Comparative studies on different enzyme preparations for (R)-phenylacetylcarbinol production

Gernalia Satianegara, BSc (Hons)

March, 2006

A thesis submitted for the degree of Doctor of Philosophy

School of Biotechnology and Biomolecular Sciences,

Faculty of Science, UNSW, Sydney, Australia

‘tuk yang tersayang: papa dan mama

A trip should not ride on a single anchor, nor life on a single hope. -Epictetus-

Acknowledgements

First of all, a huge thanks to my supervisor, Emeritus Professor Peter Rogers for giving me the opportunity to join the research group. In particular, I also deeply appreciate the help of my co-supervisor Dr. Bettina Rosche for a good start in my early studies. I am forever indebted for their invaluable vision, guidance, support, thoughtfulness and encouragement. This thesis could not have come together in the way it had without the assistance, experience and advice from them.

Thanks also go to the numerous staff members in BABS who have assisted me in many ways. Special thanks go to: Dr Martin Zarka for his generous concession on the use of HPLC, Dr Sohail Siddiqui for lending his expertise in chemical modification, Dr. Russell Cail and Mr Malcolm Noble for making it possible to use equipments vital to this study. I would like to thank all the members of PAC and ethanol groups, past and present, who have made my time here memorable. I have been very privileged to know of these amazing people and I am very grateful for their friendships. In particular I would like to thank: Allen, Noppol, Onn, Malikka and Ronachai for the friday-night’s stuffs and ski trips; Cindy and Eny for all the “Indo” news and jokes; Jae and Nick for sharing many things, organizing picnic, they had also made my early research time entertaining and fun; Charles Svenson for the wine “knowledge” and for giving me opportunity to teach; Michael, André, Richard and many others in BABS (too many to list!) for being my friends, their laughter has kept me sane here.

A special thank you goes to my parents and I thank them wholeheartedly for their unfailing support through every step of my life. I would also like to thank my brother Kris for his quirky ways to make life entertaining. Lastly but not the least, an enormous thank to my best friend and boyfriend, Allen. No words can describe how much your love, laughter and support meant to me.

i Abstract

The present study is part of a project to develop a high productivity enzymatic process for (R)-phenylacetylcarbinol (PAC), a precursor for the pharmaceuticals ephedrine and pseudoephedrine, with recent interest for a low cost and more stable biocatalyst pyruvate decarboxylase (PDC) preparation.

PDC initially added in the form of Candida utilis cells, viz. whole cell PDC, showed higher stability towards substrate benzaldehyde and temperature in comparison to the partially purified preparation in an aqueous/benzaldehyde emulsion system. Increasing the temperature from 4° to 21°C for PAC production with whole cell PDC resulted in similar final PAC levels of 39 and 43 g/L (258 and 289 mM) respectively from initial 300 mM benzaldehyde and 364 mM pyruvate. However, the overall volumetric productivity was enhanced by 2.8-fold. Enantiomeric excess values of 98 and 94% for R-PAC were obtained at 4° and 21°C respectively and benzylalcohol (a potential by-product from benzaldehyde) was not formed.

The potential of whole cell PDC was also evident in an aqueous/octanol- benzaldehyde emulsion system at 21°C with a 3-fold higher specific production compared to partially purified PDC. At 2.5 U/mL, PAC levels of 104 g/L in the organic phase and 16 g/L in the aqueous phase (60 g/L total reaction volume, 15 h) and a 99.1% enantiomeric excess for R-PAC were obtained with whole cell PDC.

The study of cell membrane components provided a better understanding for the enhanced performance of whole cell PDC in comparison to partially purified PDC. It was apparent that surfactants, both biologically-occurring (e.g. phosphatidylcholine) and synthetically manufactured (e.g. bis(2-ethyl-1-hexyl)sulfosuccinate (AOT)), enhanced PDC stability and/or PAC production in the aqueous/octanol-benzaldehyde biotransformation system with the partially purified enzyme. Addition of 50 mM AOT

ii to the biotransformation with partially purified PDC enhanced the enzyme half-life by 13-fold (19 h) and increased specific PAC production by 2-fold (36 mg/U).

Chemical modification studies targeting the amino and carboxyl groups were carried out to achieve increased stability of partially purified PDC. However these were not successful and future work could be directed at PDC protein engineering as well as optimization and scale up of the two-phase process using whole cell PDC.

iii Publications

Manuscripts

Satianegara, G., M. Breuer, B. Hauer, P.L. Rogers, B. Rosche (2006) Enzymatic (R)- phenylacetylcarbinol production in a benzaldehyde emulsion system with Candida utilis. Applied Microbiology and Biotechnology 70:170−175

Satianegara, G., P.L. Rogers, B. Rosche (2006) Comparative studies on enzyme preparations and role of cell components for (R)−phenylacetylcarbinol production in a two-phase biotransformation. Biotechnology and Bioengineering 94(6):1189−1195

Gunawan, C., Satianegara, G., Chen, A., Breuer, M., Hauer, B., Rogers, PL, Rosche, B. (2006) Yeast pyruvate decarboxylases: variation in biocatalysis characteristics for (R)-phenylacetylcarbinol production. FEMS Yeast Research doi:10.1111/j.1567- 1364.2006.00138.x

Presentations

G. Satianegara, M. Breuer, B. Hauer, P. Rogers and B. Rosche (2004) Enantioselective C-C bond formation in two-liquid phase reactor by whole cells Candida utilis. School of Biotechnology and Biomolecular Sciences Third Annual Symposium, Sydney, Australia, 5 November 2004, oral presentation IV−2, ISBN 0 7334 2162 8

G. Satianegara, M. Breuer, B. Hauer, P. Rogers and B. Rosche (2004) Enzymatic production of R-phenylacetylcarbinol in a biphasic system – whole cells versus partially purified enzyme. 6th International Conference of Protein Stabilization “ProtStab2004”. Bratislava, Slovakia, 26−29 September 2004, poster presentation 041.

B. Rosche, V. Sandford, G. Satianegara, A. Chen, N. Leksawasdi, C. Gunawan, M. Breuer, B. Hauer, and P. Rogers (2004) Carbon−bond formation by resting cells of

iv Candida utilis in an aqueous/octanol two-phase reactor. Kimball Union Academy (Gordon Research Centre of Biocatalysis), New Hampshire USA: 11−16 July

G. Satianegara, M. Breuer, B. Hauer, P. Rogers and B. Rosche (2004) Characterization of whole cell Candida utilis for biotransformation of R-phenylacetylcarbinol in the aqueous-organic two-phase system. Proceeding of Fermentation and Bioprocessing Conference, Brisbane, Australia, 5−6 July, poster presentation 20, ISBN 0 646 43707 0

G. Satianegara, C. Gunawan, A. Chen, M. Breuer, B. Hauer, P. Rogers and B. Rosche (2003) Comparison of four yeast strains for R-phenylacetylcarbinol (PAC) production. Proceeding of Fermentation and Bioprocessing Conference: Ideas into products, Sydney, Australia, 14−15 April, poster presentation 03, ISBN 0 7334 2023 0

G. Satianegara, C. Gunawan, A. Chen, M. Breuer, B. Hauer, P. Rogers and B. Rosche (2003) R-phenylacetylcarbinol (R-PAC) production and stability study with pyruvate decarboxylase from four yeast strains. Proceeding of 6th International Symposium on Biocatalysis and Biotransformation. BioTrans 2003, Palacky University, Olomouc, Czech Republic, June 28 – July 3, poster presentation P119, ISSN 0009−2770

B. Rosche, V. Sandford, N. Leksawasdi, A. Chen, G. Satianegara, C. Gunawan, M. Breuer, B. Hauer, and P. Rogers (2003) Bioprocess development of ephedrine production. Proceeding of 6th International Symposium on Biocatalysis and Biotransformation. BioTrans 2003, Palacky University, Olomouc, Czech Republic, June 28 – July 3, poster presentation P244, ISSN 0009−2770

v Table of Contents

Acknowledgements...... i Abstract...... ii Publications...... iv Table of Contents ...... vi List of Tables ...... xi List of Figures...... xiv List of Figures...... xiv 1 Literature review...... 1 1.1 Introduction...... 1 1.1.1 Trends in the biotechnology industry...... 1 1.1.2 Biotransformation and biocatalysis...... 2 1.2 Ephedrine production...... 7 1.2.1 Therapeutic use...... 8 1.2.2 Routes of synthesis...... 9 1.3 Biotransformation processes for R-PAC...... 11 1.3.1 Biocatalyst PDC...... 11 1.3.1.1 Structure of yeast PDC...... 13 1.3.1.2 Cofactors magnesium ions and TPP ...... 14 1.3.1.3 Catalytic activity...... 17 1.3.1.4 PDC stability...... 19 1.3.2 Enzymatic PAC production...... 20 1.3.2.1 Cell-based biotransformations ...... 21 1.3.2.2 Cell-free biotransformations ...... 23 1.4 New strategies for enhanced biotransformation processes ...... 25 1.4.1 Use of whole cell biocatalysts...... 25 1.4.2 Biocatalyst improvements...... 27 1.4.2.1 Rational approach...... 28

vi 1.4.2.2 Directed evolution...... 29 1.4.2.3 Stabilizing additives...... 30 1.5 Project Objectives...... 32 2 Materials and methods ...... 34 2.1 General chemicals, procedures, equipments ...... 34 2.2 Microorganisms source and maintenance ...... 35 2.3 PDC production and preparation...... 35 2.3.1 PDC production...... 35 2.3.1.1 Shake flask fermentation...... 35 2.3.1.2 3 L fermentation with pH downshift condition...... 37 2.3.2 PDC preparation...... 37 2.3.2.1 Whole cell PDC...... 37 2.3.2.2 Crude extract PDC ...... 38 2.3.2.3 Partially purified PDC...... 38 2.4 Recovery of cell debris and cell lipids from C. utilis...... 39 2.5 Analytical methods ...... 39 2.5.1 Dry cell mass (DCM)...... 39 2.5.2 Enzymatic analyses...... 39 2.5.2.1 PDC carboligase activity...... 39 2.5.2.2 Pyruvate and acetaldehyde concentrations...... 40 2.5.2.3 Glucose concentration...... 41 2.5.3 Chromatographic analyses...... 41 2.5.3.1 PAC, benzoic acid, benzaldehyde and benzylalcohol concentrations ...... 41 2.5.3.2 R- and S-PAC determination...... 42 2.5.3.3 Acetoin and ethanol concentrations ...... 42 2.5.4 Total protein quantification...... 44 2.5.5 Removal of PDC interfering substances ...... 44 2.6 PDC modification...... 44 2.6.1 Tresylated-activated methoxypolyethylene glycol ...... 44 2.6.2 Polysaccharides...... 45 2.6.3 Guanidination...... 45 2.6.4 Acylation...... 46

vii 2.6.5 Amidation...... 46 2.7 Biotransformation experiments...... 46 2.7.1 Aqueous/benzaldehyde emulsion system...... 46 2.7.1.1 General setup, sampling and sample preparations ...... 46 2.7.1.2 Comparison study of different PDC preparations...... 47 2.7.1.3 Biotransformations with whole cell PDC ...... 48 2.7.2 Aqueous/octanol-benzaldehyde emulsion system...... 48 2.7.2.1 General setup, sampling and sample preparations ...... 48 2.7.2.2 Comparison study of different PDC preparations...... 48 2.7.2.3 Effect of cell concentration ...... 49 2.7.2.4 Effect of initial enzyme activity...... 49 2.7.2.5 Effect of pH...... 51 2.7.2.6 Effect of MOPS concentration...... 51 2.7.2.7 Whole cell biotransformation with pH-control...... 52 2.7.2.8 Biotransformation in the presence of additives...... 52 2.7.3 Calculations for biotransformation ...... 52 2.8 Determination of residual PDC activity...... 53 2.8.1 Stability study in 50 – 300 mM benzaldehyde...... 53 2.8.1.1 Preparation procedure...... 53 2.8.1.2 Sampling procedure...... 54 2.8.2 Estimation of PDC release during whole cell biotransformation ...... 54 2.8.3 Stability study in an aqueous/octanol-benzaldehyde emulsion system without pyruvate ...... 54 2.8.3.1 Preparation procedure...... 54 2.8.3.2 Sampling procedure...... 55 2.8.4 Stability study of partially purified PDC in an aqueous/octanol-benzaldehyde biotransformation...... 55 3 Chemical modifications of partially purifed Candida utilis PDC ...... 56 3.1 Introduction...... 56 3.2 Results...... 58 3.2.1 Modification targeting amino-group of lysine ...... 58 3.2.1.1 Modification with polyethylene glycol ...... 58

viii 3.2.1.2 Modification with polysaccharide...... 60 3.2.1.3 Guanidination...... 62 3.2.1.4 Acylation...... 63 3.2.2 Modification targeting carboxyl-group of aspartic and glutamic acids...... 65 3.2.2.1 Amidation...... 65 3.3 Discussion and conclusions ...... 67 4 Enzymatic R-PAC production in an aqueous/ benzaldehyde emulsion system with C. utilis whole cell PDC...... 69 4.1 Introduction...... 69 4.2 Results...... 71 4.2.1 PDC stability towards benzaldehyde ...... 71 4.2.1.1 Effect of different purification grades on deactivation of C. utilis PDC ...... 71 4.2.1.2 Comparison to other yeast PDCs ...... 75 4.2.2 Biotransformation studies...... 76 4.2.2.1 Whole cells versus cell-free PDC ...... 76 4.2.2.2 Kinetics of PAC formation with whole cell PDC ...... 79 4.2.2.3 Factor affecting a decrease of R-enantiomer of PAC...... 81 4.3 Discussion and conclusions ...... 82 5 Use of whole cell PDC in a rapidly stirred aqueous/octanol- benzaldehyde emulsion system ...... 85 5.1 Introduction...... 85 5.2 Results...... 88 5.2.1 Comparative studies on different PDC preparations in an aqueous/octanol-benzaldehyde system ...... 88 5.2.1.1 PAC and by-product formation with whole cell versus cell-free PDC ...... 88 5.2.1.2 Effect of initial enzyme concentration on PAC production ...... 89 5.2.2 Further evaluation of whole cell biotransformations ...... 92 5.2.2.1 Effect of cell concentration ...... 92 5.2.2.2 Effect of pH...... 93

ix 5.2.2.3 Effect of MOPS concentration...... 95 5.2.2.4 Biotransformation time profiles...... 97 5.3 Discussion and conclusions ...... 101 6 Role of cell components for enhanced PAC Production...... 104 6.1 Introduction...... 104 6.2 Results...... 106 6.2.1 Effect of cell debris on PDC stability ...... 106 6.2.2 Effect of different additives on PAC production ...... 108 6.2.3 Effect of specific additives on PDC half-lives...... 110 6.2.4 Biotransformation in the presence of AOT...... 111 6.2.4.1 Effect of AOT concentration...... 111 6.2.4.2 Kinetics of PAC production...... 112 6.3 Discussion and conclusion ...... 115 7 Conclusions and recommendations for future studies...... 119 References ...... 125 Appendix A ...... 155 Appendix B ...... 156 Appendix C ...... 157 Appendix D ...... 162

x List of Tables

Table 1-1. Typical microbial sources of PDC or PDC activity (Candy and Duggleby, 1998; Cardillo et al., 1991; Chandra Raj et al., 2001; Iding et al., 1998; Neuser et al., 2000; Rosche et al., 2001; Talarico et al., 2001; Wiegel, 1980; Zorn et al., 2003) ...... 12 Table 1-2. Various mutations of catalytically important amino acids of yeast PDC ...... 16 Table 1-3. Recent application of microbial whole cell catalysts to industrial processes (Ishige et al., 2005) ...... 26 Table 2-1. The composition of yeast fermentation media for PDC production...... 36 Table 2-2. The reaction mixture composition for pyruvate/acetaldehyde assay ...... 40 Table 2-3. The column and operating conditions for acetoin quantification with GC...... 43 Table 2-4. The column and operating conditions for ethanol quantification with GC ...... 43 Table 2-5. Details of experimental setup for enzymatic biotransformation with different PDC preparations in an aqueous/benzaldehyde emulsion system...... 47 Table 2-6. The specific PDC activities of different whole cell PDC preparation for use in the study of effect of cell concentration in an aqueous/octanol-benzaldehyde system ...... 49 Table 2-7. Details of experimental volumes for effect of initial enzyme concentration for whole cell and partially purified PDC in the aqueous/octanol-benzaldehyde emulsion system...... 50 Table 3-1. Details of PDC modification with tresylated-PEG (initial activity 6 U/mL, 0.1

M K2HPO4/H3PO4, initial pH 7.5, 0.125 M NaCl, 1 mM MgCl2, room temperature and total reaction volume 1.5 mL) ...... 59 Table 3-2 shows the various conditions for PDC modification with oxidized polysaccharides, also known as dialdehyde polysaccharides (DAP), of different molecular weights at room temperature. The conditions were also varied for the reducing agent, pH and reaction period. Heavy precipitates were formed after 2 days in the absence of magnesium ions (one of the two cofactors for PDC) and/or in the presence of mild reducing agent Pyridine- Borane-Complex (PBC). With only traces of m-PDC activities recovered after modification (even for control in the absence of DAP), it was clear that partially purified PDC was significantly inactivated during modification, presumably due to the enzyme exposure to alkaline pH over 1 to 4 days of

xi reaction. Similar results were observed with the use of S. cerevisiae PDC.Table 3-2. Details of PDC modification with dialdehyde polysaccharides/DAP (initial activity 6

U/mL, 0.2 M K2HPO4/H3PO4 pH adjusted to 7.13 or 0.1 M boric acid/borate adjusted to pH 8.6, room temperature and total reaction volume 1.0 mL)...... 60 Table 3-3. Details of PDC guanidination with 3,5-dimethylpyrazole-1-carboximidamide (DMPC) (initial pH 9.4/NaOH and a total reaction volume 1.5 mL)...... 63 Table 3-4. Details of PDC modification with anhydrides (initial activity 6 U/mL, 0.2 M

K2HPO4/H3PO4 adjusted to pH 7.13 or 0.1 M

Na2HPO4/NaH2PO4/CH3COONa adjusted to pH 8.0 or 0.1 M boric

acid/borate adjusted to pH 8.6, 1 mM MgCl2, room temperature, total reaction volume 1.5 mL)...... 64 Table 3-5. The expected effect on enzyme surface characteristics upon the replacement of the carboxyl-groups with various side chains...... 66 Table 3-6. Details of PDC modification with carbodiimides (initial activity 6 U/mL, 0.2

M K2HPO4/H3PO4 adjusted to pH 5.2, 2 mM MgCl2, 0.5 mM TPP, 0.9 % (w/v) EDC, room temperature, total reaction volume 2.0 mL)...... 66 Table 4-1. The estimated half-lives of different preparations of C. utilis PDC from data in Figures 4-1 and 4-2...... 75 Table 4-2. Summary of biotransformation data for different PDC preparations in an aqueous/benzaldehyde emulsion system after 24 h (see also Figure 4-4)...... 78 Table 4-3. Effect of treatment for whole cell C. utilis PDC on PAC production in an aqueous/benzaldehyde emulsion system over 1 and 24 h period (initial activity

3 U/mL, 2.5 M MOPS/KOH, pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 300 mM benzaldehyde, 360 mM pyruvate) (see Appendix B for initial 1-3 h PAC formations) ...... 78 Table 4-4. Kinetic analysis of biotransformations with PDC added in the form of whole cells as detailed in Figure 4-5 ...... 81 Table 5-1. PAC and by-products formation in a rapidly stirred aqueous/octanol- benzaldehyde system at 21oC with partially purified, crude extract and whole cell PDC (initial activity 1 U/mL, 2.5 M MOPS, initial pH 6.5, 0.5 mM

MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol, 24 h and a total reaction volume of 1.8 mL)...... 89 Table 5-2. Comparison of PAC production with whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde system at 21oC with and without pH control ...... 100

xii Table 6-1. Estimated half-lives of different PDC preparations in a rapidly stirred aqueous/octanol-benzaldehyde system without pyruvate at 21oC. See data (4 U/mL) in Figure 6-1...... 107 Table 6-2. Summary of PAC production by partially purified and whole cell PDC in the absence and presence of 50 mM AOT in a rapidly stirred aqueous/octanol- benzaldehyde system (24 h) at 21oC with an initial PDC activity 1 U/mL (see Figure 6-6 and Figure 6-7) ...... 115

xiii List of Figures

Figure 1-1. Total biotechnology granted patents in USA from 1989 to 2002 (Biotechnology Industry Association, 2005)...... 2 Figure 1-2. Australia Biotech companies by sub-sectors (AusIndustry, 2004)...... 2 Figure 1-3. An overview of the production of fine chemicals by biotransformation: (a) biotransformation processes, (b) type of compound produced, (c) industrial sectors and (d) source of chirality for the products (Straathof et al., 2002)...... 4 Figure 1-4. The energy efficiencies and eco-friendly advantages of biotechnology compared to the traditional chemical technology in antibiotics intermediate processing (The European Association of BioIndustries, 2000-2005)...... 4 Figure 1-5. A typical research cycle for biocatalysis (Schmid et al., 2001)...... 5 Figure 1-6. Specific technical research needs for a competitive biocatalysis in the 21st century (blue for highest priority, green for medium priority, yellow for low priority) (Scouten, 1999) ...... 6 Figure 1-7. Photographs of Ephedra viridis (Charters, 2005) and Ephedra sinica (Anonymous, 2006) ...... 7 Figure 1-8. Six major isomers of ephedrine alkaloids (Abourashed et al., 2003) ...... 8 Figure 1-9. Preparation of ephedrine from plant material (Reti, 1953) ...... 9 Figure 1-10. Chemical synthesis of ephedrine from cyanohydrin ...... 10 Figure 1-11. Combined biotransformation and chemical steps for synthesis of ephedrine from benzaldehyde ...... 10 Figure 1-12. A ribbon drawing of a complete PDC tetramer based on crystallographic study with brewer’s yeast Saccharomyces uvarum (Furey et al., 1998). The binding site of the cofactors TPP and magnesium (represented with space filling) is also known as the active site of PDC...... 14 Figure 1-13. The structure of Thiamine PyroPhosphate (TPP) ...... 15 Figure 1-14. The decarboxylation of pyruvate through TPP reaction mechanisms in the active site of PDC (adapted from Pohl et al. (2002))...... 18 Figure 1-15. PDC-catalyzed non-oxidative decarboxylation and carboligation: R-PAC production and its associated by-products (adapted from Sergienko and Jordan (2001b)) ...... 20

xiv Figure 1-16. Comparison of R-PAC concentration in various biotransformation processes (R.j refers to Rhizopus javanicus. C.u. refers to Candida utilis) (Rosche et al., 2002b)...... 24 Figure 3-1. Chemical modification of enzyme by tresylated-methoxypolyethylene glycol (tresylated-MPEG)...... 59 Figure 3-2. Chemical modification of enzyme by dialdehyde polysaccharide...... 60 Figure 3-3. Chemical modification of enzyme by 3,5-dimethylpyrazole-1-carboximidamide (DMPC)...... 62 Figure 3-4. Chemical modification of enzyme by: (a) acetic anhydride (AA) (b) succinic anhydride (SA) and (c) pyromellitic dianhydride (PA)...... 64 Figure 3-5. Chemical modification of enzyme by carbodiimide-promoted amide formation...... 65 Figure 4-1. Profiles of residual activity of partially purified, crude extract and whole cell PDC at (a) 4oC (closed symbols) and (b) 21oC (open symbols) respectively in absence of benzaldehyde (initial activity 1.5 U/mL, 2.5 M MOPS, pH 6.5, 0.5

mM MgSO4, 0.5 mM TPP). The mean values were obtained from duplicate analyses of two vials. The error bars show lowest and highest values...... 73 Figure 4-2. Profiles of residual activity of partially purified, crude extract and whole cell PDC at (a) 4oC (closed symbols) and (b) 21oC (open symbols) respectively in the presence of 50 mM benzaldehyde (initial activity 1.5 U/mL, 2.5 M MOPS,

pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP). The mean values were obtained from duplicate analyses of two vials. The error bars show lowest and highest values...... 74 Figure 4-3. Estimated half-lives of whole cell PDC of four yeast strains at 21oC in the absence (white bars) and presence (black bars) of 50 mM benzaldehyde

(initial activity 1.5 U/mL, 2.5 M MOPS, pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP)...... 76 Figure 4-4. Initial 1-3 h PAC production by partially purified, crude extract and whole cell PDC in an aqueous/benzaldehyde emulsion system at 4oC (solid lines) and 21oC (broken lines) (initial activity 3 U/mL, 2.5 M MOPS, initial pH 6.5, 0.5

mM MgSO4, 0.5 mM TPP, 300 mM benzaldehyde, 360 mM pyruvate). The mean values were obtained from triplicate analyses of two biotransformations. The error bars show the lowest and highest values...... 77 Figure 4-5. Profiles of PAC production with PDC added in the form of whole cells of C. utilis in the aqueous/benzaldehyde emulsion system at (a) 4oC and (b) 21oC

(initial activity 3 U/mL, initial pH 6.5, 2.5 M MOPS, 0.5 mM MgSO4, 0.5 mM TPP, 300 mM benzaldehyde, 360 mM pyruvate). The mean values were

xv obtained from triplicate analyses of two biotransformations. The error bars show lowest and highest values...... 80 Figure 5-1: Biotransformation of pyruvate and benzaldehyde by enzyme PDC: an illustration of the benzaldehyde emulsion and the rapidly stirred two-phase systems...... 86 Figure 5-2. Effect of enzyme concentration on (a) initial and (b) final PAC concentrations, (c) PAC productivity and (d) specific PAC production by partially purified (◊) and by whole cells PDC (♦) in a rapidly stirred aqueous/octanol-benzaldehyde system at 21°C (2.5 M MOPS/KOH, initial pH 6.5, 1.45 M pyruvate, 0.5 mM

MgSO4, 0.5 mM TPP, 1.6 M benzaldehyde in octanol phase and a total reaction volume of 15 mL)...... 90 Figure 5-3. Effect of cell concentrations (g DCM/L) on PAC production in a rapidly stirred aqueous/octanol-benzaldehyde system (initial activity 1 U/mL, 2.5 M

MOPS, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol, 21oC and a total reaction volume of 15 mL). The mean values were obtained from triplicate analyses from two experiments. The error bars show lowest and highest values...... 93 Figure 5-4. Effect of pH on PAC production at 1h (◊) and 24h (♦) (1 U/mL, 50 mM buffer

cocktail, 0.5 mM MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol, 21oC and a total reaction volume of 100 mL) and on PDC half-life (∆) (1 U/mL, 50 mM buffer cocktail, 50 mM benzaldehyde, 10 mM octanol, 0.5 o mM MgSO4, 0.5 mM TPP, 21 C and a total reaction volume of 10 mL)...... 94 Figure 5-5. Effect of MOPS concentration on PAC production by whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde system (1 U/mL, initial pH was

set at 6.5 by addition of 1 M or 5 M KOH, 0.5 mM MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol, 21oC, 20 h). The mean values were obtained from triplicate analyses from two experiments. The error bars show lowest and highest values...... 96 Figure 5-6. Time profiles of substrates, product and by-products for whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde emulsion biotransformation at 21 ºC (initial activity 2.5 U/mL, 2.5 M MOPS/KOH initial pH 6.5, 0.5 mM

MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.6 M benzaldehyde in octanol and a total reaction volume of 15 mL). The mean values were obtained from triplicate analyses and the error bars show lowest and highest values...... 98

xvi Figure 5-7. Time profiles of substrates, product and by-products by whole cell PDC in a pH-controlled rapidly stirred aqueous/octanol-benzaldehyde emulsion biotransformation at 21 ºC (initial activity 2.5 U/mL, 2.5 M MOPS/KOH pH

6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.4 M pyruvate, 1.7 M benzaldehyde in octanol and a total reaction volume of 100 mL). The mean values were obtained from triplicate analyses and the error bars show lowest and highest values...... 99 Figure 6-1. Profiles of residual activities of partially purified, crude extract (clarified and unclarified) and whole cell PDC in a rapidly stirred aqueous/octanol- benzaldehyde system without pyruvate (initial activity 4 U/mL, 2.5 M MOPS, o pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.5 M benzaldehyde in octanol, 21 C). The mean values were determined from duplicate analyses from two vials. The error bars show lowest and highest values...... 107 Figure 6-2. Effect of additives on PAC formation in a rapidly stirred aqueous/octanol- benzaldehyde system at 21oC with partially purified PDC (1.5 U/mL, 2.5 M

MOPS, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.43 M pyruvate, 1.5 M benzaldehyde in octanol, 24 h and a total reaction volume of 1.8 mL). The mean values were determined from duplicate analyses. The error bars show lowest and highest values...... 109 Figure 6-3. Effect of the addition of 5 mM ergosterol and 50 mM AOT on PAC formation in a rapidly stirred aqueous/octanol-benzaldehyde system at 21oC with a crude extract and a whole cell PDC preparation (1.5 U/mL, 2.5 M

MOPS/KOH, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.43 M pyruvate, 1.5 M benzaldehyde in octanol, 24 h and a total reaction volume of 1.8 mL). The mean values were determined from duplicate analyses. The error bars show lowest and highest values...... 109 Figure 6-4. Profiles of residual activities of partially purified PDC in the absence and presence of 2 mg phosphatidylcholine/mL, 5 mM ergosterol, 50 mM AOT in comparison to whole cell PDC in a rapidly stirred aqueous/octanol- benzaldehyde system without pyruvate at 21oC (initial activity 1 U/mL, 2.5 M

MOPS/KOH, pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.5 M benzaldehyde in octanol). The mean values were determined from triplicate analyses. The error bars show lowest and highest values...... 110 Figure 6-5. Effect of AOT concentration on PAC production by partially purified PDC in a rapidly stirred aqueous/octanol-benzaldehyde system at 21ºC (initial activity

1.0 U/mL, 2.5 M MOPS/KOH, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP,

xvii 1.33 M pyruvate, 1.5 M benzaldehyde in octanol, 24 h and a total reaction volume of 1.8 mL). The mean values were determined from duplicate analyses. The error bars show lowest and highest values...... 111 Figure 6-6. Time profiles of substrates, product and by-products for partially purified PDC in an aqueous/octanol-benzaldehyde system at 21ºC in the absence of AOT (initial conditions: 1 U/mL, 1.3 M pyruvate, 2.5 M MOPS initial pH 6.5,

0.5 mM MgSO4, 0.5 mM TPP, 1.5 M benzaldehyde in octanol phase and a total reaction volume of 10 mL). The mean values were determined from triplicate analyses of two vials. The error bars show lowest and highest values...... 113 Figure 6-7. Time profiles of substrates, product and by-products for partially purified PDC in an aqueous/octanol-benzaldehyde system at 21ºC with 50 mM AOT addition (initial conditions: 1 U/mL, 1.3 M pyruvate, 2.5 M MOPS initial pH

6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.5 M benzaldehyde in octanol phase and a total reaction volume of 10 mL). The mean values were determined from triplicate analyses of two vials. The error bars show lowest and highest values...... 114 Figure 6-8. Chemical compositions of (a) phosphatidylcholine and (b) sodium bis(2-ethyl- 1-hexyl)sulfosuccinate (AOT or Aerosol-OT)...... 117 Figure 6-9. Possible microstructures formed in an aqueous/surfactant/organic system: the droplet model as (a) micelle and (b) reverse micelle versus (c) bicontinuous model (adapted from Lindman et al. (1989)) as pipeline structure/sponge-like phase (O = oil/organic solvent phase, W = water/aqueous phase, S = surfactant film)...... 118

xviii

Chapter 1

1 LITERATURE REVIEW Literature review

1.1 Introduction

The literature review initially concentrates on the current trends and challenges in biotransformation processes as alternatives to chemical ones. The biological production of (R)-phenylacetylcarbinol (PAC), which is a key intermediate in the production of ephedrine and its derivatives is the main area of interest, and the review then focuses specifically on a comparison of various enzymatic PAC biotransformation processes using pyruvate decarboxylase (PDC) from yeasts.

1.1.1 Trends in the biotechnology industry

Biotechnology is a collection of technologies that capitalize on the attributes of cells and their molecules to solve problems or to make useful products for human benefit (Eramian et al., 2005). Biotechnology had been traditionally used since 4000 – 2000 BC to leaven bread and ferment beer (Egypt), to produce cheese and ferment wine (Sumeria, China and Egypt) and to control plant breeding e.g. by the selective pollination of date palm trees (Babylonia). Modern biotechnology has developed since the discovery of bacteria in 1675 by Antonie van Leeuwenhoek and of proteins and enzymes in 1830s. A market capitalization of USD 331 billon was reported for the US- based biotechnology industry in 2004 and the increasing number of biotechnology patents granted from 1989 to 2002 is shown in Figure 1-1 (Biotechnology Industry Association, 2005). There are several different areas of activity in biotechnology ranging from pharmaceutical, environmental and classical fermentation to those relating to intellectual property or even terrorism. In Australia, the

1 Chapter 1. Literature review

major activities are focussed on the human therapeutics, agriculture and diagnostics applications (Figure 1-2).

Figure 1-1. Total biotechnology granted patents in USA from 1989 to 2002 (Biotechnology Industry Association, 2005)

Figure 1-2. Australia Biotech companies by sub-sectors (AusIndustry, 2004)

1.1.2 Biotransformation and biocatalysis

Industrial biotechnology is increasingly transforming the chemical industry via biotransformation/biocatalysis. By definition, biotransformation is the use of biocatalysts (or enzymes) for the transformation of non-natural man-made organic compounds (Faber, 1999). According to Bachmann and Riese (2005), industrial production is changing in three specific ways. Firstly, agricultural and waste biomass is

2 Chapter 1. Literature review

replacing fossil fuel feedstock. Secondly, bioprocesses such as fermentation and biocatalysis are replacing chemical syntheses. And thirdly, new bioproducts are emerging such as bio-based polymers, enzymes for use in textile industries or for feed and nutritional ingredients. Industrial biotechnology is already used to produce 5% of all chemicals today out of the USD 1.2 trillion total chemical sales. The enzyme market has reached USD 2 billion in sales and more are generated from the specialty chemicals for flavours, fragrances and other applications. Current industrial enzymes result in more than 500 products for more than 50 applications (Beilen and Li, 2002; Kirk et al., 2002). To date approximately 3000 enzymes have been characterized (Jaeger, 2004).

The physical state of the catalysts can be very diverse and they can exist either in free enzyme form or within whole microorganisms. The diverse specificity as well as the catalytic and biological properties of enzymes are highly desirable mainly because of the demand for enantiomerically pure compounds (Ward and Singh, 2000). According to McCoy (1999) as published in Chemical and Engineering News (Business), fermentation processes and enzyme-based catalysis are starting to challenge traditional synthetic methods of producing optically active pharmaceutical intermediates. An excellent overview on the diversity in the fine chemical industries is given by Straathof et al. (2002) based on 134 processes (Figure 1-3) and more overviews are available from Panke et al. (2004) and Kirk et al. (2002).

The use of biocatalysts in biotransformations minimizes many problems encountered in chemical processes. Biocatalysts can be used in simple and/or complex transformations without the need for tedious blocking and deblocking steps that are common in organic synthesis (Dordick et al., 1998). In addition, close to quantitative substrate conversion can be achieved under milder condition than those of chemical processes (Schmid et al., 2001). Biocatalysts are considered as environmentally friendly alternatives to chemical catalysts as they are biodegradable and meet the requirements of efficient reactions with fewer by-products. For production methods based on kinetic resolution of racemates, the development of specific bioprocesses allows the enhancement of the yield and reduction of waste and costs (Scouten, 1999). Figure 1-4 illustrates the advantage of the production of an antibiotic intermediate with a biological process in comparison to production via a traditional chemical process.

3 Chapter 1. Literature review

Figure 1-3. An overview of the production of fine chemicals by biotransformation: (a) biotransformation processes, (b) type of compound produced, (c) industrial sectors and (d) source of chirality for the products (Straathof et al., 2002).

Figure 1-4. The energy efficiencies and eco-friendly advantages of biotechnology compared to the traditional chemical technology in antibiotics intermediate processing (The European Association of BioIndustries, 2000-2005)

4 Chapter 1. Literature review

Scouten (1999) also commented that major driving forces for replacement with biocatalyst technology in “chemical-based” industries are based on pressures from society (way of living), business (profit and cost reduction), government (regulatory or environmental issues) and the need for basic research (truth and discovery). A typical biotransformation research strategy to improve and/or to produce a value-added product involves biocatalyst selection and preparation, bioprocess development and by-product removal associated with continuing economy evaluation. An illustration for the process cycle is given in Figure 1-5.

Figure 1-5. A typical research cycle for biocatalysis (Schmid et al., 2001)

However many biotransformation processes frequently result in low product concentration, low productivity and high production and recovery costs and hence, many such processes cannot compete directly with those in the chemical industry (Scouten, 1999; Willke and Vorlop, 2004). A workshop involving 50 leading scientific and industry experts was established in 1999 to produce a strategic plan for developing and utilizing a new generation of biocatalysts for the 21st Century (Scouten, 1999). The priorities and targets were divided into four areas due to the complexity for R&D

5 Chapter 1. Literature review

formulation viz. small volume fine chemicals, high volume industrial processes, screening/selection/development and whole organisms. Specific needs and technical barriers were addressed and the results are presented in Figure 1-6.

Figure 1-6. Specific technical research needs for a competitive biocatalysis in the 21st century (blue for highest priority, green for medium priority, yellow for low priority) (Scouten, 1999)

6 Chapter 1. Literature review

1.2 Ephedrine production

The sympathomimetic alkaloid ephedrine is one of 6 optical alkaloids naturally concentrated in the aerial part of Ephedra plants. In China, the species of genus Ephedra is traditionally known as Tsaopen - Ma Huang. Ephedrine is derived from several species of the genus Ephedra. While some of the Ephedra species have no alkaloid content, the Asian species E. sinica typically has the highest concentration of ephedrine. Photographs of the aerial part of E. viridis and E. sinica are shown in Figure 1-7.

Figure 1-7. Photographs of Ephedra viridis (Charters, 2005) and Ephedra sinica (Anonymous, 2006)

Ephedrine alkaloids are derivatives of 2-amino-1-phenyl-1-propanol in which the amino-group is either free, methylated or dimethylated (Abourashed et al., 2003; Hurlbut and Carr, 1998). The 3 pairs of diastereomeric alkaloids include ephedrine - pseudoephedrine, norephedrine - norpseudoephedrine and methylephedrine – methylpseudoephedrine (Figure 1-8).

7 Chapter 1. Literature review

OH CH3

CH NHCH3 3

OH NHCH3 (1R,2S)-Ephedrine (1S,2S)-Pseudoephedrine

OH CH3

CH3 N(CH3)2

OH N(CH3)2

(1R,2S)-N-Methylephedrine (1S,2S)-N-Methylpseudoephedrine

OH CH3

CH3 NH2

OH NH2

(1R,2S)-Norephedrine (1S,2S)-Norpseudoephedrine

Figure 1-8. Six major isomers of ephedrine alkaloids (Abourashed et al., 2003)

1.2.1 Therapeutic use

The use of these herbaceous perennial herbs dates back over 5000 years in China and was well documented during the time of Chinese Han Dynasty (ca. 207 BC – 220 AD) and Roman Empire (Bensky and Gamble, 1993; Chen and Schmidt, 1926; Weiss, 1988). The aerial parts of this evergreen have been used to treat symptoms of colds and asthma while the dried root is also one of the medicinal parts of the plant.

In modern medicine, ephedrine (or its derivatives) is commonly used as a nasal decongestant, and has been used therapeutically for nocturnal enuresis, dysmenorrhea, narcolepsy, diabetic neuropathic edema and myasthenia gravis (Abourashed et al., 2003). Ephedrine is valued for its calming effect on bronchial walls spasms and for its nervous system stimulatory effect as well as its effect in boosting the rate and strength of heart contractions. It acts directly and indirectly on the sympathetic nervous system and as it is metabolized to norephedrine, the central nervous system is also stimulated. Because of its indirect effect on neurotransmitter, long-term use of ephedrine is not recommended.

8 Chapter 1. Literature review

1.2.2 Routes of synthesis

Three known routes for ephedrine production are through plant extraction, chemical synthesis or combined biological (biotransformation) and chemical steps. Current commercial practice employs either plant extraction and/or the combined biological and chemical steps. For chemical and the combined routes, the common key intermediate is (R)-phenylacetylcarbinol (PAC).

Ephedrine is obtainable from dried plant material through alkali treatment and solvent extraction processes (Shukla and Kulkarni, 2000). The details of the plant extraction method are outlined in Figure 1-9. This process is tedious, cost and time- consuming and accompanied by many undesired products. The analysis of herbal products containing ephedrine alkaloids is complicated because of the potential occurrence of similar alkaloids, the presence of matrix contaminants and the occurrence of a wide range of different matrices (Hurlbut and Carr, 1998). Currently, China is the only producer that employs this route of ephedrine production and about 100,000 kg Ephedra powder and extracts were imported into the US in 1993 alone (Hurlbut and Carr, 1998).

Powdered plant materials

1. Base treatment 2. Base removal with chloroform 3. Chloroform removal 4. Treatment with dilute acid 5. Filtration 6. Base treatment of the filtrate 7. Alkaloid extraction with diethylether 8. Solvent evaporation 9. Recrystallisation with hot water

Ephedrine

Figure 1-9. Preparation of ephedrine from plant material (Reti, 1953)

9 Chapter 1. Literature review

The second route for ephedrine production is through a chemical synthesis from cyanohydrin (Abourashed et al., 2003; Shukla and Kulkarni, 2000). Cyanohydrin is chemically converted to key intermediate PAC followed by conversion to ephedrine (Figure 1-10). This process is not popular as problems such as a high level of impurities as well as harsh chemical process conditions are not industrially attractive.

OH 1. Grignard reagent OH 2. NaBH4 3. N-methylation R1 H CN NHR2

Various ephedrine analogs

Figure 1-10. Chemical synthesis of ephedrine from cyanohydrin

The most popular route employed is via the combined biotransformation and chemical steps. The first enzymatic biotransformation step takes advantage of the ability of an intracellular enzyme pyruvate decarboxylase (PDC) from fermenting yeast to produce PAC from substrates benzaldehyde and pyruvate. The second chemical step would then follow with the reduction of optically active R-PAC to form ephedrine in the presence of methylamine, hydrogen and platinum as catalysts (Figure 1-11).

OH OH CHO CH3 CH NH CH3 yeast 3 2 fermentation O NHCH H2/Pt 3 Benzaldehyde R-Phenylacetylcarbinol (1R,2S)-Ephedrine

Figure 1-11. Combined biotransformation and chemical steps for synthesis of ephedrine from benzaldehyde

This combined process is highly valued as more than 99% optically active intermediate R-PAC is produced during the biological step thereby eliminating the additional costs in racemic separation often encountered in chemical processes. To date, ephedrine is mainly synthesized chemically from the intermediate R-PAC, the latter through the microbial decarboxylation of pyruvate to acetaldehyde and subsequent

10 Chapter 1. Literature review

condensation with benzaldehyde (Mahmoud et al., 1990a) catalysed by the enzyme PDC. This biotransformation process became industrially important after the chemical synthesis of (1R)-(2S)-ephedrine from R-PAC was patented in Germany and USA (Hilderbrandt and Klavehn, 1932; Hilderbrandt and Mann, 1934).

1.3 Biotransformation processes for R-PAC

R-PAC is an important intermediate in the synthesis of (1R)-(2S)-ephedrine. The designation of the (R)-PAC enantiomer refers to the R-/S- system and is identical to (-)- PAC, rotating polarized light anticlockwise (laevorotatory) as well as to L-PAC in the D-/L- system using the Fischer projection. R-PAC is also known as 1-hydroxy-1- phenyl-2-propanone or Neuberg’s ketol (90-63-1) or 1-hydroxy-1-phenylacetone or α- hydroxybenzyl methyl ketone (Shukla and Kulkarni, 2000).

1.3.1 Biocatalyst PDC

PDC as the catalyst is recognized for its key role in PAC production. PDC is the first active enzyme following branching of metabolites (specifically pyruvate) from the glycolytic pathway in many fermentative microorganisms and its role is mainly recognized for ethanol production. During ethanol fermentation, PDC catalyses the non- oxidative decarboxylation of pyruvate to acetaldehyde, which is subsequently reduced to ethanol (Bringer-Meyer and Sahm, 1988; Ward and Singh, 2000). PDC is known also as a type of thiamine diphosphate-dependent enzymes and their use has been increasingly popular for regio- and enantio-selective reactions in other chemoenzymatic syntheses (Iding et al., 1998; Schorken and Sprenger, 1998; Sprenger and Pohl, 1999). The enzyme was identified by Neuberg in 1911 who later described its role for PAC formation in yeast through the condensation of acetaldehyde and benzaldehyde (Neuberg and Karczag, 1911). PDC is widely distributed in many prokaryotes and eukaryotes (Table 1-1) and the genes of PDC have been isolated from more than ten microorganisms (Holloway and Subden, 1993; Iding et al., 1998; Pohl, 1997; Ward and Singh, 2000).

11 Chapter 1. Literature review

Table 1-1. Typical microbial sources of PDC or PDC activity (Candy and Duggleby, 1998; Cardillo et al., 1991; Chandra Raj et al., 2001; Iding et al., 1998; Neuser et al., 2000; Rosche et al., 2001; Talarico et al., 2001; Wiegel, 1980; Zorn et al., 2003)

Acetobacter pasteurianus, A. suboxydans, Neurospora crassa, Aspergillus niger, A. nidulans, Pichia pastoris, A. parasiticus, Saccharomyces cerevisiae, Bacillus subtilis, S. carlbergensis, S. fermentatii, Candida albicans, C. utilis, C. tropicalis, S. uvarum, Clostridium botulinum, Sarcina ventriculi , Erwinia amylovorans, Schizosaccharomyces pombe, Hanseniaspora uvarum, Sporobolomyces salmonicolor, Hansenula anomala, Torulaspora delbrueckii, Hyphopichia burtoni, Zygosaccharomyces bisporus Kluyveromyces marxianus, K. lactis, Zymomonas mobilis

Extensive studies have been reported for PDC of yeasts Saccharomyces cerevisiae and Candida utilis and the bacterium Zymomonas mobilis (Iding et al., 1998; Pohl, 1997). For enzymatic PAC production, PDC of yeast C. utilis has been the best choice in term of activity and stability (Rosche et al., 2003b) compared to Z. mobilis PDC due to the low affinity of the latter to benzaldehyde as well as significant substrate inhibition (Konig, 1998; Ward and Singh, 2000) even though it is more stable at 25oC (Goetz et al., 2001). Other studies have shown that PDC can catalyse the condensation reaction (C-C bond) using alternative substrates i.e. between acetaldehyde and benzaldehyde (Torulaspora delbrueckii and Z. mobilis PDC) or pyruvate and various aliphatic/aromatic/heterocyclic aldehydes (Zygosaccharomyces bisporus PDC) to give the corresponding hydroxy ketones (Bringer-Meyer and Sahm, 1988; Fuganti and Grasselli, 1977; Fuganti et al., 1988; Kren et al., 1993; Ohata et al., 1986; Shukla and Kulkarni, 2001; Zorn et al., 2003).

12 Chapter 1. Literature review

1.3.1.1 Structure of yeast PDC

3D structures have been previously determined for S. cerevisiae PDC (Arjunan et al., 1996) and S. uvarum PDC (Dyda et al., 1993). The active structure of yeast PDC is reported to be a tetramer holoenzyme (or a dimeric dimer) consisting of a combination of α and/or β subunits. The tetramer can be purified in all possible variants viz. α4, α3β1, α2β2, α1β3 and β4, and shown to have identical activities and kinetic profiles with substrate pyruvate (Furey et al., 1998).

Each monomeric subunit has a molecular weight of around 60 kDa (Lu et al., 1997) and is formed of three domains termed α, β and γ domains (Furey et al., 1998). The N-terminal of the α-domain provides binding to the pyrimidine arm of the cofactor thiamine pyrophosphate (TPP), the β-domain is a central - regulatory domain providing intermolecular contact between the two dimeric halves of the PDC tetramer, while the C-terminal of the γ-domain provides binding to the diphosphate arm of the cofactor TPP (the γ-diphosphate binding domain) (Furey et al., 1998; Hong et al., 1998).

The α and γ domains are mainly involved in the non-covalent binding of the cofactors and substrates (Konig, 1998; Lu et al., 1997), providing tight subunit-subunit contacts within each dimer (Furey et al., 1998), at the active site (Figure 1-12). A rather loose and relaxed contact is confined to the β domains in dimer-dimer contact to form a tetramer (Furey et al., 1998) which leaves the centre of the tetramer largely open and solvent accessible. It is thus thought to be involved with activation by substrate (Lu et al., 1997). An illustration on PDC tetramer showing subunit interaction with cofactors TPP and magnesium is given in Figure 1-12.

13 Chapter 1. Literature review

Figure 1-12. A ribbon drawing of a complete PDC tetramer based on crystallographic study with brewer’s yeast Saccharomyces uvarum (Furey et al., 1998). The binding site of the cofactors TPP and magnesium (represented with space filling) is also known as the active site of PDC.

1.3.1.2 Cofactors magnesium ions and TPP

Two obligatory cofactors viz. TPP and magnesium ions are required for the PDC activity. The importance of magnesium and TPP for the catalytic activity of yeast PDC was first reported by Auhagen (1931) and by Lohmann and Schuster (1933), respectively. For active function, PDC requires 2-4 molecules of TPP and magnesium ions as obligatory cofactors (Tripathi et al., 1997). The presence of cofactors provides a stabilizing effect and shifts the association/dissociation equilibrium towards the tetrameric state (Killenberg-Jabs et al., 2001). The absolute requirement for the cofactors was clearly demonstrated by the effect of alkaline treatment (above pH 8.0) on the enzyme (Gounaris et al., 1971). Upon enzyme reconstitution, the presence of excess amounts of cofactors is required to recover total activity. The binding of the cofactors to the enzyme is highly cooperative, and the presence of one cofactor greatly increases the affinity for the second (Candy and Duggleby, 1998).

14 Chapter 1. Literature review

Green et al. (1941) proposed that magnesium serves as a bridge between PDC and TPP and this was supported in a later study using X-ray spectrocospy by Dyda et al. (1993). The presence of magnesium is essential to achieve overall stability and correct conformation for the active enzyme. In PDC, the octahedrally coordinated magnesium ions aid the binding of the diphosphate end of TPP to the γ domain of enzyme as well as the binding of 3 oxygen molecules from nearby amino acid residues and one molecule of water (Furey et al., 1998).

The structure of TPP, a biologically active form of vitamin B1, was elucidated by Lohmann and Schuster (1937). Figure 1-13 shows the structure of a TPP molecule comprised of an aminopyrimide end, a thiazolium ring and a diphophate end.

H NH2 + CH2 N S N - - O O H C N 3 - H3C CH2 CH2 OOPOP O O

Figure 1-13. The structure of Thiamine PyroPhosphate (TPP)

The binding of two ends of the TPP molecule provides a suitable “V” conformation of the thiazolium ring for the formation of a catalytically active PDC tetramer (Candy and Duggleby, 1998; Eppendorfer et al., 1993; Gounaris et al., 1975; Schellenberger, 1967). Both aromatic rings of the TPP have equal importance in catalysis (Sergienko and Jordan, 2001a). The conformation also involves the presence of essential amino acids within the contact range (Hubner et al., 1998; Killenberg-Jabs et al., 2002). Mutation studies have been carried out to investigate the role of these amino acids while crystallography studies identified a common structural binding motif shared by all TPP dependent enzymes (Candy and Duggleby, 1998) including PDC from S. cerevisiae, Kluyveromyces marxianus, Hanseniaspora uvarum and Neurospora crassa (Pohl, 1997).

15 Chapter 1. Literature review

Many studies have been carried out to elucidate the importance and the function of amino acids within the active site of yeast PDC through point mutations (site-directed mutagenesis) and the results can be summarised as follows (Arjunan et al., 1996; Baburina et al., 1994; Baburina et al., 1996; Hubner et al., 1988; Li and Jordan, 1999; Liu et al., 2001; Sergienko and Jordan, 2001a, b; Zeng et al., 1991; 1993): (1) Cys221 in the β-domain provides a trigger for substrate activation and the information is transmitted to the active centre via His92 and Glu91 in the α- domain to the Trp412 and Gly413 in the γ-domain, (2) interaction between Cys221 and His92 for substrate activation is presumably through electrostatic and steric mechanisms, (3) the four amino acids in the active centre (Asp28, His114, His115 and Glu477) are involved in substrate binding, decarboxylation and product release but not in the post-decarboxylation step (i.e. carboligation).

A list of different amino acid mutations, mostly on yeast PDC (S. cerevisiae or S. uvarum), that resulted in significant loss of decarboxylation activity is given in Table 1-2 (Baburina et al., 1994; 1996; 1998; Guo et al., 1998; Jordan et al., 1996; 1998).

Table 1-2. Various mutations of catalytically important amino acids of yeast PDC

Amino acid mutations

TPP binding site Asp28Ala, Asp28Asn, Glu51Ala, Glu51Asn, Glu51Asp, Glu51Gln, Glu477Gln, Glu477Asn, Glu477Asp, His114Phe, His115Phe, Ile415Leu, Ile415Met, Ile415Thr, Ile415Val

Substrate activation pathway Cys152Ala, Cys221Ala, Cys221Ser, Cys222Ser, Glu91Ala, Glu91Gln, Glu91Asp, His91Lys, His92Ala, His92Gly

16 Chapter 1. Literature review

1.3.1.3 Catalytic activity

PDC catalyses two simultaneous reactions i.e. the faster nonoxidative decarboxylation of α-keto acids to the corresponding aldehydes and a significantly slower carboligation side reaction forming hydroxy ketones (Goetz et al., 2001; Ward and Singh, 2000).

The first step of decarboxylation of an α-keto acid (i.e. pyruvate), involves the cleavage of a C-C bond adjacent to a keto group and results in the formation of an enzyme-bound aldehyde (Candy and Duggleby, 1998; Hubner et al., 1998). The mechanism for pyruvate decarboxylation is very dependent on the presence of cofactor TPP (Candy and Duggleby, 1998; Hubner et al., 1998; Kern et al., 1997; Kluger, 1987; Kuo and Jordan, 1983; Menon-Rudolph et al., 1992; Schellenberger and Hubner, 1968; Zeng et al., 1991). The schematic steps are shown in Figure 1-14 and can be summarized as follows: (1) the activation of TPP by PDC through deprotonation which results in the formation of C2 carbanion of the thiazolium ring of TPP, (2) the nucleophilic attack of the C2 carbanion of TPP on the α-carbonyl group of an α-keto acid (pyruvate) which forms α-lactyl-TPP,

(3) the stabilization the negative charge upon cleavage of the CO2 which results in the formation of α-carbanion-enamine (zwitterion intermediate). This step is known also as decarboxylation, (4) the protonation of the intermediate which gives an α-hydroxyethyl-TPP, also known as “enzyme-bound aldehyde” or “active acetaldehyde”.

The “active acetaldehyde” would then be released as free acetaldehyde or form a C-C bond (carboligation) with another aldehyde (see Figure 1-15 for the subsequent fate of the “active acetaldehyde” in PAC production). An interesting fact on the relative importance of PDC and TPP in catalysis is that TPP alone could catalyse the requisite reaction although it would be 1012 times less efficient than the holoenzyme itself (Furey et al., 1998).

17 Chapter 1. Literature review

- O O + + H - + H H H3C O - + +C pyruvate R1 N S R1 N S ( 1 ) H3CR2 H3CR2

O H3C ( 2 ) H acetaldehyde

"active- H OH O acetaldehyde" H3COC H H3C C - O +C + R1 N S R1 N S

H3CR2 H3CR2

( 4 ) ( 3 ) + CO H 2

H C - OH H C OH 3 C 3 C + .. R1 N S R1 NS zwitterion intermediate H3CR2 H3CR2

Figure 1-14. The decarboxylation of pyruvate through TPP reaction mechanisms in the active site of PDC (adapted from Pohl et al. (2002)).

PDC from yeasts exhibit a time-dependent regulation by substrate pyruvate i.e. activity is induced by the presence of pyruvate (Hong et al., 1998) and follow a typical sigmoidal kinetic pattern of substrate conversion. By comparison, no substrate activation was required for PDC from Z. mobilis which exhibits a hyperbolic kinetic pattern of substrate conversion with increasing substrate concentrations (Iding et al.,

18 Chapter 1. Literature review

1998). This ‘disadvantage’ of Z. mobilis PDC was compensated for by higher catalytic decarboxylation activity compared to yeast PDC (Iding et al., 1998). With respect to substrate utilization, yeast PDC decarboxylates a broader spectrum of α-keto acids as substrates (Ward and Singh, 2000). Low carboligation activity was observed for Z. mobilis PDC and successful activity improvement through site-directed mutagenesis was hampered by problems such as the reduction of conformational stability and the concomitant occurrence of a higher amount of optically inactive PAC (Goetz et al., 2001; Pohl, 1997; Ward and Singh, 2000).

1.3.1.4 PDC stability

PDC from most sources showed optimum activity between pH 6.0 and 6.5 (Candy and Duggleby, 1998; Lowe and Zeikus, 1992). The pH dependence of the PDC tetramer from S. cerevisiae was investigated using X-ray solution scattering by Konig et al. (1992). The tetramers were mostly found between pH 5.5 – 6.5 while the dimers were found at pH 9.5. Between pH 6.5 – 9.5, both forms of PDC were found to exist (Candy and Duggleby, 1998). At alkaline conditions, PDC is almost completely dissociated from TPP but not from the magnesium ions (Pohl, 1997). In brewer’s yeast S. cerevisiae, only the tetramer has been shown to be catalytically active. The reversible-pH dependent dissociation of the quaternary structure into smaller complexes was also found for PDC from other yeasts and pea seeds but was less significant for the tetrameric Z. mobilis PDC (Iding et al., 1998).

Temperature plays a critical role in PDC stability and is often considered as one of the most important parameters in R-PAC production. Most of the studied PDC exhibited a considerable instability with increasing temperature (Konig, 1998; Ward and Singh, 2000).

PDC deactivation is also very significant in the presence of substrate benzaldehyde (Chow et al., 1995; Leksawasdi et al., 2003). Increased PDC stability towards benzaldehyde in the presence of pyruvate has been observed (Rosche et al., 2005a) but the mechanism of this improvement is not understood. Product PAC and by-

19 Chapter 1. Literature review

products (acetaldehyde and acetoin) were also reported to have a deactivating and/or inhibiting effect on partially purified C. utilis PDC (Sandford et al., 2005).

1.3.2 Enzymatic PAC production

The biotransformation of benzaldehyde to optically active R-PAC is the key step in the production of (1R)-(2S)-ephedrine (Long and Ward, 1989). Figure 1-15 shows that the decarboxylation of pyruvate by PDC forms an “active acetaldehyde” (or enzyme-bound aldehyde) which reacts with benzaldehyde to produce R-PAC and other associated by-products (acetaldehyde, acetoin, benzylalcohol, benzoic acid). In general, substrate benzaldehyde may also be reduced to benzylalcohol and/or oxidized into benzoic acid while substrate pyruvate gives by-products acetaldehyde and acetoin.

H3C O + + H - - O O pyruvate

PDC CO2 H H COHC OH 3

+ R NS NAD 1

+ H CR NADH + H 3 2 benzylalcohol O oxid "active acetaldehyde" ored uct ases

benzaldehyde PDC on O idati PDC ox PDC OH

H C O H C O H3C O 3 3 benzoic acid H OH OH CH3 acetaldehyde acetoin PAC

Figure 1-15. PDC-catalyzed non-oxidative decarboxylation and carboligation: R-PAC production and its associated by-products (adapted from Sergienko and Jordan (2001b))

20 Chapter 1. Literature review

1.3.2.1 Cell-based biotransformations

PAC is commercially produced via biotransformation of benzaldehyde and glucose using live fermenting yeast S. cerevisiae at 25°C, also known as the Knoll procedure (Hilderbrandt and Klavehn, 1932). In general, the process can be divided into two stages as follows (Shin and Rogers, 1995): (1) fermentative yeast growth which involves the accumulation of yeast cells containing enzyme PDC and limited production of substrate pyruvate, (2) biotransformation stage involving (programmed) feeding of substrate benzaldehyde to produce relatively low concentrations of PAC together with some benzylalcohol production.

Quantitative conversion of pyruvate and/or benzaldehyde to higher concentrations of PAC was never achieved with the traditional yeast-based fermentation process due to the inactivating effect of benzaldehyde on PDC, insufficient pyruvate generation from glucose and substantial conversion of benzaldehyde to by-product benzylalcohol (Csuk and Glanzer, 1991; Iding et al., 1998; Long and Ward, 1989; Mahmoud et al., 1990a; Rogers et al., 1997).

A significant amount of benzaldehyde was converted to benzylalcohol due to the presence of yeast oxidoreductases and the regeneration of electron donors (e.g. NADH). Strategies to reduce benzylalcohol formation had limited success, although it was found that the addition of acetaldehyde as well as pyruvate could inhibit the formation of by- product benzylalcohol and increased the yield of PAC (Shukla and Kulkarni, 2000; Tripathi et al., 1997). The use of alcohol dehydrogenase (ADH) isoenzyme mutants of S. cerevisiae to eliminate the production of benzylalcohol also had limited success indicating the presence of many other oxidoreductases within the cells (Nikolova and Ward, 1991; Pohl, 1997; Tripathi et al., 1997). The main role of ADH in yeast fermentation is for the reduction of acetaldehyde to ethanol. Mochizuki et al. (1995) indicated that PAC can act as an inhibitor of ADH activity and the formation of benzylalcohol ceased at approximately 33 mM PAC (5 g/L) in their study with S. cerevisiae cells. According to Shin and Rogers (1995), benzylalcohol was preferentially

21 Chapter 1. Literature review

generated instead of PAC when the benzaldehyde was kept below 30 mM (approx. 3 g/L) for both free and immobilized cells.

Other by-products including acetyl benzoyl, benzoic acid, benzoin (an analog of PAC), butan-2,3-dione, 2-hydroxy-1,1-phenyl-propanone, (1R,2S)-1-phenyl-1,2- propane-diol, 1-phenyl-propan-2,3-dione, S-PAC and trans-cinnamaldehyde can also be produced during the PAC biotransformation process (Goetz et al., 2001; Groger et al., 1966; Nikolova and Ward, 1991; Smith and Hendlin, 1953; Tripathi et al., 1997; Voets et al., 1973).

Typical concentrations of PAC reported for the yeast biotransformation process are 4-12 g/L (Ward and Singh, 2000). Controlling the concentration of PDC cofactors was shown to achieve substantial improvement in the efficiency of PAC formation (Ward and Singh, 2000). Cessation of PAC production due to limitation of pyruvate near the end of biotransformation phase was also reported (Shin and Rogers, 1995).

The toxic effect of benzaldehyde could be counteracted to some extent by progressive addition of benzaldehyde, by strain selection, by using immobilized cells or by carrying out biotransformation reaction in media containing organic solvents (Ward and Singh, 2000), although the latter method may be detrimental to the cells. PAC production by cells immobilized in polymer matrices, alginates or cyclodextrins resulted in only an average of 5 g/L (Mahmoud et al., 1990a; 1990b; 1990c; Nikolova and Ward, 1994; Rogers et al., 1997; Shin and Rogers, 1995; Shukla and Kulkarni, 2000; Tripathi et al., 1991). The design of a three-stage process (biomass maximization, PDC induction, biotransformation with benzaldehyde feeding) to minimize inhibitory effects on yeast metabolism and PDC activity achieved a concentration of 10.6 g PAC/L, a productivity of 0.44 g/L/h and with a yield of 56% based on consumed benzaldehyde (Rogers et al., 1997). In a study by Wang (1994), up to 22 g PAC/L was produced through a combination of controlled fermentative metabolism of C. utilis (viz. RQ = 4- 5) and minimized PDC deactivation by using a sustained low benzaldehyde (1-2 g/L) feeding strategy.

22 Chapter 1. Literature review

1.3.2.2 Cell-free biotransformations

While the costs of enzyme isolation and purification may render the process uneconomical (Tripathi et al., 1997), it is possible that isolated enzymes can be recycled and reused during a biotransformation process (Ward and Singh, 2000). The use of isolated PDC offers the possibility of overcoming the problem of some of the by- product formation associated with cell-based biotransformation systems, i.e. by-product benzylalcohol formation. Furthermore, pyruvate limitation can be overcome and thus resulting in increased PAC yields and concentrations (Rosche et al., 2002a; 2002b; Shin and Rogers, 1996a).

The interest in characterization and improvement in enzymatic biotransformation started with the study of PAC formation by isolated bacterial and yeast PDC (Bringer-Meyer and Sahm, 1988) and with the reported higher PAC production with partially purified PDC (free and immobilized) from C. utilis (Shin and Rogers, 1996a; 1996b). A final PAC concentration of 28 g/L was achieved with both C. utilis PDC preparations at 4°C although there was increased formation of by-products of acetaldehyde and acetoin. It was also reported that higher immobilization capacity could be achieved with an inert polyacrylamide gel matrix when compared to adsorption on the surface of a cationic exchange resin. The immobilized PDC was found to have a longer half-life when compared to that of the free enzyme.

Later studies by Rosche et al. (2002a; 2002b; 2003a) with a high concentration MOPS buffer (2.5 M) and C. utilis or Rhizopus javanicus PDC resulted in approx. 50 g PAC/L (8.4 U/mL, 4°C) based on an aqueous/benzaldehyde emulsion system. Lowering the temperature from 23° to 4°C decreased initial PAC production rates but increased final PAC concentrations. The optimum pH for PDC activity was 6.5 and the maximum PAC yields were achieved at pH 6.5–7.0. The stability of PDC was shown to depend on the MOPS buffer concentration with the high concentration (2.5 M) providing enhanced PDC stability. This stabilising effect was not specific to MOPS buffer and could be achieved also by high concentrations of some alcohols (2M glycerol, 0.75M sorbitol), 10% (w/v) polyethylene glycol 6000), salts (1M KCl) or dipropyleneglycol (2.5 M) (Leksawasdi et al., 2005; Rosche et al., 2002a). An enantiomeric excess of 99% R-PAC

23 Chapter 1. Literature review

was achieved and the molar yield on consumed benzaldehyde was improved to more than 95%. Improved yields on pyruvate utilized were achieved by lowering the magnesium concentration from 20 to 0.5 mM (Rosche et al., 2003a). The enzymatic process based on cell-free PDC however still resulted in appreciable PDC deactivation by substrate benzaldehyde (Leksawasdi et al., 2003) even though deactivation was less severe when the substrate pyruvate was present (Rosche et al., 2005a). A mathematical model for PAC production in this system has also been developed (Leksawasdi et al., 2004).

To minimize PDC deactivation, an enzymatic aqueous/organic (two-phase) reactor has been designed subsequently in which the major concentrations of substrate benzaldehyde and product PAC are partitioned into the organic phase while PDC and pyruvate are located in the aqueous phase (Rosche et al., 2002b; Sandford et al., 2005). A significant improvement in final PAC concentrations and productivities has been achieved in this two-phase system using octanol as the organic phase. PAC concentrations exceeding 160 g/L have been obtained in the octanol phase with partially purified PDC from C. utilis but not with R. javanicus PDC, indicating that the latter enzyme would be less suitable for a two-phase process. A comparison of final PAC concentrations achieved with the fermentative and cell-free PDC is given in Figure 1-16 to illustrate the enhanced capability of the two-phase system.

Figure 1-16. Comparison of R-PAC concentration in various biotransformation processes (R.j refers to Rhizopus javanicus. C.u. refers to Candida utilis) (Rosche et al., 2002b)

24 Chapter 1. Literature review

1.4 New strategies for enhanced biotransformation processes

Various aspects of biotransformation processes have been gradually improved leading to potential industrial processes with areas of general focus listed below (Burton, 2001; Burton et al., 2002; Castro and Knubovets, 2003; Lye et al., 2002): (a) biocatalyst modification/preparation which includes immobilization, mutagenesis/rational design/metabolic engineering and directed evolution, (b) non conventional reaction media which includes biotransformations in two- phase systems, organic solvents, ionic liquids, solvent-free and micellar systems, (c) bioprocess development which includes bioreactor design, substrate feeding, biocatalyst recycle, in situ product removal, process monitoring / control / modeling / optimization.

In relation to the current study, the application of different biocatalyst preparations viz. whole cell and modified enzymes with improved characteristics, are reviewed in detail.

1.4.1 Use of whole cell biocatalysts

The use of whole cells (living or non-living, free or immobilized) has been an area of growing interest as such cell preparations may be more economical than purified enzymes (Dutta et al., 1994; Kiffe et al., 1995; Lee and Park, 1995; Leon et al., 1998; Matsumoto et al., 2001; Witholt et al., 1990). With the advent of recombinant DNA techniques, cells can also be hosts for enzyme overproduction, an important advantage for many industrial processes.

The advantages of whole cell processes can result from their higher catalyst stability, possible cofactor regeneration and/or use of multienzymatic pathways, as well as reduced cost for enzyme preparation (Leon et al., 1998; Matsumoto et al., 2001). Key issues for use of whole cells may include correct solvent selection, possible mass transfer limitations, selection of optimal agitation rate, potential side reactions and

25 Chapter 1. Literature review

maintenance of desired product stereoselectivity (Isken and de Bont, 1998; Leon et al., 1998; Matsumoto et al., 2001). The decision as to whether or not to use either cells or isolated enzymes however is usually made on a case by case basis.

A summary of recent successful whole cell-catalysed commercial processes is given in Table 1-3.

Table 1-3. Recent application of microbial whole cell catalysts to industrial processes (Ishige et al., 2005)

26 Chapter 1. Literature review

Due to the increasing number of potential biocatalytic conversions involving apolar substrates and products which are often toxic for whole cells, specific production conditions may need to be developed to minimize problems of whole cell enzyme deactivation and stability under these conditions. Strategies to minimize these problems include the use of two-phase system, cell entrapment or immobilization as well as reactions in alternative media, such as ionic liquids or supercritical CO2, for cell-based processes.

1.4.2 Biocatalyst improvements

The value of enzymes as analytical tools and as industrial catalysts is often limited by their requirements for mild storage and reaction conditions as well as their poor operational or long-term stability in some cases. The development of enzymes with increased stability under harsh conditions has greatly increased over the past decades (Schiraldi and De Rosa, 2002). According to Schmid et al. (2001), the key technologies of protein and enzyme engineering have contributed greatly to generating more stable enzymes. Protein engineering involves the alteration of structure, function and selectivity of enzymes and can sometime includes enhancement via molecular evolution, while enzyme engineering is designed to alter the enzyme microenvironment and thereby enhance its properties. While the expectation is that a conserved functional feature of the enzyme will be associated with one or more key residues, the plasticity of enzyme active sites is well noted. Many investigators have shown that homologous enzymes can catalyse different overall reactions possibly by different catalytic mechanisms, and it is quite evident that a fundamental understanding of the different interactions at enzyme active sites is still lacking (Todd et al., 2002; Weissman, 2004). In this Section, the different schemes for improved biocatalyst preparation are categorized to three areas viz. rational approach, directed evolution and use of stabilizing additives, as this provides a background to the enzyme modification studies that will be reported later in this thesis (Chapter 3).

27 Chapter 1. Literature review

1.4.2.1 Rational approach

The rational approach in enzyme modification means altering and changing the enzyme characteristics using a systematic approach based on an understanding of enzyme properties. The approach is best applied when a 3D enzyme structure has been obtained, preferably in the presence of substrate, with the active site identified (Weissman, 2004). Chemical modification or gene mutageneses are then used to target specific amino acids in the structure. The chemical modification may involve cross- linking of enzyme crystals (CLECs) (Govardhan, 1999) or enzyme aggregates (CLEA) (Schoevaart et al., 2004); cross-linking to a polymer matrix or to polysaccharides; bilayer encagement or immobilization to strengthen intra- or intermolecular interaction; modifications to obtain new overall enzyme properties e.g. hydrophilization or increased ionic interactions through replacement of amino acids. The chemical reactions used in amino acid modification are typically non specific (Davis, 2003).

The recent interest in thermophilic enzymes with their remarkable stabilities has contributed to new strategies for the physical optimization of an enzyme e.g. by reduction in the size of loops and in the number of cavities, reduced ratio of surface area to volume, changes in specific amino acid residues, increased hydrophobic interaction at subunit interfaces and changes in the surface area of the enzyme likely to be exposed to solvents (Demirjian et al., 2001; Jaenicke and Bohm, 1998). Modification of the surface hydrophilic/hydrophobic balance of the enzyme has also been reported to change enzyme properties such as its catalytic activity and stability. According to Mozhaev et al. (1988a; 1988b), non polar amino acids on the surface of proteins are organized as hydrophobic surface clusters. The distribution of these clusters can determine the solubility of a protein and its ability to act as a surfactant. These clusters can also play an important role in substrate binding and enzyme inactivation due to environmental stresses (Perutz, 1978).

Many enzymes have been successfully enhanced by systematic modification resulting in improved stability, altered substrate specificity and product enantioselectivity, altered cofactor requirements and even the introduction of novel catalytic activity (Bornscheuer and Pohl, 2001; Gåseidnes et al., 2003). For example,

28 Chapter 1. Literature review

the chemical modification of bovine pancreatic trypsin with different cyclodextrin derivatives successfully increased its thermostability up to 17°C. It was also more resistant to autolytic degradation at pH 9.0 (Fernandez et al., 2002a; Fernandez et al., 2002b). Surface charge substitution has provided subtilisin variants whose stability was higher than that of the native enzyme (Sears et al., 1994). Other studies reported the use of activated-PEG derivatives and n-octanol to form covalent linkages with the lysine residues of lipases. This resulted in variation of the enzyme surface hydrophobicities to improve catalytic properties in organic solvents (Koops et al., 1999). Chemical modification of lipase with PEG-5000 resulted in a significantly more stable enzyme (up to 20-fold higher than the native enzyme) at alkaline pH values (Calvo et al., 1995). Immobilization of protease on celite was also reported to cause a significant improvement of transesterification activity when compared to the suspended (not immobilized) protease (Triantafyllou et al., 1996).

1.4.2.2 Directed evolution

In contrast to the rational approach, a directed evolution strategy does not require an understanding of the relationship between protein structure and function (Weissman, 2004). Directed evolution exploits the recent availability of sequence information in enzyme/DNA libraries and applies genetic diversity/evolution strategies as usually found in nature such as errors in gene duplication, consensus mutations and gene shuffling. Enzymes resulting from this approach are then introduced back into microbial hosts followed by selection for the requisite trait and further evaluation of the mutant or chimeric enzyme. According to Arnold and Georgious (2003), the design of this selection and evaluation stage is extremely critical for the overall success.

In recent years, directed evolution has been successfully applied in the stabilization of enzymes towards thermoinactivation, tolerance to organic solvents, pH and oxidation (Bornscheuer and Pohl, 2001; Lehmann et al., 2000) as well as for the development of biocatalysts for bioremediation and detoxification and for the introduction of novel specificities and activities (Weissman, 2004). In a case study of phytase reported by Lehmann (2000), based on an alignment of homologous amino acid sequences with (closely) homologous mesophilic enzymes, and following calculation of

29 Chapter 1. Literature review

a consensus protein, the thermostability of phytase was increased in a single step by more than 20°C. The technology has made it possible even to evolve whole systems of enzymes.

1.4.2.3 Stabilizing additives

Protein stability in solution may be increased by addition of osmolytes, such as polyols, carbohydrates, polysaccharides, amino acids and/or their derivatives (Ou et al., 2001; Shimizu and Smith, 2004). Osmolytes are uncharged additives that, by acting as molecular crowders, affect protein folding/stability/function, solvent viscosity, surface tension (Davis-Searles et al., 2001; Kaushik and Bhat, 1998; Minton, 1998; Ó'Fágáin, 2003; Schein, 1990). Proteins are naturally designed to function in a crowded cellular (in vivo) environment and a volume exclusion generated by these cellular crowders is a key factor for maintaining protein function and stability in an in vitro environment. Several other studies suggested the role that osmolytes play in the event of osmotic stress with thermophilic and hyperthermophilic Archaea (Martins et al., 1997; Neves et al., 2005) and with yeast S. cerevisiae (Elbein et al., 2003; Hounsa et al., 1998; Mansure et al., 1994; van Dijck et al., 1995), presumably by preventing protein denaturation (Singer and Lindquist, 1998) or acting as membrane protectants (Hounsa et al., 1998). Increase in the refolding rate of Coprinus cinereus peroxidase was also reported in the presence of glycerol thus maintaining protein stability (Tams and Welinder, 1996). The addition of osmolytes (Anjum et al., 2000) and synthetic compensatory solutes (glycine betaine, homodeanolbetaine) can also increase the melting temperature of an enzyme (Goller and Galinski, 1999; Vasudevamurthy et al., 2004).

Increase in buffer concentrations has been reported to promote the activities of bilirubin oxidase, Chromobacterium viscosum lipase, subtilisin and chymotrypsin (Blanco et al., 1992; Skrika-Alexopoulos and Freedman, 1993; Yang et al., 1993). Lyophilization of α-chymotrypsin with buffers such as sodium phosphate, Tris-HCl and EPPS significantly increased the transesterification activity by at least 100-fold when compared to lyophilization in the absence of buffer salts (Triantafyllou et al., 1996). The same authors reported a similar observation with Candida antartica lipase. These

30 Chapter 1. Literature review

compounds, acting like a matrix support, may have preserved the optimum protein conformation and have prevented protein rigidification.

Polymers can act also to stabilize proteins. For example, the cationic polymer polyethyleneimine (PEI) has been reported to increase the shelf lives of dehydrogenases and hydrogenases at 36°C (Andersson and Hatti-Kaul, 1999). PEI protected the sulfydryl group of lactate dehydrogenase against oxidation, via metal chelation, and prevented protein aggregation at 36°C. The protection was observed also in the presence of EDTA (Andersson et al., 2000).

The stabilizing effect of surfactants on proteins has been observed in many enzymatic studies (Sarcar et al., 1991; Wu et al., 2001). Surfactants interact with the globular proteins through interaction of their charged polar groups and alkyl hydrophobic chains. The surfactants may form a micellar layer surrounding the protein molecule thereby creating a stabilizing effect on its globular structure. Most studies investigated enzymes entrapped in reverse micelles (water in oil microemulsions) (Castro and Knubovets, 2003). A preparation with extracted chymotrypsin in organic solvents with an ion-paired enzyme-surfactant (enzyme-AOT) complex has provided a stable and efficient catalyst (Paradkar and Dordick, 1994a, b).

Addition of soluble proteins such as Bovine Serum Albumin (BSA) was reported to have stabilized the enzyme β-galactosidase from Streptococcus thermophilus against thermal inactivation (Chang and Mahoney, 1995). It appears that the effect was on the enzyme-protein hydrophobic interaction as the surface hydrophobicity is a major determinant of the extent of the stabilization by a protein. In the absence of BSA, the half-life of the enzyme was 440 seconds at 64°C. In the presence of BSA and 1 M

NA2SO4, no loss of activity was detected after 8 h at 64°C.

31 Chapter 1. Literature review

1.5 Project Objectives

R-phenylacetylcarbinol (PAC) is the key chiral intermediate in the commercial production of ephedrine and pseudoephedrine. The current enzymatic biotransformation process has eliminated some of the problems associated with commercial practice (biotransformation using live-fermenting yeast). Such a cell-free process with partially purified Candida utilis pyruvate decarboxylase (PDC) has resulted in significant improvements in PAC concentrations, yields, productivities and specific PAC production per unit enzyme. However, maintaining PDC stability in the various systems is still a key concern.

The primary goal of the present work was to evaluate different PDC preparations of Candida utilis for enzymatic production of PAC. The criteria used for these evaluations are PDC stability, PAC concentration, yield and productivity, PAC per unit catalyst, enantio-specificity for R-PAC or lower by-products formation specifically acetaldehyde and acetoin. Three different C. utilis PDC preparations were used: resting cells (initially added and referred as a whole cell PDC), a crude extract obtained through physical (glass bead) attrition and a partially purified preparation obtained through solvent precipitation. Two different bioreactor systems were used: an aqueous/benzaldehyde emulsion (also referred as a benzaldehyde emulsion) and an aqueous/octanol-benzaldehyde emulsion (also referred as an aqueous/octanol emulsion or a two-phase emulsion) system.

The specific objectives can be summarized as follows: Objective (1): Chemical modifications of the partially purified C. utilis PDC - to chemically modify a partially purified PDC preparation for an improved characteristic viz. increased stability towards inactivating substrate benzaldehyde and/or improved activity for PAC formation.

Objective (2): Enzymatic R-PAC production in an aqueous/benzaldehyde emulsion system with various preparations including C. utilis whole cell PDC

32 Chapter 1. Literature review

- to evaluate the use of PDC initially added in the form of C. utilis cells (whole cell PDC) for its stability and for PAC production in an aqueous/benzaldehyde emulsion system at 4o and 21oC and to compare the results with those for crude extract and partially purified preparations.

Objective (3): Use of whole cell PDC in a rapidly stirred aqueous/octanol- benzaldehyde emulsion system - to investigate the potential of whole cell PDC in a rapidly stirred two-phase emulsion bioreactor at 21oC and identify key factors influencing PAC production.

Objective (4): Role of cell components for enhanced PAC production - to develop an understanding of the factors underlying the improved characteristics of whole cell PDC in a rapidly stirred two-phase emulsion system in comparison to the use of partially purified PDC preparations via addition of various additives similar to cell components.

33

Chapter 2

2 MATERIALS AND METHODS Materials and methods

2.1 General chemicals, procedures, equipments

All chemicals and proteins used in this project were purchased from either APS Chemicals Ltd (Australia), Boehringer Mannheim (Germany), Merck Co. (USA), Roche Diagnostic (Switzerland) or Sigma-Aldrich Co. (USA) unless otherwise stated. All chemicals were of Analytical Grade and of the highest purity. Reverse Osmosis (RO) water was used to prepare all media and buffer solutions. Milli-Q (Millipore, Australia) filtered water was used in the mobile phase preparation for HPLC analysis.

Each concentrated medium component for yeast cultivation was sterilized (121oC, 1 atm, 20 min) separately using an autoclave (Atherton, Australia) unless otherwise stated. A laminar flow chamber (Email Westinghouse, Australia) was used to carry out all aseptic work including the mixing of sterilized medium components and yeast sub-culturing. The weights for each chemical component were determined on scientific balances (Sartorius, Australia) of up to four decimals accuracy. A PHM210 Standard pH meter (Radiometer Analytical, France) was used for pH adjustments and was calibrated prior to use with colour coded buffer pH 4, 7 and 10 supplied by Merck Co. (USA). A refrigerated bench top Universal 32R Centrifuge (Hettich Zentrifugen, Germany) with interchangeable rotor was used for routine centrifugation. Centrifugation is generally carried out at 4oC unless otherwise stated.

Optical density (OD) readings of culture (660 nm), pyruvate/acetaldehyde measurements (340 nm) and total protein quantification (595 nm) were determined using an Ultraspec 2000 spectrophotometer (Amersham Pharmacia Biotech, Australia).

34 Chapter 2. Materials and Methods

Substrate (benzaldehyde), product (PAC) and by-products (acetoin and benzylalcohol) from enzymatic biotransformations were analysed either using HPLC (Shimadzu or Waters/Shimadzu, Australia) or using GC (Packard, Australia). Biotransformation and/or stability reactions in teflon screw cap glass vials were stirred with appropriate size of stirrer bars on magnetic stirrer plates (Bibby Sterilin or IKA) in constant temperature rooms.

2.2 Microorganisms source and maintenance

The yeasts C. utilis strain 709400, C. tropicalis strain 708900, K. marxianus strain 510700 and S. cerevisiae strain 102200 were obtained from the Microbiology Culture Collection of the School of Biotechnology and Biomolecular Sciences, University of New South Wales, Sydney, Australia (World Directory of Culture Collections No. 248). All yeast strains were maintained on agar plates containing glucose, 20 g/L; polypeptone, 10 g/L; KH2PO4, 1 g/L; MgSO4.7H2O, 0.5 g/L; agar, 15 g/L and stored at 4oC. Prior to PDC production in the bioreactor, the yeast was subcultured twice into fresh agar plates and incubated at 30oC for two days for each subculturing.

2.3 PDC production and preparation

All media and procedures for yeast cultivation and PDC production were as detailed in Chen et al. (2005). Methods of PDC preparation were as described by Rosche et al. (2002a) and Satianegara et al. (2006a). The detailed composition of yeast growth media is provided in Table 2-1.

2.3.1 PDC production

2.3.1.1 Shake flask fermentation

To produce sufficient PDC enzyme for strain comparison purposes, one loopful of yeast culture was inoculated into 50 mL pre-seed medium in a 0.5 L baffled

35 Chapter 2. Materials and Methods

Erlenmeyer flask. This pre-seed inoculum was grown in an orbital shaker at 30°C and

160 rpm for 12 – 16 hours to reach an exponential phase (OD660 nm = 7−9). The process was repeated with the transfer of 20 mL pre-seed inoculum into 180 mL of seed medium in a 1 L baffled flask and harvested when glucose concentration fell below 10 g/L.

Table 2-1. The composition of yeast fermentation media for PDC production

Concentration (g/L) pH downshift process Shake flask process Pre-seed a Seed a Fermentation b Pre-Seed a/Seed a

Glucose 20 20 100 90

Yeast extract ------10

Na2HPO4 ------2

(NH4)2SO4 6 6 10 10

KH2PO4 1 1 1 3

MgSO4.7H2O 0.5 0.5 0.5 1

CaCl2.2H2O 0.02 0.02 0.02 0.05

c FeSO4.7H2O 0.02 0.02 0.02 ---

ZnSO4.7H2O 0.0106 0.0106 0.0106 --- a pH was adjusted to 6.0 with 200 mM MES/KOH (sterilized with a 0.22 μm filter from Millipore, Australia) b pH was adjusted to 6.0 with 5 M NaOH c solution was prepared in acidic condition with the addition of 8 mM H2SO4 without further sterilization

36 Chapter 2. Materials and Methods

2.3.1.2 3 L fermentation with pH downshift condition

For an improved PDC production, the C. utilis pre-seed inoculum was prepared as described above at a higher agitation rate of 250 rpm. The process was repeated with the transfer of 10% (v/v) inoculum into the seed medium for a total of 50 mL and grown for 8 hours (OD660 nm=7−9). 10% (v/v) seed inocula were then transferred aseptically into the fermentor to initiate the larger scale PDC production.

Each batch of fermentation was carried out in a 5 L Biostat®A fermentor vessel (B.Braun, Germany) with 3 L working volume. The fermentation conditions were controlled at 30°C, stirrer speed 300 rpm and 0.1 vvm aeration. The pH of the fermentation was initially maintained at pH 6.0 and shifted down to pH 3.0 with addition of 5 M H2SO4 after approximately 20 g glucose/L (OD660nm = ~9) was consumed. Input gas flow rates as well as CO2 and O2 concentrations in the exit gas were measured continuously to determine the respiratory quotient (RQ) (Leksawasdi, 2004). The PDC production was stopped after glucose concentration fell below 10 g/L. Profiles of fermentation kinetics and PDC production with yeast C. utilis are given in Appendix A.

2.3.2 PDC preparation

All PDC preparations were stored at -20°C and PDC activity in the corresponding buffer was adjusted to the required value on the day of experiment.

2.3.2.1 Whole cell PDC

The yeast culture was harvested and centrifuged at 8,000 g for 15 min and washed with chilled RO water for C. utilis. For the study comparing PDC of four different yeast strains, chilled 0.9% (w/v) NaCl solution was used in the washing step for all yeast strains. The centrifugation and washing process were repeated three times. The washed pellet is referred to as a whole cell PDC preparation and was found to maintain 100% activity for at least 18 months when stored as frozen yeast cake.

37 Chapter 2. Materials and Methods

2.3.2.2 Crude extract PDC

The whole cell PDC was suspended in a breakage buffer (200 mM citrate buffer/KOH pH 6.0, 5 mM MgSO4, 5 mM TPP), adjusted to 60 g/L (OD660 nm = ~180) followed by freezing in liquid nitrogen for 5 times, with defrosting at 25oC in between each freezing. The freeze/thawed cell suspension was mixed with 0.5 mm glass beads at a volume ratio of 1 to 1 in a bead beater (Biospec, USA) for a 50 – 250 mL scale experiment or in a thick pyrex tube for a 1 mL scale study. The cells and beads were blended vigorously for five 1-min intervals with 1-min maintenance on ice in between each step. The cell supernatants were clarified by chilled centrifugation at 20,000 g for 5 min and referred to as a crude extract PDC. The preparation was found to maintain 100% activity for 1 week in the presence of 2.5 M MOPS pH 6.5, 0.5 mM TPP and 0.5 mM MgSO4 at -20°C.

2.3.2.3 Partially purified PDC

The partially purified PDC was prepared according to the previously described solvent precipitation protocol using acetone as the solvent (Sandford et al., 2005). Acetone was slowly added into a stirred solution of crude extract PDC at -10°C to reach a final concentration of 40 – 50% v/v. The precipitated protein containing PDC was then recovered through centrifugation at 8,000 g for 5 min at 0oC. The enzyme paste was freeze-dried (-50°C, 0.1 mbar) and ground to a fine powder with mortar and pestle. The enzyme powder was referred to as a partially purified PDC preparation. For the present studies, the enzyme powder was first dissolved in the corresponding biotransformation buffer (magnetically stirred for 1 h). The remaining undissolved powder was removed by centrifugation at 20,000 g for 5 min.

38 Chapter 2. Materials and Methods

2.4 Recovery of cell debris and cell lipids from C. utilis

After cell attrition and centrifugation (Section 2.3.2.2), the upper white layer of the pellet was recovered as cell debris and washed five−times with RO water to remove soluble cell components.

For lipid extraction, cell attrition (Section 2.3.2.2) was carried out for a continuous 5 min in 1 mL chloroform/methanol (2:1 v/v). After further addition of 4 mL of chloroform/methanol, the mixture was left standing for 1 h at room temperature and was then filtered through Whatman paper No.1. The filtrate was mixed with 0.88% (w/v) KCl at a volume ratio of 2:1 and allowed to separate into two phases. The lower phase was collected and volatile solvents were evaporated under vacuum (Speed Vac Savant DNA 100, USA) for 15 min at room temperature to obtain the lipid fraction. This extraction method was adapted from the protocol developed by Folch et al. (1957).

2.5 Analytical methods

2.5.1 Dry cell mass (DCM)

3 mL of culture broth was transferred into a pre-weighed glass culture tube and centrifuged at 20,000 g for 5 min. The supernatant was discarded and the cell pellet was washed twice more with RO water. The tubes were placed in an 110oC oven overnight, cooled in desiccator and weighed on a fine balance. The average dry cell mass for each time point sample was determined from 3-4 measurements.

2.5.2 Enzymatic analyses

2.5.2.1 PDC carboligase activity

Definition: One unit of carboligase activity produces one μmole PAC from benzaldehyde and pyruvate per min at 25°C and pH 6.4.

39 Chapter 2. Materials and Methods

An equal volume of enzyme solution was incubated with a 2-fold substrate solution (200 mM citric acid/KOH pH 6.4, 20 mM MgSO4, 1 mM TPP, 200 mM pyruvate, 3 M ethanol, 80 mM benzaldehyde) in a 25°C water-bath. The reaction was stopped after 20 min with 10% (v/v) TCA and incubated on ice for a further 5 min. The precipitated protein was removed by cold centrifugation at 20,000 g for 5 min and the supernatant was subjected for HPLC analysis. The linear range of the assay was 1-5 mM PAC.

2.5.2.2 Pyruvate and acetaldehyde concentrations

In the presence of reducing agent NADH, pyruvate is enzymatically reduced to lactate by lactate dehydrogenase (LDH) while acetaldehyde is enzymatically reduced to ethanol by alcohol dehydrogenase (ADH). The decrease in NADH is detected at 340 nm wavelength and it is stoichiometrically proportional to the amount of substrate consumed. The reaction mixture composition is detailed in Table 2-2.

Table 2-2. The reaction mixture composition for pyruvate/acetaldehyde assay

Volume (μL) Components Pyruvate assay Acetaldehyde assay

250 mM Triethanolamine/NaOH pH 7.6, 2.5 750 725 mM EDTA-Na2H2.2H2O (25°C water-bath)

Sample or blank (kept on ice) 25 50

0.025 g NADH-Na2, 0.05 g NaHCO3 25 25 in 5 mL of MilliQ H2O (kept on ice)

LDH 550 U/mL (kept on ice) 5 ---

ADH 5000 U/mL (kept on ice) --- 5

Total reaction volume 805 805

40 Chapter 2. Materials and Methods

The difference of absorbance over 8 min (before and after enzyme (LDH/ADH) addition) was used for pyruvate/acetaldehyde quantification based on Beer’s Law and is simplified as shown in the following equation, where c = concentration and b = light path length. The linear range was 0.25 – 4 mM for pyruvate assay and 0.125 – 2 mM for acetaldehyde assay. volumetotal sample volume A (mM)c A nm340 ×Δ= 1-1- ( ( ε NADH, 340 nm = .36 mM b()cm =× 1 )cm

Pyruvate(mM) ( nm340 ×Δ= 111.5A ) mM

Acetaldehy (mM)de ( nm340 ×Δ= 556.2A ) mM

2.5.2.3 Glucose concentration

Glucose concentration was measured using an YSI glucose analyser model 2300 Stat Plus (Yellow Springs Instrument, USA). The immobilized glucose oxidase in the analyser reacts with glucose generating glucono-δ-lactone and hydrogen peroxide. The latter component diffuses across a membrane and reacts with a Pt anode to produce current which is proportional to the glucose concentration in the sample.

The harvested cell culture was centrifuged at 20,000 g for 5 min and the clear supernatant was further diluted with RO water to less than 27.8 mM (upper linear range of assay).

2.5.3 Chromatographic analyses

2.5.3.1 PAC, benzoic acid, benzaldehyde and benzylalcohol concentrations

The concentrations of PAC, benzaldehyde and benzoic acid were quantified by HPLC at 283 nm and at 263 nm for benzylalcohol by comparing the peak areas to a set of external standards. The HPLC column Alltima C8 (5μm particle size, 150 mm length, 4.6 mm internal diameter) and guard column All-guard ™ (5μm particle size,

41 Chapter 2. Materials and Methods

7.5 mm length, 4.6 mm internal diameter) from Alltech Associates Pty. Ltd. (Australia) were used. The mobile phase used was prepared by degassing 32 % (v/v) acetonitrile, 0.5 % (v/v) acetic acid in Milli-Q water. The flow rate for isocratic operation was 1 mL/min with the run time of 20 min. The injection volume was 5 μL. The linear range of PAC detection was 1 – 19 mM. The linear range of benzaldehyde detection was 5 – 40 mM.

2.5.3.2 R- and S-PAC determination

HPLC was used also to differentiate the R- and S-PAC as well as to determine the respective concentrations. Comparison of relative peak areas of both enantiomers at 283 nm was used to calculate the enantiomeric excess (ee) of R-PAC. The HPLC column Chiracel OD (10 μm particle size, 250 mm length, 4.6 mm internal diameter) from Daicel Chemical Industries Ltd. (USA) was used. The mobile phase was prepared freshly based on the following composition 950 mL hexane, 50 mL isopropanol and 1 mL formic acid. The flow rate for isocratic operation was 0.8 mL/min with the run time of 30 min. The injection volume was 1 μL. The sample was prepared by vortexing 200 μL of the aqueous sample with 200 μL of ethylacetate. After separating the aqueous and organic phases by centrifugation, 150 μL of the top organic phase containing extracted PAC was placed into a clean 1.5 mL tube and was evaporated under vacuum (Speed Vac Savant DNA 100, USA) for 15 min at room temperature to remove ethylacetate. The sample was resuspended in 300 μL hexane, vortexed and centrifuged to remove solids. The supernatant was analysed by HPLC.

2.5.3.3 Acetoin and ethanol concentrations

Acetoin and ethanol concentrations were determined by comparison of peak areas with standards using GC in an isothermic operation. The chromatograms were recorded on a Spectra-Physics SP 4270 integrator with the peak areas given for further calculation.

A dilution of at least 10 times with RO water was required to determine acetoin concentration in the octanol phase. A set of acetoin standards of known concentrations

42 Chapter 2. Materials and Methods

ranging from 1 – 15 mM was used to correlate the area under peak to the acetoin concentration. The column and operating conditions for acetoin quantification are detailed in Table 2-3.

Table 2-3. The column and operating conditions for acetoin quantification with GC

Stainless steel, 3.6 m length, 10% carbowax® Column on chromosorb ® W-AW, 80 - 100 μm mesh range, 3.2 mm internal diameter

Carrier gas Nitrogen at 25 psig

Detector FID with hydrogen (15 psig) and air (15 psig)

Oven/injection/detector temperature 150/180/180 oC

Sample volume; total run time 3 μL ; 30 min

To measure the ethanol concentration, mainly from culture supernatant, a set of ethanol standards of known concentrations ranging from 1 – 20 g/L was used to correlate the area under peak to the ethanol concentration. The column and operating conditions for ethanol quantification are detailed in Table 2-4.

Table 2-4. The column and operating conditions for ethanol quantification with GC

¼ inch glass column, 1 m length, Porapak ® Column type Q packing material, 100 – 120 μm mesh range, 2.35 mm internal diameter

Carrier gas Nitrogen at 25 psig

Detector FID with hydrogen (15 psig) and air (15 psig)

Oven/injection/detector temperature 180/220/220 oC

Sample volume; total run time 3 μL ; 5 min

43 Chapter 2. Materials and Methods

2.5.4 Total protein quantification

Protein concentration was determined using the Coomassie® Plus Protein Assay Reagent (Pierce, USA) based on the Bradford colorimetric assay (Bradford, 1976). For quantification, 25 μL of the protein sample was mixed with 750 μL of the assay reagent and allowed to stand for 5 min at room temperature. The absorbance at 595 nm was used to determine protein concentration by correlating the reading to a standard curve of a set of albumin concentrations of 0.25, 0.5, 0.8 and 1.0 mg/mL.

2.5.5 Removal of PDC interfering substances

In the presence of biotransformation substrates, product and by-products, PDC activity cannot be measured directly. Therefore, a microchromatography column Biogel P-6/Sodium chloride (Biorad, Australia) was used to remove these interfering substances. The column was vortexed to resuspend the settled gel and allowed to stand for approx. 5 min to drain out the buffer. Excess buffer was further removed by centrifuging the column and collection tube at 1,000 g for 2 min. 75 μL samples containing interfering substances were delivered into a microfiltration column with a clean collection tube containing 55 μL collection buffer (400 mM citric acid, 40 mM

MgSO4.7H2O, 4 mM TPP, pH 6.0) and centrifuged for 4 min at 1,000 g. Approximately 110 μL of enzyme and buffer suspension were collected each time for carboligase activity determination.

2.6 PDC modification

2.6.1 Tresylated-activated methoxypolyethylene glycol

Tresylchloride-activated methoxypolyethylene glycol (MPEG with a molecular weight of 5,000 purchased from Sigma-Aldrich Co.) was added to the PDC solution. After 1 - 3 h at room temperature, 25% glycine solution was added to stop the reaction (500 μL/5 mL PDC solution).

44 Chapter 2. Materials and Methods

2.6.2 Polysaccharides

In this study, different molecular weights of dextrans (40 and 250 kDa) and Ficolls (70 and 400 kDa) were used. Dextran is a branched polysaccharide while Ficoll is a linear synthetic polymer of sucrose. The polysaccharide is first converted into dialdehyde polysaccharide (DAP) as a result of periodate (NaIO4) oxidation. A 4% (w/v) solution of each type of polysaccharide was prepared in 20 mM sodium acetate/acetic acid (0.1 M NaCl, pH 4.5). The oxidation was initiated by the addition of

50 mM NaIO4 and reaction was carried out for 2 h at pH 4-4.5 and 25°C. The reaction was stopped by addition of 1 % (v/v) ethylene glycol or via dialysis. DAP was stored as frozen aliquot for further use.

The resultant DAP could then be covalently linked to the amino-groups of amino acid lysine via Schiff’s base formation and the reaction was carried out at either pH 7.13 (0.2 K2HPO4/H3PO4) or at pH 8 (0.1 M Boric acid/Borate). The unstable Schiff’s base intermediate can be stabilized by the addition of mild reducing agent such as Pyridine Borane Complex (PBC, 60 mM) in the beginning of modification or by the addition of strong reducing agent such as sodium borohydrate (NaBH4) at the end of reaction. The reaction was stopped with the addition of 25% glycine and the m-PDC was characterized further with or without dialysis. The remaining reagents were then removed by exhaustive dialysis against 20 mM MOPS/KOH pH 6.5, 1 mM MgSO4 and 1 mM TPP.

2.6.3 Guanidination

10% (w/v) 3,5-dimethylpyrazole-1-carboxamidine (DMPC) was dissolved in Milli-Q water and the pH of solution was adjusted to 9.5 with 2 M NaOH. PDC solution was added into the DMPC solution and mixed for a period of time on a rotating wheel in a constant temperature room. At the end of incubation period, the samples were exhaustively dialyzed against 20 mM MOPS/KOH pH 6.5, 1 mM MgSO4 and 1 mM TPP.

45 Chapter 2. Materials and Methods

2.6.4 Acylation

Partially purified PDC was dissolved in the corresponding buffer with initial pH of 7.13, 8 or 8.6. The anhydride stock solution (1 M) was prepared in dimethylsulfoxide (DMSO). The concentrated anhydride solution was added stepwise (two to three-times) as an aliquot of 10 μL with each time interval of approx. 15 to 30 min. The drop-wise addition of anhydride results in pH decrease. At the end of incubation period, the samples were exhaustively dialyzed against 20 mM MOPS/KOH pH 6.5, 1 mM MgSO4 and 1 mM TPP.

2.6.5 Amidation

The nucleophile donors i.e. aniline (25 mM), arginine methyl ester (1 M), chitosan 5000 (0.2% w/v) and guanidine-2-benzimidazole chloride (50 mM) were prepared in 0.2 M K2HPO4/KH2PO4, pH 5.2, 4 mM MgCl2 and 0.5 mM TPP. 2 mL of nucleophile donor solution was added to 200 μL of partially purified PDC solution in the same buffer and the pH was adjusted to 5.2 with 12 M NaOH. The reaction was initiated by adding 0.9 % (w/v) 1-ethyl-3(3-dimethylaminopropyl) carbodiimide (EDC) to activate the carboxyl groups of enzyme. At different time periods of 20-60 min, the reaction was quenched by the addition of 0.25 M sodium acetate buffer, pH 5.7. At the end of incubation period, the samples were exhaustively dialyzed against 20 mM

MOPS/KOH pH 6.5, 1 mM MgSO4 and 1 mM TPP.

2.7 Biotransformation experiments

2.7.1 Aqueous/benzaldehyde emulsion system

2.7.1.1 General setup, sampling and sample preparations

To minimize pH increase during biotransformation, 2.5 M MOPS/KOH (the initial pH was adjusted to 6.5 with 5 M KOH) was used as the aqueous phase buffer for all reactions together with freshly added cofactors (0.5 mM MgSO4, 0.5 mM TPP). The initial pH was checked and recorded. Biotransformation experiments were started by the

46 Chapter 2. Materials and Methods

addition of concentrated PDC solution into the stirring system in a constant temperature room and run in duplicates or triplicates as detailed in the Table/Figure captions. The reaction was stopped by adding a TCA solution to the sample to a final concentration of 10% (w/v). Precipitated protein was immediately removed by cold centrifugation at 20,000 g for 5 min. The samples were stored in -20oC prior to analysis.

2.7.1.2 Comparison study of different PDC preparations

Concentrated benzaldehyde, pyruvate and PDC stock solutions were prepared as detailed in Table 2-5. All components excluding the PDC stock solution were previously stirred as an aqueous/benzaldehyde emulsion system in a 4 mL screw cap amber glass vial for at least 1 h at 4oC or at 21oC. In the control vials, 2.5 M MOPS buffer was added instead of the enzyme PDC.

Table 2-5. Details of experimental setup for enzymatic biotransformation with different PDC preparations in an aqueous/benzaldehyde emulsion system

Volume (mL) Components Partially Crude Whole Whole Whole purified extract cella cellb cellc

1.08 M benzaldehyde 0.50 0.50 0.50 0.50 0.50

1.32 M pyruvate 0.49 0.49 0.49 0.49 0.49

Absolute ethanol ------0.21 ---

2.5 M MOPS pH 6.5 0.21 0.21 0.21 --- 0.21

9 U PDC/mL 0.60 0.60 0.60 0.60 0.60 a this biotransformation system with untreated whole cell C. utilis served as a basis of comparison b this biotransformation system contained 2 M ethanol as one of the attempts to make whole cell C. utilis more permeable c the whole cell PDC was repeatedly frozen with liquid nitrogen and thawed in 25oC water-bath for three times prior to use in biotransformation

47 Chapter 2. Materials and Methods

2.7.1.3 Biotransformations with whole cell PDC

A concentrated benzaldehyde stock of 900 mM was prepared and vigorously stirred as an emulsion for 2 h. While stirring, 5 mL of the benzaldehyde emulsion was withdrawn into a 20 mL screw cap glass vial containing 5 mL of 1.08 M pyruvate. The mixture was stirred for 1 h at 4oC and at 21oC. The reaction was initiated with the addition of 5 mL of 9 U PDC/mL. In the control vials (total of 1.5 mL volume in a 4 mL screw cap glass vials), 2.5 M MOPS buffer was added instead of the enzyme PDC.

2.7.2 Aqueous/octanol-benzaldehyde emulsion system

2.7.2.1 General setup, sampling and sample preparations

A similar aqueous phase buffer as in an aqueous/benzaldehyde emulsion system was used unless otherwise stated. For the octanol (organic) phase, a 1.5 M benzaldehyde solution was prepared in pure octanol. The benzaldehyde in octanol was diluted by 100 times with RO water to measure its concentration in HPLC. Biotransformations were started by the addition of aqueous PDC stock solution into the rapidly stirred aqueous/octanol-benzaldehyde reaction mixture. The volume ratio of octanol to aqueous phase was 1 to 1. The reactions were stopped by withdrawing 1 volume of the aqueous/octanol-benzaldehyde sample into a prepared 0.1-volume of 100% (w/v) TCA solution and vortexed immediately. Precipitated protein was removed by cold centrifugation at 20,000 g for 5 min. The clear octanol and aqueous samples were then withdrawn separately taking care to avoid the interphase and the bottom pellet layers formed after centrifugation. When necessary, the aqueous phase sample was further centrifuged at a similar setting to obtain a clear sample.

2.7.2.2 Comparison study of different PDC preparations

0.9 mL aliquot of 1.5 M benzaldehyde in octanol was mixed with 0.6 mL of 1.65 M pyruvate in 2.5 M MOPS buffer containing 0.5 mM cofactors and an initial pH of 6.5. The mixture was stirred for 1 h at 4oC and at 21oC and the reaction was initiated with the addition of 0.3 mL of 6 U PDC/mL using different PDC preparations. Total

48 Chapter 2. Materials and Methods

reaction volume was 1.8 mL in a 4 mL screw cap glass vials. In the control vials, 2.5 M MOPS buffer was added instead of the enzyme PDC.

2.7.2.3 Effect of cell concentration

7.5 mL of 1.5 M benzaldehyde in octanol was mixed together with 5 mL of 1.65 M pyruvate and vigorously stirred as an aqueous/octanol-benzaldehyde emulsion for 1 h. The reaction was initiated with the addition of 2.5 mL of 6 U PDC/mL. Total reaction volume was 15 mL in a 20 mL screw cap glass vials. In the control vials (a total of 1.8 mL volume in a 4 mL screw cap glass vials), 2.5 M MOPS buffer was added instead of the enzyme PDC.

Different whole cell PDC preparations were provided by Allen Chen from various fermentation batches. The specific activities of each preparation are given in Table 2-6.

Table 2-6. The specific PDC activities of different whole cell PDC preparation for use in the study of effect of cell concentration in an aqueous/octanol-benzaldehyde system

1 2 3 4 5

Batch pH shift Fermentative seed pH shift Constant pH Constant pH

U/g DCM 448 305 203 142 116

2.7.2.4 Effect of initial enzyme activity

7.5 mL of 1.5 M benzaldehyde in octanol was mixed together with 5 mL of 1.65 M pyruvate and vigorously stirred as an aqueous/octanol-benzaldehyde emulsion for 1 h. Total reaction volume was 15 mL in 20 mL screw cap glass vials. In the control vials (a total of 1.8 mL volume in 4 mL screw cap glass vials), 2.5 M MOPS buffer was added instead of the enzyme PDC. Enzyme PDC was prepared as a concentrated stock solution and the details are shown in Table 2-7.

49 Chapter 2. Materials and Methods

Table 2-7. Details of experimental volumes for effect of initial enzyme concentration for whole cell and partially purified PDC in the aqueous/octanol-benzaldehyde emulsion system.

Volume added (mL)

Octanol phase

1.5 M benzaldehyde 7.5 7.5 7.5 7.5 7.5 7.5 7.5

Initial activity 0.2 0.5 1.0 1.5 2.5 3.5 4.3 (U/mL total reaction volume)

Aqueous phase –

(whole cell PDC)

1.65 M pyruvate 5.0 5.0 5.0 5.0 5.0 5.0 5.0

Buffer 2.38 2.21 1.91 1.62 1.03 2.44 0.0

25.5 U PDC/mL 0.12 0.29 0.59 0.88 1.47 2.06 2.5

Initial activity 0.3 0.6 1.1 1.4 2.5 3.5 4.3 (U/mL total reaction volume)

Aqueous phase –

(partially purified PDC)

1.79 M pyruvate 6.0 6.0 6.0 6.0 6.0 6.0 6.0

Buffer 1.41 1.3 1.12 1.02 0.62 0.26 0.0

42.5 U PDC/mL 0.09 0.2 0.38 0.48 0.88 1.24 1.5

50 Chapter 2. Materials and Methods

2.7.2.5 Effect of pH

A 100 mL reaction was performed in a Quickfit reactor. The reactor was equipped with a water jacket connected to a thermostat water bath (Thermoline, Australia) for temperature control. An overhead stirrer (IKA, Germany) with a custom- made glass impeller (stirrer dia. 3.5 cm, shaft dia. 0.8 cm, shaft length 33 cm) was mounted on top. A high precision digital pH controller with 20 mL autoburette equipped with an antidiffusion delivery tip (Radiometer, Denmark) was used to control pH during biotransformation. A combination pH electrode (Mettler Toledo, USA) with 1 M LiCl in acetic acid as an electrolyte was used for pH measurement in the aqueous/octanol-benzaldehyde system. The horizon parameter in the adaptive addition algorithm (AAA) of the pH-stat controller was set at 80 with the default time constant value of 2 seconds.

To accommodate a wide range of pH, buffer cocktails containing a mixture of 50 mM of each citric acid (pKa 3.15), MES (pKa 6.15) and HEPES (pKa 7.55) were used to replace 2.5 M MOPS. The buffer stock was prepared at a 1.25-times concentration (62.5 mM) with a final pH of 3 without adjustment. An enzyme solution containing 25

U PDC/mL was prepared freshly in RO water with 2.5 mM TPP and MgSO4. Prior to the start of the reaction, 50 mL of 1.5 M benzaldehyde in octanol and 44 mL of 1.25 M pyruvate in 50 mM buffer cocktail were vigorously mixed at 600 rpm. The pH was slowly adjusted with 1 M KOH due to the partitioning of benzoic acid (small amount present as contaminant in benzaldehyde) from the octanol phase into the aqueous phase. During biotransformation, the pH was maintained with the addition of 5 M acetic acid. The reaction was monitored over time and the amount of acid addition was recorded for later calculations (see Appendix D).

2.7.2.6 Effect of MOPS concentration

Each set of pyruvate and PDC stock solution was prepared separately in MOPS buffer of different concentrations (0.05, 0.1, 0.25, 0.5, 1, 1.5 and 2.5 M) with 0.5 mM cofactors and an initial pH adjusted to 6.5. A 0.9 mL aliquot of 1.5 M benzaldehyde in octanol was mixed with 0.6 mL of 1.65 M pyruvate. The mixture was stirred for 1 h at

51 Chapter 2. Materials and Methods

21oC and the reaction was initiated with the addition of 0.3 mL of 6 U PDC/mL. Total reaction volume was 1.8 mL in a 4 mL screw cap glass vials.

2.7.2.7 Whole cell biotransformation with pH-control

The general method of biotransformation followed that described in Section 2.7.2.5 with the following changes. Prior to the start of reaction, 50 mL of 1.5 M benzaldehyde in octanol and 39 mL of 1.65 M pyruvate in 2.5 M MOPS were vigorously mixed at 1,200 rpm and the pH was slowly adjusted to 6.5. The biotransformation was initiated upon addition of 10 mL of concentrated PDC solution

(25 U/mL) prepared in 2.5 M MOPS pH 6.5 with 0.5 mM TPP and 0.5 mM MgSO4.

2.7.2.8 Biotransformation in the presence of additives

All additives were weighed directly into screw cap glass vials and the weights were recorded for further adjustment of the total reaction volume. For a 1.8 mL reaction, 0.9 mL of 1.5 M benzaldehyde in octanol and 0.72 mL of the concentrated pyruvate stock solution were mixed together with the different additives and stirred as a two-phase emulsion for at least 1 h. Then 0.18 mL of 10 U partially purified PDC/mL was added to initiate the reaction. The same volume ratio was adjusted accordingly during scale-up. The washed cell wall fraction of yeast S. cerevisiae (Auxoferm MCT lot 463-5) containing 78% polysaccharides (55% glucan) plus some lipids was obtained from Deutsche Hefewerke GmbH & Co (Marl, Germany). Soybean lecithin was purchased from Nature’s Own (Australia).

2.7.3 Calculations for biotransformation

The following calculations were used to determine the kinetics in biotransformation studies.

PDC mLUsampletheinactivity )( Specific PDC production = mLgmasscelldry )(

52 Chapter 2. Materials and Methods

PAC (molformed ) perPACyieldMolar benzaldehyde = × %100 benzaldehyde consumed ()mol

PAC (molformed ) perPACyieldMolar pyruvate = × %100 pyruvate consumed ()mol

⎡ PAC (molformed ) ⎤ Benzaldehyde unaccounted %100 −= ⎢ × %100 ⎥ ⎣benzaldehyde consumed ()mol ⎦

⎡(PAC + acetaldehyde (2×+ acetoin) )(mol ) ⎤ uvatePyr unaccounted %100 −= ⎢ × %100 ⎥ ⎣ pyruvate consumed ()mol ⎦

PAC mLmgformed )( Specific PAC production = PDC mLUactivity )(

PAC Lgformed )( Volumetric PAC typroductivi = time

2.8 Determination of residual PDC activity

2.8.1 Stability study in 50 – 300 mM benzaldehyde

2.8.1.1 Preparation procedure

PDC as a partially purified enzyme, a crude extract or a whole cell preparation was prepared as a concentrated stock solution in 2.5 M MOPS pH 6.5, 0.5 mM TPP and

0.5 mM MgSO4 with an initial pH of 6.5. The stability study was initiated with the addition of PDC solution into a pre-stirred benzaldehyde solution. PDC was exposed to 50 mM benzaldehyde in stirred 4 mL teflon screw cap glass vials of 1.8 mL reaction volume at 4oC and 21oC.

In the investigation of the putative effect of protease on PDC stability, the conditions were slightly altered with respect to the buffer composition i.e. 2.5 M MOPS

53 Chapter 2. Materials and Methods

pH 6.5, 5 mM of EDTA, 1 mM of TPP with addition of 1 tablet of Complete® protease inhibitor cocktail-EDTA free per 25 mL reaction.

2.8.1.2 Sampling procedure

Residual carboligase activities for each experiment and their controls (without benzaldehyde) were determined over time by withdrawing and diluting samples for the subsequent analysis. In the study investigating the putative effect of protease, samples were diluted with the same buffer in which 15 mM MgSO4 was added instead of EDTA. The dilution factor remained constant throughout the course of each reaction. The half- life values of the various PDC preparations were estimated graphically from the deactivation profiles.

2.8.2 Estimation of PDC release during whole cell biotransformation

PDC as a whole cell preparation was exposed to 310 mM benzaldehyde and 355 mM pyruvate at 21oC. For the control, the whole cell PDC was exposed to 2.5 M buffer only. Immediately after enzyme addition, representative samples were withdrawn at different time intervals (2 min, 30 min, 60 min and 90 min) and centrifuged at 20,000 g for 5 min. The clear supernatant was passed through a Biogel P-6 microfiltration column to remove interfering substances and this was followed by carboligase activity assay. Recovered PDC activity from the biotransformation supernatant was then compared with the control (no biotransformation).

2.8.3 Stability study in an aqueous/octanol-benzaldehyde emulsion system without pyruvate

2.8.3.1 Preparation procedure

PDC preparations were prepared as concentrated stock solutions in 2.5 M MOPS buffer containing 0.5 mM TPP and 0.5 mM MgSO4. The octanol phase containing 1.5

54 Chapter 2. Materials and Methods

M benzaldehyde in octanol was rapidly stirred with the aqueous phase for at least 1 h to ensure sufficient transfer of benzaldehyde into the aqueous phase. The stability study was then initiated with the addition of PDC solution into a pre-stirred aqueous/octanol- benzaldehyde reaction mixture with a final phase volume ratio of 1 to 1. PDC was therefore exposed to 1.5 M benzaldehyde in octanol and approx. 50 mM benzaldehyde in the aqueous phase.

2.8.3.2 Sampling procedure

In the absence of pyruvate, residual activities were determined by mixing 250 μL PDC-containing aqueous/octanol-benzaldehyde samples with an equal volume of the 2-fold carboligase substrate solution saturated with octanol. The reaction was stirred in a 4 ml screw capped glass vial and stopped after 20 min with 500 μL of 20% TCA (w/v). The samples were immediately centrifuged at 20,000 g for 5 min and the aqueous phase was withdrawn for PAC quantification.

2.8.4 Stability study of partially purified PDC in an aqueous/octanol-benzaldehyde biotransformation

Samples were obtained from the biotransformation reaction (Section 2.7.2.8) with partially purified PDC. The residual activity for whole cell PDC during biotransformation could not be accurately determined due to difficulties in complete PDC recovery from the associated cell debris in the aqueous phase. During biotransformation, the aqueous/octanol-benzaldehyde samples were withdrawn, centrifuged at 20,000 g for 5 min and the clear aqueous phase containing cell-free PDC was passed through a Biogel P-6 microfiltration column followed by standard carboligase assay. For the control, partially purified PDC was added into the aqueous/octanol-benzaldehyde reaction mixture without substrates. The residual activity was calculated taking the control value as 100% and the half-life was determined graphically from the corresponding Figure.

55

Chapter 3

3 CHEMICAL MODIFICATIONS OF PARTIALLY PURIFED CANDIDA UTILIS PDC Chemical modifications of the partially purified Candida utilis PDC

3.1 Introduction

Enzymatic R-phenylacetylcarbinol (PAC) production using enzyme pyruvate decarboxylase (PDC) from Candida utilis eliminates some of the problems associated with the commercial process and has been reported to achieve higher PAC concentrations, yields and productivities. However enzyme stability is still a key concern for the overall biotransformation process (Leksawasdi et al., 2003; Rosche et al., 2002b). The use of enzyme modification techniques can be an additional approach for improving PDC stability towards benzaldehyde (substrate), PAC (product), acetaldehyde and acetoin (by-products) as well as to a pH rise during biotransformation.

Earlier modification studies with yeast PDC were mostly carried out with purified enzyme from Saccharomyces species and usually involved specific amino acid replacements at or adjacent to the active site (site-directed mutagenesis, see Table 1-2) and specific cross-linking. The cross-linking of activated-PDC with bisimidates (bifunctional reagents) of different chain lengths has been reported to eliminate the lag phase in product formation, a typical characteristic with yeast PDC (Konig et al., 1990). In another study with bisimidate dimethyladipinediimidatedihydrocloride (DMAI), a 3-

56 Chapter 3. Chemical modifications

fold increase in PDC thermostability at 40oC was reported compared to that of the native enzyme (Dobritzsch et al., 1996). A different approach using selective removal of the PDC catalytic site responsible for the release of free acetaldehyde via proteolytic degradation (Juni and Heym, 1968) resulted in an improved C-C bond formation (up to 60% increase in acetoin formation) for a crude extract PDC from S. cerevisiae mutant.

Recently, a platform technology for the development of novel enzymes via chemical modification procedures has been developed at the University of New South Wales with the aim of designing enzymes specifically optimised for dedicated processes and applications. A number of novel enzymes has been developed with this technology (Cavicchioli et al., 2006; Siddiqui et al., 1997; 1999; 2004; Siddiqui and Cavicchioli, 2005), which involves the use of a range of modification techniques followed by screening for potential improvement in the modified enzyme.

The application of the above chemical modification technology to increase the stability of partially purified PDC from C. utilis is the focus in this Chapter. Different modification reactions targeting mostly the amino-group of the amino acid lysine and the carboxyl group of amino acids aspartic and/or glutamic acid were carried out by Dr Khawar Sohail Siddiqui from the School of Biotechnology and Biomolecular Sciences (UNSW). No information on the methods and the experimental designs was provided at the time of experiments due to the confidentiality of the developed technology.

57 Chapter 3. Chemical modifications

3.2 Results

In this evaluation, PDC that has not been previously exposed to modification buffers and/or reagents is denoted as native PDC, and that which has been subjected to such buffers and reagents is designated as m-PDC. After modification, the enzyme samples were then passed on to the candidate to be further screened for (1) the (recovered) activity and when feasible (2) a stability study to compare the residual activity after exposure to benzaldehyde. The carboligase activity of m-PDC was determined using a standard carboligase assay at pH 6.4 (Section 2.5.2.1). The stability was analysed by comparing the residual activities of m-PDC and native PDC after 24 h incubation in 50 mM benzaldehyde (initial activity 0.6 U/mL in 50 mM MES, 0.5 mM o TPP, 0.5 mM MgSO4, pH 6.4, 25 C, 24 h).

3.2.1 Modification targeting amino-group of lysine

3.2.1.1 Modification with polyethylene glycol

Chemical modification with polyethylene glycol (PEG) and/or its derivatives is a procedure of growing interest for enzyme enhancement in biocatalysis as well as for therapeutic use especially in drug delivery (Hernaiz et al., 1999; Roberts et al., 2002; Salleh et al., 2002; Veronese, 2001). The process involves activation of PEG molecules followed by reaction with the amino-group of the amino acid lysine. The attachment of these amphipathic polymer arms, mostly on the surface amino-groups, is likely to alter the overall polarity and surface hydrophilicity of the enzyme. PEG-modified enzymes are very useful in many organic processes due to their increased solubility in organic solvents (Inada et al., 1995; Yang et al., 1996). In this study, a tresylchloride-activated methoxypolyethylene glycol (tresylated-MPEG) was used for subsequent PDC modification. The principle of action is detailed in Figure 3-1.

58 Chapter 3. Chemical modifications

pH 7.5 O MPEG O MPEG F COHSO [Lys]E NH2 + F3C SO2 O [Lys]E NH + 3 2

Enzyme Tresylated-MPEG Modified enzyme

Figure 3-1. Chemical modification of enzyme by tresylated-methoxypolyethylene glycol (tresylated-MPEG).

Table 3-1 shows the results for PDC modification with tresylated-MPEG (PEG) at room temperature under different modification conditions. The concentration of PEG and the reaction time were varied. It was found that more than 50% m-PDC activity was lost upon incubation in buffer (control 1) and in the presence of PEG, presumably due to high pH 7.5 and chemical inactivation, respectively. When incubated in 50 mM benzaldehyde, the residual activity for m-PDC was significantly lower (< 5%) than for control 1 (50%) indicating a less stable enzyme preparation resulted from this PEG modification.

Table 3-1. Details of PDC modification with tresylated-PEG (initial activity 6 U/mL,

0.1 M K2HPO4/H3PO4, initial pH 7.5, 0.125 M NaCl, 1 mM MgCl2, room temperature and total reaction volume 1.5 mL)

Control 1 Control 2 3 4 5 6

Tresylated-PEG (g/mL) --- 0.03 (no PDC) 0.03 0.03 0.03 0.075

Reaction time (h) 3 0 1 2 3 2

Recovered PDC activity 49% --- 12% 9% 4% 9% (%) after modification

59 Chapter 3. Chemical modifications

3.2.1.2 Modification with polysaccharide

Increase in enzyme stability (against pH or temperature) can be obtained via protein modification with various polysaccharides (Ertan et al., 1997; Gomez et al., 2000; Kobayashi and Takatsu, 1994; Marshall and Rabinowitz, 1974; Masarova et al., 2001; Siddiqui and Cavicchioli, 2005; Vina et al., 2001; Yamagata et al., 1994). The successful modification is likely to increase the hydrophilicity of the protein surface as well as its rigidity due to the attachment of polysaccharide at multiple points on the enzyme. The principle of action is detailed in Figure 3-2.

OH

O O O H 1. Oxidation (NaIO4) OH OH OH OH O OH O OH OH OH OH n OH

OH

O O O H OH OH OH O OH O OH OH OH O O n OH

OH

O O O H OH OH OH O O OH OH 2. Enzyme addition E-[Lys]-NH2 OH NH OH n OH

[Lys] 3. Reduction (NaBH4 or Pyridine - Borane Complex) E

Figure 3-2. Chemical modification of enzyme by dialdehyde polysaccharide.

Table 3-2 shows the various conditions for PDC modification with oxidized polysaccharides, also known as dialdehyde polysaccharides (DAP), of different molecular weights at room temperature. The conditions were also varied for the reducing agent, pH and reaction period. Heavy precipitates were formed after 2 days in the absence of magnesium ions (one of the two cofactors for PDC) and/or in the presence of mild reducing agent Pyridine-Borane-Complex (PBC). With only traces of m-PDC activities recovered after modification (even for control in the absence of DAP), it was clear that partially purified PDC was significantly inactivated during modification, presumably due to the enzyme exposure to alkaline pH over 1 to 4 days of reaction. Similar results were observed with the use of S. cerevisiae PDC.

60 Chapter 3. Chemical modifications

Table 3-2. Details of PDC modification with dialdehyde polysaccharides/DAP

(initial activity 6 U/mL, 0.2 M K2HPO4/H3PO4 pH adjusted to 7.13 or 0.1 M boric acid/borate adjusted to pH 8.6, room temperature and total reaction volume 1.0 mL).

DAP MgCl 60 mM PBC Reaction 25 % glycine Dialysis pH 2 (% v/v) (mM) (% v/v) time (day) (mL) (h)

No

1 --- 7.13 --- 0.03 4 0.5 ---

2 0.37 7.13 --- 0.03 4 0.5 ---

3 --- 7.13 --- 0.03 a 1 --- 6

4 0.37 7.13 --- 0.03 a 1 --- 6

5 0.37 7.13 ------1 --- 6

6 --- 8.6 --- 0.03 4 0.5 ---

7 0.15 8.6 --- 0.03 4 0.5 ---

8 --- 7.13 1.5 --- 3 0.36 6

9 0.25 7.13 1.5 --- 3 0.36 6

10 --- 7.13 1.5 0.01 3 0.36 6

11 0.25 7.13 1.5 0.01 3 0.36 6

12 0.25 7.13 1.5 0.02 3 0.36 6

b 13 0.25 7.13 1.5 --- 3 0.4 6

14c 0.1 7.13 1.5 0.03 2 0.4 --- Note: condition no. 1-7 applied for the use of dextran 40 kDa, dextran 250 kDa, Ficoll 70 kDa and Ficoll 400 kDa while condition no. 8-14 applied for the use of dextran 250 kDa and Ficoll 70 kDa only. a the precipitated protein was removed by centrifugation after 2- 3 h b strong reducing agent NaBH4 (0.38 M) was added at the end of reaction. c S. cerevisiae PDC purchased from Sigma Co. was used.

61 Chapter 3. Chemical modifications

3.2.1.3 Guanidination

In other studies greater enzyme stability has been reported when the ratio of arginine to lysine was high in the protein (Cupo et al., 1980). Since the basic (guanidino) group of arginine is known to have the capacity of forming five H-bonds and/or two salt bridges, this led to the notion that arginine may contribute to protein surface interaction and increase stabilization. Evaluation of the effect of increased ionic interactions on PDC was made in the present experiments by replacing the amino-group of lysine with a guanidino-group, a procedure known as guanidination. The principle of action is detailed in Figure 3-3.

CH3 + NH CH3 N pH > 9.5 2 E-[Lys]-NH N 2 + H3C E [Lys] NH C + N NH H3C N 2 H NH NH2 Enzyme 3,5-dimethyl-1H-pyrazole-1-carboximidamide Modified enzyme 3,5-dimethyl-1H-pyrazole

Figure 3-3. Chemical modification of enzyme by 3,5-dimethylpyrazole-1- carboximidamide (DMPC)

Table 3-3 shows the details of various conditions for PDC modification with 3,5-dimethylpyrazole-1-carboximidamide (DMPC) at 2-4oC. Protein precipitates were observed in all modification samples even with the addition of MgCl2. As expected, due to the requirement of high pH at 9.4, only traces m-PDC activities were recovered after modification. Exhaustive dialysis was necessary to avoid false positive results as a peak with a retention time close to that of product PAC appeared during HPLC analysis of a carboligase reaction sample.

62 Chapter 3. Chemical modifications

Table 3-3. Details of PDC guanidination with 3,5-dimethylpyrazole-1- carboximidamide (DMPC) (initial pH 9.4/NaOH and a total reaction volume 1.5 mL)

No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14a 15a

Conditions

Temperature (oC) 2-4 2-4 25 2-4 2-4

PDC (U/mL) 6 0 2 6 6 6 6 6

DMPC (M) 0.55 0.5 0.25 0.5 0.5 0.25

MgCl2 (mM) 5 5 0 5

Reaction time (day) 1 2 3 4 4 5 4 1 1

Dialysis (h) 6 6 6 6 2 6 ------2 -- 6 6 6 -- a S. cerevisiae PDC purchased from Sigma Co. was used

3.2.1.4 Acylation

Amino-groups of proteins can be readily acylated under conditions wherein other groups either do not react at significant rates or the products so formed are labile (Means and Feeney, 1971). The bond formations due to acylation reactions of the side chain groups of amino acids cysteine, histidine, tyrosine, serine and threonine are not very stable and can be removed by neutral hydroxylamine while the acylation of the amino-groups of lysine can give a stable amide bond. The pH at which acylation can be carried out depends greatly upon the range of pH values in which the amino-groups tend to be protonated. In this study, acetic anhydride (AA), succinic anhydride (SA) and pyromellitic dianhydride (PA) were used. While the presence of an acetic group results in a neutral hydrophobic residue of lysine, the addition of succinic and pyromellitic groups converts it into an anionic residue. The principle of action is detailed in Figure 3-5.

63 Chapter 3. Chemical modifications

O O O O pH > 8 (a) E [Lys] NH2 + + 3CH OCH3 [Lys]E NH CH3 3CH OH Enzyme acetic anhydride Modified enzyme acetic acid

O O pH > 8 HOOC (b) E [Lys] NH O - 2 + E [Lys] NH COO + COOH

O Enzyme succinic anhydride Modified enzyme succinic acid

O O O - - pH > 8 COO HOOC COO [Lys]E NH (c) E [Lys] NH2 + O O + - - O O OOC COOH OOC COOH

Enzyme Pyromellitic dianhydride Modified enzyme Pyromellitic acid

Figure 3-4. Chemical modification of enzyme by: (a) acetic anhydride (AA) (b) succinic anhydride (SA) and (c) pyromellitic dianhydride (PA).

Table 3-4 shows the PDC modification conditions with various anhydrides in pH 7.13 or 8.6 conditions. Only traces of m-PDC activities were recovered from all types of anhydrides modification, with or without addition of 1 mM thiamine pyrophosphate (TPP, another PDC cofactor) during the course of modification.

Table 3-4. Details of PDC modification with anhydrides (initial activity 6 U/mL, 0.2 M

K2HPO4/H3PO4 adjusted to pH 7.13 or 0.1 M Na2HPO4/NaH2PO4/CH3COONa adjusted to pH 8.0 or 0.1 M boric acid/borate adjusted to pH 8.6, 1 mM MgCl2, room temperature, total reaction volume 1.5 mL)

No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Conditions

Anhydrides AA SA PA concentration(mM) 14 28 28 7.5 12 15 15 20 12 13 12 20 12 13

Reaction pH 7.13 7.13 8.6 7.13 8.6

TPP (mM) 0 0 1 0 0 0 1 0 0 0 0 0 0 0

64 Chapter 3. Chemical modifications

3.2.2 Modification targeting carboxyl-group of aspartic and glutamic acids

3.2.2.1 Amidation

Carbodiimide is often used to activate carboxylic acids to form an amide or an ester and the same principle can also be applied for modification of protein carboxyl groups. The use of carbodiimide for carboxyl-group modification has been reported in several studies for different enzymes such as amylase (Kochlar and Dua, 1984), carboxymethycellulase (Siddiqui et al., 1997; Siddiqui et al., 1999), glucoamylase (Munch and Tritsch, 1990), lysozyme (Lin and Koshland, 1969) and putidaredoxin (Geren et al., 1986).

In this study, the carboxyl-groups were activated by 1-ethyl-3(3- dimethylaminopropyl) carbodimide (EDC) followed with the addition of aromatic nucleophiles (aniline and guanidine-2-benzimidazole chloride (GB)) and aliphatic nucleophiles (arginine methylester (ArgMetEst) and chitosan 5,000). EDC is a water- soluble derivative of carbodiimide. Carbodiimide catalyses the formation of amide bonds between carboxylic acids and amines by activating the carboxyl-groups to form an amine-reactive intermediate (O-acylisourea). This intermediate is unstable in the aqueous phase and reacts readily with a nucleophile (Figure 3-5). The overall effects on enzyme surface characteristics expected from the replacement of the negative charge of carboxyl-groups with various nucleophiles side chains are listed in Table 3-5.

+ O O N R1 H N R1 + pH 4 - 6 E [Asp/Glu] O C E [Asp/Glu] OH N R + 2 N R2 Enzyme Carbodiimide O-acylisourea

.. O 2NH R3 NH R1 + E [Asp/Glu] NH R + O + H 3 pH 4 - 6 NH R2 Modified Enzyme

Figure 3-5. Chemical modification of enzyme by carbodiimide-promoted amide formation.

65 Chapter 3. Chemical modifications

Table 3-5. The expected effect on enzyme surface characteristics upon the replacement of the carboxyl-groups with various side chains.

Expected effect on enzyme Nucleophiles Side chain (R ) 3 surface characteristic

Aniline .. Hydrophobic 2NH

+ N guanidine-2- NH2 Hydrophilic .. benzimidazole 2NH NH N CH3 H (increased ionic interaction)

O O CH3 Arginine + .. NH NH2 Increased ionic interaction methylester 2NH NH2

HOH C OH NH 2 + 2 O Hydrophilic Chitosan 5,000 OH O OH + n .. O (glycosylation) OH NH2 HOH2C

Table 3-6 shows the details for PDC modification via EDC-promoted amide formation. 0.9% (w/v) of EDC was added into the modification reaction to activate the carboxyl groups of proteins. After 20-60 min period, only traces of m-PDC activities were recovered even though the modification reactions were carried out at a milder pH condition for PDC (approx. pH 5.2) and in the presence of cofactors TPP and magnesium.

Table 3-6. Details of PDC modification with carbodiimides (initial activity 6 U/mL,

0.2 M K2HPO4/H3PO4 adjusted to pH 5.2, 2 mM MgCl2, 0.5 mM TPP, 0.9 % (w/v) EDC, room temperature, total reaction volume 2.0 mL)

No. 1 2 3 4 5 6 Conditions

Initial pH 5.18 5.18 5.18 5.15 5.15 5.25

Carbodiimide Aniline Arg.Met.Est. Chitosan GB

concentration 25 mM 1 M 0.2 % (w/v) 50 mM

Reaction time (min) 30 20 60 20 60 30

66 Chapter 3. Chemical modifications

3.3 Discussion and conclusions

Enzymatic biotransformation for PAC production with partially purified C. utilis PDC has many advantages over the commercial fermentative process with S. cerevisiae cells and a patent application on this process had recently been filed (Hauer et al., 2003). Improved PDC characteristics, e.g. higher stability towards substrate benzaldehyde while maintaining PAC productivity, would be valuable for the overall production.

In these preliminary studies, different conditions were tested to identify possible enzyme modification for improved PDC stability. In the absence of urea as a chaotrophic agent, only fully and partially exposed amino acids were targeted for modification reactions thus, avoiding the buried catalytic site. However no enzyme enhancement was achieved using this strategy. The results showed very low or no recovered m-PDC activity after modifications and a reduced stability for m-PDC (Section 3.2.1.1) in comparison to the native partially purified PDC. Possible factors which may account for these significant losses of PDC activity are: (1) the alkaline pH conditions required in amino-group modifications. The active tetrameric PDC dissociates reversibly into inactive dimeric subunit at pH above 7 (Gounaris et al., 1971), (2) the inactivating effects of the modification, perhaps due to reagent interaction with the active centre or due to a change of the overall enzyme conformation, (3) the particular amino acids and side chains selected for modification (e.g. those of lysine, aspartic and glutamic acids). Other authors have reported that lysine has no specific catalytic or regulatory function in the PDC structure (Konig et al., 1990), however aspartic and glutamic acids have a vital role in the active centre (see Table 1-2).

It is recommended following from the results of the present study that a mass spectroscopy analysis be further used to quantify and analyse the role of specific amino acids in influencing the activity and stability of C. utilis PDC. This knowledge will be valuable in determining the conditions e.g. the optimal concentrations of modifying agents and the most suitable sites on the enzyme for subsequent modification. Use of

67 Chapter 3. Chemical modifications

other protein stabilization techniques that are more suitable with the less purified enzyme preparations such as cross linked enzyme aggregation (CLEA) is also a further possibility for enhancement of PDC activity and stability under biotransformation conditions.

68

Chapter 4

4 ENZYMATIC R-PAC PRODUCTION IN AN AQUEOUS/ BENZALDEHYDE EMULSION SYSTEM WITH C. UTILIS WHOLE CELL PDC Enzymatic R-PAC production in an aqueous/benzaldehyde emulsion system with C. utilis whole cell PDC

4.1 Introduction

Values of 4−22 g PAC/L have been reported for fermentative biotransformations with various yeast species (Becvarova and Hanc, 1963; Netrval and Vojtisek, 1982; Rogers et al., 1997) and up to 50 g/L for enzymatic processes with partially purified PDC from C. utilis or R. javanicus based on added pyruvate and a benzaldehyde emulsion (Rosche et al., 2002a; 2003a).

The enzyme processes based on cell-free PDC, however, result in appreciable PDC deactivation by substrate benzaldehyde (Leksawasdi et al., 2003). For an improved biotransformation, lowering the temperature from 25 to 4oC extends PDC life span and increases final PAC concentration (Rosche et al., 2002a). Cell-free biotransformation thus involves a high cost of catalyst preparation and cooling to extend enzyme life.

The use of whole cells as an alternative biocatalyst preparation in an enzymatic process has the advantage of simple and economical preparation as expenses for cell

69 Chapter 4. Aqueous/benzaldehyde emulsion system

breakage and/or enzyme purification are saved. A further reported advantage is greater biocatalyst stability and protection within the intracellular environment especially in organic media, although some limitations in mass transfer of substrates and products may be associated with whole cell biocatalysis (Leon et al., 1998).

Previous enzymatic PAC production with free cells of S. cerevisiae resulted in by-product benzylalcohol formation (0.1-0.5 g/L) with only 1-3 g PAC/L produced (Nikolova and Ward, 1991; 1994). In a further study with cells of S. cerevisiae, 5 g benzylalcohol/L and 15 g PAC/L were reported (Mochizuki et al., 1995). The latter authors maintained the biotransformation at pH 6, 30oC with progressive addition of low concentrations of benzaldehyde. Benzylalcohol formation has been reported also for PAC production by cells immobilized on polymer matrices, alginates or cyclodextrin (Mahmoud et al., 1990c; Nikolova and Ward, 1994; Rogers et al., 1997).

In former studies of enzymatic PAC production in an aqueous/benzaldehyde emulsion system, C. utilis PDC has been added as a cell free preparation (Rosche et al., 2002a; 2003a; 2005a; Shin and Rogers, 1996a). In this Chapter, the use of PDC initially added in the form of C. utilis cells (whole cell PDC) is evaluated for its stability and for the enzymatic production of PAC in an aqueous/benzaldehyde emulsion at 4o and 21oC. The results are compared with those for crude extract and partially purified PDC preparations and are used to determine whether or not a whole cell PDC process offers any potential advantages in this system.

70 Chapter 4. Aqueous/benzaldehyde emulsion system

4.2 Results

4.2.1 PDC stability towards benzaldehyde

During biotransformation, continuous exposure to toxic substrate benzaldehyde results in irreversible PDC deactivation (Goetz et al., 2001; Iwan et al., 2001; Long and Ward, 1989). In addition, proton uptake during biotransformation increases the pH above 7, which is detrimental to PDC as the tetrameric PDC dissociates into inactive dimeric subunits at higher pH with a simultaneous release of cofactor TPP (Gounaris et al., 1971).

Earlier studies reported that maintaining the pH with high concentration of MOPS buffer increased PDC stability and significantly enhanced PAC production (Rosche et al., 2002a). The authors suggested that the high concentration MOPS buffer (2.5 M) possibly acts as an osmolytic additive that increases the affinity between enzyme subunits thereby stabilizing the overall protein structure, as achieved also via enzyme cross-linking/modification. These optimized conditions were used also in a kinetic deactivation study of partially purified PDC by Leksawasdi et al. (2003). In this reported study, a reaction time deactivation constant of 2.64 x 10-3/h and a benzaldehyde deactivation coefficient of 1.98 x 10-4 /(mM.h) were determined by exposing partially purified C. utilis PDC (7 U/mL) to 0-200 mM benzaldehyde.

While use of expensive MOPS buffer had been successfully replaced by the addition of glycerol (Rosche et al., 2005a) or dipropylene glycol (Leksawasdi et al., 2005) in combination with a pH-controller, a more stable catalyst still need to be identified to further improve PAC productivity.

4.2.1.1 Effect of different purification grades on deactivation of C. utilis PDC

Purification is essential where a single enzyme is extracted from cells/tissues in order to have a specific role as a biocatalyst or for therapeutic use. The degree of purity required depends on the nature of the application. PDC recovered from the yeast cells

71 Chapter 4. Aqueous/benzaldehyde emulsion system

can be purified through several steps (Section 2.3.2) and enzyme conformation as well as the activity/stability may be affected. Direct comparison of C. utilis PDC stability for different purification grades is reported in this Section.

Based on the possibility that whole cell PDC activity might be underestimated in analysis, PDC activity was initially assayed with 2 approaches viz. without treatment and with 5-times repeated freeze-thawing, before being subjected to carboligase activity assay. The second approach was intended to provide greater access of substrates into the cells after their physical disruption through the repeated formation of sharp ice crystals. However it was observed that this procedure did not result in any significant difference in PDC activity values when compared to the first approach especially for samples with benzaldehyde present (data not shown). Thus all reported values were based on the earlier approach as a standardized procedure.

In Figures 4-1 and 4-2 the time profiles of residual PDC activities in the absence and presence of 50 mM benzaldehyde are compared for a partially purified, a crude extract and a whole cell preparation at 4o and 21oC (see Section 2.8.1 for Materials and Methods). From the data it is evident that PDC activity was maintained better in the whole cell preparation. Furthermore deactivation occurred more rapidly for all preparations at the higher temperature. The deactivation kinetics revealed not only differences in stability but also in the curve shapes as not all deactivation profiles followed an approximately exponential decay.

The half-life values were estimated graphically from the profiles and the values are given in Table 4-1. Exposure to benzaldehyde and increasing the temperature from 4o to 21oC reduced PDC stability in all cases with greatest impact on the more purified PDC. A half-life of 9.5 days was estimated for whole cell PDC in comparison to 1 day for the partially purified preparation at 4oC in the presence of 50 mM benzaldehyde. When the temperature was increased to 21oC, the half-life for whole cell PDC decreased by 42%. This compared with an 85% decrease for the partially purified preparation.

72 Chapter 4. Aqueous/benzaldehyde emulsion system

a 4oC 100%

50% Residual PDC activity

0% 0 6 12 18 24 30 36 Time (days)

b 21oC 100%

50% Residual PDC activity

0% 0 6 12 18 24 30 36 Time (days)

Figure 4-1. Profiles of residual activity of partially purified, crude extract and whole cell PDC at (a) 4oC (closed symbols) and (b) 21oC (open symbols) respectively in absence of

benzaldehyde (initial activity 1.5 U/mL, 2.5 M MOPS, pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP). The mean values were obtained from duplicate analyses of two vials. The error bars show lowest and highest values.

73 Chapter 4. Aqueous/benzaldehyde emulsion system

a 4oC 100%

50% Residual PDC activity

0% 0 2 4 6 8 10 12 Time (days)

b 21oC 100%

50% Residual PDC activity

0% 024681012 Time (days)

Figure 4-2. Profiles of residual activity of partially purified, crude extract and whole cell PDC at (a) 4oC (closed symbols) and (b) 21oC (open symbols) respectively in the presence of 50 mM benzaldehyde (initial activity 1.5 U/mL, 2.5 M MOPS, pH 6.5, 0.5 mM

MgSO4, 0.5 mM TPP). The mean values were obtained from duplicate analyses of two vials. The error bars show lowest and highest values.

74 Chapter 4. Aqueous/benzaldehyde emulsion system

Table 4-1. The estimated half-lives of different preparations of C. utilis PDC from data in Figures 4-1 and 4-2.

Half-life (day) at 4oC Half-life (day) at 21oC Benzaldehyde concentration 0 mM 50 mM 0 mM 50 mM Partially purified 7 1.0 0.6 0.2 Crude extract 20 2.5 5.5 2 Cells 25 9.5 13.5 5.5

4.2.1.2 Comparison to other yeast PDCs

In a screening of 105 yeast strains from 10 genera and 40 species, it was found that the highest PDC carboligase activities combined with resistance to 40 mM benzaldehyde and 30 mM acetaldehyde for 1 h at 25oC was obtained with C. utilis and C. tropicalis (Rosche et al., 2003b). The authors noted also that the best resistance was obtained with PDC from Schizosaccharomyces pombe even though the carboligase activity was very low.

It is therefore interesting to compare PDC of C. utilis to other yeast PDCs with high aldehyde resistance or other interesting characteristics. This study compares the stability of whole cell PDC from C. utilis to that of 3 other yeast strains viz. the high aldehyde resistant strain C. tropicalis, the commercial strain S. cerevisiae and a thermotolerant strain K. marxianus. Previous reports on yeasts belonging to genus Kluyveromyces had been promising with ethanol (PDC-associated product) production up to 52oC (Banat and Marchant, 1995; Banat et al., 1998). Yeast S. pombe and a pyruvate producing strain of C. glabrata were excluded from the stability study due to insufficient amounts of PDC produced in shake flask cultures. For the purpose of comparison, the selected yeast strains were all grown in the same medium containing yeast extract (see Table 2-1). The specific PDC activities obtained for different yeast strains were 121, 254, 106 and 151 U/g DCM for C. utilis, C. tropicalis, K. marxianus and S. cerevisiae respectively.

75 Chapter 4. Aqueous/benzaldehyde emulsion system

As demonstrated in Figure 4-3, in the absence of benzaldehyde at 21oC, whole cell PDC of S. cerevisiae and C. utilis were very stable with half lives of more than two weeks. In contrast, C. tropicalis PDC was very unstable with half life of less than 1 day. The PDC half-lives of all yeast strains were reduced in the presence of 50 mM benzaldehyde, however the stability trend between the strains remained the same. C. utilis PDC was the most stable with a half life of 7 days while C. tropicalis PDC was the least stable with a half life of less than 1 day. It is also interesting to note that the second most stable PDC (S. cerevisiae) did not show an initial increase in activity before subsequent deactivation as was commonly observed in the initial period for the deactivation profile for C. utilis PDC (see Figure 4-2).

C. tropicalis

K. marxianus

S. cerevisiae

C. utilis

0 4 8 12 16 Half-life (days)

Figure 4-3. Estimated half-lives of whole cell PDC of four yeast strains at 21oC in the absence (white bars) and presence (black bars) of 50 mM benzaldehyde (initial activity 1.5

U/mL, 2.5 M MOPS, pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP).

4.2.2 Biotransformation studies

4.2.2.1 Whole cells versus cell-free PDC

In light of the higher stability of whole cell C. utilis PDC (Table 4-1), it is therefore interesting to further investigate its potential for PAC production. These results will be compared to those of the more purified PDC such as crude extract and partially purified preparations (see Section 2.7.1.2 for Materials and Methods). In

76 Chapter 4. Aqueous/benzaldehyde emulsion system

Figure 4-4, the effects of PDC preparation (3 U/mL) on initial PAC formation at 4o and 21oC with the different PDC preparations are shown. The profiles revealed that initial PAC production increased with the rise of temperature. Within a 1-hour reaction period, all PDC preparations produced at least 3-fold higher PAC concentration at 21oC.

30

20 PAC (g/L) 10

0 0123 Time (h)

Figure 4-4. Initial 1-3 h PAC production by partially purified, crude extract and whole cell PDC in an aqueous/benzaldehyde emulsion system at 4oC (solid lines) and 21oC

(broken lines) (initial activity 3 U/mL, 2.5 M MOPS, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 300 mM benzaldehyde, 360 mM pyruvate). The mean values were obtained from triplicate analyses of two biotransformations. The error bars show the lowest and highest values.

The summary of biotransformation by these different PDC preparations is further given in Table 4-2. It is apparent that similar PAC concentrations were achieved after 24 h by all PDC preparations regardless of the temperature. The use of whole cell PDC at 21oC resulted in a higher level of acetoin (32 mM) after 24 h than the partially purified PDC preparation (15 mM). No benzylalcohol and lower levels of acetaldehyde (5-11 mM) were formed. Table 4-3 shows that repeated freeze thawing of whole cells to increase cell permeability and/or release intracellular PDC did not result in any significantly higher PAC concentrations, however addition of 2 M ethanol resulted in a reduction in PAC production (possibly PDC deactivation particularly at higher temperature).

77 Chapter 4. Aqueous/benzaldehyde emulsion system

Table 4-2. Summary of biotransformation data for different PDC preparations in an aqueous/benzaldehyde emulsion system after 24 h (see also Figure 4-4).

Part. Crude Whole cell Components purified extract 4oC 21oC 4oC 21oC 4oC 21oC

Initial PAC (mM) a 3.8 11.0 3.7 12.1 4.9 16.8

PAC (mM) 254 260 254 233 243 228

Acetaldehyde (mM) 4.7 10.8 4.8 11 4.7 11.1

Acetoin (mM) 2.0 15.2 2.6 28.5 1.8 31.6

Benzylalcohol (mM) 0 0 0 0 0 0

Specific PAC production (mg/U) b 12.7 13.0 12.7 11.6 12.1 11.4

Benzaldehyde unaccounted (%, molar basis) 6 1 11 4 6 12

Pyruvate unaccounted (%, molar basis) 4 2 3 2 8 15 a based on PAC production in 1 hour b based on initial units of PDC carboligase activity

Table 4-3. Effect of treatment for whole cell C. utilis PDC on PAC production in an aqueous/benzaldehyde emulsion system over 1 and 24 h period (initial activity 3 U/mL, 2.5

M MOPS/KOH, pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 300 mM benzaldehyde, 360 mM pyruvate) (see Appendix B for initial 1-3 h PAC formations)

Temp. (oC) Cell treatment PAC 1 h (mM) Maximum PAC (mM) 4 No treatment 4.9 243 Repeated freeze-thaw 4.3 260 Addition of 2 M ethanol 4.7 185 21 No treatment 16.8 228 Repeated freeze-thaw 17.0 239 Addition of 2 M ethanol 5.9 34.2

78 Chapter 4. Aqueous/benzaldehyde emulsion system

4.2.2.2 Kinetics of PAC formation with whole cell PDC

Detailed kinetics for whole cell biotransformations at 4o and 21oC with initial concentrations of 300 mM benzaldehyde and 364 mM pyruvate in an aqueous/benzaldehyde emulsion system (see Section 2.7.1.3 for Materials and Methods) are shown in Figure 4-5. The profiles demonstrate faster initial PAC production at 21oC than at 4oC as would be expected. At 21oC, a maximum PAC concentration of 43 g/L (289 mM, ee 94 %) was achieved after 12 h in comparison to 39 g/L (258 mM, ee 98 %) after 30 h at 4oC. However a 4-fold higher acetoin level (1.6 g/L, 18 mM) was evident at 21oC when compared to the results at 4oC (0.4 g/L, 4.4 mM) at these times. The remaining time-profiles up to 48 h showed a small increase in PAC concentration at 4oC while a gradual PAC loss was noted at 21oC. In the latter case, approximately 13% PAC was lost between 20 h and 48 h with the ee value for R-PAC decreasing from 94% to 79%.

A comparison of the whole cell biotransformations at 4o and 21oC is given in Table 4-4. The results show a 12% higher maximum PAC concentration at 21oC at 12 h than at 4oC at 30 h and a 2.8-fold increase in overall volumetric productivity at the higher temperature. However, increased formation of the by-product acetoin at 21oC lowered the product yield by 7% (based on consumed substrate pyruvate). The molar balances for benzaldehyde and pyruvate were determined from the data at the times of maximum PAC concentrations for the respective temperatures. Less than 3% of substrate pyruvate and 1% benzaldehyde remained unaccounted for.

79 Chapter 4. Aqueous/benzaldehyde emulsion system

a 400 80 4oC

300 60

200 40 Pyruvate (mM)

PAC, Benzaldehyde, 100 20 Acetaldehyde, Acetoin (mM)

0 0 0 1020304050 Time (h)

b 400 80 21oC

300 60

200 40 Pyruvate (mM)

PAC, Benzaldehyde, 100 20 Acetaldehyde, Acetoin (mM) Acetoin Acetaldehyde,

0 0 0 1020304050 Time (h)

Figure 4-5. Profiles of PAC production with PDC added in the form of whole cells of C. utilis in the aqueous/benzaldehyde emulsion system at (a) 4oC and (b) 21oC (initial

activity 3 U/mL, initial pH 6.5, 2.5 M MOPS, 0.5 mM MgSO4, 0.5 mM TPP, 300 mM benzaldehyde, 360 mM pyruvate). The mean values were obtained from triplicate analyses of two biotransformations. The error bars show lowest and highest values.

80 Chapter 4. Aqueous/benzaldehyde emulsion system

Table 4-4. Kinetic analysis of biotransformations with PDC added in the form of whole cells as detailed in Figure 4-5

7.9 g DCM/L (3 U/mL)

4oC, 30 h 21oC, 12 h

Initial production rate (g/L/h)a 4.7 15.5 Maximum PAC concentration (g/L) 38.7 43.3 Specific PAC production (mg/U) b 12.9 14.4 Specific PAC production per biomass (g/g DCM) c 4.9 5.5 Overall volumetric productivity (g/L/h) 1.3 3.6 Acetoin (g/L) 0.4 1.6 Acetaldehyde (g/L) 0.3 0.6 Molar yield (%) PAC on pyruvate d 91 84 Molar yield (%) PAC on benzaldehyde d 98 99 Pyruvate unaccounted (%, molar basis) 3 2 Benzaldehyde unaccounted (%, molar basis) 1 0 Enantiomeric excess for R-PAC (%) 98 94 a based on PAC production in 1 hour b based on initial PDC carboligase activity c based on the initial amount of biomass added d molar yield (mole/mole) based on consumed substrates

4.2.2.3 Factor affecting a decrease of R-enantiomer of PAC

To investigate further the decrease in R-enantiomer of PAC with an extended biotransformation period, a study was carried out by incubating PAC racemic standard at pH 7.5 for 28 h at 21oC in the presence of whole cell and partially purified PDC. An enzyme-free condition was used as a control. The results showed that the ee value for R- PAC decreased from 98% to 82% in the presence of C. utilis cells, while the ee values with the partially purified PDC or in the absence of PDC under similar conditions remained high at 95% and 97% respectively. This suggests that components associated

81 Chapter 4. Aqueous/benzaldehyde emulsion system

with whole cell PDC may reduce the enantiomeric quality of the R-PAC over time, and indicates the importance of terminating the biotransformation at maximum R-PAC production.

4.3 Discussion and conclusions

From the results it is evident that the use of PDC in the form of C. utilis cells is preferable to cell-free (crude extract or partially purified) PDC preparations for PAC production as the former preparation achieved similar PAC levels, involved a simpler preparation and demonstrated greater stability.

Comparison of the deactivation profiles and half-lives (Figure 4-1, Figure 4-2 and Table 4-1) in the presence of benzaldehyde demonstrated that PDC stability decreased with the greater degree of enzyme purification especially when the temperature was increased from 4o to 21oC. It is possible that the biomass matrix may provide some support for improved stability or that the conformation or the micro- environment of the cell-free PDCs may have been altered during cell attrition, acetone precipitation or freeze-drying thereby resulting in a less stable enzyme. In the support of the present results, higher stabilities of trehalosyl dextrin-forming and trehalose-forming enzymes in whole cell Escherichia coli (t1/2 = 30 days) have been reported also in comparison to the purified preparations (t1/2 = 4 days) (di Lernia et al., 2002).

A possible disadvantage of using whole cells as biotransformation catalysts is putative mass transfer limitation of substrates and products across the cell envelope. However in the present study, data for PAC production over the initial hour (Figure 4-4) demonstrated higher levels for whole cells than for cell-free preparations. Furthermore, data from this study indicated cell permeabilization because 0.8 – 1.0 U PDC/mL were measured in the extracellular biotransformation medium (27-33% of the initially added PDC) after 1 h reaction. The addition of 2 M ethanol into the biotransformation system to improve cell permeabilization appeared to be detrimental to PDC activity at the higher temperature (21oC) where PAC production ceased after only 1 h biotransformation period (Appendix B) although in a previous study, PDC was shown

82 Chapter 4. Aqueous/benzaldehyde emulsion system

to be resistant to denaturation by ethanol at concentrations up to 150 g/L (3.3 M) (Scopes, 1989).

The highest published PAC productivity achieved for whole cell biotransformation using bakers’ yeast was 2.5 g/L/h (15 g/L in 6 h at 30oC) (Mochizuki et al., 1995). In the present study, the overall productivity of C. utilis cells was 3.6 g PAC/L/h (43 g/L in 12 h at 21oC). Biotransformation at the higher temperature was also found to be very significant in achieving higher productivities, as the results with whole cell PDC at 21oC were 3-fold higher than at 4oC (Figure 4-4, Table 4-2). In contrast, Shin and Rogers (1996a) found that for partially purified C. utilis PDC with a different set of conditions (different method for PDC preparation, 70 mM pyruvate, 70 mM benzaldehyde, 40 mM phosphate buffer), a similar temperature increase resulted in an approximate 50% decrease in overall productivity after 6 h. The same study also reported an increase in acetoin formation at the higher temperature which is consistent with the results of the present study.

In comparison to biotransformation at 4oC, a lower ee value for (R)-PAC was apparent at 21oC following completion of the biotransformation (20 – 48 h time period). A similar trend of decreased ee with increased temperature and exposure time had been reported in the biotransformation of benzaldehyde and HCN into R-mandelonitrile (Willeman et al., 2002). The authors suggested that the competing rates of nonenzymatic racemization and enzymatic reaction for the desired product were the main factors in determining the final ee value for this reaction. In the present study, the decrease in the ee value over time at 21oC occurred only in the presence of whole cells but not in an enzyme free control or with partially purified PDC, which indicates that cell-associated factors could have an influence, in the longer term, on enantiomeric purity of PAC.

It was observed also that PAC concentration decreased gradually after reaching a maximum at 12 - 20 h at 21oC (Figure 4-5b). The loss was not observed with the cell- free PDC. This observation could explain the higher values for the unaccounted benzaldehyde with whole cell PDC after 24 h reaction at 21oC in the preliminary biotransformation study (Table 4-2). HPLC analysis showed that the loss was

83 Chapter 4. Aqueous/benzaldehyde emulsion system

concurrent by the appearance of a peak indicating the presence of a compound which eluted immediately after benzaldehyde. Further analysis such as mass spectroscopy of the unknown peak would be needed to confirm the composition of this compound and the associated formation mechanism. One possibility was suggested by the study of Mochizuki et al. (1995) which reported a 95% conversion of PAC to PAC-diol (1- phenyl-1,2-propanediol; commercially unavailable) within a 13 h biotransformation period using bakers’ yeast under controlled pH 7.5 at 30oC. However, the presence of PAC-diol was not confirmed in our present investigation.

The use of C. utilis cells in the present experiments did not result in any benzylalcohol formation nor did this occur with the crude extract or with partially purified PDC. This is in contrast to the results of other enzymatic whole cell studies with S. cerevisiae (Mochizuki et al., 1995; Nikolova and Ward, 1991; 1994) in which appreciable benzylalcohol production occurred. It is possible in the present investigation that the relatively high initial benzaldehyde concentration (300 mM) used and the PAC concentration (33-127 mM) formed within the first hour (Figure 4-5) may have inhibited any associated formation of benzylalcohol. Mochizuki et al. (1995) indicated that PAC can act as an inhibitor of alcohol dehydrogenase (ADH) activity and in their study the formation of benzylalcohol ceased at approximately 33 mM PAC (5 g/L). Another study reported a 55% loss of ADH activity in 6 h after exposing bakers’ yeast to 13 mM PAC (2 g/L) at 30oC (Long and Ward, 1989).

In conclusion, the potential of using C. utilis cells for PAC production in an aqueous/benzaldehyde emulsion is evident from the present data. PDC added in form of whole cells showed higher stability particularly at 21oC when compared to the crude extract or partially purified PDC preparations. With no apparent mass transfer limitations, a concentration of 43 g PAC/L was achieved in 12 h with 1.6 g by-product acetoin/L. No benzylalcohol formation was evident.

84

Chapter 5

5 USE OF WHOLE CELL PDC IN A RAPIDLY STIRRED AQUEOUS/OCTANOL-BENZALDEHYDE EMULSION SYSTEM Use of whole cell PDC in a rapidly stirred aqueous/octanol- benzaldehyde emulsion system

5.1 Introduction

The particular focus in this Chapter is the use of a whole cell enzyme in an aqueous/organic (two-phase) system where the organic phase is added to the aqueous phase (Ballesteros et al., 1995; Biselli et al., 1995; Klibanov et al., 1977). The choice of solvent as a second phase is crucial as the presence of an unsuitable water-organic mixture may compromise the enzyme (Schiffer and Dotsch, 1996) resulting in decreased activity and enzyme deactivation. The main advantages of a two-phase system are improved solubility of a non-polar substrate in an organic phase, potential product extraction into this phase and protection of the enzyme in the aqueous phase, thus enhancing volumetric productivity and product recovery.

Recently, an aqueous/organic bioreactor was developed by our group where major concentrations of substrate benzaldehyde and product PAC are partitioned into the organic (octanol) phase while PDC and pyruvate are located in the aqueous phase to minimize PDC deactivation (Rosche et al., 2002b; Sandford et al., 2005). Under a rapidly stirred condition, the octanol phase containing a high concentration of benzaldehyde is dispersed as droplets and thus the bioreactor functions as an aqueous/octanol-benzaldehyde emulsion system. Figure 5-1 provides an illustration to

85 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

differentiate the state of the previously studied aqueous/benzaldehyde emulsion system (Chapter 4) from the present aqueous/octanol-benzaldehyde emulsion system (Chapter 5 and 6).

Figure 5-1: Biotransformation of pyruvate and benzaldehyde by enzyme PDC: an illustration of the benzaldehyde emulsion and the rapidly stirred two-phase systems.

The potential of a rapidly stirred aqueous/octanol-benzaldehyde biotransformation was reported in a previous study with partially purified C. utilis PDC by Sandford et al. (2005). With a phase ratio of 1:1 at 4oC and an initial overall PDC carboligase activity of 4.3 U/mL (8.5 U/mL added into the aqueous phase); the results are summarized: (1) an appreciably higher PAC concentration was achieved (141 g/L and 18 g/L in the octanol and aqueous phases respectively) when compared to that achieved in the benzaldehyde emulsion system at 4oC (51 g/L, initial PDC activity of 8.4 U/mL). (2) a 7-fold higher volumetric productivity (1.6 g/L/h) was obtained in comparison to that for a phase-separated aqueous/octanol-benzaldehyde system at 4oC (3.8 U/mL added into the aqueous phase). This is an important process advantage even though the PDC stability was lower than that in the phase-separated aqueous/octanol-benzaldehyde system. The decreased PDC stability in the former system is likely to have resulted from the greater interfacial contact

86 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

between PDC and the dispersed octanol droplets containing benzaldehyde in the rapidly stirred aqueous/octanol-benzaldehyde bioreactor.

Studies reported in Chapter 4 have demonstrated the potential of C. utilis PDC added in the form of resting cells (whole cell PDC) to replace partially purified PDC for PAC production in an aqueous/benzaldehyde emulsion system (Satianegara et al., 2006a). Whole cell PDC was very stable in comparison to the cell-free (crude extract and partially purified) preparations at both 4o and 21oC. A half-life of 9.5 days was reported for whole cell PDC in comparison to 1 day for partially purified PDC at 4oC and 50 mM benzaldehyde (Satianegara et al., 2006a). Enzymatic biotransformation with whole cell PDC also maintained the high enantiomeric excess for R-PAC as well as resulting in no formation of benzylalcohol (Satianegara et al., 2006a).

A preliminary study with whole cell PDC at 21oC in the rapidly stirred aqueous/octanol-benzaldehyde system (Rosche et al., 2005b) achieved significant increases in specific PAC production (g/g DCM) and productivity (g/g DCM/h) based on initial biomass added (5.6 g DCM/L) compared to the commercial process using cells of S. cerevisiae at 10 g DCM/L (estimated from literature). Previous use of S. cerevisae cells in organic solvents has resulted in low PAC production and significant PDC inactivation: low yield of PAC (~1 g/L) as well as the formation of a low level (<0.5 g/L) of benzylalcohol were reported (Nikolova and Ward, 1992a,b; Kostraby et al., 2002).

The present Chapter focuses on a detailed investigation of whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde system. The effect of conditions such as initial PDC activity, cell concentration, MOPS concentration and initial pH were investigated for biotransformations at 21oC. The kinetics were evaluated in detail also for a whole cell PDC preparation at a selected initial activity with and without pH control.

87 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

5.2 Results

5.2.1 Comparative studies on different PDC preparations in an aqueous/octanol-benzaldehyde system

5.2.1.1 PAC and by-product formation with whole cell versus cell- free PDC

The influence of the difference in PDC preparations on production of PAC and by-products is shown in Table 5-1 for a partially purified, a crude extract and a whole cell PDC preparation in a rapidly stirred aqueous/octanol-benzaldehyde emulsion system (see Section 2.7.2.2 for Materials and Methods). The product and by-products values in this Chapter are reported as average concentrations in the total reaction volume (overall concentrations) unless stated otherwise. Yields and molar balances are based on combined amounts in organic and aqueous phases.

The use of whole cell PDC (1 U/mL) resulted in higher PAC and acetoin concentrations at 24 h compared to the more purified preparations. No by-product benzylalcohol, a by-product of the commercial fermentative process using whole cell S. cerevisiae PDC, was detected. Overall, the molar balances for both substrates closed within approx. 10% for all PDC preparations except for 20% unaccounted pyruvate with whole cell PDC. It should be remembered that these values for unaccounted substrates were carried out in small scale experiments (1.8 mL) and more accurate values were determined later in larger scale experiments.

It is evident from this data that the use of whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde biotransformation system was more advantageous than for other PDC preparations and by comparison to its use in the aqueous/benzaldehyde emulsion system. With an initial PDC activity of 1 U/mL, a 3-fold higher specific PAC production (36 mg/U, Table 5-1) was obtained in comparison to that in the aqueous/benzaldehyde emulsion at a higher initial PDC activity of 3 U/mL (12 mg/U, Table 4-2) at 21oC. Additionally when PAC production in the rapidly stirred aqueous/octanol-benzaldehyde system was extended from 24 h to 48 h, no PAC loss

88 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

was observed with the whole cell PDC (see Appendix C-(1)) while in the aqueous/benzaldehyde emulsion system the PAC concentration declined in the later stages of the experiment (Figure 4-5b).

Table 5-1. PAC and by-products formation in a rapidly stirred aqueous/octanol- benzaldehyde system at 21oC with partially purified, crude extract and whole cell PDC

(initial activity 1 U/mL, 2.5 M MOPS, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol, 24 h and a total reaction volume of 1.8 mL).

Partially Crude Whole PDC preparations purified extract cell PAC (mM) 92 141 241 Acetaldehyde (mM) 4 7 5 Acetoin (mM) 3 9 15 Benzylalcohol (mM) 0 0 0 Specific production (mg/U) a 14 21 36 Specific productivity (g/L/h) 0.6 0.9 1.5 Benzaldehyde unaccounted (%, molar basis) b 5 10 7 Pyruvate unaccounted (%, molar basis) b 11 11 20 a based on initial units of PDC carboligase activity b based on combined amounts in organic and aqueous phases

5.2.1.2 Effect of initial enzyme concentration on PAC production

Whether or not the improved PAC production with whole cell PDC was dependent on the catalyst concentration in the rapidly stirred aqueous/octanol- benzaldehyde biotransformation system with a phase volume ratio of 1:1 was investigated and compared with results for a partially purified PDC preparation (Figure 5-2) (see Section 2.7.2.4 for Materials and Methods). For initial activities of less than 0.6 U/mL, the reactions finished after approx. 10 and 20 h for partially purified and whole cell PDC, respectively. Reactions with initial activities above 1 U/mL were completed after 15 h for both PDC preparations.

89 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

a 10

8

6

4 Initial(g/L) PAC 2

0 01234

b 80

60

40

Total PAC (g/L) PAC Total 20

0 01234

c

100

75

50 g/L/day

25

0 01234

d 40

30

20 mg PAC/U 10

0 01234 Initial carboligase activity (U/mL)

Figure 5-2. Effect of enzyme concentration on (a) initial and (b) final PAC concentrations, (c) PAC productivity and (d) specific PAC production by partially purified (◊) and by whole cells PDC (♦) in a rapidly stirred aqueous/octanol-benzaldehyde

90 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

system at 21°C (2.5 M MOPS/KOH, initial pH 6.5, 1.45 M pyruvate, 0.5 mM MgSO4, 0.5 mM TPP, 1.6 M benzaldehyde in octanol phase and a total reaction volume of 15 mL)

The initial 1 h PAC concentrations were similar for both PDC preparations and were approx. proportional to the initial enzyme concentration up to 2.5 U/mL (Figure 5-2a). Figure 5-2b shows that at lower initial activities (0.3 – 2.5 U/mL) the final PAC values were higher for whole cell PDC in comparison to those for partially purified PDC. Similar final PAC concentrations however were achieved for both preparations at initial activities above 2.5 U/mL.

Increased enzyme concentrations resulted in improved PDC stability (see Chapter 6), however the accumulation of PAC and by-products was also likely to contribute for PDC deactivation (Sandford et al., 2005). Unfortunately, the residual activity for whole cell PDC during biotransformation could not be accurately determined due to difficulties in complete PDC recovery from the associated cell debris in the aqueous phase.

With PDC levels of 3.5-4.3 U/mL total reaction volume, biotransformations with whole cell and partially purified PDC both achieved total PAC concentrations of approx. 70 g/L (Figure 5-2b) at a rate of 112 g /L/day (Figure 5-2c). For high volumetric productivities in a PAC production process, Figure 5-2c indicates that whole cell PDC would be advantageous up to an initial PDC activity of 2.5 U/mL. Analysis at 2.5 U/mL established that a 99.1% enantiomeric excess for R-PAC could be achieved for whole cell PDC and 99.4% for partially purified PDC. Thus, the use of a whole cell preparation in a rapidly stirred aqueous/octanol-benzaldehyde did not decrease the enantio-selectivity of PDC. Figure 5-2d shows that with an initial activity of 1.1 U/mL, whole cell PDC achieved a maximum specific PAC production of 42 mg/U in comparison to 13 mg/U for partially purified PDC. No benzylalcohol was detected for either PDC preparation.

91 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

5.2.2 Further evaluation of whole cell biotransformations

5.2.2.1 Effect of cell concentration

Enhancement of yeast PDC production has been extensively studied (Chen et al., 2005; Pronk et al., 1996; Sims and Barnett, 1991; Smits et al., 2000; Toh and Doelle, 1997; van Dijken et al., 1993; van Hoek et al., 1998; van Hoek et al., 2000; Weusthuis et al., 1994; Zeeman et al., 2000) as this plays an important role in improving PAC production. Different maximum values of specific PDC production (U/g DCM) have been obtained from these studies. It is therefore of interest to study whether different concentrations of cell (biomass) with the same PDC activity (U/mL) could affect the overall biotransformation efficiency.

The concentrations of C. utilis cells harvested from five fermentations with different conditions and specific PDC production values (see Section 2.7.2.3 for Materials and Methods) were adjusted to the same initial activity of 1 U/mL. The effect of these different cell concentrations is given in Figure 5-3 where profiles of PAC production are compared. The results show different profiles with increasing rates of PAC production at lower cell concentrations.

The initial 1 h PAC production rate was twice as fast (7 g/L/h) for the biotransformation with the lowest cell concentration (2 g DCM/L) and resulted in 1.3 to 2-fold higher final PAC concentration (56 g/L) in comparison to others with higher cell concentrations. The final values for by-products acetaldehyde and acetoin followed the same trend as the PAC concentrations (see Appendix C-(2)).

The results indicate that lowering the cell concentrations while maintaining the same volumetric PDC activity (U/mL) improved overall PAC production and indicate that cell concentrations as low as possible should be used to achieve the requisite volumetric PDC activity. The implication of these results is that cells with the highest possible specific PDC activity (defined as U/ g DCM) should be used; a characteristic recently maximized by Chen et al. (2005) in pH shift fermentation experiments.

92 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

60

45

30 PAC (g/L) 15

0 0 5 10 15 20 25 30 Time (h) 2 g/L 3 g/L 5 g/L 7 g/L 9 g/L

Figure 5-3. Effect of cell concentrations (g DCM/L) on PAC production in a rapidly stirred aqueous/octanol-benzaldehyde system (initial activity 1 U/mL, 2.5 M MOPS, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol, 21oC and a total reaction volume of 15 mL). The mean values were obtained from triplicate analyses from two experiments. The error bars show lowest and highest values with closely similar values being determined for PAC concentrations.

5.2.2.2 Effect of pH

The effect of pH on PAC production and PDC stability was investigated also for whole cell C. utilis PDC in a rapidly stirred aqueous/octanol-benzaldehyde system at 21oC (see Section 2.7.2.5 for Materials and Methods). A buffer cocktail (50 mM citric acid (pKa 3.15), 50 mM MES (pKa 6.15), 50 mM HEPES (pKa 7.55)) was used to cover the range of tested pH (5.5 – 7.5). The buffer stock was prepared at a 1.25-times concentration (62.5 mM) and pH was adjusted with 5 M KOH additions prior to biotransformation. The concentrations of benzaldehyde partitioned into the aqueous phase were approx. 30-40 mM under these conditions. Figure 5-4 shows that the PAC values at 24 h were highest at a pH of 6.5.

93 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

A subsequent investigation on PDC stability was carried out initially in the same aqueous/octanol-benzaldehyde emulsion condition without pyruvate. In this condition, any differences in PDC stability could not be identified due to short half-lives (≤ 3 h) at all tested pH (data not shown), presumably due to high (1.5 M) benzaldehyde and low (50 mM) buffer concentration. The experimental conditions were therefore changed to allow better differentiation. The octanol phase containing 1.5 M benzaldehyde was removed and instead, 50 mM benzaldehyde and 10 mM octanol were added into the aqueous phase. At this condition, Figure 5-4 shows that whole cell PDC was most stable at pH 6 and the half-life rapidly decreased at higher pH.

12.0 60

9.6 48

7.2 36

4.8 24 PAC (g/L)

2.4 12 PDC half-life (h)

0.0 0 55.566.577.58 pH

Figure 5-4. Effect of pH on PAC production at 1h (◊) and 24h (♦) (1 U/mL, 50 mM buffer cocktail, 0.5 mM MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol, 21oC and a total reaction volume of 100 mL) and on PDC half-life (∆) (1 U/mL, 50 mM buffer cocktail, 50 mM benzaldehyde, 10 mM octanol, 0.5 mM MgSO4, 0.5 mM TPP, 21oC and a total reaction volume of 10 mL).

94 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

5.2.2.3 Effect of MOPS concentration

After an optimal pH of 6.5 was determined for further studies, the next step in process simplification involved the selection of an optimal buffer condition. Until now, MOPS (2.5 M) had been the buffer of choice since it acts both as a buffer and a PDC stabilizing agent. Similar PAC kinetic profiles have been achieved through a combination of low MOPS concentration, pH control and addition of 2 - 2.5 M concentrations of a lower cost stabilizing agents such as glycerol or dipropylene glycol (Leksawasdi et al., 2005; Rosche et al., 2002a; 2005a). However, these studies were performed with partially purified PDC and it would be of great value if whole cell PDC did not require the addition of stabilizing agents.

The effect of MOPS concentration on PAC production by whole cell PDC (see Section 2.7.2.6 for Materials and Methods) is shown in Figure 5-5. The results show a clear trend of higher final PAC with increasing MOPS concentration. The highest PAC concentration (38 g/L) was produced from biotransformation with 2.5 M MOPS. Lowering the MOPS concentration to 1.5 M only reduced the PAC production to 36 g/L while decreasing it to 0.1 M resulted in 12 g PAC/L presumably due to a more rapid pH increase at lower buffer concentration. Lower benzaldehyde partitioning into the aqueous phase has been reported also with lower MOPS concentration (Leksawasdi, 2004). This study also found that the initial 1 h PAC production rates were similar (5% difference) for 1 to 2.5 M MOPS as presumably there was little change in pH or PDC activities in this initial period.

95 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

50

40

30

20 PAC (g/L)

10

0 0 0.5 1 1.5 2 2.5

MOPS (M)

Figure 5-5. Effect of MOPS concentration on PAC production by whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde system (1 U/mL, initial pH was set at 6.5 by addition of 1 M or 5 M KOH, 0.5 mM MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol, 21oC, 20 h). The mean values were obtained from triplicate analyses from two experiments. The error bars show lowest and highest values.

96 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

5.2.2.4 Biotransformation time profiles

To further understand the characteristics of whole cell PDC for PAC production, a detailed biotransformation profile is given in Figure 5-6 with an initial overall enzyme concentration of 2.5 U/mL (5 U/mL aqueous phase). This initial PDC concentration was selected based on the results shown in Figure 5-2. The profiles of the organic and aqueous phase concentrations are shown separately. The initial substrate concentrations were 1.6 M for benzaldehyde in the octanol phase and 1.1 M for pyruvate in the aqueous phase. The profiles show that pyruvate was entirely consumed after 28 h although 600 mM benzaldehyde remained. PAC and by-product acetaldehyde were predominantly partitioned into the octanol phase while by-product acetoin remained mostly in the aqueous phase. The highest benzaldehyde concentration in the aqueous phase was 57 mM. This declined to 41 mM towards the end of biotransformation. Only a small increase in total PAC concentration was observed between 15 – 28 h.

Under these conditions, 725 mM PAC was determined in the organic phase at 28 h with an additional 103 mM PAC in the aqueous phase. The concentrations of acetoin were 24 and 42 mM in the organic and aqueous phases respectively while 18 and 7 mM acetaldehyde was determined in the organic and aqueous phases respectively. The concentration of benzoic acid, present originally as an impurity of the added benzaldehyde, remained constant during the biotransformation and no benzylalcohol was detected.

In a further whole cell biotransformation with pH control, the initial total reaction volume was increased to 100 mL in an open bioreactor and pH was maintained at 6.5 with the addition of 5 M acetic acid (see Section 2.7.2.7 for Materials and Methods). A total of 8.8 mL acid was added over 29 h and the dilution factor of each time point was determined to calculate the biotransformation values (see Appendix D). The biotransformation profile is given in Figure 5-7. The profiles were very similar to those shown in Figure 5-6. Control of the pH value at 6.5 possibly resulted in a more sustained PDC activity although this was not reflected in any significant differences in PAC production (see Table 5-2).

97 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

300 1600 Octanol phase 240 1200 180 800 120 (mM)

400 60 Acetaldehyde and acetoin

PAC and benzaldehyde (mM) 0 0 0 5 10 15 20 25 30

300 1600 Aqueous phase 240 1200 180

800 120 Pyruvate (mM) 400 60 (mM) All others

0 0 0 5 10 15 20 25 30 Time (h)

pyruvate PAC benzaldehyde acetaldehyde acetoin

Figure 5-6. Time profiles of substrates, product and by-products for whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde emulsion biotransformation at 21 ºC

(initial activity 2.5 U/mL, 2.5 M MOPS/KOH initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.6 M benzaldehyde in octanol and a total reaction volume of 15 mL). The mean values were obtained from triplicate analyses and the error bars show lowest and highest values.

98 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

300 1600 Octanol phase 240 1200 180

800 (mM) 120 All others (mM) 400 60 Benzaldehyde, PAC

0 0 0 5 10 15 20 25 30 Time (h)

300 1600 Aqueous phase 240 1200 180

800 120 Pyruvate (mM) 400 60 All others (mM)

0 0 0 5 10 15 20 25 30 Time (h) pyruvate PAC benzaldehyde acetaldehyde acetoin

Figure 5-7. Time profiles of substrates, product and by-products by whole cell PDC in a pH-controlled rapidly stirred aqueous/octanol-benzaldehyde emulsion biotransformation at 21 ºC (initial activity 2.5 U/mL, 2.5 M MOPS/KOH pH 6.5, 0.5 mM

MgSO4, 0.5 mM TPP, 1.4 M pyruvate, 1.7 M benzaldehyde in octanol and a total reaction volume of 100 mL). The mean values were obtained from triplicate analyses and the error bars show lowest and highest values.

99 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

Detailed biotransformation analyses for both biotransformations are given in Table 5-2 at 15 h reaction time point. The similar PAC concentrations showed that scalability was not a problem as long as adequate mixing was provided. A very low concentration of by-product acetaldehyde was measured at 15 h for pH-controlled biotransformation. Increased evaporative losses of the substrate benzaldehyde and by- product acetaldehyde were anticipated in the larger scale bioreactor (open system) during biotransformation at 21oC and this is reflected in the slightly higher values for the unaccounted substrate pyruvate.

Table 5-2. Comparison of PAC production with whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde system at 21oC with and without pH control

2.5 M MOPS 2.5 M MOPS with pH control (see Figure 5-6) (see Figure 5-7) Total reaction volume (mL) 15 100 Reaction time (h) 15 15 PAC in total reaction volume (g/L) 60 60 PAC in octanol phase (g/L) 104 105 PAC in aqueous phase (g/L) 16 14 Acetoin (g/L) 2.3 1.9 Acetaldehyde (g/L) 0.8 0.04 Volumetric productivity (g/L/day) 96 95 Specific production (mg/U) a 24 26 Molar yield (%) per pyruvate b 79 78 Molar yield (%) per benzaldehyde b 91 93 Pyruvate unaccounted (%, molar basis) c 6 10 Benzaldehyde unaccounted (%, molar basis) c 4 4 a based on initial PDC carboligase activity b molar yield (mole/mole) based on consumed substrates c based on combined amounts in organic and aqueous phases

100 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

5.3 Discussion and conclusions

Two-phase liquid-liquid systems have been applied for biotransformation processes to improve productivities, yields and product recovery. Two strategies have evolved: one is to use living cells and the other is to use isolated enzymes, since organic solvents often affect the viability activity of cells (Schmid et al., 2001). The present study provides a comparison for catalyst preparations (including resting whole cells) with characteristics in-between living cells and enzymes.

Improved biotransformation characteristics were achieved with whole cell PDC as shown with higher specific PAC production (mg PAC/U) and productivity (g/L/h) compared to the values for partially purified and crude extract preparations under similar initial conditions (Table 5-1). These advantages for whole cell PDC were even more evident at lower PDC activities (Figure 5-2), presumably due to greater stability of whole cell PDC (Satianegara et al., 2006a). As such, it was likely that PAC formation slowed and ceased earlier for partially purified PDC during biotransformation. The advantages were less evident at higher initial enzyme activities (3.5 – 4.3 U/mL) as shown with similar final PAC concentrations for both PDC preparations. Presumably at the higher enzyme concentrations, the PDC activity was not limiting PAC formation for either preparation. Whole cell PDC also resulted in a high product enantiomeric excess with a similar value as for partially purified PDC.

Having established the potential of the whole cell PDC as low cost enzyme preparation, further evaluations showed its advantages at a minimal cell concentration with 1.5 – 2.5 M MOPS and pH of 6.5. The initial production rate and overall productivity was enhanced with a low cell concentration having high specific PDC activity (U/g DCM) (Figure 5-3). As enzyme PDC appears to be leaked into the biotransformation medium (Chapter 4), the enhanced initial production rate and overall productivity was likely to be a result from reduced resistance of substrates and products transfer between the two bulk liquid phases across the cell biomass. Studies by Cull et al. (2000) and Lennoury et al. (2005) had reported concerns and a relationship between reduced initial production rate and the mass transfer barrier across the bulk liquid

101 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

phases in the presence of cell biocatalyst. Thus, maximization of specific PDC activity inside cells (U/g DCM) is very important. A recent study to answer this need was reported by Chen et al. (2005) with a 6-fold improvement in specific PDC activity by shifting down the pH from 6 to 3 during yeast C. utilis fermentation.

The optimal pH of 6.5 for PAC production determined in the present investigation (Figure 5-4) has been reported also in other studies with partially purified PDC from C. utilis (Shin and Rogers, 1996a) and of R. javanicus (Rosche et al., 2002a) under different sets of conditions in the aqueous/benzaldehyde emulsion system at 4oC.

The stability of whole cell PDC appeared to be influenced by buffer species and/or buffer concentration. Rosche et al. (2002a) reported that increasing MOPS concentration resulted in longer half-life for partially purified PDC. This was observed for whole cell PDC with higher PAC produced at higher MOPS concentrations (Figure 5-5). Low PAC concentrations were obtained at lower MOPS concentration presumably due to the effect of the increasing pH. Similar PAC values were reported with 50 mM MOPS and 50 mM cocktail buffer (Figure 5-4).

One of the reported advantage of using whole cell enzyme is the potential supply of cofactor(s) (Ishige et al., 2005), which could further reduced the associated cost of biotransformation. However, it was confirmed in the current study that cofactor TPP addition is an essential requirement for high PAC production with whole cell C. utilis PDC (Appendix C-(3)). Long & Ward (1989) reported that cell permeabilization by benzaldehyde resulted in a release of cell magnesium and TPP with S. cerevisiae.

PAC and by-product concentrations after a 15 h reaction period with whole cell PDC at an initial overall activity of 2.5 U/mL (Table 5-2) were consistent with the preliminary results reported by Rosche et al. (2005b) where PAC levels of approximately 686 mM in the organic phase and 85 mM in the aqueous phase (58 g/L total reaction volume) were obtained with whole cell PDC at initial overall activities of 2.5 U/mL (C. utilis). Furthermore, the present study showed that the PAC production characteristics remained unchanged during scale up (1.8 mL to 100 mL). The increased amount of unaccounted pyruvate (up to 10%) possibly resulted from greater evaporative

102 Chapter 5. Aqueous/octanol-benzaldehyde emulsion system

loss of by-product acetaldehyde in the larger scale system. While the feasibility of the system at industry scale was not a subject in this study, future investigation on this matter should consider factors such as adequate mixing, shear rates, pH-control, compatibility of the vessel and seal materials with substrate benzaldehyde or even vessel configuration.

103

Chapter 6

6 ROLE OF CELL COMPONENTS FOR ENHANCED PAC PRODUCTION Role of cell components for enhanced PAC production

6.1 Introduction

The use of whole cell PDC instead of partially purified PDC for enzymatic PAC synthesis can reduce the cost of catalyst preparation (Rosche et al., 2005b; Satianegara et al., 2006a). Whole cell PDC exhibits higher stability towards benzaldehyde and temperature when compared with the more purified PDC preparation in an aqueous/benzaldehyde emulsion system (Satianegara et al., 2006a). This catalyst with characteristics between those of isolated enzymes and living cells used in the commercial process also maintains the advantages of the enzymatic system, viz. no formation of by-product benzylalcohol and achievement of high PAC concentrations (Rosche et al., 2005b; Satianegara et al., 2006a). It is however unclear which factors contribute to these improved characteristics of whole cell PDC.

When C. utilis PDC was added in the form of viable resting whole cells (whole cell PDC) in a rapidly stirred aqueous/octanol-benzaldehyde system, cell viability and metabolic activity were lost immediately (Rosche et al., 2005b). It has been shown also in the present study that the C. utilis cells were at least partially permeabilized in the aqueous/benzaldehyde emulsion (Chapter 4) with PDC being released into the reaction medium. In S. cerevisiae, enzyme PDC is known to be located within its cytoplasm (van Hoek et al., 1998). Other studies have shown that cellular components (fatty acids and proteins) were released from S. cerevisiae cells upon exposure to various organic

104 Chapter 6. Role of cell components

solvents in a two-phase system with a 10% v/v aqueous phase (Nikolova and Ward, 1992a, b). In particular, these authors showed that cells exposed to the more hydrophilic solvents, for which the log10 values of the partitioning coefficient were lower or equal than 2 (log P ≤ 2), appeared to be punctured. For these conditions, a spectrum of phospholipids was detected in the reaction medium. Relevant to the present study, the log P values for benzaldehyde, PAC, acetaldehyde and acetoin were 1.4, 0.6, 0.7 and - 0.5 respectively (Sandford et al., 2005) indicating their potential to cause cell damage. (Note: the partitioning coefficient P is defined as a ratio of the equilibrium concentration of a component in the aqueous phase to its equilibrium concentration in solvent octanol).

In this Chapter, some factors leading to enhanced PAC production with whole cell C. utilis PDC were investigated. Any protective effect of the cell envelope was distinguished from stabilizing effects of cell components. Various additives similar to cell components were added to partially purified PDC in order to evaluate which component(s) were likely to enhance PAC production and PDC stability when using the whole cell preparation.

105 Chapter 6. Role of cell components

6.2 Results

6.2.1 Effect of cell debris on PDC stability

The higher specific PAC production achieved with whole cell PDC compared to that for partially purified PDC in the rapidly stirred aqueous/octanol-benzaldehyde system (Figure 5-2) might be related to higher enzyme stability in the whole cell preparation. The stabilities of various PDC preparations were therefore determined and compared (see Section 2.8.3 for Materials and Methods). PDC in cells broken by cell attrition (unclarified crude extract) and PDC in clarified crude extract served as intermediate preparations between whole cells and partially purified PDC.

Figure 6-1 shows the time profiles of PDC deactivation for the different PDC preparations on exposure to 1.5 M benzaldehyde (octanol phase) without pyruvate addition. Higher stability for whole cell PDC was observed in comparison to the more purified preparations.

Table 6-1 shows the estimated half-lives of different PDC preparations and for two volumetric activities viz. 1 U/mL and 4 U/mL. PDC was more stable at increased enzyme concentration (4 U/mL). In the presence of broken cell debris, PDC as an unclarified crude extract had a 4-fold greater half-life in comparison to the clarified crude extract PDC preparation. The enzyme half-lives for the unclarified crude extract and whole cell PDC preparation were of similar magnitude at 41 and 50 hours respectively. In an additional experiment with similar conditions (this study), substantial whole cell permeabilization was apparent as residual PDC activities were measured in the aqueous phase (41%) and in the resuspended biomass pellet (45%) from the initially added 2.2 U PDC/mL aqueous phase after 3 h. Therefore it is likely that components in the cell debris rather than the maintenance of an intact cell envelope in whole cell PDC were at least partly responsible for the increased PDC stability.

106 Chapter 6. Role of cell components

100% Whole cell Crude extract (unclarified) Crude extract (clarified) Part. purified 50% Residual PDC activity (%)

0% 01234 Time (day)

Figure 6-1. Profiles of residual activities of partially purified, crude extract (clarified and unclarified) and whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde system without pyruvate (initial activity 4 U/mL, 2.5 M MOPS, pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.5 M benzaldehyde in octanol, 21oC). The mean values were determined from duplicate analyses from two vials. The error bars show lowest and highest values.

Table 6-1. Estimated half-lives of different PDC preparations in a rapidly stirred aqueous/octanol-benzaldehyde system without pyruvate at 21oC. See data (4 U/mL) in Figure 6-1.

Half-lives of PDC (h) Partially Crude Unclarified crude Whole cell purified extract extract 1.0 U/mL 1 n.d. n.d. 8 4.0 U/mL 5 9 41 50 n.d. not determined

107 Chapter 6. Role of cell components

6.2.2 Effect of different additives on PAC production

To identify possible cell components responsible for the increased stability of the unclarified crude extract PDC when compared to clarified crude extract PDC, several additives were selected based on their similarity with major cell components usually removed through centrifugation in crude extract PDC preparations.

The effects of addition of these components to partially purified PDC (1.5 U/mL) were investigated (see Section 2.7.2.8 for Materials and Methods) and PAC concentrations after 24 h biotransformations are shown in Figure 6-2. The PAC values in this Chapter are reported as average concentrations in the total reaction volume (overall concentrations) unless stated otherwise. As evident from the data, addition of 5 mg/mL of chitin, albumin, asolectin and lecithin all resulted in similar or lower PAC concentrations. Lecithin and asolectin are compounds selected primarily due to their contents of phosphatidylcholine. Recovered cell debris and cell lipids from C. utilis, as well as a commercial cell wall fraction from S. cerevisiae, resulted in 1.5 to 2-fold higher PAC production when compared to the control. Of the three cell membrane components, the addition of the fatty acyl chain palmitic acid (5 mM) did not result in any improvement. However addition of either a sterol (ergosterol, 5 mM) or a phospholipid (phosphatidylcholine, 2 mg/mL) resulted in marked increases (2.4 and 2.8- fold respectively) in PAC concentrations. A similar improvement in PAC concentration was observed with addition of the 50 mM anionic surfactant sodium bis(2-ethyl-1- hexyl)sulfosuccinate (AOT).

When 5 mM ergosterol or 50 mM AOT was added to the crude extract PDC preparation, a significant improvement (2.5-fold) in PAC production was observed also (Figure 6-3). However for whole cell PDC, no significant improvement in PAC production was evident following addition of 5 mM ergosterol or 50 mM AOT to a rapidly stirred aqueous/octanol-benzaldehyde biotransformation. The results suggest that the presence of cell components in whole cell PDC may be sufficient for maintaining PDC activity without the need for further additives.

108 Chapter 6. Role of cell components

Chitin (5 mg/mL) Bovine serum albumin (5 mg/mL) Asolectin (5 mg/mL) Lecithin from soy bean (5 mg/mL) Palmitic acid (5 mM) Control (no additive) Cell lipid from C. C. utilis utilis Cell debris from C. C. utilis utilis (5 mg/mL) Cell wall from S. S. cerevisiae cerevisiae (5 mg/mL) Ergosterol (5 mM) AOT (50 mM) Phosphatidylcholine (2 mg/mL)

0 120 240 360 Total PAC (mM)

Figure 6-2. Effect of additives on PAC formation in a rapidly stirred aqueous/octanol- benzaldehyde system at 21oC with partially purified PDC (1.5 U/mL, 2.5 M MOPS, initial

pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.43 M pyruvate, 1.5 M benzaldehyde in octanol, 24 h and a total reaction volume of 1.8 mL). The mean values were determined from duplicate analyses. The error bars show lowest and highest values.

Crude extract PDC (control)

Crude extract PDC + 5 mM ergosterol

Crude extract PDC + 50 mM AOT

Whole cell PDC (control)

Whole cell PDC + 50 mM AOT

Whole cell PDC + 5 mM ergosterol

0 150 300 450 Total PAC (mM)

Figure 6-3. Effect of the addition of 5 mM ergosterol and 50 mM AOT on PAC formation in a rapidly stirred aqueous/octanol-benzaldehyde system at 21oC with a crude extract and a whole cell PDC preparation (1.5 U/mL, 2.5 M MOPS/KOH, initial pH 6.5,

0.5 mM MgSO4, 0.5 mM TPP, 1.43 M pyruvate, 1.5 M benzaldehyde in octanol, 24 h and a total reaction volume of 1.8 mL). The mean values were determined from duplicate analyses. The error bars show lowest and highest values.

109 Chapter 6. Role of cell components

6.2.3 Effect of specific additives on PDC half-lives

The effects of additives which enhanced PAC production on stability of partially purified PDC (1 U/mL, 1.5 M benzaldehyde in octanol, no pyruvate) are compared in Figure 6-4 with the results for whole cell PDC (see Section 2.8.3 for general methods). No improvements in half-lives were evident with values of approx. 1 h obtained in the presence of phosphatidylcholine, ergosterol and AOT similar to that for the partially purified PDC. However, despite its short half-life, the residual activity for partially purified PDC with addition of AOT was maintained relatively constant (46%) after the second hour of exposure with a slow decline to 35% residual activity at 24 h. This residual activity was similar or even slightly higher compared to that of whole cell PDC at 24 h. Partially purified PDC control (no AOT addition) lost its activity faster than partially purified enzyme with AOT addition.

100% Whole cells Part. purified Part. purified + AOT Part. purified + Phosphatidycholine Part. purified + Ergosterol 50% Residual activity (%)

0% 0 5 10 15 20 25

Time (h)

Figure 6-4. Profiles of residual activities of partially purified PDC in the absence and presence of 2 mg phosphatidylcholine/mL, 5 mM ergosterol, 50 mM AOT in comparison to whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde system without o pyruvate at 21 C (initial activity 1 U/mL, 2.5 M MOPS/KOH, pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.5 M benzaldehyde in octanol). The mean values were determined from triplicate analyses. The error bars show lowest and highest values.

110 Chapter 6. Role of cell components

6.2.4 Biotransformation in the presence of AOT

6.2.4.1 Effect of AOT concentration

The effect of increased concentrations of AOT addition (0, 7, 10, 20, 50, 100 mM) on PAC production (24 h) with the partially purified PDC is shown in Figure 6-5. The results show a clear trend of increasing PAC with increasing AOT concentration even though only small differences were observed between 20-100 mM of AOT. The maximum PAC concentration (approx. 300 mM) was produced with the addition of 50 mM AOT. The PAC produced with the addition of 100 mM AOT was slightly less compared to that with 50 mM AOT, possibly due to experimental variability.

400

300

200 [PAC] (mM) 100

0 0255075100 [AOT] (mM)

Figure 6-5. Effect of AOT concentration on PAC production by partially purified PDC in a rapidly stirred aqueous/octanol-benzaldehyde system at 21ºC (initial activity 1.0

U/mL, 2.5 M MOPS/KOH, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.33 M pyruvate, 1.5 M benzaldehyde in octanol, 24 h and a total reaction volume of 1.8 mL). The mean values were determined from duplicate analyses. The error bars show lowest and highest values.

111 Chapter 6. Role of cell components

6.2.4.2 Kinetics of PAC production

Since AOT stabilized partially purified PDC, the effect of AOT on PAC production kinetics was further investigated. Figure 6-6 and Figure 6-7 show the kinetics of PAC formation for partially purified PDC at an initial activity of 1 U/mL in the absence and presence of 50 mM AOT respectively. Phase separation in samples was easily achieved by centrifugation. Similar benzaldehyde concentrations in the aqueous phase (52-63 mM) were measured for all reactions. The volumetric rates of PAC production over the initial 4 h period were slower for partially purified PDC with AOT addition (12 g/L/h) compared to the rate without AOT addition (19 g/L/h). In the absence of AOT, total PAC concentrations were constant at 95 mM after 10 h. In the presence of AOT, PAC was formed in near linear fashion for at least 20 h and reached 210 mM at this time. This prolonged PAC production was accompanied by an enhanced residual PDC activity for the partially purified PDC with AOT. This was supported by the results of an earlier stability study suggesting that AOT may be a useful additive for longer term biotransformations.

In Table 6-2 the biotransformation kinetics are summarized and compared to whole cell data in the presence and absence of AOT for similar conditions (see Appendix C-(4)). With AOT addition, the improvement was very significant for partially purified PDC both for PAC production and PDC stability. No improvement was evident for whole cell PDC. The residual activity for whole cell PDC during biotransformation could not be accurately determined due to difficulties in complete PDC recovery from the associated cell debris in the aqueous phase. However in both cases there was an unfavourable increase of inactivating by-product acetoin following AOT addition.

112 Chapter 6. Role of cell components

1500 Octanol Phase 100 1200 80 900 60

600 40 All others (mM) 300 20

Benzaldehyde, PAC (mM) 0 0 0 5 10 15 20 25 Time (h)

1500 Aqueous Phase 100

1125 80

60 750 40 Pyruvate (mM) 375 All others (mM) 20

0 0 0 5 10 15 20 25 Time (h)

pyruvate PAC benzaldehyde acetaldehyde Acetoin PDC

Figure 6-6. Time profiles of substrates, product and by-products for partially purified PDC in an aqueous/octanol-benzaldehyde system at 21ºC in the absence of AOT (initial

conditions: 1 U/mL, 1.3 M pyruvate, 2.5 M MOPS initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.5 M benzaldehyde in octanol phase and a total reaction volume of 10 mL). The mean values were determined from triplicate analyses of two vials. The error bars show lowest and highest values.

113 Chapter 6. Role of cell components

1500 Octanol Phase 100 1200 80 900 60

600 40 All others (mM) 300 20 Benzaldehyde, PAC (mM) 0 0 0 5 10 15 20 25 Time (h)

1500 Aqueous Phase 100 1200 80 900 60

600 40 Pyruvate (mM) All others (mM) 300 20

0 0 0 5 10 15 20 25 Time (h)

pyruvate PAC benzaldehyde acetaldehyde acetoin PDC

Figure 6-7. Time profiles of substrates, product and by-products for partially purified PDC in an aqueous/octanol-benzaldehyde system at 21ºC with 50 mM AOT addition

(initial conditions: 1 U/mL, 1.3 M pyruvate, 2.5 M MOPS initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.5 M benzaldehyde in octanol phase and a total reaction volume of 10 mL). The mean values were determined from triplicate analyses of two vials. The error bars show lowest and highest values.

114 Chapter 6. Role of cell components

Table 6-2. Summary of PAC production by partially purified and whole cell PDC in the absence and presence of 50 mM AOT in a rapidly stirred aqueous/octanol- benzaldehyde system (24 h) at 21oC with an initial PDC activity 1 U/mL (see Figure 6-6 and Figure 6-7)

Partially purified Whole cell -AOT +AOT -AOT +AOT Figure 6-6 and 6-7 Appendix C-(4) Initial production rate (g/L/h) a 19 12 29 28 PAC (g/L) b 15 36 42 41 Acetoin (g/L) b 0.3 1.5 1.4 1.6 Acetaldehyde (g/L) b 0.2 0.3 0.3 0.3 Pyruvate unaccounted (%) 3 1 12 11 Benzaldehyde unaccounted (%) 1 1 3 3 Molar yield PAC on pyruvate (%) c 87 85 77 77 Molar yield PAC on benzaldehyde (%) c 98 97 91 93 PDC half-lives (h) 1.5 19 ------a based on PAC production in 4 hours b based on the average concentration in the total reaction volume c molar yield (mole/mole) based on consumed substrates

6.3 Discussion and conclusion

Whole cells of C. utilis have been shown to be efficient catalysts for PAC production in an aqueous/octanol-benzaldehyde system (Chapter 5 and a preliminary study by Rosche et al. (2005b)). The present study gives more insights into possible factors behind their improved performance compared to partially purified preparations i.e. higher stability as well as higher productivity at low enzyme concentrations. The present study identifies phospholipids as cell components that enhanced PAC production and shows that the synthetic surfactant, AOT, improved both PAC production and PDC stability when using a partially purified PDC preparation (Satianegara et al., 2006b).

115 Chapter 6. Role of cell components

At lower PDC activities, whole cell PDC resulted in much higher specific productivities (mg PAC/U) than partially purified PDC (Figure 5-2). This was further confirmed in the present study, with a higher stability of whole cell PDC in the presence of 1.5 M benzaldehyde (octanol phase) when compared to that of partially purified PDC (Figure 6-1 and Table 6-1). It has been suggested that the barrier of the cell envelope could give physical protection to enzymes inside cells (Stratford, 1994). Interestingly the results of the present study suggest that protection was not conferred by any cell envelope barrier since PDC in broken cells (unclarified crude extract) was nearly as stable as whole cell PDC. Proteins are also known for their potential to stabilize enzymes (Scopes, 1994). However, the soluble proteins present in the clarified crude extract did not appear to protect PDC from the strong deactivation that was evident. Stability in the present study appeared to be associated with components in the cell debris (which may also include proteins of lower solubility).

Results which showed enhanced PAC production with partially purified PDC in the presence of phosphatidylcholine and ergosterol indicated a possible role of cell membrane components in the enhanced performance of whole cell PDC (Figure 6-2). However the mixture of phospholipids contained in lecithin and asolectin did not increase PAC concentrations although this might have been due to their compositions or the presence of impurities. Phospholipids are very important components of cell membranes in yeast (Lagovic et al., 2005) and various organic solvents have been shown to cause rapid breakdown of this permeability barrier function of cell membranes (de Smet et al., 1978; Ingram and Buttke, 1982; Silver and Wendt, 1967). The enhancement effect of ergosterol shown in this study is interesting as it has been reported to have a structural role, in coexistence with phospholipids, in the maintenance of cellular membranes (Smith et al., 1996).

Of a particular significance in the present investigation was the finding that PDC activity could be enhanced by the addition of the synthetic surfactant AOT. Phosphatidylcholine is an amphoteric surfactant while AOT is classified as an anionic surfactant - surfactants being molecules with a hydrophilic head and a hydrophobic tail. Both AOT and phosphatidylcholine have long chain structures. The chemical compositions for AOT and phosphatidylcholine are given in Figure 6-8.

116 Chapter 6. Role of cell components

CH2CH3 O O O O + N (CH3)3 P H3CH2CH2CH2CCH2O H3C(H2C)14 OOO- O H3CH2CH2CH2CCH2O CHSO3Na (CH ) CHCHCH CHCH(CH ) CH 2 7 2 2 4 3 O CH2CH3

(a) (b)

Figure 6-8. Chemical compositions of (a) phosphatidylcholine and (b) sodium bis(2- ethyl-1-hexyl)sulfosuccinate (AOT or Aerosol-OT).

The protective effect of AOT appeared to be enhanced when both AOT and pyruvate were present in the system (Figure 6-7). Higher PDC stability towards benzaldehyde in the presence of pyruvate has been observed in an earlier study (Rosche et al., 2005a) however the mechanism of this enhancement is not understood. A similar observation regarding the role of surfactants in the presence of a substrate was made by Smolders et al. (1991) in a study of steroid bioconversion by cells of Arthrobacter simplex. In the presence of the surfactant phosphatidylethanolamine, the authors reported 100% residual activity (in the presence of substrate) and 50% residual activity (in the absence of substrate) after similar reaction times. Addition of phosphatidylcholine or AOT has been studied in many other biocatalytic systems, mostly for activity and stability enhancement or for ease in enzyme and product recoveries (Fadnavis et al., 1989; Hirakawa et al., 2003; Larsson et al., 1990; Pinto- Sousa et al., 1996; Robl et al., 2000; Sarcar et al., 1991; Wu et al., 2001).

Addition of AOT improved the final PAC concentrations in the biotransformation with partially purified PDC (Figure 6-5). With AOT, PAC was produced at a lower but steady rate for at least 20 h and with much lower PDC inactivation (Figure 6-7 and Table 6-2). Since the benzaldehyde concentration in the aqueous phase was similar both in the presence and absence of AOT, the lower rate was not due to this substrate concentration. However benzaldehyde transfer into the enzyme microenvironment might have been slowed by the presence of AOT and this could have resulted in the reduced initial reaction rates although increasing the overall enzyme stability. In this latter regard, it has been suggested that a surfactant layer at the

117 Chapter 6. Role of cell components

aqueous/organic interphase can protect enzymes from deactivation at this interphase (Carvalho and Cabral, 2000). Most studies of enzymatic aqueous/organic/surfactant systems have investigated enzymes entrapped in reverse micelles (water in oil microemulsions) with very low water content (Castro and Knubovets, 2003). By comparison the present study employs equal water content (1:1 phase ratio) for the reaction between the low solubility benzaldehyde and high solubility pyruvate. The arrangement of AOT molecules in the present aqueous/octanol-benzaldehyde emulsion system remains undetermined since many factors such as temperature, water content, pH, degree of agitation, the ionic strength of aqueous phase, phase ratio and the formed product/by-products can dynamically contribute in the arrangement and formation of possible micellar structures. Different possible surfactant arrangements in an aqueous/organic system are given in Figure 6-9.

Figure 6-9. Possible microstructures formed in an aqueous/surfactant/organic system: the droplet model as (a) micelle and (b) reverse micelle versus (c) bicontinuous model (adapted from Lindman et al. (1989)) as pipeline structure/sponge-like phase (O = oil/organic solvent phase, W = water/aqueous phase, S = surfactant film).

118

Chapter 7

7 CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE STUDIES Conclusions and recommendations for future studies

Objective: Chemical modifications of the partially purified C. utilis PDC Conclusions: Partially purified PDC was subjected to five different chemical modifications. Various conditions were applied in each modification to screen for modified enzymes with enhanced stability towards inactivating substrate benzaldehyde. However, significant activity losses were observed after each of the modification reactions presumably due to the requirement for non-acidic pH (7.2−9.4) in most reactions, the inactivating effects of the modification reagents (the chemical composition or the excess concentration) or reaction sites targeted may not have been most suitable.

Objective: Enzymatic R-PAC production in an aqueous/benzaldehyde emulsion system with various preparations including C. utilis whole cell PDC Conclusions: Recent progress in enzymatic PAC production has established the need for low cost biocatalyst preparation while maintaining the current high concentrations, productivities and yields. The use of whole cell PDC was therefore evaluated in an aqueous/benzaldehyde emulsion and was compared to the more purified preparations viz. crude extract and partially purified. Whole cell PDC showed higher stability towards benzaldehyde and temperature in comparison to partially purified preparations. Increasing the temperature from 4o to 21oC for PAC production with whole cell PDC resulted in similar final PAC concentrations of 39 and 43 g/L (258 and 289 mM)

119 Chapter 7. Conclusions and recommendations

respectively from initial 300 mM benzaldehyde and 364 mM pyruvate. The overall volumetric productivity was enhanced 2.8-fold, which reflected the 60% shorter reaction time at the higher temperature. Enantiomeric excess values of 98 and 94% for R-PAC were obtained at 4o and 21oC respectively and benzylalcohol (a potential by- product from benzaldehyde) was not formed.

Objective: Use of whole cell PDC in a rapidly stirred aqueous/octanol- benzaldehyde emulsion system Conclusions: The potential of the less expensive whole cell PDC preparation was also shown in an aqueous/octanol-benzaldehyde (two-phase) bioreactor at 21oC. Enhanced PAC production with whole cell PDC was observed when compared to the partially purified PDC especially for a lower range of initial activities (0.3 – 2.5 U/mL). With an initial activity of 1.1 U/mL and at a 1:1 phase volume ratio, whole cell PDC achieved a maximum specific PAC production of 42 mg/U in comparison to 13 mg/U for partially purified PDC. PAC levels of 104 g/L in the organic phase and 16 g/L in the aqueous phase (60 g/L total reaction volume, 15 h) were obtained with whole cell PDC at an initial activity of 2.5 U/mL. The values corresponded to a specific PAC production of 9 g/g DCM which is 1.7-fold higher than those achieved in the aqueous/benzaldehyde emulsion system (whole cell PDC, 3 U/mL, 12 h, 21oC). No benzylalcohol was formed and a 99.1% enantiomeric excess for R-PAC was determined. The initial production rate and overall productivity appeared to be enhanced at low cell concentrations with the highest specific PDC activities (U/g DCM) achieved for these conditions. PAC loss which occurred during prolonged biotransformation with whole cell PDC in the aqueous/benzaldehyde emulsion system was not observed in the two-phase process.

Objective: Role of cell components for enhanced PAC production Conclusions: Studies to understand the factors behind enhanced performance of whole cell PDC in the aqueous/octanol-benzaldehyde system at 21oC established that some components in the cell debris rather than the intact cell envelope were responsible for the increased PDC stability. The enhanced PAC production with partially purified PDC in the presence of phosphatidylcholine and ergosterol indicated a possible role of cell

120 Chapter 7. Conclusions and future recommendations

membrane components. It was apparent that surfactants, both biologically-occurring (e.g. phosphatidylcholine) or synthetically manufactured (e.g. AOT), could enhance PDC stability and PAC production with the partially purified enzyme. Addition of 50 mM AOT to the biotransformation with partially purified PDC enhanced the PDC half- life by 14-fold (19 h) and increased specific PAC production by 2-fold (36 mg/U). When AOT was added to whole cell PDC, no improvement of PAC production was evident.

Overall conclusions From the current studies it can be concluded that whole cell PDC is preferable to cell-free (crude extract or partially purified) preparations for PAC production as the former preparation is a less expensive catalyst, achieved similar PAC levels and showed higher stability, most possibly due to the advantageous presence of its cell membrane components. Furthermore, the use of whole cell PDC maintained low by-product formation (acetaldehyde and acetoin) and a high enantiomeric excess value for R-PAC (~ 99%) as for the cell-free processes. The use of whole cell PDC in the present studies also did not result in any benzylalcohol formation nor did this occur with the crude extract or with partially purified PDC. Benzylalcohol is a major by-product associated with other previous whole cell enzyme-based processes (Mochizuki et al., 1995; Nikolova and Ward, 1991; 1994). The relatively high initial substrate benzaldehyde and the resultant faster PAC formation during the initial biotransformation periods in the present investigation may have inhibited the formation of benzylalcohol. A comparison between the enzymatic whole cell biotransformation and the in vivo analogue to the industrial process, showing improvements in final product concentration, final productivity, specific production and specific productivity, has been detailed in Rosche et al. (2005b). Further improvements for a successful commercial process could be achieved by reducing the volume ratio of octanol or replacing the use of commercial pyruvate with biological pyruvate to lower the operational costs as well as reducing by- product acetoin formation to improve yield on pyruvate and to minimize problem in downstream product recovery.

121 Chapter 7. Conclusions and future recommendations

Recommended future work

Future research following from the present study could include: 1. Improved whole cell PDC preparations • maximization of PDC production (PDC per cell) through gene overexpression with the model yeast S. cerevisiae or by varying culture conditions (e.g. by pH shifts). Higher PDC activities might contribute to reduced enzyme production cost and further enhance the product concentrations, productivities and specific productions of the whole cell biotransformation,

• investigation of the characteristics of whole cell C. utilis PDC harvested at different fermentation time points for possible different isoenzymes showing improved PAC production and/or PDC stability,

• studies on cell storage methods to preserve enzyme conformation and/or increase catalytic activity such as the addition of stabilizing additives. Gradual loss of PDC activity stored as frozen yeast cake has been observed after 18 months storage at -20oC,

• cross-linking, immobilization and/or encapsulation of cells using new developed technologies, such as Cross-Linked Enzyme Aggregates (CLEAs), Eupergit© beads and/or sol-gel technology respectively, to increase PDC stability in the presence of benzaldehyde and/or for the purpose of enzyme recycling while maintaining activity and productivity, Previous immobilization studies used calcium alginate beads with no significant advantages (Shin and Rogers, 1995).

• investigation of PDC from other microorganisms with interesting characteristics, such as those with lower by-product acetoin formation (e.g. yeast C. tropicalis, see Gunawan et al. (2006)) or thermotolerant PDC, while applying the current knowledge for optimal whole cell process.

122 Chapter 7. Conclusions and future recommendations

2. Improved PDC characteristics • PDC engineering to alter (remove or reduce) the catalytic formation of by- product acetoin. Elimination of acetoin would contribute to increased product yield on pyruvate, may reduce enzyme inactivation and improve downstream product recovery,

• PDC engineering to increase the carboligation activity of yeast PDC e.g. through a directed evolution approach following comparison of amino acid sequences between PDC of yeast S. cerevisiae and bacteria Z. mobilis, as the latter has been shown to have a higher carboligation activity,

• PDC engineering to increase the affinity of yeast PDC for acetaldehyde as an alternative substrate to pyruvate. Use of acetaldehyde would contribute to a significantly lower substrate cost and eliminate the problem of pH rise during biotransformation (pyruvate decarboxylation).

3. Process developments • development of a repeated batch or continuous biotransformation process for constant removal of inactivating substances (PAC and by-products) while maintaining enzyme availability to maximize PAC production and volumetric productivity e.g. an immobilized-enzyme membrane reactor with a higher PDC stability,

• further mathematical modelling for enzymatic whole cell biotransformation (see Leksawasdi et al. (2004)) in the rapidly stirred aqueous/octanol- benzaldehyde system to determine the optimal strategies for maximum PAC production in such a repeated batch or continuous process,

123 Chapter 7. Conclusions and future recommendations

• replacing the requirement of high MOPS concentration to maintain PDC stability and pH control in the aqueous/octanol-benzaldehyde system with lower buffer concentration, addition of optimal concentration surfactant (biological or synthetic) and pH control through acid addition while maintaining PDC stability, PAC production and high productivity.

(3) Economic analysis • economic evaluation of enzymatic whole cell biotransformation in the aqueous/octanol-benzaldehyde bioreactor and comparison with the other process options to identify the most cost effective process for further development and scale-up studies.

124 References

References

Abourashed, E. A., El-Alfy, A. T., Khan, I. A. and Walker, L. (2003). Ephedra in perspective - a current review. Phytotherapy Research 17, 703-712.

Andersson, M. M. and Hatti-Kaul, R. (1999). Protein stabilizing effect of polyethyleneimine. Journal of Biotechnology 72, 21-31.

Andersson, M. M., Breccia, J. D. and Hatti-Kaul, R. (2000). Stabilizing effect of chemical additives against oxidation of lactate dehydrogenase. Biotechnology and Applied Biochemistry 32, 145-153.

Anjum, F., Rishi, V. and Ahmad, F. (2000). Compatibility of osmolytes with Gibbs energy of stabilization of proteins. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1476, 75-84.

Anonymous (2006). Ephedra herb.

Arjunan, P., Umland, T., Dyda, F., Swaminathan, S., Furey, W., Sax, M., Farrenkopf, B., Gao, Y., Zhang, D. and Jordan, F. (1996). Crystal structure of the thiamin diphosphate-dependent enzyme pyruvate decarboxylase from the yeast Saccharomyces cerevisiae at 2.3 A resolution. Journal of Molecular Biology 256, 590- 600.

Arnold, F. H. and Georgiou, G. (2003). Directed enzyme evolution: screening and selection methods. Totowa, NJ: Humana Press Inc.

125 References

Auhagen, E. (1931). Co-Carboxylase, ein neues Co-Enzyme der alkoholischen Gärung. Hoppe-Seyler's Zeitschrift fur Physiologische Chemie 204, 149-167.

AusIndustry (2004). Biotech Business Indicators (Report). Department of Industry Tourism and Resources: Canberra, Australia

Baburina, I., Gao, Y., Hu, Z., Jordan, F., Hohmann, S. and Furey, W. (1994). Substrate activation of brewer's yeast pyruvate decarboxylase is abolished by mutation of cysteine 221 to serine. Biochemistry 33, 5630-5635.

Baburina, I., Moore, D. J., Volkov, A., Kahyaoglu, A., Jordan, F. and Mendelsohn, R. (1996). Three of four cysteines, including that responsible for substrate activation, are ionized at pH 6.0 in yeast pyruvate decarboxylase: evidence from Fourier transform infrared and isoelectric focusing studies. Biochemistry 35, 10249-10255.

Baburina, I., Li, H., Bennion, B., Furey, W. and Jordan, F. (1998). Interdomain information transfer during substrate activation of yeast pyruvate decarboxylase: the interaction between cysteine 221 and histidine 92. Biochemistry 37, 1235-1244.

Bachmann, R. and Riese, J. (2005). From promise to profit. Practical considerations for adopting industrial biotechnology platforms. Industrial Biotechnology 1, 9-15.

Ballesteros, A., Bornscheuer, U., Capewell, A., Combes, D., Condoret, J. S., Koenig, K., Kolisis, F. N., Marty, A., Menge, U., Scheper, T., Stamatis, H. and Xenakis, A. (1995). Enzyme in non-conventional phases. Biocatalysis and Biotransformation 13, 1-42.

Banat, I. M. and Marchant, R. (1995). Characterization and potential industrial applications of five novel, thermotolerant, fermentative, yeast strains. World Journal of Microbiology and Biotechnology 11, 304-306.

126 References

Banat, I. M., Nigam, P., Singh, D., Marchant, R. and McHale, A. P. (1998). Review: Ethanol production at elevated temperatures and alcohol concentrations: Part I - Yeasts in general. World Journal of Microbiology and Biotechnology 14, 809-821.

Becvarova, H. and Hanc, O. (1963). Production of phenylacetylcarbinol by various yeast species. Folia Microbiology (Praha) 8, 42-47.

Beilen, J. B. v. and Li, Z. (2002). Enzyme technology: an overview. Current Opinion in Biotechnology 22, 338-344.

Bensky, D. and Gamble, A. (1993). Chinese herbal medicine: materia medica. Seattle, Washington: Eastland Press.

Biotechnology Industry Association (2005). Biotechnology industry facts.

Biselli, M., Kragl, U. and Wandrey, C. (1995). Reaction engineering for enzyme- catalyzed biotransformations. Enzyme catalysis in organic synthesis: A comphrehensive handbook. pp. 89-155. edited by K. Drauz and H. Waldmann. New York: Weinheim VCH.

Blanco, R. M., Halling, P., Bastida, A., Cuesta, C. and Guisan, J. M. (1992). Effect of immiscible organic solvents of activity/stability of native chymotrypsin and immobilized-stabilized derivatives. Biotechnology and Bioengineering 39, 75-84.

Bornscheuer, U. T. and Pohl, M. (2001). Improved biocatalysts by directed evolution and rational protein design. Current Opinion in Chemical Biology 5, 137-143.

Bradford, M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72, 249-254.

127 References

Bringer-Meyer, S. and Sahm, H. (1988). Acetoin and phenylacetylcarbinol formation by the pyruvate decarboxylases of Zymomonas mobilis and Saccharomyces cerevisiae. Biocatalysis 1, 321-331.

Burton, S. G. (2001). Development of bioreactors for application of biocatalysts in biotransformations and bioremediation. Pure and Applied Chemistry 73, 77-83.

Burton, S. G., Cowan, D. A. and Woodley, J. M. (2002). The search for the ideal biocatalyst. Nature Biotechnology 20, 37-45.

Calvo, M. V., Plou, F. J., Pastor, E. and Ballesteros, A. (1995). Effect of chemical modification of isoenzymes A and B from C. rugosa on their activity and stability. Biotechnology Letters 17, 171-176.

Candy, J. M. and Duggleby, R. G. (1998). Structure and properties of pyruvate decarboxylases and site directed mutagenesis of the Zymomonas mobilis enzyme. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1385, 323-338.

Cardillo, R., Servi, S. and Tinti, C. (1991). Biotransformation of unsaturated aldehydes by microorganisms with pyruvate decarboxylase activity. Applied Microbiology and Biotechnology 36, 300-303.

Carvalho, C. M. L. and Cabral, J. M. S. (2000). Reverse micelles as reaction media for lipases. Biochimie 82, 1063-1085.

Castro, G. R. and Knubovets, T. (2003). Homogeneous biocatalysis in organic solvents and water-organic mixtures. Critical Reviews in Biotechnology 23, 195-231.

Cavicchioli, R., Curmi, P. M. G., Siddiqui, K. S. and Thomas, T. (2006). Proteins from psychrophiles. In Extremophiles-Methods in Microbiology. edited by F. A. Rainey and A. Oren. London: Elsevier.

128 References

Chandra Raj, K., Ingram, L. O. and Maupin-Furlow, J. A. (2001). Pyruvate decarboxylase: a key enzyme for the oxidative metabolism of lactic acid by Acetobacter pasteurianus. Archives in Microbiology 176, 443-451.

Chang, B. S. and Mahoney, R. R. (1995). Enzyme thermostabilization by bovine serum albumin and other proteins: evidence for hydrophobic interactions. Biotechnology and Applied Biochemistry 22, 203-214.

Charters, M. L. (2005). Ephedra viridis.

Chen, A. K. L., Breuer, M., Hauer, B., Rogers, P. L. and Rosche, B. (2005). pH shift enhancement of Candida utilis pyruvate decarboxylase production. Biotechnology and Bioengineering 92, 183-188.

Chen, K. K. and Schmidt, C. F. (1926). The action and clinical use of ephedrine. Journal of American Pharmacy Association 87, 836-842.

Chow, Y. S., Shin, H. S., Adesina, A. A. and Rogers, P. L. (1995). A kinetic model for the deactivation of pyruvate decarboxylase (PDC) by benzaldehyde. Biotechnology Letters 17, 1201-1206.

Csuk, R. and Glanzer, B. I. (1991). Baker's yeast mediated transformations in organic chemistry. Chemical Reviews 91, 49-97.

Cull, S. G., Holbrey, J. D., Vargas-Mora, V., Seddon, K. and Lye, G. J. (2000). Replacements for organic solvents in multiphase bioprocess operations. Biotechnology and Bioengineering 69, 227-233.

Cupo, P., El-Deiry, W., Whitney, P. L. and Awad, W. M. J. (1980). Stabilization of proteins by guanidination. Journal of Biological Chemistry 255, 10828-10833.

129 References

Davis, B. G. (2003). Chemical modification of biocatalysts. Current Opinion in Biotechnology 14, 1-8.

Davis-Searles, P. R., Saunders, A. J., Erie, D. A., Winzor, D. J. and Pielak, G. J. (2001). Interpreting the effects of small uncharged solutes on protein-folding equilibria. Annual Review of Biophysic and Biomolecular Structure 30, 271-306. de Smet, M. J., Kingma, J. and Withold, B. (1978). The effect of toluene on the structure and permeability of the outer cytoplasmic cell membranes of E. coli. Biochimica et Biophysica Acta (BBA) 506, 64-80.

Demirjian, D. C., Moris-Varas, F. and Cassidy, C. S. (2001). Enzymes from extremophiles. Current Opinion in Chemical Biology 5, 144-151. di Lernia, I., Schiraldi, C. and De Rosa, M. (2002). Trehalose production at high temperature exploiting an immobilized cell bioreactor. Extremophiles 6, 341-347.

Dobritzsch, D., Konig, S. and Hubner, G. (1996). Attempts to crystallize chemical modified pyruvate decarboxylase from brewer's yeast. In Biochemistry and physiology of thiamin diphosphate enzymes, pp. 177-185. edited by H. Bisswanger and A. Schellenberger. Prien, Blaubeuren, Germany: A. & C. Intemann, Wissenschaftlicher.

Dordick, J. S., Khmelnitsky, Y. L. and Sergeeva, M. V. (1998). The evolution of biotransformation technologies. Current Opinion in Microbiology 1, 311-318.

Dutta, A., Pedersen, H. and Chin, C. K. (1994). Two-phase culture system for plant cells. Annals of the New York Academy of Sciences 745, 251-260.

Dyda, F., Furey, W., Swaminathan, S., Sax, M., Farrenkopf, B. and Jordan, F. (1993). Catalytic centers in the thiamin diphosphate dependent enzyme pyruvate decarboxylase at 2.4 A resolution. Biochemistry 32, 6165-6170.

130 References

Eppendorfer, S., Konig, S., Golbik, R., Neef, H., Lehle, K., Jaenicke, R., Schellenberger, A. and Hubner, G. (1993). Effects of metal ions, thiamin diphosphate analogues and subunit interactions on the reconstitution behaviour of pyruvate decarboxylase from brewer`s yeast. Biological Chemistry Hoppe-Seyler 374, 1129- 1134.

Eramian, D., Strickland, D., Dry, L., Coghill, K., Winters, P., Fischer, S., Schulz, D., Carstoiu, D. S., Resse, E. and Keflemarian, Y. (2005). The guide to biotechnology. Washington DC: Biotecnology Industry Association.

Ertan, H., Kazan, D. and Erarslan, A. (1997). Cross-linked stabilization of Escherichia coli penicillin G acylase against pH by dextran-dialdehyde polymers. Biotechnology Techniques 11, 225-229.

Faber, K. (1999). Biotransformations in Organic Chemistry 4th edition. Germany: Springer-Verlag.

Fadnavis, N. W., Reddy, N. P. and Bhalerao, U. T. (1989). Reverse micelles, an alternative to aqueous medium for microbial reactions: Yeast-mediated resolution of α− amino acids in reverse micelles. Journal of Organic Chemistry 54, 3218-3221.

Fernandez, M., Fragozo, A., Cao, R., Banos, M., Villalonga, M. L. and Villalonga, R. (2002a). Stabilization of trypsin by chemical modification with β-cyclodextrin monoaldehyde. Biotechnology Letters 24, 1455-1459.

Fernandez, M., Fragozo, A., Cao, R., Banos, M. and Villalonga, R. (2002b). Chemical conjugation of trypsin with monoamine derivatives of cyclodextrins: Catalytic and stability properties. Enzyme and Microbial Technology 31, 543-548.

Folch, J., Lees, M. and Sloane Stanley, G. H. (1957). A simple method for the isolation and purification of total lipides from animal tissues. Journal of Biological Chemistry 226, 497-509.

131 References

Fuganti, C. and Grasselli, P. (1977). Transformation of non conventional substrates by fermenting baker's yeast: production of optically active methyl diol from aldehyde. Chemistry and Industry (London) 24, 983-988.

Fuganti, C., Grasselli, P., Poli, G., Servi, S., Stefano, S. and Zorzella, A. (1988). Decarboxylation incorporation of α-oxobutyrate and α-oxovalarate into R-(α)- hydroxyethyl- and n-propyl ketones on reaction with aromatic and α,β-unsaturated aldehyde in baker's yeast. Journal of Chemical Society Chemical Communications 24, 1619-1621.

Furey, W., Arjunan, P., Chen, L., Sax, M., Guo, F. and Jordan, F. (1998). Structure- function relationships and flexible tetramer assembly in pyruvate decarboxylase revealed by analysis of crystal structure. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1385, 253-270.

Gåseidnes, S., Synstad, B., Nielsen, J. E. and Eijsink, V. G. H. (2003). Rational engineering of the stability and the catalytic performance of enzymes. Journal of Molecular Catalysis B: Enzymatic 21, 3-8.

Geren, L., Tuls, J., O'Brien, P., Millett, F. and Peterson, J. A. (1986). The involvement of carboxylate groups on putidaredoxin in the reaction with putidaredoxin reductase. Journal of Biological Chemistry 261, 15491-15495.

Goetz, G., Iwan, P., Hauer, B., Breuer, M. and Pohl, M. (2001). Continuous production of (R)-phenylacetylcarbinol in an enzyme-membrane reactor using a potent mutant of pyruvate decarboxylase from Zymomonas mobilis. Biotechnology and Bioengineering 74, 317-325.

Goller, K. and Galinski, E. (1999). Protection of a model enzyme (lactate dehydrogenase) against heat, urea and freeze-thaw treatment by compatible solute additives. Journal of Molecular Catalysis B: Enzymatic 7, 37-45.

132 References

Gomez, L., Ramírez, H. and Villalonga, R. (2000). Stabilization of invertase by modification of sugar chains with chitosan. Biotechnology Letters 22, 347-350.

Gounaris, A. D., Turkenkoff, I., Buckwald, S. and Young, A. (1971). Pyruvate decarboxylase. I. Protein dissociation into subunits under conditions in which thiamine pyrophosphate is released. The Journal of Biological Chemistry 246, 1302-1309.

Gounaris, A. D., Turkenkopf, I., Civerchia, L. L. and Greenlie, J. (1975). Pyruvate decarboxylase III. Specificity restrictions for thiamine pyrophosphate in the protein association step, sub-unit structure. Biochimica et Biophysica Acta 405, 492-499.

Govardhan, C. P. (1999). Crosslinking of enzymes for improved stability and performance. Current Opinion in Biotechnology 10, 331-335.

Green, D. E., Herbert, D. and Subrahmanyan, V. (1941). Carboxylase. Journal of Biological Chemistry 138, 327.

Groger, G., Schmauder, H. P. and Mothes, K. (1966). Untersuchung uber die gewinnung von L-phenylacetylcarbinol. Zeitschrift fur Allgemeine Mikrobiologie 6, 275-287.

Gunawan, C., Satianegara, G., Chen, A. K. L., Breuer, M., Hauer, B., Rogers, P. L. and Rosche, B. (2006). Yeast pyruvate decarboxylases: variation in biocatalytic characteristics for (R )-phenylacetylcarbinol production. FEMS Yeast Research DOI:10.1111/j.1567-1364.2006.00138.x.

Guo, F., Zhang, D., Kahyaoglu, A., Farid, R. S. and Jordan, F. (1998). Is a hydrophobic amino acid required to maintain the reactive V conformation of thiamin at the active center of thiamin diphosphate-requiring enzymes? Experimental and computational studies of isoleucine 415 of yeast pyruvate decarboxylase. Biochemistry 37, 13379-13391.

133 References

Hauer, B., Breuer, M., Rogers, P. L., Sandford, V. and Rosche, B. (2003). Process for production of R-phenylacetylcarbinol by an enzymatic process in a two-phase system. WO 03/020942

Hernaiz, M. J., Sanchez-Montero, J. M. and Sinisterra, J. V. (1999). Modification of purified lipases form Candida rugosa with polyethyleneglycol. Enzyme and Microbial Technology 24, 181-190.

Hilderbrandt, G. and Klavehn, W. (1932). Verfahren zur Herstellung von 1-1-Phenyl- 2-methylaminopropan-1-ol. German patent 548 459

Hilderbrandt, G. and Mann, W. K. (1934). Manufacture of Laevo-1-phenyl-2- methylaminopropanol-1. USPTO 1956950

Hirakawa, H., Kamiya, N., Yata, T. and Nagamune, T. (2003). Regioselective reduction of a steroid in a reversed micellar system with enzymatic NADH- regeneration. Biochemical Engineering Journal 16, 35-40.

Holloway, P. and Subden, R. E. (1993). The isolation and nucleotide sequence of the pyruvate decarboxylase gene from Kluyveromyces marxianus. Current Genetics 24, 274-277.

Hong, J., Sun, S., Derrick, T., Larive, C., Schowen, K. B. and Schowen, R. L. (1998). Transition-state theoretical interpretation of the catalytic power of pyruvate decarboxylases: the roles of static and dynamical considerations. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1385, 187-200.

Hounsa, C.-G., Brandt, E. V., Thevelein, J., Hohmann, S. and Prior, B. A. (1998). Role of trehalose in survival of Saccharomyces cerevisiae under osmotic stress. Microbiology 144, 671-680.

134 References

Hubner, G., Konig, S. and Schellenberger, A. (1988). The functional role of thiol groups of pyruvate decarboxylase from brewer's yeast. Biomedica Biochimica Acta 47, 9-18.

Hubner, G., Tittman, K., Killenberg-Jabs, M., Schaffner, J., Spinka, M., Neef, H., Kern, D., Kern, G., Schneider, G., Wikner, C. and Ghisla, S. (1998). Activation of thiamin diphosphate in enzymes. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1385, 221-228.

Hurlbut, J. A. and Carr, J. R. (1998). Solid-phase extraction cleanup and liquid chromatography with ultraviolet detection of ephedrine alkaloids in herbal products. Journal of AOAC International 81, 1121-1127.

Iding, H., Siegert, P., Mesch, K. and Pohl, M. (1998). Application of α-keto acid decarboxylases in biotransformations. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1385, 307-322.

Inada, Y., Furukawa, M., Sasaki, H., Kodera, Y., Hiroto, M., Nishimura, H. and Matsushima, A. (1995). Biomedical and biotechnological applications of PEG- and PM-modified proteins. Trends in Biotechnology 13, 86-91.

Ingram, L. O. and Buttke, T. M. (1982). Effects of alcohols on micro-organisms. Advances in Microbial Physiology 25, 253-300.

Ishige, T., Honda, K. and Shimizu, S. (2005). Whole organism biocatalysis. Current Opinion in Chemical Biology 9, 174-180.

Isken, S. and de Bont, J. A. M. (1998). Bacteria tolerant to organic solvents. Extremophiles 2, 229-238.

Iwan, P., Goetz, G., Schmitz, S., Hauer, B., Breuer, M. and Pohl, M. (2001). Studies on the continuous production of R-(−)-phenylacetylcarbinol in an enzyme-membrane reactor. Journal of Molecular Catalysis B: Enzymatic 11, 387-396.

135 References

Jaeger, K.-E. (2004). Protein technologies and commercial enzymes. White is the hype - biocatalysts on the move. Current Opinion in Biotechnology 15, 269-271.

Jaenicke, R. and Bohm, G. (1998). The stability of proteins in extreme environments. Current Opinion in Structural Biology 8, 738-748.

Jordan, F., Baburina, I., Gao, Y., Kahyaoglu, A., Nemeria, N., Volkov, A., Yi, J. Z., Zhang, D., Machado, R., Guest, J. R., Furey, W. and Hohmann, S. (1996). New insights to the regulation of thiamin diphosphate decarboylases by substrate and ThDP.Mg(II). In Biochemistry and physiology of thiamin diphosphate enzymes, pp. 53. edited by H. Bisswanger and A. Schellenberger. Prien, Blaubeuren, Germany: A. & C. Intemann, Wissenschaftlicher.

Jordan, F., Nemeria, N., Guo, F., Baburina, I., Gao, Y., Kahyaoglu, A., Li, H., Wang, J., Yi, J. Z., Guest, J. R. and Furey, W. (1998). Regulation of thiamin diphosphate-dependent 2-oxo acid decarboxylases by substrate and thiamin diphosphate.Mg (II) - evidence for tertiary and quarternary interactions. Biochimica et Biophysica Acta 1385, 287-306.

Juni, E. and Heym, G. A. (1968). Properties of yeast pyruvate decarboxylase and their modification by proteolytic enzymes. II. Selective alteration by yeast proteases. Archives of Biochemistry and Biophysics 127, 89-100.

Kaushik, J. K. and Bhat, R. (1998). Thermal stability of proteins in aqueous polyol solutions. Journal of Physical Chemistry B 102, 7058-7066.

Kern, D., Kern, G., Neef, H., Tittman, K., Killenberg-Jabs, M., Wikner, C., Schneider, G. and Hubner, G. (1997). How thiamin diphosphate is activated in enzymes. Science 275, 67-70.

136 References

Kiffe, M., Nokihara, K. and Matsunag, T. (1995). Purification of docohexaenoic acid produced by marine microalga Isochrysis galbana. Journal of Marine Biotechnology 2, 139-142.

Killenberg-Jabs, M., Jabs, A., Lilie, R., Golbik, R. and Hubner, G. (2001). Active oligomeric states of pyruvate decarboxylase and their functional characterization. European Journal of Biochemistry 268, 1698-1704.

Killenberg-Jabs, M., Kern, G., Hubner, G. and Golbik, R. (2002). Folding and stability of different oligomeric states of thiamin diphosphate dependent homomeric pyruvate decarboxylase. Biophysical Chemistry 96, 259-271.

Kirk, O., Borchert, T. V. and Fuglsang, C. C. (2002). Industrial enzyme applications. Current Opinion in Biotechnology 13, 345-351.

Klibanov, A. M., Samokhin, G. P., Martinek, K. and Berezin, I. V. (1977). A new approach to preparative enzymatic synthesis. Biotechnology and Bioengineering 19, 1351-1361.

Kluger, R. (1987). Thiamin diphosphate: a mechanistic update on enzymic and nonenzymic catalysis of decarboxylation. Chemical Reviews 87, 863-876.

Kobayashi, M. and Takatsu, K. (1994). Cross-linked stabilization of trypsin with dextran-dialdehyde. Bioscience, Biotechnology, and Biochemistry 58, 275-278.

Kochlar, S. and Dua, R. D. (1984). An active-site carboxyl group in liquefying α- amylase: specific chemical modification. Bioscience Reports 4, 613-619.

Konig, S., Hubner, G. and Schellenberger, A. (1990). Cross-linking of pyruvate decarboxylase. Characterization of the native and substrate-activated enzyme states. Biomedica Biochimica Acta 49, 465-471.

137 References

Konig, S., Svergun, D., Koch, M. H., Hubner, G. and Schellenberger, A. (1992). Synchrotron radiation solution X-ray scattering study of the pH dependence of the quaternary structure of yeast pyruvate decarboxylase. Biochemistry 31, 8726-8731.

Konig, S. (1998). Subunit structure, function and organization of pyruvate decarboxylases from various organisms. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1385, 271-286.

Koops, B. C., Verheij, H. M., Slotboom, A. J. and Egmond, M. R. (1999). Effect of chemical modification on the activity of lipases in organic solvents. Enzyme and Microbial Technology 25, 622-631.

Kren, V., Crout, D., Dalton, H., Hutchinson, D., Konig, W., Turner, M., Dean, G. and Thomson, N. (1993). Pyruvate decarboxylase: a new enzyme for the production of acyloins by biotransformation. Journal of the Chemical Society, Chemical Communications 4, 341-343.

Kuo, D. J. and Jordan, F. (1983). Direct spectroscopic observation of a brewer's yeast pyruvate decarboxylase-bound enamine intermediate produced from a suicide substrate. Evidence for nonconcerted decarboxylation. Journal of Biological Chemistry 258, 13415-13417.

Lagovic, B. B., Upcic, J. R., Esaric, M. M. and Aric, V. M. (2005). Lipid analysis of the plasma membrane and mitochondria of brewer's yeast. Folia Microbiology 50, 24- 30.

Larsson, K. M., Adlercreutz, P. and Mattiasson, B. (1990). Enzymatic catalysis in microemulsions: enzyme reuse and product recovery. Biotechnology and Bioengineering 36, 135-141.

Lee, S. H. and Park, T. H. (1995). Two-phase cultivation of insect cells for production of recombinant protein. Biotechnology and Bioengineering 9, 719-724.

138 References

Lehmann, M., Pasamontes, L., Lassen, S. F. and Wyss, M. (2000). The consensus concept for thermostability engineering of proteins. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1543, 408-415.

Leksawasdi, N., Breuer, M., Hauer, B., Rosche, B. and Rogers, P. L. (2003). Kinetics of pyruvate decarboxylase deactivation by benzaldehyde. Biocatalysis and Biotransformation 21, 315-320.

Leksawasdi, N. (2004). Kinetics and modelling of enzymatic process for R- phenylacetylcarbinol (PAC) production. (PhD Thesis). University of New South Wales: Sydney

Leksawasdi, N., Chow, Y. S., Breuer, M., Hauer, B., Rosche, B. and Rogers, P. L. (2004). Kinetic studies and model validation of enzymatic (R)-phenylacetylcarbinol batch biotransformation process. Journal of Biotechnology 111, 179-189.

Leksawasdi, N., Rogers, P. L. and Rosche, B. (2005). Improved enzymatic two-phase biotransformation for (R)-phenylacetylcarbinol: effect of dipropylene glycol and modes of pH control. Biocatalysis and Biotransformation 23, 445-451.

Lennoury, A., Gardiner, J. M. and Stephens, G. (2005). Hydrogenation of C-C double bonds in an ionic liquid reaction system using the obligate anaerobe, Sporomusa termitida. Biotechnology Letters 27, 161-165.

Leon, R., Fernandes, P., Pinheiro, H. M. and Cabral, J. M. S. (1998). Whole-cell biocatalysis in organic media. Enzyme and Microbial Technology 23, 483-500.

Li, H. and Jordan, F. (1999). Effect of substitution of tryptophan 412 in the substrate activation pathway of yeast pyruvate decarboxylase. Biochemistry 38, 10004-10012.

Lin, T. Y. and Koshland, D. E. (1969). Carboxyl group modification and the activity of lysozyme. Journal of Biological Chemistry 244, 505-508.

139 References

Lindman, B., Shinodo, K., Olsson, U., Anderson, D., Karlstrom, G. and Wennerstrom, H. (1989). On the demonstration of bicontinuous structures in microemulsions. Colloids and Surfaces 38, 205-224.

Liu, M., Sergienko, E. A., Guo, F., Wang, J., Tittman, K., Hubner, G., Furey, W. and Jordan, F. (2001). Catalytic acid-base groups in yeast pyruvate decarboxylase. I. Site-directed mutagenesis and steady-state kinetic studies on the enzyme with the D28A, H114F, H115F, and E477Q substitutions. Biochemistry 40, 7355-7368.

Lohmann, K. (1933). Uber Phosphorylierung und Dephosphorylierung. Bildung der natürlichen Hexosemonophosphorsäure aus ihren Komponenten. Biochemische Zeitschrift 262, 137-151.

Lohmann, K. and Schuster, P. (1937). Untersuchungen uber die cocarboxylase. Biochemische Zeitschrift 294, 188-214.

Long, A. and Ward, O. P. (1989). Biotransformation of benzaldehyde by Saccharomyces cerevisiae. Characterization of fermentation and toxicity effects of substrate and products. Biotechnology and Bioengineering 34, 933-941.

Lowe, S. E. and Zeikus, J. G. (1992). Purification and characterization of pyruvate decarboxylase from Sarcina ventriculi. Journal of General Microbiology 138, 803-807.

Lu, G., Dobritzsch, D., Konig, S. and Schneider, G. (1997). Novel tetramer assembly of pyruvate decarboxylase from brewer's yeast observed in a new crystal form. FEBS Letters 403, 249-253.

Lye, G. J., Dalby, P. A. and Woodley, J. M. (2002). Better biocatalytic processes faster: new tools for the implementation of biocatalysis in organic synthesis. Organic Process Research and Development 6, 434-440.

140 References

Mahmoud, W. M., El-Sayed, A.-H. M. M. and Coughlin, R. W. (1990a). Production of L-phenylacetylcarbinol by immobilized yeast cells: I. Batch fermentation. Biotechnology and Bioengineering 36, 47-54.

Mahmoud, W. M., El-Sayed, A.-H. M. M. and Coughlin, R. W. (1990b). Production of L-phenylacetylcarbinol by immobilized yeast: II. Semicontinuous fermentation. Biotechnology and Bioengineering 36, 55-63.

Mahmoud, W. M., El-Sayed, A.-H. M. M. and Coughlin, R. W. (1990c). Effect of β- cyclodextrin on production of L-phenylacetylcarbinol by immobilised cells of Saccharomyces cerevisiae. Biotechnology and Bioengineering 36, 256-262.

Marshall, J. J. and Rabinowitz, M. L. (1974). Enzyme stabilization by covalent attachment of carbohydrate. Archives of Biochemistry and Biophysics 167, 777-779.

Martins, L. O., Huber, R., Huber, H., Stetter, K. O., Da Costa, M. S. and Santos, H. (1997). Organic solutes in hyperthemophilic Archaea. Applied and Environmental Microbiology 63, 896-902.

Masarova, J., Mislovicova, D., Gemeiner, P. and Michalkova, E. (2001). Biotechnology and Applied Biochemistry 34, 127-133.

Matsumoto, T., Takahashi, S., Kaieda, M., Ueda, M., Tanaka, A., Fukuda, H. and Kondo, A. (2001). Yeast whole-cell biocatalyst constructed by intracellular overproduction of Rhizopus oryzae lipase is applicable to biodiesel fuel production. Applied Microbiology and Biotechnology 57, 515-520.

McCoy, M. (1999). Biocatalysis grows for drug synthesis. Chemical and Engineering News 77, 1-64.

Means, G. E. and Feeney, R. E. (1971). Chemical modification of protein. California: Holdeh-Day, Inc.

141 References

Menon-Rudolph, S., Nishikawa, S., Zeng, X. and Jordan, F. (1992). Rate of decarboxylation, monitored via the key enzyme-bound enamine, of conjugated .alpha.- keto acids by pyruvamide activated pyruvate decarboxylase is kinetically competent with turnover. Journal of American Chemical Society 114, 10110-10112.

Minton, A. P. (1998). Molecular crowding: analysis of effects of high concentrations of inert cosolutes on biochemical equilibria and rates in term of volume exclusion. Methods in Enzymology 295, 127-149.

Mochizuki, N., Hiramatsu, S., Sugai, T., Ohta, H., Morita, H. and Itokawa, H. (1995). Improved conditions for the production and characterization of 1-arylpropane- 1,2-diols and related compounds. Bioscience, Biotechnology and Biochemistry 59, 2282-2291.

Mozhaev, V. V., Berezin, I. V. and Martinek, K. (1988a). Structure-stability relationship in proteins: fundamental tasks and strategy for the development of stabilzied enzyme catalysts for biotechnology. Critical Reviews in Biochemistry 3, 235- 281.

Mozhaev, V. V., Melik-Nubarov, N. S., Galkantaite, N. S., Denis, G. J., Butkus, E. P., Zaslavsky, B. Y., Mestechkina, M. and Martinek, K. (1988b). Protein stabilization via hydrophilization. Covalent modification of trypsin and α- chymotrypsin. European Journal of Biochemistry 173, 147-154.

Munch, O. and Tritsch, D. (1990). Irreversible thermoinactivation of glucoamylase from Aspergillus niger and thermostabilization by chemical modification of carboxyl groups. Biochimica et Biophysica Acta 1041, 111-116.

Netrval, J. and Vojtisek, V. (1982). Production of phenyl acetyl carbinol in various yeast species. European Journal of Applied Microbiology and Biotechnology 16, 35-38.

Neuberg, C. and Karczag, L. (1911). Uber zuckerfreie Hefegarungen. IV. Carboxylase, ein neues Enzym der Hefe. Biochemische Zeitschrift 36, 68-81.

142 References

Neuser, F., Zorn, H., Richter, U. and Berger, R. G. (2000). Purification, characterisation and cDNA sequencing of pyruvate decarboxylase from Zygosaccharomyces bisporus. Biological Chemistry 381, 349-353.

Neves, C., Da Costa, M. S. and Santos, H. (2005). Compatible solutes of the hyperthermophile Palaeococcus ferrophilus: osmoadaptation and thermoadaptation in the order Thermococcales. Applied and Environmental Microbiology 71, 8091-8098.

Nikolova, P. and Ward, O. O. (1991). Production of L-phenylacetylcarbinol by biotransformation: product and by-product formation and activities of the key enzymes in wild-type and ADH isoenzyme mutants of Saccharomyces cerevisiae. Biotechnology and Bioengineering 38, 493-398.

Nikolova, P. and Ward, O. P. (1992a). Whole cell yeast biotransformation in two- phase systems: effect of solvent on product formation and cell structure. Journal of Industrial Microbiology 10, 169-177.

Nikolova, P. and Ward, O. P. (1992b). Production of phenylacetylcarbinol by biotransformation using baker's yeast in two-phase systems. In Biocatalysis in non- conventional media, pp. 675-680. edited by J. Tramper. Amsterdam: Elsevier Science Publishers B.V.

Nikolova, P. and Ward, O. P. (1994). Effect of support matrix on ratio of product to byproduct formation in L-phenylacetylcarbinol synthesis. Biotechnology Letters 16, 7- 10.

Ó'Fágáin, C. (2003). Enzyme stabilization - recent experimental progress. Enzyme and Microbial Technology 33, 137-149.

Ohata, H., Ozaki, K., Konishi, J. and Tsuchihashi, G. (1986). Reductive C2 homologation of substituted benzaldehyde by fermenting baker's yeast. Agricultural and Biological Chemistry 50, 1261-1266.

143 References

Ou, W.-B., Park, Y.-D. and Zhou, H.-M. (2001). Molecular mechanism for osmolyte protection of creatine kinase against guanidine denaturation. European Journal of Biochemistry 268, 5901-5911.

Panke, S., Held, M. and Wubbolts, M. (2004). Trends and innovations in industrial biocatalysis for the production of fine chemicals. Current Opinion in Biotechnology 15, 272-279.

Paradkar, V. M. and Dordick, J. S. (1994a). Mechanism of extraction of chymotrypsin into isooctane at very low concentrations of aerosol OT in the absence of reverse micelles. Biotechnology and Bioengineering 43, 529-540.

Paradkar, V. M. and Dordick, J. S. (1994b). Aqueous-like activity of α-chymotrypsin dissolved in nearly anhydrous organic solvents. Journal of American Chemical Society 116, 5009-5010.

Perutz, M. F. (1978). Electrostatic effects in proteins. Science 201, 1187-1191.

Pinto-Sousa, A. M., Cabral, J. M. S. and Aires-Barros, M. R. (1996). Stability of a Fusarium solani pisi recombinant cutinase in phosphatidylcholine reversed micelles. Biotechnology Letters 18, 583-586.

Pohl, M. (1997). Protein design on pyruvate decarboxylase (PDC) by site-directed mutagenesis: Application to mechanistical investigations and tailoring PDC for the use in organic synthesis. Advances in Biochemical Engineering and Biotechnology 58, 15- 43.

Pohl, M., Lingen, B. and Muller, M. (2002). Thiamin-diphosphate-dependent enzymes: new aspects of asymmetric C-C bond formation. Chemistry A European Journal 8, 5288-5295.

144 References

Pronk, J. T., Steensma, H. Y. and van Dijken, J. P. (1996). Pyruvate metabolism in Saccharomyces cerevisiae. Yeast 12, 1607-1633.

Reti, L. (1953). Ephedra Bases. In The Alkaloids, pp. 339-362. edited by R. H. F. Manske and H. L. Holmes. New York: Academic Press.

Roberts, M. J., Bentley, M. D. and Harris, J. M. (2002). Chemistry for peptide and protein PEGylation. Advanced Drug Delivery Reviews 54, 459-476.

Robl, I., Grazzl, R., Tanner, W. and Opekarova, M. (2000). Properties of a reconstituted eukaryotic hexose/proton symporter solubilized by structurally related non-ionic detergents: specific requirement of phosphatidylcholine for permease stability. Biochimica et Biophysica Acta (BBA)/Biomembranes 1463, 407-418.

Rogers, P. L., Shin, H. S. and Wang, B. (1997). Biotransformation of L-ephedrine production. Advances in BIochemical Engineering and Biotehcnology 56, 33-59.

Rosche, B., Sandford, V., Breuer, M., Hauer, B. and Rogers, P. L. (2001). Biotransformation of benzaldehyde into (R)-phenylacetylcarbinol by filamentous fungi or their extracts. Applied Microbiology and Biotechnology 57, 309-315.

Rosche, B., Leksawasdi, N., Sandford, V., Breuer, M., Hauer, B. and Rogers, P. L. (2002a). Enzymatic (R)-phenylacetylcarbinol production in benzaldehyde emulsions. Applied Microbiology and Biotechnology 60, 94-100.

Rosche, B., Sandford, V., Breuer, M., Hauer, B. and Rogers, P. L. (2002b). Enhanced production of R-phenylacetylcarbinol (R-PAC) through enzymatic biotransformation. Journal of Molecular Catalysis B: Enzymatic 19, 109-115.

Rosche, B., Breuer, M., Hauer, B. and Rogers, P. L. (2003a). Increased pyruvate efficiency in enzymatic production of (R)-phenylacetylcarbinol. Biotechnology Letters 25, 847-851.

145 References

Rosche, B., Breuer, M., Hauer, B. and Rogers, P. L. (2003b). Screening of yeasts for cell-free production of (R)-phenylacetylcarbinol. Biotechnology Letters 25, 841-845.

Rosche, B., Breuer, M., Hauer, B. and Rogers, P. L. (2005a). Role of pyruvate in enhancing pyruvate decarboxylase stability towards benzaldehyde. Journal of Biotechnology 115, 91-99.

Rosche, B., Breuer, M., Hauer, B. and Rogers, P. L. (2005b). Cells of Candida utilis for in vitro (R)-phenylacetylcarbinol production in an aqueous/octanol two-phase reactor. Biotechnology Letters 27, 575-581.

Salleh, A. B., Basri, M., Taib, M., Jasmani, H., Rahman, R. N. Z. A., Rahman, M. B. A. and Razak, C. N. A. (2002). Modified enzymes for reactions in organic solvents. Applied Biochemistry and Biotechnology 102-103.

Sandford, V., Breuer, M., Hauer, B., Rogers, P. L. and Rosche, B. (2005). (R)- phenylacetylcarbinol production in aqueous/organic two-phase systems using partially purified pyruvate decarboxylase from Candida utilis. Biotechnology and Bioengineering 91, 190-198.

Sarcar, S., Jain, T. K. and Maitra, A. (1991). Activity and stability of yeast alcohol dehydrogenase (ADH) entrapped in aerosol OT reverse micelles. Biotechnology and Bioengineering 39, 474-478.

Satianegara, G., Breuer, M., Hauer, B., Rogers, P. L. and Rosche, B. (2006a). Enzymatic (R)-phenylacetylcarbinol production in a benzaldehyde emulsion system with whole cell Candida utilis. Applied Microbiology and Biotechnology 70, 170-175.

Satianegara, G., Rogers, P. L. and Rosche, B. (2006b). Comparative studies on enzyme preparations and role of cell components for (R)-phenylacetylcarbinol production in a two-phase biotransformation. Biotechnology and Bioengineering 94, 1189-1195.

146 References

Schein, C. H. (1990). Solubility as a function of protein structure and solvent components. Biotechnology 8, 308-317.

Schellenberger, A. (1967). Struktur und wirkungsweise des aktiven zentrums der hefe- pyruvatdecarboxylase. Angewandte Chemie 79, 1050-1061.

Schellenberger, A. and Hubner, G. (1968). Binding of the substrate in yeast pyruvate decarboxylase. Angewandte Chemie International Edition 7, 68-69.

Schiffer, C. A. and Dotsch, V. (1996). The role of protein-solvent interactions in protein unfolding. Current Opinion in Biotechnology 7, 428-432.

Schiraldi, C. and De Rosa, M. (2002). The production of biocatalysts and biomolecules from extremophiles. Trends in Biotechnology 20, 515-521.

Schmid, A., Dordick, J. S., Hauer, B., Kiener, A., Wubbolts, M. and Witholt, B. (2001). Industrial biocatalysis today and tomorrow. Nature 409, 258-268.

Schoevaart, R., Wolbers, M. W., Golubovic, M., Ottens, M., Kieboom, A. P. G., van RAntwijk, F., van der Wielen, L. A. M. and Sheldon, R. (2004). Preparation, optimization, and structures of cross-linked enzyme aggregates (CLEAs). Biotechnology and Bioengineering 87, 754-762.

Schorken, U. and Sprenger, G. (1998). Thiamin-dependent enzymes as catalysts in chemoenzymatic syntheses. Biochimica et Biophysica Acta 1385, 229-243.

Scopes, R. K. (1989). Effect of ethanol on glycolytic enzymes. In Alcohol toxicity in yeast and bacteria, pp. 89-100. edited by N. van Unden. Boca Raton, FL: CRC Press Inc.

Scopes, R. K. (1994). Protein purification: principles and practice, 3rd edn. New York: Springer-Verlag.

147 References

Scouten, W. (1999). New biocatalysts: essential tools for a sustainable 21st century chemical industry (Report). Council for Chemical Research: Palo Alto, CA

Sears, P., Schuster, M., Wang, P., Witte, K. and Wong, C. H. (1994). Engineering subtilisin for peptide coupling: studies on the effects of counterions and site-specific modifications on the stability of the enzyme. Journal of American Chemical Society 116, 6521-6530.

Sergienko, E. A. and Jordan, F. (2001a). Catalytic acid-base groups in yeast pyruvate decarboxylase. 3. A steady-state kinetic model consistent with the behaviour of both wild-type and variant enzymes at all relevant pH values. Biochemistry 40, 7382-7403.

Sergienko, E. A. and Jordan, F. (2001b). Catalytic acid-base groups in yeast pyruvate decarboxylase. 2. Insights into the specific roles of D28 and E477 from the rates and stereospecificity of formation of carboligase side products. Biochemistry 40, 7369-7381.

Shimizu, S. and Smith, D. J. (2004). Preferential hydration and the exclusion of cosolvents from protein surfaces. Journal of Chemical Physics 121, 1148-1154.

Shin, H. S. and Rogers, P. L. (1995). Biotransformation of benzaldehyde to L- phenylacetylcarbinol, an intermediate in L-ephedrine production, by immobilized Candida utilis. Applied Microbiology and Biotechnology 44, 7-14.

Shin, H. S. and Rogers, P. L. (1996a). Production of l-phenylacetylcarbinol (L-PAC) from benzaldehyde using partially purified pyruvate decarboxylase (PDC). Biotechnology and Bioengineering 49, 429-436.

Shin, H. S. and Rogers, P. L. (1996b). Kinetic evaluation of biotransformation of benzaldehyde to L-phenylacetylcarbinol by immobilized pyruvate decarboxylase from Candida utilis. Biotechnology and Bioengineering 49, 429-436.

148 References

Shukla, V. B. and Kulkarni, P. R. (2000). L-Phenylacetylcarbinol (L-PAC): biosynthesis and industrial applications. World Journal of Microbiology and Biotechnology 16, 499-506.

Shukla, V. B. and Kulkarni, P. R. (2001). Process parameters and reusability of the free cell mass of Torulaspora delbrueckii for the production of L-phenylacetylcarbinol (L-PAC). World Journal of Microbiology and Biotechnology 17, 301-306.

Siddiqui, K. S., Saqib, N., Rashid, M. H. and Rajoka, M. I. (1997). Thermostabilization of carboxymethylcellulase from Aspergillus niger by carboxyl group modification. Biotechnology Letters 19, 325-329.

Siddiqui, K. S., Shemsi, A. M., Anwar, M. A., Rashid, M. H. and Rajoka, M. I. (1999). Partial and complete alteration of surface charges of carboxylmethylcellulase by chemical modification: thermostabilization in water-miscible organic solvent. Enzyme and Microbial Technology 24, 599-608.

Siddiqui, K. S., Poljak, A. and Cavicchioli, R. (2004). Improved activity and stability of alkaline phosphatases from psychrophilic and mesophilic organisms by chemically modifying aliphatic or amino groups using tetracarboxy-benzophenone derivatives. Cellular and Molecular Biology 50, 657-667.

Siddiqui, K. S. and Cavicchioli, R. (2005). Improved thermal stability and activity in the cold-adapted lipase B from Candida antartica following chemical modification with oxidized polysaccharides. Extremophiles 9, 471-476.

Silver, S. and Wendt, L. (1967). Mechanism of action of phenethyl alcohol: breakdown of the cellular permeability barrier. Journal of Bacteriology 93, 560-566.

Sims, A. P. and Barnett, J. A. (1991). Levels of activity of enzymes involved in anaerobic utilization of sugars by six yeast species: Observations towards understanding the Kluyver effect. FEMS Microbiology Letters 77, 295-298.

149 References

Singer, M. A. and Lindquist, S. (1998). Multiple effect of trehalose on protein folding in vitro and in vivo. Molecular Cell 1, 639-648.

Skrika-Alexopoulos, E. and Freedman, R. B. (1993). Factors affecting enzyme characteristics of bilirubin oxidase suspensions in organic solvents. Biotechnology and Bioengineering 41, 887-893.

Smith, P. F. and Hendlin, D. (1953). Mechanism of phenylacetylcarbinol synthesis by yeast. Journal of Bacteriology 65, 440-445.

Smith, S. J., Crowley, J. H. and Parks, L. W. (1996). Transcriptional Regulation by Ergosterol in the Yeast Saccharomyces cerevisiae. Molecular and Cellular Biology 16, 5427-5432.

Smits, H. P., Haur, J., Muller, S., Hobley, T. J., Zimmermann, F. K., Hahn- Hagerdal, B., Nielsen, J. and Olsson, L. (2000). Simultaneous overexpression of enzymes of the lower part of glycolysis can enhance the fermentative capacity of Saccharomyces cerevisiae. Yeast 16, 1325-1334.

Smolders, A. J. J., Pinheiro, H. M., Noronha, P. and Cabral, J. M. S. (1991). Steroid bioconversion in a microemulsion system. Biotechnology and Bioengineering 38, 1210- 1217.

Sprenger, G. and Pohl, M. (1999). Synthetic potential of thiamin diphosphate- dependent enzymes. Journal of Molecular Catalysis B: Enzymatic 6, 145-159.

Straathof, A. J. J., Panke, S. and Schmid, A. (2002). The production of fine chemicals by biotransformations. Current Opinion in Biotechnology 13, 548-556.

Stratford, M. (1994). Another brick in the wall? Recent developments concerning the yeast cell envelope. Yeast 10, 1741-1752.

150 References

Talarico, L. A., Ingram, L. O. and Maupin-Furlow, J. A. (2001). Production of the gram-positive Sarcina ventriculi pyruvate decarboxylase in Eschericia coli. Microbiology 147, 2425-2435.

Tams, J. W. and Welinder, K. G. (1996). Unfolding and refolding of Coprinus cinereus peroxidase at high pH, in urea and at high temperature. Effect of organic and ionic additives on these processes. Biochemistry 35, 7573-7579.

The European Association of BioIndustries (2000-2005). White Biotech.

Todd, A. E., Orengo, C. A. and Thornton, J. M. (2002). Plasticity of enzyme active sites. Trends in Biochemical Sciences 27, 419-426.

Toh, H. and Doelle, H. (1997). Changes in the growth and enzyme level of Zymomonas mobilis under oxygen-limited conditions at low glucose concentrations. Archives of Microbiology 168, 46-52.

Triantafyllou, A. O., Wehtje, E., Adlercreutz, P. and Mattiasson, B. (1996). How do additives affect enzyme activity and stability in nonaqueous media? Biotechnology and Bioengineering 54, 67-76.

Tripathi, C. M., Basu, S. K., Vora, V. C., Mason, J. R. and Pirt, S. J. (1991). Biotransformation of benzaldehyde to L-phenylacetylcarbinol by immobilized yeast cell. Research and Industry 36, 159-160.

Tripathi, C. M., Agarwal, S. C. and Basu, S. K. (1997). Production of L- phenylacetylcarbinol by fermentation. Journal of Fermentation and Bioengineering 4, 487-492. van Dijken, J. P., Weusthuis, R. A. and Pronk, J. T. (1993). Kinetics of growth and sugar consumption in yeasts. Antonie Leeuwenhoek 63, 343-352.

151 References

van Hoek, P., Flikweert, M. T., van der Aart, Q., Steensma, H. Y., van Dijken, J. P. and Pronk, J. T. (1998). Effects of pyruvate decarboxylase overproduction on flux distribution at the pyruvate branch point in Saccharomyces cerevisiae. Applied and Environmental Microbiology 64, 2133-2140. van Hoek, P., van Dijken, J. P. and Pronk, J. T. (2000). Regulation of fermentative capacity and levels of glycolytic enzymes in chemostat cultures of Saccharomyces cerevisiae. Enzyme and Microbial Technology 26, 724-736.

Vasudevamurthy, M. K., Weatherley, L. R. and Lever, M. (2004). Enzyme stabilization using synthetic compensatory solutes. Biocatalysis and Biotransformation 23, 285-291.

Veronese, F. M. (2001). Peptide and protein PEGylation: a review of problems and solutions. Biomaterials 22, 405-417.

Vina, I., Karsakeich, A. and Bekers, M. (2001). Stabilization of anti-leukemic enzyme L-asparaginase by immobilization on polysaccharide levan. Journal of Molecular Catalysis B: Enzymatic 11, 551-558.

Voets, J. P., Vandamme, E. J. and Vlerick, C. (1973). L-phenylacetylcarbinol biosynthesis by Saccharomyces cerevisiae. Zeitschrift fur Allgemeine Mikrobiologie 13, 355-365.

Wang, B. (1994). Kinetic study of fed-batch and continuous bioconversion processes for L-Phenylacetylcarbinol (L-PAC) production by the yeast Candida utilis. (PhD Thesis). University of New South Wales: Sydney-Australia

Ward, O. P. and Singh, A. (2000). Enzymatic asymmetric synthesis by decarboxylases. Current Opinion in Biotechnology 11, 520-526.

Weiss, R. F. (1988). Herbal medicine. Gothenburg, Sweden: AB Arcanum and Beaconsfield.

152 References

Weissman, K. (2004). Rational or random? Chemistry World 1, 41-42.

Weusthuis, R. A., Visser, W., Pronk, J. T., Scheffers, W. A. and van Dijken, J. P. (1994). Effects of oxygen limitaion of sugar metabolism in yeast: A continuous-culture study of the Kluyver effect. Microbiology 140, 703-715.

Wiegel, J. (1980). Formation of ethanol by bacteria. A pledge for the use of extreme thermophile anaerobic bacteria in industrial ethanol fermentation processes. Experentia 36, 1434-1446.

Willeman, W. F., Straathof, A. J. J. and Heijnen, J. J. (2002). Reaction temperature optimization procedure for the synthesis of (R)-mandelonitrile by Prunus amygdalus hydroxynitrilase lyase usign a process model approach. Enzyme and Microbial Technology 30, 200-208.

Willke, T. and Vorlop, K.-D. (2004). Industrial bioconversion of renewable resources as an alternative to conventional chemistry. Applied Microbiology and Biotechnology 66, 131-142.

Witholt, B., de Smet, M. J., Kingma, J., van Bailen, J. B., Kok, M. and Lageveen, R. G. (1990). Bioconversion of aliphatic compounds by Pseudomonas oleovorans in multiphase bioreactors; background and economic patented. Trends in Biotechnology 8, 46-52.

Wu, J. C., He, Z., Yao, C. and Yu, K. (2001). Increased activity and stability of Candida rugosa lipase in reverse micelles formed by chemically modified AOT in isooctane. Journal of Chemical Technology and Biotechnology 76, 949-953.

Yamagata, Y., Arakawa, K., Yamaguchi, M., Kobayashi, M. and Ichishima, E. (1994). Functional changes of dextran-modified alkaline proteinase from alkalophilic Bacillus sp. Enzyme and Microbial Technology 16, 99-103.

153 References

Yang, Z., Zacherl, D. and Russell, A. J. (1993). pH dependence of subtilisin dispersed in organic solvents. Journal of American Chemical Society 115, 12251-12257.

Yang, Z., Domach, M., Auger, R., Yang, F. X. and Russell, A. J. (1996). Polyethylene glycol-induced stabilization of subtilisin. Enzyme and Microbial Technology 18, 82-89.

Zeeman, A.-M., Kuyper, M., Pronk, J. T., van Dijken, J. P. and Steensma, H. Y. (2000). Regulation of pyruvate metabolism in chemostat culture of Kluyveromyces lactis CBS2359. Yeast 16, 611-620.

Zeng, X., Chung, A., Haran, M. and Jordan, F. (1991). Direct observation of the kinetic fate of a thiamin diphosphate bound enamine intermediate on brewers' yeast pyruvate decarboxylase. Kinetic and regiospecific consequences of allosteric activation. Journal of the American Chemical Society 113, 5842-5849.

Zeng, X., Farrenkopf, B., Hohmann, S., Dyda, F., Furey, W. and Jordan, F. (1993). Role of cysteines in the activation and inactivation of brewers' yeast pyruvate decarboxylase investigated with a pdc1-pdc6 fusion protein. Biochemistry 32, 2704- 2709.

Zorn, H., Schuler, M. and Berger, R. G. (2003). Pyruvate decarboxylase catalysed formation of terpenoid α-hydroxy ketones. Biocatalysis and Biotransformation 21, 341- 347.

154 Appendices

Appendix A C. utilis fermentation for PDC production

This Appendix shows the bioreactor conditions, cell growth and PDC activity profiles of a yeast C. utilis fermentation involving a pH shift from 6.0 to 3.0 (Chen et al., 2005). Two sets of parallel fermentations (3 L) were carried out to produce sufficient enzyme for this study and both batches achieved similar final specific PDC activities of 379 and 382 U per gram dry cell mass.

a 100 20 DO (%) 80 pH 16 RQ 60 12

40 8 RQ, pH

20 4 Dissolved Oxygen (%)

0 0 0 5 10 15 20 25 30 35 40 Time (h)

b 100 15

80 12 glucose 60 DCM 9 pyruvate 40 ethanol 6 Glucose (g/L) Glucose 20 3 (g/L) others All

0 0 0 5 10 15 20 25 30 35 40 Time (h)

c 400 3.0

2.5 300 2.0

200 1.5

PDC (U/g) 1.0 100 PDC (U/g biomass) PDC (U/mL) 0.5 culture) (U/mL PDC 0 0.0 0 5 10 15 20 25 30 35 40 Time (h)

Figure A. C. utilis PDC production (T = 30oC, 0.1 vvm, 300 rpm) with a decrease of pH from 6.0 to 3.0 at 15 h by adding 5 M H2SO4: (a) bioreactor conditions, (b) cell growth profiles and (c) PDC activity profiles.

155 Appendices

Appendix B Effect of freeze-thawing and ethanol addition on PAC production by whole cell PDC in an aqueous/benzaldehyde emulsion system

Table 4-3 (in Section 4.2.2.1) shows that repeated freeze thawing of whole cells to increase cell permeability and/or release intracellular PDC did not result in any significantly higher PAC concentrations. However addition of 2 M ethanol resulted in a reduction in PAC production (possibly PDC deactivation particularly at higher temperature). The following Figure B shows the detail for initial (1-3 h) PAC formation profiles with whole cells (untreated, freeze-thawed and with the addition of 2M ethanol) in an aqueous/benzaldehyde emulsion system at 4o and 21oC.

30

20 PAC (g/L) PAC 10

0 0123 Time (h)

Figure B. Initial 1-3 h PAC production by whole cell PDC as untreated cells (■,□), freeze and thawed cells (●,○) and with addition of 2 M ethanol into the reaction (▲,∆) in an aqueous/benzaldehyde emulsion system at 4oC (solid lines) and 21oC (broken lines) (initial activity 3 U/mL, 2.5 M MOPS, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 300 mM benzaldehyde, 360 mM pyruvate).

156 Appendices

Appendix C Further studies on the aqueous/octanol-benzaldehyde emulsion system at 21oC

(1) Effect of prolonged biotransformation on PAC concentrations with different PDC preparations

The following Figure C-1 shows that PAC losses which occurred during prolonged biotransformation with whole cell PDC in the aqueous/benzaldehyde emulsion system (see Section 4.2.2.2) were not observed in the aqueous/octanol- benzaldehyde process (from 24 to 48 h).

300 24 h 48 h

200

100 Total PAC (mM) PAC Total

0 Partially purified Crude extract Whole cells

PDC preparations

Figure C-1. Effect of prolonged biotransformation on PAC concentrations in a rapidly stirred aqueous/octanol-benzaldehyde system at 21oC with partially purified, crude extract

and whole cell PDC (initial activity 1 U/mL, 2.5 M MOPS, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol, 24 h and a total reaction volume of 1.8 mL).

157 Appendices

(2) Effect of cell concentrations on by-products (acetaldehyde and acetoin) formation

Following from the effect of cell concentrations on PAC formation in Section 5.2.2.1, Figure C-2 shows that the final (24 h) values for by-products acetaldehyde and acetoin also followed the same trend as the PAC concentrations (higher PAC formation with lower cell concentration).

16 acetaldehyde acetoin 12

8

4 Acetaldehyde, acetoin (mM) 0 23579 Cells concentration (g/L)

Figure C-2. Effect of cell concentrations on by-products (acetaldehyde and acetoin) formation in a rapidly stirred aqueous/octanol-benzaldehyde system (initial activity 1

U/mL, 2.5 M MOPS, initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol, 21oC and a total reaction volume of 15 mL, 24 h).

(3) PAC formation by whole cell PDC in the absence of cofactor TPP

The current study also confirms that cofactor TPP addition is an essential requirement for high PAC production with whole cell PDC. Figure C-3 shows that PAC levels of 84 mM in the octanol and 12 mM in the aqueous phase (7 g/L total reaction volume, 34 h, initial activity of 4.3 U/mL) were obtained with whole cell PDC in the absence of TPP.

158 Appendices

100

80

60 octanol phase 40 aqueous phase PAC (mM)

20

0 0 5 10 15 20 25 30 35 Time (h)

Figure C-3. Time profiles of PAC formation by whole cell PDC in a rapidly stirred aqueous/octanol-benzaldehyde emulsion biotransformation at 21ºC without addition of cofactor TPP (initial activity 4.3 U/mL, 2.5 M MOPS/KOH pH 6.5, 0.5 mM MgSO4, 1.1 M pyruvate, 1.5 M benzaldehyde in octanol and a total reaction volume of 10 mL).

(4) Enzymatic whole cell biotransformation in the absence and presence of 50 mM AOT

Since the addition of 50 mM AOT increased PAC production with partially purified PDC (see Section 6.2.4.2), the effect of AOT on PAC production kinetics was further studied for whole cell PDC. Figure C-4 and C-5 show the kinetics of PAC formation for whole cell PDC at an initial activity of 1 U/mL in the absence and presence of 50 mM AOT respectively. Similar profiles were observed for both conditions and no improvement in PAC concentrations was evident for whole cell PDC with the addition of AOT. The residual activity for whole cell PDC during biotransformation could not be accurately determined due to difficulties in complete PDC recovery from the associated cell debris in the aqueous phase.

159 Appendices

1500 Octanol Phase 100 1200 80 900 60

600 40 All others (mM) 300 20 Benzaldehyde, PAC (mM) 0 0 0 5 10 15 20 25 Time (h)

1500 Aqueous Phase 100 1200 80 900 60

600 40 Pyruvate (mM) Pyruvate All others (mM) 300 20

0 0 0 5 10 15 20 25 Time (h)

pyruvate PAC benzaldehyde acetaldehyde acetoin

Figure C-4. Time profiles of substrates, product and by-products for whole cell PDC in an aqueous/octanol-benzaldehyde system at 21oC without 50 mM AOT addition (initial

conditions: 1 U/mL, 1.3 M pyruvate, 2.5 M MOPS initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.5 M benzaldehyde in octanol phase and a total reaction volume of 10 mL). The mean values were determined from duplicate analyses. The error bars show lowest and highest values

160 Appendices

1500 Octanol Phase 100 1200 80 900 60

600 40 All others (mM) 300 20 Benzaldehyde, PAC (mM) 0 0 0 5 10 15 20 25 Time (h)

1500 Aqueous Phase 100 1200 80 900 60

600 40 Pyruvate (mM) Pyruvate All others (mM) 300 20

0 0 0 5 10 15 20 25 Time (h)

pyruvate PAC benzaldehyde acetaldehyde acetoin

Figure C-5. Time profiles of substrates, product and by-products for whole cell PDC in an aqueous/octanol-benzaldehyde system at 21oC with 50 mM AOT addition (initial

conditions: 1 U/mL, 1.3 M pyruvate, 2.5 M MOPS initial pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.5 M benzaldehyde in octanol phase and a total reaction volume of 10 mL). The mean values were determined from triplicate analyses. The error bars show lowest and highest values.

161 Appendices

Appendix D Dilution factors for whole cell biotransformation with pH-control In the whole cell biotransformation with pH control (see Section 5.2.2.4), pH was maintained at 6.5 with the addition of 5 M acetic acid. A total of 8.8 mL acid was added over 29 h and the dilution factor of each time point is given in the following Table D.

Table D. Dilution factors for each of the sampling time points in a biotransformation with whole cell PDC in a pH-controlled rapidly stirred aqueous/octanol-benzaldehyde emulsion system at 21 ºC (initial activity 2.5 U/mL, 2.5 M

MOPS/KOH pH 6.5, 0.5 mM MgSO4, 0.5 mM TPP, 1.4 M pyruvate, 1.7 M benzaldehyde in octanol and a total reaction volume of 100 mL) (see Figure 5-7).

Time (h) Acid addition (ml) Total dilution factor 0.1 0 1.00 1 0.098 1.00 3 1.064 1.02 5 2.103 1.04 10 4.302 1.09 12 5.424 1.11 15 6.41 1.13 20 7.65 1.15 24 8.084 1.16 29 8.811 1.18

162