Directed evolution of a Subtilisin Carlsberg Variant towards ...Tag der mündlichen Prüfung:...

“Directed Evolution of a Subtilisin Carlsberg Varia nt

towards Improved Perhydrolytic Activity for Industr ial

Application”

Von der Fakultät für Mathematik, Informatik und Naturwissenschaften der RWTH

Aachen University zur Erlangung des akademischen Grades einer Doktorin der

Naturwissenschaften genehmigte Dissertation

vorgelegt von

Diplom -Biochemikerin

Dragana Despotovic

aus Sabac, Republic of Serbia

Berichter: Universitätsprofessor Dr. Ulrich Schwaneberg

Universitätsprofessor Dr. Martin Zacharias

Tag der mündlichen Prüfung: 12.04.2012

Diese Dissertation ist auf den Internetseiten der Hochschulbibliothek online verfügbar.

Table of Contents

i

Table of contents

Acknowledgments ........................................................................................................... 1

Abstract ........................................................................................................................... 3

Abbreviations .................................................................................................................. 4

Part I: General Introduction ............................................................................................. 5

1. Protein Engineering (directed evolution and rational design) ................................. 5

2. Perhydrolases and peroxycarboxylic acids .......................................................... 10

2.1 Enzyme promiscuity ..................................................................................... 12

2.2 Applications of the perhydrolase-catalyzed synthesis of peroxycarboxylic

acids ................................................................................................................... 13

3. Project objective ................................................................................................... 15

Part II: Assay Development for Detection of Peroxycarboxylic Acids ............................ 16

1. Introduction .......................................................................................................... 16

1.1 Aim of this work ............................................................................................ 16

1.2 Detection of peroxycarboxylic acids ............................................................. 16

1.3 Flow cytometry screening system ................................................................. 19

1.4 Enzyme surface display on Bacillus subtilis cells ......................................... 20

2. Materials .............................................................................................................. 22

2.1 Reagents ...................................................................................................... 22

2.2 Apparatus ..................................................................................................... 22

2.3 Bacterial strains ............................................................................................ 23

2.4 Plasmids ....................................................................................................... 23

2.5 Oligonucleotides ........................................................................................... 23

3. Methods ............................................................................................................... 24

3.1 Synthesis of 7-(4-aminophenoxy)-3-carboxy coumarin (APCC) ................... 24

3.2 Sample preparation and analysis of peroxyacetic acid (PAA) and 7-

hydroxycoumarin-3-carboxylic acid (HCC) ......................................................... 27

Table of Contents

ii

3.3 Expression of subtilisin Carlsberg ................................................................. 28

3.4 Determination of enzymatic activity .............................................................. 28

3.5 APCC assay performed in double emulsions ............................................... 29

3.6 Emulsification of Bacillus cells ...................................................................... 29

3.7 Fusion of CWBc domain to Perhydrolase ..................................................... 30

3.8 Proteolytic activity ......................................................................................... 30

4. Results ................................................................................................................. 31

4.1 APCC assay principle ................................................................................... 31

4.2 Effect of bromide concentration .................................................................... 32

4.3 Effect of APCC concentration ....................................................................... 33

4.4 Effect of hypobromite and peroxycarboxylic acid on fluorescence intensity of

HCC .................................................................................................................... 34

4.5 Sensitivity ..................................................................................................... 35

4.6 Peroxycarboxylic detection in the presence of hydrogen-peroxide ............... 36

4.7 Detection of enzymatic activity ..................................................................... 37

4.8 Future development – optimization for high-throughput screening ............... 40

4.9 Surface display of protease .......................................................................... 44

5. Discussion ........................................................................................................... 45

6. Conclusion ........................................................................................................... 46

Part III: Directed Evolution of Subtilisin Carlsberg for Improved Perhydrolytic Activity .. 47

1. Introduction .......................................................................................................... 47

1.1 Aim of this work ............................................................................................ 47

1.2 Proteases ..................................................................................................... 47

1.2.1 Serine proteases and subtilisins ..................................................... 50

1.2.2 Protein engineering in subtilisin proteases ...................................... 56

1.2.3 Subtilisin Carlsberg ......................................................................... 58

1.2.4 Promiscous activity of subtilisin Carlsberg and Thr59Ala/Leu217Trp

variant ...................................................................................................... 59

2. Materials .............................................................................................................. 63

Table of Contents

iii

2.1 Chemicals ..................................................................................................... 63

2.2 Bacterial strains ............................................................................................ 63

2.3 Plasmids ....................................................................................................... 64

2.4 Oligonucleotides ........................................................................................... 64

2.5 Cell culture media and cultivation ................................................................. 66

3. Methods ............................................................................................................... 67

3.1 Cloning ......................................................................................................... 67

3.2 Transformation of B. subtilis WB600 and DB104 .......................................... 67

3.3 epPCR library generation ............................................................................. 68

3.4 SeSaM library generation ............................................................................. 69

3.5 Site Saturation Mutagenesis of Perhydrolase ............................................... 71

3.6 Protein expression ........................................................................................ 71

3.7 Catalase inhibition ........................................................................................ 72

3.8 Screening for decreased proteolytic activity ................................................. 72

3.9 Screening for improved perhydrolytic activity ............................................... 72

3.10 Purification of serine proteases with affinity chromatography ..................... 73

3.11 Protein purification ...................................................................................... 74

3.12 Proteolytic activity with natural substrate – skim milk ................................. 75

3.13 Proteolytic activity – suc-AAPF-pNA ........................................................... 75

3.14 Perhydrolytic activity of the variants ........................................................... 75

3.15 Esterolytic activity of the variants ................................................................ 76

3.16 Modeling studies ......................................................................................... 77

4. Results ................................................................................................................. 78

4.1 Expression system ....................................................................................... 78

4.2 Perhydrolytic (APCC) assay optimization for microtiter plate screening ....... 79

4.3 Diversity generation ...................................................................................... 80

4.4 Screening ..................................................................................................... 82

4.5 Diversity generation and screening optimization .......................................... 83

4.6 Rational design approach ............................................................................. 88

Table of Contents

iv

4.6.1 SSM at positions included in catalysis and close to active site ....... 89

4.6.2 SSM of positions in the substrate binding pocket ........................... 91

4.7 Purification of selected variants .................................................................... 95

4.8 Characterization of the variants .................................................................... 99

4.8.1 Characterization with suc-AAPF-pNA as a substrate ...................... 99

4.8.2 Skim milk activity ........................................................................... 100

4.8.3 Perhydrolytic activity: kinetic parameters for hydrogen-peroxide and

methyl-propionate .................................................................................. 101

4.8.4 Perhydrolytic activity: kinetic parameters for methyl-butyrate and

methyl-pentanoate ................................................................................. 103

4.8.5 Esterolytic activity of the variants .................................................. 105

5. Discussion ......................................................................................................... 107

5.1 Suc-AAPF-pNA specificity .......................................................................... 110

5.2 Ester specificity ........................................................................................... 113

6. Conclusion ......................................................................................................... 120

Future prospects ......................................................................................................... 121

References .................................................................................................................. 122

Publication list ............................................................................................................. 129

Curriculum vitae .......................................................................................................... 130

Appendix ......................................................................................................................... A

Acknowledgements

1

Acknowledgments

During my PhD research at Jacobs University Bremen and RWTH Aachen

University many people influenced my scientific and personal development.

First, I would like to thank my supervisor Prof. Dr. Ulrich Schwaneberg for giving

me the opportunity to work in this group, as well for his guidance and support during my

PhD studies. I admire his scientific thoughts and his dedication to his work. He was an

excellent life and science teacher.

I would like to thank my PhD committee members Prof. Dr. Martin Zacharias and

Prof. Dr. Lothar Elling for accepting to be in my thesis committee. I appreciate their time

to read this thesis and Prof. Zacharias for making his trip from Munich to Aachen. I am

thankful to Prof. Zacharias for helping us with the modeling part of the project, hosting

our visit to Munich and Technische Universität München where we were introduced to

the group of great scientist and their interesting work.

I acknowledge BMBF for financial support and thank our collaborators Dr.

Hendrik Hellmuth, Dr. Timothy O’Connell, Nina Mussman and Inken Prueser at Henkel

AG & Co KGaA. It was a great pleasure to be part of their team even for at least a short

period. I learned a lot about science and collegiality. I also thank Prof. Dr. Karl-Heinz

Maurer for his support on this project; I truly enjoyed scientific discussions with him.

I thank my co-supervisors Dr. Radivoje Prodanovic and Dr. Ronny Martinez. My

gratitude to Ronny, who supported me from the beginning of my PhD. I learned a lot

from him, enjoyed our discussions and I am very thankful for his help during manuscript

writing and for careful reading of my thesis. It was also great pleasure to be part of HTS

subgroup together with my colleagues: Ronny Martinez, Ljubica Vojcic, Felix Jakob,

Christian Lehmann, Georgette Wirtz and Christian Pitzler.

I thank to my students Ranjan Shrestha and Jana Ognjenovic; I hope they

learned from me as much I learned from them. They were all a great help in the lab and

I really enjoyed working with them.

I am thankful to all members of Schwaneberg group for their support; it was a

great pleasure working and learning from them. Multicultural environment was a big

Acknowledgements

2

experience for me as a person and a scientist. I appreciate the help from Daniela,

Gisela, Wolfgang and Brigitta in lab and Marina, Monika and Anita’s for help with

administrative work.

I thank to my collegues and friends Dr. Milan Blanusa, Raluca Ostafe, Dr.

Hemanshu Mundhada and Dr. Amol Shivange for their support, scientific discussions,

advices and small talks which I enjoyed especially. I am also very thankful to Arsenovic

family who helped me a lot during my stay in Aachen, their house was a second home

to me.

At last, I would like to thank my friends and family: Jelena, Aleksandra, Milena,

Ana, aunt Julijana and cousin Danica for their love and support. Thank you for believing

in me, had time and patience to listen and provide advices. My gratitude and love goes

to my father, my mother and sister who always believed in me. My mother helped us to

achieve our goals and I wish we made her proud. She is the strongest person I know

and the best teacher I ever had. I am thankful to my father for his love and guidance; I

believe that he would be happy to see how much we are committed to our family and

work.

All this work would have not been possible without the financial support of the

German government through the Bundesministerium für Bildung und Forschung

(BMBF) and Henkel AG & Co KGaA.

Dragana Despotovic

Abstract

3

Abstract

Adapting biocatalysts to non-conventional substrates and environments is of high

academic and industrial interest. Directed evolution and rational design methods offer

opportunities to tailor protein properties by generating diversity on gene level and

screening for new variants on the protein level.

Subtilisin Carlsberg is one of the first described serine proteases from the

subtilisin family. Due to its stability under production conditions and broad specificity,

this enzyme has become industrially important, especially in the detergent industry.

Subtilisin Carlsberg also catalyzes ester perhydrolysis, generating peroxycarboxylic

acids, compounds well known as bleaching and disinfection agents.

The bottleneck for most experiments focused on finding and improving

peroxycarboxylic acid producing enzymes is the availability of a medium and

high‐throughput screening system. To overcome this obstacle, a sensitive, fluorescent

assay with detection limit below 1 µM of peroxyacetic acid and suitable for detection of

perhydrolytic activity without previous enzyme purification was developed. The assay

was optimized for screening in 96 and 384-well microtiter plate system, with the

possibility to further upscale throughput by using flow cytometry screening technology.

The screening platform was developed in this work to improve protease and likely other

hydrolases (e.g. lipases or esterases) having perhydrolytic activity or find new enzymes

with peroxycarboxylic acid production by screening of metagenome libraries.

Subtilisin Carlsberg was furthermore the first protease engineered through

directed evolution towards increased perhydrolytic activity which offers novel

opportunities for bleaching applications. The protease variant, T59A/G166L/L217W,

identified in the current work showed a 9.4-fold increased specificity constant of

perhydrolytic activity compared to the wild type. Position 166 is part of the substrate

binding pocket and changes in the side-chain volume at this position were found to

modulate substrate specificity of subtilisin Carlsberg.

Abbreviations

4

Abbreviations

ABTS 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid)

APCC 7-(4-aminophenoxy)-3-carboxy coumarin

B. subtilis Bacillus subtilis

CMC Carboxymethylcellulose

Da Dalton

DMSO Dimethyl-sulfoxide

DNA Deoxyribonucleic acid

E. coli Escherichia coli

epPCR Error prone PCR

HCC 7-hydroxycoumarin-3-carboxylic acid

HPLC High Pressure Liquid Chromatography

HTS High Throughput Screening

LB Luria-Bertani medium

MEGAWHOP Megaprimer PCR of Whole Plasmid

MTP Microtiter plate

OD Optical Density

PAA Peroxyacetic acid

PBS Phosphate buffered saline

PCR Polymerase Chain Reaction

PDB Protein Data Bank

pI Isoelectric point

PMSF Phenylmethylsulfonyl fluoride

SSM Site Saturation Mutagenesis

StEP Staggered Elongation Process

Suc-AAPF-pNA N-succinyl-Ala-Ala-Pro-Phe-p-nitroanilide

TCA Trichloroacetic acid

Tet Tetracycline

w/o/w Water-in-oil-in-water

WT Wild type

Part I: General Introduction

5


1. Protein Engineering (directed evolution and rati onal

design)

“There are two fundamental conditions for life. First, the living entity must be able

to self-replicate; second, the organism must be able to catalyze chemical reactions

efficiently and selectively.” (Lehninger, Nelson et al. 2005).

Enzymes, highly specialized proteins, are the most efficient catalysts having

efficiency often greater than that of synthetic or inorganic ones. Enzymes can perform

regio-selective and/or enantio-selective chemical transformations and accelerate

reaction rates by factors up to 1012, all under mild conditions. Although highly attractive

for chemical synthesis, enzymes are usually not well suited for bioindustrial applications

(Arnold 2001). Limitations include low catalysis of nonnatural substrates, low stability

(towards temperature, pH, and oxidative agents), poor activity in non-aqueous media,

requirements for expensive cofactors and of course, nature has not provided catalysts

for every reaction. The goal of protein engineering research is to alter nature catalysts

making them suitable for industrial technology. (Farinas, Bulter et al. 2001). The most

successful methods for optimization and redesign of biocatalysts are rational design and

directed evolution.

Rational design is the strategy of modifying biocatalysts that requires not only

knowledge of enzyme crystal structure and function, but also of the property which

should be improved. Many properties are well understood and in some of the cases

general rules might be applied, for example termostability can be improved by

introduction of disulfide bridges, ionic interactions and salt bridges that increase the

rigidity of the protein scaffold, reducing the possible conformations of the unfolded

protein (Korkegian, Black et al. 2005). On the other hand, there are cases in which

specific knowledge is necessary to change enzyme specificity or to increase catalytic

efficiency (Lassila, Keeffe et al. 2005). These general rules are derived from studying

the effect of certain mutations on

methods accelerate the generation of new biocatalysts

small and high-quality mutant libraries based on protein structure and/or interactions

protein-substrate (Reetz, Bocola et al. 2005

the structure and/or the molecular basis for a

enzymes.

Directed protein evolution

to create improvements within

presented in Figure 1.1, starting from diversity generation where mutations are

randomly introduced in the gene of interest

expression in appropriate host.

agar plate and/or microtiter plate assays.

usually used for next round of evolution until desired property

achieved.

Figure 1.1. General overview of a directed evolution experimentProdanovic et al. 2010).


6

the effect of certain mutations on the protein structure and/or activit

generation of new biocatalysts by the generation of

quality mutant libraries based on protein structure and/or interactions

Reetz, Bocola et al. 2005; Wong, Roccatano et al. 2007

the molecular basis for a desired function are not available for most

Directed protein evolution is an approach which requires an outside intelligence

improvements within species. Major steps in directed evolution experiment are

, starting from diversity generation where mutations are

ndomly introduced in the gene of interest (Step I) to library

in appropriate host. In Step II library is screened for improved variants

agar plate and/or microtiter plate assays. Identified variants are isolated (Step II

usually used for next round of evolution until desired property or level of improvement

General overview of a directed evolution experiment. Taken over from

t I: General Introduction

protein structure and/or activity. Rational design

the generation of relatively

quality mutant libraries based on protein structure and/or interactions

Wong, Roccatano et al. 2007). However,

not available for most

requires an outside intelligence

Major steps in directed evolution experiment are

, starting from diversity generation where mutations are

to library transformation and

for improved variants using

isolated (Step III) and

or level of improvement is

. Taken over from (Güven,


7

Among the main challenges in directed evolution campaigns are the generation

of unbiased diversity and development of screening systems which enable scanning

through a meaningful fraction of the generated sequence space (Wong, Zhurina et al.

2006). A large number of chemical and molecular biology methods are now available for

generating random mutation in gene sequences, site specific and recombination of

mutations. The most commonly used random mutagenesis techniques are the PCR

methods under low fidelity conditions (epPCR) (Cirino, Mayer et al. 2003). One of the

latest random mutagenesis methods is Sequence Saturation Mutagenesis (SeSaM),

which overcomes limitations caused by polymerase bias (transitions over

transversions), saturates every position in a protein and is able to introduce subsequent

mutations (Wong, Tee et al. 2004). The most used recombination methods in directed

evolution used in the last step to combine beneficial mutations from different variants

are DNA shuffling (Stemmer 1994) and Staggered Extension Process (StEP) (Zhao,

Giver et al. 1998). Which of these methods will be used for directed evolution

experiments usually depends on the property which is aimed to be improved and

screening capabilities. Engineering termostability usually demands higher number of

mutations and substitution with chemically different amino acids leading to SeSaM as

method of choice for diversity generation. While, for enantioselectivity changes to

similar amino acids are more appropriate with epPCR as a method of choice for

diversity generation (Reetz and Jaeger 1999; Arnold 2001).

A large variety of methods for the generation of genetic diversity is available,

although currently the bottleneck of directed evolution is the availability of screening

systems for a desired property. Screening technology should reflect as close as

possible the property of interest, with substrate identical or similar to the target

substrate. Furthermore, the assay should be optimized for medium and/or high

throughput screening and sensitive over the final application range of substrate

concentration, enabling identification of desired hits in the library (Aharoni, Griffiths et al.

2005).

The most common screening methods are solid phase screening assays and

microtiter plate assays. Solid phase screening systems can be performed directly on the

agar plate, on filter paper or on membrane. These types of methods have a high


8

throughput, of 104‐106. The crucial factor when using this screening technique is that

substrate conversion generates a visual signal, such as fluorescence or color, to identify

colonies expressing an enzyme with desirable properties. The drawback is that they

have low accuracy and that’s why they are usually used as a pre-screen method.

Microtiter plate based screenings are the most commonly applied method in identifying

desirable enzyme variants. Microtiter plate assay enables screening 96, 384 or 1536

clones at the same time and these assays are considered as medium throughput.

Single transformants are grown in microtiter plates and variants are usually assessed in

a second plate, while the original plate is stored as back-up. Supernatant or cell lysate

is used for detection of enzymatic activity and generated signal measured by

spectrophotometer or fluorometer. High-throughput screening (HTS) systems are

Fluorescent Activated Cell Sorter (FACS) and microfluidic devices which can routinely

screen 107 clones per hour (Farinas 2006). This system is based on detection of

fluorescent product generated by enzyme reaction and linked to the cell surface or

entrapped in double emulsion compartments/microbeads, enabling the link between

genotype and phenotype (Mastrobattista, Taly et al. 2005). FACS greatly accelerates

HTS, but this technique is still in development, allowing for now only selection between

positive and negative variants. Figure 1.2 summarizes screening strategies currently

used.


9

Figure 1.2. Overview of screening technologies. (A) Experimental steps from obtaining the gene library as PCR product to the actual screen. (B) explanation of symbols, (C) Cell growth/survival selection and agar plate screening, (D) microtiter plate screening, (E) cell as micro-reactor, (F) cell surface display, (G) cell-in-droplet, and (H) in vitro compartmentalization. Taken from (Leemhuis, Kelly et al. 2009).

2. Perhydrolases and per

Perhydrolases catalyze the

CO-O-OH) from carboxylic acids (or

had been known as metal free haloperoxidases which

triad and structure very similar to

catalyze hydrolysis of chemical bonds.

al. 2006).

Figure 1.3. Reaction scheme for enzyme catalyzed perhydrolysis

Perhydrolysis presumably t

which a carboxylic acid first reacts with the

enzyme intermediate, which

intermediate (TI) leading to release of free enzyme and per

(Hofmann, Tölzer et al. 1998

mechanism, Bugg suggested that acetic acid is directly attacked by hydrogen

without formation of acetyl-

is not covalently bound for protein

perhydrolytic reaction are presented at Figure


10

and per oxycarboxylic acids

catalyze the reversible formation of peroxycarboxylic acids

from carboxylic acids (or esters) and hydrogen peroxide

had been known as metal free haloperoxidases which contain a Ser

and structure very similar to serine hydrolyses-esterases

catalyze hydrolysis of chemical bonds. (Pelletier, Altenbuchner et al. 1995

Reaction scheme for enzyme catalyzed perhydrolysis.

Perhydrolysis presumably takes place with an ping-pong bi

which a carboxylic acid first reacts with the active site serine group to form an acyl

enzyme intermediate, which reacts with hydrogen peroxide forming second tetrahedral

leading to release of free enzyme and per

Hofmann, Tölzer et al. 1998). This ping-pong mechanism is not the only proposed

mechanism, Bugg suggested that acetic acid is directly attacked by hydrogen

-enzyme intermediate where second tetrahedral intermediate

t covalently bound for protein (Bugg 2004). Possible mechanism

rhydrolytic reaction are presented at Figure 1.4.

t I: General Introduction

reversible formation of peroxycarboxylic acids (R-

esters) and hydrogen peroxide, Figure 1.3. They

contain a Ser-His–Asp catalytic

and lipases which

Pelletier, Altenbuchner et al. 1995; Song, Ahn et

pong bi-bi mechanism in

active site serine group to form an acyl-

ing second tetrahedral

leading to release of free enzyme and peroxycarboxylic acid

pong mechanism is not the only proposed

mechanism, Bugg suggested that acetic acid is directly attacked by hydrogen-peroxide

nd tetrahedral intermediate

Possible mechanisms of the


11

Figure 1.4. Mechanisms for pehydrolysis of acetic acid catalyzed by P. fluorescens esterase (PFE), a) the ping-pong bi-bi mechanism which includes acetyl-enzyme intermediate, b) bi-bi mechanism proposed by Bugg. The numbering corresponds to the active site of PFE. Figure taken from (Yin, Bernhardt et al. 2010).

Although perhydrolases and hydrolases have similar structure and reaction

mechanism, it is not clear why hydrolases have much lower perhydrolytic activity than

hydrolytic and why perhydrolases have much lower hydrolytic activity compared to


12

perhydrolytic. For example, Pseudomonas fluorescens esterase (PFE) shows low

perhydrolytic activity and has similar tertiary structure with the perhydrolases and Ser-

His–Asp catalytic triad. Bernhardt et al. (Bernhardt, Hult et al. 2005) identified residues

around the catalytic center which are only conserved for pehydrolases by aligning the

sequences from six hydrolases and six perhydrolases. Residues close to the active site

of PFE were identified and exchanged. Using this approach, perhydrolytic activity of

PFE was increased 28-fold with only one substitution (Leu29Pro). The proposed

explanation of this change was that this amino acid substitution shifts the backbone of

the neighboring Trp which could form hydrogen bond with perhydroxy-group in the

second tetrahedral intermediate, stabilizing the state, Figure 1.3a. Specific stabilization

of the second tetrahedral intermediate for the perhydrolysis reaction is of key

importance for completion of the reaction.

2.1 Enzyme promiscuity

Currently, only seven bacterial perhydrolases have been cloned, sequenced and

characterized (Wiesner, van Pée et al. 1988; Bantleon, Altenbuchner et al. 1994;

Pelletier, Pfeifer et al. 1994; Burd, Yourkevich et al. 1995; Kirner, Krauss et al. 1996;

Song, Ahn et al. 2006). Thus, there is a need to expand the group of enzymes which

can be used as a perhydrolases. Esterases, lipases and some proteases beside their

natural hydrolytic activity also show perhydrolytic activity (Bornscheuer and Kazlauskas

2004; Lee, Vojcic et al. 2010). This ability of enzymes to catalyze different synthetic

reactions, which are more or less far from their natural role, is called catalytic

promiscuity. In many cases the origin of such promiscuity can be found in the evolution

of the enzymes from a common ancestor, which could have generated/maintained

different residual side-reactions.

Catalytic promiscuity has been classified in three groups: (1) condition

promiscuity which is the catalytic activity in various reaction conditions different from

natural ones, (2) substrate promiscuity which is shown by enzymes with broad substrate

specificity (3) catalytic promiscuity, exhibited by enzymes catalyzing distinctly different

chemical transformations with different transition states. Enzyme catalytic promiscuity


13

can be either accidental, a side reaction catalyzed by the wild-type enzyme; or induced,

a new reaction established by one or several mutations changing the reaction catalyzed

by the wild-type enzyme (Hult and Berglund 2007). Rational understanding of the

causes could allow the development of new and useful enzymatic reactions by using

“well-known” enzymes (Bornscheuer and Kazlauskas 2004; Hult and Berglund 2007).

Promiscuous enzymes are then protein of interest for directed evolution, and new

knowledge regarding their mechanisms has to be generated in order to make them

applicable in industrial processes.

In the case of enzymes with perhydrolytic activity, the switch from water to

hydrogen peroxide as the nucleophilic agent could be considered as a change in

substrate selectivity. However, some serine hydrolases do not exhibit perhydrolase

activity which suggests that the Ser–His–Asp catalytic triad is not the only determinant

for perhydrolase activity, also better stabilization of hydrolytic or perhydrolytic second

tetrahedral intermediate will favor one or another reaction (Bernhardt, Hult et al. 2005).

2.2 Applications of the perhydrolase-catalyzed synthesis of peroxycarboxylic

acids

The use of hydrogen peroxide as a bleaching agent (because of its oxidizing

power) is well known since a long time. Due to its instability, it was produced in situ by

using sodium perborate and high temperature. Thus, in the 1950s peroxycarboxylic

acids and peroxycarboxylic acids-generating bleaches were recognized for their

potential to provide efficient bleaching at lower washing temperatures (Stanley and

Gianfranco 1983). Medium chain alkyl peroxycarboxylic acids are very effective for the

bleaching system, they have a broad application in various industry branches such as

synthesis of epoxides (Zhao, Wu et al. 2011), alcohols oxidation (Cella, Kelley et al.

1975), nylon manufacturing (Dicosimo, Gavagan et al. 2008). As antimicrobial agents,

since they can damage virtually all types of macromolecules, which ultimately leads to

cell lysis and microbial death (Baldry 1983). Due to these properties, peroxycarboxylic

acids are widely used in agricultural industry, food establishments, and medical facilities

(Block 2001). However, neither hydrogen peroxide, nor the organic peroxycarboxylic


14

acids are stable over time in aqueous solution. Therefore, an in situ generation is almost

mandatory for an effective bleaching procedure. The combination of an ester-

hydrolyzing enzyme, an acyl-alkyl ester, and a hydrogen-peroxide/peroxide donor, is an

effective bleaching platform (Minning, Weiss et al. 1999). Since the process takes place

in aqueous environment, a close interaction between the peroxycarboxylic acid,

hydrogen-peroxide and the hydrolase, is expected. In such conditions the catalyst is

highly sensitive to denaturation, and consequently, it is important to find a biocatalyst

able to work in that environment (Carboni-Oerlemans, Domínguez de María et al. 2006).

Additionally, there is a competition between water and hydrogen-peroxide for the acyl–

enzyme complex of the hydrolase. Therefore, the ratio of perhydrolysis- to hydrolysis-

rate (P/H) in a reaction mixture determines the efficiency of the enzymatic bleaching

system. Hydrolases also perform the reverse reaction, and form hydrogen peroxide and

carboxylic acids from the peroxycarboxylic acids in aqueous solutions. Related to this,

most of the commercial hydrolases in aqueous media have low a P/H value which limits

their practical applications. Thus, there is a need to improve hydrolases to obtain

perhydrolases of industrial interest (Carboni-Oerlemans, Domínguez de María et al.

2006).


15

3. Project objective

The aim of this work is to generate new perhydrolytic enzymes by altering the

catalytic properties of a serine protease (subtilisin Carlsberg) from hydrolase to

perhydrolase by means of directed evolution. The work is divided in two parts: the first

part is the development of a medium throughput detection system for peroxycarboxylic

acids, and the second part is directed evolution of subtilisin Carlsberg. The aim of the

first part is to generate a sensitive, simple assay able to follow continuous enzymatic

production of peroxycarboxylic acids in expression media. This assay should be

optimized for medium and high throughput screening system for identification of

subtilisin Carlsberg variants with improved perhydrolytic activity.

The aim of the second part is to identify by the screening of mutagenesis

libraries, subtilisin Carlsberg variants with increased perhydrolytic activity for application

in washing and cosmetic industry using the screening system previously developed in

this work. The analysis of amino acid substitutions present in variants with increased

perhydrolytic/hydrolytic activity ratio should provide new insights in understanding the

mechanism of promiscuous enzymatic activity by the identification of common features

that could be extrapolated to other hydrolases for the generation of new perhydrolases.

Part II: Assay Development

16

Part II: Assay Development for Detection of Peroxy-

carboxylic Acids

1. Introduction

1.1 Aim of this work

The goal of this work was to develop a new detection system for

peroxycarboxylic acids with different length of alkyl chain, which is sensitive (detection

limit below 1 µM), simple and suitable for screening in microtiter plate (96 and 384-well

format) without previous enzyme purification.

1.2 Detection of peroxycarboxylic acids

Peroxycarboxylic acids (R-CO-O-O-H) are oxidants which are used in bleaching,

disinfection (Rüsch gen. Klaas, Steffens et al. 2002), synthesis of epoxides (Zhao, Wu

et al. 2011), alcohols oxidation (Cella, Kelley et al. 1975) and antimicrobial agents

(Block 2001).

In particular, peroxyacetic acid (PAA) and medium chain (C7–C12)

peroxycarboxylic acids, have been introduced in the detergent industry, in addition to

hydrogen peroxide, as bleaching agents (Akerman, Phillips et al. 2003). Despite their

multiple uses, peroxycarboxylic acids are not widespread in industry due to their

instability in aqueous solution (Yuan, Ni et al. 1997), and the explosive character of their

pure form (Swern 1949). As an alternative, peroxycarboxylic acids can be produced in

situ by hydrolysis of esters or amides in the presence of hydrogen peroxide directly into

their final application.

Current methods for determination and quantification of peroxycarboxylic acids

are summarized in the Table 2.1.

The first approach for detection of peroxycarboxylic acids includes analytical

methods based on two step titration (methods 1 and 2 in the table) (D'Ans and Frey

1913; Greenspan and MacKellar 1948), where hydrogen-peroxide titration is followed by


17

a iodometric determination of peroxycarboxylic acids. With some modifications (method

3) (Sully and Williams 1962) such is detection under low acidic condition in which the

rate of equilibration between peroxyacids and peroxide is negligible, these methods are

still widely used for PAA monitoring. Method 4 (Di Furia, Prato et al. 1984) is a gas-

liquid chromatography including pre-column derivatization, while methods 5 and 6 (Kirk,

Damhus et al. 1992; Effkemann, Pinkernell et al. 1998) are liquid chromatographic

(HPLC) methods with direct separation or post-column derivatization. The development

of chromatographic methods, significantly increased sensitivity; reaching a detection

limit below 1 µM (Kirk, Damhus et al. 1992). HPLC analysis is also used as a reference

technique to calibrate other methods for peroxycarboxylic acid determination.

Current colorimetric assays (methods 7-9) suitable for the screening of enzymes

catalyzing perhydrolytic reaction are based on the oxidation of azo-dyes (Minning,

Weiss et al. 1999), chromogens (aromatic amines, phenols) (Fischer, Arlt et al. 1990),

iodine (Davies and Deary 1988) or 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic

acid) (ABTS) in the presence of peroxycarboxylic acids (Pinkernell, Luke et al. 1997;

Harms and Karst 1999; Binder and Menger 2000). ABTS assay (method 9) is an end

point assay for determination of peroxycarboxylic acids, reported to be used in microtiter

plates (Pinkernell, Luke et al. 1997). Beside spectrophotometric assays, fluorescent

detection methods offer high sensitivity and simplicity. Method 10 (Jacks 2004) is the

only fluorescent assay reported for detection of hypohalites and peroxycarboxylic acids.

This method was validated by activity studies on purified and semi-purified heme and

non-heme haloperoxidase activities (Jacks 2005).

The fluorescence of 7-hydroxycoumarin in aqueous media is completely lost in 7-

aryloxycoumarins. O-dearylation of coumarins can be used as a trigger in detection of

highly reactive oxygen species, resulting in fluorescent compounds. Nonfluorescent 7-

arylcoumarins can also be O-dearylated by hypohalites, forming highly fluorescent 7-

hydroxycoumarin (Setsukinai, Urano et al. 2000).


18

Table 2.1. Current methods for determination and quantification of peroxycarboxylic acids. Methods labeled in gray could be optimized in a screening platform for directed evolution.

Reference Method type Principle Notes

(D'Ans and Frey 1913)

Titrimetric

H2O2 is titrated using K2MnO4, followed by iodometric determination of

peroxycarboxylic acids (iodide reacts with peroxycarboxylic acids forming

iodine,generated iodine is titrated with thiosulphate solution)

Detection limit: not reported, endpoint assay

Decomposition of peroxycarboxylic acid by Mn2+

(Greenspan and

MacKellar 1948)

Titrimetric H2O2 is titrated using ceric sulphate,

followed by iodometric determination of peroxycarboxylic acids


(Sully and Williams

1962) Titrimetric

Direct iodometric determination of peroxycarboxylic acid and hydrogen peroxide in acidic conditions (pH 3-5)

based on greater reactivity of iodide with peroxycarboxylic acids than with H2O2


(Di Furia, Prato et al.

1984)

Gas-liquid chromatographic

Peroxycarboxylic acids oxidize methyl p-tolylsulphide to methyl p-tolylsulphoxide, residual sulphide or produced sulphoxide

are determined by gas-liquid chromatography


(Kirk, Damhus et al. 1992)

Liquid chromatography

Peroxycarboxylic acids are separated by reversed-phase chromatography followed

by amperometric detection

Detection limits of 0.1-0.6 µM; linear range of 0.05-5 mM.

(Effkemann, Pinkernell

et al. 1998)

Liquid chromatography with postcolumn

derivatization

Peroxycarboxylic acids with chain lengths from C2 to C12 are separated by HPLC on a

reversed-phase C18 column. After separation, peroxycarboxylic acids were

detected using ABTS method

Detection limit: 5 µM, linear range - 10-250 µM for C2 and

10-500 µM for C3 to C12 peroxycarboxylic acids,

endpoint assay

(Minning, Weiss et al.

1999) Colorimetric

Peroxycarboxylic acids oxidize methylsulfanyl-groups of azo-dye leading in

change of color from orange to yellow.

Detection limit: ~ 4 ppm (50 µM), detection range: 4-40 ppm

(50-500 µM) semi-quantitative agar plate method

(Davies and Deary 1988)

Spectrophoto-metric

Measurement of iodine produced in the reaction between peroxycarboxylic acid

and iodide at 352 nm. Linear range: 4.9-68.8 µM.

(Pinkernell, Luke et al.

1997)

Spectrophoto-metric

Peroxycarboxylic acids in the presence of iodide as a catalyst oxidize ABTS forming

green colored radical (pH 3.5)

Detection limit: 1 µM, linear range: 2.5-100 µM. End point assay

(Harms and Karst 1999)

Spectrophoto-metric

Flow injection analysis system for measuring PAA samples based on ABTS

method

Detection limit: 0.7 µM, linear range: 2-700 µM

(Binder and Menger 2000)

Spectrophoto-metric

H2O2 is decomposed by catalase and residual peroxycarboxylic acid determined under the action of peroxidase and ABTS

Detection limit: 45 µM, linear range: 45-550 µM

endpoint assay

(Jacks 2004)

Spectrofluoro-metric

Peroxycarboxylic acids oxidize bromide to hypobromite which reacts with kojic acid

forming fluorescent product. Linear range: 0-100 mM


19

Hydrolases (proteases, lipases or esterases) which are mainly used in

detergents, are reported to generate peroxycarboxylic acids as a result of a secondary

catalytic function in the presence of carboxylic esters and hydrogen peroxide (Björkling,

Frykman et al. 1992; Kirk, Christensen et al. 1994). Indentifying and improving enzymes

for production of peroxycarboxylic acids is a constant subject of interest in detergent

industry. Lipases (Bernhardt, Hult et al. 2005) and proteases (Wieland, Polanyi-Bald et

al. 2009) have been engineered in order to increase their perhydrolytic activity. Despite

significant advance, an efficient enzymatic system for in situ production of

peroxycarboxylic acids is yet to be reported. One of the main challenges is developing

an analytical method capable of monitoring peroxycarboxylic acid generation in the

presence of a large excess of hydrogen peroxide which at the same time is suitable for

high throughput screening in microtiter plates in order to identify new enzyme variants

either by directed evolution or from metagenome libraries.

In this work a novel, highly sensitive fluorescent assay (µM range) is described

as part of a high throughput screening platform for peroxycarboxylic acids producing

enzymes. This continuous assay relies on the detection of coumarin compounds. After

enzymatic reaction, the generated peroxycarboxylic acids oxidize bromide producing

hypobromite, which reacts with 7-(4’-aminophenoxy)-3-carboxy coumarin (APCC)

releasing the highly fluorescent 7-hydroxycoumarin-3-carboxylic acid (HCC) together

with iminocyclohexadienon.

Assay was validated by detection of perhydrolytic activity of protease subtilisin

Carlsberg and T58A/L217W variant expressed in B. subtilis WB600.

1.3 Flow cytometry screening system

One way of accelerating HTS is to use Fluorescence-Activated Cell Sorting

(FACS), which can sort >107 clones per hour. FACS is successfully used for selection of

proteins with high binding affinity such as antibodies. Additionally, FACS can be used

for catalysis, but only if certain requirements are fulfilled: diffusion of product out of the

cell has to be prevented, or to capture the fluorescent product on: the surface of the

cells, into microbeads or in double emulsion compartments, in order to keep the link


20

between phenotype and genotype (Mastrobattista, Taly et al. 2005). Conversion inside

water-in-oil-in-water (w/o/w) double emulsions enabled sorting the compartments by

FACS, and the isolation of living bacterial cells and their enzyme-coding genes. High

enzyme concentration in the reaction droplets (>104 enzyme molecules in <10

femtoliter) increases sensitivity of the method (Aharoni, Amitai et al. 2005).

Figure 2.1. Single-cell compartmentalization and selection by in vitro compartmentalization in w/o/w emulsions (1) A gene library is transformed into expression host, and (2) proteins are expressed in the cytoplasm, or on the surface of the cells, (3) single cells are compartmentalized in the aqueous droplets of a primary (w/o) emulsion, (4) and secondary w/o/w emulsion is formed by emulsification of the primary w/o emulsion, (5) compartments containing the fluorescent product are sorted by FACS. Taken from (Aharoni, Amitai et al. 2005).

1.4 Enzyme surface display on Bacillus subtilis cells

Regardless of the screening or selections system used, the linkage between the

genotype and the phenotype must be somehow maintained in order to isolate the

improved mutants. In vitro compartmentalization in double emulsions of the cells or in

vitro translation machinery in emulsions enables this link. Using in vitro translation

system for compartmentalization has many disadvantages such are: 1) presence of

protease inhibitors in the kit, which will influence on activity of subtilisin Carlsberg; 2)

product of perhydrolytic reaction is peracid, strong oxidative agent which can deaminate

nucleotide bases and mutate the sequence of the gene (a better variant has higher

Fluorescent

product


21

probability that gene will be damaged); 3) delivery of H2O2 through oil into the droplets;

4) very expensive commercial kits.

However, subtilisin Carlsberg is an extracellular protease and prior to

compartmentalization of the cells expressed protease is secreted in the medium, where

only small number of the molecules stays on the surface of the cells. Entrapping

protease on the cell wall would significantly increase the sensitivity of flow cytometry

screening system.

In Bacillus subtilis there are many proteins non-covalently bound on the cell

surface. Major cell wall-binding proteins are peptidoglycan hydrolases, some of which

consists of two domains: catalytic and cell wall-binding domain (CWB). A small cell wall-

binding domain (CWBc) of the Bacillus subtilis peptidoglycan hydrolase CwlC fused to

B. subtilis lipase B was able to be localized on the cell wall of B. subtilis (Kobayashi,

Fujii et al. 2002). Although there are several candidates as an anchoring protein, the C-

terminal domain of CwlC (CWBc) is one of the smallest cell wall-binding domains. CWBc

consists of only 79 amino acids with two direct repeats (each containing 36 amino

acids) and it will be fused to the C-terminus of the Perhydrolase gene.


22

2. Materials

2.1 Reagents

All chemicals were of analytical-reagent grade or higher quality and were

purchased from Fluka, Sigma-Aldrich (Steinheim, Germany) and AppliChem

(Darmstadt, Germany), except APCC which was synthesized. The peroxyacetic acid

(PAA) solution was purchased from Sigma-Aldrich with a labeled concentration of 32 %.

Subtilisin Carlsberg protease was purchased from Sigma-Aldrich. All enzymes were

purchased from New England Biolabs GmbH (Frankfurt, Germany) and Fermentas

GmbH (St. Leon-Rot, Germany).

TLC plates coated with silica gel 60 were from Merck (Haar, Germany). UV was

used as a detection system, TLC plates were exposed to UV light (366 nm and 254

nm).

PD-10 Desalting Column was from GE Healthcare, Munich, Germany.

Silica gel, used for chromatography, was type 60 and purchased from Roth

(Karlsruhe, Germany).

2.2 Apparatus

Rotary evaporator (Rotavap) was from Heidolp Instruments (Schwabach,

Germany). Infinite M1000 (Tecan Group AG, Zürich, Switzerland) and a flow cytometer

(CyFlow® space, Partec GmbH, Muenster, Germany) were used for the detection of

fluorescence. H+ NMR spectra were recorded using Bruker AMX 300 (Bruker BioSpin,

Germany). Thermal cycler (Mastercycler gradient; Eppendorf) and thin-wall PCR tubes

(Multi-ultra tubes; 0.2 ml; Carl Roth, Germany) were used in all PCRs. The PCR volume

was always 50 µl. The amount of DNA in cloning experiments was quantified using a

NanoDrop photometer (NanoDrop Technologies, Germany).


23

2.3 Bacterial strains

The bacterial strains used in this work were Bacillus subtilis DB104 (Yang,

Ferrari et al. 1984) and Bacillus subtilis WB600 (Wu, Lee et al. 1991).

2.4 Plasmids

Plasmids pBC-Car (pBC carrying the subtilisin Carlsberg gene) and pBC-Per

(pBC carrying the subtilisin Carlsberg double mutant T59A/L217W) were provided by

the collaborator Henkel AG & Co. KGaA (Düsseldorf, Germany).

Other plasmids used in this work are pHY300PLK (Takara BIO) and pHY300Per

produced in this work.

2.5 Oligonucleotides

The oligonucleotides used in this work are summarized in Table 2.2.

Table 2.2. List of oligonucleotides used in this work.

Name Sequence (5’ →3’) Description

CWBc_fp GCGCCTCGAGCTTAAAAAGACTTCCAGCTC Amplification of CwlC domain

CWBc_rp GCGCTCTAGACCGTCTGGAAACATGGTC Amplification of CwlC domain

pHY300Per_fp GCGCCTCGAGTTGAGCGGCAGCTTCG Amplification of vector with Perhydrolase

pHY300Per_rp GCGCTCTAGAAGCTTGGGCAAAGCGTTTTTCC Amplification of vector with Perhydrolase


24

3. Methods

3.1 Synthesis of 7-(4-aminophenoxy)-3-carboxy coumarin (APCC)

APCC was synthetized as described in literature (Khanna, Chang et al. 1991).

Step 1 - synthesis of 7-hydroxy-3-ethyl-carboxy-coumarin. 8.4 g of 2,4-

dihydroxybenzaldehyde was dissolved in 45 ml of anhydrous methanol. To the stirred

solution 10.5 g of diethylmalonate was added and the solution was brought to reflux

temperature (80°C). 0.45 g of morpholine and 0.15 g of acetic acid were added to 2 ml

of methanol and mixed until the precipitate was dissolved. That solution then was added

to a refluxed reaction mixture and refluxing continued during 3 h. After cooling, 200 ml

of methanol was added; at this point reaction product precipitated. The precipitate was

heated again until it was dissolved. The resulting solution was left to precipitate again.

Crystals were recovered by filtering, washed with 10-15 ml of cold methanol and dried.

4.2 g of yellow crystals was obtained. Chemical reaction is presented in Figure 2.2.

Figure 2.2. Synthesis of 7-hydroxy-3-ethyl-carboxy-coumarin.

Step 2 – 7-hydroxy-3-ethyl-carboxy-coumarin salt preparation. 1.5 g of 7-

hydroxy-3-ethyl-carboxy-coumarin was resuspended in 50 ml of toluene (it is not

soluble) and heated to 120°C until about 5 ml of to luene was evaporated, to be sure that

water is removed (about 30-60 min) with stirring. After cooling, 0.3 g of NaH was added

to the mixture (at this point gas was released). When gas releasing decreased (after 20

min), mixing and heating to 120°C continued until t oluene evaporated (about 1-2 h). The

obtained salt was dried in vacuum overnight. 1.7 g of orange salt was obtained.


25

Step 3 – Benzoilation of 7-hydroxy-3-ethyl-carboxy-coumarin salt. 1.7 g of

step 2 product was dissolved in 100 ml of dry dimethyl-formamide (DMF), (DMF was

dried with molecular sieves, 1-2 h). The mixture was heated to 120°C. During heating

(when 50°C was reached) 0.82 g of 1-fluoro-4-nitrob enzene in 3 ml of dry DMF was

added slowly adding the last portion when the temperature reached 120°C. The mixture

(black color) was mixed and heated for 2 h at 120°C and additionally 0.41 g of 1-fluoro-

4-nitrobenzene in 3 ml of DMF was added. After 4-6 h of heating at 120°C, mixture was

left to cool overnight and diluted to 300 ml with ice water (200 ml), a precipitate was

formed. Before ice was melted completely, the precipitate was filtrated washed with cold

water to remove DMF, and dried. After drying, the pellet was extracted twice with 100 ml

of dichloromethane. Water was removed with anhydrous Na-sulphate, dichlormethane

was evaporated using Rotavap and the remaining pellet was dried in vacuum. 0.7 g of

orange compound was obtained. Chemical reaction is presented in Figure 2.3.

Figure 2.3. Synthesis of 7-(4-nitrophenoxy)-3-ethyl-carboxy-coumarin.

0.7 g of product was dissolved in 2 ml of dichloromethane (DCM) and loaded into

a column with 30 g of silica gel. Column was washed with 150 ml of DCM and 7-(4-

nitrophenoxy)-3-ethyl-carboxy-coumarin was eluted using a 20:1 mixture of DCM and

ethyl-acetate. Fractions were analyzed on TLC using a 20:1 mixture of DCM and ethyl-

acetate as developing solution. 7-(4-nitrophenoxy)-3-ethyl-carboxy-coumarin was visible

under UV light at 366 nm. Fractions containing the product were pooled, solvent was

evaporated using Rotavap and product was dried. 0.28 g of white-yellow product was

obtained and structure was confirmed by H-NMR (Appendix, Figure A1).


26

Step 4 – reduction of 7-(4-nitrophenoxy)-3-ethyl-carboxy-coumarin. 0.28 g of

p-nitro-benzyl-ethyl-carboxy-coumarin was resuspended in 19 ml of pure ethanol and 1

ml of distilled water. To this suspension 50 µl of concentrated HCl was added and the

mixture was stirred at 100°C. During heating, a mix ture of 0.6 g of Fe and 0.06 g of

FeCl3x6H2O was added and the solution was stirred during 1h at 100°C. The reduction

was monitored by thin-layer chromatography (TLC) using DCM/ethyl-acetate (20:1) as a

mobile phase and UV for visualization. Product was less soluble in mobile phase than 7-

(4-nitrophenoxy)-3-ethyl-carboxy-coumarin. After cooling, the mixture was diluted with

40 ml of water and extracted four times, each time with 50 ml of diethyl ether. Ether

extracts were dried over anhydrous Na-sulphate, filtered and evaporated to dryness.

0.25 g of yellow amino compound was obtained. Chemical reaction is presented in

Figure 2.4.

Figure 2.4 . Reduction of 7-(4-nitrophenoxy)-3-ethyl-carboxy-coumarin to 7-(4-aminophenoxy)-3-ethyl-carboxy-coumarin.

Step 5 – deesterification of 7-(4-aminophenoxy)-3-ethyl-carboxy-coumarin.

0.25 g of 7-(4-aminophenoxy)-3-ethyl-carboxy-coumarin was dissolved in 30 ml of

dioxane and then 30 ml of 20 % (v/v) sulfuric acid in water was added. Mixture was

stirred at 80°C overnight. The mixture was cooled d own and neutralized using 1M

NaOH to pH 5.0 when the formation of precipitate was visible. Precipitate was filtered,

washed with water and dried. Yield was 140 mg of white-yellow 7-(4-aminophenoxy)-3-

carboxy coumarin (APCC). Structure of the product was confirmed by H NMR

(Appendix, Figure A2). Chemical reaction is presented in Figure 2.5.


27

Figure 2.5. Deesterification reaction of 7-(4-aminophenoxy)-3-ethyl-carboxy-coumarin.

3.2 Sample preparation and analysis of peroxyacetic acid (PAA) and 7-

hydroxycoumarin-3-carboxylic acid (HCC)

Standard solutions and samples were prepared in 96-well microplates

(polystyrene black flat 96-well no. 655076, Greiner Bio-One GmbH, Frickenhausen,

Germany). PAA was diluted to a concentration range from 0.0005 to 1 mM, the

reactions were performed in Na-phosphate buffer (100 mM, pH 7.5) containing 10 µl

PAA standard solution, varied concentration of NaBr (10, 50, 100, 200 or 400 mM) and

varied concentration of APCC (0.05, 0.25 or 0.5 mM) in a final volume of 100 µl. The

PAA calibration curve in presence of hydrogen peroxide contained: 10 µl of PAA

standard solution, NaBr (100 mM), APCC (0.5 mM) and varied concentration of H2O2 (1,

10, 30 or 100 mM). The reaction was measured at 30°C by monitoring the fluorescence

intensity (excitation/emission 382/445 nm, gain 100) using Infinite M100. Reported

values are the average of three measurements with average deviation from the mean

value after 10 min of reaction when fluorescent intensity is constant. The PAA solutions

were used within 1 h to prevent any disproportionate reactions.

Calibration curve in 384-well microplates (polystyrene black flat 384-well no.

781076, Greiner Bio-One GmbH, Frickenhausen, Germany) was performed in Na-

phosphate buffer (100 mM, pH 7.5) containing 4 µl PAA standard solution, 100 mM

NaBr and 0.5 mM APCC in a final volume of 40 µl.


28

Reactions for HCC standard curves were performed in Na-phosphate buffer (0.1

M, pH 7.5) containing 0.05-25 µM HCC, 200 mM NaBr, 100 mM H2O2 and 0.05-25 µM

PAA. The reaction was measured at 30°C by monitorin g the fluorescence intensity

(excitation/emission 382/445 nm, gain 100) using Infinite M100.

3.3 Expression of subtilisin Carlsberg

Expression of subtilisin Carlsberg and subtilisin Carlsberg double mutant

Thr59Ala/Leu217Trp (in further text addressed as Perhydrolase) was in B. subtilis

DB104 strain. Test tubes containing 4 ml of LB media supplemented with tetracycline

(15 µg/ml) were inoculated with a 1:250 dilution of overnight culture (B. subtilis DB104

harboring pBC-Car and pBC-Per) grown in LB media. After 24 hr of expression (250

rpm, 37°C) culture was centrifugated (Eppendorf 581 0R, 4°C, 3220 g, 20 min) and

supernatant of the cells was used for determination of enzymatic activity.

3.4 Determination of enzymatic activity

Before measurement of enzymatic activity, catalase was inhibited in the cell

culture supernatant. Inhibition reaction was performed with 100 mM 3-amino-1,2,4-

triazole and 30 mM H2O2 at RT for 2 hr.

After inhibition supernatant was diluted 5 times and used for determination of

perhydrolytic activity.

The assay was carried out in 96-well microplates (polystyrene black flat 96-well

no. 655076, Greiner Bio-One GmbH, Frickenhausen, Germany). Reactions were

performed in Na-phosphate buffer (100 mM, pH 7.5) containing 50 mM methyl-butyrate,

30 mM H2O2, 100 mM NaBr, 0.5 mM APCC and 10 µl of catalase inhibited cell culture

supernatant (commercial enzyme) in a final volume of 100 µl. The enzymatic activity

was measured at 30°C by monitoring the increase of fluorescence intensity using

microtiter plate reader Infinite M1000.

One unit of enzyme activity is defined as the amount of enzyme that produces 1

µmol of peroxycarboxylic acid per minute. Concentration of produced peroxycarboxylic

acid was determined from calibration curve for peroxyacetic acid.


29

3.5 APCC assay performed in double emulsions

For the preparation of water in oil in water (w/o/w) double emulsions protocol

(Miller, Bernath et al. 2006) was followed. The reaction mixture (W1) was prepared in

100 mM Na-phosphate buffer pH 9.0 by mixing 20 µl of purified Perhydrolase

(0.01mg/ml), 25 µl of 800 mM NaBr, 20 µl 313 mM of H2O2 and 1 µl of 50 mM APCC in

80 µl reaction volume. Samples were vortexed and kept in ice. Primary emulsion was

prepared by mixing 80 µl of reaction mixture and 80 µl of Oil solution (2.9 % (w/w) ABIL

EM90 in light mineral oil). Components were emulsified using Ultra Turax at 10000 rpm

during 5 minutes, on ice. Substrate (methyl-butyrate) was added directly in the primary

emulsion in a final concentration of 50 mM and 250 mM. After addition of methyl-

butyrate, primary emulsion was incubated at 37°C fo r 2 h. Secondary emulsion was

prepared by adding W2 solution (1.5 % (w/v) CMC and 2 % (w/v) Tween 20 in PBS).

Components were emulsified using Ultra Turax at 8000 rpm for 3 minutes, on ice.

Secondary emulsions were diluted 1:100 in 1x PBS and analyzed on CyFlow Space

flow cytometer (Partec). The fluorescence of the product was detected at 375/440 nm.

3.6 Emulsification of Bacillus cells

Cells expressing the Perhydrolase gene (pBC-Per and pHY300Per) and cells

with pHY300PLK vector were grown and expressed as described in Part II, Chapter 3.3.

Bacillus cultures were harvested using a bench centrifuge (Eppendorf centrifuge 5414D,

Eppendorf AG, Hamburg, Germany) at 8000 g for 3 min and rinsed three times in ice-

cold 1x PBS. Approximately 5*108 cells were resuspended in 1 ml Na-phosphate buffer

(100 mM, pH 9.0). Prior to emulsification, catalase is inhibited on the surface of the

cells. Bacillus subtilis cells were incubated in the presence of 100 mM 3-amino-1,2,4-

triazole and 60 mM H2O2 for 2 h at 800 rpm. Emulsified reaction mixture contained 50 µl

of cells (5*108 cells/ml), 20 µl of 600 mM H2O2, 10 µl of 800 mM NaBr and 1 µl of 50 mM

APCC. Reaction mixture is emulsified according to the protocol described in Part II,

Chapter 3.5.


30

3.7 Fusion of CWBc domain to Perhydrolase

Cloning of Perhydrolase in pHY300PLK is described in Part III, Chapter 3.1.

All gene cloning and DNA manipulation steps were carried out according to

standard molecular cloning protocols (Sambrook and Russell 2001). Preparation of E.

coli competent cells and transformation were carried out as standard heat shock

transformation (Inoue, Nojima et al. 1990).

To amplify the cell wall-binding domain region of the B. subtilis peptidoglycan

hydrolase gene CWBc from genomic DNA, the primers CWBc_fp (5’-gcgc-

CTCGAGCTTAAAAAGACTTCCAGCTC-3’, the underlined indicates XhoI restriction

site, the 3’ end corresponds to the position 545 bp downstream of the CwlC start codon)

and CWBc_rp (5’-gcgc-TCTAGACCGTCTGGAAACATGGTC-3’, the underlined

indicates XbaI site, the 3’ end corresponds to the position 856 downstream of the cwlC

start codon) were used. Perhydrolase with the vector backbone was amplified with

primers pHY300Per_fp (5’-gcgcCTCGAGTTGAGCGGCAGCTTCG-3’, underlined

indicates XhoI restriction site) and pHY300Per_rp (5’-gcgcTCTAGAAGCTTGGGCAAA-

GCGTTTTTCC-3’, underlined indicates XbaI restriction site). Amplified fragments were

digested with restriction enzymes, ligated and transformed in E. coli DH5α. Plasmid was

isolated and transformed in B. subtilis WB600.

3.8 Proteolytic activity

Proteolytic activity in the supernatant of the cells: 5 µl of the supernatant was

added to 95 µl of suc-AAPF-pNA solution (1.1 mM in reaction) in Na-phosphate buffer

(100 mM, pH 9.0). Proteolytic activity for samples was defined as the increase in

absorbance at 405 nm per second across 5 min of reaction.

Proteolytic activity on the surface of the cells: 1 ml of expression cultures (B.

subtilis WB600 pHYPer and B. subtilis WB600 pHYPer_ CWBc), were centrifugated at

8000 rpm for 5 min at RT, supernatant was discarded and cells washed 3 times with

PBS. After washing, cells were re-suspended in 1ml of Na-phosphate buffer (100 mM,

pH 9.0). 5 µl of the re-suspended cells was added to 95 µl of suc-AAPF-pNA solution

(1.1 mM in reaction) in Na-phosphate buffer (100 mM, pH 9.0).


31

4. Results

4.1 APCC assay principle

The main reaction on the assay is the oxidation of sodium bromide to

hypobromite by the presence of peroxycarboxylic acids, Figure 2.6. Subsequently, the

fluorogenic substrate 7-(4-amino-phenyl)-3-carboxy coumarin (APCC) is O-dearylated

by the generated hypobromite, releasing the fluorescent 7-hydroxycoumarin-3-

carboxylic acid (HCC) and the side product iminocyclohexadienon.

Figure 2.6. Schematic representation of APCC assay.

The excitation and emission spectra of fluorescent product 7-hydroxycoumarin-3-

carboxylic acid (HCC) has excitation and emission peaks at 382 and 445 nm,

respectively, Figure 2.7. The same spectrum was obtained by simulating the O-

dearylation reaction by the addition of sodium bromide and 7-(4-aminophenoxy)-3-

carboxy coumarin (APCC) to a peroxyacetic acid solution. Fluorescent intensity of HCC

increases towards alkaline pH with maximum values at pH 9. Because of this

characteristic, HCC is also used as pH indicator (Offenbacher, Wolfbeis et al. 1986).

APCC assay can be used for detection of peroxycarboxylic acids in the pH range 5-9.

Figure 2.7. 3D fluorescence spectra of acid (HCC).

In order to study theof each compound (bromide and APCC) was varied.

4.2 Effect of bromide concentration

Experimental data showed that p

dearylate APCC; and dearylation

where fluorescent product is gener

bromide (10 to 400 mM) with constant concentration of APCC (0.5 mM) in a calibration

curve for peroxyacetic acid was investigated,

bromide higher than 50 mM, a dramatic s

reaction was not observed, suggesting that the concentration of bromide is in excess in

this reaction setup. Hundred mM sodium bromide were finally selected as standard for

the APCC assay and for recording the ca


32

3D fluorescence spectra of the reaction product 7-hydroxycoumarin

In order to study the substrate effect on the sensitivity of the assay (bromide and APCC) was varied.

concentration

Experimental data showed that peroxycarboxylic acids cannot

dearylation only takes place in the presence of hypobromite

where fluorescent product is generated. The influence of different concentration of


acid was investigated, Figure 2.8. With a concentration of

bromide higher than 50 mM, a dramatic shift on the calibration curves after 10 min


Hundred mM sodium bromide were finally selected as standard for

the APCC assay and for recording the calibration curves.


hydroxycoumarin-3-carboxylic

sensitivity of the assay concentration

acids cannot directly O-

presence of hypobromite

The influence of different concentration of


. With a concentration of

hift on the calibration curves after 10 min


Hundred mM sodium bromide were finally selected as standard for

Figure 2.8. Calibration curves for 7.5) with different concentrations of NaBr and 0.5 mM APCC. Concentration range of PAA is 0.05-100 µM. The linearity of each calibration curve

4.3 Effect of APCC concentration

For the evaluation the influence

calibration curve for peroxyacetic

APCC substrate, (0.05-0.5 m

concentration of APCC in aqueous solutions

Figure 2.9. Calibration curves for 7.5) containing different concentrations PAA is 0.05-100 µM.


33

Calibration curves for peroxyacetic acid (PAA) in Na-phosphate buffer (0.1 M, pH 7.5) with different concentrations of NaBr and 0.5 mM APCC. Concentration range of PAA is

100 µM. The linearity of each calibration curve was with R2 > 0.992.

.3 Effect of APCC concentration

evaluation the influence of the secondary substrate

peroxyacetic acid was performed using different concentrations of

.5 mM) in presence of 100 mM bromide. The highest

aqueous solutions due to solubility limitation was 0.5 m

Calibration curves for peroxyacetic acid (PAA) in Na-phosphate buffer (0.1 M, pH ent concentrations of APCC and 100 mM NaBr. Concentration range of


phosphate buffer (0.1 M, pH 7.5) with different concentrations of NaBr and 0.5 mM APCC. Concentration range of PAA is

of the secondary substrate concentration, a

ifferent concentrations of

00 mM bromide. The highest

due to solubility limitation was 0.5 mM.

phosphate buffer (0.1 M, pH

00 mM NaBr. Concentration range of

Using 0.05 mM, APCC fluorescence intensit

assay is decreased to less than 50

4.4 Effect of hypobromit

of HCC

Peroxycarboxylic acids

substrates or products in APCC assay. Due to that

each reagent present in the assay on the fluorescent product

peroxyacetic acid, bromide and APCC,

calibration curve for HCC in the presence of

the reaction present in the same concentration as HCC)

together (forming hypobromite) these compounds

effect on HCC, Figure 2.10.

Figure 2.10. Calibration curves for HCC in NaHCC, ▲ HCC with 200 mM NaBr, 200 mM NaBr and peroxyacetic

Peroxyacetic acid in the 20 molar excess compared to HCC without sodium

bromide had no effect on fl

to 5 molar excess with sodium bromide (100 mM) showed no effect on HCC

fluorescence intensity, while

decreased the fluorescence intensit


34

APCC fluorescence intensity is higher, however sensitivity of the

assay is decreased to less than 50 µM PAA, after which linear range is lost, Figure 2.9

bromite and peroxycarboxylic acid on fluorescence intensity

eroxycarboxylic acids and hypohalites are bleaching agents, both are also

substrates or products in APCC assay. Due to that, it’s important to investigate effect of

the assay on the fluorescent product formed in

acid, bromide and APCC, 7-hydroxycoumarin-3-carboxylic acid

HCC in the presence of 200 mM bromide or peroxyacetic

he same concentration as HCC) showed that

together (forming hypobromite) these compounds don’t have a quenching or bleaching

.

Calibration curves for HCC in Na-phosphate buffer (0.1 M, pH 7.5) cont HCC with 200 mM NaBr, ▼ HCC with 200 mM NaBr and 100 mM H

peroxyacetic acid in the same concentration as HCC.


bromide had no effect on fluorescence intensity. Concentrations of peroxyacetic acid


while further increasing concentration of PAA to 10 molar excess

uorescence intensity of HCC by 30 %, Figure 2.11.


y is higher, however sensitivity of the

er which linear range is lost, Figure 2.9.

acid on fluorescence intensity

and hypohalites are bleaching agents, both are also

it’s important to investigate effect of

formed in the reaction of

carboxylic acid (HCC). A

peroxyacetic acid (in

showed that independently and

a quenching or bleaching

phosphate buffer (0.1 M, pH 7.5) containing ■

HCC with 200 mM NaBr and 100 mM H2O2, ♦ HCC with


peroxyacetic acid up


further increasing concentration of PAA to 10 molar excess


35

Figure 2.11. Fluorescence intensity of 1 µM HCC in the presence of sodium bromide (100 mM) and different molar excess of PAA (0, 1, 5, 10 and 20 µM PAA).

4.5 Sensitivity

The detection limit of the assay is 0.05 µM PAA and the linear range 0.05-100

µM PAA under optimal conditions (100 mM bromide and 0.5 mM APCC), Figure 2.12.

Linear range in 384-well microplates was from 0.25-100 µM PAA. It is also possible to

optimize the assay according to specific needs (i.e. enzymatic reactions) by decreasing

concentration of bromide and APCC or changing the reaction pH with little or no effect

on detection limit and linear range.


36

Figure 2.12. (A) Correlation of fluorescence (Y-axis) to peroxyacetic acid concentration (0.05-100 µM PAA, R2 value = 0.9954) in sodium phosphate buffer (100 mM, pH 7.5; supplemented with sodium bromide (100 mM) and APCC (0.5 mM)). (B) Enlarged Figure 3A to show the linearity of fluorescence to peroxyacetic acid concentration (0.05-2.5 µM PAA, R2 value = 0.9993) which are of high relevance for directed evolution experiments.

4.6 Peroxycarboxylic detection in the presence of hydrogen-peroxide

Peroxycarboxylic acids are mainly used in industry as bleaching agents, due to

their instability there is a need for in situ peroxycarboxylic acid generation. Current

enzymatic methods relay on the production of peroxycarboxylic methods in the

presence of an excess of hydrogen-peroxide. To be able to perform under these

conditions, and in order to be used as screening system for perhydrolytic enzymes,

APCC assay has to be reliable in the presence of hydrogen-peroxide.

Hydrogen-peroxide is a strong oxidizing agent with very similar characteristics to

peroxycarboxylic acids, due to this it is necessary to develop an assay which can detect

peroxycarboxylic acids in presence of excess of hydrogen-peroxide. Surprisingly, the

HCC product formation is reduced when hydrogen peroxide concentration is increased

for all investigated PAA concentrations (0.05

peroxide concentrations (1 mM, 10 mM, 30 mM and 100 mM) a linear trend is observed

between fluorescence and PAA concentration.

fluorescence intensity is also

2.13. Since hydrogen-peroxide doesn’t have a direct effect on the fluorescent intensity

of the product (HCC), it is not clear the reason for

with high excess of hydrogen

to oxidation of bromide to hypobromite by hydrogen

is much slower compared to reaction rate

Results showed that APCC

presence of 10000 times excess of hydrogen

fluorescent intensity.

Figure 2.13. Calibration curves for 7.5) containing different concentrations of Hof each calibration curve was with R

4.7 Detection of enzymatic activity

To determine whether APCC assay can be used for

perhydrolase activity, ester perhydrolysis catalyzed by s

variant subtilisin Carlsberg

Perhydrolase) was investigated


37


for all investigated PAA concentrations (0.05-100 µM). Still, for all four hy


between fluorescence and PAA concentration. In the presence of 100 mM H

fluorescence intensity is also 2.4-fold lower compared to signal without peroxide,

peroxide doesn’t have a direct effect on the fluorescent intensity

it is not clear the reason for decrease in fluorescence intensity

with high excess of hydrogen-peroxide. Also increased background

to oxidation of bromide to hypobromite by hydrogen-peroxide; however

is much slower compared to reaction rate against peroxycarboxylic acids.

ed that APCC assay can detect peroxycarboxylic acids in the

s excess of hydrogen-peroxide without influence

Calibration curves for peroxyacetic acid (PAA) in Na-phosphate buffer (0.1 M, pH 7.5) containing different concentrations of H2O2, 100 mM NaBr and 0.5 mM APCC. Theof each calibration curve was with R2 > 0.991.

.7 Detection of enzymatic activity

To determine whether APCC assay can be used for

ster perhydrolysis catalyzed by subtilisin Carlsberg

subtilisin Carlsberg Thr59Ala/Leu217Trp (in further text addressed as

investigated.



or all four hydrogen


In the presence of 100 mM H2O2

ed to signal without peroxide, Figure

peroxide doesn’t have a direct effect on the fluorescent intensity

decrease in fluorescence intensity

increased background was observed due

however this reaction rate

peroxycarboxylic acids.

assay can detect peroxycarboxylic acids in the

peroxide without influence in the

phosphate buffer (0.1 M, pH

00 mM NaBr and 0.5 mM APCC. The linearity

To determine whether APCC assay can be used for the detection of

ubtilisin Carlsberg and improved

Trp (in further text addressed as

At first, sensitivity of the assay and response to different amount of enzyme was

investigated. Figure 2.14 shows the response of the assay to different

commercial subtilisin Carlsberg p

Figure 2.14. Enzyme catalyzed perhyCarlsberg detected using APCC continuous assaybuffer (0.1M, pH 7.5) containing 50 mM methylmM APCC.

The difference in change of fluorescent signal per minute of reactions with

different enzyme concentration

assay responds at different enzyme concentration.

Although assay responds

the expression media can be changed due to interference of media components.

perhydrolytic activity was followe

expressing protease subtilisin Carlsberg (WT)

molecules from media such are salts by passing sample

column according to protocol provided by man


38


shows the response of the assay to different

commercial subtilisin Carlsberg present in the reaction mixture.

Enzyme catalyzed perhydrolysis of methyl-butyrate by APCC continuous assay. Reaction mixture (40 µl) in Na

buffer (0.1M, pH 7.5) containing 50 mM methyl-butyrate, 30 mM H2O2, 1

ifference in change of fluorescent signal per minute of reactions with

concentration correlates with the enzyme dilution,

different enzyme concentration.

responds to different concentration of pure enzyme, sensitivity in

expression media can be changed due to interference of media components.

perhydrolytic activity was followed in the supernatant of B. subtilis

expressing protease subtilisin Carlsberg (WT) and in the sample after removing small

molecules from media such are salts by passing sample through

according to protocol provided by manufacturer, Figure 2.15



shows the response of the assay to different concentrations of

butyrate by commercial subtilisin

0 µl) in Na-phosphate , 100 mM NaBr and 0.5

ifference in change of fluorescent signal per minute of reactions with

enzyme dilution, which proves that

to different concentration of pure enzyme, sensitivity in

expression media can be changed due to interference of media components. The

B. subtilis cell culture

and in the sample after removing small

through PD-10 desalting

, Figure 2.15.


39

Figure 2.15. Perhydrolytic activity in the supernatant of the cells and in the sample after desalting. Reaction mixture (40 µl) in Na-phosphate buffer (0.1M, pH 7.5) containing 50 mM methyl-butyrate, 30 mM H2O2, 100 mM NaBr and 0.5 mM APCC.

When perhydrolytic activity was measured directly in the supernatant of the cells,

a lag phase was visible in the first 10 minutes of reaction (without increase of

fluorescence intensity) suggesting interferences some of the media components,

probably by reacting with generated peroxycarboxylic acid. When additional

components were removed by exchanging media with buffer lag phase disappeared.

However, in the screening process, desalting every sample would be time consuming,

robust and increase standard deviation. Desalting process was substituted with the

dilution of supernatant which also decreased the influence of interfering compounds.

The optimal dilution, where lag phase disappears and fluorescence intensity is the

highest, was 5-fold. Figure 2.16 shows perhydrolysis of methyl-butyrate catalyzed by

subtilisin Carlsberg and Perhydrolase after catalase inhibition and dilution of the

supernatant. The enzymatic reaction was followed in the supernatant of B. subtilis cell

culture expressing protease obtaining an enzyme activity of 30 mU/ml for subtilisin

Carlsberg and 80 mU/ml for Perhydrolase. Purified Perhydrolase also showed 2.7-fold

higher perhydrolytic activity compared to pure subtilisin Carlsberg under the same

conditions. Perhydrolytic activity ratio between two variants is in agreement with

reported values (Wieland, Polanyi-Bald et al. 2009), which confirms the validity of the

APCC assay.


40

Figure 2.16. Enzyme catalyzed perhydrolysis of methyl-butyrate by subtilisin Carlsberg and Perhydrolase detected using APCC continuous assay. Reaction mixture (100 µl) in Na-phosphate buffer (0.1M, pH 7.5) containing 50 mM methyl-butyrate, 30 mM H2O2, 100 mM NaBr and 0.5 mM APCC.

4.8 Future development – optimization for high-throughput screening

High throughput screening methods developed until now have a throughput

between 103 and 106, meaning that a significant part of the diversity is lost due to

limitation of screening systems. Recently, methods based on flow cytometry and

fluorescent activated cell sorting (FACS) have been developed (Aharoni, Griffiths et al.

2005; Miller, Bernath et al. 2006), these methods can routinely sort >107 clones per

hour. The core of these methods is the activity detection assay, which needs to be

reproducible and adaptable to a flow cytometry-based screening.

The main requirements for such an assay in order to be used in double

emulsions are: (a) the presence of the charge on the detected molecule, since non-

charged molecules will diffuse through lipid layer leading to cross-contamination, and

(b) excitation wavelength has to be in the range of the laser employed by the flow

cytometry device. Since APCC and the product of the reaction (HCC) carry a positive

charge and fluorescent signal of HCC can be detected using our in-house flow

cytometer equipped with a cell sorter, APCC assay was tested for application in flow

cytometry based screening system.

The first approach was

double emulsions using the

water droplets containing the

with the delivery of ester substra

Perhydrolase was tested using

the water phase containing the rest of the reaction mixtur

Figure 2.17. Flow cytometry recordings of a blankcontaining Perhydrolase in double emulsions. FL3droplets. RN1 – gated region with fluorescence droplets. A) 50mM methylmethyl-butyrate was added in the prima

Results presented in Figure 2.17

reaction and droplets with high fluorescent signal.

intensity values were gated

the number of droplets in RN1 region wa

to control, while in the reaction with 250 mM methyl

higher for positive reaction


41

first approach was to detect the perhydrolytic activity of

using the flow cytometer. To prevent cross-contamination between

the reaction mixture, the perhydrolytic reaction

substrate through primary emulsions. Perhydrolytic

using different concentration of methyl-butyrate

water phase containing the rest of the reaction mixture.

Flow cytometry recordings of a blank reaction (without enzyme) and reaction containing Perhydrolase in double emulsions. FL3-DAPI represents blue fluorescence of

gated region with fluorescence droplets. A) 50mM methylbutyrate was added in the primary emulsions and incubated at 37°C for 2h.

Results presented in Figure 2.17 show a clear difference between

reaction and droplets with high fluorescent signal. Droplets with highest fluorescence

values were gated in RN1 region. In the reaction with 50 mM methyl

er of droplets in RN1 region was 7-fold higher for positive

n the reaction with 250 mM methyl-butyrate this number wa

reaction compared to control. According to this data


of purified enzyme in

contamination between

perhydrolytic reaction was started

Perhydrolytic activity of

butyrate, delivered in

reaction (without enzyme) and reaction DAPI represents blue fluorescence of

gated region with fluorescence droplets. A) 50mM methyl-butyrate, B) 250 mM C for 2h.

clear difference between the control

with highest fluorescence

reaction with 50 mM methyl-butyrate,

positive reaction compared

this number was 12-fold

According to this data, we could


42

conclude that APCC assay is suitable for the detection of perhydrolytic activity in double

emulsions.

Since this assay is suitable for flow cytometry screening, the next step is the

development of a whole-cell screening system using Bacillus cells as an expression

system. B. subtilis secretes subtilisin Carlsberg into the extracellular medium, whereas

~1000 molecules are always present at the surface of the cells (Power, Adams et al.

1986). We expected that after expression, washing and emulsification, 1000 molecules

of Perhydrolase are still on the surface of the cells and can catalyze perhydrolytic

reaction to give fluorescent signal strong enough for detection by flow cytometer. Cells

with empty pHY300PLK vector and cells carrying Perhydrolase gene in two different

vectors (pBC-Per and pHY300Per) were emulsified and activity (fluorescent intensity) in

the droplets was analyzed using flow cytometer device, Figure 2.18.

Figure 2.18. Flow cytometer recw/o/w double emulsions. A) B. subtilispHY300Per and C) B. subtilisthe released HCC product of perhydrolytic reaction.

The product of the p

expressing Perhydrolase in

compared to pHY300PLK vector.

2.18C, the number of fluorescent droplets wa

reaction, Figure 2.18A. This difference in signal between positive and negative reaction

is not significant and in the case of

variants a significant difference compared to

However, these results show that

expression host and APCC assay can be used for devel

screening system based on flow cytometry technology for dire

Perhydrolase.

Further optimization

concentration of Perhydrolase in


43

Flow cytometer recordings of fluorescence signals of B. subtilis B. subtilis WB600 harboring pHY300PLK, B)

B. subtilis WB600 harboring pBC-Per. FL3 represents blue fluorescence of sed HCC product of perhydrolytic reaction.

The product of the perhydrolyic reaction was detectable only in

expressing Perhydrolase in pBC vector, which has a higher level of expression

compared to pHY300PLK vector. In the RN1 region of that reaction mixture

number of fluorescent droplets was ~3.5-fold higher than in t

This difference in signal between positive and negative reaction

is not significant and in the case of a library which contains low percentage of positive

difference compared to the negative control cannot be achieved

these results show that an in vivo system which employs

expression host and APCC assay can be used for development of high throughput

screening system based on flow cytometry technology for dire

Further optimization of this system should consider: i)

concentration of Perhydrolase in the reaction droplets by attaching protein molecules for


B. subtilis cells entrapped in

WB600 harboring pHY300PLK, B) B. subtilis harboring Per. FL3 represents blue fluorescence of

only in B. subtilis cells

higher level of expression

at reaction mixture, Figure

fold higher than in the control

This difference in signal between positive and negative reaction

ntains low percentage of positive

cannot be achieved.

system which employs B. subtilis as an

opment of high throughput

screening system based on flow cytometry technology for directed evolution of

i) including a higher

hing protein molecules for


44

the surface of the cells (surface display), ii) higher expression of protease and iii)

constant delivery of hydrogen-peroxide through the emulsions since most of the

peroxide is destroyed by cells protective mechanism (catalases).

4.9 Surface display of protease

The cell wall binding (CWBc) domain fused to the C-terminus of Perhydrolase

should localize the protein on the surface of the Bacillus cells due to non-covalent

interactions between CWBc domain and structures on the cell wall. Increased number of

protease molecules on the Bacillus cells will increase the sensitivity of flow cytometry

screening system for detection of perhydrolytic activity in double emulsions.

CWBc domain was successfully fused to the C-terminus of Perhydrolase gene

and transformed in B. subtilis WB600. If the construct is functional, proteolytic activity at

the surface of the cells should be increased compared to cells which are expressing

Perhydrolase without this domain. In the Table 2.3 is presented activity in the

supernatant and on the surface of the cells after.

Table 2.3. Proteolytic activity of Perhydrolase and Perhydrolase_CWBc constructs in the supernatant and on the surface of the cells.

Expression time (h)

Supernatant (U/ml) Surface of the cells (U/ml)

Perhydrolase Perhydrolase_CWBc Perhydrolase Perhydrolase_CWBc

8 0.035 0.03 0.043 0.043

10 0.283 0.25 0.04 0.038

24 3.125 3.655 0.12 0.1

There was no difference in activity on the surface of the cells of Perhydrolase

and Perhydrolase_CWBc constructs. Similar level of activity was detected in the

supernatant of the cells, suggesting that Perhydrolase cleaves this small domain. SDS-

PAGE analysis of the supernatant confirmed the hypothesis that the small binding


45

domain is cleaved, since the band which corresponds to Perhydrolase without CWBc

domain was detected.

5. Discussion

The first fluorescent based screening system for detection of peroxycarboxylic

acids in microtiter plate was developed as part of a high throughput screening platform

for peroxycarboxylic acid producing enzymes. APCC allows a continuous detection of

peroxycarboxylic acids and relies on the detection of a generated fluorescent coumarin

compound (see Fig. 2.6). Main performance criteria for a high throughput screening

systems are: a low standard deviation, reproducibility, simplicity, high sensitivity and low

interference of media or reaction compounds on enzymatic activity. In order to

determine the sensitivity of the APCC assay, peroxyacetic acid concentrations were

quantified in the presence of varied concentrations of sodium bromide and APCC.

Additionally, the fluorescent detection product, HCC, was “treated” with bleaching

agents (hydrogen peroxide, peroxyacetic acid and hypobromite) to determine its

stability.

A main challenge of peroxycarboxylic acid detection systems for directed enzyme

evolution is the interference of cell lysate/supernatant components with the detection

system. Most peroxycarboxylic acid detection methods require enzyme purification

before peroxycarboxylic acid determination and only a few are downscaled to a

microtiter plate format, Table 2.1 (Pinkernell, Luke et al. 1997) and (Binder and Menger

2000)). The validated APCC screening system for directed perhydrolase evolution is the

first reported continuous assay which enables detection of perhydrolytic activity without

enzyme purification and which has been optimized for medium and high throughput

screening in microtiter plate format.


46

6. Conclusion

APCC assay is a spectrofluorometric assay for quantification of in situ

peroxycarboxylic acid production. This is the first reported continuous assay which

enables detection of perhydrolytic activity without previous purification of enzyme

optimized for medium and high throughput screening. The wide pH range (5-9) and

robustness (sodium bromide concentration, oxidative HCC stability) of the APCC

detection system offers a broad flexibility in terms of assay conditions. This simple,

highly sensitive assay is suitable for finding new or modifying existing enzymes with

perhydrolytic activity (lipases, esterases, proteases, nonheme haloperoxidases) to

develop efficient enzymatic bleaching system for industrial applications. APCC assay

enables detection of perhydrolytic activity in double emulsions, opening the possibility to

develop in vivo (cell host expression) and in vitro (cell free expression) high throughput

screening systems.

Part III: Directed Evolution of Subtilisin Carlsber g

47

Part III: Directed Evolution of Subtilisin Carlsberg for

Improved Perhydrolytic Activity

1. Introduction

1.1 Aim of this work

The aim of this work was to generate and identify variants of subtilisin Carlsberg

with increased perhydrolytic activity and simultaneously decreased proteolytic/hydrolytic

activity. In order to achieve this, random mutagenesis techniques (epPCR, SeSaM) and

rational design approach (Saturation Mutagenesis) were used employing high

throughput screening methods. Improved variants were characterized and structure-

function analysis was performed. Changed specificity from peptides toward ester

substrates and from water to hydrogen-peroxide will allow us to study enzymatic

promiscuity and gain more general knowledge about tailoring subtilisin catalysis in

direction of perhydrolases.

1.2 Proteases

Proteases (also named peptidases or proteinases) are enzymes that hydrolyze

peptide bonds in proteins and polypeptides. Different proteases have its specificity in

“choosing” which peptide bonds hydrolyze depending in neighboring amino acid

residues, but all proteases have in common that they can catalyze the hydrolysis of

ester bonds, since the chemical mechanism of hydrolysis of amide bonds and ester

bonds is similar (Hedstrom 2002).

Proteases are classified into six broad groups: serine, threonine, cysteine,

aspartate, glutamic acid proteases and metalloproteases (Barrett, Rawlings et al. 1998).

The mechanism used to cleave a peptide bond involves a nucleophilic attack to the

carbonyl group of the peptide bond by an amino acid residue (in serine, cysteine and

threonine proteases) or a water molecule (aspartic acid, metallo- and glutamic acid

proteases). This nucleophilic attack is induced by a catalytic triad, where a histidine


48

residue is used to activate serine, cysteine or threonine as a nucleophile. Within each of

these groups, proteases are classified into families of related proteases. Alternatively,

proteases can be classified by the optimal pH in which they are active: acid, neutral and

alkaline proteases and peptide bond specificity: endopeptidase, exopeptidase

(carboxypeptidase and aminopeptidase) and amino acid specific peptidase (Barrett,

Rawlings et al. 1998).

Proteases were the first enzyme class to be discovered because of their

abundance in the digestive system. Additionally, they have a wide range of important

biological activities, as they are also involved in processing other proteins. For example,

the blood clothing cascade and activation of immune system involves serine proteases.

Metalloproteases and cysteine proteases are involved in the self destruction of cells

(apoptosis). Bacteria also secrete proteases to hydrolyze (digest) the peptide bonds in

proteins and therefore break the proteins down into their constituent monomers and use

them as nutrients. A secreted bacterial protease may also act as an exotoxin which

destroys extracellular structures (Fersht 1999).

Proteases find applications in various industrial sectors: medicine, pharmacology

and drug manufacture, hard surface cleaning formulations, contact lens cleaning

formulations, waste treatment, fermentation, animal feed additives, digestive

supplements, food and food processing applications (Kirk, Borchert et al. 2002). Today,

proteases are dominant in total enzyme sales and is expected that their dominance will

increase. Most of these proteases are microbial, since they can be produced in large

quantities in a short time. Microbial alkaline proteases hold the largest part of enzyme

market. These proteases are used in food industry, peptide synthesis, leather industry,

industrial and household waste management, photographic industry, medical usage, silk

degumming and the largest field is detergent industry (Gupta, Beg et al. 2002).

Commercial alkaline proteases are summarized in Table 3.1.


49

Table 3.1. Commercial bacterial alkaline proteases, sources, applications and their industrial suppliers. n.s.- not specified, PE - protein engineered. Table taken from (Gupta, Beg et al. 2002)

Supplier Product trade name Microbial source Application

Novo Nordisk, Denmark

Alcalase Savinase Esperase

Biofeed pro Durazym

Novozyme 471MP Novozyme 243

Nue

B. licheniformis Bacillus sp.

B. lentus B. licheniformis

Bacillus sp. n.s

B. lentus Bacullus sp.

Detergent, silk degumming Detergent, textile

Detergent, food, silk degumming Feed

Detergent Photographic gelatin hydrolysis

Denture cleaners Leather

Genencor Inter-national, USA

Purafact Primatan

B. lentus Bacterial source

Detergent Leather

DSM Gist Brocades, The Netherlands

Subtilisin Maxacal

Maxatase

B. alcalophilus Bacillus sp. Bacillus sp.

Detergent Detergent Detergent

Solvay Enzymes, Germany

Opticlean Optimase Maxapem

HT-proteolytic

Protease

B. alcalophilus B. licheniformis PE variant of Bacillus sp. B. subtilis

B. licheniformis

Detergent Detergent Detergent

Alcohol, baking, brewing, feed,

Food, waste

Amano Enzyme Inc, Japan

Proleather Collagenase

Amano protease S

Bacillus sp. Clostridium sp.

Bacillus sp.

Food Technical

Food

Enzyme Development,

USA

Enzeco alkaline protease

Enzeco alkaline protease-L FG Enzeco high

alkaline protease

B. licheniformis

B. licheniformis

Bacillus sp.

Industrial

Food

Industrial

Nagase Biochemicals,

Japan

Bioprase concentrarte Ps. protease Ps. elastase

Cryst. protease Cryst. protease

Bioprase Bioprase SP-10

B. subtilis

P. aeruginosa P. aeruginosa B. subtilis (K2)

B. subtilis (bioteus) B. subtilis B. subtilis

Cosmetic, Pharmaceuticals

Research Research Research Research

Detergent, cleaning Food

Godo Shusei, Japan

Godo-Bap B. licheniformis Detergent, food

AB Enzymes, Germany

Corolase 7089 B. subtilis Food

Wuxi Synder Bioproducts,

China

Wuxi Bacillus sp. Detergent

Advance Biochemicals,

India

Protosol Bacillus sp. Detergent


50

1.2.1 Serine proteases and subtilisins

Serine proteases are the most thoroughly studied class of proteases, and

perhaps in all of enzymology. Serine endo- and exo-peptidases widespread from

bacteria to mammals and have diverse functions, such as digestive and degradative

processes in blood clothing, immune system, tissue remodeling (Krem and Di Cera

2001).

The catalytic mechanism of serine proteases is based on nucleophilic attack of

the targeted peptide bond by a serine. Side chains of serine, histidine and aspartate

build the catalytic triad common to most serine proteases. Figure 3.1 present catalytic

mechanism of serine proteases on example of chymotripsin-like enzymes. The

deprotonated His 57 acts as a general base to subtract a proton from Ser 195,

enhancing its nucleophilicity as it attacks the electrophilic carbon of the amide or ester

link, forming the oxyanion tetrahedral intermediate. Asp 102 stabilizes the positive

charge on the His. His, as a general acid and base catalyst, stabilizes charges in the

transition state and provides a path for proton transfer, without which reactions would

have difficulty in proceeding. The first intermediate is the Michaelis complex. This is

followed by a tetrahedral intermediate in which a covalent bond is formed between the

substrate carbonyl carbon atom and serine O atom. This transition state is stabilized by

formation of hydrogen bonds with residues in the oxianion hole. A hydrogen atom is

transferred from histidine N atom to the leaving group of the substrate. When this

transfer is finished, the first product of the reaction is released and the acyl-enzyme

intermediate is formed. In the next step water comes as a second substrate, His

accepts a proton from the water and covalent bond between oxygen from water and the

carbon of acyl-enzyme intermediate is formed, this is the second tetrahedral

intermediate. At the end, the second tetrahedral intermediate collapses leading to the

release of the second product of the reaction and the regenerated active site of the

enzyme.


51

Figure 3.1. Reaction mechanism of catalysis of serine proteases. Modified from (Garrett and Grisham 1999).

SSuubbssttrraattee

BBiinnddiinngg ooff

ssuubbssttrraattee

FFoorrmmaattiioonn ooff ccoovvaalleenntt EESS ccoommpplleexx

PPrroottoonn ddoonnaattiioonn bbyy

HHiiss5577

CC--HH bboonndd cclleeaavvaaggee

RReelleeaassee ooff aammiinnoo pprroodduucctt

NNuucclleeoopphhiilliicc aattttaacckk bbyy wwaatteerr

CCoollllaappssee ooff tteettrraahheeddrraall

iinntteerrmmeeddiiaattee

CCaarrbbooxxyyll pprroodduucctt rreelleeaassee


52

Additional amino residues forming the oxyanion hole are involved in the

stabilization of tetrahedral intermediates. The negative oxygen ion of the intermediates

fits in the oxyanion hole, forming hydrogen bonds with the backbone of the residues in

the hole, Figure 3.2. This structure is favored, which means that activation energy of the

reaction is lowered and this binding is responsible for the catalytic efficiency of the

enzyme.

Figure 3.2. The “oxyanion hole” of chymotrypsin stabilizes the tetrahedral oxyanion transition states of chymotrypsin. Taken from (Garrett and Grisham 1999).

This type of catalysis is called ping-pong, where the first substrate binds to the

enzyme (peptide), then the first product is released (the N-terminus of peptide), the

second substrate binds (water) and at the end the final product is released (C-terminus

of peptide). The proof for this type of catalytic pathway is that three intermediates (the

Michaelis complex, tetrahedral intermediate and acyl-enzyme complex) have been

isolated and characterized by X-ray chystallography (Kraut 1977).

Many distinct families of serine proteases exist; they have been grouped into six

clans (Chemotrypsines, Subtilisin-like, Carboxipeptidase C, Alapeptidase E, Represor

Lexa-like, ATP-dependent serine peptidases) (Barrett and Rawlings 1995). The two

largest clans of serine proteases are the Trypsin-like and Subtilisin-like clans. Subtilisins

are a non-specific, extracellular proteases initially obtained from Bacillus subtilis. They

are physically and chemically well characterized enzymes. The catalytic triad of

Gly 193

Ser 195

The oxyanion hole

subtilisins consists of aspartic acid,

90 kDa, but most have a size of approximately 27 kDa

Subtilisins are produc

sequence) leads the protein

between the signal sequence and the N

folding of the subtilisin into its

1987). The pro-region also

membrane until it is cleaved off autocatalyt

Today, crystal structures of the most important representatives of this family are

available: Bacillus licheniformis

Bacillus amyloliquefaciens

alkaline protease Savinase

protease PB92 (van der Laan, Teplyakov et al

(Gros, Betzel et al. 1989),

Bacillus TA41 Subtilisin S41

structure of these enzyme

amino acids) has a globular shape and

active site is located in the larger

Figure 3.3. Three-dimensional modelsubtilisin Carlsberg (PDB code 1c3l).


53

subtilisins consists of aspartic acid, histidine and serine, the size varies fro

ze of approximately 27 kDa (Maurer 2004

Subtilisins are produced as pre-proproteins, where the pre

sequence) leads the protein to the extracellular environment. The p

between the signal sequence and the N-terminal end of mature su

folding of the subtilisin into its enzymatically active conformation (Ikemura, Takagi et al.

also holds the pro-subtilisin molecule associated with the

membrane until it is cleaved off autocatalytically by subtilisin (Egnell and Flock 1992

rystal structures of the most important representatives of this family are

licheniformis subtilisin Carlsberg (Bode, Papamokos et al. 1986

subtilisin BPN’ (Wright, Alden et al. 19

alkaline protease Savinase (Betzel, Klupsch et al. 1992) , Bacillus alcalophilus

van der Laan, Teplyakov et al. 1992), Thermus vulgaris

, Thermus album proteinase K (Betzel, Pal et al. 1988

Subtilisin S41 (Almog, González et al. 2009). The secondary

enzymes is highly conserved. The central core of the protein

has a globular shape and consists of 7 β-sheet and nine

in the larger subdomain next to the central β-sheet, Figure 3.3

dimensional model of subtilisin Carlsberg derived from crystal structure of

subtilisin Carlsberg (PDB code 1c3l). Catalytic triad residues are Asp 32, His 64 and Ser 221.


he size varies from 18 kDa to

Maurer 2004).

proteins, where the pre-sequence (signal

pro-region is located

terminal end of mature subtilisin, guiding the

Ikemura, Takagi et al.

subtilisin molecule associated with the

Egnell and Flock 1992).

rystal structures of the most important representatives of this family are

Bode, Papamokos et al. 1986),

Wright, Alden et al. 1969), Bacilus lentus

Bacillus alcalophilus alkaline

Thermus vulgaris thermitase

Betzel, Pal et al. 1988) and

The secondary and tertiary

entral core of the protein (194

sheet and nine α-helices, the

sheet, Figure 3.3.

derived from crystal structure of atalytic triad residues are Asp 32, His 64 and Ser 221.

Crystal structures of

involved in substrate binding

pockets from P6 to P3’. P is

residues extending from the cleaved bond toward

nomenclature for the substrate amino acid residues

toward COOH-terminus-leaving

corresponding binding sites on the enzyme. E

that the inhibitor binds in a surface channel of the enzyme interacting with

residues from P4 to P2’, Figure 3.4

Figure 3.4. Interaction of oligopeptide substrate with catalytic residues of serine proteasesurface interactions (A) and chemicalacid residues is P, where P1binding site. Part B taken from

Subtilisins are highly stable

a wide substrate specificity which makes them suitab

Additionally, as extracellular enzymes,

expression in Bacillus species enables secretion of enzymes in a short period of time

into the fermentation media

Although members of

continuously in many organisms

used in the detergent industry


54

Crystal structures of the enzyme-inhibitor complex helped to identify residues

involved in substrate binding. More than nine amino acids residues form binding

P is the standard nomenclature for the substrate amino acid

residues extending from the cleaved bond toward NH2-terminus-acyl

nomenclature for the substrate amino acid residues extending from the cleaved bond

leaving-group side, S and S’ is the

corresponding binding sites on the enzyme. Every enzyme-inhibitor complex showed

inhibitor binds in a surface channel of the enzyme interacting with

Figure 3.4, (Perona and Craik 1995).

Interaction of oligopeptide substrate with catalytic residues of serine proteaseface interactions (A) and chemical interactions (B). Nomenclature for the substrate amino

acid residues is P, where P1-P1’ denotes the hydrolyzed bond, S is nomenclature for enzyme aken from (Perona and Craik 1995).

Subtilisins are highly stable in the presence of solvents and detergents

substrate specificity which makes them suitable for industrial application

extracellular enzymes, they are easily separated from the biomass

species enables secretion of enzymes in a short period of time

into the fermentation media (Maurer 2004).

embers of subtilase superfamily of serine proteases are

organisms, subtilisins from Bacillus species are

used in the detergent industry, Table 3.2. Since detergent industry demands high


or complex helped to identify residues

residues form binding

nomenclature for the substrate amino acid

acyl-group side, P’ is

extending from the cleaved bond

the nomenclature for

inhibitor complex showed

inhibitor binds in a surface channel of the enzyme interacting with protein

Interaction of oligopeptide substrate with catalytic residues of serine protease:

Nomenclature for the substrate amino P1’ denotes the hydrolyzed bond, S is nomenclature for enzyme

in the presence of solvents and detergents and have

le for industrial application.

ily separated from the biomass and

species enables secretion of enzymes in a short period of time

proteases are identified

are still the only ones

Since detergent industry demands high


55

amount of these enzymes, subtilisins are among the first produced using recombinant

strains. They were also among the first enzymes which were modified using protein

engineering techniques to satisfy needs of growing detergent industry for enzymes

efficient under different non-natural conditions (high or low temperature, presence of

denaturizing agents). Subtilisin BPN’ is the most engineered subtilisin protease and by

1996 substitutions at almost every position of mature subtilisin BPN’ have been

patented. This enzyme is today used as a model system for protein engineering of other

homologous subtilisins (Maurer 2004).

Table 3.2. Subtilisin variants used in detergent industry. Legend: aExclusive molecules for specific customer, bExlusive molecules for captive use, PE protein engineered, WT wild type. Taken from (Maurer 2004).

Trade mark Producer Origin WT/PE Production strain Synonim

Alcalase Novozymes B. licheniformis WT B. licheniformis subtilisin Carlsberg

FNAa Genencor B. amyloliquefaciens PE B. subtilis

Savinase Novozymes B. clausii WT B. clausii Subtilisin 309

Purafect Genencor B. lentus WT B. subtilis

KAPb Kao B. alkalophilus WT B. alkalophilus

Everlase Novozymes B. clausii PE B. clausii

Purafect OxP

Genencor B. lentus PE B. subtilis

FN4a Genencor B. lentus PE B. subtilis

BLAP Sb Henkel B. lentus PE B. licheniformis

BLAP Xb Henkel B. lentus PE B. licheniformis

Esperase Novozymes B. halodurans WT B. halodurans Subtilisin 147

Kannase Novozymes B. clausii PE B. clausii

Properase Genencor B. alkalophilus PB92 PE B. alkalophilus


56

1.2.2 Protein engineering in subtilisin proteases

Almost every property of subtilisin has been altered by protein engineering

including its catalysis, substrate specificity, stability to oxidative, thermal and alkaline

inactivation and pH/rate profile. In this work, the main focus was on altered substrate

specificity of subtilisin and substrate-assisted catalysis.

Subtilisin BPN’ was, in most of the cases, subject of protein engineering

investigation. Amino acid substitution at positions Glu 156 and Gly 166 which form S1

site (P1 binding site) showed that charged substitutions increase the catalytic efficiency

toward complementary charged P1 substrates and decrease toward similarly charged

P1 substrates (Wells, Powers et al. 1987). Altering the electrostatic potential of the S1

site by introducing or removing Arg, Lys, Glu and Asp at positions 156 and 166 kcat/KM

1000-fold increases toward complementary charged substrates. This enzyme also

shows preference for the substrates with hydrophobic residues at position P1. To

investigate the effect of different hydrophobicity levels of S1 site, position Gly 166 was

substituted with 12 different amino acids. Results showed that an increase in the side-

chain volume at position 166 decrease the size of S1 gap leading to reductions in

kcat/KM toward large amino acids up to 5000-fold. The reason for reduced activity is due

to steric repulsion, which is more dominant than hydrophobic interactions. At the same

time catalytic efficiencies toward small P1 side chains were increased up to 10-fold in

these variants (Estell, Graycar et al. 1986). Results showed that specificity is easily

changed by replacing directly amino acids which are in contact with substrate. This

brought the idea of changing one protease into a related one by protein engineering

(Wells, Cunningham et al. 1987). Subtilisin Carlsberg and subtilisin BPN’ differ by 31 %

in sequence and by factor > 60 in catalytic efficiency toward different substrates.

Despite these differences in sequence and catalytic efficiency only 3 substitutions lie

within 7 Ǻ of the S1 pocket. Those substitutions are at positions 156, 217 and 169;

positions 156 and 217 are within 4 Ǻ of modeled substrate where position 156 is part of

S1 pocket, while position 217 is in the S1’ site. A third residue at position 169 is behind

the S1 pocket, 7 Ǻ from the substrate. In subtilisin BPN’ the corresponding amino acids

are Glu 156, Gly 169 and Tyr 217; these were replaced with analogous of subtilisin

Carlsberg Ser 156, Ala 169 and Leu 217. Substrate specificity of BPN’ triple mutant was


57

tested with seven different substrates that vary in charge, size and hydrophobicity was

similar to wild type of subtilisin Carlsberg.

Subtilisin BPN’ can also function as a peptide ligase. This reaction happens

when a peptide carrying a free amino-terminal group can compete with water for the

nucleophilic attack on the acyl-enzyme intermediate. The substitution of the active site

Ser 221 by Cys improved ligase activity of BPN’ 1000-fold (Nakatsuka, Sasaki et al.

1987). The additional mutation Pro225Ala improved ligase activity additionally 10-fold

(Abrahmsen, Tom et al. 1991). Subtiligase, engineered variant of subtilisin BPN’, was

used to synthesize ribonuclease A and active-site variants of this enzyme by stepwise

ligation of six esterified peptide fragments (each 12 to 30 residues long) at yields

averaging 70 % per ligation (Jackson, Burnier et al. 1994).

Considerable specificity toward substrate residues depends from the interaction

between S4 and P4. The S4 site of subtilisin BPN’ and Bacillus lentus alkaline protease

(BLAP) is formed of two structural elements: residues 100-107 which are part of α-helix

in the small subdomain and residues 125-132 in a surface loop (Perona and Craik

1995). For example, introducing mutations at position Tyr 104 and Ile 107 in subtilisin

BPN’ changed specificity toward residues with large side chains at P4. Mutations

increased the size of the P4 site by replacing Ile 107 with smaller amino acids and to

disrupt the hydrogen bond between Tyr 104 and Ser 130 by introducing Phe at position

104. The largest improvement in specificity for P4-Phe relative to P4-Ala was 200-fold

for the variant which had Ile107Gly substitution (Rheinnecker, Baker et al. 1993).

Engineering of S4 site draw as a conclusion that both steric and hydrophobic effects

determinate the S4 specificity profile.

Substitutions within S1 and S4 site showed that only the local environment of

amino acids contacting substrate should be considered in designing substrate specificity

and polar and hydrophobic interactions enzyme-substrate are important since each of

them generates a different specificity.


58

1.2.3 Subtilisin Carlsberg

Subtilisin Carlsberg is a serine protease produced by B. licheniformis. The size of

mature protein is approximately 27 kDa with a pH optimum 8-10. The original subtilisin

is described in 1954 (Guntelberg and Ottesen 1954) and later called subtilisin

Carlsberg. In 1968 subtilisin Carlsberg was sequenced, characterized and the

relationship with other proteinases determined (Smith, DeLange et al. 1968). Due to its

termostability and broad specificity, this enzyme has become industrially important. In

order to elucidate the evolutionary relationship between the subtilisins and simplify the

production of this enzyme, in 1985 the gene encoding subtilisin Carlsberg was isolated,

sequenced and cloned into Bacillus expression vectors (Jacobs, Eliasson et al. 1985).

Subtilisin Carlsberg consists of N-terminal signal peptide including a 29 residues

followed by a 76 residue highly charged pro-region and 274 residues of mature protein.

The signal sequence has similar properties to other signal sequences from Gram

positive bacteria, which includes a basic N-terminal segment followed by a stretch of

uncharged residues and cleavage site most frequently follows the (-3,-1) rule. Most

probably that cleavage of signal sequence takes place after the residues Ala-Ser-Ala.

Homology plot analysis revealed that C-terminus of the pro region is conserved, which

means that there is an evolutionary pressure to keep the structure of this region since is

most probably recognized by a proteolytic enzyme during maturation. For example, the

mature enzyme subtilisin Carlsberg is 70% homologous to BPN’ with identical catalytic

triad, while pro regions are less than 50 % homologous. Still, the secondary structure of

the whole pro region is very similar leading to conclusion that most probably tertiary

structure is also the same (Jacobs, Eliasson et al. 1985).

Maturation and release of subtilisins in an extracellular environment is an

autocatalytic process. Experimental data also confirmed this theory, when inactive

variants of B. amyloliquefaciens subtilisin were expressed in Bacillus protease deficient

strains, only preproenzyme was detected in the cell membrane, whereas the mature

form of the enzyme was absent in the media. Maturation can be carried out with

addition of active subtilisin during fermentation or co-expression together with the

inactive form (Power, Adams et al. 1986). However, when subtilisin Carlsberg variant

which has a deletion of amino acids 97-101 in the mature part of the enzyme was


59

expressed in B. subtilis DB104 (protease deficient strain), the processing and secretion

of enzyme was blocked. Only the pre-proenzyme (molecular weight on SDS gel

corresponds to 42 kDa) was detected in the membrane fraction. When the deletion

variant was expressed in the protease-proficient Bacillus strain only slight increase in

proteolytic activity was detected which can be due to increased expression of the host

strain own proteases (Schulein, Kreft et al. 1991). Experimental data also showed that

the autocatalytic cleavage of the pro sequence and its recognition is independent of the

amino acid sequence around that site, suggesting that the cleavage specificity most

likely depends on the secondary structure of that region (Egnell and Flock 1992).

As of today, all the amino acid positions of subtilisin Carlsberg have been

modified either by site-directed mutagenesis based on rational design or by various

methods of random mutagenesis.

1.2.4 Promiscous activity of subtilisin Carlsberg and Thr59Ala/Leu217Trp variant

It is observed that subtilisin Carlsberg, besides proteolytic (hydrolytic) activity,

also shows perhydrolytic activity. As a result of random mutagenesis, it has been

discovered that modification of the subtilisin Carlsberg by point mutations can generate

perhydrolases with the potential to be used in bleaching systems. The variant

Thr59Ala/Leu217Trp (Perhydrolase) has an 8-fold decreased proteolytic activity and 3-

fold increased perhydrolytic activity compared to wild type subtilisin Carlsberg (Wieland,

Polanyi-Bald et al. 2009).

This results and the previously mentioned switching catalysis from esterase to

perhydrolase motivated to study the second tetrahedral intermediate during hydrolytic

and perhydrolytic reaction in order to understand how single mutation shifts activity

toward perhydrolysis (Lee, Vojcic et al. 2010). The initial structure used for Molecular

Dynamics (MD) simulation corresponded to the coordinates of the crystal structure of

wild type subtilisin Carlsberg, while mutations were introduced in silico. Thr59Ala

substitution is far from the catalytic Ser 221 so it was assumed that this mutation

doesn’t influence on activity of Carlsberg. Leu217Trp substitution is close to the active

site and the Trp 217 variant was used as a starting structure for MD simulations.

Results showed that Trp 217 is very important for the stabilization of the second


60

tetrahedral intermediate (TI) of the perhydrolytic reaction. Trp 217 forms a hydrogen

bond with oxygen of the perhydroxyl group from the second TI, Figure 3.5, while in the

case of wild type there is no hydrogen bonding which can stabilize the second TI state

which can explain for lower perhydrolytic activity compared to Perhydrolase.

Figure 3.5. Hydrogen bonding between N-H in Trp217 and the hydroxyl oxygen of the 2nd tetrahedral intermediate (TI) of the hydrolysis (A) and perhydrolysis reaction (B). Taken over from (Lee, Vojcic et al. 2010).

MD simulation studies also showed differences between the hydrogen bonds

formed by the hydrogen attached to Nε of the catalytically important His 64 and the

second TI. In case of the hydrolysis reaction, this hydrogen interacts with the oxygen of

Ser 221 and any competing hydrogen bond may interfere with the catalysis. In the case

of wild type there were no interfering hydrogen bonds, but in case of the Perhydrolase

there was interfering hydrogen bonds which can be explanation for reduced hydrolytic

activity. During the perhydrolytic reaction, a similar effect was observed, just this time

the interfering hydrogen bond was present in a higher extent for wild type which could

explain the lower perhydrolytic activity compared to the mutant, Figure 3.6.


61

Figure 3.6. Hydrogen bonding between His64 and oxygen atoms of the 2nd TI state of hydrolysis (A) and perhydrolysis (B). Hydrogen bonds are indicated with double arrows, where with dashed arrows are represented interfering hydrogen bonds. Taken over from (Lee, Vojcic et al. 2010).

We used Perhydrolase as a starting point for further directed evolution

experiments in order to increase perhydrolytic activity of subtilisin Carlsberg.

Table 3.3 shows the ratio of hydrolytic/perhydrolytic activity of enzymes present

in nature or engineered.

The values for the ratio of hydrolytic/perhydrolytic activity in the Table 3.3 are not

the representing the real ratio of activities since in most of the cases characterization of

hydrolytic and perhydrolytic activity was done with different substrates. For example,

hydrolytic activity of subtilisin Carlsberg and it’s variant Thr59Ala/Leu217Trp was

determined with tetrapeptide suc-AAPF-pNA as a substrate, while perhydrolytic activity

was determined with ester methyl-butyrate.


62

Table 3.3. Perhydrolytic enzymes and ratio of their hydrolytic and perhydrolytic activites

Enzyme source Hydrolytic/perhydrolytic

activity Reference

P. fluorescens esterase mutant L29P

0.02 (Bernhardt, Hult et al. 2005)

T. maritime CE-7 0.2 (Dicosimo, Gavagan et al.

2008)

T. lettingae CE-7 0.3 (Dicosimo, Gavagan et al.

2008)

T. neapolitana CE-7 1.0 (Dicosimo, Gavagan et al.

2008)

C. parapsilosis polypeptide 1.5 (Dubreucq, Weiss et al. 2007)

P. mendocina lipase mutant Ser207

2.0 (Poulouse 1994)

P. putida lipase mutant Arg 127 2.0 (Poulose and Anderson

1992)

B. subtilis 31954 CE-7 4.5 (Dicosimo, Gavagan et al.

2008)

B. licheniformis subtilisin Carlsberg Thr59Ala/Leu217Trp

7.3 (Wieland, Polanyi-Bald et al.

2009)


63

2. Materials

2.1 Chemicals

All chemicals used were of analytical reagent grade or higher quality and

purchased from Sigma-Aldrich (Taufkirchen, Germany), Applichem (Darmstadt,

Germany) or Carl Roth (Karlsruhe, Germany). All enzymes were purchased from New

England Biolabs (Frankfurt, Germany) or Fermentas GmbH (St. Leon-Rot, Germany).

The substrate for detection of proteolytic activity, Suc-AAPF-pNA, was purchased from

Sigma-Aldrich (Taufkirchen, Germany) and Bachem (Bubendorf, Switzerland). The

protease inhibitor PMSF was purchased from Sigma-Aldrich. The fluorogenic substrate

APCC was synthesized according to protocol described in Part II, Chapter 3.1.

2.2 Bacterial strains

The bacterial strains used in this work are listed in Table 3.4.

Table 3.4. Bacterial strains used in this work.

Strain Description References

E. coli DH5α

F'/endA1 hsdR17(rK-mK+) supE44

thi-1 recA1 gyrA (Nalr) relA1

D(laclZYA-argF)U169 deoR

(F80dlacD(lacZ)M15)

Strategene

B. subtilis DB104 nprE aprE (Yang, Ferrari et al. 1984)

B. subtilis WB600 nprE nprB aprE epr mpr bpr (Wu, Lee et al. 1991)


64

2.3 Plasmids

The plasmids used and generated during this work are listed in Table 3.5.

Table 3.5. Plasmids used and generated in this work.

Plasmid Description References

pBC-Car pBC carrying subtilisin

Carlsberg Provided by Henkel AG & Co. KGaA, Düsseldorf, Germany

pBC-Per pBC carrying Perhydrolase Provided by Henkel AG & Co. KGaA, Düsseldorf, Germany

pHY300PLK Shuttle vector (E. coli, Bacillus

sp.), Apr, Tetr Takara BIO

pHY300Car pHY300PLK carrying subtilisin

Carlsberg This work

pHY300Per pHY300PLK carrying

Perhydrolase This work

2.4 Oligonucleotides

The oligonucleotides used in this work are summarized in Table 3.6.

Table 3.6. Oligonucleotides used in this work.


Car_BamHI_fp GATGGATCCCCGGGACCTCTTTC Cloning SC and

T59A/L217W variant in pHY300PLK vector

Car_XbaI_rp TAGTCTAGATTATTGAGCGGCAGCTTCG Cloning SC and

T59A/L217W variant in pHY300PLK vector

pHY_shuttle_fp CAGATTTCGTGATGCTTGTCAGG Sequencing primer

pHY300PLK

pHY_shuttle_rp CGTTAAGGGATCAACTTTGGGAG Sequencing primer

pHY300PLK

ep_Per_mature_fp CATGTGGCCCATGCGCTAGCG epPCR library generation


65


pHYPer_pro_fw CGATGGCATTCAGCGATTCCGCTTC epPCR and SeSaM library

ep_Per_rp TGCCCAGCTTCTAGATTATTGAGCGG epPCR library generation

SeFpPer_pro_fp CACACTACCGCACTCCGTCGCGATGGCATTCAGCGATTCCGCTTC SeSaM library generation

pHYPer_rp GATTTCGTGATGCTTGTCAGGGGGC SeSaM library generation

SeRp_Per_rp GTGTGATGGCGTGAGGCAGCGATTTCGTGATGCTTGTCAGGGGGC SeSaM library generation

BioSeSaM_fp [Biotin]CACACTACCGCACTCCGTCG 5’ Biotin primer for SeSaM

BioSeSaM_rp [Biotin]GTGTGATGGCGTGAGGCAGC 5’ Biotin primer for SeSaM

SSM32_Per_fp GTAAAAGTAGCCGTCCTGNNNACAGGAATCCAAGCTTC SSM of Perhydrolase

SSM32_Per_rp GAAGCTTGGATTCCTGTNNNCAGGACGGCTACTTTTAC SSM of Perhydrolase

SSM33_Per_fp AAAGTAGCCGTCCTGGATNNKGGAATCCAAGCTTCTCAT SSM of Perhydrolase

SSM33_Per_rp ATGAGAAGCTTGGATTCCMNNATCCAGGACGGCTACTTT SSM of Perhydrolase

SSM68_Per_fp CGGACACGGCACACATNNKGCCGGTACAGTAGCTG SSM of Perhydrolase

SSM68_Per_rp CAGCTACTGTACCGGCMNNATGTGTGCCGTGTCCG SSM of Perhydrolase

SSM125_Per_fp GGATGTTATCAATATGNNKCTTGGGGGAGCATCAG SSM of Perhydrolase

SSM125_Per_rp CTGATGCTCCCCCAAGMNNCATATTGATAACATCC SSM of Perhydrolase

SSM127_Per_fp TATCAATATGAGCCTTNNKGGAGCATCAGGCTCGA SSM of Perhydrolase

SSM127_Per_rp TCGAGCCTGATGCTCCMNNAAGGCTCATATTGATA SSM of Perhydrolase

SSM155_Per_fp GTAGCTGCAGCAGGGNNNAGCGGATCTTCAGG SSM of Perhydrolase

SSM155_Per_rp CCTGAAGATCCGCTNNNCCCTGCTGCAGCTAC SSM of Perhydrolase

SSM156_Per_fp GTGTTTCCTGAAGATCCNNNGTTCCCTGCTGCAGCTAC SSM of Perhydrolase

SSM156_Per_rp GTAGCTGCAGCAGGGAACNNNGGATCTTCAGGAAACAC SSM of Perhydrolase

SSM166_Per_fp GAAACACGAATACAATTNNKTATCCTGCGAAATACGA SSM of Perhydrolase

SSM166_Per_rp TCGTATTTCGCAGGATAMNNAATTGTATTCGTGTTTC SSM of Perhydrolase

SSM169_Per_fp TACAATTGGCTATCCTNNKAAATACGATTCTGTCA SSM of Perhydrolase

SSM169_Per_rp TGACAGAATCGTATTTMNNAGGATAGCCAATTGTA SSM of Perhydrolase

SSM217_Per_fp GCCATTGACGTTCCGTTNNNTGTTGCATAAGTGTTCG SSM of Perhydrolase

SSM217_Per_rp CGAACACTTATGCAACANNNAACGGAACGTCAATGGC SSM of Perhydrolase

SSM218_Per_fp GAAGCCATTGACGTTCCNNNCCATGTTGCATAAGTG SSM of Perhydrolase

SSM218_Per_rp CACTTATGCAACATGGNNNGGAACGTCAATGGCTTC SSM of Perhydrolase

SSM219_Per_fp GGAGAAGCCATTGACGTNNNGTTCCATGTTGCATAAG SSM of Perhydrolase

SSM219_Per_rp CTTATGCAACATGGAACNNNACGTCAATGGCTTCTCC SSM of Perhydrolase

SSM222_Per_fp CATGGAACGGAACGTCANNKGCTTCTCCTCATGTAGC SSM of Perhydrolase

SSM222_Per_rp GCTACATGAGGAGAAGCMNNTGACGTTCCGTTCCATG SSM of Perhydrolase


66

2.5 Cell culture media and cultivation

Cells were cultivated in Luria Broth (LB) medium supplemented with the

appropriate antibiotics using a shaking incubator (Multitron II, Infors GmbH, Einsbach,

Germany) at 37°C and 250 rpm for 24 h. Antibiotics used for cell culture are listed in

Table 3.7.

LB medium: 1 % (w/v) peptone from casein, 0.5 % (w/v) yeast extract, 1 % (w/v)

NaCl.

1.5 x LB medium buffered: 1.5 % (w/v) peptone from casein, 0.75 % (w/v) yeast

extract, 1 % (w/v) NaCl and 90 mM K-phosphate buffer pH 7.6.

LB skim milk (LBM) medium: LB medium plus 1 % (w/v) skim-milk. LB or LBM

agar plates contain 1.5 % (w/v) agar.

Table 3.7. Antibiotics used for cell culture in this work.

Antibiotic Stock (mg/ml) Solvent Working

concentration (µg/ml)

Ampicilin 100 water 100

Erythromycin 2.5 75 % ethanol 5

Lincomycin 2.5 water 5

Tetracycline 15 50 % ethanol 15


67

3. Methods

3.1 Cloning

All gene cloning and DNA manipulation steps were carried out according to

standard molecular cloning protocols (Sambrook and Russell 2001). Preparation of E.

coli competent cells and transformation were carried out as standard heat shock

transformation (Inoue, Nojima et al. 1990). Subtilisin Carlsberg (Car) and subtilisin

Carlsberg Thr59Ala/Leu217Trp variant (Per) genes along with its pre-pro-sequence

including the Bacillus promoter were cloned into pHY300PLK shuttle vector (Takara Bio

Inc, Shiga, Japan) by using BamHI and XbaI restriction sites. The generated construct

was named pHY300Car and pHY300Per (6221 bp). Plasmids constructs were verified

by PCR, restriction enzyme digestion and sequence analysis. Sequence results were

analyzed by Vector NTI 9 (Invitrogen). Plasmid isolation, gel purification and PCR

purification kit were purchased from Macherey-Nagel (Düren, Germany).

3.2 Transformation of B. subtilis WB600 and DB104

For preparation of Bacillus competent cells and transformation two protocols

were used: modified (Matsuno, Ano et al. 1992) and protocol developed during work on

this project (Vojcic 2012).

Solutions for preparation of B. subtilis competent cells: T base (150 mM

ammonium sulfate, 800 mM potassium hydrogen phosphate (K2HPO4), 440 mM

potassium dihydrogen phosphate (KH2PO4) and 35 mM Na citrate); SpC media (10 ml T

base, 1 ml 50 % glucose, 1.5 ml 1.2 % MgSO4, 2.0 ml 10% yeast extract, 2.5 ml 1%

casamino acids in final volume 100 ml) and SpII media (10 ml T base, 1 ml 50 %

glucose, 7.0 ml 1.2 % MgSO4, 1.0 ml 10% yeast extract, 1.0 ml 1% casamino acids, 0.5

ml 100 mM CaCl2 in final volume 100 ml).

Freshly streaked cells were inoculated in 10 ml of SpC media overnight at 37°C,

250 rpm. Overnight culture was diluted in SpC media until OD600 was adjusted to 0.5.

Culture is grown for 3 hours 10 minutes in flask shaker (37°C, 200 rpm). Cell cultures

are diluted 1:1 with starvation medium (essential amino acid was added in final


68

concentration 10 µg/ml), and then incubated again for 2 h in flask shaker (37°C, 300

rpm). After incubation, cultures are maximally competent for about an hour.

Transformation of Bacillus competent cells: 4 µl of plasmid (DNA concentration:

>20 ng/ul) was mixed with 500 µl of competent cells in 2 ml Eppendorf tube, incubated

for 30 min at 37°C, 200 rpm. 300 µl of LB was added for the recovery of competent

cells. Culture was incubated for another 30 min at 37°C, 200 rpm. 250 µl of cells were

plated on LB agar plates containing selective antibiotic.

3.3 epPCR library generation

The random mutagenesis library was constructed using standard epPCR

(Cadwell and Joyce 1994) and megaprimer PCR of whole plasmid (MEGAWHOP)

(Miyazaki 2003) as described. The epPCR library was generated by PCR (95°C for 1

min, one cycle; 95°C for 30 s, 60°C for 30 s, 72°C for 75 s, 30 cycles; 72°C for 5 min,

one cycle) using dNTPs mix (0.2 mM), primers (ep_Per_mature_fp: 5´-

CATGTGGCCCATGCGCTAGCG-3´; ep_Per_rp: 5´-TGCCCAGCTTCTAGATTATTGA-

GCGG-3´; 0.5 µM each), plasmid DNA template (pHY300Per, 5 ng/50 µL), Taq

polymerase (5 U/50 µL) and MnCl2 (0.1 mM). epPCR library for Perhydrolase gene with

pro-sequence was generated according to same protocol, using ep_proPer_fp primer

(5’- CGA TGG CAT TCA GCG ATT CCG CTT C -3’). The MEGAWHOP PCR (72°C for

5 min, one cycle; 98°C for 1 min, one cycle; 98°C f or 30 s, 55°C for 45 s, 72°C for 3

min, 24 cycles; 72°C for 10 min, one cycle) was per formed in a final volume of 50 µL

using dNTPs mix (0.2 mM), megaprimer (purified epPCR amplification product, 500 ng),

plasmid DNA template (pHY300Per, 60 ng) and PfuS polymerase (1.5 U). The amplified

MEGAWHOP product was digested by DpnI (20 U) at 37°C overnight, purified using

Macherey-Nagel kit and transformed into E. coli strains. Plasmids were then pooled and

isolated using the Qiagen kit.


69

3.4 SeSaM library generation

SeSaM library of Perhydrolase was generated based on the advanced protocol

(Mundhada, Marienhagen et al. 2011) which was modified to attain tunable mutational

load and transition/transversion bias.

Preliminary step: Generation of forward and reverse step 1 template

For the template-generation of of SeSaM-TvTv step 1-forward DNA-template, a

50 µl PCR-mixture contained: 1x Phusion HF Buffer; (New England Biolabs); 0.2 mM of

each dNTP; 12.5 pmol of each primer (SeFpPer_pro_fp and pHYPer_rp); 2.5 U of

Phusion polymerase (New England Biolabs) and 50 ng of pHY300Per plasmid as

template. PCR protocol: 98°C for 40 s (1 cycle); 98 °C for 10 s; 64°C for 30 s; 72°C for

40 s (20 cycles); 72°C for 3 min (1 cycle). The sam e PCR setup and protocol was

applied for the generation of the SeSaM-TvTv reverse template, only different primers

(pHYPer_pro_fw and SeRp_Per_rp) were used. After PCR, reactions were column-

purified using the Nucleospin kit from Macherey-Nagel to obtain pure PCR products.

Step 1: Generating a ssDNA fragment pool with random length distribution

The ‘G-Forward’ and ‘-Reverse’ library was constructed by linear amplification of

the pro-Perhydrolase gene in the presence of 0.07 mM of dGTPαS (35 %), 0.130 mM of

dGTP and 0.2 mM each of dATP, dTTP and dCTP. For forward library apart from the

nucleotides, 50 µl PCR mix (6 x 50 µl) contained: 1x buffer (Qiagen PCR), 20 pmol

primer BioSeSaM_fp, 100 ng of the step 1 forward template and 2.5 U Taq Polymerase.

Reverse ‘G’ library (6 x 50 µl) on the other hand contained 100 ng of reverse step 1

template and 20 pmol of primer BioSeSaM_rp, 1x buffer (Qiagen PCR) and 2.5 U Taq

Polymerase. On the similar lines, ‘A-Forward’ library 50 µl (6 x 50 µl) contained 0.07 mM

of dATPαS (35 %), 0.13 mM of dATP and 0.2 mM each of dGTP, dTTP and dCTP; 1x

buffer (Qiagen PCR), 20 pmol primer BioSeSaM_rp, 100 ng of the step 1 forward

template and 2.5 U Taq Polymerase. Whereas 50 µl ‘A-Reverse’ (6 x 50 µl) constituted

0.07 mM of dATPαS (35 %) and 0.130 mM of dATP; 0.2 mM each of dGTP, dTTP and

dCTP; 1x buffer (Qiagen PCR), 20 pmol primer BioSeSaM_rp, 100 ng of the step 1


70

reverse template and 2.5 U Taq Polymerase. Common PCR protocol: 94°C for 2 min (1

cycle); 94°C for 30 s; 60°C for 30 s; 72°C for 35 s (25 cycles); 72°C for 3 min (1 cycle).

300 µl of PCR product was then column purified by Nucleospin kit and was eluted in 80

µl elution buffer. 10 µl of 10x Qiagen Taq buffer was added to provide the alkaline

condition followed by addition of 10 µl of cleavage solution (20 mM solution of Iodine in

99 % ethanol). The cleavage reaction was vortexed immediately after setup, and

incubated at 70°C for 1 h. Biotinylated-DNA fragmen ts were subsequently isolated

using, magnetic streptavidin beads. For better yield of single stranded DNA template

NTC buffer was used for binding on Nucleospin columns for all following steps of

SeSaM-TvTv.

Step 2: Enzymatic addition of dPTP αS to the DNA fragment pool

300 ng (2 x 150) each of ‘G’ Forward, ‘G’ reverse, ‘A’ ‘Forward’ and ‘A’ ‘Reverse’

step 1 DNA fragments were subjected for ‘tailing’ reaction with TdT. Fragments were

elongated at the 3´-OH groups by TdT-catalyzed incorporation of dPTPαS. The reaction

mixture contained: 1x NEB buffer 4 (New England Biolabs); 150 ng DNA fragments; 40

U TdT; 60 pmol dPTPαS and 0.05 mM CoCl2 in a total reaction volume of 25 µl (2 x 25

µl). After incubation at 37°C for 2 h and heat inac tivation for 20 min at 65°C, DNA

fragments were subsequently purified with the Nucleospin Kit.

Step 3: Full length gene synthesis

Full-length of pro-Perhydrolase genes were generated by combining ‘tailed’

‘forward’ and ‘reverse’ library fragments. The reaction mixture contained: 150 ng G and

A Forward and 150 ng of G and A Reverse step 2 fragments 0.25 mM of each dNTP; 1x

Super Taq buffer (MoBiTec) and 10 U 3D1 Polymerase in 100 µl (2 x 50 µl). PCR

Protocol: 94°C for 2 min (1 cycle); 94°C for 30 s; 51.5°C for 1 min; 72°C for 1 min (29

cycles); 72°C for 5 min (1 cycle). Fragments were s ubsequently purified using the

Nucleospin kit.


71

Step 4: Degenerate base replacement

The universal bases were replaced by standard nucleotides by PCR-amplification

using primers pHYPer_pro_fw and pHYPer_rp. 100 µl of reaction contained 12.5 pmol

each of primer pHYPer_pro_fw and pHYPer_rp; 1x Qiagen Taq buffer; 5 U Taq

polymerase (Qiagen); 0.2 mM of each dNTP and 20 ng of purified step 3 PCR product.

PCR program: 94°C for 2 min (1 cycle); 94°C for 30 s; 60°C for 30 s; 72°C for 40 s (15

cycles); 72°C for 3 min (1 cycle). The resulting PC R product representing the final

mutant library was purified with the Nucleospin kit and cloned by MEGAWHOP in E. coli

cells.

3.5 Site Saturation Mutagenesis of Perhydrolase

Site Saturation Mutagenesis (SSM) of Perhydrolase was performed according to

the published method (Wang and Malcolm 2002). The oligonucleotides for SSMs of

Perhydrolase were listed as in Table 3.5. The procedure consists of two stages: In

stage one, two extension reactions are performed in separate tubes; one containing the

forward primer and the other containing the reverse primer. Subsequently, both

reactions were mixed and the stage two was carried out. For the mutagenic PCR (First

stage: 98°C for 30 s, one cycle; 98°C for 10 s, 55° C for 30 s, 72°C for 3 min, four

cycles. Second stage: 98°C for 30 s, one cycle; 98° C for 10 s, 55°C for 30 s, 72°C for 3

min, 14 cycles; 72°C for 3 min, one cycle), PfuS DNA polymerase (1 U), dNTP mix (0.2

mM), each primer (0.2 µM) together with plasmid template (10 ng) were used in 50 µL

reaction volume. Following the PCR, DpnI (20 U) was added, and the mixture was

incubated at 37°C overnight, purified using Machere y-Nagel kit, transformed into E. coli

strains, clones were pooled and plasmid isolated using Macherey-Nagel kit.

3.6 Protein expression

Plasmid library was transformed into B. subtilis WB600 and DB104 and plated

into LB agar plates containing 15 µg/ml tetracycline and 1% skim milk. Colonies

generating a clearance halo on skim milk were picked and resuspended in separate

wells of a 96-well flat bottom microtiter plate (Greiner, Frickenhausen, Germany)


72

containing 150 µl of LB medium and 15 µg/ml tetracycline (master plates) and incubated

during 24 h at 37°C, 900 rpm and 70% humidity. Expr ession was carried out by

replicating the master plates into 96-well flat bottom microtiter plates (expression plates)

containing 200 µl of LB medium and 15 µg/ml tetracycline. Expression was carried out

during 24 h at 37°C, 900 rpm and 70% humidity. Afte r expression, the microtiter plates

were centrifuged at 4000 rpm during 20 min at 4°C a nd the supernatant was used for

activity testing.

3.7 Catalase inhibition

Since is known that Bacillus cells express catalase during stationary growth

phase (Naclerio, Baccigalupi et al. 1995) which was experimentally confirmed in this

work, the first step after cell centrifugation was the inhibition of catalase to prevent

degradation of hydrogen-peroxide, which is substrate in the perhydrolytic reaction.

Catalase was inhibited using 100 mM 3-amino-1,2,4-triazole (AT) and 30 mM hydrogen

peroxide at RT, 800 rpm (using microtiter plate shaker) for 2 h (Tephly, Mannering et al.

1961). After inhibition, the supernatant was diluted 5-fold and used for detection of

proteolytic and perhydrolytic activity.

3.8 Screening for decreased proteolytic activity

Screening was performed in 96-well microtiter plate by adding 5 µl catalase

inhibited supernatant from each clone to 95 µl of suc-AAPF-pNA solution (0.22 mM in

reaction) in 100 mM Na-phosphate buffer pH 7.5. Proteolytic activity for samples within

the same microtiter plate was defined as the increase in absorbance at 405 nm per

second across 5 min of reaction.

3.9 Screening for improved perhydrolytic activity

Screening was performed in 384-well plate by adding 4 µl catalase inhibited

supernatant to 36 µl of reaction mixture containing: 4 µl 313 mM hydrogen-peroxide, 5

µl 800 mM NaBr, 10 µl of 200 mM methyl-butyrate in buffer with 0.1 % SDS and 0.4 µl

APCC in DMSO, in 100 mM Na-phosphate buffer pH 7.5. Perhydrolytic activity was


73

defined as increase in fluorescence (ex/em 360/465 nm) per second across 10 min of

reaction at 30°C.

3.10 Purification of serine proteases with affinity chromatography

The purification of proteases form supernatant can be carried out in a one-step

affinity chromatography with bacitracin as biospecific ligand. Bacitracin is a decapeptide

with antibiotic properties produced by Bacillus species. It is a competitive inhibitor for

serine proteases and binds in the active site via a D-glutamic acid. CNBr-activated

Sepharose couples preferentially at ε-lysine groups forming an iso-urea derivative,

provided that amino group on the ligand are in the unprotonated form and therefore

nucleophile. Thus bacitracin can be coupled directly to the support matrix. This method

enables purification of 96 samples at the same time by using 96-well plate column (2 ml,

Zinsser Analytic, Frankfurt, Germany) connected to a vacuum system. The protocol was

provided by the collaborator Henkel AG & Co. KGaA, Düsseldorf, Germany.

After washing CNBr-activated Sepharose 4B (Sigma-Aldrich) in 1 mM HCl,

bacitracin (100mg/g dry sepharose) was added to the soaked gel and incubated in 100

mM NaHCO3 , 500 mM NaCl while rotating (4°C, 14-15 h). The ge l was washed with

100 mM NaHCO3, 500 mM NaCl pH 10 first and then washed with destilled water. To

block remaining active groups of the sepharose, the washed gel was incubated with 1M

ethanolamine pH 8 while rotating (4°C, 2 h). Afterw ards, the gel was washed with 100

mM NH4-acetate, 10 mM CaCl2 pH 6.5 in steps using a membrane (regenerated

cellulose 0.45 µm) while applying vacuum to remove the washing solution. This gel can

be stored at 4oC. The gel was diluted 1:4 with 100 mM NH4-acetate; 10 mM CaCl2 pH

6.5 and 800 µl of this suspension was pipetted into each well of a Microelute plate and

used as Mini Columns (2 ml, Zinsser Analytic, Frankfurt, Germany).

The column matrix was regenerated by washing with 100 mM Na-acetate, 500

mM NaCl pH 4 and 100mM Tris/HCl, 500 mM NaCl pH 8. Afterwards column was

equilibrated with 100 mM NH4-acetate 10 mM CaCl2 pH 6.5 and 300 µl of the

supernatant was applied. Not bound protein was washed with 100 mM NH4-acetate pH

6.5 and 100 mM NH4-acetate, 50 mM NaCl pH 6.5. Protease was eluted with 100 mM


74

NH4-acetate, 1 M NaCl pH 6.5, 25 % isopropanol. The matrix was washed and stored in

100 mM Na-acetate pH 4.

Total protein concentration of the final samples was determined using the Micro

BCA assay according to the protocol provided by manufacturer.

3.11 Protein purification

Pehydrolase and its variants were expressed in B. subtilis DB104 strain at 37°C

for 60 h in 250 ml of LB media. Cell pellet was separated from supernatant by

centrifugation at 8000 g for 30 min at 4°C. The resulting supernatant was filtered using a

#30 glass fiber filter (Schleicher & Schuell Microscience, Dassel, Germany) and again

by using a 0.2 µm Polyethersulfone (PES) membrane filter (Sartorius, Göttingen,

Germany). The supernatant was concentrated to 10 ml using a Millipore ultra-filtration

stirred cell with a 5 kDa cut off regenerated cellulose ultra-filtration membrane (Millipore,

Billerica, MA, USA). The concentrated supernatant was diluted 10 times with 10 mM

HEPES buffer pH 7.0, re-concentrated to 10 ml and loaded into a Toyopearl Super Q

650c (TOSOH) anion exchange chromatography column (25 ml) previously equilibrated

with 10 mM HEPES pH 7.0 buffer. The flow rate and pressure were managed by an

ÄKTA prime system (Amersham, GE Healthcare Europe, Freiburg, Germany) using the

Unicorn software package. The perhydrolase did not bind to this column at pH 7.0 and

was collected on the buffer front fractions. Pulled fractions are concentrated to 2 ml and

diluted 5 times with distilled water to decrease conductivity (< 1.5 mS/cm) and loaded

into a Toyopearl Super SP-650c (TOSOH) cation exchange chromatography column

(20 ml) previously equilibrated with 20 mM HEPES pH 7.0 and Perhydrolase was

binding for negatively charged matrix. Elution was achieved with 4 % 20 mM HEPES,

1 M NaCl pH 7.0. The collected fractions were analyzed by SDS-PAGE.

The optimized protocol was used for purification of all variants. Total protein

concentration was determined using the BCA assay, and purity was assessed by

running Experion Pro260 Analysis Kit (Bio-Rad, Munich, Germany) to confirm validity of

protocol. The fractions with the highest purity were pooled together, concentrated in

0.1 M Na-phosphate buffer pH 7.5 up to volume of 500 µl and stored at -80°C.


75

3.12 Proteolytic activity with natural substrate – skim milk

Proteolytic activity of the variants was determined using skim milk as a substrate.

The reaction was started by adding 10 µl of enzyme (3.66 µM, final 0.18 µM) to 190 µl

of skim milk in 100 mM TRIS/HCl buffer pH 8.6 at 30°C and kinetic the reaction was

followed by the decrease in absorbance at 650 nm across 10 min of reaction. Change in

absorbance per minute (slope of the reaction) was compared between the variants.

3.13 Proteolytic activity – suc-AAPF-pNA

Kinetic constants for proteolytic activity were determined using suc-AAPF-pNA as

substrate. The amount of released product was determined from Lambert-Beer’s law

using extinction coefficient of pNA ε410 = 8480 M-1 cm-1 provided by manufacturer.

The reaction was started by adding 5 µl (0.37 µM, final 18.3 nM, concentration of

M2 variant was 9.2 µM, final 0.46 µM) of enzyme to 95 µl of suc-AAPF-pNA reaction

mix in 100 mM Na-phosphate buffer pH 7.5 preheated at 30°C. The change in

absorbance was followed every 15 s during 5 min of reaction at 30°C. Kinetic

parameters (KM, kcat) were calculated from initial rates (Vmax) at different concentrations

of substrate (the range from 0.05 to 3 mM).

3.14 Perhydrolytic activity of the variants

Specificity of subtilisin Carlsberg and its variants for three different ester

substrates and hydrogen-peroxide was determined through perhydrolytic activity of the

enzyme with these substrates.

Kinetic parameters (KM, kcat) were determined for methyl-propionate, methyl-

butyrate and methyl-pentanoate in presence of 100 mM hydrogen-peroxide. The

reaction was started by adding 10 µl of enzyme (1.50 µM, final 0.15 µM) to 90 µl of

reaction mixture preheated at 30°C. The reaction mi xture contains: 12.5 µl of 0.8 M

NaBr, 10 µl of 1 M H2O2, 1 µl of 0.05 M APCC and ester substrates of different

concentration in 100 mM Na-phosphate buffer pH 7.5. Concentration range of: methyl-

propionate ranged from 10 to 600 mM; methyl butyrate from 10 to 200 mM and methyl-


76

pentanoate from 5 to 90 mM in reaction. Kinetic parameters (KM, kcat) for hydrogen-

peroxide were determined in the presence of 500 mM methyl-propionate. The reaction

started by adding 10 µl of enzyme (1.50 µM, final 0.15 µM) to 90 µl of reaction mixture

preheated at 30°C. Reaction mixture contains: 12.5 µl of 0.8 M NaBr, 1 µl of 0.05 M

APCC and H2O2 in different concentration in 0.1 M Na-phosphate buffer pH 7.5.

Concentration range of H2O2 was from 1 to 60 mM.

The amount of produced peroxycarboxylic acid per minute was calculated from a

peroxyacetic acid calibration curve. The change in fluorescence was followed every 15

s during 5 min of reaction at 30°C. Kinetic parameters (KM, kcat) were calculated from

initial rates (V0) at different concentrations of substrate.

3.15 Esterolytic activity of the variants

Subtilisin Carlsberg catalyzes hydrolysis of methyl esters generating methanol

and carboxylic acid as a product of the reaction. Amount of generated carboxylic acids

was determinate by HPLC.

Esterolytic reaction: 5 µl of enzyme (concentration 1 mg/ml) was added to 195 µl

of methyl-butyrate (reaction concentration 200 mM) in 100 mM Na-phosphate buffer pH

7.5. Reaction was carried at 30°C for 10 min in mic rotiter plate shaker at 650 rpm.

Reaction was stopped by addition of 1 µl of phosphoric acid. Samples were analyzed by

reversed-phase HPLC.

The HPLC instrument was from Beckman Coulter. Data acquisition was

performed using Beckman Coulter software. The column (Macherey-Nagel, Düren,

Germany) was reversed-phase C18 Nucleosil 100-5 (50 x 4.6 mm). The injection volume

for all measurements was 50 µl, column was heated at 50°C and the UV detection

wavelength was 210 nm. The following gradient of acetonitrile and 20 mM KH2PO4 pH

3.0 was selected to achieve separation: 0-0.2 min 25 % acetonitrile, 0.2-5.8 min applied

gradient from 25 to 100 % acetonitrile, 5.8-9.3 min 100 % acetonitrile and 9.3-16 min

25 % acetonitrile. The flow rate of the mobile phase was 1 ml/min. Concentration of

butyric acid was determined from calibration curve generated with standard solutions.


77

Standard solutions contained butyric acid with concentration range from 0.25 to 25 mM

and 200 mM methyl-butyrate, area below peak for butyric acid was calculated.

3.16 Modeling studies

Analyses are based on the crystal structure of subtilisin Carlsberg (PDB code

1uy6) obtained from RCSB Protein Data Bank. Program Yasara Structure

(http://www.yasara.org/) was used for docking studies and generating the images of the

molecular structures.

Amino acid numeration in the thesis is according to numeration in PDB file (PDB

code 1uy6), where all amino acids starting from position 56 are shifted by one position

(missing position 56).

4. Results

In this work Directed Evolution was performed

subtilisin Carlsberg Thr59A

perhydrolytic activity and decreased proteolytic activity.

4.1 Expression system

Carlsberg and Perhydrolase

shuttle vector pHY300PLK obtaining pHY300Car

3.7. The presence of insert in

Constructs were transformed in expression strains

and expression was confirmed on skim milk agar plates by formation of halos

subtilis colonies. Proteolytic acti

detected in liquid culture.

Figure 3.7 . pHY300PLK shuttle vector containing Carlsberg(pHY300Per) genes for expression and dive


78

In this work Directed Evolution was performed to generate and identify a

59Ala/Leu217Trp (Perhydrolase) variant having increased

perhydrolytic activity and decreased proteolytic activity.

Carlsberg and Perhydrolase genes including the promoter

shuttle vector pHY300PLK obtaining pHY300Car and pHY300Per,

resence of insert in the shuttle vector was confirmed by sequencing analysis.

Constructs were transformed in expression strains Bacillus subtilis

expression was confirmed on skim milk agar plates by formation of halos

Proteolytic activity with artificial substrate suc-AAPF

. pHY300PLK shuttle vector containing Carlsberg (pHY300Car)genes for expression and diversity library generation.


to generate and identify a

variant having increased

promoter were cloned into a

respectively, Figure

shuttle vector was confirmed by sequencing analysis.

DB104 and WB600,

expression was confirmed on skim milk agar plates by formation of halos around B.

AAPF-pNA was also

(pHY300Car) and Perhydrolase


79

4.2 Perhydrolytic (APCC) assay optimization for microtiter plate screening

Screening of random mutagenesis libraries for increased perhydrolytic activity of

Perhydrolase was performed using APCC assay in 384-well microtiter plates,

measuring the increase in fluorescence at ex/em 360/465 nm over time at 30°C. Methyl-

butyrate was chosen as substrate, since Perhydrolase showed the highest performance

with this ester whereas background activity (autoperhydrolysis of methyl-butyrate) was

not detectable under screening conditions. Perhydrolytic reaction is presented in Figure

3.8.

Figure 3.8. Enzyme catalyzed perhydrolysis of methyl-butyrate.

Subtilisin Carlsberg is an alkaline protease with pH optimum 8-10, however,

peroxycarboxylic acids are more stable at a lower pH, where the fluorescent product

(HCC) also shows higher fluorescence intensity. The screening was performed at pH

7.5 in 100 mM phosphate buffer. At this pH, the proteolytic activity of Perhydrolase

measured with the artificial substrate suc-AAPF-pNA was 1.5-fold decreased compared

to the activity at pH 9.0. On the other hand, fluorescent intensity generated in

perhydrolytic reaction was 1.5-fold higher at pH 7.5. Standard deviation of APCC assay,

calculated for 94 clones under these conditions, was approximately 14 %, as shown in

Figure 3.9.


80

Figure 3.9. Validation of APCC assay for screening in MTP. The assay was performed in 100 mM Na-phosphate buffer pH 7.5 with 50 mM methyl-butyrate, 31.3 mM H2O2, 100 mM NaBr, 0.9 mM SDS and 0.5 mM APCC in 40 µl reaction volume using supernatant of B. subtilis WB600 cells after catalase inhibition containing Perhydrolase. The coefficient of variation was calculated for the change of relative fluorescence intensity (RFU) per second.

4.3 Diversity generation

Diversity was generated using error prone PCR (epPCR). Different

concentrations of MnCl2 (0.05 to 0.2 mM) were used in epPCR library generation to

adjust mutation frequency at which the ratio of active to inactive clones is 50 %. Only

the mature Perhydrolase was used following described protocol, PCR product was

cloned into pHY300Per using MEGAWHOP and transformed into competent E. coli

DH5α cells, Figure 3.10.


81

Figure 3.10. Library generation and cloning, epPCR product has size 0.8 kb (line 1) and MEGAWHOP product has 6.2 kb size (line 2).

Transformants were pooled together; plasmid DNA was isolated and transformed

into the expression host B. subtilis WB600 cells. Active/inactive ratio was determined

according to proteolytic activity on skim milk agar plates and perhydrolytic activity in

MTP, Table 3.8.

Table 3.8. Percentage of inactive clones (inactive ratio) with different concentration of MnCl2.

MnCl 2 concentration (mM)

Inactive ratio (%) on skim milk Inactive ratio (%) for perhydrolytic activity

0.05 17 24

0.1 35 60

0.15 62 80

0.2 70 85

A concentration of 0.1 mM MnCl2 was chosen for library generation, since

libraries with similar mutational load have often been used in directed evolution

improving enzymes by gradual changes.


82

4.4 Screening

Approximately 2250 B. subtilis WB600 clones were picked into 96-well flat

bottom MTP containing 150 µl of 1.5 x LB buffered medium, after 24 h colonies were

replicated in 96-well flat bottom MTP containing 200 µl of 1.5 x LB buffered medium and

expression was carried out for 24 h, at 37°C, shaki ng 800 rpm and 70 % humidity.

Plates were centrifuged for 20 min at 4000 rpm and 4°C. After centrifugation, catalase

activity was inhibited by adding 2 µl of 2.5 M 3-amino-1,2,4-triazole and 1 µl of 3.13 M

hydrogen-peroxide to 47 µl of supernatant and incubated for 2 h at RT with shaking at

800 rpm. 4 µl of catalase inhibited supernatant from each well was transferred in 384-

well black, flat MTP for activity assay at 30°C.

Figure 3.11 shows activity profile of epPCR library, where in descendent order, is

presented ratio of active clones versus Perhydrolase.

Figure 3.11. MTP screening profile of active clones from error prone library plotted in descendent order. Activity ratio was calculated versus wild type.

After screening, 15 clones which showed activity at least 1.3-fold higher than

Perhydrolase were rescreened and besides perhydrolytic activity; proteolytic activity

was determined. In the rescreening, 4 clones showed all three activities higher than

Perhydrolase; plasmid DNA was isolated from the variants and again transformed in B.

subtilis WB600. After re-transformation activity was at the level of Perhydrolase,

suggesting that increased activity measured in the screening step was due to higher

protein expression. Additionally, one clone showed 3 silent mutations and another


83

showed a change Ser252Asn which is at the surface of the protein and distant from the

active site.

4.5 Diversity generation and screening optimization

Since after screening more than 2000 clones, there was no improved variant

mutational load of the libraries was confirmed by sequencing random clones. SeSaM

library covering the pro- and mature sequence of Perhydrolase and epPCR libraries

were generated following described protocols, resulting PCR products were cloned in

pHY300Per using MEGAWHOP and transformed into E. coli DH5α cells. Mutational

load of SeSaM and 0.15 mM MnCl2 epPCR library was evaluated by sequencing 3

random clones from each library, Table 3.9.

Table 3.9. Sequencing results for 3 clones from Perhydrolase epPCR library with 0.15 mM MnCl2 and 3 clones from Perhydrolase SeSaM library.

Clone Position (prepro -Perhydrolase) Mutation Type

ep library_1

431 A→G Transition 443 A→T Transversion 483 A→C Transversion 564 T→C Transition 613 A→G Transition 642 A→G Transition 857 C→A Transversion 863 G→C Transversion 891 A→G Transition 970 A→G Transition 975 A→G Transition

1012 A→G Transition ep library_2 none none None

ep library_3

422 A→G Transition 546 A→G Transition 798 G→A Transition 870 A→G Transition 929 A→T Transversion

1013 T→C Transition 1090 T→A Transversion

SeSaM library_1 none none None

SeSaM library_2 152 C→T Transition 518 T→A Transversion 841 T→C Transition


84

Clone Position (prepro -Perhydrolase) Mutation Type

SeSaM library_3 345 C→T Transition 843 T→C Transition 948 C→A Transversion

Analysis of sequencing data showed that the average number of mutations per

gene was 6.3 in epPCR library and 2 in SeSaM library.

To rule out expression mutants, proteolytic activity of Perhydrolase variants was

measured in addition to perhydrolytic activity and the perhydrolytic/proteolytic activity

ratio was calculated for each variant. Proteolytic activity was measured using the

tetrapeptide suc-AAPF-pNA. Although skim milk would reflect the real macromolecular

substrate for proteases, it was not used for screening due to inability to detect activity in

cell supernatant. Additionally, at pH 7.5 skim milk has an increased autohydrolysis,

making it difficult to get accurate measurements.

Screening conditions for perhydrolytic activity were the same as described in Part

III, Chapter 4.4. The proteolytic assay was performed using 0.22 mM suc-AAPF-pNA in

Na-phosphate buffer pH 7.5 using 5 µl of catalase inhibited supernatant containing

Perhydrolase. Standard deviation of the ratio perhydrolytic/proteolytic activity calculated

for 94 clones under these conditions was approximately 13 %.

In the second round, 3350 B. subtilis WB600 colonies were re-picked on fresh

LBtet skim milk agar plates. After 16 h incubation at 37°C colonies producing halos on

skim milk were inoculated into 96-well flat bottom MTP containing 150 µl of 1.5 x LB

buffered medium, after 24 h colonies were replicated in 96-well flat bottom MTP

containing 200 µl of 1.5 x LB buffered medium and expression was carried out for 24 h,

at 37°C, shaking 800 rpm and 70 % humidity. Perhydr olytic/proteolytic activity ratio

versus Perhydrolase ratio of active clones in epPCR (0.1 and 0.15 mM MnCl2) library is

summarized in Figure 3.12.


85

Figure 3.12. Ratio perhydrolytic/proteolytic activity versus Perhydrolase ratio of active clones from error prone library plotted in descendent order.

All clones from epPCR and SeSaM libraries with higher ratio of

perhydrolytic/proteolytic activity and clones with just increased perhydrolytic activity

were rescreened. In the rescreening only the clones which had both activities increased

were found, most likely due to higher expression or substitution which leads to improved

both activities. Plasmids were isolated from 6 variants and re-transformed into

competent B. subtilis WB600 cells to avoid variants with increased expression due to

host. Two clones from each variant were re-screened by measuring proteolytic and

perhydrolytic activity while expression level was analyzed by SDS-PAGE. Proteolytic

and perhydrolytic activity of the variants versus activity of Perhydrolase (E1 to E6) and

SDS analysis are presented in Figure 3.13.

Figure 3.13. Perhydrolytic (Perhydrolase activity and expression in means empty vector, Per is for Perhydrolase.for SDS analysis, 250 µl of the supernatant was TCA precipitated and loaded on the gelmature form of Perhydrolase variants

After re-transformation

however, SDS-PAGE analysis showed

variants, while decreased total protein content in the supernatant of E6 variant can be

due to sample handling during TCA precipitation.

revealed that there was no mutation in the Perhydrolase gene, suggesting that

increased expression could be


86

Perhydrolytic (blue)/proteolytic (red) activity of the variants compared to and expression in the supernatants of the cells from selected

means empty vector, Per is for Perhydrolase. Proteins in the supernatant were TCA precipitated SDS analysis, 250 µl of the supernatant was TCA precipitated and loaded on the gel

form of Perhydrolase variants (27 kDa) is labeled in red square.

transformation, four variants (E3, E4, E5 and E6) had increased activity,

analysis showed an increased expression level for E3, E4 and E5


sample handling during TCA precipitation. DNA sequencing of E4 variant

was no mutation in the Perhydrolase gene, suggesting that

could be due to a mutation in the vector backbone during


of the variants compared to

from selected variants. EV Proteins in the supernatant were TCA precipitated

SDS analysis, 250 µl of the supernatant was TCA precipitated and loaded on the gel, and

four variants (E3, E4, E5 and E6) had increased activity,

evel for E3, E4 and E5


DNA sequencing of E4 variant

was no mutation in the Perhydrolase gene, suggesting that the

vector backbone during


87

MEGAWHOP amplification. E4 variant (in further text addressed as pHY300PerE4) was

used as a parent in the next round of diversity generation.

Screening conditions were optimized for B. subtilis DB104 pHY300PerE4 and

standard deviation calculated for perhydrolytic and proteolytic activity. Expression host

was changed from WB600 to DB104 due to higher transformation efficiency of DB104

strain with new method described in Chapter 3.2 of Part III. Expression conditions were

the same as for WB600 and carried out in 96-well microtiter plate for 24 h at 37°C, 900

rpm and 70 % humidity. Supernatant was separated from the biomass by centrifugation

and catalase inhibited, after inhibition supernatant was diluted 5 times to decrease the

influence of interfering compounds from media. The coefficient of variation for

perhydrolytic activity calculated for the change of relative fluorescence intensity (RFU)

per second (slope of the reaction) was 11 %. Standard deviation of proteolytic activity

calculated for the slope of the reaction was 9 % and standard deviation of ratio

perhydrolytic/proteolytic reaction slopes was 11 %, Figure 3.14.

Figure 3.14. MTP screening validation of proteolytic and perhydrolytic activity. On the graph is presented ratio of perhydrolytic/proteolytic reaction slopes plotted in descendent order. Perhydrolytic reaction was carried out in 100 mM Na-phosphate buffer pH 7.5 in the presence of 31.3 mM H2O2, 100 mM methyl-butyrate, 100 mM NaBr, 0.9 mM SDS and 0.5 mM APCC in 40 µl reaction mixture. Proteolytic activity was carried out in 100 mM Na-phosphate buffer pH 7.5 in the presence of 0.22 mM suc-AAPF-pNA in 100 µl reaction mixture.

4.6 Rational design approach

Overall, 6000 clones

improved ratio perhydrolytic/proteolytic activity

obtained. Considering the limitations of medium thr

approach was employed to reduce sequence space and screening efforts. Furthermore

the motivation for this work was to define

pockets (S1, S3 and S4) of

catalytic triad in order to improve perhydrolytic a

are divided into 3 groups: 1) positions dire

155), 2) positions close to the active site (T

Gly 219 and Met 222) and 3) positions forming binding pockets (Gl

166 and Ala 169), Figure 3.15

Figure 3.15. Saturated position in Perhydrolase surrounding active site and forming substrate binding pocket. Active site residuesand Ser 221 – yellow), orange (group 2), purple – amino acids in substrate binding pockets (gr


88

4.6 Rational design approach

6000 clones were screened for improved perhydrolytic activity

ratio perhydrolytic/proteolytic activity and only expression mutant

. Considering the limitations of medium throughput screening system a rational


ion for this work was to define important residues in subs

) of the protein and interaction of the ester substrates with

catalytic triad in order to improve perhydrolytic activity of the protein. Saturated position

are divided into 3 groups: 1) positions directly involved in catalysis (Asp

s close to the active site (Thr 33, Val 68, Ser 125, Trp

222) and 3) positions forming binding pockets (Gl

169), Figure 3.15.

Saturated position in Perhydrolase surrounding active site and forming substrate nding pocket. Active site residues with surface presentation (Asp 32 – red, His

w), orange – oxyanion hole (Asn 155), blue – amino acids around active site amino acids in substrate binding pockets (group 3).


screened for improved perhydrolytic activity or

and only expression mutants were

oughput screening system a rational


important residues in substrate binding

of the ester substrates with

ctivity of the protein. Saturated position

ctly involved in catalysis (Asp 32 and Asn

125, Trp 217, Asn 218,

222) and 3) positions forming binding pockets (Gly 127, Ser 156, Gly

Saturated position in Perhydrolase surrounding active site and forming substrate

red, His 64 – light green amino acids around active site


89

4.6.1 SSM at positions included in catalysis and close to active site

Proteolytic activity of subtilisin Carlsberg is essential for maturation and enzyme

expression in extracellular environment. However, the goal of this project was to

decrease proteolytic (hydrolytic) and increase promiscuous, perhydrolytic activity.

Motivation was to decrease the self-digestion of the enzyme, and instead of ester

hydrolysis to favor the perhydrolytic reaction and production of peroxycarboxylic acids.

Since both reactions are catalyzed by the same active site, it is very likely that residues

involved in the interaction with ester substrate and the peptide chain are the same, as

well as residues involved in stabilization of hydrolytic and perhydrolytic tetrahedral

intermediate. This fact significantly limits diversity, since variants with improved

perhydrolytic activity can be lost due to inability to self-activate and express in the

supernatant.

Saturation of the positions in the catalytic triad (Asp 32) and oxyanion hole

(Asn155) was performed to study relationship maturation-expression and enzyme

processing by addition of active subtilisin Carlsberg in expression media.

Site Saturation libraries were generated in pHY300PerE4 vector and expressed

in B. subtilis DB104, 400 clones was re-picked on fresh LBtet skim milk agar plates. The

oligonucleotides used for SSM are listed in Table 3.5. Active ratio on skim milk was

15 % in both libraries; ~60 active clones and 20 inactive were inoculated in microtiter

plate for expression. Screening results showed that all active variants have ratio of

activities in the range of standard deviation. SDS analysis of the supernatant some of

the inactive variants showed that the mature form of subtilisin Carlsberg was not

present. Maturation carried by extracellular protease was tested; where two inactive

clones from each library were expressed with and without commercial subtilisin

Carlsberg in media. After 24 h of expression perhydrolytic and proteolytic activity was

tested in the supernatant and 100 µl of the supernatant after TCA precipitation was

analyzed by SDS-PAGE, Figure 3.16.

Figure 3.16. SDS analysis of maturation process with addition of subtilisin Carlsberg (1 µg/ml) in expression media. Supernatant fractions without (are shown for: EV - empty vector, P1 and P2 clones, N3 and N4 – Asp 155 negative clones. Mature protein is labeled with

On the SDS-PAGE

with and without subtilisin Carlsberg, except in the case N2, but since whole profile was

different cannot be concluded that activation was present. It is also possible that amount

of mature protein generated by extr

SDS-PAGE. As a conclusion

was not possible or amount and activity of enzyme was below detection limit under

screening conditions.

Motivated with results were substitution close to active site improved ratio of

perhydrolytic/hydrolytic activity by increasing stabilization of tetrahedral intermediate

seven position surrounding act

each position and none of the variants had improved perhydrolytic activity or improved

ratio perhydrolytic/proteolytic activity. Active ratio for each saturation li

summarized in Table 3.10.


90

SDS analysis of maturation process with addition of subtilisin Carlsberg (1 µg/ml) in expression media. Supernatant fractions without (-) and with (+) subtilisin Carlsberg in media

empty vector, P1 and P2 - positive clones, N1 and N2 155 negative clones. Mature protein is labeled with

analysis there was no difference in the supernatant profile



generated by extracellular activation was below detection limit on

As a conclusion, activation with addition of wild type in expression media

or amount and activity of enzyme was below detection limit under

sults were substitution close to active site improved ratio of


seven position surrounding active site were saturated. 400 clones were

one of the variants had improved perhydrolytic activity or improved

ratio perhydrolytic/proteolytic activity. Active ratio for each saturation li


SDS analysis of maturation process with addition of subtilisin Carlsberg (1 µg/ml)

with (+) subtilisin Carlsberg in media ositive clones, N1 and N2 – Asn 32 negative

155 negative clones. Mature protein is labeled with the red box.

o difference in the supernatant profile



s below detection limit on

, activation with addition of wild type in expression media

or amount and activity of enzyme was below detection limit under

sults were substitution close to active site improved ratio of


400 clones were screened for

one of the variants had improved perhydrolytic activity or improved

ratio perhydrolytic/proteolytic activity. Active ratio for each saturation library is


91

Table 3.10. Activity profile of Saturation Mutagenesis libraries on skim milk agar and in microtiter plate.

Position Active ratio for proteolytic activity (%)

Active ratio for perhydrolytic activity (%)

Thr 33 50 18

Val 68 40 14

Ser 125 40 12

Trp 217 95 43

Asn 218 65 41

Gly 219 7 7

Met 222 72 26

Modifications of the active site and substrate binding pocket of subtilisin

Carlsberg lead to a decrease of the proteolytic activity, which also had, as a

consequence, decreased perhydrolytic activity below detection limit.

4.6.2 SSM of positions in the substrate binding pocket

To further investigate amino acid substitutions that could affect the activity of

subtilisin Carlsberg, positions in S1 binding pocket (Ser 156 and Gly 166), which are in

direct contact with the first residue of the peptide substrate (phenylalanine of suc-AAPF-

pNA substrate) and with small substrates such as ester compounds, were saturated.

Although it was reported that position 156 has significant influence in

determination of substrate specificity (Wells, Powers et al. 1987), (Wells, Cunningham

et al. 1987), screening results of Saturation Mutagenesis library showed that out of 364

screened clones 88 % had perhydrolytic activity, suggesting that this position is not

affecting binding of small substrates such as esters.

When SSM was performed at position Gly 166, out of 184 screened clones 93 %

active variants were identified, and 90 clones had increased ratio

perhydrolytic/proteolytic activity. Screening results are summarized in Figure 3.17.


92

Figure 3.17. SSM library profile described by comparison perhydrolytic/proteolytic activity ratio of the variants versus Perhydrolase.

Seven variants with the highest ratio were further analyzed. Variants were re-

screened and expression level of the best variants was analyzed by Experion system,

Figure 3.18. Re-screening data for perhydrolytic and proteolytic activity compared to

Perhydrolase are summarized in Table 3.11.

Table 3.11. Re-screening results (assay conditions for perhydrolytic activity - 31.3 mM H2O2, 100 mM methylbutyrate, 100 mM NaBr, 0.5 mM APCC in 100 mM Na – phosphate buffer pH 7.5 with 0.9 mM SDS; assay conditions for proteolytic activity – 0.22 mM AAPF-pNA in 0.1 M Na-phosphate buffer pH 7.5). M- mutant, P - Perhydrolase

Variant Proteolytic activity M/P

Perhydrolytic activity M/P

Ratio perhydrolytic/proteol

ytic activity M/P

Amino acid substitution

P1B7 0.003 1.3 > 100 Gly→Ile

P1B12 0.152 2.8 18.4 Gly→Leu

P1C2 0.734 1.5 2.0 Gly→Phe

P1C10 0.260 1.6 6.2 Gly→Val

P1D12 0.008 1.2 > 100 Gly→Ile

P1F1 0.006 1.5 > 100 Gly→Ile

P1H2 0.192 3.0 15.6 Gly→Leu

Figure 3.18. Expression analysis of Perhydrolase and G166 variants byµl of supernatant after 24 h of expression was TCA precipitated and loaded on SDS gel. empty vector, 2 – Perhydrolase, 3 Perhydrolase, 9 – P1C2, 10 –

Selected variants were purified using bacitracin affinity chromatography

described in Chapter 3.10, Part II to confirm data

screening conditions. Bacitracin, decapeptide with antibiotic properties, is a competitive

inhibitor for serine proteases and binds in the active site via a D

ligand is coupled for CNBr Sepharose

plate column. 1.5 ml of expression culture

fractions were washed with 100 mM NH

mM NaCl pH 6.5. Protease was eluted from column with 100 mM

NaCl pH 6.5, 25 % isopro

interactions and isopropanol

substrate. Concentration of the variants was in µg/ml range (50

perhydrolytic and proteolytic


93

pression analysis of Perhydrolase and G166 variants by

µl of supernatant after 24 h of expression was TCA precipitated and loaded on SDS gel. Perhydrolase, 3 – P1B7, 4 – P1B12, 5 – P1D12, 6 –

– P1C10. Black arrow indicates mature subtilisin Carlsberg.

variants were purified using bacitracin affinity chromatography

described in Chapter 3.10, Part II to confirm data using purified

Bacitracin, decapeptide with antibiotic properties, is a competitive

inhibitor for serine proteases and binds in the active site via a D

coupled for CNBr Sepharose to prepare an affinity matrix and loaded in 96

1.5 ml of expression culture was applied on the column; non

fractions were washed with 100 mM NH4-acetate pH 6.5 and 100 mM NH

mM NaCl pH 6.5. Protease was eluted from column with 100 mM

isopropanol, where high salt concentration

isopropanol disturbs hydrophobic interactions between protein and

oncentration of the variants was in µg/ml range (50-100 µg/ml).

perhydrolytic and proteolytic activity of the variants is presented in Table 3.12


Experion system. 100 µl of supernatant after 24 h of expression was TCA precipitated and loaded on SDS gel. 1 –

P1F1, 7 – P1H2, 8 – Black arrow indicates mature subtilisin Carlsberg.

variants were purified using bacitracin affinity chromatography

using purified samples under

Bacitracin, decapeptide with antibiotic properties, is a competitive

inhibitor for serine proteases and binds in the active site via a D-glutamic acid. This

affinity matrix and loaded in 96-well

was applied on the column; non-bound

acetate pH 6.5 and 100 mM NH4-acetate, 50

mM NaCl pH 6.5. Protease was eluted from column with 100 mM NH4-acetate, 1 M

affects electrostatic

hydrophobic interactions between protein and

100 µg/ml). Specific

iants is presented in Table 3.12.


94

Table 3.12. Specific activities of the variants. Perhydrolytic activity of the variants in the presence of 100 mM methyl-butyrate and 31.3 mM hydrogen-peroxide and proteolytic activity of the variants with 0.22 mM suc-AAPF-pNA in 100 mM Na-phosphate pH 7.5 buffer.

Variant Perhydrolytic activity (U/mg) Proteolytic activity (U/mg)

Perhydrolase 2.0 60.0

P1B7 2.2 0.9

P1B12 4.1 8.6

P1C2 2.4 31.1

P1C10 2.1 10.5

Results using purified enzyme confirmed that variant P1B7 (Gly166Leu) has the

highest performance when methyl-butyrate was used as substrate, while proteolytic

activity with suc-AAPF-pNA was 7-fold decreased compared to Perhydrolase. However,

the Mutant/Perhydrolase activity ratio was slightly different compared to screening

results, which can be due to process of purification where bacitracin (ligand in affinity

chromatography) could be eluted simultaneously with protease and since it is an

inhibitor it will decrease the measured activity.

Single amino acid substitution in the S1 binding pocket changed the

perhydrolytic/proteolytic ratio more than 100-fold. Further modification of the pocket

included Saturation Mutagenesis at positions Gly 127/Ala 169, and Gly 127/Gly 166. Gly

127 as a part of S3 pocket is opposite to Gly 166 and Ala 169, substituting Gly/Gly and

Gly/Ala with more robust amino acids would additionally decrease the size of S1 pocket

and ‘’push’’ the ester substrate toward the active site. However, screening data showed

that 85 % of the variants were inactive, suggesting that position 127 is important for

autoprocessing of the enzyme. Identified variants with increased perhydrolytic activity

had substitutions only at Gly 166. Beside Gly166Leu and Gly166Ile, an additional

substitution Gly166Tyr was found. This variant showed 2-fold increased perhydrolytic

activity and 3-fold decreased proteolytic activity.

Protein purification on large scale (250 ml expression media) was necessary to

determine kinetic parameters and compare the variants with wild type subtilisin

Carlsberg. Additionally, variants Thr59Ala and Leu217Trp were generated and purified


95

in order to understand the influence of each position on proteolytic and perhydrolytic

activity.

4.7 Purification of selected variants

The calculated isoelectric point (pI) of the parent is 7.1 whereas reported values

are 6.7 (Vitale and Gamulin 1975) and 9.4 (Ottesen and Spector 1960). If pI of subtilisin

Carlsberg is around 7.0, protein at pH 7.8 should bind for anion exchange column,

however results showed that at this pH subtilisin Carlsberg does not bind to the anion or

cation exchange column. Binding to the anion exchange column was achieved at pH

8.6, while at pH 7.0 subtilisin Carlsberg binds to the cation exchange column, indicating

that pI value is around 7.8.

Purification of the variants was optimized at pH 7.0, where in the first step anion

exchange column was used to remove protein impurities from the cell supernatant.

Subtilisin Carlsberg did not bind to this column, while most of the impurities remained in

the column. Protease fractions were pooled together and loaded on the cation

exchange column. The detailed purification protocol is described in Chapter 3.11 of Part

III.

Figure 3.19 shows the purification profile of subtilisin Carlsberg. Panel A shows

the resulting fractions and SDS-PAGE analysis after the concentrated cell culture

supernatant was loaded into the anion exchange column. The blue line on the

chromatogram represents the protein content in each fraction related to the absorbance

of the sample at 280 nm. The black arrow points the fractions exhibiting proteolytic

activity, corresponding to subtilisin Carlsberg. After increasing the ionic strength of the

running buffer, the remaining protein content retained in the column was eluted. The

fractions with proteolytic activity were collected, concentrated and loaded into the cation

exchange column.

Panel B shows the result of the cation exchange chromatography; subtilisin

Carlsberg was retained in the column while other proteins present in the sample passed

through. These fractions did not show proteolytic activity. The retained protein was

eluted by increasing ionic strength and collected as a single protein peak.


96

The proteolytic activity of the eluted fractions was confirmed and the positive

fractions were collected, pooled together, desalted and reconcentrated. The protein

purity was confirmed by the presence of a single band at 28 kDa corresponding to

subtilisin Carlsberg, using the Experion Pro260 Analysis Kit, Figure 3.20. Before

Experion analysis protein content of pure samples was inhibited with PMSF to avoid

degradation during sample preparation. Obtained purity of the variants was above 90 %.

Figure 3.19. Two step ion exchange chromatography purification of subtilisin Carlsberg. (ml of concentrated supernatant was loaded into an anion exchange column (Toy650c), active fractions (3, 4, 5, 6, 7, 8) weexchange chromatography Toyopearl SPconcentrated.


97

Two step ion exchange chromatography purification of subtilisin Carlsberg. (rated supernatant was loaded into an anion exchange column (Toy

e fractions (3, 4, 5, 6, 7, 8) were pooled together, concentrated and loaded on cation exchange chromatography Toyopearl SP-650c (B) fractions 16, 17, 18 and 19 were p


Two step ion exchange chromatography purification of subtilisin Carlsberg. (A) 10 rated supernatant was loaded into an anion exchange column (Toyopearl SuperQ-

re pooled together, concentrated and loaded on cation fractions 16, 17, 18 and 19 were pooled and

Figure 3.20 . Experion analysis of purified variants. 1 3 – Leu217Trp, 4 – Perhydrolase, 5Carlsberg.

Total protein concent

standard for calibration curve. Concentration of the each variant was calculated

according to the Experion analysis respect to the total protein concentration, val

summarized in Table 3.13.

Table 3.13. Calculated specific subtilisin Carlsberg concentration for each variant after purification.

Variant

Subtilisin Carlsberg

Thr59Ala variant

Leu217Trp variant

Perhydrolase

Thr59Ala/Gly166Leu/Leu217Trp (

Thr59Ala/Gly166Ile/Leu217Trp (

Thr59Ala/Gly166Tyr/Leu217Trp (


98

nalysis of purified variants. 1 – subtilisin Carlsberg, 2

Perhydrolase, 5 - M1, 6 - M2, 7 - M3. Black arrow indicates mature subtilisin

concentration was determined using the BCA kit, using BSA as a


according to the Experion analysis respect to the total protein concentration, val

Calculated specific subtilisin Carlsberg concentration for each variant after

Variant Specific concentration (mg/ml)

Subtilisin Carlsberg (wild type) 1.0

Thr59Ala variant 0.6

Leu217Trp variant 0.6

Perhydrolase 0.6

6Leu/Leu217Trp (M1) 1.9

Thr59Ala/Gly166Ile/Leu217Trp (M2) 4.5

Thr59Ala/Gly166Tyr/Leu217Trp (M3) 1.6


subtilisin Carlsberg, 2 - Thr59Ala variant, Black arrow indicates mature subtilisin

BCA kit, using BSA as a


according to the Experion analysis respect to the total protein concentration, values are

Calculated specific subtilisin Carlsberg concentration for each variant after

Specific concentration (mg/ml)

4.8 Characterization of the variants

Kinetic parameters

activity using suc-AAPF-pNA as a substra

different ester substrates (methyl

Variants Thr59Ala and Leu217Trp were

showed that substitution Thr59Ala does no

4.8.1 Characterization with suc

Although suc-AAPF-

most widely used substrate for characterization of proteolytic activity.

kinetic parameters for subtilisin Carlsberg and four variants are summarized in Figure

3.21 and Table 3.14.

Enzyme

Wild type

Perhydrolase

M1

M2

M3

Figure 3.21/Table 3.14. Kinetic parameters of the selected variants for proteolytic activity using suc-AAPF-pNA as a substrate.


99

4.8 Characterization of the variants

inetic parameters of the selected variants were determined for proteolytic

pNA as a substrate and for perhydrolytic activity with three

different ester substrates (methyl-propionate, methyl-butyrate and methyl

Variants Thr59Ala and Leu217Trp were excluded from characterization, since r

substitution Thr59Ala does not have any effect on activity

4.8.1 Characterization with suc-AAPF-pNA as a substrate

-pNA does not represent “real” protease

most widely used substrate for characterization of proteolytic activity.

tic parameters for subtilisin Carlsberg and four variants are summarized in Figure

KM (mM) kcat (min -1) k

0.59 ± 0.06 10353 ± 342

0.15 ± 0.02 1988 ± 43

0.42 ± 0.03 908 ± 21

0.54 ± 0.04 104 ± 2

0.2 ± 0.02 1370 ± 27

Kinetic parameters of the selected variants for proteolytic activity using substrate.


were determined for proteolytic

te and for perhydrolytic activity with three

butyrate and methyl-pentanoate).

excluded from characterization, since results

t have any effect on activity.

protease substrate it is the

most widely used substrate for characterization of proteolytic activity. The calculated

tic parameters for subtilisin Carlsberg and four variants are summarized in Figure

kcat /KM (mM-1min -1*103)

17.5

13.2

2.2

0.2

6.8

Kinetic parameters of the selected variants for proteolytic activity using

As already described, Perhydrolase has

surprisingly KM value is also decreased

Perhydrolase only 1.3-times compared to WT

and M3 variants is lower compared to Perhydrolase which is in acco

screening results. Additionaly, M3 variant has

M1 and M2 variants have increased K

The best variant is M2 with

87.5-fold decreased specificity constant (

4.8.2 Skim milk activity

Proteolytic activity of the variants was tested

the activity towards complex

Figure 3.22. Skim milk activityabsorbance per minute at 650 nm

As expected, difference in activity between

obtained by characterization with suc

decreased proteolytic activity compared to WT and M2 variant

than Perhydrolase and 4.6-


100

described, Perhydrolase has a decreased kcat co

value is also decreased, decreasing specificity constant

times compared to WT. The catalytic efficiency (k

and M3 variants is lower compared to Perhydrolase which is in acco

screening results. Additionaly, M3 variant has similar KM value to

M1 and M2 variants have increased KM, but still slightly lower than in the case of WT.

The best variant is M2 with 99.5-fold decreased catalytic efficiency compared to WT and

fold decreased specificity constant (kcat/KM).

Proteolytic activity of the variants was tested using skim milk i

complex substrate.

Skim milk activity of the selected variants. Activity is represented as decrease in at 650 nm caused by 5 ng/ml of enzyme with 2 % skim milk

As expected, difference in activity between the variants is different from data

rization with suc-AAPF-pNA. Perhydrolase has only 2

eolytic activity compared to WT and M2 variant has 1.7

-lower activity compared to WT, while activity with


compared to WT, but

specificity constant (kcat/KM) of

The catalytic efficiency (kcat) of M1, M2

and M3 variants is lower compared to Perhydrolase which is in accordance with the

Perhydrolase, while

, but still slightly lower than in the case of WT.

cy compared to WT and

skim milk in order to compare

of the selected variants. Activity is represented as decrease in

ng/ml of enzyme with 2 % skim milk at pH 8.6.

variants is different from data

Perhydrolase has only 2.8-fold

has 1.7-fold lower activity

activity with suc-AAPF-


101

pNA is 87.5-fold lower. In this case, M1 is the best variant having 5-times lower activity

than WT.

4.8.3 Perhydrolytic activity: kinetic parameters for hydrogen-peroxide and methyl-

propionate

Amount of produced peroxycarboxylic acid in perhydrolytic reaction was

calculated from a calibration curve for peroxyacetic acid generated in the presence of

hydrogen-peroxide and an ester substrate.The relationship between concentration of

peroxyacetic acid in the reaction mixture and intensity of fluorescent signal is presented

in Figure 3.23.

Figure 3.23. Calibration curve for peroxyacetic acid

Kinetic constants for perhydrolysis of methyl-propionate were obtained by varying

each substrate independently. The fixed methyl-propionate concentration was 500 mM,

which is also the highest solubility of this substrate and fixed concentration of hydrogen-

peroxide was 100 mM. Kinetic constants are presented in Figure 3.23 and Table 3.15.

Enzyme Varied substrate

Wild type Hydrogen

Methyl-propio

Perhydrolase Hydrogen

Methyl-propionate

M1 Hydrogen

Methyl-propionate

M2 Hydrogen

Methyl-propionate

M3 Hydrogen

Methyl-propionate

Figure 3.23/Table 3.15. Kinetic parameters of the propionate.

As expected from the

2.4-fold increased kcat value,

and M3 variants have slightly i

fold decreased KM values for methyl

peroxide to Perhydrolase. The best vari

constant (kcat/KM) compared to

comparison with WT. The values of k

were completely saturated with both constant substrates. The K

propionate was high for all variants so the

did not completely saturate the enzyme.


102

Varied substrate KM (mM) kcat (min -1

Hydrogen-peroxide

propionate

41.0 ± 8.3

478.2 ± 55.7

66.7 ± 7.0

67.8 ± 4.3

Hydrogen-peroxide

propionate

16.4 ± 2.4

566.2 ± 75.3

115.8 ± 6.3

164.6 ± 12.7

Hydrogen-peroxide

propionate

18.8 ± 2.5

265.1 ± 27.1

173.3 ± 9.1

182.4 ± 8.0

Hydrogen-peroxide

propionate

20.9 ± 0.9

255.2 ± 27.7

183.8 ± 3.4

204.0 ± 9.3

Hydrogen-peroxide

propionate

18.7 ± 2.0

349.4 ± 42.8

209.8 ± 8.6

238.4 ± 14.1

Kinetic parameters of the selected variants for perhydrolysis

the reported data (Lee, Vojcic et al. 2010), Perhydrolase has

value, and decreased KM value for hydrogen

and M3 variants have slightly increased kcat values compared to Perhydrolase, but

values for methyl-propionate and similar KM values for hy

. The best variant, M3, shows 1.6-fold increased

compared to Perhydrolase and 7-fold increased

he values of kcat would be the same in both cases, if the enzyme

were completely saturated with both constant substrates. The K

propionate was high for all variants so the methyl-propionate concentration of 500 mM

did not completely saturate the enzyme.


1) kcat/KM (mM-1min -1)

7.0

4.3

1.6

0.14

6.3

12.7

7.1

0.29

9.1

8.0

9.2

0.69

3.4

9.3

8.8

0.80

8.6

14.1

11.2

0.68

iants for perhydrolysis of methyl-

, Perhydrolase has a

ydrogen-peroxide. M1, M2

values compared to Perhydrolase, but ~2-

values for hydrogen-

fold increased specificity

fold increased specificity in

would be the same in both cases, if the enzyme

were completely saturated with both constant substrates. The KM value for methyl-

propionate concentration of 500 mM

4.8.4 Perhydrolytic activity: kinetic parameters for

pentanoate

Since position Gly 166 is part of S1 binding pocket, different substitutions at

position can change enzyme specificity toward different

perhydrolytic activity were determined

which was substrate of choice for screening and already mentioned methyl

variants are also characterized with methyl

and methyl-pentanoate are present

Enzyme Substrate

Wild type Methyl-butyrate

Methyl-pentanoate

Perhydrolase Methyl-butyrate

Methyl-pentanoate

M1 Methyl-butyrate

Methyl-pentanoate

M2 Methyl-butyrate

Methyl-pentanoate

M3 Methyl-butyrate

Methyl-pentanoate

Figure 3.24/Table 3.16. Kinetic parameters omethyl-butyrate and methyl-pentanoate


103

Perhydrolytic activity: kinetic parameters for methyl-butyrate

166 is part of S1 binding pocket, different substitutions at

can change enzyme specificity toward different substrates.

perhydrolytic activity were determined for three ester substrates. Beside methyl

which was substrate of choice for screening and already mentioned methyl

variants are also characterized with methyl-pentanoate. Kinetic data for methyl

pentanoate are presented in Figure 3.24 and Table 3.16

Substrate KM (mM) kcat (min

butyrate

pentanoate

117.7 ± 26.8

35.2 ± 4.2

124.6 ± 13.7

69.5 ± 3.6

butyrate

pentanoate

122.6 ± 36.9

38.3 ± 5.1

258.3 ± 38.1

170.8 ± 12

butyrate

pentanoate

64.7 ± 12.5

33.3 ± 5.5

676.0 ± 49.4

382.7 ± 27.2

butyrate

pentanoate

93.1 ± 10.4

29.2 ± 3.4

605.5 ± 29.7

378.8 ± 18.2

butyrate

pentanoate

88.8 ± 20.1

26.3 ± 4.6

525.8 ± 51.2

182.3 ± 12.7

Kinetic parameters of the selected variants for perhydrolysispentanoate.


butyrate and methyl-

166 is part of S1 binding pocket, different substitutions at that

Kinetic constants for

Beside methyl-butyrate

which was substrate of choice for screening and already mentioned methyl-propionate,

pentanoate. Kinetic data for methyl-butyrate

ed in Figure 3.24 and Table 3.16.

(min -1) kcat/KM

(mM-1min -1)

13.7

3.6

1.1

2.0

38.1

12.9

2.1

5.7

49.4

27.2

10.4

11.5

29.7

18.2

6.5

13.0

51.2

12.7

5.9

6.9

ected variants for perhydrolysis activity of

Perhydrolase has a ~2.5

ester substrates and similar K

perhydrolytic reaction variants M1, M2 and M3 have increased k

2.0-fold, respectively and decreased K

the highest affinity for methyl

compared to WT.

In the case of methyl

M2 variant have ~2.2-fold increased k

Perhydrolase. KM values are up to 1.5

substrate M2 variant has the highest specificity constant which is 11.8

compared to WT.

Due to decreased KM

length of the acyl chain, Figure 3.25

Figure 3.25: Effect of acyl chain length on the catalytic efficiency of the variants.


104

has a ~2.5-fold increased kcat for perhydrolytic activity

ester substrates and similar KM to WT. With methyl-butyrate as a substrate fo

drolytic reaction variants M1, M2 and M3 have increased kcat

fold, respectively and decreased KM values compared to Perhydrolase.

the highest affinity for methyl-butyrate is M1 with specificity constant

In the case of methyl-pentanoate as a substrate in perhydrolytic reaction M1 and

fold increased kcat, while kcat of M3 variant is

values are up to 1.5-fold decreased. With methyl

substrate M2 variant has the highest specificity constant which is 11.8

M, the catalytic efficiency of all the variants increases

h of the acyl chain, Figure 3.25.

chain length on the catalytic efficiency of the variants.


for perhydrolytic activity with all three

butyrate as a substrate for

cat values 2.6, 2.3 and

values compared to Perhydrolase. Variant with

specificity constant 9.5-fold increased

pentanoate as a substrate in perhydrolytic reaction M1 and

of M3 variant is similar to kcat of

ethyl-pentanoate as a

substrate M2 variant has the highest specificity constant which is 11.8-fold higher

fficiency of all the variants increases with the

chain length on the catalytic efficiency of the variants.


105

4.8.5 Esterolytic activity of the variants

Substrate (methyl-butyrate) and product (butyric acid) in the esterolytic reaction

were separately passed through a reversed-phase C18 column under described

conditions (Chapter 3.15 of Part III) to determine retention time for each component.

After identification of each component on chromatogram, a calibration curve for butyric

acid was generated. Relationship between concentration of butyric acid and peak area

on the chromatogram is presented in Figure 3.26.

Figure 3.26. Calibration curve for butyric acid.

Esterolytic activity of the variants was carried with 200 mM methyl-butyrate at

30°C for 10 min. The reaction was stopped by decrea sing the pH value to 3 with

addition of phosphoric acid. Samples were analyzed by reversed-phase

chromatography and amount of produced butyric acid in esterolytic reaction was

calculated from calibration curve generated in the presence of methyl-butyrate. Specific

activity (U/mg) was calculated for all variants, where 1 U presents amount of enzyme

which generates 1 µmol of product per minute.


106

Table 3.17. Esterolytic activity of the variants with 200 mM methyl-butyrate in 100 mM Na-phosphate buffer pH 7.5. Product was separated from substrate by HPLC on a reversed-phase C18 column with acetonitrile/K-phosphate buffer (20 mM, pH 3.0) gradient elution and detected at 210 nm.

Variant Esterolytic activity (U/mg)

Wild type 0.96 ± 0.02

Perhydrolase 0.63 ± 0.06

M1 1.72 ± 0.08

M2 1.06 ± 0.02

M3 1.46 ± 0.02

Results for esterolytic activity also confirmed that Perhydrolase has decreased

specificity for water compared to WT, since has lower esterolytic activity with methyl-

butyrate, while perhydrolytic activity with the same substrate is increased. This data

additionally confirms Molecular Dynamics simulation results (Lee, Vojcic et al. 2010)

which suggested a better stabilization of the perhydrolytic tetrahedral intermediate over

hydrolytic intermediate in Perhydrolase variant compared to WT. Variants M1, M2 and

M3 besides increased perhydrolytic activity also have an increased esterolytic activity

compared to Perhydrolase, leading to the conclusion that a position 166, as a part of S1

binding pocket, interacts with both peptide and ester substrates.


107

5. Discussion

Subtilisin proteases are among the first engineered enzymes and until today

most of the amino acids have been exchanged to increase activity, termostability, alter

specificity and pH profile or to understand the function of each position in the active site

and substrate binding pocket. However, there has not been a study about esterolytic

and perhydrolytic activity of this proteases or protein engineering experiments which

alter their ester specificity. Wieland at al. (Wieland, Polanyi-Bald et al. 2009) for the first

time reported that one of the serine proteases, subtilisin Carlsberg, besides it’s natural

role catalyzes the reaction of ester perhydrolysis. They also performed directed

evolution experiments obtaining variant (Thr59Ala/Leu217Trp) which has 2.7-fold

increased perhydrolytic activity with methyl-butyrate and 8-fold decreased proteolytic

activity with suc-AAPF-pNA. We continued to engineer variant Thr59Ala/Leu217Trp in

order to further increase perhydrolytic activity.

Despite subtilisins having broad substrate specificity, they can differ significantly

from each other in catalytic efficiency against selected substrates. Specificity is the

result of chemical binding forces which includes hydrogen bonding, electrostatic

interactions, hydrophobic and steric effects (Fersht 1999). Strength and specificity of the

enzyme-substrate binding also determines the kinetic parameters of the reaction such

are kcat, KM and the catalytic constant (kcat/KM). KM is defined by the binding energy of

unaltered substrate for the enzyme, while kcat is determined by activation energy

(energy difference between ES and ES‡). In the reaction of peptide bond hydrolysis, the

limiting step is formation of the first transition state, which determines kcat of the

reaction, while limiting step in the reaction of ester hydrolysis is degradation of acyl-

enzyme intermediate and second product formation (Wells and Estell 1988), Figure

3.27.


108

Figure 3.27: The rate-limiting acylation step (which defines kcat) for hydrolysis of peptide bonds by subtilisin BPN’. Taken over from (Wells and Estell 1988).

Position Gly 166 is conserved in five known subtilisins: subtilisin BPN’ (Wells,

Ferrari et al. 1983), subtilisin Carlsberg (Guntelberg and Ottesen 1954), subtilisin DY

(Nedkov, Oberthür et al. 1985), subtilisin Amylosacchariticus (Kurihara, Markland et al.

1972), subtilisin E (Wong, Price et al. 1984). Still, engineering of this position produced

variants with changed substrate specificity and same or similar catalytic efficiency

(Estell, Graycar et al. 1986). Gly 166 is located at the bottom of S1 substrate binding

cleft, Figure 3.28. This pocket is a hydrophobic cleft which can accept any amino acid at

P1 position of the substrate, providing broad substrate specificity for subtilisin

Carlsberg. We identified variants with substitutions at position 166 which not only have

changed specificity towards suc-AAPF-pNA, but also for different ester substrates with

increasing size and hydrophobicity. Proteolytic activity of these variants using the

artificial substrate suc-AAPF-pNA and a complex substrate was decreased, while

perhydrolytic and hydrolytic activities with ester substrate were increased.

Figure 3.28: Model of the substrate, sucactive site of subtilisin Carlsberg


109

substrate, suc-AAPF-pNA (element representation),active site of subtilisin Carlsberg (gray).


t representation), bound to the

5.1 Suc-AAPF-pNA specificity

To understand the nature of substrate binding pocket of subtilisin

interactions with suc-AAPF

Figure 3.29: Kinetic parameters of the variantswhite bars, KM 10-5(M) – gray bars and k

Even though kcat of Perhydrolase is 5.2 times less compared to WT,

efficiency did not change significantly, since K

already proposed by Lee at al.

destabilization of second tetrahedral intermediate by introducing substitution

Leu217Trp. Additionally, the

pocket where introduction of “bulkier”

nitroanilide group of the substrate to

By introducing Leu, Ile and Tyr at

S1 pocket are changed, suggesting that specificity is determined by steric and

hydrophobic effects. The nature of p

AAPF-pNA) binding in the S1 pocket is analyz

parameters. Increasing the hydrophobicity of S1 binding pocket by amino acid

substitution at position 166 may increase binding of hydrophobic P1 substrates.


110

pNA specificity

To understand the nature of substrate binding pocket of subtilisin

AAPF-pNA, the observed kinetic parameters were

Kinetic parameters of the variants for hydrolysis of suc-AAPFgray bars and kcat/KM*105(s-1*M-1) – black bars.

of Perhydrolase is 5.2 times less compared to WT,

change significantly, since KM decreased 4-fold, Figure 3.29

Lee at al. (Lee, Vojcic et al. 2010), kcat is decreased due to a

econd tetrahedral intermediate by introducing substitution

the decreased kcat value might be due to a

pocket where introduction of “bulkier” tryptophan could prevent

nilide group of the substrate to the S1’ pocket.

By introducing Leu, Ile and Tyr at position 166, the size and hydrophobicity of the


ophobic effects. The nature of phenylalanine (Phe is at position P1 in the suc

pNA) binding in the S1 pocket is analyzed according to the obtained kinetic




To understand the nature of substrate binding pocket of subtilisin Carlsberg and

kinetic parameters were analyzed.

AAPF-pNA: kcat (1/s) –

of Perhydrolase is 5.2 times less compared to WT, the catalytic

fold, Figure 3.29. As

is decreased due to a

econd tetrahedral intermediate by introducing substitution

might be due to a changes in S1’ the

could prevent binding of the p-

position 166, the size and hydrophobicity of the


henylalanine (Phe is at position P1 in the suc-

ed according to the obtained kinetic




111

However, since hydrophobicity is proportional to its surface area (Rose, Geselowitz et

al. 1985), increasing the hydrophobicity of the S1 pocket may sterically hinder binding of

larger substrates.

Catalytic efficiency of M3 variant (Gly→Tyr) is 2-fold decreased, M1 (Gly→Leu)

has a 6-fold decreased kcat/KM, while M2 (Gly→Ile) has a 65-fold decreased efficiency

compared to Perhydrolase (Gly 166). Although, M3 has introduced a Tyr at position

166, which has a large residue, the efficiency is not as affected as when Leu and Ile

were introduced. This result suggests that the steric effect of Tyr is lower, which can be

due to the orientation of the residue in the S1 pocket. Figure 3.30 shows interaction of

suc-AAPF-pNA bound to the active site residue Ser 221 with amino acid at position 166.

The phenol ring of Tyr residue is orientated towards the solvent, while Leu and Ile

residues remain in the pocket, further decreasing the S1 pocket size. Decreased pocket

size will increase steric repulsion between the Phe residue of substrate and Ile 166,

possibly explaining why M2 (Gly→Ile) is less efficient when large aromatic substrates

are at position P1.

Figure 3.30: Interaction of sucdifferent variants (Perhydrolase, M1, M2 and M3). The closest interactionsposition 166 and phenol ring was observed in M2 variant (G166I).

Substitution of Gly 166

the variants when a complex substrate


112

Interaction of suc-AAPF-pNA bound to Ser 221 with amino acid at positiondifferent variants (Perhydrolase, M1, M2 and M3). The closest interactions

166 and phenol ring was observed in M2 variant (G166I).

Substitution of Gly 166 narrows substrate specificity, explain

complex substrate such skim milk is used, Figure 3.22.


1 with amino acid at position 166 in

different variants (Perhydrolase, M1, M2 and M3). The closest interactions between residue at

explaining lower activity of

, Figure 3.22.


113

5.2 Ester specificity

Perhydrolytic activity of the variants was determined using three ester substrates

which differ in the length of acyl chain (from propyl to pentanoyl group). As the size of

the acyl chain increases, the KM value of each variants decreases, suggesting a

stronger substrate binding to the hydrophobic S1 pocket as substrate hydrophobicity

increases. Each additional methylene group theoretically changes the Gibbs free energy

by 2.84 KJ/mol (Fersht 1999). Docking results of each substrate showed increase in

binding energy with increase of substrate acyl chain, which indicates stronger binding of

the substrate to the enzyme. Table 3.18 summarizes binding energies of substrate

conformations for each substrate to all the variants of subtilisin Carlsberg.

Table 3.18. Binding energies of the substrates (only the score with the highest energies are included, A and B) to each variant of subtilisin Carlsberg. In bolded are presented binding energies of substrate in the position which enables conversion (productive binding of the substrate).

Binding energy (kcal/mol) Variant Methyl -propionate Methyl -butyrate Methyl -pentanoate

A B A B A B

Wild type 3.34 3.21 3.7 3.58 4.09 3.98

Perhydrolase 3.36 3.21 3.77 3.61 4.22 3.99

M1 3.27 3.68 4.23

M2 3.34 3.78 4.11

M3 3.27 3.68 3.6 4.25 3.84

Next, we analyzed the effect of each mutation on the catalytic efficiency of each

variant using the same substrate.

Methyl-propionate: the increased hydrophobicity (Tyr→Leu/Ile) and decreased

size of S1 pocket decrease KM value of the variants from 1.6 to 2-fold. Catalytic

constant (kcat) of the variants is increased up to 1.4-fold (for M3) compared to parent.

Methyl-propionate binds in the S1 pocket of Perhydrolase in two conformations

(productive and nonproductive), Figure 3.31. Non-productive conformation (A) has

stronger binding energy than productive (B), meaning that at saturation less than half of

the substrate is bound productively, see Table 3.18.

Figure 3.31. Docking result of methylbinds in two conformation: A –

Introduction of Leu, Ile and Tyr in the S1 pocket instead of Gly

resulted in only one possible

is the productive conformation, Figure 3.32. This result also explains

since at saturation all the substrate


114


the substrate is bound productively, see Table 3.18.

Docking result of methyl-propionate in the active site of Perhydrolase. Substrate – non-productive binding and B – productive binding.

Introduction of Leu, Ile and Tyr in the S1 pocket instead of Gly

d in only one possible binding mode of methyl-propionate in the active site, which

productive conformation, Figure 3.32. This result also explains

t saturation all the substrate should be bound productively.



propionate in the active site of Perhydrolase. Substrate

productive binding.

Introduction of Leu, Ile and Tyr in the S1 pocket instead of Gly at position 166

propionate in the active site, which

productive conformation, Figure 3.32. This result also explains the increased kcat,

Figure 3.32: Docking result of methylvariant (Ile 166) and M3 variant

Methyl-butyrate: KM

case of M1, while kcat is significantly increas

increased kcat of M1, M2 and M3 variant, respectively

propionate, methyl-butyrate

(productive and non-productive), Figure 3.33

has stronger binding energy than productive


115

Docking result of methyl-propionate in the active site of M1 variant M3 variant (Tyr 166).

values of the variants were decreased,

is significantly increased for all the variants (

of M1, M2 and M3 variant, respectively). As in the case of methyl

butyrate also shows two binding conformations in the S1 pocket

productive), Figure 3.33. Again, non-productive conformation (A)

has stronger binding energy than productive conformation (B), see Table 3.18.


M1 variant (Leu 166), M2

values of the variants were decreased, up to 1.8-fold in the

ed for all the variants (2.6-, 2.3- and 2-fold

As in the case of methyl-

binding conformations in the S1 pocket

productive conformation (A)

(B), see Table 3.18.

Figure 3.33. Docking result of methylbinds in two conformation: A –

Docking results show that m

one conformation that favors

and M2). In the case of M3 variant, methyl

productive (A) and non-productive

conformations of the substrate are

When substrate binding is productive

site, inside the S1 pocket, while in the case of non

oriented towards the solvent, having

(M3(A) and M3(B)). The presence of the non

decreased kcat value for M3 variant compared to M1.


116

king result of methyl-butyrate in the active site of Perhydrolase. Substrate – nonproductive binding and B – productive binding.

Docking results show that methyl-butyrate binds for M1 and M2 variant only in

one conformation that favors catalysis, resulting in a increased kcat

of M3 variant, methyl-butyrate docks in two

productive (B), Figure 3.34 (M3(A) and M3(B)

substrate are possible due to ineraction of Tyr 166 with substrate.

is productive Tyr 166 is more oriented in the direction of active

site, inside the S1 pocket, while in the case of non-productive binding Tyr 166 is

solvent, having less influence on the size of S1 pocket, Figure 3.34

resence of the non-productive conformation explains slightly

for M3 variant compared to M1.


in the active site of Perhydrolase. Substrate

productive binding.

for M1 and M2 variant only in

cat, Figure 3.34, (M1

ocks in two conformations,

M3(A) and M3(B)). Two different

Tyr 166 with substrate.

ented in the direction of active

productive binding Tyr 166 is

less influence on the size of S1 pocket, Figure 3.34

ive conformation explains slightly

Figure 3.34: Docking result of methylvariant (Ile 166) and M3 variant (Tyr 166).conformation for M3 variant andconformation for M3 variant. Conformation A has higher binding energy

Methyl-pentanoate: K

hydrophobicity of the substrate, where further increase in the hydrophobicity of the

pocket does not influence significantly on the binding energy of the substrate. Again, M1

and M2 show higher catalytic constant values, which

to parent, while M3 variant has the same k

methyl-pentanoate also binds in two conformations for parent and M3 variant, which

explains similar kcat values

M3(A) and M3(B)). Docking results

is bound only in the productive conformation, Figure 3.35


117

king result of methyl-butyrate in the active site of M1 variant (Leu 166), M2 variant (Ile 166) and M3 variant (Tyr 166). A present methyl-butyrate bou

or M3 variant and B present methyl-butyrate bound in . Conformation A has higher binding energy than conformation

KM values of the variants are very similar, most



and M2 show higher catalytic constant values, which are 2.2-fold in

M3 variant has the same kcat as parent. Docking results showed that

pentanoate also binds in two conformations for parent and M3 variant, which

values between these two variants, Figure 3.35

Docking results for M1 and M2 variant show that

productive conformation, Figure 3.35 (M1 and


M1 variant (Leu 166), M2

butyrate bound in productive butyrate bound in nonproductive

than conformation B.

values of the variants are very similar, mostly defined by



fold increased compared

. Docking results showed that

pentanoate also binds in two conformations for parent and M3 variant, which

, Figure 3.35 (Per(A), Per (B),

variant show that methyl-pentanoate

and M2). At saturation

concentration of substrate,

catalysis, resulting in a kcat

Figure 3.35: Docking result of methyl(Leu 166), M2 variant (Ile 166) and M3 variant (Tyr 16present methyl-pentanoate conformationbinding energy than conformation B.


118

all the molecules are bound in the conforma

value more than 2-fold increased compared to parent.

king result of methyl-pentanoate in the active site of Perhydrolase, (Leu 166), M2 variant (Ile 166) and M3 variant (Tyr 166). A (productive) and B (nonproductive)

pentanoate conformation bound in the S1 pocket. Conformation A has higher binding energy than conformation B.


the conformation which favors

compared to parent.

Perhydrolase, M1 variant

A (productive) and B (nonproductive) bound in the S1 pocket. Conformation A has higher


119

Decreased size of the S1 pocket allows binding of the ester substrate

preferentially in the productive conformation which leads to catalysis. Leu, Ile and Tyr at

position 166 “push” the ester substrate toward active site, affecting the susceptibility of

the scissile ester bond to the nucleophilic attack by the hydroxyl group of the catalytic

Ser 221.


120

6. Conclusion

In summary, an expression system for subtilisin Carlsberg in B. subtilis WB600

and B. subtilis DB104 was developed in microtiter plates suitable for perhydrolytic and

proteolytic activity screening. APCC assay was optimized for detection of perhydrolytic

activity in the supernatant of Bacillus cells and used for final characterization of the

variants.

Screening site saturation libraries generated at positions near the active site and

in the substrate binding pockets identified position 166 as residue that determines

substrate specificity of subtilisin Carlsberg. An increase in the side-chain volume at

position 166 decreased the size of S1 gap leading to a reduction in kcat/KM value

towards large amino acids at position P1. At the same time, catalytic efficiencies

towards ester substrates were increased up to 10-fold in these variants. Results showed

that the specificity of subtilisin Carlsberg could be changed by replacing directly amino

acids that are in contact with the substrate, which brings the idea of changing protease

into esterase and perhydrolase.

Important substitutions for increasing specificity toward hydrogen-peroxide

instead water were not identified by screening random mutagenesis libraries and SSM

libraries at positions identified by rational design. Same catalytic center for hydrolytic

and perhydrolytic activity, similar structure of hydrogen-peroxide with water leads to

similar structure of perhydrolytic and hydrolytic tetrahedral intermediate, suggesting that

the same residues in subtilisin Carlsberg are included in interactions with substrates

and intermediates. This limits the possibilities to find residues that will favor one type

reaction over the other. Additionally, diversity is lost since variants without hydrolytic

activity can’t autocatalytically process into mature form, although it is possible that once

the pro-sequence is cleaved they still can have perhydrolytic activity.

Future prospects

121

Future prospects

Results of the directed evolution and modeling studies gave insight how to

engineer protease into esterase and perhydrolase preserving the level of proteolytic

activity necessary for protease maturation.

However, interesting questions come from the results and discussion of this

work:

1. Is it possible to make better perhydrolase if we avoid prescreening step

caused by automaturation?

2. How much of diversity is lost during this self caused prescreening?

3. Can serine protease be converted into effective perhydrolase?

4. Why only some subtilisins show perhydrolytic activity?

5. Will engineering of pro-sequence induce conformational changes that alter

substrate specificity and catalytic activity?

To tackle some of these questions it is necessary to develop an expression

system without autocatalytic processing. Development of in vitro translation systems of

the mature protease form in the presence of pro-sequence to guide folding or wild type

protease to activate variants which can not process combined with high throughput

screening system might provide a platform to address this questions.

References

122

References

Abrahmsen, L., J. Tom, J. Burnier, K. A. Butcher, A. Kossiakoff and J. A. Wells (1991). "Engineering subtilisin and its substrates for efficient ligation of peptide bonds in aqueous solution." Biochemistry 30(17): 4151-4159.

Aharoni, A., G. Amitai, K. Bernath, S. Magdassi and D. S. Tawfik (2005). "High-throughput screening of enzyme libraries: Thiolactonases evolved by fluorescence-activated sorting of single cells in emulsion compartments." Chemistry & Biology 12(12): 1281-1289.

Aharoni, A., A. D. Griffiths and D. S. Tawfik (2005). "High-throughput screens and selections of enzyme-encoding genes." Current Opinion in Chemical Biology 9(2): 210-216.

Akerman, D., D. A. S. Phillips and J. A. Taylor (2003). "Interaction of oxidative bleach containing detergents with dyes. Part 1: Preparation and resistance of 1-amino-2-arylazonaphthalene-4-sulphonic acid dyes to hydrogen peroxide and m-chloroperbenzoic acid." Dyes and Pigments 59(3): 285-292.

Almog, O., A. González, N. Godin, M. de Leeuw, M. J. Mekel, D. Klein, S. Braun, G. Shoham and R. L. Walter (2009). "The crystal structures of the psychrophilic subtilisin S41 and the mesophilic subtilisin Sph reveal the same calcium-loaded state." Proteins: Structure, Function, and Bioinformatics 74(2): 489-496.

Arnold, F. H. (2001). "Combinatorial and computational challenges for biocatalyst design." Nature 409(6817): 253-257.

Baldry, M. G. C. (1983). "The bactericidal, fungicidal and sporicidal properties of hydrogen peroxide and peracetic acid." Journal of Applied Microbiology 54(3): 417-423.

Bantleon, R., J. Altenbuchner and K. H. van Pee (1994). "Chloroperoxidase from Streptomyces lividans: isolation and characterization of the enzyme and the corresponding gene." Journal of Bacteriology 176(8): 2339-2347.

Barrett, A. J. and N. D. Rawlings (1995). "Families and clans of serine peptidases." Archives of Biochemistry and Biophysics 318(2): 247-250.

Barrett, A. J., N. D. Rawlings and J. F. Woessner (1998). Handbook of proteolytic enzymes. San Diego [etc.], Academic Press.

Bernhardt, P., K. Hult and R. J. Kazlauskas (2005). "Molecular basis of perhydrolase activity in serine hydrolases." Angewandte Chemie International Edition 44(18): 2742-2746.

Betzel, C., S. Klupsch, G. Papendorf, S. Hastrup, S. Branner and K. S. Wilson (1992). "Crystal structure of the alkaline proteinase Savinase(TM) from Bacillus lentus at 1.4 Å resolution." Journal of Molecular Biology 223(2): 427-445.

Betzel, C., G. P. Pal and W. Saenger (1988). "Three-dimensional structure of proteinase K at 0.15-nm resolution." European Journal of Biochemistry 178(1): 155-171.

Binder, W. H. and F. M. Menger (2000). "Assay of peracid in the presence of excess hydrogen peroxide." Analytical Letters 33(3): 479-488.

Björkling, F., H. Frykman, S. E. Godtfredsen and O. Kirk (1992). "Lipase catalyzed synthesis of peroxycarboxylic acids and lipase mediated oxidations." Tetrahedron 48(22): 4587-4592.

Block, S. (2001). Disinfection, Sterilisation and Preservation. Philadelphia, PA, USA, Lippincott, Williams and Wilkins. 1481: 185-204.

Bode, W., E. Papamokos, D. Musil, U. Seemueller and H. Fritz (1986). "Refined 1.2 A crystal structure of the complex formed between subtilisin Carlsberg and the inhibitor eglin c. Molecular structure of eglin and its detailed interaction with subtilisin." The EMBO Journal 5(4): 813-818.

References

123

Bornscheuer, U. T. and R. J. Kazlauskas (2004). "Catalytic promiscuity in biocatalysis: Using old enzymes to form new bonds and follow new pathways." Angewandte Chemie International Edition 43(45): 6032-6040.

Bugg, T. D. H. (2004). "Diverse catalytic activities in the [alpha][beta]-hydrolase family of enzymes: activation of H2O, HCN, H2O2, and O2." Bioorganic Chemistry 32(5): 367-375.

Burd, W., O. Yourkevich, A. J. Voskoboev and K. H. van Pée M (1995). "Purification and properties of a non-haem chloroperoxidase from Serratia marcescens." FEMS Microbiology Letters 129(2-3): 255-260.

Cadwell, R. C. and G. F. Joyce (1994). "Mutagenic PCR." Genome Research 3(6): S136-S140. Carboni-Oerlemans, C., P. Domínguez de María, B. Tuin, G. Bargeman, A. van der Meer and R. van

Gemert (2006). "Hydrolase-catalysed synthesis of peroxycarboxylic acids: Biocatalytic promiscuity for practical applications." Journal of Biotechnology 126(2): 140-151.

Cella, J. A., J. A. Kelley and E. F. Kenehan (1975). "Nitroxide-catalyzed oxidation of alcohols using m-chloroperbenzoic acid. New method." The Journal of Organic Chemistry 40(12): 1860-1862.

Cirino, P. C., K. M. Mayer and D. Umeno (2003). Generating mutant libraries using error-prone PCR. Directed Evolution Library Creation. F. H. Arnold and G. Georgiou, Humana Press. 231: 3-9.

D'Ans, J. and W. Frey (1913). "Untersuchungen über die Bildung von Persäuren aus organischen Säuren und Hydroperoxyd." Zeitschrift für anorganische Chemie 84(1): 145-164.

Davies, D. M. and M. E. Deary (1988). "Determination of peracids in the presence of a large excess of hydrogen peroxide using a rapid and convenient spectrophotometric method." Analyst 113(9): 1477-1479.

Di Furia, F., M. Prato, U. Quintily, S. Salvagno and G. Scorrano (1984). "Gas-liquid chromatographic method for the determination of peracids in the presence of a large excess of hydrogen peroxide." Analyst 109(8): 985-987.

Dicosimo, R., J. E. Gavagan, M. S. Payne and F. B. Cooling (2008). Production of peracids using an enzyme having perhydrolysis activity. US. 11743354.

Dubreucq, E., A. Weiss and G. Moulin (2007). Polypeptides with perhydrolase activity. US. 11709604. Effkemann, S., U. Pinkernell, R. Neumüller, F. Schwan, H. Engelhardt and U. Karst (1998). "Liquid

chromatographic simultaneous determination of peroxycarboxylic acids using postcolumn derivatization." Analytical Chemistry 70(18): 3857-3862.

Egnell, P. and J. I. Flock (1992). "The autocatalytic processing of the subtilisin Carlsberg pro-region is independent of the primary structure of the cleavage site." Molecular Microbiology 6(9): 1115-1119.

Egnell, P. and J. I. Flock (1992). "The autocatalytic processing of the subtilisin Carlsberg pro-region is independent of the primary structure of the cleavage site." Molecular Microbiology 6(9): 1115-1119.

Estell, D. A., T. P. Graycar, J. V. Miller, D. B. Powers, J. A. Wells, J. P. Burnier and P. G. Ng (1986). "Probing steric and hydrophobic effects on enzyme-substrate interactions by protein engineering." Science 233(4764): 659-663.

Farinas, E. T. (2006). "Fluorescence activated cell sorting for enzymatic activity." Combinatorial Chemistry and High Throughput Screening 9(4): 321-328.

Farinas, E. T., T. Bulter and F. H. Arnold (2001). "Directed enzyme evolution." Current Opinion in Biotechnology 12(6): 545-551.

Fersht, A. (1999). Structure and mechanism in protein science: Guide to enzyme catalysis and protein folding New York, W.H. Freeman and Co.

Fischer, W., E. Arlt and B. Brabander (1990). Method and reagent for the determination of peracids. US. 07285700.

References

124

Garrett, R. H. and C. M. Grisham (1999). Biochemistry, Harcourt Brace, Custom Publishers Greenspan, F. P. and D. G. MacKellar (1948). "Analysis of aliphatic per acids." Analytical Chemistry

20(11): 1061-1063. Gros, P., C. Betzel, Z. Dauter, K. S. Wilson and W. G. J. Hol (1989). "Molecular dynamics refinement of a

thermitase-eglin-c complex at 1.98 Å resolution and comparison of two crystal forms that differ in calcium content." Journal of Molecular Biology 210(2): 347-367.

Guntelberg, A. V. and M. Ottesen (1954). "Purification of the proteolytic enzyme from Bacillus subtilis." Comptes Rendus des Travaux du Laboratorie Carlsberg. Serie Chimique 29(3-4): 36-48.

Gupta, R., Q. K. Beg and P. Lorenz (2002). "Bacterial alkaline proteases: molecular approaches and industrial applications." Applied Microbiology and Biotechnology 59(1): 15-32.

Güven, G., R. Prodanovic and U. Schwaneberg (2010). "Protein engineering – an option for enzymatic biofuel cell design." Electroanalysis 22(7-8): 765-775.

Harms, D. and U. Karst (1999). "Rapid and selective determination of peroxyacetic acid in disinfectants using flow injection analysis." Analytica Chimica Acta 389(1-3): 233-238.

Hedstrom, L. (2002). "Serine protease mechanism and specificity." Chemical Reviews 102(12): 4501-4524.

Hofmann, B., S. Tölzer, I. Pelletier, J. Altenbuchner, K. H. van Pée and H. J. Hecht (1998). "Structural investigation of the cofactor-free chloroperoxidases." Journal of Molecular Biology 279(4): 889-900.

Hult, K. and P. Berglund (2007). "Enzyme promiscuity: mechanism and applications." Trends in Biotechnology 25(5): 231-238.

Ikemura, H., H. Takagi and M. Inouye (1987). "Requirement of pro-sequence for the production of active subtilisin E in Escherichia coli." Journal of Biological Chemistry 262(16): 7859-7864.

Inoue, H., H. Nojima and H. Okayama (1990). "High efficiency transformation of Escherichia coli with plasmids." Gene 96(1): 23-28.

Jacks, T. J. (2004). "Spectrofluorometric determination of hypohalites and peroxyacids using kojic acid." Analytical Letters 37(6): 1177 - 1184.

Jacks, T. J. (2005). "Evaluation of kojic acid for determining heme and nonheme haloperoxidase activities spectrofluorometrically." Analytical Letters 38(6): 921-927.

Jackson, D. Y., J. Burnier, C. Quan, M. Stanley, J. Tom and J. A. Wells (1994). "A designed peptide ligase for total synthesis of ribonuclease A with unnatural catalytic residues." Science 266(5183): 243-247.

Jacobs, M., M. Eliasson, M. Uhlén and J. I. Flock (1985). "Cloning, sequencing and expression of subtilisin Carlsberg from Bacillus licheniformis." Nucleic Acids Research 13(24): 8913-8926.

Khanna, P. L., C. C. Chang and E. F. Ullman (1991). Activated aryl capped fluorescent compounds, and assay methods and systems using them as fluorogenic precursors. US.

Kirk, O., T. V. Borchert and C. C. Fuglsang (2002). "Industrial enzyme applications." Current Opinion in Biotechnology 13(4): 345-351.

Kirk, O., M. W. Christensen, T. Damhus and S. E. Godtfredsen (1994). "Enzyme catalyzed degradation and formation of peroxycarboxylic acids." Biocatalysis and Biotransformation 11(1): 65-77.

Kirk, O., T. Damhus and M. Würtz Christensen (1992). "Determination of peroxycarboxylic acids by high-performance liquid chromatography with electrochemical detection." Journal of Chromatography A 606(1): 49-53.

Kirner, S., S. Krauss, G. Sury, S. T. Lam, J. M. Ligon and K. H. van Pée (1996). "The non-haem chloroperoxidase from Pseudomonas fluorescens and its relationship to pyrrolnitrin biosynthesis." Microbiology 142(8): 2129-2135.

References

125

Kobayashi, G., K. Fujii, M. Serizawa, H. Yamamoto and J. Sekiguchi (2002). "Simultaneous display of bacterial and fungal lipases on the cell surface of Bacillus subtilis." Journal of Bioscience and Bioengineering 93(1): 15-19.

Korkegian, A., M. E. Black, D. Baker and B. L. Stoddard (2005). "Computational thermostabilization of an enzyme." Science 308(5723): 857-860.

Kraut, J. (1977). "Serine proteases: structure and mechanism of catalysis." Annual Review of Biochemistry 46: 331-358.

Krem, M. M. and E. Di Cera (2001). "Molecular markers of serine protease evolution." The EMBO Journal 20(12): 3036-3045.

Kurihara, M., F. S. Markland and E. L. Smith (1972). "Subtilisin Amylosacchariticus." Journal of Biological Chemistry 247(17): 5619-5631.

Lassila, J. K., J. R. Keeffe, P. Oelschlaeger and S. L. Mayo (2005). "Computationally designed variants of Escherichia coli chorismate mutase show altered catalytic activity." Protein Engineering Design and Selection 18(4): 161-163.

Lee, W., L. Vojcic, D. Despotovic, R. Prodanovic, K.-H. Maurer, U. Schwaneberg and M. Zacharias (2010). "Rationalizing perhydrolase activity of aryl-esterase and subtilisin Carlsberg mutants by molecular dynamics simulations of the second tetrahedral intermediate state." Theoretical Chemistry Accounts: Theory, Computation, and Modeling (Theoretica Chimica Acta) 125(3): 375-386.

Leemhuis, H., R. M. Kelly and L. Dijkhuizen (2009). "Directed evolution of enzymes: Library screening strategies." IUBMB Life 61(3): 222-228.

Lehninger, A. L., D. L. Nelson and M. M. Cox (2005). Lehninger principles of biochemistry, W.H. Freeman. Mastrobattista, E., V. Taly, E. Chanudet, P. Treacy, B. T. Kelly and A. D. Griffiths (2005). "High-throughput

screening of enzyme libraries: In vitro evolution of a [beta]-galactosidase by fluorescence-activated sorting of double emulsions." Chemistry & Biology 12(12): 1291-1300.

Matsuno, Y., T. Ano and M. Shoda (1992). "High-efficiency transformation of Bacillus subtilis NB22, an antifungal antibiotic iturin producer, by electroporation." Journal of Fermentation and Bioengineering 73(4): 261-264.

Maurer, K.-H. (2004). "Detergent proteases." Current Opinion in Biotechnology 15(4): 330-334. Miller, O. J., K. Bernath, J. J. Agresti, G. Amitai, B. T. Kelly, E. Mastrobattista, V. Taly, S. Magdassi, D. S.

Tawfik and A. D. Griffiths (2006). "Directed evolution by in vitro compartmentalization." Nature Methods 3(7): 561-570.

Minning, S., A. Weiss, U. T. Bornscheuer and R. D. Schmid (1999). "Determination of peracid and putative enzymatic peracid formation by an easy colorimetric assay." Analytica Chimica Acta 378(1-3): 293-298.

Miyazaki, K. (2003). "Creating random mutagenesis libraries by megaprimer PCR of whole plasmid (MEGAWHOP)." Methods in Molecular Biology 231: 23-28.

Mundhada, H., J. Marienhagen, A. Scacioc, A. Schenk, D. Roccatano and U. Schwaneberg (2011). "SeSaM-Tv-II generates a protein sequence space that is unobtainable by epPCR." ChemBioChem 12(10): 1595-1601.

Naclerio, G., L. Baccigalupi, C. Caruso, M. De Felice and E. Ricca (1995). "Bacillus subtilis vegetative catalase is an extracellular enzyme." Applied and Environmental Microbiology 61(12): 4471-4473.

Nakatsuka, T., T. Sasaki and E. T. Kaiser (1987). "Peptide segment synthesis catalyzed by the semisynthetic enzyme thiolsubtilisin." Journal of the American Chemical Society 109(12): 3808-3810.

References

126

Nedkov, P., W. Oberthür and G. Braunitzer (1985). "Determination of the complete amino-acid sequence of subtilisin DY and its comparison with the primary structures of the Subtilisins BPN’, Carlsberg and Amylosacchariticus." Biological Chemistry Hoppe-Seyler 366(1): 421-430.

Offenbacher, H., O. S. Wolfbeis and E. Fürlinger (1986). "Fluorescence optical sensors for continuous determination of near-neutral pH values." Sensors and Actuators 9(1): 73-84.

Ottesen, M. and A. Spector (1960). "A comparison of two proteinases from Bacillus subtilis." Comptes-rendus des travaux du Laboratoire Carlsberg 32: 63-74.

Pelletier, I., J. Altenbuchner and R. Mattes (1995). "A catalytic triad is required by the non-heme haloperoxidases to perform halogenation." Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology 1250(2): 149-157.

Pelletier, I., O. Pfeifer, J. Altenbuchner and K.-H. van Pée (1994). "Cloning of a second non-haem bromoperoxidase gene from Streptomyces aureofaciens ATCC 10762: sequence analysis, expression in Streptomyces lividans and enzyme purification." Microbiology 140(3): 509-516.

Perona, J. J. and C. S. Craik (1995). "Structural basis of substrate specificity in the serine proteases." Protein Science 4(3): 337-360.

Pinkernell, U., H. J. Luke and U. Karst (1997). "Selective photometric determination of peroxycarboxylic acids in the presence of hydrogen peroxide." Analyst 122(6): 567-571.

Poulose, J. A. and A. S. Anderson (1992). Enzymatic peracid bleaching system with modified enzyme. US. 07286353.

Poulouse, J. A. (1994). Selection and method of making enzymes for perhydrolysis system and for altering substrate specificity, specific activity and catalytic efficiency. US. 07908596.

Power, S. D., R. M. Adams and J. A. Wells (1986). "Secretion and autoproteolytic maturation of subtilisin." Proceedings of the National Academy of Sciences 83(10): 3096-3100.

Reetz, M. and K.-E. Jaeger (1999). Superior biocatalysts by directed evolution. Biocatalysis - from discovery to application, Springer Berlin / Heidelberg. 200: 31-57.

Reetz, M. T., M. Bocola, J. D. Carballeira, D. Zha and A. Vogel (2005). "Expanding the range of substrate acceptance of enzymes: Combinatorial active-site saturation test." Angewandte Chemie International Edition 44(27): 4192-4196.

Rheinnecker, M., G. Baker, J. Eder and A. R. Fersht (1993). "Engineering a novel specificity in subtilisin BPN'." Biochemistry 32(5): 1199-1203.

Rose, G. D., A. R. Geselowitz, G. J. Lesser, R. H. Lee and M. H. Zehfus (1985). "Hydrophobicity of amino acid residues in globular proteins." Science 229(4716): 834-838.

Rüsch gen. Klaas, M., K. Steffens and N. Patett (2002). "Biocatalytic peroxy acid formation for disinfection." Journal of Molecular Catalysis B: Enzymatic 19-20: 499-505.

Sambrook, J. and D. W. Russell (2001). "Molecular cloning: a laboratory manual." Schulein, R., J. Kreft, S. Gonski and W. Goebel (1991). "Preprosubtilisin Carlsberg processing and

secretion is blocked after deletion of amino acids 97-101 in the mature part of the enzyme." Molecular and General Genetics 227(1): 137-143.

Setsukinai, K., Y. Urano, K. Kikuchi, T. Higuchi and T. Nagano (2000). "Fluorescence switching by O-dearylation of 7-aryloxycoumarins. Development of novel fluorescence probes to detect reactive oxygen species with high selectivity." Journal of the Chemical Society, Perkin Transactions 2(12): 2453-2457.

Smith, E. L., R. J. DeLange, W. H. Evans, M. Landon and F. S. Markland (1968). "Subtilisin Carlsberg." Journal of Biological Chemistry 243(9): 2184-2191.

Song, J. K., H. J. Ahn, H. S. Kim and B. K. Song (2006). "Molecular cloning and expression of perhydrolase genes from Pseudomonas aeruginosa and Burkholderia cepacia in Escherichia coli." Biotechnology Letters 28(12): 849-856.

References

127

Stanley, Y. C. and L. S. Gianfranco (1983). Bleaching compositions. US. 06472683. Stemmer, W. P. (1994). "Rapid evolution of a protein in-vitro by DNA shuffling." Nature 370(6488): 389-

391. Sully, B. D. and P. L. Williams (1962). "The analysis of solutions of per-acids and hydrogen peroxide."

Analyst 87(1037): 653-657. Swern, D. (1949). "Organic peracids." Chemical Reviews 45(1): 1-68. Tephly, T. R., G. J. Mannering and R. E. Parks (1961). "Studies on the mechanism of inhibition of liver and

erythrocyte catalase activity by 3-amino-1,2,4-triazole (AT)." The Journal of Pharmacology and Experimental Therapeutics 134(1): 77-82.

van der Laan, J. M., A. V. Teplyakov, H. Kelders, K. H. Kalk, O. Misset, L. J. S. M. Mulleners and B. W. Dijkstra (1992). "Crystal structure of the high-alkaline serine protease PB92 from Bacillus

alcalophilus." Protein Engineering 5(5): 405-411. Vitale, L. and S. Gamulin (1975). "Characterization of subtilisins by isoelectric focusing and

polyacrylamide gel electrophoresis." International Journal of Biochemistry 6(3): 165-171. Vojcic, L. (2012). Directed evolution of a subtilisin Carlsberg variant towards improved oxidative stability.

Fakultät für Mathematik, Informatik und Naturwissenschaften Aachen, RWTH Aachen University.

Wang, W. and B. A. Malcolm (2002). Two-stage polymerase chain reaction protocol allowing introduction of multiple mutations, deletions, and insertions, using QuikChange site-directed mutagenesis, Humana Press. 182: 37-43.

Wells, J. A., B. C. Cunningham, T. P. Graycar and D. A. Estell (1987). "Recruitment of substrate-specificity properties from one enzyme into a related one by protein engineering." Proceedings of the National Academy of Sciences 84(15): 5167-5171.

Wells, J. A. and D. A. Estell (1988). "Subtilisin — an enzyme designed to be engineered." Trends in Biochemical Sciences 13(8): 291-297.

Wells, J. A., E. Ferrari, D. J. Henner, D. A. Estell and E. Y. Chen (1983). "Cloning, sequencing, and secretion of Bacillus amyloliquefaciens subtillisin in Bacillus subtilis." Nucleic Acids Research 11(22): 7911-7925.

Wells, J. A., D. B. Powers, R. R. Bott, T. P. Graycar and D. A. Estell (1987). "Designing substrate specificity by protein engineering of electrostatic interactions." Proceedings of the National Academy of Sciences 84(5): 1219-1223.

Wieland, S., L. Polanyi-Bald, I. Prueser, R. Stehr and K. H. Maurer (2009). Subtilisin variants with improved perhydrolase activity. US, CA 2 306 376 (Oct, 2000). 11156062.

Wiesner, W., K. H. van Pée and F. Lingens (1988). "Purification and characterization of a novel bacterial non-heme chloroperoxidase from Pseudomonas pyrrocinia." Journal of Biological Chemistry 263(27): 13725-13732.

Wong, S. L., C. W. Price, D. S. Goldfarb and R. H. Doi (1984). "The subtilisin E gene of Bacillus subtilis is transcribed from a sigma 37 promoter in vivo." Proceedings of the National Academy of Sciences 81(4): 1184-1188.

Wong, T. S., D. Roccatano and U. Schwaneberg (2007). "Steering directed protein evolution: strategies to manage combinatorial complexity of mutant libraries." Environmental Microbiology 9(11): 2645-2659.

Wong, T. S., K. L. Tee, B. Hauer and U. Schwaneberg (2004). "Sequence saturation mutagenesis (SeSaM): a novel method for directed evolution." Nucleic Acids Research 32(3): e26.

Wong, T. S., D. Zhurina and U. Schwaneberg (2006). "The diversity challenge in directed protein evolution." Combinatorial Chemistry and High Throughput Screening 9(4): 271-288.

References

128

Wright, C. S., R. A. Alden and J. Kraut (1969). "Structure of subtilisin BPN' at 2.5 angstrom resolution." Nature 221(5177): 235-242.

Wu, X. C., W. Lee, L. Tran and S. L. Wong (1991). "Engineering a Bacillus subtilis expression-secretion system with a strain deficient in six extracellular proteases." Journal of Bacteriology 173(16): 4952-4958.

Yang, M. Y., E. Ferrari and D. J. Henner (1984). "Cloning of the neutral protease gene of Bacillus subtilis and the use of the cloned gene to create an in vitro-derived deletion mutation." Journal of Bacteriology 160(1): 15-21.

Yin, D. L., P. Bernhardt, K. L. Morley, Y. Jiang, J. D. Cheeseman, V. Purpero, J. D. Schrag and R. J. Kazlauskas (2010). "Switching catalysis from hydrolysis to perhydrolysis in Pseudomonas

fluorescens esterase." Biochemistry 49(9): 1931-1942. Yuan, Z., Y. Ni and A. R. P. Van Heiningen (1997). "Kinetics of peracetic acid decomposition: Part I:

Spontaneous decomposition at typical pulp bleaching conditions." The Canadian Journal of Chemical Engineering 75(1): 37-41.

Zhao, H., L. Giver, Z. Shao, J. A. Affholter and F. H. Arnold (1998). "Molecular evolution by staggered extension process (StEP) in vitro recombination." Nature Biotechnology 16(3): 258-261.

Zhao, X., R. Wu and D. Liu (2011). "Production of pulp, ethanol and lignin from sugarcane bagasse by alkali-peracetic acid delignification." Biomass and Bioenergy 35(7): 2874-2882.

Publications

129

Publication list

1. Wook Lee, Ljubica Vojcic, Dragana Despotovic , Radivoje Prodanovic, Karl-

Heinz Maurer, Ulrich Schwaneberg, Martin Zacharias (2010): "Rationalizing

perhydrolase activity of aryl-esterase and subtilisin Carlsberg mutants by molecular

dynamics simulations of the second tetrahedral intermediate state." Theoretical

Chemistry Accounts 125(3-6): 375-386.

2. Dragana Despotovic , Ljubica Vojcic, Radivoje Prodanovic, Ronny Martinez,

Karl-Heinz Maurer, Ulrich Schwaneberg (2012): “Fluorescent assay for directed

evolution of perhydrolases.” Journal of Biomolecular Screening. Accepted

3. Ljubica Vojcic, Dragana Despotovic , Ronny Martinez, Karl-Heinz Maurer, Ulrich

Schwaneberg (2012). “An efficient transformation method for Bacillus subtilis DB104.”

Applied Microbiology and Biotechnology. Accepted

4. Ljubica Vojcic, Dragana Despotovic , Karl-Heinz Maurer, Ronny Martinez, Ulrich

Schwaneberg (2012). “Directed evolution of a double mutant subtilisin Carlsberg,

Thr59Ala/Leu217Trp towards increased oxidative stability”. Under preparation

5. Dragana Despotovic , Ljubica Vojcic, Milan Blanusa, Karl-Heinz Maurer, Ronny

Martinez, Ulrich Schwaneberg (2012). Switching substrate specificity of a subtilisin

Carlsberg mutant towards ester perhydrolysis. Under preparation

Curriculum vitae

130

Curriculum vitae

Personal data

Name: Dragana Despotovic

Gender: Female

Date of birth: 24th February, 1983

Place of birth: Sabac, Republic of Serbia

Marital Status: Single

Nationality: Serbian

Education

2009 - present PhD fellow of Biotechnology, RWTH Aachen University, Aachen,

Germany

2008 – 2009 PhD fellow of Engineering and Science, Jacobs University Bremen,

Bremen, Germany

2002 – 2008 Diploma in Biochemistry, Faculty of Chemistry, University of

Belgrade, Belgrade, Serbia

1998 – 2002 High school Gimnazija “Vera Blagojevic”, Sabac, Serbia

1991 – 1998 Elementary school, Vladimirci, Serbia

Appendix

A

Appendix

Figure A1. H-NMR spectra of 7-(4-nitrophenoxy)-3-ethyl-carboxy-coumarin

Appendix

B

Figure A2. H-NMR spectra of 7-(4-aminophenoxy)-3-carboxy-coumarin

Directed evolution of a Subtilisin Carlsberg Variant towards ...Tag der mündlichen Prüfung:...

Documents

Transcript of Directed evolution of a Subtilisin Carlsberg Variant towards ...Tag der mündlichen Prüfung:...