 1.1 ± 0.1; triangles, RQ = 1.5 ± 0.1; filled symbols, PEA; open symbols, ethanol.

VOL. 112, 2011

ENHANCED NATURAL 2-PHENYLETHANOL BIOPRODUCTION

29

TABLE 2. Summary of reactor performance in various operation modes. Operation mode

Strain

Batch Fed-batch

Fed-batch with ISPR

Semi-continuous process with ISPR

b c d e

Space–time yield a (g L− 1 h− 1)

YPEA/Glu (g/g)

YPEA/Phe (mol/mol )

2.8 2.35 4.38 3.8 4.5 13.7 12.6 10.2 13.7 20.38 10.9 c 21.1 d 32.5 e

0.12 0.048 b 0.31 0.18 0.065 0.39 0.14 0.33 0.38 0.43 0.618 c 0.616 d 0.45 e

0.049 0.081 b 0.094 – – 0.096 – – – – – – 0.101 e

0.81 0.91 0.80 0.83 1.1 0.80 0.91 0.82 0.93 0.91 – – 0.80 e

In situ extractant

Reference

– – – – – FD0816 resin Oleic acid PPG 1200 Hytrel® 8206 polymer Hytrel® 8206 polymer FD0816 resin

Present work 6 Present work 6 17 Present work 8 9 11 11 Present work

Calculated from t = 0 until the maximum product concentrations were reached. Calculated from Stark et al. (6). Data of the first cyclic run. Average data of the first two cyclic runs. Average data of the whole process.

cell and aqueous L-Phe concentrations (g/L)

PEA being enriched in the resin made a considerable improvement in the overall PEA concentration. In order to further enhance the space-time yield of the bioconversion of L-Phe to PEA, a semi-continuous culture with ISPR was carried out, in which PEA was continuously produced by the strain RUV3. In the previous experiments we found that R-UV3 still retained about 80% of glucose metabolism activity at the end of a bioconversion process, although the PEA production rate was severely inhibited by PEA of higher concentration (data not shown). According to the experimental results shown in Fig. 1, it was reasonable to renew the resin when the aqueous PEA concentration rose to 2.5–3.0 g/L and continued the bioconversion process till the bioactivity of the cells declined. A typical semi-continuous bioconversion process with three cyclic runs was shown in Fig. 4. The aqueous PEA concentration in the first cyclic run reached 2.7 g/L within 17 h. When the resin was renewed, two-thirds of broth was pumped out and equal volume of medium B with additional 20 g/L L-Phe was fed into the reactor in order to obtain a proper cell growth rate and a high PEA production rate. High cell growth rate and PEA production rate continued in the second run. The average space-time yield in the first two cyclic runs increased to 0.616 g L− 1 h− 1, the highest in the reported current literature to our knowledge, due to decrease in the

inhibitory effect of PEA compared with the fed-batch process with ISPR. The cell growth rate and the PEA production rate were halved in the third cyclic run, possibly due to decline of the bioactivity of the cells after a long-time run. Even so, the average space-time yield of the whole semi-continuous process was 0.45 g L− 1 h− 1 which was still 15.4% higher than that in the fed-batch process with ISPR and near to that reported by Gao et al. (Table 2). The final overall PEA concentration in the semi-continuous process based on the total reactor volume of broth and resin was 32.5 g/L. Different kinds of culture technologies would influence the kinetic characteristics of the bioprocess. A summary of the reported reactor performance in various operation modes was listed in Table 2. The lowest PEA production rate and bioconversion efficiency were observed in the batch culture mode because of synergistic inhibition effect of PEA and ethanol on the production of PEA. In fed-batch cultures, the ethanol accumulation could be prevented by control of glucose concentration so the overall PEA concentrations and the space-time yields were greatly higher than those in batch cultures. Application of the fed-batch with ISPR technology was the most important advance in enhancing the bioconversion process. The spacetime yield was enhanced 30% more than that without ISPR. The overall PEA concentrations increased more than three-fold since PEA was concentrated in the separation phase. A high overall product

50

5

45

4.5

40

4

35

3.5

30

3

25

2.5

20

2

15

1.5

10

1 0.5

5 0

0

5

10

15

20

25

30

35

aqueous PEA concentration (g/L)

a

S. cerevisiae R-UV3 S. cerevisiae Giv 2009 S. cerevisiae R-UV3 S. cerevisiae Giv 2009 S. cerevisiae Ye9-612 S. cerevisiae R-UV3 S. cerevisiae Giv 2009 K. marxianus CBS 600 K. marxianus CBS 600 K. marxianus CBS 600 S. cerevisiae R-UV3

Final overall PEA concentration (g/L)

0

Time (h) FIG. 3. The fed-batch bioconversion of L-Phe to PEA with 10% resin FD0816 at RQ = 1.1 ± 0.1. Symbols: squares, cell; triangles, aqueous L-Phe; diamonds, aqueous PEA.

WANG ET AL.

J. BIOSCI. BIOENG., 50

3.5

cell and aqueous L-Phe concentrations (g/L)

45

3

40 35

2.5

30

2

25 20

1.5

15

1

10 0.5

5 0

0

10

20

30

40

50

60

70

80

0

aqueous PEA concentration (g/L)

30

Time (h) FIG. 4. The semi-continuous bioconversion of L-Phe to PEA with 10% resin FD0816 at RQ = 1.1 ± 0.1. When aqueous PEA concentration accumulated to 2.7 g/L, a part of broth was discharged and equal volume of Medium B with additional 30 g/L L-Phe was supplied and all the resins were also renewed. Symbols: squares, cell; triangles, aqueous L-Phe; diamonds, aqueous PEA.

concentration meant not only high space efficiency of the bioconversion but also high efficiency of the product separation. In the semi-continuous process, the space-time yield could be further improved for a long-time run, in which the aqueous PEA concentration was limited to a proper level by renewing the resin. The replacement of the resin could be operated repeatedly until the bioconversion activity of the yeast cells declined. This technique showed a prospect of industrial scale production of PEA. The overall PEA concentration and the space-time yield obtained in this work were quite close to those reported by Gao et al. (11) in a solid–liquid two-phase partitioning bioreactor system, which were superior to those achieved by Stark et al. (8) and by Etschmann et al. (9) in an organic–aqueous two-phase system (Table 2). It seemed that better results could be achieved in solid–liquid system than in aqueous/organic system (Table 2). Resin FD0816 showed little effect on the tolerance of R-UV3 to PEA so that the final aqueous PEA concentration with resin hardly decreased compared with that without resin (Figs. 2 and 3). However oleic acid showed inhibitory effect on the bioconversion and made the final aqueous PEA concentration decrease from 3.8 g/L to 2.1 g/L (8). Moreover the solid medium is easier to be separated from the aqueous broth than the liquid one. YPEA/Phe (mol/mol), the molar yields of PEA based on L-Phe consumed, were quite consistent in different operation modes in this study. However, they were lower than others reported. It could be attributed to the difference of the strains. Since the glucose was restricted from being converted to ethanol in the fed-batch culture, YPEA/Glu (g/g), the yield of PEA based on glucose consumed, increased by 92% compared with that in the batch culture. The YPEA/Glu obtained in the fed-batch process with and without resin was close to each other indicating that the resin has little influence on glucose metabolism. The YPEA/Glu in the semi-continuous process was higher than that in any other process. A high YPEA/Glu was a benefit to decrease the cost of the desired product PEA. In conclusion, the critical strategies of enhancing the bioconversion are to limit PEA to a fit level by ISPR technique and to limit ethanol below 10 g/L by controlling glucose concentration. A semicontinuous technique with ISPR by renewing resin and controlling glucose feed rate via RQ demonstrated very promising results for PEA production by S. cerevisiae in this work. Nomenclature qPhe specific L-Phe consumption rate, g g− 1 h− 1 cPEA exogenous PEA concentration, g/L cethanol exogenous ethanol concentration, g/L

qm Phe cm Phe cm ethanol ni YPEA/Glu YPEA/Phe

maximum specific L-Phe consumption rate, g g− 1 h− 1 inhibitory threshold concentration of PEA, g/L inhibitory threshold concentration of ethanol, g/L power exponent yield of PEA based on glucose consumed, g/g molar yield of PEA based on L-Phe consumed, mol/mol

ACKNOWLEDGMENTS The authors gratefully acknowledge the financial support of Fujian Natural Science Foundation (No. 2008J0293), Fujian province, P.R. China. References 1. Etschmann, M. M., Bluemke, W., Sell, D., and Schrader, J.: Biotechnological production of 2-phenylethanol, Appl. Microbiol. Biotechnol., 59, 1–8 (2002). 2. Krings, U. and Berger, R. G.: Biotechnological production of flavours and fragrances, Appl. Microbiol. Biotechnol., 49, 1–8 (1998). 3. Ingram, L. O. and Buttke, T. M.: Effects of alcohols on micro-organisms, Adv. Microb. Physiol., 25, 253–300 (1984). 4. Jordi, W., Nibbeling, R., and de Kruijff, B.: Phenethyl alcohol disorders phospholipid acyl chains and promotes translocation of the mitochondrial precursor protein apocytochrome c across a lipid bilayer, FEBS Lett., 261, 55–58 (1990). 5. Seward, R., Willetts, J., Dinsdale, M., and Lloyd, D.: The effects of ethanol, hexan-1-ol, and 2-phenylethanol on cider yeast growth, viability, and energy status: synergistic inhibition, J. Inst. Brew., 102, 439–443 (1996). 6. Stark, D., Zala, D., Münch, T., Sonnleitner, B., Marison, I. W., and von Stockar, U.: Inhibition aspects of the bioconversion of L-phenylalanine to 2-phenylethanol by Saccharomyces cerevisiae, Enzyme Microb. Technol., 32, 212–223 (2003). 7. Stark, D. and von Stockar, U.: In situ product removal (ISPR) in whole cell biotechnology during the last twenty years, Adv. Biochem. Eng. Biotechnol., 80, 149–175 (2003). 8. Stark, D., Munch, T., Sonnleitner, B., Marison, I. W., and von Stockar, U.: Extractive bioconversion of 2-phenylethanol from L-phenylalanine by Saccharomyces cerevisiae, Biotechnol. Prog., 18, 514–523 (2002). 9. Etschmann, M. M. and Schrader, J.: An aqueous-organic two-phase bioprocess for efficient production of the natural aroma chemicals 2-phenylethanol and 2phenylethylacetate with yeast, Appl. Microbiol. Biotechnol., 71, 440–443 (2006). 10. Sendovski, M., Nir, N., and Fishman, A.: Bioproduction of 2-phenylethanol in a biphasic ionic liquid aqueous system, J. Agr. Food Chem., 58, 2260–2265 (2010). 11. Gao, F. and Daugulis, A. J.: Bioproduction of the aroma compound 2-phenylethanol in a solid–liquid two-phase partitioning bioreactor system by Kluyveromyces marxianus, Biotechnol. Bioeng., 104, 332–339 (2009). 12. Mei, J. F., Min, H., and Lü, Z. M.: Enhanced biotransformation of L-phenylalanine to 2-phenylethanol using an in situ product adsorption technique, Process Biochem., 44, 886–890 (2009). 13. Fabre, C. E., Blanc, P. J., Marty, A., Goma, G., Souchon, I., and Voilley, A.: Extraction of 2-phenylethyl alcohol: by techniques such as adsorption, inclusion, supercritical CO2, liquid–liquid and membrane separations, Perfum. Flavor., 21, 27–40 (1996).

VOL. 112, 2011 14. Serp, D., von Stockar, U., and Marison, I. W.: Enhancement of 2-phenylethanol productivity by Saccharomyces cerevisiae in two-phase fed-batch fermentations using solvent immobilization, Biotechnol. Bioeng., 82, 103–110 (2003). 15. Stark, D., Kornmann, H., Munch, T., Sonnleitner, B., Marison, I. W., and von Stockar, U.: Novel type of in situ extraction: use of solvent containing microcapsules for the bioconversion of 2-phenylethanol from L-phenylalanine by Saccharomyces cerevisiae, Biotechnol. Bioeng., 83, 376–385 (2003).

ENHANCED NATURAL 2-PHENYLETHANOL BIOPRODUCTION

31

16. Etschmann, M. M., Sell, D., and Schrader, J.: Production of 2-phenylethanol and 2phenylethylacetate from L-phenylalanine by coupling whole-cell biocatalysis with organophilic pervaporation, Biotechnol. Bioeng., 92, 624–634 (2005). 17. Eshkol, N., Sendovski, M., Bahalul, M., Katz-Ezov, T., Kashi, Y., and Fishman, A.: Production of 2-phenylethanol from L-phenylalanine by a stress tolerant Saccharomyces cerevisiae strain, J. Appl. Microbiol., 106, 534–542 (2009).