Time-optimization of a chromatographic method for monitoring a multistep enzymatic conversion of epoxides to 1,2-diols

MAHA NAIM Time-optimization of a chromatographic method for monitoring a multistep enzymatic conversion of epoxides to 1,2-diols Thesis Bachelor of Science in Chemistry March 2014 Bachelor Degree project in Chemistry (15 hp) Department of Chemistry BMC Supervisor: Mikael Widersten and Åsa Janfalk Carlsson

Abstract The enzymatic conversion of selected epoxides into diols and ketones was monitored by time optimized reversed phase chromatography. Reaction samples were first analyzed isocratically to identify the components of different peaks. In the next step the samples were run with a gradient elution to shorten the retention time between the compounds in the multistep reaction with retained separation of a good quality. The use of a 100-2 mm reversed phase C-18 column with an optimized water and acetonitrile gradient allowed an analysis time of the compounds in the multistep reaction could be optimized to 20 min, with good resolution. This result represents an improvement compared to earlier protocols that required 45 min and 74 min respectively. Svensk sammanfattning Den enzymatiska omvandlingen av utvalda epoxider till dioler och ketoner övervakades med en tidsoptimerad reversed phase kromatografi. Reaktionsprover analyserades först sokratiskt för att identifiera föreningarna med olika toppar. I nästa steg analyserades proverna med en gradienteluering för att kunna förkorta analystiden mellan föreningarna i multistegsreaktionen med bibehållen god separation Användningen av en 100-2 mm reversed phase C-18-kolonn med optimerad vatten- och acetonitrilgradient tillät att analystiden av föreningarna i multistegsreaktionen kunde optimeras till 20 min med en god upplösning. Detta resultat är en förbättring jämfört med tidigare protokoll som krävde 45 respektive 74 min. 2

List of abbreviations AdhA Rhodococcus ruber alcohol dehydrogenase A Diol 1 Diol 2 2,3-EPB HPLC Ketone 1 ketone 2 NAD + PDA SO StEH1 1-phenyl-1,2-ethanediol 3-phenyl-1,2-propanediol 2,3-epoxyprobylbenzene high performance liquid chromatography 2-hydroxyacetophenone 1-hydroxy-3-phenylpropan-2-one nicotinamide adenine dinucleotide photodiode array detector styrene oxide Solanum tuberosum epoxide hydrolase 1a 3

Table of contents Abstract...2 Svensk sammanfattning...2 List of abbreviations...3 Table of contents...4 Introduction...5 Aim of the project...5 Materials and Methods...7 Equipment...7 Chemicals and materials...7 Sample preparation...7 Gradient experiment...8 The substrates...9 Results... 11 The isocratic separation of SO, 2,3-EPB and the reaction products of SO... 11 Shortening of the analysis time by gradient elution... 12 Discussion... 17 Conclusion... 18 Personal reflection... 18 Acknowledgements... 18 References... 19 Appendix... 20 4

Introduction Aim of the project The main aim of the project was to develop a time optimized analysis of the compounds in the multistep reaction (Figure 1). The research project was first run by Professor Mikael Widersten and his coworkers to study enzymatic conversion of epoxides, using a multistep enzyme approach. As a starting point, selected epoxides are hydrolyzed by the enzyme epoxide hydrolase 1a from Solanum tuberosum (StEH1) to form the corresponding diols. These diols are further converted into ketones by alcohol dehydrogenase A from Rhodococcus ruber (AdhA), (Figure 1). StEH1 H 2O AdhA NAD + Figure 1- Styrene oxide is hydrolyzed by the enzyme StEH1 to form a diol. The diol is then converted into a ketone by alcohol dehydrogenase AdhA. The epoxides used here are styrene oxide (SO) and 2,3-epoxypropylbenzene (2,3-EPB). The hydrolysis of 2,3-EPB cannot be measured by spectrophotometer due to similar absorbance spectra for the epoxide and the diol. As a Consequence HPLC was used to monitor the reaction shown in figure 1. In early protocol based on reversed phase chromatography each sample required 80 minutes. Thus, it is desirable to reduce the analysis time. The actual separation principle used here was a reversed phase chromatography that employs a nonpolar stationary phase, in this case a short C-18 column was used. The C-18 name refers to a stationary phase based on microporous silica particles that are derivatized with 18- carbon hydrocarbon chains. The requirement was to maintain a sufficiently good resolution, which 5

means not only to obtain as short analysis time as possible, but also to get a good separation between the compounds. The packed column efficiency increases as the particles size of the stationary phase decreases. The experimental factors that could influence the resolution were the flow rate, gradient elution and the column height. The resolution is often described by the plate number (N) which corresponds to the number of discrete distributions that would give the same results. In isocratic elution, the plate number can be estimated from the peak width (W) and the elution volume (Ve) as N = 16 (Ve/W) 2. The height of the theoretical plate is L/N, where the N increases linearly with L, meaning that the resolution increases with L. The Van Deemter equation (eq. 1) tells how the flow rate and the column affects the plate height (H). The term A stands for longitudinal diffusion, the term B for the equilibrium between the stationary phase and the mobile phase, and the term C is a constant and due to the properties of the column packing. H = A v + B v + C eq. 1 In packed columns the terms A, B and C all contribute to band broadening. A decrease of the particle size will normally influence both the B and C terms thus decreasing the plate height (H), resulting in sharper peaks. [1]. An isocratic elution was used in the experiment to identify the peak of each compound. Isocratic elution represent the case where the mobile phase composition is unchanged during the elution process. If one solvent mixture could not give a rapid elution of all components a gradient elution is often used (Figure 2). Gradient elution provides continuous change of the composition of the solvent to increase the eluent strength. We designed a stepwise linear gradient with two different slopes, one until the ketone was eluted and one to elute the epoxide. Short analysis time can be obtained in many ways, for example by using the shortest possible nonpolar C-18-column, changing the flow rate and the composition of the mobile phase, and by using a gradient elution. 6

Materials and Methods Equipment High performance liquid chromatography (HPLC) was carried out using a Chromolith performance RP-18 e column with dimensions of 100x2 mm, using H2O/acetonitrile adjusted to ph3.0 with formic acid is the mobile phase. The system also consisted of a sampler (Shimadzu SIL-10AF), pumps (LC 20AD) and detector (Shimadzu SPDM20A photometric unit). Our experiment was carried out at a flow rate of 0.38 ml/min [2]. The detector was a photodiode array (PDA). Chemicals and materials Styrene oxide, 1-Phenyl-1-2-ethanediol and 2-hydroxyacetophenone were obtained from Sigma Aldrich. 2,3-epoxyprobylbenzene was obtained from TCI. The epoxide hydrolase StEH1 was expressed in E.coli transformed by the plasmid pgtacsteh1-5h, that is a pgtacplasmid with a gene for StEH1 that was extended by 5 histidines by the researchers. Alcohol dehydrogenase AdhA extended by 5 histidines was expressed in E.coli strain BL21-A1. All other chemicals were of analytical grades. Sample preparation Reference sample The compounds styrene oxide, 1-Phenyl-1-2-ethanediol (diol 1) and 2-hydroxyacetophenone (ketone 2) were dissolved in H2O to 3 mm concentration. The incubation samples were prepared as described in appendix 1, Table A1. The compound 2,3-epoxyprobylbenzene (2,3-EPB) was first mixed with only the enzyme StEH1 to obtain the corresponding 3-phenyl-1,2-propanediol (diol 2), to identify the peak of the diol 2. Then 2,3-EPB was mixed with both the enzymes StEH1 and AdhA to obtain the corresponding 3-phenyl-1,2-propanediol (diol 2) and 1-Hydroxy-3-phenylpropan-2-one (ketone 2). The compound 2,3-EPB and the enzymes were dissolved in 0.1M sodium phosphate ph 7.9 to 30 mm concentration. The incubation samples were prepared as described in appendix 2, Table A5. 7

Gradient experiment Figure 2- An example of a gradient employed. Gradient elution (dark blue) with acetonitrile (%) plotted over time (min). The gradient was increased from 15 % acetonitrile to 30 % acetonitrile between 5 min and 7 min. And then increased to 90 % acetonitrile between 7 min and 20 min and maintained at 90% acetonitrile for 5 min. Between the time 25:00 and 25:01 the gradient decreased from 90% to 15% to end the experiment. The same general method was used to all the experiments but with different gradients (Table 3). 8

The substrates 1- Separation of reaction products with styrene oxide as a starting substrate StEH1 SO H2O 1-Phenyl-1-2-ethanediol (diol 1) AdhA NAD + 2-hydroxyacetophenone (ketone 1) Figure 3- A reaction scheme with SO as starting substrate. SO, diol 1 and ketone 1 were bought from Sigma Aldrich, the compounds were mixed together at 3 mm each in H2O. The sample then run in HPLC and a gradient was used to obtain a shorter analysis time with maintained separation (Table 3 and 4). The experiments in Table 3 were run in duplicate to document the reproducibility. 9

2- Separation of reaction products with 2,3-epoxypropylbenzene as a starting substrate StEH1 2,3-EPB H2O 3-phenyl-1,2-propanediol (diol 2) AdhA NAD+ 1-Hydroxy-3-phenylpropan-2-one (ketone 2) Figure 4- A reaction scheme with 2,3-EPB as starting substrate. 2,3-EPB was bought from TCI thus could be analyzed directly whereas diol 2 and ketone 2 were made by the enzymatic reaction in Figure 4. The peaks to the diol 2 and the ketone 2 in this reaction were identified by preparing two samples, one containing only 2,3-EPB (30 mm) and StEH1 to form diol 2 [3] and one containing epoxide 2 (3 mm), StEH1; and AdhA to form the ketone 2 (Table A5). The experiment 3 in table 7 was run in duplicate. 10

Results The isocratic separation of SO, 2,3-EPB and the reaction products of SO In order to develop a time optimized analysis of the compounds SO, 2,3-EPB, diol 1 and ketone 1 in the multistep reaction, their analysis times in the unoptimized system were needed to be known. The molecular weight (Mw), concentration, and volume of the compounds are given in Table A1, Appendix 1. The samples were run isocratically with 15% acetonitrile / 85% H2O at a flow rate of 0.38ml/min (Table 2). Table 2- The compounds and their retention time Compounds Retention time (min) SO 23.3 2,3-EPB ~30 Diol 1 4.9 Ketone 1 6.5 When the epoxides where run with the longer HPLC C18-column with dimensions of 25 cm x 4.6 mm, 5 µm, the SO had analysis time of 45 min and 2,3-EPB, 75 min, thus by changing the column, the analysis times were decreased. 11

Shortening of the analysis time by gradient elution Separation of the substrate styrene oxide and its products A mixture of compounds was prepared as follows: 0.36 µl SO (3 mm) was mixed with 30 µl diol 1 (3 mm) and 30µl ketone 1 (3 mm) in water to a final volume of 1 ml. The sample was then analyzed by a gradient elution (Figure 2), (Table 3 and 4). Table 3- The conditions that were used during isocratic elution and gradient elution. Experiment 1 (isocratic elution) 2 (gradient elution) 3 (gradient elution) 4 (gradient elution) Conditions 15% acetonitrile / 85% H 2O At time 15:00 15:01 min the gradient increased from 15% to 30% acetonitrile. 15:01 20:00 min the gradient increased to 80% acetonitrile. 20:00-20:01 min the gradient decreased to 0% acetonitrile. At time 5:00 7:00 min the gradient increased from 15% to 30% acetonitrile. 7:00 20:00 min the gradient increased to 90% acetonitrile and maintained at it for more 5 min. 25:00-25:01 min the gradient decreased to 15% acetonitrile. At time 00:00 15:00 min the gradient increased from 0% to 90% acetonitrile. 15:00 20:00 min the gradient was decreased to 15% acetonitrile. 20:00-20:01 min the gradient increased to 90% acetonitrile. 20:01-20:03 min the gradient decreased to 15% acetonitrile Table 4- The retention time of the compounds with isocratic elution and gradient elution. Experiment Flow rate (ml/min) Ret. time of diol 1 (min) Ret. time of ketone 1 (min) Ret. time of SO (min) Graph 1 0.38 4.9 6.5 23.3-2 0.40 5.4 6.8 19.2 Figure 5 3 0.38 4.9 6.2 9.7 Figure 6 4 0.40 7.3 8.2 10.4 Figure 7 12

The traces of experiments 2, 3 and 4 with absorption at 220nm (mau) plotted versus time (min) are shown in Figures 5, 6 and 7. 3000 2500 2000 1500 1000 500 0 m AU Extract-220nm4nm (1.00) 5.427/16488165 6.832/17277406 19.227/127824327-500 0.0 5.0 10.0 15.0 20.0 25.0 m in Figure 5- Retention times of, from left to right, diol 1, ketone 1 and SO in experiment 2. 2000 1500 m AU Extract-220nm4nm (1.00) 9.696/71354156 1000 500 4.948/20044252 6.251/19098091 0 0.0 5.0 10.0 15.0 20.0 25.0 m in Figure 6- Retention times of diol 1, ketone 1 and SO in experiment 3. 13

1000 m AU Extract-220nm4nm (1.00) 750 500 7.294/22526899 8.239/16808703 10.370/51885187 250 0-250 0.0 5.0 10.0 15.0 20.0 25.0 m in Figure 7- Retention times of diol 1, ketone 1 and SO in experiment 4. The analysis time was decreased from 45 min to 20 min by using a shorter C-18 column for SO, changing the flow rate and using an optimized gradient. Separation of 2,3-epoxypropylbenzene and its products The diol and the ketone formed from 2,3-EPB, Figure 3, are not commercially available, but could be obtained from the reaction shown in Figure 1. To obtain the ketone, 1-hydroxy-3-phenylpropan-2-one (ketone 2), 2,3-EPB has to be incubated with epoxide hydrolase (StEH1) to obtain phenyl-1.2-propanediol (diol 2). Thereafter, the diol is converted by alcohol dehydrogenase (AdhA) to give ketone 2. Zink is needed to stabilize AdhA and for the enzyme to function. The co-factor NAD + is also needed. To identify the peaks that correspond to the diol and the ketone in this reaction, two samples were prepared (Table A5). Sample 1 contained only 2,3-EPB (30 mm) and StEH1 to obtain diol 2, and sample 2 contained 2,3-EPB (3 mm), StEH1, and AdhA to obtain ketone 2. The retention time of epoxide 2 is approximately 33 min (Table 6). Samples 1 and 2 were first analyzed with isocratic elution, each sample was run for 100 min with 15% acetonitrile/85% H2O (Table 6). Table 6- The retention time of the substrates with isocratic elution. Compounds Retention time (min) 2,3-EPB 32.9 Diol 2 6.7 Ketone 2 - Ketone 2 was not obtained because only a very small amount of it was formed, that did not give a clearly visible peak. Since no ketone 2 peak was obtained, a gradient optimized for 14

separation of the diol 2 and epoxide 2 only. The mixture was prepared right before the experiment was run to prevent degradation of 2,3-EPB (Tables 7 and 8). Table 7- Conditions used during isocratic and gradient elution. Experiment 1 (isocratic elution) 2 (gradient elution) 3 (gradient elution) Conditions 15% acetonitrile / 85% H 2O At time 5:00 7:00 min the gradient increased from 15% to 30% acetonitrile. 7:00 10:00 min the gradient increased to 80% acetonitrile. 20:00-30:00 min the gradient was increased to 90% acetonitrile. 30:00-30:03 min the gradient was decreased to 15 % acetonitrile. At time 5:00 7:00 min the gradient increased from 15% to 30% acetonitrile. 7:00 10:00 min the gradient increased to 80% acetonitrile. 25:00-30:00 min the gradient was increased to 90% acetonitrile. 30:00-30:03 min the gradient was decreased to 15 % acetonitrile. Table 8- Retention times of compounds with isocratic and gradient elution. Experiment Flow rate (ml/min) Ret. time of diol 2 (min) Ret. time of 2,3-EPB (min) Graph 1 0.38 6.7 33.9-2 0.38 6.2 11.8 Figure 8 3 0.38 6.3 11.9 Figure 9 Good separation between the diol 2 and 2,3-EPB was obtained, and good reproducibility was also obtained. 15

Chromatograms of experiment 2 and 3, where peaks are detected at 225 nm (figure 8 and 9). 500 m AU Extract-225nm4nm (1.00) 400 300 11.893/13082669 200 100 0 6.297/4308599-100 0.0 5.0 10.0 15.0 20.0 25.0 m in Figure 8- Retention times of, from left to right, diol 2 and 2,3-EPB in experiment 2. 300 m AU Extract-225nm4nm (1.00) 11.935/8659210 200 100 6.327/5522478 0-100 0.0 5.0 10.0 15.0 20.0 25.0 m in Figure 9- Retention times of diol 2 and 2,3-EPB in experiment 3. 2,3-EPB has absorption maximum of 225 nm where also the diol 2 appears clearly. As can be seen in Figure 8, there is a peak that coelutes with the epoxide peak. It was identified as a degradation product that was formed in the sample. The analysis time was decreased from 75 min to 20 min by using the shorter C-18 column for 2,3-EPB, using gradient elution and could not be decreased more without sacrificing the resolution. 16

Discussion The HPLC was chosen because many compounds are not volatile enough for gas chromatography. Since the compounds that we wanted to elute are nonpolar, a nonpolar stationary phase was used with polar mobile phase and a gradient with less polar compound [1]. The mobile phase used here consisted of MilliQ-H2O and acetonitrile adjusted to ph 3 with formic acid. Acetonitrile dissolve both ionic and nonpolar compounds and has a high eluent strength that elutes all compounds rapidly. Acetonitrile was also added to the samples due to solubility problems [4]. The reason why a shorter non-polar C-18 column was used is to obtain good separation at shorter time [5]. The gradient was used to elute tight bound compounds, in this case epoxides, faster. Epoxides are nonpolar compounds why they bound tightly to the nonpolar column. When the acetonitrile concentration is increased, it elutes the epoxides faster because it has a high eluent strength. The diols and the ketones are weakly bound to the column because they are polar, more polar than the epoxides, why they elute faster through the column. 150µl was suitable for a good detectability [6]. The starting substrate styrene oxide By using a shorter C-18 column, changing the flow rate and the gradient, the analysis time of the important compounds could be decreased from 45 min, when a longer C-18 column was used, to 20 min. When the flow rate was increased, sharper peaks with retained separation between the compounds were obtained (Figures 5, 6 and 7). The gradient allowed a good separation for diol 1 and ketone 1 together with a shorter analysis time for SO. The flow rate could not be increased higher than 0.40 ml/min since the shorter C-18 column cannot tolerate higher pressure than 180 bar. The results were as expected, by using higher gradient, as high as 80% or 90% acetonitrile with flow rate of 0.38 ml/min SO was eluted earlier. The analysis times of the compounds could not be further decreased for this system. The reproducibility was checked to confirm the results from the first experiments (Table 3). The results showed good reproducibility for experiment 3 and 4. The reproducibility for experiment 2 could not be tested due storage problems. The start substrate 2,3-epoxypropylbenzene By using the shorter C-18 column and gradient elution, the compounds analysis times could be decreased. The analysis time was decreased from 80 min, when a longer C-18 column was used, to 20 min with good separation. 17

Conclusion A gradient and a shorter nonpolar C18-column with dimensions of 100x2 mm were used to shorten the analysis times of the compounds in a multistep reaction. As an example, it would save 42-92 h if 100 samples were analyzed. A further advantage of shorter analysis times of the compounds is that less solvents would be used. Good reproducibility could also be obtained with the protocol that was used. Personal reflection Shortening the retention time of the epoxides, diols and ketones with retained separation is highly important, since many samples need to be run in HPLC, which is very time consuming. Reducing the analysis times of the compounds helps the researchers to detect the properties of the compounds faster and thus they may detect a treatment. This is why this project was important. To improve the experiment further I would change the flow rate and gradient to obtain better separation between the compounds. I would also try to obtain the ketone 1- Hydroxy-3-phenylpropan-2-one by increasing the concentration of the substrate 2,3-EBP even more. Acknowledgements A big thank You to PhD. Student Åsa Janfalk Carlsson and Prof. Mikael Widersten for supervision. 18

References [1] Daniel C. Harris and Michelson Laboratory, High Performance liquid Chromatography, in Quantitative Chemical Analysis, W. H. Freeman and Company, 2010. [2]. Janfalk Carlsson. Å, Bauer P., Ma H. and Widersten M., Obtaining optical purity for product diols in enzyme-catalyzed epoxide hydrolysis: contributions from changes in both enantioand regioselectivity, Biochemistry 51, 7627 7637, 2012. [3] G. Ann, M. Widersten, Modification of Substrate Specificity Resulting in an Epoxide Hydrolase with Shifted Enantiopreference for (2,3-Epoxypropyl)benzene, Volume 11, 1422-1429, 2010. [4] Goldberg A.P., Nowakowska E., Antle P.E., Retention-optimization strategy for the highperformance liquid chromatographic ion-pair separation of samples containing basic compounds, Journal of chromatography 316, 241-260, 1984. [5] Bahowick J.T., Synovec E.R., Correlation of quantitative analysis Precision to retention time precision and chromatographic resolution for rapid, short-column analysis, Anal. Chem. 67, 631-640, 1995. [6] L.R. Snyder, J.W. Dolan, J.R. Grant, Gradient elution in high-performance liquid chromatography : I. Theoretical basis for reversed-phase systems, Journal of chromatography Volume 165, 3-30, 1979. 19

Appendix Appendix 1. The molecular weight (Mw), concentration, and volume of the compounds SO, diol 1 and ketone 1, see table A1. Table A1- The Mw, mass, concentration, volume and retention time of the compounds. Sample Substance Physical state Mixed with Molecoular weight (g/mol) Mass Conc. (mm) Volume (ml) Retention time (min) 1 SO Liquid Water 120.15 0.36µl 3 1 23.3 2 1-phenyl-1.2- Solid Acetonitrile 138.16 13.0mg 3 1 4.9 ethanediol 3 2- hydroxyacetophenone Solid Acetonitrile 136.144 14.4mg 3 1 6.5 Acetonitrile was added due to solubility problems. Appendix 2. The volumes that were used for sample 1 and sample 2, with 2,3-EPB as a start substrate, see Table A5. Table A5- The volume of the substrates and enzymes that were used for sample 1 and 2. Sample Epoxide 2.3-EPB (µl) StEH1 (µl) AdhA (µl) NAD + (µl) Zink (µm) Buffer 0.1M Sodium phosphate ph 7.9 (µl) 1 4 5 - - - 991 2 0.41 1 10 150 1 837 20

Appendix 3. Some of the gradient experiments that were run with SO styrene oxide as a start substrate, but failed to give good and shorter retention time. See Table A9 and A10. Table A9- The conditions that were used during isocratic elution and gradient elution. Experiment Conditions 1 15% acetonitrile / 85% H 2O (isocratic elution) 2 At time 15:00 15:01 min the gradient increased from 0% to 15% acetonitrile. 15:01 20:00 min the gradient increased to 80% acetonitrile. 20:00 20:03 min the gradient was still at 80% acetonitrile (gradient elution) 3 At time 7:00 9:00 min the gradient increased from 0% to 15% acetonitrile. 9:00 21:00 min the gradient increased to 30% acetonitrile and maintained at it for more 4 min. 25:00-25:01 min the gradient was decreased to 15% acetonitrile. (gradient elution) 4 At time 7:00 9:00 min the gradient increased from 0% to 15% acetonitrile. 9:00 21:00 min the gradient increased to 90% acetonitrile and maintained at it for more 4 min. 25:00-25:01 min the gradient was decreased to 15% acetonitrile. (gradient elution) 5 At time 8:00 8:01 min the gradient increased from 0% to 20% acetonitrile. 8:01 11:00 min the gradient increased to 90% acetonitrile. 11:01-11:03 min the gradient was decreased to 20% acetonitrile. (gradient elution) Table A10- The retention time of the compounds with isocratic elution and gradient elution. Experiment Flow rate (ml/min) Ret. time of 1-Phenyl-1-2- ethanediol (min) Ret. time of 2- hydroxyacetophenone (min) Ret. time of SO (min) Graph 1 0.38 4.9 6.5 23.3-2 0.38 - - 21.3 Figure 10 3 0.38 - - - Figure 11 4 0.38 - - 12.1 Figure 12 5 0.40 - - 12.2 Figure 13 21

The graph of experiments 2, 3 and 4 with absorption at 220nm (mau) are plotted versus time (min). The diol 1 and ketone 1 did not have time to form, see figure A10. 60 m AU Extract-220nm,4nm (1.00) 50 40 21.348/7249473 30 20 10 0 3.162/161421 6.548/64192-10 0.0 5.0 10.0 15.0 20.0 25.0 m in Figure A10- Retention times of, from left to right, diol 1 and SO in experiment 2. The graph of experiment 3 with absorption at 220 nm (mau) are plotted versus time (min). The peaks of the diol 1, ketone 1 and SO could either be stuck together or did not have enough time to form, see figure A11. m AU 25 Extract-220nm4nm (1.00) 20 15 10 6.146/1462687 9.712/730502 11.745/1352738 13.908/161341 5 0-5 -10 28.853/346154-15 -20 0.0 5.0 10.0 15.0 20.0 25.0 m in Figure A11- Retention times of the compounds in experiment 3 are unknown. 22

The graph of experiment 4 with absorption at 220 nm (mau) are plotted versus time (min). The diol 1 and ketone 1 did not have enough time to form, see figure A12. 2500 2000 m AU Extract-220nm4nm (1.00) 12.172/70329979 1500 1000 500 0 6.147/1259633 9.672/698555-500 0.0 5.0 10.0 15.0 20.0 25.0 m in Figure A12- Retention times of the diol 1 and SO in experiment 4. The graph of experiment 5 with absorption at 220nm (mau) are plotted versus time (min). The diol 1 and ketone 1 did not have enough time to form and give a good absorbance. Good absorbance is 500 mau, see figure A13. m AU Extract-220nm4nm (1.00) 1750 1500 1250 1000 12.232/34570127 750 500 250 0-250 8.996/2489908 10.794/3350412 0.0 5.0 10.0 15.0 20.0 25.0 m in Figure A13- Retention times of the diol 1 and SO in experiment 5. The reason why these experiments failed may depends on that the gradient was too high that the diol 1 and ketone 1 did not have enough time to form, or the peaks got stuck together because of that. 23