Ann Microbiol (2016) 66:1049–1055 DOI 10.1007/s13213-015-1190-2

ORIGINAL ARTICLE

L-phenylacetylcarbinol production by yeast petite mutants

Mohsen Doostmohammadi1 & Mohammad Ali Asadollahi1 & Iraj Nahvi2 & Davoud Biria1 & Gholam Reza Ghezelbash3 & Maryam Kheyrandish1

Received: 19 September 2015 /Accepted: 28 December 2015 /Published online: 20 January 2016 # Springer-Verlag Berlin Heidelberg and the University of Milan 2016

Abstract The yeast Saccharomyces cerevisiae is able to Introduction biotransform into L-phenylacetylcarbinol (L-PAC), a key intermediate in the production of Ephedrine and can be extracted and isolated and pseudoephedrine, by the action of pyruvate decarobxylase from numerous plant species of the genus Ephedra. These (PDC) enzyme. This biotransformation can alternatively be alkaloids exhibit anti-asthmatic and decongestant properties performed by acetohydroxyacid synthase (AHAS) which is (Borchardt 2003). Moreover, their potential application in a mitochondrial enzyme. In the yeast petite mutants, AHAS obesity control has been reported (Astrup et al. 1992). Since accumulates in the cytosol. In the current study, wild-type extraction from plants involves tedious, costly, and time- yeast cells and yeast petite mutants were examined for consuming steps, attempts have been made to find alternative L-PAC biosynthesis. The results showed higher L-PAC titers methods for industrial production of these alkaloids. Chemical in the yeast petite mutants. In addition, the effect of cell synthetic routes which involve resolution of racemic mixtures immobilization and carbon source ( or molasses) on of alkaloids are also not appropriate for commercial produc- L-PAC production was investigated. It was found that cell tion (Abourashed et al. 2003). The current large-scale process immobilization enhances L-PAC formation. The highest for the production of ephedrine is via a combined fermentation L-PAC concentration (2.4 g/l) was obtained at 2 g/l of and chemical synthesis process (Fig. 1). In this process, benzaldehyde using the immobilized petite mutants grown benzaldehyde is first converted to L-phenylacetylcarbinol on molasses. (L-PAC) as the key intermediate in the production of ephed- rine and pseudoephedrine. Ephedrine is subsequently obtain- ed by reductive amination of the carbinol (Abourashed et al. Keywords L-Phenylacetylcarbinol . Acetohydroxyacid 2003;Doostmohammadietal.2015). synthase . Saccharomyces cerevisiae . Petite mutants . Pyruvate decarboxylase (PDC) is a homotetrameric Immobilization . Benzaldehyde enzyme that catalyzes the non-oxidative decarboxylation of thiamine pyrophosphate (TPP)-bound pyruvate to acetalde- hyde and carbon dioxide. As the second function, PDC can catalyze carboligation reaction of bound decarboxylated * Mohammad Ali Asadollahi pyruvate to yield acetoin via an aldol type condensation. [email protected] This ability is exploited for the commercial production of L-PAC using a combination of benzaldehyde with the acetal- dehyde obtained from pyruvate decarboxylation reaction 1 Department of Biotechnology, Faculty of Advanced Sciences and (Fig. 2; Agarwal et al. 2015). In parallel, alcohol dehydroge- Technologies, University of Isfahan, Isfahan 81746-73441, Iran nase (ADH) converts benzaldehyde to benzyl alcohol as a 2 Department of Biology, Faculty of Science, University of Isfahan, by-product (Neuberg and Hirsch 1921; Netrval and Vojtíšek Isfahan 81746-73441, Iran 1982; Tripathi et al. 1997). 3 Department of Biology, Faculty of Science, Shahid Chamran The efficiency of the process is limited by factors such as University, Ahvaz, Iran by-product formation, benzaldehyde toxicity toward cells, and 1050 Ann Microbiol (2016) 66:1049–1055

Fig. 1 Semi-synthetic process for large-scale production of Unreacted Benzaldehyde ephedrine a yeast fermentation for L-PAC the biotransformation of Ephedrine endogenous pyruvate and exogenously supplied benzaldehyde to L-PAC; b vessel for extraction of L-PAC from the Metal culture medium; c reductive catalyst amination of L-PAC using metal catalyst in the presence of H2 and H2 methylamine Methylamine

a) b) c) irreversible inactivation of the enzyme by acetaldehyde decreases significantly. This prevents import of many (Rosche et al. 2001). Many attempts have been made to preproteins into mitochondria. As such, cytosolic accumula- improve the efficiency of the fermentation process. Some tion of AHAS precursor in the yeast petite mutants has been researchers have employed the isolated PDC enzyme instead reported (Dasari and Kölling 2011). of the whole cell as a biocatalyst to prevent by-product In this study, we have compared the ability of wild-type formation (Shin and Rogers 1996; Gunawan et al. 2008). and petite mutants of the yeast S. cerevisiae for L-PAC Acetohydroxyacid synthase (AHAS) is an enzyme present production. Both free and immobilized yeast cells were in plants, bacteria, and fungi which catalyzes the first commit- employed. Molasses and glucose were used as carbon sources. ted step in the biosynthesis of branched-chain amino acids (Calvo et al. 1969). In contrast to PDC, AHAS is an enzyme with intrinsic carboligase activity. AHAS decarboxylates Materials and methods pyruvate and catalyzes the condensation of an active acetal- dehyde moiety derived from pyruvate with another molecule Materials of pyruvate to form acetolactate or with ketobutyrate to form acetohydroxybutyrate (Chipman et al. 1998). All solvents and chemicals used were of reagent grade unless The ability of AHAS to catalyze condensation of pyruvate otherwise indicated. Standard L-PAC was a gift from Embio with benzaldehyde and formation of L-PAC has been reported Limited (Mumbai, India). (Engel et al. 2003). In the yeast Saccharomyces cerevisiae, AHAS is a mitochondrial enzyme encoded by ILV2 (Falco Microorganism, culture conditions and cell et al. 1985). immobilization In the yeast petite mutants, generation of adenosine triphos- phate (ATP) by oxidative phosphorylation ceases and S. cerevisiae (CEN.PK113-7D) was grown on yeast peptone membrane potential across the inner mitochondrial membrane dextrose (YPD) agar plates. The medium used for the

Fig. 2 Mechanism of L-PAC formation by PDC enzyme Ann Microbiol (2016) 66:1049–1055 1051 production of L-PAC contained 30-g/l glucose and 6-g/l pep- carried out using standard curves generated after each analysis tone. For cultivation of encapsulated cells, CaCl2 at a concen- run (Doostmohammadi et al. 2015). tration of 5 g/l was added to glucose–peptone medium. The composition of the molasses medium was similar to glucose– Experimental methodology peptone medium except that glucose was replaced by 30-g/l molasses. For cell immobilization, 1000 ml of overnight cul- A two-level factorial design of experiments was conducted to ture at an optical density (OD ) of 18 was centrifuged 600 600 investigate the L-PAC production by both the wild-type and and after removing supernatant, the cells (about 3 g of wet petite cells at two concentrations of benzaldehyde and in cells) were added to 35 ml of 4 % (w/v) sterile sodium alginate immobilized and free cell systems. The rationale for selecting solution. This solution was added drop wise into a 2 % (w/v) the two-level factorial design was to have a reputable method CaCl solution under gentle stirring using a magnet stirrer. 2 to compare L-PAC production by the wild-type and the petite cells in various conditions. All of the experiments (Table 1) Induction of petites with UV light were run in duplicate. Analysis of variance (ANOVA) was the statistical method used in this research. The significance of the To generate petite mutants, cells from an overnight YPD culture data was judged by the p value being closer to zero (0.00). For (OD600 of 18) were diluted 1:50 in YPD and incubated at 30 °C a 95 % confidence level, the p value should be less than or for 3 h. Grown cells were exposed to UV light radiation equal to 0.05, indicating a statistically significant effect. The (CROSSLINKER CL-E508.G) with a 254-nm UV source at individual and interactive effects of the benzaldehyde concen- 2 an intensity of 10 J/cm h for 20 min. Cell suspension was then tration (1 and 2 g/l), state (free or immobilized cell), and strain cultivated on YPD medium containing triphenyltetrazolium (wild-type or petite mutant) on L-PAC production as response chloride. A color detection method was used to measure mito- variables were studied as well. chondrial dysfunction in the yeast cells. Wild-type yeasts reduce triphenyltetrazolium chloride to the insoluble red Formosan pig- ment, whereas strains lacking the ability to carry out respiration (mitochondrial function) turn white (Stotz and Linder 1990). Results and discussion

Comparison between wild-type cells and yeast petite Biotransformation using free and immobilized yeast cells mutants

Cells harvested from 1000 ml of overnight culture medium at PDC is the major enzyme responsible for the biosynthesis of an OD600 of 18 (about 3 g of wet cells) were added to 100 ml L-PAC. However, this enzyme suffers from some disadvan- of biotransformation medium (glucose or molasses medium) tages such as formation of benzyl alcohol as an unwanted by- and incubated for 1 h on a shaker at 30 °C and 200 rpm for product and sensitivity to benzaldehyde concentration and adaptation of cells to the medium. Benzaldehyde and acetal- temperature (Velmurugan et al. 1997; Engel et al. 2003). dehyde were each added at concentrations of 1 and 2 g/l and Alternatively, AHAS can be used to catalyze condensation flasks were incubated at 30 °C and 200 rpm for 28 h. of benzaldehyde with pyruvate leading to L-PAC formation. AHAS is a mitochondrial enzyme in the yeast S. cerevisiae Analysis of L-PAC concentrations (Falco et al. 1985), but in the yeast petite mutants, this enzyme accumulates in the cytosol (Dasari and Kölling 2011). Concentrations of L-PAC were determined by gas chromatog- Therefore, in petite mutants, two parallel pathways should raphy (GC). Samples were taken after 21 h and extracted with exist for the biosynthesis of L-PAC. We compared the ability dichloromethane. 0.2 ml of the supernatant was mixed with of yeast petite mutants and wild-type strain for L-PAC biosyn- 0.6 ml of dichloromethane and vortexed for 2 min. One μlof thesis. The comparison was made using free and immobilized the bottom organic layer was injected into a GC instrument cells. The results showed that petite mutants were superior to (Agilent 6890) equipped with a flame ionization detector wild-type cells for L-PAC formation (Table 1). For all cases, (FID). Analytes were separated on an HP-5 capillary column higher amounts of L-PAC were produced using petite mutants (30 m × 0.25 mm i.d., 0.25 μm film thicknesses) using helium as compared to the corresponding wild-type cells (Fig. 3). as carrier gas at a flow rate of 1.5 ml/min and a split ratio of L-PAC titers using immobilized petite mutants reached 20:1. The injector and detector temperatures were set at 200 2.4 g/l, whereas for immobilized wild-type cells, the maxi- and 250 °C, respectively. The initial oven temperature was mum L-PAC concentration obtained was 2 g/l. Considering 50 °C and was ramped to 150 °C at a rate of 10 oC/min, then the theoretical yield of L-PAC on substrate which is approxi- to 210 °C at 20 °C/min. The oven temperature was then main- mately 1.4 g L-PAC per g of benzaldehyde, immobilized pe- tained at 210 °C for 5 min. Quantification of compounds was tite mutants resulted in 87 % of the theoretical yield. 1052 Ann Microbiol (2016) 66:1049–1055

Table 1 The series of experiments and measured Runs State Cells Carbon Benzaldehyde L-PAC L-PAC concentrations source concentration concentration (g/l) (g/l)

1 Free Petite Glucose 2 1.30 2 Immobilized Wild type Glucose 2 1.80 3 Immobilized Wild type Glucose 1 0.85 4 Free Petite Molasses 2 1.58 5 Immobilized Petite Glucose 1 1.03 6 Free Wild type Glucose 1 0.60 7 Immobilized Petite Molasses 1 1.09 8 Free Wild type Molasses 1 0.65 9 Free Petite Glucose 1 0.73 10 Immobilized Petite Molasses 2 2.40 11 Free Wild type Molasses 2 1.30 12 Immobilized Wild type Molasses 1 0.90 13 Free Wild type Glucose 2 1.10 14 Free Petite Molasses 1 0.78 15 Immobilized Petite Glucose 2 2.10 16 Immobilized Wild type Molasses 2 2.00

The impact of benzaldehyde concentration between 0.4 and 1.7 g/l (Gupta et al. 1979). The highest L- PAC production yield was obtained at a concentration of Benzaldehyde is the essential substrate for L-PAC produc- 2 g/l benzaldehyde using the immobilized Candida utilis tion, but it also exhibits toxic effects on cell growth rate cells (Shin and Rogers 1995). and enzymatic functions (Long and Ward 1989). The effect In this study, the effect of two concentrations of benz- of benzaldehyde on L-PAC production by the yeast S. aldehyde (1 and 2 g/l) on L-PAC production yield was cerevisiae has been studied (Gupta et al. 1979;Agarwal investigated. As demonstrated in Table 2, there is a sig- et al. 1987). It has been reported that benzaldehyde con- nificant relationship between benzaldehyde concentration centration higher than 1.7 g/l reduces the biotransformation and L-PAC production yield. Regardless of conditions rate and when the residual benzaldehyde concentration de- (free or immobilized cells, wild-type or petite mutant clines below 0.4 g/l, benzyl alcohol production is predom- cells, glucose or molasses as carbon sources), a benzalde- inant over L-PAC (Rogers et al. 1997). As reported in pre- hyde concentration of 2 g/l resulted in maximum L-PAC vious studies, the optimum benzaldehyde concentration is titers (Table 1 and Fig. 3).

Fig. 3 Main effects of the studied State Cells factors: a effect of culturing state 1.6 (immobilized vs. free cells), b a) b) Cell type (mutant vs. wild-type 1.4 cells), c benzaldehyde concentration 1.2

1.0

0.8 Immobilized Free Wild Type Petite BA Con.

concentration (g/l) 1.6 c) 1.4

L-PAC L-PAC 1.2

1.0

0.8 1 2 Ann Microbiol (2016) 66:1049–1055 1053

Table 2 The analysis of variance of experiments Source DF Seq SS Adj SS Adj MS F p

Main effects 3 4.28972 4.28972 1.42991 88.16 0.000 State 1 1.06606 1.06606 1.06606 65.73 0.000 Cells 1 0.20476 0.20476 0.20476 12.62 0.007 BA conc. 1 3.01891 3.01891 3.01891 186.14 0.000 two-way interactions 3 0.25372 0.25372 0.08457 5.21 0.028 State*Cells 1 0.00681 0.00681 0.00681 0.42 0.535 State*BA conc. 1 0.22801 0.22801 0.22801 14.06 0.006 Cells*BA conc. 1 0.01891 0.01891 0.01891 1.17 0.312 three-way interactions 1 0.00076 0.00076 0.00076 0.05 0.834 State*Cells*BA conc. 1 0.00076 0.00076 0.00076 0.05 0.834 Residual error 8 0.12975 0.12975 0.01622 Pure error 8 0.12975 0.12975 0.01622 Total 15 4.67394

DF Degrees of freedom, Seq SS Sequential sums of squares, Adj SS-adjusted sums of squares, AdJ MS-adjusted mean of squares, F Fvalue,ppvalue, BA Benzaldehyde

Immobilized cells versus free cells interaction plots (Fig. 4) for state–cells and cells–benzalde- hyde concentration effects. However, the interaction of benz- It has been shown that immobilization of yeast cells could aldehyde concentration and the growth state was evaluated to protect cells against toxic effects of benzaldehyde by virtue be significant because the immobilized cells can tolerate a of divisional limitations and the toxic substrate gradients that higher concentration of benzaldehyde in the system. In fact, are established within the immobilizing matrix (Mahmoud et alginate protects cells from a high concentration of benzalde- al. 1990). Shin and Rogers (1995) observed increased hyde in the bulk aqueous phase. resistance to substrate using C. utilis cells immobilized in calcium alginate beads. Also, enhanced L-PAC production Checking the analysis assumptions has been reported upon immobilization of yeast cells (Mahmoud et al. 1990;ShinandRogers1995). One of the key assumptions for the statistical analysis of data Calcium alginate was used for immobilizing yeast cells in from experiments is that the data come from a normal distri- the current study. Cell immobilization reduced the toxic bution. The normality of the data can be checked by plotting a effects of benzaldehyde on yeast cells and, in most of the normal probability of the residuals (Fig. 5a). If the points on cases, immobilized cells produced higher amounts of L-PAC the plot fall fairly close to a straight line, then the data are (Table 1 and Fig. 3). For both wild-type cells and petite normally distributed (Antony 2003). It is evident from mutants, the highest L-PAC amounts were observed for the Fig. 5a that all the points fall fairly close to the straight line, immobilized cells. suggesting that data from the experiments come from a nor- mally distributed population. On the other hand, it is assumed Statistical analysis that the variances of different experimental runs are almost equal in the ANOVA calculations. This assumption can be Table 2 demonstrates the ANOVA for the experimental re- checked by the residuals versus fits plot, as can be seen in sults. The P value is the probability value used to determine Fig. 5b. In this figure, the distances of residuals from horizon- the effects in the model that are statistically significant. tal axes for all the obtained results were nearly the same. ANOVA (Table 2) suggests that all the studied factors are meaningful. The main effect of each studied factor has been discussed in the following sections. The interactive effects are Conclusion evaluated to be non-significant with the exception of the two- way interaction of the benzaldehyde concentration with the The ability of yeast petite mutants for the biosynthesis of growth state (immobilized or free cells). In fact, the influences L-PAC was demonstrated. The L-PAC titers were measured of the growth state and benzaldehyde concentration are the for the free and immobilized yeast petite mutants on glucose same on the L-PAC production by wild-type cells and petite and molasses as carbon sources and at different benzaldehyde mutants, and the petite mutants have a greater production at a concentrations and compared with those obtained using wild- similar condition. This can be observed as parallel lines in the type yeast cells. Regardless of conditions (free or immobilized 1054 Ann Microbiol (2016) 66:1049–1055

Fig. 4 Interactive effects of the Wild Type Petite 1 2 studied factors on L-PAC 2.0 State production: a culturing state vs. a) b) Immobilized cell type, b culturing state vs. State benzaldehyde concentration, c 1.5 Free cell type vs. benzaldehyde concentration 1.0

2.0 c) Cells Cells 1.5 Wild Type Petite

1.0

BA Con. cells, wild-type or petite mutant cells, glucose or molasses as L-PAC formation. The highest L-PAC titer (2.4 g/l) was carbon sources), yeast petite mutants resulted in higher L-PAC observed for the immobilized petite mutants on molasses as titers than those for wild-type yeast cells. Immobilized cells the carbon source at 2 g/l of benzaldehyde. This corresponds (petite and wild-type) were superior to the free cells in terms of to 87 % of the theoretical yield.

Fig. 5 The residual plots to Normal Probability Plot check the analysis assumptions: a a) (response is LPAC) normal probability plot and b 99 residual vs. fits

95 90

80 70 60 50 40 Percent 30 20

10 5

1 -0.2 -0.1 0.0 0.1 0.2 Residual

Versus Fits b) (response is LPAC) 0.15

0.10

0.05

0.00 Residual -0.05

-0.10

-0.15

0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 Fitted Value Ann Microbiol (2016) 66:1049–1055 1055

References Gunawan C, Breuer M, Hauer B, Rogers PL, Rosche B (2008) Improved (R)-phenylacetylcarbinol production with Candida utilis pyruvate decarboxylase at decreased organic to aqueous phase volume ratios. Abourashed EA, El-Alfy AT, Khan IA, Walker L (2003) Ephedra in Biotechnol Lett 30:281–286 – perspective- a current review. Phytother Res 17:703 712 Gupta K, Singh J, Sahni G, Dhawan S (1979) Production of phenyl acetyl Agarwal SC, Basu SK, Vora VC, Mason JR, Pirt SJ (1987) Studies on the carbinol by yeasts. Biotechnol Bioeng 21:1085–1089 production of L‐acetyl phenyl carbinol by yeast employing benzal- Long A, Ward O (1989) Biotransformation of benzaldehyde by dehyde as precursor. Biotechnol Bioeng 29:783–785 Saccharomyces cerevisiae: characterization of the fermentation Agarwal PK, Uppada V, Swaminathan AG, Noronha SB (2015) and toxicity effects of substrates and products. Biotechnol Bioeng Engineering of yeast pyruvate decarboxylase for enhanced selectiv- 34:933–941 ity towards carboligation. Bioresour Technol 192:90–96 Mahmoud WM, El‐Sayed AHM, Coughlin RW (1990) Production of L‐ Antony J (2003) Design of experiments for engineers and scientists. phenylacetyl carbinol by immobilized yeast cells: I. batch fermen- Butterworth-Heinemann, Burlington tation. Biotechnol Bioeng 36:47–54 Astrup A, Breum L, Toubro S, Hein P, Quaade F (1992) The effect and š safety of an ephedrine/caffeine compound compared to ephedrine, Netrval J, Vojtí ek V (1982) Production of phenylacetylcarbinol in vari- – caffeine and placebo in obese subjects on an energy restricted diet. a ous yeast species. Eur J Appl Mirobiol Biotechnol 16:35 38 double blind trial. Int J Obes Relat Metab Disord 16:269–277 Neuberg C, Hirsch J (1921) An enzyme which brings about union into – Borchardt JK (2003) Traditional Chinese drug therapy. Drug News carbon chains (Carboligase). Biochem Z 115:282 310 Perspect 16:698–702 Rogers PL, Shin HS, Wang B (1997) Biotransformation for L-ephedrine Calvo J, Freundlich M, Umbarger H (1969) Regulation of branched-chain production. Adv Biochem Eng Biotechnol 56:33–59 amino acid biosynthesis in Salmonella typhimurium: isolation of Rosche B, Sandford V, Breuer M, Hauer B, Rogers P (2001) regulatory mutants. J Bacteriol 97:1272–1282 Biotransformation of benzaldehyde into (R)-phenylacetylcarbinol Chipman D, Barak Z, Schloss JV (1998) Biosynthesis of 2-aceto-2- by filamentous fungi or their extracts. Appl Microbiol Biotechnol hydroxy acids: acetolactate synthases and acetohydroxyacid 57:309–315 synthases. Biochim Biophys Acta 1385:401–419 Shin HS, Rogers PL (1995) Biotransformation of benzaldehyde to L- Dasari S, Kölling R (2011) Cytosolic localization of acetohydroxyacid phenylacetylcarbinol, an intermediate in L-ephedrine production, synthase Ilv2 and its impact on diacetyl formation during beer fer- by immobilized Candida utilis. Appl Microbiol Biotechnol 44:7–14 mentation. Appl Environ Microbiol 77:727–731 Shin HS, Rogers PL (1996) Production of L-phenylacetylcarbinol (L- Doostmohammadi M, Esfandiari M, Asadollahi MA, Kamali M, Nahvi I PAC) from benzaldehyde using partially purified pyruvate decar- (2015) Biotransformation of benzaldehyde to L-phenylacetyl carbi- boxylase (PDC). Biotechnol Bioeng 49:52–62 Saccharomyces cerevisiae nol using immobilized cells of .Minerva Stotz A, Linder P (1990) The ADE2 gene from Saccharomyces – Biotechnol 27:43 49 cerevisiae: sequence and new vectors. Gene 95:91–98 Engel S, Vyazmensky M, Geresh S, Barak Z, Chipman D (2003) Tripathi CM, Agarwal SC, Basu SK (1997) Production of L- Acetohydroxyacid synthase: a new enzyme for chiral synthesis of phenylacetylcarbinol by fermentation. J Ferment Bioeng 84:487– ‐ – R phenylacetylcarbinol. Biotechnol Bioeng 83:833 840 492 Falco SC, Dumas KS, Livak KJ (1985) Nucleotide sequence of the yeast Velmurugan S, Lobo Z, Maitra PK (1997) Suppression of pdc2 regulating ILV2 gene which encodes acetolactate synthase. Nucleic Acids Res pyruvate decarboxylase synthesis in yeast. Genetics 145:587–594 13:4011–4027