Substrate scope of the re-engineered enzyme, FucO D93

(1)

Substrate scope of the

re-engineered enzyme, FucO D93

Martin Friberg

Degree project in chemistry, Bachelor of science, 2015 Department of chemistry

(2)

Abstract

In this report we take a deeper look at the engineered protein FucO D93 and its ability to efficiently catalyze the redox reactions of primary alcohols and aldehydes. FucO is an enzyme native to E. coli where it catalyses the reduction of lacaldehyde to (S )-1,2-propanediol. The D93 variant carries a leucine to valine change at position 259, this mutation was introduced to expand the substrate scope of the enzyme to include bulkier diols. It was proven to work since the enzyme showed activity with 3-phenyl-(S )- 1,2-propanediol. In this study we examine the closely related compound (S )-1-phenyl-1,2-ethanediol but find no activity. This further supports the theory that the crowded active site of FucO is a major factor when it comes to hindering the catalysis of bulkier substrates.

(3)

1 Abbreviations

Amp Ampicillin

FucO Propane-1,2-diol oxidoreductase IPTG Isopropyl β-D-1-thiogalactopyronoside

LA Lysogeny Agar

OD₆₀₀ Optical density at 600 nm

NADH/NAD⁺ Nicotinamide adenine dinucleotide

ON Over Night

SDS-PAGE Sodiumdodecylsulfate polyacrylamide gel electrophoresis

Wt Wild type

(5)

2 Introduction

Environmental pollution is one of mankinds largest current problems. Though it was ignored in the beginning, now many solutions are being proposed. One such is the use of enzymes to make industrial production of chemicals safer and more efficient. There are several properties of enzymes that makes them interesting for use in large scale catalysis. First of all they are biodegradable and easy to produce. Second they work in water solution, outside extreme pH ranges and does not require the use of heavy metals. Last but not least enzymes are very specific, minimizing the need for protecting reactive groups and even distinguishing between enantiomers. This said they seldom catalyze exactly the reactions that are interesting for the industry, thus a need for custom designed enzymes arise. [7]

2.1 Catalysis

To fully grasp the potential of biocatalysis it is necessary to first go through the basics of catalysis. Atkins and De Paulo [1] defines catalyst as follows

”A catalyst is a substance that accelerates a reaction but undergoes no net chemical change”. This is done by lowering the activation energy of a reaction, either by stabilizing intermediates or by changing the reaction pathways to avoid slow rate-limiting steps, see Fig 2.1. [1]

Catalysis is usually divided into two categories, homogeneous catalysis and heterogeneous catalysis, depending on whether the catalyst is in the same phase or not. For example the use of a strong acid to catalyse the formation of an ester from a carboxylic acid and an alcohol would be homogeneous catalysis since the reactants and the acid are all in solution. On the other hand, the use of a metal to catalyze the hydrogenation of an alkene would be heterogeneous catalysis since the alkene is in solution and the catalyst is a solid surface to which the reactants, alkene and hydrogen, is adsorbed. [1]

It is important to note that even though the catalyst accelerates the reaction it does not affect the equilibrium. Thus catalysis can never be used to increase the amount of product obtained but only to decrease the time it takes to obtain that product.

2.2 Kinetics

Kinetics is the study of reaction rates, the speed at which reactions proceed.

Usually defined as a product of the reactants to a power and a rate constant denoted k, Eq 1, the rate constant is specific for that reaction, at that temperature and has to be experimentally determined.

(6)

Figure 1: This figure shows an energy coordination diagram for a hypotheti- cal reaction with and without catalyst. As can be seen neither start nor finish is affected by the catalyst but, the energy barrier in between is lowered.

v = k[A]^a[B]^b (1)

The two most simple cases for this would be the 0th order reactions, where the rate of the reaction is not dependent on any concentration, and the 1st order reactions, where reaction rate is dependent on a single concentration raised to the power of 1. First order reactions are very simple to deal with since they are linear, this is not the case for reactions of higher rates.

Therefore it is common to create a pseudo first order reaction when studying reactions of higher order. By doing measurements with all reactants but one in large excess the contribution of that chemical to the over all rate of the reaction can be determined and if this process is repeated for all reactants the rate law for the entire reaction can be approximated. To ensure that

(7)

depletion of reactants does not affect the reaction rate it is very common to measure only the initial rate but with several different starting concentrations. This also allows the rate to be plotted as a function of concentrations instead of time since start which is very helpful. [1]

That is the basics, but for reactions with several steps things quickly get a bit complicated. To simplify these complex expressions something called the steady-state approximation is used. The steady state approximation assumes that after a short period of time the concentration of intermediates will stay constant at a level where it is produced and consumed at the same rate. This allows for considerable simplification of most reaction sequences. [1]

2.3 Enzymes

Enzymes are catalytically active proteins, they are not the only catalysts within living things but by far the most common. The structural diversity of enzymes allow them to catalyse a myriad of reactions as varied as the fixation of atmospheric nitrogen into ammonia and the decomposition of sugar to yield energy for our own cells. This in combination with the ability of most enzymes to work well in water solutions, without strong acids or bases, makes them powerful tools when looking for environmentally friendly alternatives for mass production of chemicals. [9]

The catalytical part of a protein is called the active site. This is often a pocket or fold in the enzyme where the microenvironment is strictly reg- ulated. The amino acids that form the wall of the active site help to bind the substrate and orient it correctly so that catalysis may take place. These amino acids can also help with the catalysis by ensuring the pH is correct or act as donors or acceptors for bonds or leaving groups. For some reactions this is not enough, the enzyme needs some thing that helps it. This is called a co-factor, some of the most common ones are NAD⁺, NADH and a variation of metal ions. [9]

Diving a bit deeper into the effect of structure, there are two models that explain how the structure of the enzyme effect its catalytical efficiency and selectivity. The first one is the ”lock and key” model, it was introduced in the late 1800’s by Emil Fischer. His model states that if the enzyme is formed so that the substrate fits perfectly then only the substrate will bind and be catalysed. This model was considered good but there is a problem, if the substrate fits perfectly into the enzyme there is nothing to help it transition into the transition state and on to the product. Instead the idea was put forth that the enzyme has some ability to bind the substrate but that the transition state binds better. This would mean stabilization of the intermediate which would lower the activation energy, thus catalysing the reaction. This is called

(8)

”induced fit” and is the leading theory at the moment. Together with the effect of amino acid side chains in the active site the induced fit model forms the current explanation of enzyme catalysis. [9]

The most common method to compare the efficiency of different enzymes is through kinetic parameters, the most basic of which is the Michaelis- Menten constant (K_m). The constant derives from the Michaelis-Menten equation (Eq.2) where V_max is the maximum rate at which the enzyme can work, [S] is substrate concentration, K_m the Michaelis-Menten constant, and v₀ the initial rate.

v₀ = Vmax[S]

K_m+ [S] (2)

What K_m really represent, is the substrate concentration at which the enzyme works at half its theoretical maximum velocity, it is a measurement of how well the enzyme binds the substrate. This is easy to measure and calculate using initial rate experiments at different concentrations and plotting rate as a function of initial concentration to create a saturation curve.

Though K_m is useful it describes only half the problem, for example an enzyme can bind the substrate so tightly that it loses activity. This is called a substrate trap and though showing a very ”good” K_m they will function poorly as catalysts. To avoid this problem one can use k_cat. The turnover number, k_cat, is a measure of how many molecules a single enzyme can catalyse per unit of time. This constant is more intuitive, but also tells only half the story. Unlike K_m it says nothing about how much substrate you need to reach maximum velocity, it can be argued that this is not essential for the industry since reactants can be supplied in large excess it is however very relevant in biological systems or when reactants are expensive. To sum up this duality k_cat/K_m is often used, this ratio will rise if the substrate binds well (low K_m) or the enzyme catalyses the reaction well (high k_cat), thus being an excellent tool to estimate the over all efficiency of an enzyme. [9]

2.4 Enzyme re-engineering

Since enzymes are not well enough understood to be designed from the ground up another technique for making synthetic enzymes is needed. This is usually done by picking an enzyme that catalyses a reaction very similar to the desired reaction and subjecting it to rounds of alternating mutagenesis and selection. Mutagenesis can be either random or semi-rational, meaning you can allow random mutation at any position or allow random mutation at set position. The later is more common since computer methods and structure analysis is getting better, this allows for better predictions of which positions

(9)

will matter and thus the generation of non-functional can be minimized and thus screening time can be shaved of. Screening is the second half of the cycle;

the generated library is tested for desirable properties, most often increased activity for a new substrate, any promising clones are sequenced and are used as parents for the next cycle. This process can then be repeated until the analysed enzyme have achieved the desired properties. [4, 8]

2.5 FucO

Propane-1,2-diol oxidoreductase (FucO), is a native enzyme to E. coli where it is involved in fermentation of unusual sugars such as L-fucose. FucO catalyses the reduction L-lactaldehyde into (S )-1,2-propanediol using an Fe(II) ion and NADH as cofactors, which allows for regeneration of NAD⁺. [2, 3, 6]

Structures that use zinc as a co-factor instead of iron have been published, Fig 2, it is a homodimer where each subunit has a molecular mass of 41kD.

Figure 2: a)The entire FucO protein (generated from PDB file 1RRM), the two monomers are coloured differently. b)Close up of a single monomer with the co-factors shown in blue and orange

2.5.1 D93

D93 is an engineered variant of FucO constructed by [4], it contains a leucin to valine substitution at position 259 (L259V), Fig 3. It was created in an attempt to make FucO catalyse bulkier substrates, as can be seen in the figure below the change from leucin to valine shortens the side chain of the amino acid leaving more space in the active site. [4]

(10)

Figure 3: Left: The active site of the wild-type enzyme (generated from PDB file 2BL4). Right: The active site of D93, computer-modelled from the same structure of wild type enzyme. Position 259 is coloured orange in both pictures.

2.6 Aims

The aims of this study was to further investigate the effect of the L259V mutation on the ability of FucO to accept bulkier substrates. The first substrate to be investigated was (S )-1-phenyl-1,2-ethanediol and further investigations were performed to explain the results of this first test.

(11)

3 Methods

3.1 Chemicals and reagents

The following substrates were used, purity in brackets. 1-Propanol (un- known), propionaldehyde (97%), phenylacetaldehyde (≥ 90%), 2-phenylethanol (≥ 99%), (S )-1,2-propanediol (96%), (S )-1-phenyl-1,2-ethanediol (99%), (R)-1-phenyl-1,2-ethanediol (99%). Substrates were prepared as 2 M stock solutions in buffer (pH 7 phosphate buffer for aldehydes and pH 10 glycine-NaOH buffer for alcohols). Phenylacetaldehyde and 2-phenylethanol were dissolved in acetonitrile instead, since they show lower solubility in water.

NADH and NAD⁺were prepared from powder as 20 mM stock solution in water. New stock was prepared each day since these molecules are unstable if stored for prolonged time.

3.2 Expression

The E. coli strain XL1-Blue was used as host for expression. The protein expression plasmid was pGTac:FucO-5H, it was prepared by Blikstad and Widersten [3], it contains the FucO gene with an IPTG inducible promoter and confers ampicillin resistance. The FucO gene also has C-terminal 5-His tag for purification. Aliquots of 50 µL competent cells were transformed with 2 µL plasmid preparation ( 100 ng/µL) using electroporation, 12.5 kV/cm, 25µF. After electroporation the cells were removed from the electroporation cuvette by suspension in 1 mL 2TY media. This mixture was then transferred to a culture tube before incubation at 37 with shaking (200 rpm) for 1h.

The cells were diluted (10⁰, 10⁻², 10⁻⁴, 10⁻⁶) and 100µL of each concentration were plated on LA-plates with 100 µg/mL Amp before incubation 37 over night. Two single colonies were re-streaked on to either half of a new plate and were left at 37 over night. The plate was then stored in the cold until purification was started.

3.3 Purification

To start of purification a single colony was picked from a re-streaked plate and inoculated into 2 mL 2TY media, grown at 37 with 200 rpm shaking overnight. From this ON-culture, 350µL was transferred into an E-flask with 35 mL fresh 2TY + 100 µg/mL Amp. This new subculture was incubated at 30 with 200 rpm shaking over night.

(12)

To each of six 2-L E-flask was added the following: 0.5 L 2TY + 50 µg/mL Amp, 5 mL ON-culture, and FeCl2 to a final concentration of 100µM.

These cultures were grown at 30 with 200 rpm shaking until OD600 = 0.4, one E-flask was measured approximately every 30 min. When OD₆₀₀ = 0.4 was reached, protein expression was induced by addition of IPTG to a final concentration of 1 mM. The culture was left at 30 with 200 rpm shaking overnight.

Cells were harvested by centrifugation at 4 , 3,825 g in a J-14 rotor, for 15 min. The supernatant was discarded and the cells re-suspended in just below 30 ml binding buffer (20 mM imidazole, 0.5 M NaCl, 20 mM sodium-phosphate, pH 7.5) with added protease inhibitor cocktail (Roche).

The cells were then lysed using ultrasound (5·20 s pulses with 30 s pause in between each pulse to avoid overheating). The cell lysate was transferred into centrifuge tubes and cell debris was pelleted at 4, 30,966 g for 30 min in a JA20 rotor.

During centrifugation a nickel-IMAC column was prepared. Chelating Sepharose was added to 2-3 cm in an empty PD-10 column and was washed with at least 10 volumes of water. After this the beads were charged by addition of 0.1 M NiCl₂. This was repeated until flow-through kept its greenish colour. The column was then washed with another 10 volumes of water.

After centrifugation the supernatant was collected and filtered through two Kleenexr tissues into a 50 ml conical tube. The charged beads were removed from the PD-10 column and were added to the same tube. This mix was left for 45-60 in 4 , with mixing by inversion. The gel-sample mix was then poured back onto the column and the flow-through was collected into a clean conical tube. After this point, all eluent was collected manually as 1-1.5 mL fractions in Eppendorf tubes. Protein concentration was checked using absorbance at 280 nm in a spectrophotometer. The column was washed with binding buffer until no protein absorption was detected. The column was then washed using wash buffer (60 mM imidazole, 0.5 M NaCl, 20 mM sodium-phosphate, pH 7.5) until no protein was detected. Bound protein was eluted in 6 fractions using elution buffer (300 mM imidazole, 0.5 M NaCl, 20 mM sodium-phosphate, pH 7.5) and the protein concentration was measured for all fractions.

The elution fractions with the highest measured concentrations were pooled to a total volume of 2.5 ml, this was called the Peak elution. The highest remaining fractions were also pooled to a total volume of 2.5 and called Side elution. These two fractions were desalted (one at a time) using a PD-10 column and 0.1 M phosphate buffer (pH = 7.4). For every wash step the fraction with the highest absorbance was saved together with a sample from

(13)

the original flowthrough and the two elution samples.

SDS-PAGE was used to assess the purity of the elution fractions and to check other fractions for loss of product. First a separation gel was cast (12

% (w/v) acrylamide, pH 8.8). It was left to polymerize for 30 min with a layer of water on top to create an even edge and prevent drying. On top of this gel was cast a stacking gel (4 % /w/v) acrylamide, pH 6.5) with a 10-well comb inserted into it. This was also left to set for 30 minutes. In the meantime the samples and a protein size ladder were diluted in H₂O to a final volume of 10 µL and suitable concentrations. The samples were mixed 1:1 with loading dye and was heated at 95 for 5 min. After polymerization and removal of the comb the wells were washed carefully with H₂O. The gel was then put into the electrophoresis machine, buffer was added, and the samples loaded. The gel was run at 200 V for 40 min.

The gel was put in staining solution (PhastaGelr Blue R) and mi- crowaved until boiling, left with gentle stirring for 10 minutes. The staining solution was removed and the procedure repeated thrice with destaining solution.

3.4 Kinetic measurements

Kinetic measurements were done for both the reduction of aldehyde and the oxidation of alcohols, both reactions were measured using the change of NADH at 340 nm in a spectrophotometer (UV-1700 PharmaSpec, Shi- madzu). For the oxidation reaction 20 µL NAD⁺ (20 mM), 20 µL enzyme solution, and substrate (1-20 µL) was added to a final volume of 1 ml in pH 10 glycine-NaOH buffer in a 1 cm quartz cuvette. For the reduction reaction, 20µL NADH (20 mM), 20 µL enzyme solution, and substrate (1-20 µL) was added to a final volume of 1 ml in pH 7 phosphate buffer in a 0.5 cm quartz cuvette. Measurements were done over 1-2 min and in triplicates. Raw data was obtained as ∆Abs/min and can be found in Appendix 1. The raw data was transformed to s⁻¹ by division with the extinction coefficient for NADH ( = 6.22mM⁻¹cm⁻¹), 60 (converting minutes to seconds), and the enzyme concentration (c = 20µM, described below) to give Vmax = k_cat. V_max, K_M, and kcat/KM were calculated using SimFit programs mmfit and rffit.

Enzyme concentration was measured using a spectrum (400 nm-200 nm), Protein absorbtion was calculated from the peak at 280 nm after being corrected for noise using the absorbance at 400 nm. This corrected value was divided by the extinction coefficient of FucO ( = 41mM⁻¹cm⁻¹) to give the concentration of enzyme.

(14)

3.5 Figure generation and software

All figures in this report are generated by the author. Imaging was done using paint.NET by dotPDN LLC, protein images were generated using PyMOL 0.99rc6 (DeLano Scientific, LLC) and original files were imported using the PDB loading service plug in. Molecular formulas were drawn using Marvin sketch 15.5.18 (ChemAxon), saturation curves were done using simfit pro- gram simplot, Figure 4 is a photocopy of the SDS-PAGE gel and last of all the entire report was written using L^ATEX.

4 Results

First the enzyme was purified using a nickel column. Two samples were created, the first one by pooling the most concentrated elution fractions, the second sample using the most concentrated fractions that remained after the first pool. The first sample was named ”Peak elution” and the second ”Side elution”, both samples were desalted and the concentration was checked at 280 nm using a spectrophotometer, Table 1.

Table 1: Concentration of pooled elution fraction Sample Concentration (mM)

Peak elution 0.3122 Side elution 0.0317

The purity was checked using SDS-PAGE, Fig 4. Both samples were deemed pure and concentrated enough to use. The ”Side elution” was used for all measurements since the more concentrated ”Peak elution” could be sent for crystallization studies.

To start of the kinetic studies, the activity with 1-propanol was measured to see if the enzyme was active and comparable to previous studies, Table 2. After this the substrates in Fig 5 were measured in turn. The choice of alcohols are motivated in the discussion.

(15)

Figure 4: SDS-PAGE to test purity. The samples are 1) Flow-through 2) 1st Fraction, 1st wash 3) 27th Fraction, 1st wash 4) 4th Fraction, 2nd wash L) Protein size ladder 5) Side elution 6) Peak elution

Table 2: Kinetic properties of FucO for all measured substrates

Substrate k_cat K_M k_cat/K_M ref. k_cat/K_M

(s⁻¹) (mM ) (s⁻¹M⁻¹) (s⁻¹M⁻¹) Alcohols

1-Propanol 0.58(3.5 · 10⁻²) 7.7(2.0) 76(17) 110 [5]

(S)-1,2-Propanediol 0.63(8.5 · 10⁻²) 32(14) 27(8.5) 77 [5]

2-Phenylethanol 2.4 · 10⁻³(4.9 · 10⁻⁴) 11(4.5) 0.23(0.050) This work

(S)-1-Phenyl-1,2- NDA NDA NDA This work

Propanediol

(R)-1-Phenyl-1,2- NDA NDA NDA This work

Propanediol Aldehydes

Propionaldehyde 9.0(0.47) 5.6(1.1) 1700(270) This work

Phenylacetaldehyde SNR SNR 130(8.5) 13 [5]

After each value is given the Standard Error in paren- thesis, NDA = No Detected Activity, SNR = Saturation Not Reached

(16)

Figure 5: Tested substrates and their structural similarity. All substrates are essentially mono- or 1,2-di-substituted ethanol or acetaldehyde derivatives.

At first no activity could be detected for 2-phenylethanol but a second try was done using concentrated peak fraction and 5 min measuring time. Using this new approach a saturation curve was done and all kinetic parameters calculated.

In a continued effort to visualize the results of this study a thermodynamic cycle was created by first calculating ∆∆G, using Eq. 3, for three of the derivatives as compared to 1-propanol. And then plotting the information in Fig. 6.

∆∆G = −RT ln(k_cat/K_m)_1−propanol

(k_cat/K_m)_derivative (3)

(17)

Figure 6: Thermodynamic cycle for 1-phenyl-S-2,3-propanediol. Numbers given are ∆∆G as compared to 1-propanol.

5 Discussion

In the SDS-PAGE you can clearly see a band corresponding to FucO at around 41 kD for both the peak and tail elution. There is also a rather large band at the same height for the second wash (sample 4, Fig4). This seem to indicate that the His tag does not bind as strongly to the matrix as expected and that the protein is partly washed away at 60 mM imidazole.

It was already know that the protein binds poorer than usual which is the reasoning behind using 60 mM imidazole instead of the usual 100 mM for washing.

After the first trial using 1-propanol it was found that the activity of the purified FucO was comparable to previously published data, Table 2.

Next were the 1-phenyl-1,2-ethanediols, these were the primary interest for this study. They are interesting because the enzyme shows activity with 1- phenyl-S-2,3-propanediol but is believed to suffer from structural restraints in the active site. Since the phenyl ring of 1-phenyl-1,2-ethanediol can’t rotate freely around the chiral coal it is believed to be bulkier and harder to catalyse than the phenylpropanediol. This theory was further supported since no activity was detected for either phenyldiol. After this discovery S- 1,2-propanediol was tested, this is the same alcohol with a methyl substituent instead of the phenyl ring, activity was observed which suggests that this reduction of size does indeed allow the substrate to catalyzed.

To further investigate the interplay of size and free rotation in 1-phenyl- 1,2-ethanediol catalysis, a close relative was investigated. 2-phenylethanol

(18)

lacks the secondary alcohol and thus the chiral center, this should allow the phenyl ring to move more freely in the active site. Since it is know that the corresponding aldehyde (Phenylaceticaldehyde) is catalyzed one would expect the reverse reaction to also happen. This proved to be the case, to visualize the interplay of the Phenyl ring and the secondary alcohol a thermodynamic cycle was created (Fig 6). This figure tells us that the phenyl ring causes an increase in ∆G that is more than twice the increase from adding the secondary hydroxyl group, thus it seems that the phenyl ring is the primary cause for the loss of activity.

It is however worth noting that 3-phenyl-(S )-1,2-propanediol is a substrate to the enzyme and this molecule carries a phenyl ring at the end just like 1-phenyl-1,2-ethanediol. Why this is can not be explained by this study, it is possible that the extra coal of the propane chain allows the molecule to be arranged to fit into the active site. It would have been interesting to search for activity with the aldehyde corresponding to (S )-1-phenyl-1,2-ethanediol sice this activity is usually higher. However the chemical could not be found at any company and thus could not be measured.

To investigate this deeper one would need to go on to computer modelling, however finding out why something does not work is often more complicated than finding out how it works. Docking studies might be used to get an understanding of how the molecule fits, or rather does not fit in the active site, one must however be somewhat careful since these methods are still a bit primitive.

To sum up, a deeper investigation into the substrate scope of FucO D93 was conducted within this study. It was found that although activity for 3-phenyl-(S )-1,2-propanediol was confirmed in a previous study [5], (S )-1- phenyl-1,2-ethanediol is not catalysed by the enzyme. The exact explanation was not found but it is probable that it is due to the steric phenyl group not fitting in the crowded active site of FucO.

(19)

References

[1] Peter Atkins and Julio De Paulo.

[2] L Baldom and J Aguliar. Metabolism of l-fucose and l-rhamnose in Ex- herichia coli : Aerobic-anaerobic regulation of l-lactaldehyde dissimila- tion. Journal of Bacteriology, 1988.

[3] Cecilia Blikstad and Mickael Widersten. Functional characterization of stereospecific diol dehydrogenasse, fuco, from Escherichia coli : Substrate specificity, ph dependence, kinetic isotope effects and influence of solvent viscosity. Journal of Molecular Catalysis B: Enzymatic, 2010.

[4] Cecilia Blikstad, Kathe M. Dahlstrom, Tiina A. Salminen, and Mikael Widersten. Stereoselective oxidation of aryl-substituted vicinal diols into chiral α-hydroxy aldehydes by re-engineered propanediol oxidoreductase.

ACS Catalysis, 2013.

[5] Cecilia Blikstad, Kthe M. Dahlstrm, Tina A. Salminen, and Widersten Widersten. Effects on substrate scope and selectivity in offspring to an enzyme subjected to directed evolution. FEBS Journal, 2013.

[6] G. T. Cocks, J. Aguilar, and E. C. C. Lin. Evolution of l-1,2-propanediol catabolism in Escherichia coli by recruitment of enzymes for l-fucose and l-lactate metabolism. Journal of bacteriology, 1974.

[7] Maria Gavrilescu and Yusuf Chisti. Biotechnology-a sustainable alterna- tive for chemical industry. Biotechnology Advances, 2005.

[8] Moore JC and Arnold FH. Directed evolution of a para-nitrobenzyl es- terase for aqueous-organic solvents. Nature biotechnology, 1996.

[9] David L. Nelson and Michael M. Cox.

(20)

Appendix

1-Propanol

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70

0.0 20.0 40.0 60.0 80.0 100.0 120.0

Saturation curve 1-propanol

[S] (mM)

v (1/s)

[S](mM) 1/s

2 0.048826962

2 0.1037572943

2 0.0793438133

5 0.1800494224

5 0.1739460522

5 0.1281707753

10 0.3448404192 10 0.3845123258 10 0.3723055853 20 0.5737168036 20 0.4699595093 20 0.4180808622 40 0.4913213052 40 0.5004763606 40 0.5004763606 60 0.6408538764 60 0.6622156723

(21)

60 0.5157347862 80 0.3845123258 80 0.4241842325 80 0.4760628796 120 0.460804454 120 0.4760628796 120 0.6011819698

(22)

(S)-1,2-Propanediol

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70

0.0 50.0 100.0 150.0 200.0

Saturation curve (S)-1,2-propanediol

[S] (mM)

v (1/s)

[S](mM) 1/s

1 0.0305168513

1 0.0213617959

5 0.1037572943

5 0.0671370728

5 0.1007056091

20 0.1281707753 20 0.1678426819 20 0.1800494224 40 0.3600988448 40 0.2685482911 40 0.2777033464 60 0.5004763606 60 0.3906156961 60 0.5493033226 80 0.5737168036 80 0.6347505061 80 0.5462516375 120 0.5737168036

(23)

120 0.631698821 120 0.4943729904 160 0.3845123258 160 0.36315053 160 0.3570471597

(24)

2-Phenylethanol

0.0 0.5 1.0 1.5 2.0

×10^-3

0.0 5.0 10.0 15.0 20.0

Saturation curve 2-phenylethanol

[S] (mM)

v (1/s)

[S](mM) 1/s

2 0.000329582

2 0.0003845123

2 0.000494373

4 0.0006042337

4 0.0003845123

4 0.000494373

7.5 0.0009338156 7.5 0.0007140943 7.5 0.0012084673 10 0.0015380493

10 0.001153537

10 0.0015380493 20 0.0017028403 20 0.0014281886

20 0.001318328

(25)

(S )1-Phenyl-1,2-propanediol

[S](mM) δAbs/min

40 0.0007

40 0.0001

40 0.0019

40 0.0029

10 0.0011

10 0.0001

10 0.0012

10 0.0011

2 0.0020

2 0.0001

2 0.0014

(R)1-Phenyl-1,2-propanediol

[S](mM) δAbs/min

4 0.0017

4 0.0005

20 0.0004

20 0.0084

20 0.0004

20 0.0003

40 0.0004

40 0.0002

40 0.0008

40 0.0016

(26)

Propionaldehyde

0.0 2.0 4.0 6.0 8.0 10.0

0.0 10.0 20.0 30.0 40.0 50.0 60.0

Saturation curve propionaldehyde

[S] (mM)

v (1/s)

[S](mM) 1/s

1 1.5014290818

1 1.2450875313

1 1.3183279743

2.5 2.4596582113 2.5 2.4840716923 2.5 2.404727879

5 4.0770513279

5 4.1258782899

5 3.9793974038

20 8.2517565797 20 7.4278015958 20 8.1541026557 40 7.2996308205 40 8.8804037156 40 8.7095093486

60 7.214183637

60 6.5489162796

(27)

Phenylacetaldehyde

[S](mM) 1/s

0.4 0.1892044778 0.4 0.1892044778 0.4 0.1098606645

2 0.3356853638

2 0.2441348101

2 0.2807550316

4 0.3295819936

4 0.4760628796

4 0.4394426581

6 0.5615100631

6 0.7995415029

6 0.7568179112

10 1.5746695248

Substrate scope of the re-engineered enzyme, FucO D93