Understanding the Substrate Binding Mechanism of Streptomyces Clavuligerus Deacetoxycephalosporin C Synthase for Biosynthesis of Novel Cephalosporins

Home , Nocardicin A, Penam

UNDERSTANDING THE SUBSTRATE BINDING MECHANISM OF STREPTOMYCES CLAVULIGERUS DEACETOXYCEPHALOSPORIN C SYNTHASE FOR BIOSYNTHESIS OF NOVEL CEPHALOSPORINS

GOO KIAN SIM

NATIONAL UNIVERSITY OF SINGAPORE

2011 UNDERSTANDING THE SUBSTRATE BINDING MECHANISM OF STREPTOMYCES CLAVULIGERUS DEACETOXYCEPHALOSPORIN C SYNTHASE FOR BIOSYNTHESIS OF NOVEL CEPHALOSPORINS

GOO KIAN SIM (B. Sc. Hons., NUS)

A THESIS SUBMITTED

FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF MICROBIOLOGY

NATIONAL UNIVERSITY OF SINGAPORE

2011 Acknowledgements

Scientific research has been and will always be a part of my life. Other than carrying out the experiments, documenting the results is equally fulfilling, while putting them into a report has a greater meaning. This project has been a work of sweat and blood, not just from a single person. Hence, I want to express my heartfelt gratitude to the following people who have made this project and thesis possible:

Firstly, to my supervisor, Associate Professor Sim Tiow Suan, for her patience, guidance and encouragement during my candidature. These eleven years of working experience with her since my undergraduate attachement in her lab, has been enriching and nothing can exchange for it.

To Madam Seah Keng Ing, for her excellent technical support in this project. To Mr.

Chin Hong Soon, my mentor and friend, for his guidance and advice, which have provided me a good start to this challenging project. To the lab’s camaraderie Dr.

Doreen Tan, Dr. Goh Liuh Ling and Dr. Maurice Chan, for their valuable advice and frank opinions.

To my fellow and ex-colleagues in the lab: Wenjie, Chun Song, Huiyu, Jun Ji, Cher

Siong and Yan Hao, for their ideas, opinions and assistance in this project, and for their friendships and interesting discussions.

To my family, Linda and friends, for being there for me during this period of challenging, demanding and yet, fulfilling research work. To everyone who has helped me in one way or another in completing this project and thesis.

Thank you!

Table of Contents

Acknowledgements i Table of Contents ii Summary iv List of Tables vi List of Figures viii List of Abbreviations xi List of Units xii Chapter 1 Introduction 1-18 1.1. Antibiotics and emerging resistant pathogens 1.2. β-Lactam antibiotics 1.3. Biosynthesis of cephalosporins 1.4. Objectives of this study Chapter 2 Literature review 19-88 2.1. Antibiotics: secondary metabolites of microorganisms 2.2. Streptomyces species: well known antibiotic producers 2.3. β-Lactam antibiotics 2.4. Penam and cephem antibiotics 2.5. Biosynthesis of penicillin and cephalosporin 2.6. Deacetoxycephalosporin C synthase 2.7. Deacetylcephalosporin C synthase 2.8. Enzyme improvement strategies for improving DAOCS Chapter 3 Materials and methods 89-114 3.1. Materials 3.2. Methods Chapter 4 Improving the ring-expansion activity of scDAOCS 115-126 4.1. Identification of potential sites for the improvement of scDAOCS in hydrophobic penicillin conversions 4.2. In silico structural analyses of scDAOCS mutants 4.3. Generation and verification of mutant scDAOCS sequences 4.4. Heterologous expression and purification of wild-type and mutant scDAOCSs Chapter 5 A complete library of amino acid alterations at R306 in scDAOCS demonstrates its structural role in the ring- expansion activity 127-137 5.1. A complete library of amino acid alterations at R306 of scDAOCS

5.2. The role of amino acid residue R306 in the activity of scDAOCS Chapter 6 Relevance of single mutations at amino acid positions 275, 281, 304 and 305 of scDAOCS leading to improved penicillin conversion activities 138-153 6.1. Kinetic studies of N304 mutant scDAOCSs 6.2. The importance of amino acid position 304 of scDAOCS 6.3. Relevance of V275I, C281Y and I305M single mutations leading to improved penicillin conversion activities 6.4. The importance of the small sub-domain where V275 and C281 reside Chapter 7 Relevant double mutations of scDAOCS result in higher binding specificities which improve penicillin bioconversion 154-159 7.1. Double mutation analyses of scDAOCS 7.2. Combinatory mutagenesis as an approach in understanding the substrate binding mechanism of scDAOCS Chapter 8 Characterisation of scDACS 160-177 8.1. Cloning of scDACS 8.2. Heterologous expression of recombinant scDACS 8.3. Functional and mutational studies of recombinant scDACS 8.4. Proposed differential substrate recognition of scDAOCS and scDACS Chapter 9 Bioactivities of expanded cephalosporin products 178-190 9.1. Antimicrobial spectra of expanded cephalosporin products 9.2. Efficacies of penicillin substrates and their respective expanded cephalosporin moieties 9.3. Assessing additional properties conferred on expanded cephalosporin products: inhibition of elastase 9.4. Potential applications of expanded cephalosporin products Chapter 10 Concluding remarks and future directions 191-193

References 194-219

Appendices 220-224 List of Publications 225-227

iii

Summary

Streptomyces clavuligerus deacetoxycephalosporin C synthase (scDAOCS) is an important industrial enzyme for the production of 7- aminodeacetoxycephalosporanic acid, a precursor for the production of semi-synthetic cephalosporins. However, the substrate affinity of scDAOCS, in the conversion of penicillins to corresponding expanded cephalosporins, has limitations. Thus, the main objective in this study was to manipulate scDAOCS via understanding the molecular basis of its functionality and to forge it to accommodate a wider spectrum of substrates for enhanced conversion into new antimicrobial products. To this end, a cohort of site-directed mutated scDAOCS proteins were made and investigated for changes to their prowess in transforming penicillins, including ampicillin, carbenicillin and phenethicillin, to cephalosporin moieties. Results of the study on the involvement of amino acid residues, Y184, L186, V275, C281, W297, I298, V303 and the C-terminal residues between N304-R307, were found to be most prominent in affecting the entry of selected substrates, enhancing substrate binding affinities and improving catalytic activities up to 283% of the wild-type enzyme activity . As such, the possibility of transforming existing penicillins into new cephalosporins, as potential antibacterials, looks promising.

To understand the effects of combining various beneficial mutations, selected mutations were also paired and analysed. The bioassay results showed that the C- terminal mutations, i.e., N304A, N304L, N304M, N304K, N304R, I305M, R306L and R307L, in combination with C281Y or V275I substantially increased the catalytic activities further by up to 1,347% of the wild-type enzyme activity. Based on these

studies, it was clear that double mutants had enhanced substrate binding affinities compared to single mutants.

The understanding of the substrate binding mechanism of scDAOCS was also exploited for the investigation of the substrate selectivity of another biosynthetic enzyme, deacetylcephalosporin C synthase (DACS), which catalyses a hydroxylation reaction. Genetic modifications of its C-terminal residues T308 and H310 conferred a ring-expansion activity to provide this additional function to DACS. This finding suggested that the C-terminal of DACS is likely to be involved in substrate selectivity leading to the type of enzymatic activity it catalyses. This study supports the implication of mobilizing the substrate binding properties of biosynthetic enzymes to tailor desired bioactive products.

List of Tables

Chapter 1 Table 1.1. Kinetic parameters of wild-type scDAOCS in penicillin N and penicillin G conversions...... 14 Chapter 2 Table 2.1. Major classes of antibiotics...... 20 Table 2.2. Examples of industrial application of enzymes...... 23 Table 2.3. Summary of DAOCS-related patents...... 55 Table 2.4. A range of penicillin substrates catalysed by scDAOCS...... 59 Table 2.5. DAOCS orthologues cloned and characterised...... 61 Table 2.6. Crystal structures of DAOCS deposited in PDB...... 68 Table 2.7. Summary of annotated deacetylcephalosporin C synthases...... 69 Table 2.8. Relative ring-expansion and hydroxylation activities of wild-type and mutant DAOC/DACSs using penicillin G and DAOC as respective substrates...... 74 Table 2.9. Examples of structure-aided mutagenesis studies in improving enzymes for industrial applications...... 78 Table 2.10. Kinetic parameters of bioengineered DAOCSs in penicillin G conversion...... 83 Table 2.11. Summary of mutagenesis studies of scDAOCS substrate binding pocket...... 88 Chapter 3 Table 3.1. List of bacterial strains used...... 89 Table 3.2. List of plasmid vectors used...... 91 Table 3.3. Mutagenic primers used in generating scDAOCS mutants...... 92 Table 3.4. List of oligonucleotide primers used in gene amplification and cycle sequencing...... 93 Table 3.5. List of test kits used...... 94 Chapter 4 Table 4.1. Ligand-Protein Contacts analyses of scDAOCS complexed with (a) penicillin G (1UOB) and (b) ampicillin (1UNB)...... 122 Table 4.2. Comparison between the predicted substrate binding residues in scDAOCS crystal structures complexed with penicillin G (1UOB) and ampicillin (1UNB)...... 123 Chapter 5 Table 5.1. Relative activities of the wild-type and mutant scDAOCSs determined by hole-plate bioassay...... 130

Table 5.2. Specific activities of the wild-type and mutant scDAOCSs (R306I, R306L and R306M) for ampicillin conversion determined by HPLC analyses...... 132 Table 5.3. Kinetic parameters of the wild-type and mutant scDAOCSs (R306I, R306L and R306M) for ampicillin conversion...... 132 Chapter 6 Table 6.1. Kinetic parameters of wild-type and N304 mutant scDAOCSs for ampicillin conversion...... 139 Table 6.2. Relative activities of wild-type and mutant scDAOCSs determined by bioassay...... 146 Table 6.3. Specific and relative activities of wild-type and mutant scDAOCSs for ampicillin conversion determined by HPLC...... 147 Table 6.4. Kinetic parameters of wild-type and mutant scDAOCSs for ampicillin conversion...... 148 Table 6.5. Kinetic studies of wild-type and mutant scDAOCSs in penicillin N and penicillin G conversions...... 152 Chapter 8 Table 8.1. Sequence homologies between various annotated DAOCSs and DACSs...... 167 Table 8.2. Specific and relative activities of scDAOCS, wild-type and mutant scDACSs in penicillin G conversion determined using a hole-plate bioassay...... 172 Table 8.3. Relative ring-expansion and hydroxylation activities of wild-type and mutant AchDAOC/DACSs using penicillin G and DAOC as respective substrates...... 176 Chapter 9 Table 9.1. Affinities of various penicillins and cephalosporins for the PBPs of E. coli DC0...... 190 Chapter 10 Table 10.1. Kinetic parameters of wild-type and mutant scDAOCSs in penicillin N, penicillin G and ampicillin conversions...... 193

vii

List of Figures

Chapter 1 Figure 1.1. Newly developed cephalosporins...... 6 Figure 1.2. Basic chemical structures of penicillin (6-APA) and cephalosporin (7-ADCA) nuclei...... 9 Figure 1.3. Penicillin N is converted to DAOC by DAOCS in the β-lactam biosynthesis pathway...... 14 Figure 1.4. DAOCS is converted to DAC by DACS in the β-lactam biosynthetic pathway...... 16 Figure 1.5. Overall scheme of work for this project...... 18 Chapter 2 Figure 2.1. Basic structure of various β-lactam classes...... 27 Figure 2.2. A reaction scheme for the kinetic interaction of a PBP with a β- lactam...... 34 Figure 2.3. Crystal structures of S. pneumoniae PBPs...... 34 Figure 2.4. Crystal structures and active site residues of (A) S. pneumoniae PBP2x (1QME) and (B) E. coli TEM-1 β-lactamase (1BTL)...... 35 Figure 2.5. Chemical structures of penicillin nucleus 6-APA and major groups of semi-synthetic penicillins...... 40 Figure 2.6. Chemical structures of cephalosporin nucleus 7-ADCA and different generations of semi-synthetic cephalosporins...... 44 Figure 2.7. Industrial production of cephalosporin precursors, 7-ACA and 7- ADOCA...... 47 Figure 2.8. Industrial production of cephalosporin precursor, 7-ADCA...... 47 Figure 2.9. Biosynthesis of penicillin, cephalosporin C, cephamycin C and cephabacin in respective microorganisms...... 53 Figure 2.10. Multiple sequence alignment of annotated DAOCSs from S. clavuligerus NRRL3585 (Scl), S. chartreusis 102SH3 (Sch), S. ambofaciens 29SA4 (Sam), S. jumonjinensis NRRL5741 (Sju), N. lactamdurans (Nla), L. lactamgenus YK90 (Lla) and A. chrysogenum (Ach)...... 62 Figure 2.11. Crystal structure of scDAOCS (PDB: 1UNB)...... 67 Figure 2.12. Cephamycin C biosynthetic gene cluster of S. clavuligerus NRRL3585...... 71 Figure 2.13. Spatial locations of useful mutations in scDAOCS (PDB: 1UNB) identified by using directed evolution approaches...... 85 Chapter 3 Figure 3.1. Inoculation format of test bacteria for hole-plate bioassay...... 110

viii

Chapter 4 Figure 4.1. Spatial location of amino acid residues V275, C281, N304, I305, R306 and R307 in scDAOCS (1UNB)...... 117 Figure 4.2. Schematic diagrams of scDAOCS substrate binding pocket bound with (a) penicillin N and (b) penicillin G...... 120 Figure 4.3. Close-up view of substrate binding residues in scDAOCS crystal structures complexed with (a) penicillin G (1UOB) and (b) ampicillin (1UNB)...... 120 Figure 4.4. LIGPLOT analyses of scDAOCS crystal structures complexed with (a) penicillin G (1UOB) and (b) ampicillin (1UNB)...... 121 Figure 4.5. SDS-PAGE analysis of cell-free extracts and purified fractions of E. coli BL21 (DE3) clone expressing recombinant wild-type and mutant GST-scDAOCSs...... 126 Chapter 5 Figure 5.1. HPLC chromatogram of wild-type scDAOCS reaction (red) vs time 0 (blue) for ampicillin conversion...... 131 Figure 5.2. Close-up view of residue R306 (purple) surrounded by five hydrophobic residues Y184, L186, W297, I298, and V303 (green).... 137 Figure 5.3. Part of the multiple alignments of annotated DAOCS primary sequences...... 137 Chapter 6 Figure 6.1. Schematic diagrams of predicted residues interacting with ampicillin substrate in scDAOCS (1UNB) using SwissPdb Viewer (left column) and LIGPLOT program (right column)...... 140 Figure 6.2. Schematic diagrams of predicted residues interacting with ampicillin substrate in scDAOCS (PDB accession number: 1UNB) using SwissPdb Viewer (left column) and LIGPLOT program (right column)...... 149 Figure 6.3. Spatial orientation of residues V275 and C281 (orange) in scDAOCS (PDB accession number: 1UNB)...... 150 Figure 6.4. Spatial location of small sub-domain where residues V275 and C281 reside...... 153 Chapter 8 Figure 8.1. Agarose gel electrophoresis of PCR-amplified product of cefF gene from S. clavuligerus genomic DNA...... 162 Figure 8.2. Verification of PCR-amplified cefF gene of S. clavuligerus NRRL 3858...... 162 Figure 8.3. SDS-PAGE analysis of cell-free extracts and purified fractions of E. coli BL21 (DE3) clone expressing recombinant GST-scDACS...... 163 Figure 8.4. Multiple sequence alignment of annotated DAOCSs and DACSs of various microorganisms depicted with the secondary structure of scDAOCS...... 166 Figure 8.5. Crystal structure of scDAOCS (1UOB) and homology model of scDACS...... 168 ix

Figure 8.6. C-termini of scDAOCS crystal structure (1UOB) and scDACS in silico model...... 168 Figure 8.7. Superimposition of three crystal structures of scDAOCS complexed with ampicillin (brown) (1UNB), penicillin G (orange) (1UOB) and DAOC (gold) (1UOG)...... 177 Figure 8.8. Spatial orientation of the methyl group (circled) at the C7 position of dihydrothiazine ring of DAOC. It is approximately 4.5Å away from the iron centre of scDAOCS...... 177 Chapter 9 Figure 9.1. Antimicrobial spectra of (A) ampicillin and cephalexin, (B) penicillin G, (C) carbenicillin, (D) phenethicillin and their respective ring-expanded cephalosporin products...... 181 Figure 9.2. Efficacies of penicillin G, ampicillin, carbenicillin, phenethicillin and their respective ring-expanded cephalosporin products against (A) untransformed E. coli BL21 (DE3) and (B) transformed E. coli BL21 (DE3) expressing ampicillin-resistance gene...... 184 Figure 9.3. Elastase inhibitory assay of penicillins and their respective expanded cephalosporin moieties...... 186

List of Abbreviations

ACA aminocephalosporanic acid ACV δ-(L-α-aminoadipyl)-L-cysteinyl-D-valine ACVS ACV synthetase ADCA aminodeacetoxycephalosporanic acid ADOCA aminodeacetylcephalosporanic acid aIPNS Aspergillus nidulans isopenicillin N synthase APA aminopenicillanic acid APS ammonium persulphate ATP adenosine triphosphate DAC deacetylcephalosporin C DACS deacetylcephalosporin C synthase DAOC deacetoxycephalosporin C DAOCS deacetoxycephalosporin C synthase DMSO dimethyl sulphoxide DNA deoxyribonucleic acid dNTP deoxyribonucleotide triphosphate DTT dithiothreitol EDTA ethylenediamine tetra-acetic acid GST glutathione S-transferase HPLC high performance liquid chromatography IPN isopenicillin N IPNS isopenicillin N synthase IPTG isopropyl-β-D-thiogalactopyranoside LB Luria-Bertani NRPS non-ribosomal peptide synthetase O.D. optical density OL oligonucleotide PBP penicillin-binding protein PCR polymerase chain reaction PKS polyketide synthase scDACS Streptomyces clavuligerus deacetylcephalosporin C synthase scDAOCS Streptomyces clavuligerus deacetoxycephalosporin C synthase SDM site-directed mutagenesis SDS-PAGE sodium dodecyl sulphate-polyacrylamide gel electrophoresis SOPMA self-optimised prediction program TEMED N, N, N’, N’-tetramethyl-ethylenediamine Tris Tris-(hydroxymethyl)-aminomethane UV ultraviolet

List of Units

Å Ångström bp base pair ºC degree Celsius Da Dalton g gram h hour kbp kilobase pair kDa kiloDalton L liter M molar Mbp megabase pair mg milligram min minute mL milliliter mm millimeter mM millimolar ng nanogram nm nanometer nmol nanomole pmol picomole rpm revolution per minute s second µg microgram µL microliter µm micrometer µmol micromole U units V volt % (w/v) gram per 100 milliliter % (v/v) milliliter per 100 milliliter

xii

Chapter 1: Introduction

1.1. Antibiotics and emerging resistant pathogens

Antibiotics are widely known as compounds produced by microorganisms to inhibit the growth of other organisms in their environment (Davies, 1990; Davies,

2006; Waksman, 1947; Yim et al., 2006). The first and most well-known antibiotic is penicillin, which was serendipitously discovered and reported by Sir Alexander

Fleming in 1928 (Mann, 2004; Demain and Elander, 1999). However, it was only successfully purified from the Penicillium culture by Sir Howard Florey and Sir Ernst

Chain in 1942. These three scientists were awarded the Nobel Prize (Physiology or

Medicine) in 1945 “for the discovery of penicillin and its curative effect in various infectious diseases” (Levinovitz and Ringertz, 2001). Due to the many epidemics like

“Black Death” that took numerous lives in the pre-antibiotic era, the successful treatment of several infectious diseases, e.g., meningitis, tuberculosis, gonorrhoea and pneumonia, through the use of antibiotics resulted in them being labelled as “miracle drugs” in the late 1940s (Mann, 2004).

From the discoveries of “pioneer” antibiotics, such as penicillin and streptomycin, to the current development of semi-synthetics, countless lives have been saved from infectious diseases. However, the use of these antibiotics to inhibit and kill the microorganisms has concomitantly selected for pathogens that become resistant to them. It was clear then that “acquired resistance to penicillins” was a new challenge to overcome (Chain, 1972; Kirby, 1944). However, this was ignored or overlooked probably due to the lack of understanding in the molecular biology of the

bacteria that produced the antibiotics and of the mechanisms negating the efficacies of antibiotics.

1.1.1. Paradox of antibiotics

Due to the prevalent use of antibiotics, many resistant pathogens have emerged and colonised various habitats (French, 2005; Jones and Pfaller, 1998). Of great concern are the nosocomial infections. Instead of receiving treatments for their illnesses in the hospital and recovering from them, many patients were infected by these resistant pathogens in hospitals, with cases resulting in mortality. These pathogens were collectively known as the ESKAPE pathogens, which refer to

Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species (Boucher et al.,

2009). Incidences of infections by these bacteria are becoming increasingly common, with these bacteria also developing greater resistance to existing antibiotics (Boucher et al., 2009; Rice, 2008). For example, a study was conducted by the Asian Network for Surveillance of Resistant Pathogens to characterise the macrolide resistance mechanism of Streptococcus pneumoniae isolates collected from 10 Asian countries

(Korea, China, Hong Kong, Thailand, Taiwan, India, Sri Lanka, Singapore, Malaysia and Vietnam) between 1998 and 2001, which reported that about 60% of the 555 isolates studied were resistant to erythromycin. Vietnam, Taiwan, Korea, Hong Kong and China had very high incidences of erythromycin-resistant S. pneumonia, with more than 70% of the isolates from these countries being highly resistant to erythromycin (Song et al., 2004). Both Singapore and Malaysia were reported to have increased erythromycin resistance based on comparing the data collected from 1996 to 2001 (Song et al., 2004). However, it was not clear whether the high prevalence of 2

erythromycin resistance in these Asian countries was due to drug selection through the use of erythromycin.

Other than the selection of resistant bacteria through the use of antibiotics, globalisation also played an important part in disseminating these resistant pathogens around the world (So et al., 2010). Medical tourism is also likely to promote the dissemination of resistance. Recently, a Swedish patient was reported to be infected with a carbapenem-resistant Klebsiella pneumoniae, which harboured a new metallo-

β-lactamase gene blaNDM-1, after travelling to India for medical treatment (Yong et al.,

2009). This gene encodes a metallo-β-lactamase that degrades carbapenems, which are often regarded as the last line of defence against multi-drug resistant pathogens

(Wilcox, 2009; Schwaber and Carmeli, 2008). Subsequent reports on the detection of this metallo-β-lactamase in other countries followed (Tijet et al., 2011; Poirel et al.,

2011; Mulvey et al., 2011; Pfeifer et al., 2011; Samuelsen et al., 2011; Moellering,

2010). This has continued to raise concerns on the prevalent use of antibiotics and the selection of strains resistant to the antibiotics used (Bush, 2010). However, the correlation between the use of antibiotics and the selection of antibiotic resistance in different bacterial species is influenced by subjective factors and cannot be easily established. To this end, a study was conducted by the Network for Antimicrobial

Resistance Surveillance (Singapore) to determine the correlation between antibiotic prescription and antibiotic resistance of Gram negative bacteria in Singapore hospitals. This study showed that there was a correlation between the prescription of several key antibiotics and the resistance of Gram negative bacteria but not for cephalosporins (Hsu et al., 2010).

1.1.2. Efforts in generating “better” antibiotics

Since the discovery of penicillins as “miracle drugs”, antagonism against these antibiotics was not unknown (Chain, 1972; Kirby, 1944). Failures in treatments were not uncommon. With the advancement in research and technology, the mechanism of antibiotic resistance becomes clearer and we are better equipped with the knowledge of countering antibiotic resistance. Global efforts of genome sequencing and proteome analysis of many bacterial pathogens have also revealed more targets for drug intervention (Moellering, 2011; Coates et al., 2002). Retrospectively, the completion of genome sequencing of various actinomycetes, a family of well-known antibiotic producers, is likely to aid in the production of existing and novel antibiotics

(Katz and Khosla, 2007). For example, the completion of genomic sequencing of

Streptomyces coelicolor (Bentley et al., 2002), S. avermitilis (Ikeda et al., 2003;

Omura et al., 2001), S. clavuligerus (Song et al., 2010) and S. griseus (Ohnishi et al.,

2008) revealed a number of biosynthetic gene clusters of previously uncharacterised antibiotics. This has allowed genetic manipulation of various gene clusters to produce new antibiotics (Hranueli et al., 2005). By studying and understanding the biosynthetic enzymes of antibiotics, more opportunities will also be available for modifying the chemical structure of antibiotics to improve their efficacies.

With the large number of different antibiotic classes already discovered, the challenge of finding new antibiotic classes becomes greater. By understanding the structure-activity relationship of an antibiotic, it will be easier to modify the important regions of the antibiotic that could improve its efficacy and reduce its toxicity to humans. This has been applied to many natural antibiotics to improve their antimicrobial spectra and reduce their toxicities (Newman and Cragg, 2009; Newman,

2008). 4

Despite the finding that the prevalent use of antibiotics has resulted in the selection for resistant pathogens, some even harder to treat, antibiotics remain the most important “weapons” to fight against them (Demain and Sanchez, 2009). Hence, existing antibiotics need to be replaced by more efficacious ones. Historically, as soon as a new antibiotic became available for clinical use, resistant bacteria would subsequently emerge (Kirst, 2010). This has dampened the drive for many pharmaceutical companies to develop better antibiotics (Demain and Sanchez, 2009;

Vicente et al., 2006; Projan, 2003). However, recent efforts in modifying cephalosporins have continued to shed some hope in the fight against β-lactam- resistant bacteria (Figure 1.1). Two cephalosporins, namely ceftobiprole (Llarrull et al., 2009; Deresinski, 2008) and ceftaroline (Llarrull et al., 2009; Villegas-Estrada et al., 2008), were developed and shown to be efficacious against methicillin-resistant

Staphylococcus aureus (MRSA). Ceftobiprole has also been demonstrated to have activity against many Gram-negative bacteria in vitro, including P. aeruginosa

(Barbour et al., 2009; Zhanel et al., 2008). Other cephalosporin derivatives, with structurally different C3 side chains from ceftobiprole and ceftaroline, were also synthesised and demonstrated to be effective against MRSA and vancomycin-resistant

S. aureus (Fuda et al., 2006). By modifying the C3 side chain of ceftazidime, a new cephalosporin CXA-101 has been demonstrated to resist the action of P. aeruginosa

AmpC β-lactamase (Murano et al., 2008; Toda et al., 2008). More recently, it was also tested to be effective against various β-lactam-resistant strains of P. aeruginosa

(Zamorano et al., 2010; Livermore et al., 2009; Giske et al., 2009).

OH O N H N N S N O O H N N O N 2 N O N S O O O O N COOH H OH N S N N Ceftobiprole medocaril HO P N H N O N S S S O F O O N COOH H Ceftaroline fosamil N S N NH H N 2 N O N N S N N O F O N COOH NH H ABRI-1175 N S NH2 N NH O F H N 2 N O N N N S N N H O H N S N N ABRI-2901 COOH H N 2 N O N N S O N N COOH H2N ABRI-1974 H3C COOH H C H C 3 NH 3 COOH H N 2 O O H3C N NH O H N S N N H NH2 H2N N N S N O N N NH S N H N 2 O 2 N O N N CH3 S N COOH O CH3 FR259647 COOH CXA-101 (FR264205)

Figure 1.1. Newly developed cephalosporins. Ceftobiprole medocaril, ceftaroline fosamil, ABRI-1175, ABRI-2901 and ABRI-1974 are anti-MRSA cephalosporins. FR259647 and CXA-101 are anti-pseudomonal cephalosporins.

1.2. β-Lactam antibiotics

It has been estimated that the global antibiotics market generated sales of

US$42 billion in 2009. The class of antibiotics that forms the largest proportion of total sales is the cephalosporins, followed by the penicillins, generating US$11.9 billions and US$8.8 billions, respectively. They made up almost 50% of the antibiotics market, illustrating the usefulness and relevance of these related classes of

β-lactam antibiotics (Hamad, 2010).

1.2.1. Penicillin and cephalosporin

Penicillin and cephalosporin antibiotics have been very successful as chemotherapeutics due to their high target specificity and low toxicity to humans

(Demain and Elander, 1999). These are partly due to the attributes of their drug targets, penicillin binding proteins (PBPs). The PBPs, which are responsible for the synthesis of the bacterial cell wall, are not found in mammalian cells (Bugg et al.,

2011; Schneider and Sahl, 2010). Many early penicillin and cephalosporin derivatives were developed with wide antimicrobial spectrum for clinical applications (Silver,

2007; Vicente et al., 2006). However, due to the prevalent use of these antibiotics, resistant bacteria have emerged, leading to the increase in nosocomial infections

(French, 2005; Jones and Pfaller, 1998).

Both penicillin and cephalosporin are important classes of antibiotics

(Collignon et al., 2009). However, the structural relationship between penicillin and cephalosporin was only realised in the late 1950s, about a decade after the discovery of cephalosporin C in Cephalosporium acremonium (presently known as Acremonium chrysogenum) (Hamilton-Miller, 2000; Abraham and Newton, 1961). Penicillin and cephalosporin possess a four-membered β-lactam ring that binds covalently with 7

nucleophilic active site serine residues in bacterial cell wall biosynthetic enzymes, leading to their inactivation and indirectly, lysis of the bacterial cell walls (Zapun et al., 2008a; Essack, 2001; Ghuysen, 1991). The basic difference between the nuclei of penicillin and cephalosporin lies in the additional cyclic structure fused to the β- lactam ring, i.e., penicillin has a five-membered thiazolidine ring and cephalosporin has a six-membered dihydrothiazine ring (Figure 1.2). This difference allows the latter to evade destruction by penicillinase, which is produced by penicillin-resistant bacterial strains. Thus, the cephalosporins are generally considered to have a broader antimicrobial spectrum than the penicillins (Marshall and Blair, 1999). In addition, the cephalosporin nucleus has two sites for modification, as compared to the penicillin nucleus which has only one, thus enabling the synthesis of a wider variety of cephalosporins.

The production of cephalosporin derivatives stemmed from the feasibility to synthesise the penicillin nucleus, 6-aminopenicillanic acid (6-APA), and to add different side chains to this nucleus in order to create a variety of penicillin derivatives with medicinal value (Hamilton-Miller, 2008, 2000). Hence, it was clear that progressive development of various cephalosporin derivatives as useful therapeutics depended on the ability to obtain the cephalosporin nucleus. This was probably one of the factors that initiated interest in deciphering the biosynthetic pathway of cephalosporin. The availability of the cephalosporin nuclei, 7- aminodeacetoxycephalosporanic acid (7-ADCA) and 7-aminocephalosporanic acid

(7-ACA), led to the production of many semi-synthetic cephalosporins which differ widely in antimicrobial spectrum and pharmacokinetics (Barber et al., 2004; Elander,

2003; Demain and Elander, 1999; Marshall and Blair, 1999). However, these semi- synthetic cephalosporins are expensive due to high production cost. For example, 7- 8

ADCA is synthesised by multi-step chemical ring-expansion of penicillin G, which is an expensive and polluting process (Velasco et al., 2000). This has led to serious considerations for synthesis of the cephalosporin nucleus by enzymatic processes, which are cleaner and more economical (Bruggink et al., 2003; Wegman et al., 2001).

H 1 H 1 N N 7 6 S 6 5 S 2 R 2 CH3 R1 N CH N 3 7 3 3 8 O 4 O 5 4 R2 COOH COOH 6-APA 7-ADCA

Figure 1.2. Basic chemical structures of penicillin (6-APA) and cephalosporin (7- ADCA) nuclei.

1.2.2. Chemotherapeutic and non-medicinal applications of cephalosporins

The initial interest on cephalosporin C was mainly due to its ability to inhibit some Gram negative bacteria and simultaneously resist the action of penicillinase.

Furthermore, with the ability to modify the structure of cephalosporin to improve its efficacy, a plethora of cephalosporin derivatives was developed and marketed. It was estimated that at least 50 different cephalosporins were available in the market by

2003 (Hamilton-Miller, 2008, 2000; Elander, 2003). They have been classified based on their antimicrobial properties and grouped into generations, which is considered more useful for clinical prescription. The first generation cephalosporins are made up of cephalosporins that have similar antimicrobial spectra as cephalexin. Cephalexin is one of the earlier semi-synthetic cephalosporin available and has the same side chain as ampicillin. Its antimicrobial spectrum encompasses that of ampicillin, except that its resistance to penicillinase bestowed upon itself a wider spectrum than ampicillin

(Hansch et al., 1990). The next two generations of cephalosporins have a wider 9

spectrum against many Gram negative bacteria and are more resistant to the evolving

β-lactamases, but with decreasing inhibitory activity against Gram positive bacteria.

In contrast to the first three generations, fourth-generation cephalosporins have a broader antimicrobial spectrum, encompassing those of the first generation cephalosporins and many Gram negative bacteria, including pathogenic P. aeruginosa. They are also generally more resistant to the evolved β-lactamases active against earlier generations of cephalosporins (Marshall and Blair, 1999).

Other than using cephalosporins as antibacterial drugs, a recent screening effort has also identified ceftriaxone and other β-lactams as activators for the expression of the excitatory amino acid transporter in humans, which is important for preventing glutamate neurotoxicity (Rothstein et al., 2005). This has invoked intensive discussion on the “re-discovery” of other possible applications of β-lactam antibiotics, with the obvious advantage that many of these β-lactam antibiotics have already been demonstrated to be non-toxic to humans (Secko, 2005; Beghi et al.,

2005; Eisenstein, 2005; Mao, 2005; Miller and Cleveland, 2005). Another chemotherapeutic application of cephalosporins was as inhibitors against serine proteases, which have been implicated in certain diseases, e.g. human elastase in pulmonary emphysema (Doherty et al., 1986). Since they are acylating inhibitors of bacterial serine proteases (PBPs), it is very likely that they could also inhibit mammalian serine proteases (Veinberg et al., 2003; Davies et al., 1991; Doherty et al.,

1986). In fact, various cephalosporin derivatives have been synthesised and demonstrated to inhibit human leukocyte elastase (Alpegiani et al., 1994; Knight et al., 1992; Doherty et al., 1986).

A third application of cephalosporins is towards diagnostic and research purposes. Since they are substrates of β-lactamases, they are used for the detection of 10

β-lactamase-producing pathogens in hospitals and also β-lactamase reporter activity in protein complementation assay for protein-protein interaction studies. This can be achieved by attaching fluorogenic or chromogenic moieties to the C3 or C7 positions of the cephalosporin core structure (Remy et al., 2007).

1.2.3. Industrial production of cephalosporins: a move towards “Green

Chemistry”

Due to the difficulties in synthesising cephalosporin from the three amino acid precursors in vitro, it is still relevant to obtain cephalosporins from the fermentation culture of producing organisms. Many existing semi-synthetic cephalosporins are produced using 7-ACA or 7-ADCA as a precursor. 7-ACA and 7-ADCA are derived from deacylation of cephalosporin C and G-7-ADCA, respectively (Elander, 2003;

Demain and Elander, 1999). Industrial production of cephalosporin C involves the fermentation of Acremonium chrysogenum. However, the problem with cephalosporin

C fermentation is the intrinsic chemical instability of cephalosporin C. It is readily degraded in long-cycle cephalosporin C fermentations, which can account up to 40% loss of the cephalosporin C produced (Elander, 2003; Usher et al. 1988). Hence, chemical ring-expansion of penicillin G was deemed as an alternative since penicillin

G is cheap and easily available from the fermentation culture of Penicillium chrysogenum (Wegman et al., 2001). Silyl protection chemistry was used to chemically ring-expand penicillin G to produce G-7-ADCA. However, this method involves multiple steps, is expensive and polluting (Velasco et al., 2000). An alternative “greener” method to synthesise G-7-ADCA would be to exploit the cephalosporin synthesising enzyme from the natural biosynthetic pathway of cephalosporin. 11

1.3. Biosynthesis of cephalosporins

1.3.1. Deacetoxycephalosporin C synthase: ring-expansion of penicillin to cephalosporin

Deacetoxycephalosporin C synthase (DAOCS) (EC number 1.14.20.1) is the key enzyme in the β-lactam biosynthetic pathway. It is an iron-dependent and α- ketoglutarate-requiring dioxygenase encoded by cefE gene. It catalyses the ring- expansion of penicillin N to DAOC and hence, it is also known as expandase (Figure

1.3). It was previously thought to have a very narrow substrate specificity, in that it can only convert its natural substrate penicillin N but not other penicillin analogues

(Maeda et al., 1995; Dotzlaf and Yeh, 1987). However, Cho and co-workers (1998) derived a new reaction condition that allows the conversion of penicillin G and 14 other penicillins into cephalosporins. Subsequently, Sim and Sim (2001) further refined the catalytic conditions by omitting superfluous reagents, resulting in higher enzymatic activity. Hence, this optimised reaction condition has provided the milieu for extensive studies on DAOCS and its possible application in the industry for producing “better” cephalosporins.

Although studies have shown that DAOCS possesses wide substrate specificity (Chin et al., 2003; Cho et al., 1998), its catalytic efficiency in converting penicillin G is rudimentary (Table 1.1) (Wei et al., 2005; Hsu et al., 2004). So far,

Streptomyces clavuligerus DAOCS (scDAOCS) is the most active enzyme in converting penicillin G, compared to orthologues from other organisms (Hsu et al.,

2004; Chin et al., 2003). Various studies have suggested that the C-terminal of scDAOCS is important in guiding and orientating the penicillin substrate into its catalytic pocket (Chin and Sim, 2002; Lee et al., 2002; Lee et al., 2001). However, the 12

molecular mechanism of this interaction is still not clear. Mutagenesis studies have indicated that mutations of C-terminal residues, e.g. N304L, I305M, R306L and

R307L, improved the penicillin G conversion activity of the enzyme (Wei et al., 2003;

Chin and Sim, 2002; Lee et al., 2001; Lloyd et al., 1999). The rationale behind these mutations could not be comprehended in the absence of a crystal structure complex with a penicillin substrate (Lee et al., 2001; Valegård et al., 1998). Fortunately, such a structure has been timely resolved (Valegård et al., 2004) for this study and has provided the opportunity to virtually view the catalytic properties of this enzyme and understand its substrate binding mechanism, for the prediction and improvement of its catalytic functions via protein engineering (Barends et al., 2004). To achieve this, site- directed mutagenesis (SDM) was used to construct novel proteins with desired properties, such as enhanced activity and/or alteration in substrate specificity, for industrial applications. For example, Lee et al. (2001a) showed that a single replacement of R258 with glutamine in scDAOCS altered its co-substrate specificity.

This technique has shaped the basis of protein engineering and has been applied widely. Thus, in this study, scDAOCS was targeted as a potential platform for bettering cephalosporins.

H H HOOC N S HOOC N S DAOCS

NH2 O N NH2 O N O α-KG, Succinate, O C OOH O CO , H O COOH Penicillin N 2 2 2 DAOC

Figure 1.3. Penicillin N is converted to DAOC by DAOCS in the β-lactam biosynthesis pathway.

Table 1.1. Kinetic parameters of wild-type scDAOCS in penicillin N and penicillin G conversions (Wei et al., 2005).

-1 -1 -1 Substrate Chemical structure Km (mM) kcat (s ) kcat/Km (M s ) Relative kcat/Km H H2N N S CH3

Penicillin N COOH O N CH3 0.014 0.307 22,000 1.0 O COOH H N S CH3 -4 Penicillin G O N CH3 2.58 0.0453 18 8.2x10 O COOH

1.3.2. Deacetylcephalosporin C synthase: hydroxylation of cephalosporin

Deacetylcephalosporin C synthase (DACS) (EC number 1.14.11.26) is the enzyme that catalyses the conversion of DAOC to DAC in the β-lactam biosynthetic pathway of cephalosporin producers (Figure 1.4). It is also known as deacetoxycephalosporin C hydroxylase or 3'-methylcephem hydroxylase (Coque et al., 1996; Baker et al., 1991). Its enzymatic reaction is essentially an oxidation reaction similar to the enzymatic reaction of DAOCS; it is also a 2-oxoglutarate dioxygenase enzyme, which has an absolute requirement for ferrous iron and 2- oxoglutarate (also known as α-ketoglutarate) as co-factor and co-substrate, respectively, for catalysis.

Based on primary sequence alignment, DACS has high resemblance to

DAOCS. For example, S. clavuligerus DACS (scDACS) has been reported to possess

59% and 71% identity with scDAOCS at the protein and nucleotide sequence level, respectively (Kovacevic and Miller, 1991). Hence, it has also been postulated to have arisen from a gene duplication event during the evolution of streptomycetes

(Kovacevic and Miller, 1991). However, studies on this enzyme have been limited due to the difficulty in synthesising the native substrate DAOC. Nevertheless, biochemical studies have revealed its highly specific hydroxylation activity despite sharing high protein sequence identity, similar catalytic platform and reactants with

DAOCS, which demonstrated dominant ring-expansion activity (Baker et al., 1991).

Intriguingly, A. chrysogenum possesses an enzyme with both ring-expansion and hydroxylation activities. It is known as deacetoxy/deacetylcephalosporin C synthase (DAOC/DACS). Although studies on this bifunctional enzyme have revealed some important sites for either ring-expansion or hydroxylation activities (Lloyd et al., 2004), this is not well understood in the monofunctional enzymes. By elucidating 15

the determinants and mechanism behind these structure-function differences in conferring the enzyme with either or both activities, it will become easier to tailor- make the enzyme with desired properties for industrial application.

H H HOOC N S HOOC N S DACS

NH2 O N NH2 O N OH O α-KG, Succinate, O COOH O CO , H O COOH DAOC 2 2 2 DAC

Figure 1.4. DAOCS is converted to DAC by DACS in the β-lactam biosynthetic pathway.

1.4. Objectives of this study

Preceding studies on the characteristics of scDAOCS have exemplified the potential of modifying it to accept a range of industrially preferred penicillin G and other non-native penicillins as substrates. Understanding the mechanisms behind enzyme-substrate interaction and those mutations that led to improved ring-expansion activity of scDAOCS for hydrophobic penicillin conversions will provide a basis for improving scDAOCS, thus leading to a more efficient and cleaner process for the production of cephalosporin and possibly, the generation of novel cephalosporins. The objectives of this project will entail the following:

1. Map the functional domains of scDAOCS involved in substrate binding via

biocomputational tools and predict useful substitutions.

2. Validate the predicted substitutions using SDM and activity studies, and

elucidate the substrate binding mechanism of scDAOCS.

3. Investigate the functionality of S. clavuligerus DACS via biocomputational

analyses, SDM and enzymatic assays using scDAOCS as a model.

4. Evaluate the antimicrobial activities and efficacies of the biotransformed

cephalosporin products

5. Investigate additional properties conferred on the biotransformed

cephalosporin products, e.g., against elastase.

The overall scheme of work for this project is summarised in Figure 1.5.

Understanding the substrate binding mechanism of Streptomyces clavuligerus deacetoxycephalosporin C synthase for biosynthesis of novel cephalosporins

Mapping functional domains of scDAOCS

Multiple sequence alignment

Co-factor binding motif

Co-substrate binding motif

Substrate binding site

ClustalW Substrate binding residues

Ligand-Protein Contacts Hydrophobicity index

LIGPLOT Non-covalent interactions

SwissPdbViewer Solvent accessibility

In silico mutations

Generation and evaluation of mutant scDAOCSs

Bioassay & HPLC analyses: Kinetic analyses: •Relative specific activity •Substrate binding affinity: Km •Bioactivity •Turnover rate: kcat •Authentication of products •Catalytic efficiency: kcat/Km

Mutagenesis studies of Evaluation of biotransformed NH2 H scDACS using scDAOCS cephalosporins N S O N HO O CH3 as a model: COOH

Efficacies & Unusual functions: antimicrobial Elastase inhibition activities: CFU assay assay

Figure 1.5. Overall scheme of work for this project.

Chapter 2: Literature review

2.1. Antibiotics: secondary metabolites of microorganisms

Secondary metabolites are low molecular weight compounds that are non- essential to the growth and proliferation of the host organisms. They are structurally and taxonomically diverse with ambiguous functions, and noticeably characteristic to specific types of organisms (Berdy, 2005).

Although the definition of antibiotic in dictionaries and textbooks may differ slightly, antibiotic is generally considered a secondary metabolite produced by microorganisms that exhibit growth inhibition on other microorganisms. Examples of major classes of antibiotics are summarised in Table 2.1. However, the physiological role or evolutionary biology of these naturally occurring low molecular weight biomolecules remains poorly understood (Davies et al., 2006). On the basis that they exhibit antibiosis at certain concentration, they were deemed as “weapons” produced by the host organism against competitors in the harsh environment to enhance its own survival. However, under normal physiological or environmental conditions, inhibitory concentrations of these naturally occurring antibiotics are seldom achieved

(Davies et al., 2006). Recently, antibiotics at sub-inhibitory concentrations were recognised to have physiological effects, leading to changes in 5-10% of the host’s transcriptome. This concentration-dependent effect has been described as hormesis

(Calabrese and Blain, 2005; Davies et al., 2006). In turn, this has very important implications for the use of antibiotics as chemotherapeutics, particularly in the development and selection of antibiotic resistance (Davies et al., 2006).

Table 2.1. Major classes of antibiotics.

Class of Example of natural Mode of action Producer References antibiotics antibiotic Kanamycin Streptomyces (Okami et al., 1959) kanamyceticus Interfere with protein Aminoglycosides synthesis Gentamicin Micromonospora sp. (Weinstein et al., 1963)

Geldanamycin Streptomyces (BeBoer and Dietz, Ansamycins Inhibit Hsp90 hygroscopicus 1976)

Penicillin G Penicillium chrysogenum (Peterson and Tornqvist, 1956)

Cephalosporin C Acremonium chrysogenum (Smith et al., 1967) Inhibit bacterial β-Lactams peptidoglycan Cephamycin C Nocardia lactamdurans, (Nagarajan et al., biosynthesis Streptomyces clavuligerus 1971; Stapley et al., 1972)

Thienamycin Streptomyces cattleya (Kahan et al., 1979) (carbapenem)

Inhibit bacterial Glycopeptides peptidoglycan Vancomycin Amycolatopsis orientalis (Levine, 2006) biosynthesis

Interfere with protein (Macleod et al., Lincosamides Lincomycin Streptomyces lincolnensis synthesis 1964)

Bind to bacterial cell membrane and cause rapid depolarisation and (McHenney et al., Lipopeptides loss of membrane Daptomycin Streptomyces roseosporus 1998) potential leading to inhibition of protein, DNA and RNA synthesis

Interfere with protein Saccharopolyspora (Vanden Boom, Macrolides Erythromycin synthesis erythraea 2000)

2.1.1. Industrial production of antibiotics: “Green Chemistry”

Some natural antibiotics have limited clinical uses because of undesirable side-effects, poor pharmacokinetics or low antimicrobial activity, especially against resistant pathogens (Chin et al., 2006; Riva, 2001). Hence, modifications were often made to the core structure of the natural antibiotics to improve their antimicrobial or pharmacokinetic properties. These derivatives, including the active pharmaceutical ingredients, are often produced via chemical means using reagents and by-products that are mostly hazardous (Bruggink et al., 2003; Wohlgemuth, 2010).

With increasing knowledge of antibiotic synthesis enzymes, many of these chemical processes could actually be replaced by environmentally friendly enzymatic methods, which are also more specific in their reactions. In turn, this can effectively reduce the number of production steps (Gotor, 2002). For example, in the industrial production of cephalexin, the original chemical procedure requiring ten steps was reduced to six by replacing some steps with enzymatic processes (Gotor, 2002).

The remarkable feature of enzyme is its lately recognised plasticity and amenability to modification in order to alter its functionalities. Many enzymes have the ability to accept a wide range of substrate derivatives, while catalysing specific biochemical reactions. Hence, the use of enzymes in an industrial process has the advantage of avoiding the need for blocking or deblocking steps that are common in many chemical syntheses to prevent side-reactions (Singh et al., 2006). This will result in lower by-product formation and consequently, avoid the need for the purification of the desired products. Hence, enzymatic process is often deemed as a

“greener” alternative to conventional chemical method.

2.1.2. Microbial enzymes for secondary metabolite production and modification

There are several advantages in using enzymes over chemical catalysts.

Enzymes are derived from renewable resources and are biodegradable. These attributes make them an attractive substitute to chemical catalysts. They generally function under relatively mild conditions of pH and temperature, hence, the enzymatic process is generally less hazardous compared to a chemical reaction (Bruggink et al.,

2003; Riva, 2001; Schmidt, 2010; Wohlgemuth, 2010). Furthermore, the reactions they catalyse are often specific and yet, their structural plasticity allows selectivity in both reactant and product stereochemistry. In addition, improvement to substrate selectivity and/or reactivity can be achieved by protein engineering (Bottcher and

Bornscheuer, 2010). With increasing awareness of the effects of industrial pollution to the environment, the need to find a “greener” alternative for the production of chemicals and pharmaceuticals has led to further interest in enzymes (Table 2.2)

(Tucker, 2006; Woodley, 2008).

Secondary metabolites are products of secondary metabolism in the host organism. Hence, biosynthetic enzymes involved in the synthesis of the secondary metabolites and their intermediates are useful molecular tools for industrial application (Riva, 2001). They serve as a readily available source of biocatalysts for biotransformation of preferred industrial substrates into desired products. In fact, with the advent of molecular techniques and protein engineering, enzymes from unrelated biosynthesis of these biomolecules could also be exploited for biotransformation of other compounds by engineering their substrate specificities (Chaput et al., 2008).

Table 2.2. Examples of industrial application of enzymes.

Enzymes Applications References Acyltransferase LovD Conversion of monacolin J to simvastatin (Xie and Tang, 2007)

Amylases Production of high-fructose syrups and ethanol (Pandey et al., 2000) from starch

D-amino acid oxidase Deacylation of cephalosporin C to yield (Pollegioni et al., 2008) cephalosporin nucleus 7-ACA

D-carbamoylase Production of D-p-hydroxyphenylglycine, a side- (Ogawa and Shimizu, chain precursor for semi-synthetic penicillins and 2002) cephalosporins

Lipolase Production of pregabalin, an active (Martinez et al., 2008) pharmaceutical ingredient of Lyrica1® for the treatment of neuropathic pain and epilepsy

Nitrilase Production of acrylamide from acrylonitrile (Singh et al., 2006)

Penicillin acylase Production of penicillin nucleus 6-APA (Sio and Quax, 2004)

Proline 4-hydroxylase Production of trans-4-hydroxy-L-proline, an (Ogawa and Shimizu, important precursor of antiphlogistics, 2002) carbapenem antibiotics, and angiotensin- converting-enzyme inhibitors

Subtilisin Active ingredient of detergents (Gupta et al., 2002)

2.2. Streptomyces species: well known antibiotic producers

Bacterial species from the genus Streptomyces are well known for the diversity of secondary metabolites they produce, particularly antibiotics (Demain,

1999). A survey on the number of antimicrobial compounds discovered in various species of this genus showed a steady increase from the late 1940s and peaked in the

1970s (Watve et al., 2001). However, there was a substantial decline in the late 1980s and 1990s. By assuming that the antibiotic screening efforts were encouraged by a previous year’s success and the probability of finding a new antibiotic is a function of the fraction of antibiotics undiscovered so far, a mathematical model was proposed by

Watve and co-workers (2001) to estimate the total number of antimicrobial compounds Streptomyces species are capable of producing and the number that has yet to be discovered. Based on the model, the number of antibiotics potentially produced by streptomycetes was in the order of 100,000. Comparing to the estimated number of 3,000 natural antibiotics discovered by then, there were many more compounds to be uncovered from streptomycetes. Importantly, they believed that the decline in the number of antimicrobial compounds discovered was due to a decline in screening efforts rather than an exhaustion of compounds (Watve et al., 2001).

2.2.1. Streptomyces clavuligerus NRRL 3585

Streptomyces clavuligerus NRRL 3585 (ATCC 27064) is a Gram positive bacterium belonging to the family Streptomycetaceae (Higgens and Kastner, 1971). It is a soil bacterium isolated in South America. It is named as such due to the morphology of its spore-forming hyphae that resemble little clubs, i.e., clavula in

Latin means little club and the suffix -igerus means bearing. Its spores are grey to

greyish green and it produces at least four antibiotics, namely cephamycin C, clavulanic acid, holomycin and tunicamycin (Song et al., 2010).

S. clavuligerus became an important industrial microorganism because of its ability to produce cephamycin C and clavulanic acid, which is a β-lactamase inhibitor.

Hence, it was also selected for genome sequencing in the hope of providing more information for further research on this bacterium. A draft genome sequence of S. clavuligerus has recently been reported (Song et al., 2010). Its genome consists of a linear chromosome of 6.7 Mbp, which has a G+C content of 72.69%, and four plasmids. Its genome is the smallest among the Streptomyces species whose genomes have been completely sequenced to date. For example, S. coelicolor has approximately 8.7 Mbp (Bentley et al., 2002) and S. avermitilis has about 9 Mbp

(Ikeda et al., 2003). A total of 7,898 protein-coding genes were annotated, contrasting with the 7,825 and 7,574 predicted genes in S. coelicolor and S. avermitilis, respectively (Bentley et al., 2002; Ikeda et al., 2003; Song et al., 2010). With the completion of the whole genome and full annotation of the genes in S. clavuligerus, further advancement in drug discovery and exploitation of this bacterium for antibiotic production could be expected.

2.3. β-Lactam antibiotics

The class of β-lactam antibiotics is made up of structurally diverse molecules possessing a common β-lactam ring. Drug discovery efforts since the serendipitous discovery of penicillin have found many natural products with antibiotic property and a number of them contain a β-lactam ring. These natural products were found to possess structures that are different from penicillin.

2.3.1. General properties and sub-classifications

The β-lactam ring in β-lactam antibiotics is the bioactive portion of the biomolecules. This class of antibiotics mainly interferes with the growth of bacteria by inhibiting the activity of penicillin-binding proteins (PBPs) responsible for bacterial cell wall biosynthesis. These PBPs are bacterial serine DD-peptidases. Since

β-lactams are acylating inhibitors of PBPs, it is likely that β-lactams could also inhibit other serine proteases (Davies et al., 1991; Doherty et al., 1986; Veinberg et al.,

2003). In fact, this is possible by modifying the side chains, leaving the β-lactam ring intact. Hence, the modest growth inhibition observed on other microorganisms, e.g., fungi, was probably due to inhibition of the host’s proteases, leading to growth retardation or even cell death (Sanyal et al., 1992).

Although the mechanism of action for all β-lactam antibiotics have yet to be completely understood, they have been divided into different groups based on their basic core structures (Figure 2.1). Penams are those that have a thiazolidine ring attached to the β-lactam ring, which are mainly the penicillins. Cephems are those with a dihydrothiazine ring attached to the β-lactam ring, which consist of cephalosporin, cephamycin and cephabacin. Clavams are similar to the penams, except that they possess an oxygen atom instead of a sulphur atom. An example of 26

clavams is clavulanic acid, which is a potent β-lactamase inhibitor with low antibacterial activity. Carbapenems also have a similar structure to the penams and clavams, except that they possess a carbon atom instead of a sulphur or oxygen atom in the five-membered ring. Furthermore, they have a double bond between C2 and C3.

For those with a single bond, they are the carbapenams, which are often the intermediates in the carbapenem biosynthetic pathway.

From these natural structural variations, synthetic derivatives were generated, e.g., penem, carbacephem and oxacephem, and assessed for their bioactivities (Figure

2.1). However, due to their poor antimicrobial activities, they were not further developed (Hamilton-Miller, 1999).

Natural β-lactams

S O S

NH N N N N N O O O O O O Monobactam Penam Carbapenam Carbapenem Clavam Cephem

Synthetic β-lactams

S O

N N N O O O Penem Carbacephem Oxacephem

Figure 2.1. Basic structure of various β-lactam classes.

2.3.2. Targets of β-lactams: penicillin binding proteins (PBPs)

PBPs are a group of membrane-bound enzymes involved in bacterial cell wall synthesis. They are named as such due to their ability to bind penicillin although they may play different roles in the process of cell wall synthesis. Collectively, they are involved in the formation of peptidoglycan, which is an important component of the bacterial cell wall in determining the final shape of the host bacteria. A peptidoglycan is made up of glycan chains, where each chain is composed of alternating N- acetylglucosamine and N-acetylmuramic acid, cross-linked by oligopeptides attached to the N-acetylmuramic acid (Sauvage et al., 2008a).

PBPs are classified into three main classes, A, B and C, and are sub-divided within each class based on sequence differences (Sauvage et al., 2008a). Class A and

B PBPs, which are also known as high molecular mass PBPs, are multi-modular enzymes responsible for peptidoglycan formation and insertion into the pre-existing cell wall. Generally, class A PBPs are bifunctional enzymes that possess an N- terminal glycosyltransferase domain, which catalyses the polymerisation of glycan chains, and a C-terminal penicillin-binding domain, which possesses a transpeptidase activity responsible for peptide cross-linking between two glycan chains. Hence, the main function of class A PBPs is the synthesis of peptidoglycan. However, the exact role and significance of each class A PBP in a bacterium possessing more than one class A PBP remains poorly understood although some studies have provided some clues. For example, mutation and deletion studies of the four class A PBPs, namely

PBP1, PBP2c, PBP2d and PBP4, in Bacillus subtilis illustrated functional redundancy of these enzymes (Popham and Setlow, 1996). However, mutation of PBP1 alone increased its average cell length and decreased its cell diameter (Popham and Setlow,

1995; Popham and Young, 2003). Interestingly, when all the class A PBPs were 28

deleted, B. subtilis was still viable and has a relatively normal peptidoglycan structure

(McPherson and Popham, 2003). In contrast, loss of both PBP1a and PBP1b in E. coli is fatal, even though it has a third class A PBP, i.e., PBP1c (Denome et al., 1999).

Like class A PBPs, Class B PBPs also have a C-terminal penicillin-binding domain with transpeptidase activity, while the functions of their N-terminal domains are still unclear although they are believed to be involved in cell morphogenesis by interacting with other proteins in the cell cycle (Macheboeuf et al., 2006; Sauvage et al., 2008a). The function of class B PBPs in a bacterium is mainly cross-linking of the peptidoglycan chains via the stem pentapeptide covalently attached to N- acetylmuramyl portion of the glycan chain, but the exact role of each class B PBP remains vague due to redundancy. Nevertheless, some studies have provided a better understanding of the functions of class B PBPs in cell division. For example, in E. coli, which possesses two class B PBPs, PBP2 has been shown to be important for cell elongation and may have a role in the initiation of cell division (Den Blaauwen et al., 2003; Spratt and Pardee, 1975), whereas PBP3 is involved in septation (Di Lallo et al., 2003; Karimova et al., 2005; Spratt, 1975). In B. subtilis, there are six class B

PBPs, where both PBP2a and PBPH have been shown to play redundant roles in defining its shape (Carballido-Lopez and Formstone, 2007; Wei et al., 2003b).

Class C PBPs are also known as low molecular mass PBPs, containing a penicillin-binding domain with either a DD-endopeptidase or DD-carboxypeptidase activity (Popham and Young, 2003). In general, class C PBPs are involved in the regulation of peptidoglycan formation (Macheboeuf et al., 2006; Morlot et al., 2004).

For example, PBP4 and PBP7 of E. coli are DD-endopeptidases that cleave the peptide cross-bridges between two glycan chains in the peptidoglycan cell wall

(Sauvage et al., 2008a; Vollmer and Holtje, 2004; Vollmer et al., 2008). E. coli PBP5 29

is a DD-carboxypeptidase involved in cleaving the terminal amino acid residue of the stem pentapeptide in a peptidoglycan chain, preventing peptidogylcan cross-linking

(Sauvage et al., 2008a; Spratt and Strominger, 1976).

Most bacterial species have at least one PBP from each class and they are usually named using numbers, followed by an alphabet for sub-classes, if any.

However, PBPs from two different bacterial species with the same numbering may not belong to the same class because they were initially numbered based on their migration patterns when separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) (Sauvage et al., 2008a). For example, Staphylococcus aureus PBP2 encoded by pbp2 is a class A PBP, while E. coli PBP2 encoded by pbpA is a class B PBP. Hence, this can be confusing and a standardised nomenclature will be useful, especially for phylogenetic studies. Interestingly, it has been noted that rod- shaped bacteria tend to have higher total number of PBPs compared to the cocci bacteria and bacteria of similar shapes tend to have similar sets of PBPs (Zapun et al.,

2008b). As such, efforts in understanding the role of each set of PBPs in different bacteria and their phylogenies may help in designing new drugs that are efficacious against multiple PBPs in closely-related bacteria, thus achieving more effective treatment.

2.3.3. Inhibitory action of β-lactam antibiotics and penicillin-binding domains of PBPs

Penicillin and cephalosporin antibiotics are able to inhibit the activity of PBPs because of their structural similarity to the terminal D-Ala-D-Ala stem peptide of the peptidoglycan precursor and they bind to the penicillin-binding domain of PBPs

(Strominger and Tipper, 1965). This subsequently leads to the deregulation of 30

bacterial cell wall synthesis, followed by activation of cell lysis. This hypothesis was strengthened by the recent crystallisation of E. coli PBP6 with a peptidoglycan substrate fragment containing the full pentapeptide, N-acetylmuramyl-(L-Ala-D- isoGlu-L-Lys-D-Ala-D-Ala) (Chen et al., 2009). Superimposition of this structure and

Streptomyces strain R61 DD-peptidase complexed with a non-covalently bound β- lactam (1PW1) revealed a similar spatial orientation of the C3 carboxylate of the β- lactam with that of the terminal D-Ala of the peptidoglycan substrate fragment, which forms hydrogen bonds with residues S106 and T210 of E. coli PBP6. This study has provided empirical evidence for the theory of molecular mimicry of β-lactam antibiotics to the terminal D-Ala-D-Ala moiety of the peptidoglycan in inhibiting bacterial PBPs (Chen et al., 2009).

The inhibitory action of β-lactam antibiotics occurs in three steps and this can be described using the kinetic model depicted in Figure 2.2 (Frere et al., 1975). The formation of a non-covalent complex between PBP and β-lactam is described by the dissociation constant Kd, followed by the formation of the acylated PBP at the rate k2.

Subsequently, deacylation occurs at the rate k3 to yield PBP and an inactivated β- lactam. Often, k3 is slower than k2 and this leads to inhibition of the PBP by the β- lactam since the active site of the PBP is occupied.

The penicillin-binding domain of PBPs bears three conserved motifs, i.e.

SXXK, SXN and K(S/T)G, present in the penicilloyl serine transferase superfamily, which includes class A, C and D β-lactamases (Macheboeuf et al., 2006; Massova and

Mobashery, 1998). Although the primary sequences of the penicillin-binding domain of PBPs are very diverse, the structural folds of these domains are much conserved

(Figure 2.3) (Sauvage et al., 2008a). The penicillin-binding domain is made up of two sub-domains, i.e., one is an anti-parallel five-stranded β-sheets surrounded by three α- 31

helices and another is an all-helical domain (Sauvage et al., 2008a). The SXXK motif bears the active site serine residue responsible for catalysis and is positioned at the beginning of helix α2. The SXN motif is situated in a loop between helices α4 and α5, and the K(S/T)G motif is localised in β3 (Figure 2.4A) (Sauvage et al., 2008a).

Many structural and molecular dynamic simulation studies have been attempted to understand the catalytic mechanism of PBPs upon binding of substrate analogues or β-lactam antibiotics (Contreras-Martel et al., 2009; Massova and

Kollman, 2002; Oliva et al., 2003; Sauvage et al., 2008b; Zhang et al., 2007). The activation of the active site serine by abstraction of a proton is achieved by either the lysine residue of the SXXK motif or the carboxylate group of the β-lactam (Gordon et al., 2000). Recent studies on E. coli PBP5, which is a class C PBP, suggested that the former hypothesis is more likely (Zhang et al., 2007). The reaction begins with the binding of penicillin to the penicillin-binding domain, followed by the nucleophilic attack of the deprotonated active site serine on the β-lactam ring. This results in the formation of a stable acyl-enzyme complex. In contrast to class A TEM-1 β- lactamase, PBPs generally lack the equivalent residue to the conserved residue E166 in TEM-1 responsible for deacylation of β-lactams (Figure 2.4B). This provides an explanation for the inability or extremely slow deacylation rate of PBPs in hydrolysing β-lactams (Morlot et al., 2005). This is further supported by an approximately 10-100 fold increment in the deacylation rate of penicillin G when residue F450 of P. pneumoniae PBP2x was mutated to either an aspartic acid or a glutamic acid to mimic the corresponding position of residue E166 in TEM-1 β- lactamase (Chesnel et al., 2002).

Retrospectively, this is inadequate to explain the deacylation of peptidoglycan stem peptide by class C PBPs, which is a native reaction in these enzymes (usually 32

DD-carboxypeptidase or DD-endopeptidase activity). Structural studies using a deacylation-defective G105D mutant of E. coli PBP5 suggested the involvement of the SXN motif in the deacylation activity of PBPs for its native substrate (Davies et al., 2001; Nicholas et al., 2003). Further studies using a substrate analogue inhibitor covalently bound to the active site of E. coli PBP5 suggested a hydrogen-bonding network involving residues S110 of the SXN motif, K213 of the KTG motif, and a bridging water molecule to activate the hydrolytic water molecule as a mechanism of deacylation of native substrate by class C PBPs (Nicola et al., 2005). Since some class

C PBPs are strictly carboxypeptidases, while others may possess endopeptidase activities, there remains the question of what determines their functional differences.

It is even more puzzling for transpeptidase, which links two stem peptides together.

This could be inferred by the substrate preference of carboxypeptidase for linear stem peptide in peptidoglycan, that of endopeptidase for interlinked stem peptide between two glycan chains of the peptidoglycan, and that of transpeptidase for two stem peptides to be interlinked. Hence, it is likely that the active site of carboxypeptidases cannot accommodate a bulky structure due to steric hindrance (Sauvage et al., 2007).

Structural studies on Bacillus subtilis PBP4a, which is a class C PBP, led Sauvage and co-workers (2007) to propose the plausible role of residue T394 in determining its activity. This residue is hydrogen-bonded to residue T412 and indirectly contributes to the hydrogen-bond network that strengthens the active site. Interestingly, in DD- carboxypeptidases, e.g. E. coli PBP5, PBP6 and PBP6b, a bulkier arginine residue is located at the analogous position of residue T394 in B. subtilis PBP4a, whereas in

DD-endopeptidases, e.g. E. coli PBP4 and PBP7, it is either a threonine or a serine residue. Furthermore, majority of classes A and B PBPs have a threonine or a serine in the analogous position. Hence, it is likely that the type of residue at this analogous 33

position of 394 in B. subtilis PBP4a plays an important role in determining the substrate specificity, leading to the type of peptidase activity using a similar catalytic platform (Sauvage et al., 2007). Nevertheless, the exact mechanism of native substrate processing by the various classes of PBPs remains to be revealed. It has also been suggested that a different acylation-deacylation mechanism may be used by PBPs to interact with their native substrates and β-lactam antibiotics (Nicola et al., 2005).

k 1 k2 k3 PBP + β-lactam PBP β-lactam PBP-β-lactam PBP + Inactivated β-lactam

k-1

Kd = k-1/ k1

Figure 2.2. A reaction scheme for the kinetic interaction of a PBP with a β-lactam. The dissociation constant for the reversible formation of the non-covalent PBP-β- lactam complex is described by Kd. The acylation rate is described by k2 and the deacylation rate is described by k3.

Penicillin-binding domain

Figure 2.3. Crystal structures of S. pneumoniae PBPs. These PBPs possess a common penicillin-binding domain in PBPs of class A (PBP1a and PBP1b), B (PBP2b and PBP2x) and C (PBP3). (PDB accession numbers: PBP1a – 2C5W; PBP1b – 2UWY; PBP2b – 2WAF; PBP2x – 1RP5; PBP3 – 1XP4)

(A)

(B)

Figure 2.4. Crystal structures and active site residues of (A) S. pneumoniae PBP2x (1QME) and (B) E. coli TEM-1 β-lactamase (1BTL). Crystal structures and active site residues of these proteins are depicted in the left column and right column, respectively. The α2 helix, the loop between α4 and α5 helices, and β3 strand containing the SXXK, SXN and K(T/S)G motifs are coloured green, orange and red, respectively, in S. pneumoniae PBP2x. The corresponding folds in serine β-lactamase TEM-1 are similarly coloured with the loop possessing the conserved residue E166 coloured blue. The atoms in the right column are in CPK colours; red - oxygen atom, blue - nitrogen atom, yellow - sulphur atom, white - carbon atom. The hydrogen bonds are depicted as green dotted lines.

2.4. Penam and cephem antibiotics

Both penams and cephems are clinically important classes of antibiotics.

However, the structural relationship between them was only realised in the late 1950s, about a decade after the discovery of cephalosporin C in C. acremonium (Abraham and Newton, 1961; Hamilton-Miller, 2000). The basic difference between penicillin and cephalosporin nuclei lies in the additional cyclic structure fused to the β-lactam ring, which allows the latter to resist degradation by penicillinase produced by penicillin-resistant bacterial strains. Hence, the cephalosporins are generally considered to have a broader antimicrobial spectrum than the penicillins (Marshall and Blair, 1999). In addition, a cephalosporin nucleus which has two sites for modification, compared with the penicillin nucleus which only has one, enables the synthesis of a wider variety of cephalosporins. These probably explain why cephalosporin is generally preferred over penicillin in the synthesis of new β-lactam antibiotic.

2.4.1. Development of semi-synthetic penicillins

Natural penicillins are structurally diverse and their biosyntheses are dependent on the types of side chain precursors available. The first clinically useful penicillin G was produced by supplying phenylacetic acid in the fermentation culture of P. chrysogenum. However, it was used only as a parenteral chemotherapeutic because of its instability under the acidic gastric conditions. Penicillin V was found to be more acid-stable and hence, could be administered orally. This attribute was mainly due to the C6 phenoxyacetyl side chain in penicillin V compared to penicillin

G. However, both penicillins have narrow antimicrobial spectra and were only effective in treating diseases caused by Gram positive bacteria (Wright, 1999). 36

Following the use of penicillin G and penicillin V, the emergence of penicillin- resistant staphylococci became a serious clinical problem (Chain, 1972). This resistance factor was found to be an inactivator of penicillin and was termed as penicillinase (Kirby, 1944). Thereafter, many studies were pursued with the intention of improving the antimicrobial spectrum of penicillin to include Gram negative bacteria, as well as penicillin-resistant staphylococci.

Recognising the importance of C6 side chain of penicillin for its antibacterial property, successes in the development of more efficacious penicillins were brought about by the discovery of 6-APA production in the fermentation culture of P. chrysogenum (Rolinson and Geddes, 2007) (Figure 2.5). By chemically varying the structure of its C6 side chain, a structurally diverse group of semi-synthetic penicillins could be synthesised with different antimicrobial spectra. The semi-synthetic penicillins are mainly classified by structural similarity, which also reflects their antimicrobial spectra.

Methicillin is the first semi-synthetic penicillin that has an ortho- dimethoxyphenyl group as a side chain and was resistant to the action of staphylococcal penicillinase (Rolinson, 1998; Wright, 1999). It was clinically useful in inhibiting penicillin-resistant S. aureus in the early 1960s but its inhibitory activity was not as high as penicillin G against similarly tested susceptible pathogens. It was eventually replaced by other penicillinase-resistant penicillins, e.g., oxacillin and cloxacillin, which have higher bioactivity against staphylococci (Wright, 1999).

However, strains of S. aureus that were resistant to methicillin and other penicillinase- resistant penicillins eventually emerged and till now, still pose a problem to the healthcare industry (Appelbaum, 2006). They are known as methicillin-resistant S. aureus (MRSA) and the resistance mechanism they possessed was due to the 37

acquisition of a low affinity PBP to β-lactam antibiotics known as PBP2a encoded by mecA, instead of evolved β-lactamases (Llarrull et al., 2009; Matsuhashi et al., 1986;

Ubukata et al., 1985; Ubukata et al., 1989).

Ampicillin is the first semi-synthetic penicillin with a broader antimicrobial spectrum than penicillin G, in that it also inhibits a variety of Gram negative bacteria

(Rolinson, 1998). It has a phenylglycyl side chain, which structurally differs from penicillin G by an amino group in the α-carbon of the side chain. However, it has a poor absorption property in the alimentary canal. Subsequently, the addition of a hydroxyl group in the para position of the aromatic ring of ampicillin converted it to a more efficacious amoxicillin. Amoxicillin had better absorption from the gut and resulted in higher levels in the blood than equivalent doses of ampicillin (Wright,

1999). However, resistance to amoxicillin due to β-lactamases also became widespread in the 1970s. Fortunately, a β-lactamase inhibitor, which has a highly similar chemical structure to penicillin, was discovered. It was mainly produced by S. clavuligerus and was named clavulanic acid. This discovery led to the first combinatorial administration of amoxicillin and clavulanic acid, a β-lactamase inhibitor, to overcome the proliferation of β-lactamase-producing pathogens (Miller et al., 2001; Wright, 1999).

Although both ampicillin and amoxicillin were considered to have broad antimicrobial spectrum, they do not have good inhibitory activity against P. aeruginosa, which is one of the leading causes of nosocomial infections, especially in immunocompromised and cystic fibrosis patients (Mesaros et al., 2007; Strateva and

Yordanov, 2009). By replacing the amino group with a carboxyl group in the side chain of ampicillin, improved activity against P. aeruginosa was observed and this penicillin was named carbenicillin. Subsequently, by replacing the phenyl ring with a 38

thiophenyl ring in carbenicillin, ticarcillin was found to possess better antipseudomonal activity than carbenicillin (Rolinson, 1998). By virtue of the carboxyl group in the side chain of both penicillins, they are classified as carboxypenicillins.

Next, the addition of urea moieties to the α-amino group in the side chain of ampicillin yielded ureidopenicillins (azlocillin, piperacillin and mezlocillin), which have even higher activities against P. aeruginosa, while retaining its inhibitory activity against streptococci, enterococci and non-β-lactamase-producing staphylococci (Bush and Johnson, 2000; Tan and File, 1995; Wright, 1999). However, the development of semi-synthetic cephalosporins eventually superseded that of the penicillins probably due to the amenability of the cephalosporin nucleus for multiple chemical modifications.

H 1 N 6 5 S R 2 CH3 N CH 7 4 3 3 O COOH 6-APA β-Lactamase-resistant Penicillins

Cl O CH3 N Cl H O H N N H S N S N O S H C O O N 3 O N CH O N 3 O O CH3 COOH O COOH COOH Methicillin Oxacillin Dicloxacillin

Aminopenicillins

NH2 NH2 H H N S N S

O N O N O HO O COOH COOH Ampicillin Amoxicillin

Carboxypenicillins

COOH COOH H H N S N S

O N S O N O O COOH COOH Carbenicillin Ticarcillin

Ureidopenicillins

O O S N O NH N N O N O N O

O O NH O NH NH H H H N N S N S S

O N O N O N O O O COOH COOH COOH Azlocillin Mezlocillin Piperacillin

Figure 2.5. Chemical structures of penicillin nucleus 6-APA and major groups of semi-synthetic penicillins.

2.4.2. Development of semi-synthetic cephalosporins

Although cephalosporin C is the first cephalosporin discovered, its narrow antimicrobial spectrum limited its clinical application. Nevertheless, the discovery of cephalosporin C led to the development of numerous cephalosporin derivatives. It was recognised that, similar to the development of semi-synthetic penicillins, the availability of the cephalosporin core structure was important in modifying the side chain of cephalosporin to improve its antimicrobial spectrum and efficacy. The production of the cephalosporin core structure, i.e., 7-aminocephalosporanic acid (7-

ACA), was made possible by deacylation of the aminoadipyl side chain of cephalosporin C. Hence, 7-ACA was used as a precursor for the synthesis of many earlier semi-synthetic cephalosporins (Hamilton-Miller, 2008).

Other than having a similar chemical structure and mode of action as penicillin, cephalosporins were found to be naturally resistant to penicillinase. These two attributes made it a more attractive lead compound in the development of new β- lactam antibiotics (Hansch et al., 1990). Furthermore, cephalosporin has two alkyl sites amenable to chemical modification compared to penicillin, which has only one

(Figure 2.6). This allows a larger variety of cephalosporin derivatives to be synthesised and assessed for better antimicrobial properties suitable for clinical applications. Hence, a plethora of β-lactam derivatives were developed and marketed, with an estimation of at least 50 different cephalosporins available in the market in

2003 (Elander, 2003; Hamilton-Miller, 2000; Hamilton-Miller, 2008).

Cephalexin is one of the earlier semi-synthetic cephalosporins available and has the same phenylglycyl side chain as ampicillin. Its antimicrobial spectrum encompasses that of ampicillin, except that its resistance to penicillinase bestowed upon itself a wider spectrum than ampicillin (Hansch et al., 1990). Similarly, 41

cefadroxil, which is a p-hydroxyl derivative of cephalexin, also possesses similar pharmacokinetic properties but has a wider spectrum than its penicillin counterpart, amoxicillin (Hansch et al., 1990). It became clear that in order to further improve the efficacy of cephalosporin, the structure of the side chains of cephalosporin has to be different from that of the semi-synthetic penicillins. Thereafter, the pool of structurally diverse semi-synthetic cephalosporins with different antimicrobial spectra and susceptibilities to the β-lactamases increased tremendously.

The semi-synthetic cephalosporins are generally grouped into ‘generations’ based on their antimicrobial spectra and stability against β-lactamases, which is considered more useful for administrative reference (Figure 2.6). The first generation cephalosporins are made up of early cephalosporins that have similar antimicrobial spectra as that of cephalexin. They were considered broad spectrum antibiotics with the advantage of being more stable against penicillinase. The second and third generations have extended spectrum of the first generation cephalosporins, including many Gram negative bacteria, and are more resistant to the evolving β-lactamases.

However, these improved cephalosporins suffered lower inhibitory activity against

Gram positive bacteria. In contrast to the first three generations, the fourth generation cephalosporins, e.g., cefepime, have a broad antimicrobial spectrum including many

Gram positive and Gram negative bacteria, as well as P. aeruginosa. They are also generally more resistant to the evolved β-lactamases that are active against earlier generation cephalosporins (Marshall and Blair, 1999). However, the issue with

MRSA remained problematic until recently, when two new cephalosporins, namely ceftobiprole (Deresinski, 2008; Llarrull et al., 2009) and ceftaroline (Llarrull et al.,

2009; Villegas-Estrada et al., 2008), were developed. Both cephalosporins were tested and shown to be efficacious against MRSA due to higher binding affinity to the 42

resistant PBP2A. Ceftobiprole has also been shown to have activity against many

Gram negative bacteria in vitro including P. aeruginosa (Barbour et al., 2009; Walkty et al., 2008). Both ceftobiprole and ceftaroline have been approved by the United

States Food and Drug Administration for the treatment of skin and skin structure infections (Laudano, 2011; Stein et al., 2009).

Other than MRSA, P. aeruginosa has also been reported with evolved resistance towards the fourth generation cephalosporins. This is primarily due to

AmpC β-lactamase degrading these cephalosporins. By modifying the C3 side chain of ceftazidime, a new cephalosporin CXA-101 was designed and demonstrated to be resistant to the action of P. aeruginosa AmpC β-lactamase (Murano et al., 2008; Toda et al., 2008). Recently, it was also shown to be effective against various β-lactam- resistant strains of P. aeruginosa (Giske et al., 2009; Livermore et al., 2009;

Zamorano et al., 2010). Its efficacy was also tested in the presence of a β-lactamase inhibitor, tazobactam, against extended-spectrum β-lactamase-producing E. coli and

Klebsiella pneumoniae. This study showed that the combination CXA-201 (consisting of an antibiotic and an inhibitor) was able to inhibit 96% of the test bacteria with a breakpoint of ≤ 1 mg/L (Titelman et al., 2011). Currently, CXA-201 is undergoing

Phase 3 trial (http://www.cubist.com/products/cxa_201).

H 1 N 7 6 S R1 2 3 8 N O 5 4 R2 COOH 7-ADCA

1st Generation Cephalosporins

NH2 NH2 H H H N S N S N S

O N O N S O N O CH3 O CH3 HO O CH3 O Cephalexin COOH Cefadroxil COOH Cephalothin COOH O

nd CH3 2 Generation Cephalosporins O N CH3 NH2 H H O H O N S N S N S

O N O NH2 S O N O NH2 O N O O HO O CH3 COOH O COOH O COOH Cefuroxime Cefoxitin Cefprozil

3rd Generation Cephalosporins

H3C COOH CH3 H3C CH3 O O N O H N N H N N S H N N S N N S H N 2 O N O CH H N H N S 3 2 O N N 2 O N S N O O S S O O COOH O COOH COOH N Cefotaxime Ceftazidime Ceftriaxone H3C N O H 4th Generation Cephalosporins

CH3 CH3 O O N N H H N N S N N S H N H N 2 O N N 2 O N N S S O O COOH COOH Cefepime Cefpirome

New Cephalosporins H2N H3C COOH N H3C NH OH O O O N N N H H H NH N N S N N N S N N S N NH H N N H N H N 2 2 N O N 2 N O N 2 N O N N S S S S N O O S O O CH COOH COOH COOH 3 Ceftaroline Ceftobiprole CXA-101

Figure 2.6. Chemical structures of cephalosporin nucleus 7-ADCA and different generations of semi-synthetic cephalosporins.

2.4.3. Industrial production of penicillin and cephalosporin nuclei

Industrial production of semi-synthetic penicillins and cephalosporins began in the 1960s. The idea of producing semi-synthetic cephalosporins stemmed from the ability to synthesise a variety of semi-synthetic penicillins, using the penicillin nucleus 6-APA, with improved antimicrobial spectrum (Hamilton-Miller, 2000;

Hamilton-Miller, 2008). However, unlike 6-APA, the cephalosporin nucleus is not readily produced in the cephalosporin producers (Rolinson and Geddes, 2007). It was clear then that the development of semi-synthetic cephalosporins as useful chemotherapeutics depended on the ability to obtain the cephalosporin nucleus.

Naturally, cephalosporin C became the initial choice to obtain the cephalosporin nucleus. By using mild acid hydrolysis, cephalosporin C could be chemically deacylated to remove the aminoadipate side chain and yield 7-ACA

(Morin et al., 1962). However, the amount of 7-ACA produced was very low. By using formic acid, a yield increment from 7% to 40% could be achieved (Morin et al.,

1962). This initial high yield preparation of 7-ACA allowed the synthesis of various semi-synthetic cephalosporins, i.e., thiophene-2-acetyl and furan-2-acetyl side-chains at C7β of cephem nucleus, that were tested for better efficacy than benzylpenicillin in the 1960s (Chauvette et al., 1962). The selected semi-synthetic cephalosporin was named cephalothin and was then further evaluated clinically as a broad spectrum cephalosporin (Chauvette et al., 1962). However, this chemical deacylation method required organic solvents and produced toxic chemical wastes, which are undesirable.

Hence, chemical deacylation of cephalosporin C was subsequently replaced by enzymatic methods.

Currently, there are two approaches to enzymatic deacylation. The ideal method is the direct deacylation of the aminoadipate side chain using a cephalosporin 45

acylase. However, the natural cephalosporin acylases isolated from natural sources were not efficient (Barber et al., 2004). With the advancement in protein engineering, the activity of cephalosporin acylase could be improved, allowing it to be used industrially (Oh et al., 2003; Otten et al., 2002; Pollegioni et al., 2005; Sonawane,

2006). Another approach is the oxidative deamination of cephalosporin C to keto- adipyl-7-ACA by D-amino acid oxidase (Barber et al., 2004). In this reaction, hydrogen peroxide would be produced and lead to spontaneous oxidative decarboxylation of keto-adipyl-7-ACA, which would then formed glutaryl-7-ACA.

Glutaryl-7-ACA is then deacylated using glutaryl acylase to finally yield 7-ACA

(Figure 2.7). For semi-synthetic cephalosporins that require 3-vinyl substituted intermediates, e.g., cefdinir, 7-ACA has to be deacetylated to 7- aminodeacetylcephalosporanic acid (7-ADOCA) using cephalosporin acetyl hydrolase or cephalosporin acetyl esterases (Figure 2.7) (Barber et al., 2004; Sonawane, 2006).

Another setback with cephalosporin C production is its chemical instability in the fermentation cultures of A. chrysogenum. This instability leading to degradation could account for approximately 40% loss of the cephalosporin C produced (Usher et al., 1988). To compete with the market of 7-ACA production, penicillin G became the preferred substrate for the production of cephalosporin nucleus because it is easily available in bulk from a high yielding strain of P. chrysogenum. Penicillin G could be converted into a cephalosporin DAOC G by a chemical ring-expansion procedure.

Subsequently, its side chain could be enzymatically deacylated to yield the cephalosporin nucleus, 7-ADCA (Figure 2.8). However, this chemical ring-expansion method involved multiple steps and produced toxic wastes. Hence, cheaper and greener alternatives were preferred and highly sought after.

H HOOC N S Cephalosporin C H2N S Enzymatic H2N S acylase deacetylation

NH2 O N O CH3 N O CH3 N OH O O O COOH O COOH O COOH

Cephalosporin C 7-ACA 7-ADOCA

O2 D-amino acid oxidase Glutaryl acylase NH3 + H2O2

H Spontaneous oxidative H HOOC N S decarboxylation N S HOOC

O O N O CH3 O N O CH3 O O H O CO COOH O 2 2 2 COOH O Keto-adipyl-7-ACA Glutaryl-7-ACA

Figure 2.7. Industrial production of cephalosporin precursors, 7-ACA and 7-ADOCA.

Chemical ring- H H Enzymatic expansion H N S N S N S deacylation 2

O N O N N O O O CH3 C OOH COOH COOH Penicillin G DAOC G/G-7-ADCA 7-ADCA

Figure 2.8. Industrial production of cephalosporin precursor, 7-ADCA.

2.5. Biosynthesis of penicillin and cephalosporin

The primary interest in understanding β-lactam biosynthesis originated from the structural similarity between penicillin and cephalosporin. Resolution of the chemical structure of cephalosporin C bearing a β-lactam ring fused with a dihydrothiazine ring further affirmed the possible genetic linkage between their biosyntheses (Abraham and Newton, 1961). Studies on the biosynthetic pathways of penicillin and cephalosporin producers mainly concentrated on P. chrysogenum, A. chrysogenum, S. clavuligerus, N. lactamdurans and L. lactamgenus. Despite 30 years of investigation, our understanding of the biosynthetic pathway of β-lactams is not complete because of the branching into varied pathways leading to different end- products in different microorganisms, especially the cephamycin and cephabacin producers (Liras, 1999; Liras and Demain, 2009; Martin and Gutierrez, 1995; Martin et al., 2010). Nevertheless, the major pathway leading to the formation of the penicillin and cephalosporin core structures have been deciphered, allowing genetic manipulation of industrial strains for high yield production (Figure 2.9).

2.5.1. β-Lactam biosynthetic pathways of fungi and bacteria

Biosynthesis of penicillin begins with the fusion of three amino acids, i.e., L- aminoadipic acid, L-cysteine and L-valine, into a tripeptide known as δ-(L-α- aminoadipyl)-L-cysteinyl-D-valine (ACV), by ACV synthetase (ACVS) encoded by pcbAB (Figure 2.9) (Coque et al., 1991; Diez et al., 1990; Gutierrez et al., 1991;

Loder and Abraham, 1971; Tobin et al., 1991). ACV is cyclised to isopenicillin N by isopenicillin N synthase (IPNS), which is encoded by pcbC gene (Castro et al., 1988;

Diez et al., 1989). This is the formation of the first bioactive compound in the biosynthetic pathway. IPNS is a non-haem ferrous iron-dependent oxygenase, which 48

bears the conserved iron-binding motif HXD/EXnH. It also possesses the RXS motif, which is critical for the coordination of the co-substrate 2-oxoglutarate in the enzymatic reaction of 2-oxoglutarate dependent dioxygenases. Interestingly, IPNS does not require 2-oxoglutarate for its reactivity. Instead, the RXS motif is important for binding the valine carboxylate of ACV substrate (Roach et al., 1997; Roach et al.,

1995).

In penicillin-producing organisms, e.g., P. chrysogenum, the end-product of its

β-lactam biosynthetic pathway can be altered by feeding phenylacetic acid or phenoxyacetic acid to the fermentation culture to derive penicillin G or penicillin V, respectively, which are the earliest clinically useful penicillins. This is achieved by acyl-coenzyme A:IPN acyltransferase, which exchanges the aminoadipyl side chain of isopenicillin N to the respective coenzyme A-activated feeding substrate (Barredo et al., 1989) (Figure 2.9). This enzyme not only possesses IPN acyltransferase activity, it also bears amidohydrolase and 6-APA acyltransferase activities, and accepts a wide range of acyl-CoA derivatives and various non-CoA thioesters, highlighting a potential application in the synthesis of semi-synthetics (Alvarez et al., 1993;

Whiteman et al., 1990). Although this enzyme is encoded by a single penDE gene in both fungi, P. chrysogenum IPN acyltransferase is auto-processed from a 40 kDa inactive proenzyme into a heterodimer consisting of α (29 kDa) and β (11 kDa) subunits, which differs from Aspergillus nidulans IPN acyltransferase that is active in its unprocessed form. This difference in functionality has been attributed to the protein sequence disparity in the proenzyme, which determines the autocatalytic activity (Fernandez et al., 2003).

Currently, there are at least three different classes of natural cephem antibiotics. Cephalosporin C is the first cephem discovered and so named after the 49

earlier genus name of the producer A. chrysogenum, i.e., Cephalosporium acremonium (Campos Muñiz et al., 2007). Cephamycin C deviates from cephalosporin C as it possesses a methoxy group at C7α and a carbamoyl group at C3 instead of an acetyl group (Figure 2.9). So far, it is mainly produced by actinomycetes viz Nocardia lactamdurans and Streptomyces species (Nagarajan et al., 1971; Stapley et al., 1972). The third group of cephem is cephabacin. Similar to cephamycin C, cephabacin differs from cephalosporin C at its C7α and C3 positions. It possesses a formylamino group at C7α and an oligopeptide moiety at its C3 position. Presently, cephabacins are only known to be produced by Gram-negative bacteria Lysobacter lactamgenus (Ono et al., 1984), Xanthomonas lactamgena (Ono et al., 1984) and

Flavobacterium sp. (Palissa et al., 1989). Both cephamycin C and cephabacin have been demonstrated to be more resistant to the degradation of β-lactamases than cephalosporin C (Nozaki et al., 1985; Nozaki et al., 1984).

In cephalosporin C-producing A. chrysogenum, isopenicillin N is converted into penicillin N by a two-component isopenicillin N epimerase system. This two- component protein system is encoded by two genes, namely cefD1 and cefD2, which correspond to isopenicillinyl-CoA ligase and isopenicillinyl-CoA epimerase, respectively (Martin et al., 2004; Ullan et al., 2002). Penicillin N is then ring- expanded to deacetoxycephalosporin C (DAOC) by a bifunctional deacetoxy/deacetylcephalosporin C synthase (DAOC/DACS) enzyme, where the five- membered thiazolidine ring in penicillin N is converted to the six-membered dihydrothiazine ring in DAOC (Dotzlaf and Yeh, 1989). DAOC is subsequently hydroxylated at C3 position by the same enzyme to yield deacetylcephalosporin C

(DAC). Then an acetyl group is added to the C3 hydroxyl group of DAC by DAC

acetyltransferase, which is encoded by cefG, to finally produce cephalosporin C

(Gutierrez et al., 1992).

In S. clavuligerus and N. lactamdurans, isopenicillin N is converted to penicillin N by a monomeric pyridoxal phosphate-dependent isopenicillin N epimerase, which is different from the A. chrysogenum epimerase system (Martin et al., 2004; Usui and Yu, 1989). Penicillin N is ring-expanded to DAOC, followed by hydroxylation to DAC using two separate monofunctional DAOCS and deacetylcephalosporin C synthase (DACS) enzymes (Coque et al., 1996; Cortes et al.,

1987; Jensen et al., 1985). These two steps are similar to that of the cephalosporin C biosynthetic pathway. Thereafter, a carbamoyl group is added to the C3 hydroxyl of

DAC by DAC-carbamoyl transferase (Coque et al., 1995b), followed by methoxylation at C7α by O-carbamoyl-DAC-7-methoxyl transferase (Coque et al.,

1995a) (Figure 2.9). Studies have shown that the hydroxyl and methyl groups are derived from molecular oxygen and methionine, respectively (O'Sullivan et al., 1979;

Whitney et al., 1972). This last step of methoxylation is carried out by two enzymes,

CmcI and CmcJ, separately encoded by the cmcI and cmcJ genes, respectively. These two enzymes function as a duo-protein component in hydroxylating the C7α position of the cephem nucleus followed by methylation, although in N. lactamdurans, the

CmcI enzyme alone has weak hydroxylation activity in the presence of 2-oxoglutarate

(Coque et al., 1995a; Enguita et al., 1996). Interestingly, N. lactamdurans CmcI was also found to exhibit hydroxylase activity in the presence of NADH although no methyltransferase activity was detected for the CmcI-CmcJ complex (Coque et al.,

1995a). Fluorescence spectroscopic analysis of CmcI and CmcJ interaction illustrated conformational changes in CmcI with increasing concentrations of CmcJ, and the

binding of substrates S-adenosylmethionine and cephalosporin C was higher in the

CmcI-CmcJ complex compared to the individual proteins (Enguita et al., 1996).

Similar to the biosynthesis of cephamycin C, the biosynthesis of cephabacin differs from that of cephalosporin C after DAC is formed although it is still not clear if the formylamino or the oligopeptide moiety was added first (Figure 2.9). Recent efforts in deciphering the functionalities of enzymes involved in the synthesis of the oligopeptide moiety have shed light on their identities. The cpbI gene encodes the non-ribosomal peptide synthetase-polyketide synthase-hybrid complex (NRPS-PKS), in which the NRPS component is composed of three modules. Each of these modules has three domains responsible for catalysing condensation, adenylation and thiolation reactions. The cpbK gene encodes only a single NRPS module (Demirev et al., 2006).

These two genes encode enzymes necessary for the formation of the oligopeptide moiety. However, the enzymes responsible for the addition of the formylamino and oligopeptide moieties to the C7α and C3 position of cephabacin F, respectively, have yet to be elucidated.

H2N H2N SH H N COOH + + 2 COOH COOH COOH L-aminoadipic acid L-cysteine L-valine

pcbAB ACV synthetase H H2N N SH

COOH O N O H COOH L-aminoadipyl-L-cysteinyl-D-valine

pcbC Isopenicillin N synthase (IPNS)

H 1 H H H2N N 6 5 S N S N S 2 IPN acyltransferase O C O N O N O N OOH 3 or O 7 4 O O Isopenicillin N C OOH penDE C OOH COOH cefD/cefD1 Penicillin G Penicillin V IPN epimerase & cefD2 H HOOC N S Penicillium chrysogenum

NH2 O N O Penicillin N C OOH cefE/cefEF Deacetoxycephalosporin C synthase (DAOCS) H 1 HOOC N S 7 6 2 3 NH2 O N 8 5 O 4 DAOC COOH cefF/cefEF Deacetylcephalosporin C synthase (DACS)

H H HOOC N S DAC acetyltransferase HOOC N S

NH2 O N OH NH2 O N O CH3 O cefG O DAC COOH Cephalosporin C COOH O DAC-carbamoyl transferase Acremonium chrysogenum CpbI & cmcH cpbI & cpbK CpbK H HOOC N S

NH2 O N O NH2 O COOH O

H R1 HOOC N S R2 O-Carbamoyl-DAC (OCDAC) HN H NH2 O N O N NH O OCDAC-7-methoxyl Cephabacin F/H COOH O OH NH2 cmcI & cmcJ transferase R1 R2

F1 H -L-Ala CH3 N O -L-Ala-L-Ala H O F2 C HOOC N S F3 -L-Ala-L-Ala-L-Ala NH2 O N O NH2 H1 -L-Ala O COOH O H2 -H -L-Ala-L-Ala Cephamycin C H3 -L-Ala-L-Ala-L-Ala Lysobacter lactamgenus Streptomyces clavuligerus Nocardia lactamdurans

Figure 2.9. Biosynthesis of penicillin, cephalosporin C, cephamycin C and cephabacin in respective microorganisms. In A. chrysogenum, the epimerisation of isopenicillin N to penicillin N involves two different enzymes encoded by cefD1 and cefD2 genes, and the ring-expansion followed by hydroxylation of penicillin N to DAC is catalysed by a bifunctional enzyme DAOC/DACS encoded by cefEF gene.

2.6. Deacetoxycephalosporin C synthase

The committed step in the biosynthesis of cephem antibiotics is executed by monofunctional DAOCS in prokaryotes and bifunctional DAOC/DACS in A. chrysogenum. This enzymatic reaction is highly specific and has been proposed to replace the existing chemical ring-expansion of penicillin G in the industrial production of semi-synthetic cephalosporins (Table 2.3) (Andersson et al., 2001;

Barber et al., 2004; Barends et al., 2004). However, the limitation in using DAOCS for enzymatic ring-expansion of penicillin G is its low catalytic activity since its native substrate is penicillin N. Hence, many studies have focused on its biochemical and structural characteristics in order to facilitate bioengineering of DAOCS for industrial application.

Table 2.3. Summary of DAOCS-related patents.

Country/Patent Applicants/Inventors Title Summary of inventions/Claims number DSM/Nieboer M, Method to The invention describes a method for US/ 6,995,003 Bovenberg RAL. localise an improved in vivo production of expandase in acylated cephalosporins; the host cell the cytosol has expandase activity localised in the cytosol instead of the usual localisation in peroxisomes or microbodies of the host cell.

Synmax Biochemical Mutated Mutated expandase having expansion US/ 6,905,854 Co., Ltd./Yang Y-B, penicillin activities 2-32 fold higher on penicillin Wei C-L, Hsu J-S, expandase G than wild-type expandase; mutated Tsai Y-C. and process penicillin expandase, which comprises for preparing one or more specific amino acid 7-ADCA substitutions selected from the group consisting of M73T, S79E, V275I, L277K, C281Y, G300V, N304K, I305L and I305M mutations

Isis Innovation Modified Modified DAOC/DACS with UK/ WO Ltd/Lloyd MD, DAOC/DACS hydroxylase activity inactivated and 2005/019468 A1 Schofield CJ, enzymes improved ring-expanding activity. Hewitson KS, Lipscomb S.

Massachusetts Penicillin The inventive conversion system US/ 6,383,773 Institute of conversion allows biological ring expansion of Technology penicillin substrates such as penicillin (Cambridge, G, and replaces the multi-step chemical MA)/Demain, AL, ring expansion process currently Cho H, Piret JM, performed in industry; the inventive Adrio JL, Fernandez system can utilise growing or resting M-JE, Baez-Vasquez cells, or isolated expandase, and is MA, Hintermann G. capable of converting exogenously- added penicillins; the inventive system can be applied to any penicillin substrate, including natural penicillins, biosynthetic penicillins, semisynthetic penicillins, and/or synthetic penicillins.

2.6.1. Biochemical characteristics of deacetoxycephalosporin C synthase

Enzyme reaction conditions for the detection of ring-expansion activity have been modified over the last two decades. To achieve optimum conditions for the conversion of penicillin to a cephalosporin, Shen and co-workers (1984) derived a reaction mixture containing 50 mM Tris–HCl (pH 7.4), 10 mM KCl, 10 mM MgSO4,

0.6 mM ascorbic acid, 0.8 mM ATP, 0.04 mM FeSO4, 0.6 mM 2-oxoglutarate, and

0.28 mM penicillin N for detection of the ring-expansion activity of DAOCS from A. chrysogenum. They showed that an effective order of co-factor addition to the reaction mixture is important, i.e., simultaneous addition of iron (II) ion, ascorbic acid, and ATP followed by 2-oxoglutarate and then penicillin N, were obligatory to achieve the highest detectable activity (Shen et al., 1984). However, it was found that

DAOCS was unable to convert penicillin G and other penicillin N analogues under the optimum conditions described for penicillin N (Dotzlaf and Yeh, 1989). Maeda and co-workers (1995) tested various conditions such as temperature, duration of reaction, buffers and concentration of penicillin G, but were still unable to catalyse the conversion of penicillin G to its respective cephem moiety. They were only able to detect ring-expansion activity when penicillin N or D-carboxymethylcysteinyl-6-APA was used as the substrate. As such, DAOCS isolated from various sources was thought to have a very narrow substrate specificity (Cortes et al., 1987; Dotzlaf and

Yeh, 1989; Maeda et al., 1995). Subsequently, Cho and co-workers (1998) discovered reaction conditions that enabled the ring-expansion of penicillin N analogues. This was made feasible by increasing the concentration of iron(II) and the co-substrate, 2- oxoglutarate in the reaction mixture, which enabled the conversion of penicillin G and

14 other penicillin N analogues to their respective cephem moieties by S. clavuligerus

NP1 resting cells (Table 2.4). Thereafter, Sim and Sim (2001) showed that omission 56

of KCl, MgSO4, dithiothreitol and ascorbic acid from Cho’s reaction conditions could significantly improve the conversion of penicillin G and ampicillin by GST- scDAOCS fusion protein (Sim and Sim, 2001). Using these improved reaction conditions, Chin and Sim (2002) showed that the substrate specificity of scDAOCS is wider than previously thought (Table 2.4).

Several assays have been used for the detection of cephalosporin formation by

DAOCS. These include a hole-plate bioassay (Chin et al., 2001; Sim and Sim, 2001), thin-layer chromatography (Jensen et al., 1983), and reverse-phase high-pressure liquid chromatography (HPLC) (Chin et al., 2001; Fernandez et al., 1999; Jensen et al., 1983). A hole-plate bioassay relies on the use of agar plate seeded with β-lactam- sensitive E. coli and penicillinase, which will degrade any remaining penicillin substrate. A well is made in the centre of the agar plate, followed by loading of the enzyme reaction mixture. The antibiotic compounds present in the mixture will diffuse into the agar and inhibit the growth of E. coli, resulting in a zone of growth inhibition, which reflects the concentration of cephalosporin present. Thin-layer chromatography uses a cellulose paper as the solid phase and n-butanol-acetic acid- water as the mobile phase to separate penicillin substrate and cephalosporin product based on their solubility differences in the mobile phase. After development, the paper is dried and placed face-down on an agar slab containing β-lactam-sensitive E. coli and penicillinase to detect for growth inhibition. Reverse-phase HPLC is similar to thin-layer chromatography, except that it uses a C18 column instead of cellulose paper and the separation of compounds is under high pressure. The elution of respective penicillin substrate and cephalosporin product is monitored using a spectrophotometer, which will read out the level of absorption at a specific wavelength of light against retention time. However, these assays are laborious, 57

cumbersome, and yield limited data points. Furthermore, a large number of samples are required for kinetic studies. A continuous and direct spectrophotometric assay for the determination of penicillin N conversion by DAOCS was first reported by

Baldwin and Crabbe (1987). However, this assay contained sodium azide as one of the components in the reaction mixture, which probably limited its application.

Subsequently, another spectrophotometric assay was devised by Dubus and co- workers (2001) to probe the kinetic properties and substrate and co-substrate selectivities of the recombinant scDAOCS. They showed that scDAOCS has a Km

-1 value of 0.89 mM and a kcat value of 0.079 s for penicillin G conversion, and a Km

-1 value of 4.86 mM and a kcat value of 0.120 s for ampicillin conversion. Interestingly, they also showed that the substrate specificity of scDAOCS included acetyl-6-APA,

-1 with a Km value of 3.28 mM and a kcat value of 0.060 s for its conversion.

Although many studies have shown that the substrate specificity of scDAOCS is broad, kinetic studies have illustrated its low catalytic efficiency towards penicillin

G compared with its natural substrate, penicillin N (Lee et al., 2002; Lee et al., 2001a;

Lipscomb et al., 2002; Lloyd et al., 1999b; Wei et al., 2003a; Wei et al., 2005). In a recent study by Wei and co-workers (2005), the Km and kcat values for penicillin G conversion were reported to be 2.58 mM and 0.0453 s-1, and for penicillin N conversion, the reported values were 0.014 mM and 0.307 s-1, respectively. Hence, further studies on this enzyme have moved towards understanding its structure- function relationship and improving its catalytic efficiency in converting penicillin G.

Table 2.4. A range of penicillin substrates catalysed by scDAOCS. H H N S N S R CH3 DAOCS R

N CH3 N CH3 O 2OG, O2 Succinate, O COOH CO2, H2O COOH Penicillin substrate Side chain, R References Penicillin N H2N Maeda et al., 1995 COOH O H2N D-carboxymethylcysteinyl-6-APA S Maeda et al., 1995 COOH O Acetyl-6-APA Dubus et al., 2001 O Butyryl-6-APA Cho et al., 1998 O Valeryl-6-APA Cho et al., 1998 O Hexanoyl-6-APA Cho et al., 1998 O Penicillin F Cho et al., 1998 O Adipyl-6-APA HOOC Cho et al., 1998 O Heptanoyl-6-APA Cho et al., 1998 O Octanoyl-6-APA Cho et al., 1998 O Nonanoyl-6-APA Cho et al., 1998 O Decanoyl-6-APA Cho et al., 1998 O 2-Thiophenylacetyl-6-APA Cho et al., 1998 S O Penicillin X Cho et al., 1998 O HO Penicillin mX HO Cho et al., 1998 O Penicillin G Chin and Sim, 2002; O Dubus et al., 2001; Cho et al., 1998; Sim and Sim 2001 Ampicillin NH2 Chin and Sim, 2002; Dubus et al., 2001; Cho O et al., 1998; Sim and Sim 2001 Amoxicillin NH2 Chin and Sim, 2002

O HO CH2 Metampicillin N Chin and Sim, 2002

O Penicillin V Chin and Sim, 2002; Cho O et al., 1998 O Phenethicillin CH3 Chin and Sim, 2002 O CH O Carbenicillin COOH Chin and Sim, 2002

O Ticarcillin COOH Unpublished

S O

2.6.2. Limitations of native deacetoxycephalosporin C synthase

A few DAOCS genes have been cloned and characterised. Among these, only one is of eukaryotic origin, while the rest are from prokaryotes (Table 2.5). Earlier studies on DAOCS were mainly performed using those from C. acremonium, S. clavuligerus, or N. lactamdurans. Although S. cattleya NRRL 8057 (Khaoua et al.,

1991; Williamson et al., 1985) and Streptomyces katsurahamanus T-272 (Ward and

Hodgson, 1993) were reported to produce cephamycin C, and Streptomyces sp. strain

DRS-1 (Roy et al., 1997) possessed ring-expansion activity, the genes encoding

DAOCS from these streptomycetes have yet to be cloned and characterised. Three cefE genes from Streptomyces jumonjinensis, Streptomyces ambofaciens, and

Streptomyces chartreusis (Hsu et al., 2004) were cloned and characterised within the last decade (Table 2.5). Although DAOCSs of these three organisms have high protein sequence identities with that of S. clavuligerus, they were less active than that of S. clavuligerus in converting penicillin G under the respective prescribed assay conditions despite the presence of highly conserved amino acid residues reported to be involved in binding penicillin G (Figure 2.10) (Hsu et al., 2004; Chin et al., 2003).

This suggests that the binding of penicillin G does not solely depend on these conserved amino acid residues. Furthermore, penicillin G used in these studies is not a natural substrate of these enzymes and their activities in converting their natural substrate have yet to be demonstrated. Thus, it is difficult to rationalise the differences in activities observed among the DAOCSs from various sources, especially when scDAOCS is the only enzyme with its three-dimensional structure resolved.

Comparison of the catalytic efficiency of scDAOCS in converting its native substrate penicillin N and penicillin G revealed a striking difference of approximately

1200-fold. The Km and kcat values of scDAOCS for penicillin G was about 180-fold 60

higher and 7-fold lower than that of scDAOCS for penicillin N, respectively (Table

1.1). It is very likely that the substrate binding pocket of scDAOCS is not ideal for penicillin G since the substrate binding affinity for penicillin G, as reflected by the Km value, was much lower.

Table 2.5. DAOCS orthologues cloned and characterised.

GenBank Protein Sequence Organism accession no. length identitya Activity References S. clavuligerus NRRL 3585 P18548 311 100% Active (Kovacevic et al., 1989)

S. jumonjinensis NRRL 5741 AF317908 311 81.7% Active (Chin et al., 2003)

S. chartreusis (102SH3) AY318743 311 76.5% Active (Hsu et al., 2004)

S. ambofaciens (29SA4) AY318742 311 81.4% ND (Hsu et al., 2004)

N. lactamdurans Z13974 314 70.1% Active (Coque et al., 1993)

L. lactamgenus YK90 CAA39984 319 56.1% Active (Kimura et al., 1996)

A. chrysogenumb P11935 332 53.8% Active (Samson et al., 1987) ND-not determined aPrimary protein sequence identities were obtained using MatGAT (Campanella et al., 2003) and expressed as percentage of S. clavuligerus DAOCS sequence bA bifunctional enzyme encoded by cefEF, which possesses both deacetoxycephalosporin C synthase and deacetylcephalosporin C synthase activities

Figure 2.10. Multiple sequence alignment of annotated DAOCSs from S. clavuligerus NRRL3585 (Scl), S. chartreusis 102SH3 (Sch), S. ambofaciens 29SA4 (Sam), S. jumonjinensis NRRL5741 (Sju), N. lactamdurans (Nla), L. lactamgenus YK90 (Lla) and A. chrysogenum (Ach). The accession numbers of these sequences are in Table 2.5. The substrate binding residues reported by Hsu and co-workers (2004) are highlighted in grey colours. The secondary structure of scDAOCS is presented above its primary sequence based on the structural results reported by Valegard and co- workers (2004) (PDB: 1UNB). The secondary structures of residues 81-90 and 167- 177 were highly disordered and were not determined (Valegard et al., 2004). The cylinders, arrows and black lines represent α-helices, β-sheets and coil structures, respectively.

2.6.3. Structural studies of deacetoxycephalosporin C synthase

DAOCS belongs to a family of enzymes that utilises an iron (II) ion as a co- factor and 2-oxoglutarate as a co-substrate to catalyse oxidative reactions, i.e., 2- oxoglutarate dependent dioxygenases. The enzymes in this family have a characteristic “jelly-roll” structure made up of eight β-sheets, with four β-sheets on each side of the barrel (Figure. 2.11). The β-barrel accommodates the conserved

HXD/EXnH and RXS motifs essential for iron (II) ion and 2-oxoglutarate binding, respectively (Hausinger, 2004); in the case of scDAOCS, they are H183, D185, H243,

R258, and S260 (Sim and Sim, 2000; Sim et al., 2003). Other antibiotic-producing enzymes in this family, for example clavaminate synthase (CAS) and carbapenem synthase, also have similar structural features (Clifton et al., 2003; Lloyd et al.,

1999a). Furthermore, a number of flavonoid synthesising enzymes also belongs to this family (Turnbull et al., 2004), e.g., flavanone 3β-hydroxylase (Owens et al., 2008), flavone synthase (Britsch, 1990), flavonol synthase (Chua et al., 2008) and anthocyanidin synthase (Wilmouth et al., 2002).

The iron-binding centre in this family of enzymes essentially provides versatility and flexibility in catalysing a variety of oxidative reactions that include desaturation, ring-expansion, and hydroxylation. These are achieved by the generation of a reactive iron species that is required to attack the substrates. The resulting oxidative reaction would depend on the internal architecture of the enzyme, where the spatial arrangements of the substrate relative to the activated iron centre are important considerations for eliciting reactions. The sequence of events from iron (II) ion and 2- oxoglutarate binding to substrate oxidation may also differ slightly between enzymes, for similar reasons. For example, in CAS, simultaneous 2-oxoglutarate and substrate binding are necessary to activate the iron centre, whereas in DAOCS, binding of 2- 63

oxoglutarate and penicillin substrates are mutually exclusive, as supported by structural studies (Valegard et al., 2004; Zhou et al., 2001).

To date, a total of 17 crystal structures of scDAOCS have been reported and deposited in the Protein Data Bank (PDB) (Table 2.6). The first structure published by

Valegard and co-workers (1998) provided an important view of the tertiary folding of the enzyme and physical evidence of residues involved in the iron centre and interaction with the co-substrate. The study revealed that scDAOCS apoenzyme exists as a trimeric unit, where the C-terminus of one molecule (residues 308–311) penetrates the active site of its neighbouring molecule in a cyclical fashion. However, upon addition of iron (II) and 2-oxoglutarate, scDAOCS dissociates into monomers.

These observations were supported by gel filtration and dynamic laser light scattering studies, which revealed the existence of an equilibrium between the monomeric and trimeric scDAOCS in the absence of iron (II), and a shift towards the monomeric form upon addition of the co-factor (Lloyd et al., 1999b). These findings implied that monomeric scDAOCS is the catalytically active form facilitating the oxidative reaction.

The lack of a substrate in the crystal structure has, then, limited understanding on the catalytic mechanism of scDAOCS. The stable native trimeric structure of scDAOCS made crystallisation of an enzyme–substrate complex difficult, as the active site is partially occupied by the C-terminal tail of the neighbouring enzyme.

Truncation of the C-terminal tail helped to prevent trimer formation but the method resulted in a loss of enzymatic activity. Oster and co-workers (2004) reported that the attachment of a His-tag at the N-terminus of scDAOCS disrupts trimer formation and liberates the C-terminal arm of the apoenzyme. However, analyses of the biochemical and structural data revealed significant changes in the substrate-binding affinity and 64

orientation of the penicillin nucleus in the His-tagged scDAOCS–substrate complex structure compared with those in the native enzyme–substrate structure published by the same group (Valegard et al., 2004). These discrepancies suggest caution when selecting the appropriate structural template for subsequent studies. More importantly, the data illustrated that the activity of scDAOCS very much depends on the stability of the enzyme maintained by equilibrium in a pool of monomeric and trimeric forms of scDAOCS, which can be disrupted with the introduction of a relatively small His- tag at the N-terminus of the enzyme. Nevertheless, in both types of structures, the C- terminus remained disordered in the enzyme–substrate complex. This implies a degree of flexibility that may be crucial for catalytic activity, as supported by mutagenesis studies (Chin and Sim, 2002; Chin et al., 2001; Lee et al., 2002; Lee et al., 2001a).

The elucidation of scDAOCS structures with iron (II), 2-oxoglutarate, and penicillins was achieved by soaking the crystals of apo-scDAOCS with iron (II), followed by soaking in solution containing 2-oxogluatrate and either penicillin G or ampicillin. The structural information obtained provides the basis of the current

“booby-trap” mechanism hypothesis (Valegard et al., 2004). The postulated mechanism involves initial activation of the iron centre by 2-oxoglutarate binding, followed by oxidative decarboxylation of 2-oxoglutarate to generate oxidising ferryl species with succinate and carbon dioxide as by-products. Succinate remains bound and stabilises the reactive iron species until the penicillin substrate enters the active site. Binding of penicillin to the reactive iron species via the sulphur group subsequently expels succinate and leads to oxidative ring-expansion of the penicillin substrate. This study also illustrated the sharing of overlapping binding sites for substrate and co-substrate, which implies mutual exclusivity in substrate and co- 65

substrate binding. This requirement of an ordered sequential binding for catalysis supported earlier findings on the importance of the order of adding reagents in the enzyme (Shen et al., 1984). Furthermore, the difference in the orientation of penicillin substrates in DAOCS complexes formed with or without 2-oxoglutarate suggests that

2-oxoglutarate is necessary to influence the orientation of penicillin binding. This is in accordance to the requirement of 2-oxoglutarate to activate the iron centre in order to elicit oxidative ring-expansion. Hence, the orientation of the penicillin substrate in

DAOCS formed in the presence of 2-oxoglutarate, which also resembles the orientation of product DAOC co-crystallised with DAOCS, advocated a productive ring-expansion of penicillin (Valegard et al., 2004).

Apart from providing insights to the mechanism of catalysis, the crystal structures of DAOCS with penicillins have also served as platforms to identify important residues that interact with the substrate. Some of these residues have the potential to improve the enzymatic activity of DAOCS by mutational alteration.

Figure 2.11. Crystal structure of scDAOCS (PDB: 1UNB). This figure depicts the characteristic β-barrel and conserved HXD/EXnH and RXS motifs in the Fe (II)/2- oxoglutarate dependent dioxygenase enzyme family. Residues H183, D185 and H243 are involved in binding Fe (II) ion whereas S260 and R258 stabilise the binding of 2- oxoglutarate (2OG) through hydrogen bond interactions (Valegard et al., 2004).

Table 2.6. Crystal structures of DAOCS deposited in PDBa.

Enzyme Crystallised form Resolution (Ǻ) PDBb References scDAOCS Apoprotein 1.3 1DCS (Valegard et al., 1998) Fe (II) 1.5 1RXF Fe (II), 2-oxoglutarate 1.5 1RXG

scDAOCSc R258Q Fe (II), α-keto-β-methyl-butanoate 1.5 1HJG (Lee et al., 2001b) R258Q Fe (II), α-keto-β-methyl-pentanoate 1.6 1HJF

scDAOCS ΔR306 Fe (II), 2-oxoglutarate 2.1 1E5I (Lee et al., 2001a) ΔR307A Fe (II), succinate, CO2 1.96 1E5H

scDAOCS Fe (II), succinate 1.5 1UO9 (Valegard et al., 2004) Fe (II), penicillin G 1.6 1UOF Fe (II), deacetoxycephalosporin C 1.7 1UOG Fe (II), 2-oxoglutarate, penicillin G 1.7 1UOB Fe (II), 2-oxoglutarate, ampicillin 1.5 1UNB

scDAOCSd Apoprotein 2.3 1W28 (Oster et al., 2004) Fe (II) 2.51 1W2A Fe (II), ampicillin 2.7 1W2N Fe (II), deacetoxycephalosporin C 3.0 1W2O

scDAOCS 5-hydroxy-4-keto valeric acid 1.53 2JB8 (Cicero et al., 2007) aThe list of crystal structures are arranged according to chronological order bPDB accession number cscDAOCS with mutational alteration dscDAOCS with N-terminal His-tag

2.7. Deacetylcephalosporin C synthase

In the β-lactam biosynthetic pathway, deacetoxycephalosporin C (DAOC) is hydroxylated to deacetylcephalosporin C (DAC) by deacetylcephalosporin C synthase

(DACS), which is also known as DAOC hydroxylase. This enzyme adds a hydroxyl group to C3 of the dihydrothiazine ring of DAOC. A similar reaction is also carried out by DAOC/DACS from A. chrysogenum. However, DAOC/DACS is a bifunctional enzyme that also possesses the ring-expansion activity of DAOCS. Currently, only four monofunctional bacterial DACSs have been annotated (Table 2.7). Since scDAOCS and A. chrysogenum DAOC/DACS show a high degree of structural similarity, it is difficult to predict the amino acid residues responsible in conferring the ring-expansion and hydroxylation activities. Hence, by characterising the S. clavuligerus DACS (scDACS), we may get further insight into the regions required for activity or substrate binding and how they differ from those involved in DAOCS activity.

Table 2.7. Summary of annotated deacetylcephalosporin C synthases.

GenBank Protein Organism Activity References accession no. length (Baker et al., 1991; Kovacevic S. clavuligerus NRRL 3858 M63809; P42220 318 Active and Miller, 1991)

Streptomyces 65PH1 AY318744 312 ND (Hsu et al., 2004)

N. lactamdurans P42219 310 Active (Coque et al., 1996)

L. lactamgenus YK90 CAA39985 313 UD (Kimura et al., 1996)

A. chrysogenuma P11935 332 Active (Samson et al., 1987) ND-not determined UD-undetected aA bifunctional enzyme encoded by cefEF, which possesses both deacetoxycephalosporin C synthase and deacetylcephalosporin C synthase activities

2.7.1. Biochemical and structural relationships of scDACS and scDAOCS

The open-reading frame of cefF gene of S. clavuligerus NRRL 3858 has been deciphered and deposited in GenBank (Kovacevic and Miller, 1991). It is located in the cephamycin C biosynthetic gene cluster but transcribed in an opposite direction from cefE gene (Figure 2.12). It encodes a protein with a molecular mass of 34, 584

Da (Kovacevic and Miller, 1991). Its protein primary sequence bears approximately

59% identity to that of scDAOCS. At the DNA level, cefF gene bears approximately

71% identity to that of cefE gene (Kovacevic and Miller, 1991). Hence, it was postulated that both cefE and cefF genes might have evolved from a gene duplication event in S. clavuligerus.

Earlier studies on scDACS depended on the purification of the native protein from cell-free extracts of S. clavuligerus by three anion-exchange chromatographies, ammonium sulphate fractionation and two gel filtrations. Similar to scDAOCS, scDACS requires ferrous iron, molecular oxygen and 2-oxoglutarate for its activity

(Baker et al., 1991). The Km values of scDACS for DAOC and 2-oxoglutarate were 59 and 10 µM, respectively (Baker et al., 1991). Other than hydroxylating DAOC to

DAC, it could also hydroxylate a DAOC analogue 3-exomethylenecephalosporin C to

DAC at approximately 37% relative activity of DAOC hydroxylation (Baker et al.,

1991). Although scDACS is structurally homologous to scDAOCS with 74% amino acid sequence similarity, it could only ring-expand penicillin N at about 3% activity of DAOC/DACS (Baker et al., 1991). Interestingly, the DAOC hydroxylation activity of scDACS was inhibited in the presence of penicillin N with a Ki of 0.12 mM, indicating a competitive occupancy of the substrate binding site of scDACS (Baker et al., 1991).

pcbC pcbAB lat blp ORF11 ccaR cmcH cefF cmcJ cmcI cefD cefE

Figure 2.12. Cephamycin C biosynthetic gene cluster of S. clavuligerus NRRL3585. The arrows refer to the promoters and direction of gene transcription of respective polycistronic genes.

2.7.2. Bifunctional DAOC/DACS enzyme of Acremonium chrysogenum

Cephalosporin C is the first cephem antibiotic to be discovered. It was produced by A. chrysogenum, a fungus discovered by Giuseppe Brotzu, an Italian scientist, in 1945. It was found to inhibit the growth of a number of Gram negative bacteria, e.g. Salmonella typhi, Salmonella paratyphi B, Yersinia pestis, Brucella melitensis and Vibrio cholera, where such bioactivity was not found in earlier penicillins (Hamilton-Miller, 2000). However, due to the lack of expertise and finances to purify the bioactive compound produced by A. chrysogenum, this fungus was subsequently brought to Oxford University where studies on the bioactive compound were continued.

The first bioactive compound isolated from A. chrysogenum was named cephalosporin P due to its narrow antibacterial spectrum against mainly Gram positive bacteria. Only the second bioactive compound isolated was able to account for the antibacterial spectrum observed by Brotzu. This compound was named cephalosporin

N due to its antibacterial spectrum including Gram negative bacteria, but was later changed to penicillin N since its structure was determined to be a penicillin.

Interestingly, a third compound was found in the purest preparation of penicillin N, which has significantly different physical properties from penicillin N, e.g., ultraviolet

absorption spectrum and pH stability. It was also more resistant to penicillinase than penicillin N although it has a weaker antibacterial activity. Being the third antibiotic isolated from A. chrysogenum, this compound was named cephalosporin C (Hamilton-

Miller, 2000).

Studies on the biosynthesis of cephalosporin led to the discovery of the enzymes involved. Purification of the native enzyme involved in the biosynthesis of cephalosporin led to the postulation that the enzyme was a bifunctional enzyme since the physical enzyme could not be separated. However, in S. clavuligerus, ring- expansion and hydroxylation is carried out by two separate enzymes.

Interestingly, DAOC/DACS is structurally homologous to monofunctional

DAOCS and DACS; it bears 57% and 54% amino acid sequence identities to scDAOCS and scDACS, respectively (Lloyd et al., 2004). It also possesses an additional 20 amino acid residues at its C-terminus that has been shown to be important for its hydroxylation activity but not for its ring-expansion activity, via truncation studies using penicillin G as a substrate (Chin et al., 2003; Lloyd et al.,

2004).

Mutational studies of DAOC/DACS have also provided some clues to the determinants of ring-expansion and hydroxylation in DAOC/DACS (Table 2.8). This was aided by multiple sequence alignment of DAOCSs and DACSs of S. clavuligerus and N. lactamdurans. Amino acid residues that were common in DAOCs and differ consistently from DACS were selected for analyses (Lloyd et al., 2004). For example, methionine is common among DACSs at position 306 of DAOC/DACS, while it is isoleucine in the corresponding position in DAOCSs. By replacing residue M306 with isoleucine, which corresponds to residue I305 in scDAOCS, hydroxylation of DAOC by DAOC/DACS was abolished accompanied with the halving of its ring-expansion 72

activity (Lloyd et al., 2004). This indicated the importance of the methionine residue in hydroxylation activity and for higher ring-expansion activity. Another example is residue W82 of DAOC/DACS, where the corresponding residue in DAOCSs and

DACSs is consistently serine and alanine residues, respectively. Upon mutation of residue W82 to either serine or alanine residues, predominant ring-expansion or hydroxylation activity was retained (Lloyd et al., 2004).

In another study, a similar approach to the mutational study of the C-terminal residues N304, R306 and R307 of scDAOCS was adopted for understanding the importance of C-terminal residues N305, R307 and R308 of DAOC/DACS (Chin and

Sim, 2002; Wu et al., 2005). These three residues were each mutated to leucine residue and only N305L and R308L single mutations showed significant improvement in the conversion of penicillin G and other analogues (Wu et al., 2005).

Further analyses were carried out on residue R308 of DAOC/DACS to understand its role in controlling substrate specificity by mutation to the other 19 amino acids (Wu et al., 2011). Only R308L, R308I, R308T and R308V mutations showed significant improvement in the conversion of penicillin G and other analogues, with R308I mutation showing more than a seven-fold increase in the relative activity and a three-fold increase in kcat/Km value of penicillin G conversion when compared to the wild-type DAOC/DACS. It is postulated that the four amino acid residues, which all possess short aliphatic side-chains, may improve hydrophobic interactions with the substrates, thus enhancing the ring-expansion activity of

DAOC/DACS (Wu et al., 2011). However, the determining factors in defining the ring-expansion and hydroxylation activities of monofunctional DAOCSs and DACSs, respectively, have yet to be validated.

Table 2.8. Relative ring-expansion and hydroxylation activities of wild-type and mutant DAOC/DACSs using penicillin G and DAOC as respective substrates (Lloyd et al., 2004).

Enzyme Relative ring-expansion activity (%) Relative hydroxylation activity (%) Wild-type 100 100

W82A 5.5 71

W82S 44 18

N305L 107 85

M306I 59 <2

∆310 203 50

∆328 64 45

∆310/M306I 185 3

∆310/N305L/M306I 370 2

2.8. Enzyme improvement strategies for improving DAOCS

Biocatalysis has become increasingly important for the pharmaceutical industries especially in the last decade to reduce hazardous waste and to be socially responsible in protecting the environment (Tucker, 2006). However, the industrial application of enzymes has been hampered by several reasons, namely limited enzyme availability, substrate specificity, catalytic rate and operational stability

(Chaput et al., 2008). Advances in protein engineering have provided many options to overcome these factors, and many strategies have been proposed. For example, one common aim in protein engineering is to improve the thermostability of an enzyme for industrial production of chemical compounds (Bottcher and Bornscheuer, 2010).

This enzyme has to function at elevated temperatures, which can reduce the chances of microbial contamination during the process. Many studies demonstrated that a rigid enzyme will be more stable at higher temperature. Hence, specific stabilising interactions such as disulphide bonds can be rationally introduced into the structure of the enzyme if its tertiary structure is available (Kazlauskas and Bornscheuer, 2009).

Importantly, the technique of introducing changes in the protein sequence was revolutionised from earlier chemical or physical random mutagenesis methods to

SDM at the genetic level (Peracchi, 2001).

2.8.1. Mutagenesis: a key technology in improving enzyme’s functions

The advent and rapid progress of recombinant DNA technology has overcome the need to purify the native protein from the organism-of-interest and allowed the possibility to produce a large amount of recombinant protein in a heterologous host, e.g., E. coli, which is safer to handle and has high proliferation rate. A significant methodology was devised and reported by Michael Smith in 1978, which earned him 75

a Nobel Prize in Chemistry in 1993 together with Kary Mullis, who invented polymerase chain reaction (Shampo and Kyle, 2003). This method introduces specific mutations at desired positions of a gene, resulting in changes in the translated protein sequence (Hutchison et al., 1978). This allowed for a precise study on the effect of changes at a specific position of a protein to provide clues on the function of the protein. This method has also led to a new area of research known as “protein engineering”, which exploits this technique to modify an existing protein such that it possesses desired properties such as improved stability and activity or alteration in substrate specificity. The field of protein engineering has also expanded intensively due to the development of bioinformatics, X-ray crystallography and nuclear magnetic resonance (NMR), which provides important structural information of the protein for protein design and engineering. Together, this advancement in technology has provided the opportunity for the pharmaceutical industries to exploit enzymatic processes for greener production of pharmaceutical products.

2.8.2. Two main approaches that drive protein engineering: rational versus directed evolution

Natural enzymes have evolved to function in vivo and not outside their natural environment. Hence, they are often poorly suited for industrial application, particularly for industrially preferred substrates and industrial processes (Chaput et al., 2008). Protein engineering thus becomes a very important tool to overcome these limitations.

Generally, there are two approaches in protein engineering, i.e., rational approach and directed evolution approach (Bornscheuer and Pohl, 2001; Jestin and

Vichier-Guerre, 2005). The directed evolution approach mainly exploits random 76

mutagenesis techniques to generate at least one mutation in the protein and couple with a screening or selection process to obtain the mutant protein with desired functionality. Hence, it can be performed without prior knowledge of the enzyme tertiary structure but often requires a large recombinant library to increase the chance of obtaining a clone with the desired property. The rational approach uses SDM to incorporate specific mutation(s) at relevant site(s) in a protein in order to confer a desired property. In this approach, structural information of the protein is necessary to allow precise sequence modification of the enzyme to achieve the desired property

(Table 2.9). Often, this approach is hampered by inadequate knowledge in the structure-function relationship of the protein, resulting in a lower success rate of obtaining desired protein function compared to the directed evolution approach. In addition, the knowledge available in protein science has yet to allow the precise selection of beneficial amino acid substitutions to achieve desired properties.

Table 2.9. Examples of structure-aided mutagenesis studies in improving enzymes for industrial applications.

Properties of protein Protein Applications References Related patents altered α-Amylase Thermostability Industrial processes of starch Declerck et al. 6,008,026 (USPTO*) hydrolysis 1995, 1997, 2000.

β-Amylase Thermostability Industrial processes of starch Okada et al. 1995. 5,726,057 (USPTO) hydrolysis

Cephalosporin Substrate specificity: Enzymatic production of 7- Oh et al. 2003; WO/2005/014821 acylase from glutaryl-7- aminocephalosporanic acid Otten 2004. (WIPO*) aminocephalosporanic from cephalosporin C acid to cephalosporin C

DAOCS Substrate specificity: Enzymatic production of 7- Chin et al. 2001, 6,905,854 (USPTO) from penicillin N to ADCA from penicillin G 2004; Wei et al. penicillin G 2003, 2005; Hsu et al. 2004.

DNA polymerase Improvement of the 3'-5' Polymerase chain reaction Park et al. 1997. - exonuclease activity

Glucoamylase Thermostability Industrial processes of starch Chen et al. 1996. 6,352,851 (USPTO) hydrolysis 6,537,792 (USPTO)

Laccase Shifting optimal pH of Degradation of lignin; Madzak et al. 2006. - activity development of oxygen cathodes in biofuel cells, green biodegradation of xenobiotics, biosensors, organic synthesis

Lactate oxidase Thermostability and A component of Hamamatsu et al. - catalytic efficiency electrochemical sensor systems 2006; Kaneko et al. for monitoring lactate 2005. concentration in blood

Lipase Increase activity at oil- Detergent lipase Martinelle et al. 7,312,062 (USPTO) water interface 1996; Svendsen A. 1996.

Penicillin acylase Catalytic efficiency Enzymatic production of 6- Gabor and Janssen. WO/2004/111241 APA from penicillins 2004; Arroyo et al. (WIPO) 2003.

Subtilisin Stability Detergent protease; model Almog et al. 1998, 6,541,234 (USPTO) system for protein engineering 2002; Bryan et al. studies 1992; Bryan PN. 2000; Takagi et al. 2000; Maurer KH. 2004. *USPTO: United States Patent and Trademark Office; WIPO: World Intellectual Property Organisation

2.8.3. Enzyme modification to improve the substrate binding and catalytic activity of DAOCS: directed evolution approach

The directed evolution approach mimics the evolutionary process of proteins by exploiting random mutagenesis techniques, for example homologous and non- homologous recombinations, error-prone PCR, and chemical and physical mutagenesis methods, coupled with an efficient screening or selection method to obtain proteins with the desired properties. The first law of this approach, “you get what you screen for”, always applies. Hence, the screening method is an important part of the whole experiment. The design of the screening method has to ensure efficient detection of a product, which is a reflection of the desired property of the protein, and yet, the outcome is not always guaranteed. Often, a large library of recombinants is required and the number of desired recombinants could be very low.

Nevertheless, this approach has also provided some important information on the structure–function relationship of DAOCS.

2.8.3.1. Homologous recombination

The first reported attempt to improve the activity of DAOCS by the directed evolution approach was probably the construction of hybrid DAOCS of S. clavuligerus and N. lactamdurans by Adrio and co-workers using in vivo homologous recombination (Adrio and Demain, 2002; Adrio et al., 2002). This technique exploited the advantage of natural recombination at homologous regions of closely related sequences in a host organism to generate a recombinant enzyme. Rapid visual screening was set up by exploiting a tyrosinase melC gene from S. glaucescens, which encodes the ability to secrete the black pigment, melanin. The melC gene was inserted between the two cefE genes in a pJA680 vector and recombination of the two cefE 79

genes would result in loss of pigment production by the host organism. Using S. lividans strain 1326, a recombinant strain W25 was generated and found to have higher penicillin G conversion activity than the S. clavuligerus NP1 strain (Adrio and

Demain, 2002; Adrio et al., 2002). Further analyses revealed that this recombinant strain preferred a lower reaction temperature (<30°C), a lower iron (II) concentration, and a higher level of 2-oxoglutarate for penicillin G conversion than S. clavuligerus

NP1 strain (Gao et al., 2003).

A similar method was adopted by Hsu and co-workers (2004) to improve the substrate specificity of DAOCS for penicillin G, except that it is an in vitro recombination method and more than two gene sequences were used for shuffling.

The cefE genes from S. clavuligerus (ATCC 27064), N. lactamdurans (ATCC 27382),

S. jumonjinensis (ATCC 29864), S. ambofaciens (102SH3), and S. chartreusis

(29SA4), the cefF gene from S. clavuligerus (ATCC 27064), which encodes for deacetylcephalosporin C synthase, and the cefEF gene from A. chrysogenum (ATCC

11550), were used for the shuffling experiment. These shuffled genes were heterologously expressed in E. coli cells and the transformants were screened for penicillin G conversion activity. After two rounds of gene shuffling and screening, two clones, FF1 and FF8, showed no substrate inhibition and were subjected to kinetic and sequence analyses. FF1 had an increment in kcat/Km value of approximately nine-fold when compared with scDAOCS and FF8 had a 118-fold increment, which is probably the highest improvement recorded so far for penicillin G conversion. Comparison of the amino acid sequence of FF8 with that of scDAOCS revealed that only two of the 35 substitutions, i.e., residues M73 and I305, are at positions known to have direct interactions with the substrate (Hsu et al., 2004).

2.8.3.2. Chemical mutagenesis

Chemical mutagenesis using hydroxylamine was attempted by Wei and co- workers (2003) with the objective of modifying scDAOCS for optimum ring- expansion activity towards penicillin G. The mutant library was first screened by a simple high-throughput paper disk-based bioassay, followed by a second screening method using thin-layer chromatography with radio-labelled penicillin G as a substrate. From 5,500 clones screened, six were found to have increased conversion activity and these were because of three point mutations, G79E, V275I, and C281Y.

As these three sites are located away from the substrate binding pocket, the improvement achieved could have been due to a global effect of these mutations on the conformation of the active site (Wei et al., 2003a).

2.8.3.3. Error-prone PCR combined with gene shuffling

Error-prone PCR exploits the altered fidelity of DNA polymerase in the presence of manganese ions to incorporate mismatched nucleotides during PCR amplification of the gene of interest. Wei and co-workers (2005) used a ratio of

Mg(II) ions to Mn(II) ions of 12:1 in the error-prone PCR process to induce mutations. Approximately 6,000 clones were subjected to a screening process, where six types of point mutations (M73T, T91A, M188V, M188I, H244Q, G300V) and six types of double mutations (D53H/C281Y, M73T/P145L, A106T/C155Y,

A106T/M188V, D107G/L277Q, Y184H/C281R) were identified. Subsequently, single mutations based on the double mutations were constructed to assess the effect of each substitution. Among these mutations, the L277Q mutant had the highest catalytic efficiency for penicillin G (six-fold increment), whereas the M73T, G79E,

T91A, Y184H, M188V, and H244Q mutants had approximately two- to five-fold 81

increment in kcat/Km values when compared with the wild-type enzyme. Other mutations either had activities similar to that of the wild-type enzyme or exhibited substrate inhibition. Gene shuffling of these single mutants, including those reported previously, i.e., G79E, V275I, C281Y, N304K, I305L, and I305M mutants (Wei et al., 2003a), was then conducted to generate random combinations. Approximately

10,000 clones were screened and only one quaternary mutant

(C155Y/Y184H/V275I/C281Y) with enhanced activity and without substrate inhibition was subjected to kinetic analysis. For this mutant, the kcat/Km value for penicillin G conversion was a further seven-fold higher than for the C281Y single mutant, which had the highest kcat/Km value among these four single mutants (six-fold higher than that of the wild-type enzyme). Table 2.10 summarises the kinetic data for selected examples of mutant scDAOCSs with improved activities obtained by Wei and co-workers (2003, 2005) and Hsu and co-workers (2004).

Table 2.10. Kinetic parameters of bioengineered DAOCSs in penicillin G conversion (Hsu et al., 2004; Wei et al., 2003a; Wei et al., 2005).

-1 -1 -1 Mutations Km (mM) kcat (s ) kcat/Km (M s ) Relative kcat/Km Wild-type 2.58 0.0453 18 1

M73T 0.74 0.0627 85 4.7

G79E 0.75 0.0315 42 2.3

T91A 1.36 0.0522 38 2.1

C155Y 1.76 0.0476 27 1.5

Y184H 0.95 0.0798 84 4.7

M188V 0.76 0.0456 60 3.3

H244Q 1.39 0.0698 50 2.8

V275I 1.68 0.0502 30 1.7

L277Q 1.02 0.1036 102 5.7

C281Y 0.68 0.0744 109 6.1

N304K 0.22 0.0564 256 14.2

I305L 0.66 0.0759 115 6.4

I305M 0.75 0.1452 194 10.8

V275I/I305M 0.25 0.1458 583 32.4

C155Y/Y184H/V275I/C281Y 0.19 0.1398 736 40.9

FF8a 0.014 0.0297 2,121 117.8 aRecombinant derived from gene shuffling (Hsu et al., 2004).

2.8.3.4. Comparison of random mutagenesis methods

Comparing these techniques, hydroxylamine mutagenesis generated the least number of mutants with improved penicillin G conversion activities. This chemical targets mainly the cytosine residues in DNA, which limits the repertoire of mutations

(Tessman et al., 1965). In contrast, error-prone PCR generated a reasonably large number of mutants with improved activities. However, this method has errors biased toward AT to GC changes (Cirino et al., 2003). Furthermore, codon redundancy and possible introduction of stop codons lower the efficiency of this method in generating desired mutations. Gene shuffling has been successfully employed in generating an efficient recombinant DAOCS with catalytic efficiency in penicillin G conversion comparable with that of scDAOCS in penicillin N conversion. Although it is still not clear which amino acid differences between this recombinant enzyme and scDAOCS are responsible for this substantial improvement, this technique is a promising route in generating recombinants with desired substrate specificity, especially when more

DAOCS orthologues are available for shuffling.

Together, these studies suggest that the substrate specificity of DAOCS is not solely governed by the substrate-binding residues. Amino acid residues outside the substrate-binding site may also affect the substrate specificity of the enzyme and these residues are often difficult to identify via the rational approach (Figure 2.13).

However, the effects of the mutations identified cannot be rationalised without structural information about the enzyme. In order to understand the molecular basis of such mutations in influencing the function of the enzyme, further analyses requiring the rational approach must be performed.

Figure 2.13. Spatial locations of useful mutations in scDAOCS (PDB: 1UNB) identified by using directed evolution approaches. Among the mutations listed in Table 2.11, only M73T, N304K and I305M/L mutations are localised in the substrate binding site with the side chains of the amino acid residues interacting with the penicillin substrate.

2.8.4. Enzyme modification to improve the substrate binding and catalytic activity of DAOCS: rational approach

The three-dimensional structure of the DAOCS-Fe(II)-2-oxoglutarate complex was resolved in the late 1990s (Lloyd et al., 1999b; Valegard et al., 1998) but the lack of bound substrate hindered many subsequent studies. It was recognised that the crystal structure of DAOCS resembles that of IPNS (Roach et al., 1997; Valegard et al., 1998). The crystal structure of the DAOCS–Fe(II)–2-oxoglutarate complex was superimposed on that of the IPNS–Fe(II)–ACV complex to predict the possible substrate binding pocket of scDAOCS (Chin et al., 2001; Valegard et al., 1998).

Subsequently, various mutational studies aimed at understanding the substrate binding mechanism of DAOCS have revealed certain amino acid residues that are important for catalysis and others that are amenable to modification to improve the functions of the enzyme (Table 2.11).

In one of the studies, residues R74, R160, R266, and N304 were selected for

SDM studies based on the close proximity to the L-α-aminoadipyl branch of ACV and their hydropathic properties (Chin et al., 2001). This study showed that only the

N304L mutation increased its penicillin analogue conversion activity. The R74L,

R160L, and R266L mutations abolished the activity of scDAOCS and this indicated the importance of these residues in binding the penicillin substrate. Similarly, Lee and co-workers (2002) also showed that the N304A mutation was able to increase the activity of the enzyme. Thereafter, Wei and co-workers (2003) also adopted the same approach and reported that the N304K, I305L, and I305M mutations were able to increase the activity of scDAOCS (Table 2.10). Although results from various structural studies suggested that the C-terminal of scDAOCS (residues 301–311) might be involved in guiding the substrate into the substrate binding pocket (Lee et 86

al., 2001a), the rationale behind these mutations could not be clearly explained in the absence of a crystal structure of DAOCS complexed with a penicillin substrate.

In order to elucidate the molecular basis of substrate interaction by residue

N304, Chin and co-workers (2004) performed a study using a complete library of amino acid alterations and found that mutation to basic residues (arginine and lysine) has better effects than mutation to aliphatic residues (alanine, leucine, and methionine) in the conversion of seven hydrophobic penicillins. In the absence of the structural information on substrate interaction, these replacements harbouring a strictly basic or aliphatic side chain were postulated to likely incorporate favourable hydrophobic or charged interaction between these mutant enzymes and their prime substrates, which could enhance the efficacy of these mutant enzymes (Chin et al.,

2004).

Table 2.11. Summary of mutagenesis studies of scDAOCS substrate binding pocket.

Mutation Substrate Assay method Findings References R74L Penicillin G Bioassay Penicillin conversion abolished Chin et al., 2001

M180F Penicillin N & HPLC Penicillin N conversion reduced Lee et al., 2002 penicillin G by half; penicillin G conversion abolished

R160Q Penicillin N & HPLC Both penicillin conversions Lipscomb et al., penicillin G abolished 2002

R160L Penicillin G Bioassay Penicillin conversion abolished Chin et al., 2001

R162A, Penicillin N & HPLC Both penicillin conversions Lipscomb et al., R162Q penicillin G abolished 2002

R266L Penicillin G Bioassay Penicillin conversion abolished Chin et al., 2001

N304A, Penicillin G Bioassay ~120 to 260% of wild-type Chin et al., 2004 N304L, activity N304K, N304M, N304R

I305L, Penicillin G HPLC 1.7- and 3.2-fold increased in Wei et al., 2003 I305M kcat, and 3.9- and 3.2 fold decreased in Km, respectively

Chapter 3: Materials and methods

3.1. Materials

3.1.1. Bacterial cultures

The bacterial strains used for cloning, expression of recombinant proteins and bioassay are listed in Table 3.1.

Table 3.1. List of bacterial strains used.

Strain Characteristics Reference/Source S. clavuligerus NRRL Cephamycin C and clavulanic acid American Type Culture 3585 (ATCC 27064) producer Collection Escherichia coli BL21 Expression host, F-, carries T7 RNA Studier and Moffatt, 1986, (DE3) polymerase gene under control of the Novagen lacUV5 promoter E. coli Ess Mutant of E. coli, a β-lactam Aharonowitz and Demain, supersensitive strain 1978; a gift from Prof. Demain A. (Drew University) E. coli TOP10 Cloning host, F-, used in TOPO cloning Invitrogen Micrococcus luteus Gram positive bacterium, spherical, American Type Culture ATCC 381 catalase positive, form bright yellow Collection punctiform colonies of <1 mm E. coli K12 Gram negative bacterium, straight rods, American Type Culture lactose fermenter with acid production, Collection methyl red positive, Voges-Proskauer negative, Simmons citrate negative Staphylococcus aureus Gram positive bacterium, spherical, American Type Culture ATCC 9144 occurring in clusters, catalase positive, Collection form light yellow colonies of ~1 mm Enterobacter aerogenes Gram negative bacterium, straight rods, Department of Microbiology, lactose fermenter with acid production, laboratory isolate methyl red negative, Voges-Proskauer positive, Simmons citrate positive Salmonella cholerasuis Gram negative bacterium, straight rods, American Type Culture subsp. cholerasuis (Smith) non-lactose fermenter, methyl red Collection Weldin serovar Infantis positive, Voges-Proskauer negative, ATCC 51741 Simmons citrate positive Enterococcus faecalis Gram positive bacterium, spherical, American Type Culture ATCC 29212 occurring in pairs, catalase negative Collection

3.1.2. Chemicals

Chemicals of analytical/reagents grade were obtained from BDH Limited, UK

(BDH), Bio-Rad Laboratories Private Limited, Singapore (Bio-Rad) and Sigma-

Aldrich Private Limited, Singapore (Sigma). HPLC-grade solvents were purchased from Fisher-Scientific. Chemicals of DNA and electrophoresis grade were obtained from UltraPure™, GIBCO Bethesda Research Laboratories, USA (BRL). Culture media were obtained from Oxoid Limited, UK. Antibiotics were purchased from

Sigma-Aldrich Private Limited, Singapore. Unless otherwise stated, all buffers and culture media were sterilised by autoclaving for 15 min at 121ºC. Thermolabile components were filter-sterilised using Millipore™ syringe filters with a pore size of

0.22 µm, and were stored at -20ºC. They were thawed prior to use and were aseptically added to autoclaved solutions.

3.1.3. Plasmid vectors

The plasmid vectors used in this study are listed in Table 3.2. The restriction maps of pCR®-Blunt, pGEX-6P-1 and pGK vectors are illustrated under Appendix I.

3.1.4. Oligonucleotide primers

The oligonucleotide primers used for SDM (Table 3.3), gene amplification and sequencing experiments (Table 3.4) were synthesised by 1st Base Private Limited,

Singapore, or Sigma-Aldrich Private Limited, Singapore.

Table 3.2. List of plasmid vectors used.

Vector Characteristics Reference/Source pGEX-6P-1 Ampr, lacIq, tac promoter preceding gluthathione- GE Healthcare S-transferase (GST) open-reading frame, PreScission™ protease recognition site (LEVLFQ↓GP) at C-terminus of GST, preceding MCS pGK A derivative of pGEX-6P-1, Kanr Loke and Sim, 2000 pscEXP-GST pGK vector carrying scDAOCS gene in the MCS Sim and Sim, 2001 pscEXPN304L-GST pGK vector carrying N304L scDAOCS mutant Chin et al., 2001 gene in the MCS pscEXPN304A-GST pGK vector carrying N304A scDAOCS mutant Chin et al., 2004 gene in the MCS pscEXPN304K-GST pGK vector carrying N304K scDAOCS mutant Chin et al., 2004 gene in the MCS pscEXPN304M-GST pGK vector carrying N304M scDAOCS mutant Chin et al., 2004 gene in the MCS pscEXPN304R-GST pGK vector carrying N304R scDAOCS mutant Chin et al., 2004 gene in the MCS pscEXPR306L-GST pGK vector carrying R306L scDAOCS mutant Chin et al., 2002 gene in the MCS pCR®-Blunt II Cloning vector, Kanr, topoisomerase I covalently Invitrogen TOPO® bound to the vector allows direct insertion of blunt-end PCR products

Table 3.3. Mutagenic primers used in generating scDAOCS mutants. Gene Mutations Direction Sequence of mutagenic primers* scDAOCS V275I Forward 5’-CTTCACCTTCTCCATCCCGCTGGCGCGC-3’ Reverse 5’-GCGCGCCAGCGGGATGGAGAAGGTGAAG-3’ V275L Forward 5’-CACCTTCTCCCTCCCGCTGGCGCGC-3’ Reverse 5’-CGCCAGCGGGAGGGAGAAGGTGAAGTC-3’ C281F Forward 5’-CTGGCGCGCGAGTTCGGCTTCGATGTC-3’ Reverse 5’-GACATCGAAGCCGAACTCGCGCGCCAG-3’ C281Y Forward 5’-CTGGCGCGCGAGTACGGCTTCGATGTC-3’ Reverse 5’-GACATCGAAGCCGTACTCGCGCGCCAG-3’ I305M Forward 5’-GCAACTACGTGAACATGCGCCGCACATCC-3’ Reverse 5’-GGATGTGCGGCGCATGTTCACGTAGTTGC-3’ R306A Forward 5'-CTACGTGAACATCGCCCGCACATCCAAGG-3' Reverse 5'-CCTTGGATGTGCGGGCGATGTTCACGTAG-3' R306K Forward 5'-CTACGTGAACATCAAGCGCACATCCAAGG-3' Reverse 5'-CCTTGGATGTGCGCTTGATGTTCACGTAG-3' R306I Forward 5’-AACTACGTGAACATCATCCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGATGATGTTCACGTAGTT-3’ R306V Forward 5’-AACTACGTGAACATCGTCCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGACGATGTTCACGTAGTT-3’ R306F Forward 5’-AACTACGTGAACATCTTCCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGAAGATGTTCACGTAGTT-3’ R306C Forward 5’-AACTACGTGAACATCTGCCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGCAGATGTTCACGTAGTT-3’ R306W Forward 5’-AACTACGTGAACATCTGGCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGCCAGATGTTCACGTAGTT-3’ R306Y Forward 5’-AACTACGTGAACATCTACCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGTAGATGTTCACGTAGTT-3’ R306H Forward 5’-AACTACGTGAACATCCACCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGTGGATGTTCACGTAGTT-3’ R306M Forward 5’-AACTACGTGAACATCATGCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGCATGATGTTCACGTAGTT-3’ R306N Forward 5’-AACTACGTGAACATCAACCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGTTGATGTTCACGTAGTT-3’ R306D Forward 5’-AACTACGTGAACATCGACCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGTCGATGTTCACGTAGTT-3’ R306E Forward 5’-AACTACGTGAACATCGAGCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGCTCGATGTTCACGTAGTT-3’ R306Q Forward 5’-AACTACGTGAACATCCAGCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGCTGGATGTTCACGTAGTT-3’ R306S Forward 5’-AACTACGTGAACATCAGCCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGCTGATGTTCACGTAGTT-3’ R306T Forward 5’-AACTACGTGAACATCACCCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGGTGATGTTCACGTAGTT-3’ R306G Forward 5’-AACTACGTGAACATCGGCCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGGCCGATGTTCACGTAGTT-3’ R306P Forward 5’-AACTACGTGAACATCCCGCGCACATCCAAGGCA-3’ Reverse 5’-TGCCTTGGATGTGCGCGGGATGTTCACGTAGTT-3’ N304AR306L Forward 5’-GGGCAACTACGTGGCCATCCTCCGCACATC-3’ Reverse 5’-GATGTGCGGAGGATGGCCACGTAGTTGCCC-3’ N304KR306L Forward 5’-GGGCAACTACGTGAAGATCCTCCGCACATC-3’ Reverse 5’-GATGTGCGGAGGATCTTCACGTAGTTGCCC-3’ N304LR306L Forward 5’-GGGCAACTACGTGCTCATCCTCCGCACATC-3’ Reverse 5’-GATGTGCGGAGGATGAGCACGTAGTTGCCC-3’ N304MR306L Forward 5’-GGGCAACTACGTGATGATCCTCCGCACATC-3’ Reverse 5’-GATGTGCGGAGGATCATCACGTAGTTGCCC-3’ N304RR306L Forward 5’-GGGCAACTACGTGAGGATCCTCCGCACATCCAAGG-3’ Reverse 5’-CCTTGGATGTGCGGAGGATCCTCACGTAGTTGCCC-3’ scDACS T308N Forward 5’-CAACTACGTCAACATGCACGCGAAGAAC-3’ Reverse 5’-CTTCGCGTGCATGTTGACGTAGTTGGTG-3’ T308R Forward 5’-CAACTACGTCAGGATGCACGCGAAGAAC-3’ Reverse 5’-CTTCGCGTGCATCCTGACGTAGTTGGTG-3’ H310R Forward 5’-CGTCACCATGCGCGCGAAGAACGAGCCG-3’ Reverse 5’-GTTCTTCGCGCGCATGGTGACGTAGTTG-3’ H310L Forward 5’-CGTCACCATGCTCGCGAAGAACGAGCCG-3’ Reverse 5’-GTTCTTCGCGAGCATGGTGACGTAGTTG-3’ T308NH310R Forward 5’-CTACGTCAACATGCGCGCGAAGAACGAGCCG-3’ Reverse 5’-GTTCTTCGCGCGCATGTTGACGTAGTTGGTG-3’ T308RH310R Forward 5’-CAACTACGTCAGGATGCGCGCGAAGAACGAG-3’ Reverse 5’-GTTCTTCGCGCGCATCCTGACGTAGTTGGTG-3’ *The underlined bold triplets indicate the mutated codons. 92

Table 3.4. List of oligonucleotide primers used in gene amplification and cycle sequencing.

Primer ID Sequences* Description OL74 5'-GCTAGTTATTGCTCAGCGG-3' pGK T7 terminator region reverse primer, for sequencing

OL334 5'-GGGCTGGCAAGCCACGTTTGGTG-3' pGK tac promoter region forward primer, for sequencing

OL111 5’-GACTACTCGATGTGCTACT-3’ scDAOCS internal forward sequence primer, for sequencing

OL112 5’-ACTGGAGGTGCGGCTGCTG-3’ scDAOCS internal reverse sequence primer, for sequencing

OL963 5’-G↓GATCCATGGCGGACACGCCCGTACCG-3’ Cloning of cefF, forward primer with BamHI restriction site

OL964 5’-G↓AATTCTCATCCGGCCTGCGGCTCGTTC-3’ Cloning of cefF, reverse primer with EcoRI restriction site

OL993 5’-AAGGACATCGAGTAGTCCGTGTAG-3’ scDACS internal reverse sequence primer, for sequencing

M13F 5’-GTAAAACGACGGCCAG-3’ pCR®-Blunt II TOPO® M13 forward primer, for sequencing

M13R 5’-CAGGAAACAGCTATGAC-3 pCR®-Blunt II TOPO® M13 reverse primer, for sequencing *The underlined sequences indicate specific restriction endonuclease recognition sites.

3.1.5. Enzymes

Restriction endonucleases were purchased from Promega Private Limited,

Singapore, Stratagene, USA, Fermentas International Inc., Canada and New England

Biolabs Inc., USA. PreScission™ protease was purchased from GE Healthcare

Private Limited, Singapore. Pyrococcus furiosus (Pfu) DNA polymerase was purchased from Promega Private Limited, Singapore, and Difco™ Penase was purchased from Becton Dickinson, USA. T4 DNA ligase was purchased from

Invitrogen Singapore Private Limited, Singapore. All enzymes were used according to the manufacturer’s specifications.

3.1.6. Test kits

The test kits used in this study are listed in Table 3.5. The test kits were used according to manufacturer’s instructions unless otherwise stated.

Table 3.5. List of test kits used.

Test kit Purpose Source QuikChange™ SDM kit Site-directed mutagenesis Stratagene

Wizard™ Plus SV Minipreps DNA Plasmid purification Promega Purification System

ABI PRISM™ BigDye® Terminator DNA sequencing Applied Biosystems v3.1 cycle sequencing kit

MicroSpin™ GST purification module Fusion protein purification GE Healthcare

QIAquick® Gel Extraction kit Extraction of DNA from agarose gel Qiagen

GFX™ Genomic Blood DNA Genomic DNA extraction Amersham Pharmacia purification kit

Calbiochem® InnoZyme™ Human Assessment of protease inhibitors Merck Pte Ltd Neutrophil Elastase Immunocapture on human neutrophil elastase Activity Assay kit

3.1.7. Analytical instruments

The liquid chromatography system used was ÄKTA purifier (GE Healthcare

Private Limited, Singapore) consisting of a UV-900 detector and an Autosampler

A900, and a 5 µm, 4.6 x 150 mm Sun-Fire™ C18 column from Waters Asia Limited,

USA. The operation and data acquisition of the chromatographic process were controlled by UNICORN 5.0 (GE Healthcare Private Limited, Singapore).

A filter-based GENios™ (with a 260 or 595 nm absorbance filter) or Safire™ microplate readers from Tecan Group Limited, Switzerland, was used for determination of protein concentration, protease inhibition assay and enzyme kinetic studies. The operation and data acquisition of the spectrophotometric processes were controlled by Magellan™ software (Tecan). For the determination of nucleic acid concentration, a NanoDrop-1000 spectrophotometer from Thermo Scientific, USA, was used.

3.2. Methods

3.2.1. Growth and maintenance of bacterial cultures

Purity of the bacterial cultures is of utmost importance to ensure the accuracy of the results. Hence, aseptic technique was employed in the handling of all bacterial cultures. Colony morphology and Gram staining was used to identify the bacterial cultures based on the characteristics described in Table 3.1.

All E. coli strains were grown in liquid Luria-Bertani (LB) medium containing

1% (w/v) tryptone, 0.5% (w/v) yeast extract and 1% (w/v) NaCl, and LB agar plates with 1.5% (w/v) agar. Other bacterial cultures were grown in nutrient broth (NB) or on nutrient agar (NA) plates, unless otherwise stated. All bacterial liquid cultures were grown at 37°C on a rotary shaker set at 250 rpm while all bacterial solid cultures 95

were grown at 37°C in an oven incubator. For the selection of E. coli clones harbouring pGK or pCR®-Blunt vectors, kanamycin was included in the culture medium at a final concentration of 50 µg/mL. For the selection of clones harbouring pGEX-6P-1 vector, ampicillin was included in the culture medium at a final concentration of 100 µg/mL. Stocks of bacterial strains were stored in 20% (v/v) glycerol at -80°C. S. clavuligerus NRRL 3585 was grown in a liquid culture medium containing 0.4% (w/v) yeast extract, 1% (w/v) malt extract and 0.4% (v/v) glycerol, and incubated at 28°C on a rotary shaker set at 250 rpm for 3 days.

3.2.2. Transformation of E. coli cells

The preparation of electrocompetent cells with the highest competence generally requires the cells to be in the early to mid-logarithmic growth phase.

Avoiding the late log phase is crucial, as cell competence will be significantly reduced. A single well-isolated E. coli colony was inoculated into 5 mL of sterile LB broth and incubated overnight at 37°C with shaking at 250 rpm. 1 mL aliquot of the overnight culture was then added to a sterile flask containing 100 mL of LB broth.

This is incubated at 37°C with agitation at 250 rpm until the O.D.600 reached 0.5-0.6 absorbance units, which corresponds to the early to mid-logarithmic phase of E. coli

BL21 (DE3) strain.

Following that, the flask containing the bacterial culture was then chilled on ice for 30 min and centrifuged at 6000 rpm for 5 min at 4°C. The supernatant was discarded and then, the cell pellet was resuspended in 50 mL of ice-cold 10% (v/v) glycerol. This washing step was repeated twice with the cell pellet resuspended in 10 mL of ice-cold 10% (v/v) glycerol after the first wash, followed by a final resuspension volume of 400 µL of ice-cold 10% glycerol after the second wash. The 96

resuspended competent cells were then aliquoted into 50 µL aliquots. The competent cells were stored frozen at -80ºC.

Electrocompetent E. coli cells were allowed to thaw on ice and 1-2 µL of plasmid DNA or ligation mixture was added to 50 µL of cells in a chilled 0.2 cm

Gene Pulse cuvette (Bio-Rad). The Bio-Rad Gene Pulser® was adjusted to 2.5kV and a short pulse of current was rendered across the cuvette. Then, 1 mL of LB broth was added to the mixture. This step is crucial to ensure the maximal recovery of transformants. The cells were then allowed to recover at 37ºC for 1 h with shaking at

250 rpm. Selection of transformed colonies was carried out by plating 100 µL of the transformation mixture onto LB plates containing appropriate antibiotics. The plates were subsequently incubated overnight at 37°C and transformed colonies were isolated the following day.

3.2.3. Agarose gel electrophoresis of DNA

Electrophoresis using agarose gels is a simple technique that uses an electric field to separate and resolve DNA from approximately 200 bp to 50 kbp in size depending on the concentration of the agarose gel prepared. For this study, analysis of

DNA samples was done using standard agarose gels (BRL, electrophoresis grade) of

0.8% (w/v) concentration.

For each preparation, a desired amount of agarose was dissolved in TAE buffer (40 mM Tris pH 8.0, 40 mM sodium acetate and 1 mM EDTA pH 8.0) by heating in a microwave oven. When molten agarose was cooled to lukewarm temperature, ethidium bromide was added to a final concentration of 0.1 µg/mL before it was poured into a casting tray and allowed to set for at least 30 min. Samples of DNA were mixed with DNA loading dye (Fermentas International Inc., Canada) 97

before loading into the agarose gel and subjected to a voltage of 80-100 volts for a preferred time. Examination of the separated DNA bands was conducted under UV illumination and an image of the agarose gel was captured using Bio-Rad Molecular

Imager Gel Doc XR system, which is controlled by the Quantity One software (Bio-

Rad).

3.2.4. Generation of mutants

3.2.4.1. Generation of single mutants: site-directed mutagenesis

SDM was performed using the QuikChange™ SDM kit (Stratagene) with modifications as previously reported by Loke and Sim (2001). Sequences of the oligonucleotide primers used for SDM are shown in Table 3.3. Both primers were designed to carry the desired mutation and anneal to the same sequence on opposite strands of the template. The cefE gene from S. clavuligerus NRRL 3585 was previously cloned into pGK expression vector (pscEXP-GST) (Sim and Sim, 2001) and this was used as the mutagenesis template in this study. Pfu DNA polymerase was used in the SDM process to replicate both plasmid strands with high fidelity without displacing the mutagenic primers. The SDM reaction mixture contained 2.5 µL of

10X reaction buffer, 1 µL of each primer (100 pmole/µL), 1 µL of DNA template, 1

µL of dNTPs, 1 µL of Pfu DNA polymerase and deionised water to give a final reaction volume of 25 µL. These reaction mixtures were then subjected to thermal cycling using a Perkin-Elmer GeneAmp® PCR System 2400 in which the parameter included 30 s at 95ºC to activate the DNA polymerase, followed by 28 cycles of temperature changes: 30 s at 95ºC, 60 s at 57.5ºC, 14 min at 72ºC. Thereafter, the product was digested with 10 U of DpnI at 37ºC for 3 h to remove the methylated parental plasmid template. The DpnI-treated mixture was then transformed into 98

competent E. coli BL21 (DE3) cells and selected using appropriate antibiotics. The transformants were then grown in larger quantity and harvested for DNA plasmid extraction using Wizard® Plus SV Minipreps DNA Purification System (Promega).

The site-specific nucleotide modifications in the recombinant plasmids were verified via DNA sequencing described under Chapter 3.2.4.3.

3.2.4.2. Generation of double mutants: restriction enzyme digestion and

DNA ligation

Double mutations consisting of either V275I or C281Y mutation with the C- terminal single mutations were generated by restriction endonuclease digestion, followed by DNA ligation. Recombinant vectors carrying V275I, C281Y, and the C- terminal single mutations (N304A, N304K, N304L, N304M, N304R, I305M, R306L, and R307L) were digested using the PfoI and BamHI restriction endonucleases

(Fermentas International Inc., Canada) according to manufacturer’s instructions. The digested samples were separated by agarose gel electrophoresis. After this, the desired

DNA fragments were excised and purified from the agarose gel using QIAquick® Gel

Extraction kit (Qiagen Singapore Private Limited, Singapore). The purified fragments were then ligated using T4 DNA ligase (Invitrogen) according to manufacturer’s recommendations. The ligated products were then transformed into the electrocompetent E. coli BL21 (DE3) cells following the procedure described in

Chapter 3.2.2.

3.2.4.3. DNA sequencing

DNA sequencing was done using ABI PRISM™ BigDye® Terminator v3.1 cycle sequencing kit (Applied Biosystems), following manufacturer’s 99

recommendations. Briefly, about 250-500 ng of the mutagenised plasmid DNA was mixed with 4 µL of BigDye® Terminator Ready Reaction Mix, 3.2 pmole of the sequencing primer, and deionised water added to a final volume of 10 µL in a 0.2 mL microfuge tube. The tubes were then placed in a Perkin-Elmer GeneAmp® PCR

System 2400 and heated for 1 min at 96°C. Thermal cycling using the following conditions, 96ºC for 10 s, 50°C for 5 s and 60°C for 4 min for 30 cycles, was performed. After the reaction was completed, the sample tubes were placed on hold at

4°C.

To purify the extension products, the entire 10 µL reaction content was transferred to a 1.5 mL microfuge tube containing 1.5 µL of 3 M sodium acetate (pH

4.6), 7.25 µL of deionised water and 31.25 µL of 95% (v/v) ethanol. The mixture was then vortexed and incubated at room temperature for 15 min, after which the mixture was centrifuged at 12000 rpm for 20 min. The ethanol was aspirated and the retained pellet was rinsed with 250 µL of 70% (v/v) ethanol to remove the presence of salts before drying at 50ºC for 10 min. The dried extension products were then sent to sequence-analysis service provided in the Department of Microbiology, National

University of Singapore, using an ABI PRISM™ 3100 Genetic Analyzer (Applied

Biosystems).

3.2.5. Recombinant protein expression and purification

3.2.5.1. Induction of recombinant protein expression

Recombinant clones of wild-type and mutant scDAOCSs were inoculated into

5 mL of LB broth containing the appropriate antibiotic and grown at 37°C overnight with agitation at 250 rpm. Thereafter, 1 mL of the overnight culture was inoculated into 50 mL of LB broth containing the appropriate antibiotic and grown at 37°C with 100

agitation at 250 rpm until the O.D.600 of the culture reached a value of 0.6-0.8 absorbance units. This is followed by the addition of IPTG to a final concentration of

100 µM to induce the expression of wild-type and mutant scDAOCS genes. The cultures were incubated at room temperature with agitation at 250 rpm for 15 h.

3.2.5.2. Affinity purification of recombinant protein

The induced cultures were harvested by centrifugation at 8000 rpm for 5 min at 4ºC and the supernatant was discarded. The cell pellet was rinsed twice with 10 mL

1X PBS buffer and finally resuspended in 2 mL of 1X PBS buffer. The cell suspension was then transferred into a Bijou bottle prior to sonication. Lysis of the cell was performed by sonication with a 1/8” probe while keeping the cells on ice.

The cells were disrupted at an intensity of 10 µm for eight 10 sec pulses, with a 10 sec rest period between pulses. The cell lysates were then centrifuged at 12000 rpm for 20 min at 4ºC to remove the cell debris. The clear supernatant obtained was referred to as soluble protein fraction and transferred into clean microfuge tubes. The soluble protein fraction was then stored at -20ºC.

The wild-type and mutant scDAOCS fusion proteins were purified from soluble protein fraction by affinity chromatography using MicroSpin™ GST purification module (GE Healthcare Private Limited) according to the manufacturer’s instructions. Protein concentrations were measured by the Bio-Rad protein assay and bovine serum albumin (Sigma-Aldrich) was used as the reference protein standard.

The purified proteins and soluble protein fractions collected were then analysed on

SDS-PAGE.

101

3.2.5.3. SDS-PAGE analysis of recombinant protein

The protein fractions were separated and analysed using SDS-PAGE. The procedure of separating proteins on a polyacrylamide gel was based on the method reported by Laemmli (1970). The glass plates, alumina plates and spacers were assembled according to the manufacturer’s instructions (Hoefer® Inc, USA). The composition of the various buffers used for gel preparation is shown in Appendix II.

In an Erlenmeyer flask, the 10% acrylamide solution for the resolving gel without

SDS, TEMED and ammonium persulphate (APS) was prepared and degassed for 15 min. Subsequently, a prescribed amount of SDS was added, followed by TEMED and

APS. Without delay, the solution was mixed thoroughly and poured into the gap between the glass plates. The acrylamide was carefully overlaid with 70% (v/v) ethanol to get a level-loading surface. The gel was left to polymerise at room temperature for at least 1 h.

Thereafter, the ethanol overlay was discarded and any remaining unpolymerised gel was washed away with deionised water. The 4% stacking gel solution was prepared in the same manner as the resolving gel solution. The stacking gel solution was poured onto the polymerised resolving gel and a clean Teflon comb was immediately inserted to create wells for loading protein samples. The gel was then left to polymerise at room temperature for at least 1 h. When the polymerisation of acrylamide was completed, the comb was removed and the wells were rinsed with deionised water to remove any unpolymerised acrylamide. After mounting the gel onto the electrophoresis apparatus (SE 260 Mighty Small II System, Hoefer®), Tris- glycine electrophoresis buffer (25 mM Tris-base, 250 mM glycine, 0.1% (w/v) SDS) was added to fill up the top and bottom reservoirs. Each protein sample was mixed with 1X SDS-PAGE gel loading buffer (50 mM Tris-HCl pH 6.8, 100 mM DTT, 2% 102

(w/v) SDS, 0.1% (w/v) bromophenol blue, 10% (v/v) glycerol) and denatured by boiling for 5 min before loading into the wells. Precision Protein Standards™ (Bio-

Rad) was used as a marker.

The electrophoresis was carried out at a constant current of 20 mA per gel until the bromophenol blue dye front reached the bottom of the resolving gel. The gel was then removed from the electrophoresis setup and stained with Coomassie Blue staining solution (0.25% (w/v) Coomassie Brilliant Blue R-250, 45% (v/v) methanol,

10% (v/v) acetic acid) for 30 min. The gel was destained by soaking in destaining solution (30% methanol, 10% acetic acid) until distinct protein bands appeared. The gel images were captured using Bio-Rad Molecular Imager Gel Doc XR system, which is controlled by the Quantity One software (Bio-Rad). This program was also used to estimate the relative expression levels of the wild-type and mutant scDAOCS recombinant fusion proteins. Hence, to determine the percentage of recombinant fusion protein expressed in the soluble fraction, the protein band containing the recombinant fusion protein was marked out and compared to the entire lane containing the respective soluble protein fraction. Protein concentrations were measured by the Bio-Rad protein assay, following the manufacturer’s recommendation and bovine serum albumin (Sigma) was used as the reference protein standard.

3.2.6. Enzyme activity determination

3.2.6.1. Bioassay of ring-expanded products from DAOCS reaction

The standard reaction mixtures reported by Sim and Sim (2001) were adopted because these conditions are optimal for the ring-expansion of the thiazolidine ring in penicillin nucleus to a dihydrothiazine ring in cephalosporin nucleus. Briefly, the 103

enzyme reaction mixture contained 50 mM Tris (pH 7.4), 1.8 mM iron (II) sulphate,

0.8 mM ATP, 1.28 mM α-ketoglutarate, 4 mg/mL penicillin substrate and 0.8 mg/mL purified GST-scDAOCS. The penicillin substrates used were penicillin G, ampicillin, phenethicillin, and carbenicillin. The individual reagents were added in the order described previously (Sim and Sim, 2001). The reaction mixtures were incubated at

30°C, shook at 200 rpm for about 15 min. The reaction was stopped by adding an equal volume of methanol. The activities of wild-type and mutant scDAOCSs were determined via hole-plate bioassay as previously described (Sim and Sim, 2001). In the hole-plate bioassay, E. coli Ess strain was used as the test organism (a gift from

Professor Arnold Demain, Drew University) and seeded to a final concentration of

1% (v/v) using a culture diluted to an O.D.600 of 0.1 absorbance unit. Difco™ Penase was added to a final concentration of 100,000 IU/mL in the bioassay plates to remove the unconverted penicillin substrates. Cephalosporin C was used as the reference standard. One unit of enzyme activity is defined as the amount of enzyme that forms the equivalent of 1 µg of cephalosporin C per minute under the prescribed conditions.

The specific activity of each enzyme was determined from at least three independent experiments with duplicates.

3.2.6.2. Reverse-phase HPLC analysis

A more sensitive and quantitative non-biological high performance liquid chromatography (HPLC) was used to affirm the results obtained from bioassays. The liquid chromatography system used was AKTA purifier (Amersham Biosciences) consisting of a UV-900 detector and an Autosampler A900, and a 5 µm 4.6 X 150

2 mm Sun-Fire™ C18 column from Waters (USA). The operation and data acquisition of the chromatographic process were controlled by UNICORN 5.0 (Amersham 104

Biosciences). The chromatographic condition described by Chin and Sim (2002) was adopted. Briefly, each injection volume was 20 µL and the mobile phase flow-rate was set at 1 mL/min. Separation was effected at approximately 4°C using 10 mM

KH2PO4 (pH 2.0, adjusted with concentrated phosphoric acid) containing 0.06 mM tetrabutylammonium bromide), mixed with methanol at a ratio of 80:20 (v/v) in the isocratic mode for 5 min followed by a 15 min linear gradient from 100% of the initial solution to 100% methanol. Detection of cephalosporin products was performed at a wavelength of 260 nm. The availability of the cephem moiety of ampicillin

(cephalexin) permitted only the conversion of ampicillin to be precisely determined via the HPLC method. One unit of enzyme activity is defined as the amount of enzyme that forms 1 µg of cephalexin per minute under the prescribed conditions.

The specific activity of each enzyme was determined from at least three independent experiments with duplicates.

3.2.6.3. Enzyme kinetic analysis

The kinetic assays were performed using the spectrophotometric method reported by Dubus et al. (2001), assuming the molar extinction co-efficient, ε, to be

6100 M-1 cm-1 for the conversion of ampicillin to cephalexin. The final reaction mixture contained 1 mM Tris(2-carboxyethyl)phosphine hydrochloride (TCEP), 50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) (pH 7.5), 50 mM

(NH4)2SO4, 100 µM ascorbic acid, 25 µM FeSO4⋅7H2O and 100 µM 2-oxoglutarate.

The final concentration of enzyme used was 2.5 µM. The final concentration of ampicillin substrates used ranged from 0.1 to 10 mM. The kinetic assay was performed using a 96-well format filter-based spectrophotometer from Tecan

(GENios™), with a 260 nm absorbance filter and temperature fixed at 30°C. The 105

kinetic parameters were determined from at least three independent measurements with triplicates and calculated using the Hanes-Woolf plot.

3.2.7. Cloning of cefF encoding DACS from S. clavuligerus

3.2.7.1. Extraction of genomic DNA of S. clavuligerus NRRL 3585

Liquid culture of S. clavuligerus NRRL 3585 was prepared as described in

Chapter 3.2.1 and was harvested by first breaking up the mycelia using a glass homogeniser, followed by centrifugation at 12000 rpm for 5 min to pellet all the cells.

Thereafter, the cell pellet was washed with TE buffer pH 8.0 (10 mM Tris-HCl, 1 mM

EDTA) and resuspended in 40 µL of lysozyme buffer (10 mM Tris-HCl (pH 8.0), 1 mM EDTA, 100 mM NaCl, 5% (v/v) Triton X-100). Then, 10 µL of 10 mg/mL lysozyme solution was added and the sample was incubated at 37°C for 15 min before heat inactivation at 55°C for 15 min. 5 µL of 5 mg/mL RNase solution was then added and incubated at room temperature for another 15 min. Finally, the genomic

DNA of S. clavuligerus was extracted using GFX™ Genomic Blood DNA purification kit (Amersham Pharmacia) following the manufacturer’s recommended protocol.

3.2.7.2. PCR amplification of cefF gene

The gene amplification primers were designed based on the 5’- and 3’-ends of cefF open reading frame encoding deacetylcephalosporin C synthase from S. clavuligerus ATCC 27064 (synonymous to NRRL 3585, GenBank accession number

M37186) reported by Kovacevic and Miller (1991). Restriction sequences of BamHI and EcoRI were added at the 5’-ends of the designed cefF forward primer OL963 and reverse primer OL964, respectively (Table 3.4). This allows cloning of the amplified 106

DNA into the expression vector pGK and expression of scDACS as recombinant GST fusion protein. A PCR reaction mixture containing 1X Pfu DNA polymerase PCR buffer with MgSO4 (Promega), 100 µM dNTP mix, 400 nM forward primer OL963,

400 nM reverse primer OL964, 8% (v/v) DMSO, 0.06 U/µL Pfu DNA polymerase and 1 µL of extracted S. clavuligerus genomic DNA, was prepared and subjected to a touch-down PCR amplification programme starting with an initial denaturation at

95°C for 1 min, followed by eight cycles of 95°C (45 s), 65-61°C (30 s) and 72°C (2 min), 27 cycles of 95°C (45 s), 60°C (30 s) and 72°C (2 min), and a final extension step at 72°C for 5 min. A negative control replacing genomic DNA with sterile deionised water was included. After PCR amplification, agarose gel electrophoresis was used to detect for the amplified product.

3.2.7.3. TOPO® cloning of PCR-amplified cefF gene

The PCR product of approximately 1 kb was excised and extracted from the agarose gel using QIAquick® Gel Extraction kit (Qiagen). The purified PCR product was then ligated into pCR®-Blunt II TOPO® vector using Zero Blunt® TOPO® PCR

Cloning kit (Invitrogen), following manufacturer’s recommendation. The ligated product was then transformed into One Shot® TOP10 Chemically Competent E. coli cells (Invitrogen) according to manufacturer’s instruction. The kanamycin-selected transformants were grown in LB broths for the extraction of recombinant plasmids.

The recombinant plasmids were extracted using Wizard® Plus SV Minipreps DNA

Purification System (Promega) according to the manufacturer’s instruction. The extracted DNA plasmids were then subjected to restriction endonuclease analysis using BamHI/EcoRI (NEB) double digestion and analysed by agarose gel electrophoresis. 107

3.2.7.4. Sub-cloning of cefF gene into expression vector

Three samples that showed a release of approximately 1 kb DNA after

BamHI/EcoRI double digestion were excised and extracted using QIAquick® Gel

Extraction kit (Qiagen). The purified DNA was then ligated into an expression vector pGK using T4 DNA ligase (Invitrogen). A ligation reaction mixture containing 2 µL

5X Invitrogen DNA ligase buffer, 5 µL DNA insert, 3 µL BamHI/EcoRI double- digested pGK and 0.5 µL of 1 U/µL Invitrogen T4 DNA ligase was prepared and incubated at 16°C for 24 h. Thereafter, the ligated product was transformed into the electrocompetent E. coli BL21 (DE3) cells following the procedure described in

Chapter 3.2.2. The kanamycin-selected transformants were screened for the presence of cefF DNA insert using RE analysis described in Chapter 3.2.7.3. Those clones harbouring the recombinant plasmid were fully sequenced to verify the authenticity of cefF gene insert using the sequencing primers listed in Table 3.4.

3.2.7.5. Protein expression and purification

Since scDACS was cloned into the same expression vector as scDAOCS, the expression of recombinant scDACS was determined by following the procedure described in Chapter 3.2.5. PreScission™ Protease (GE Healthcare) was used to cleave the recombinant scDACS from the fusion partner GST, according to manufacturer’s recommendations.

3.2.7.6. Hydroxylation activity of scDACS

Since the reaction mechanism of DACS is very similar to that of DAOCS

(Lloyd et al. 2004; Baker et al. 1991), except for the difference in their substrate 108

preferences, the standard reaction mixture reported by Sim and Sim (2001) for the detection of scDAOCS activity was adopted in this study except that the substrate was replaced by cephalexin. HPLC analysis was done as described in Chapter 3.2.6.2, to detect for the hydroxylated product formed.

3.2.8. Bioactivities of expanded cephalosporin products

3.2.8.1. Determination of the antimicrobial spectra of penicillin substrates and their respective expanded cephalosporin moieties

In order to determine the efficacies of the penicillin substrates and their respective bioconverted cephalosporin products, the amount of cephalosporin products formed has to be determined. This was achieved by using reverse-phase

HPLC method, which separates the penicillin substrate and its respective bioconverted cephalosporin product based on their hydrophobicity differences, and the detection of the cephem ring at OD600.

A similar hole-plate bioassay format, described in Chapter 3.2.6.1., was employed with slight modification. Mueller-Hinton II agar (MHA) (Difco™) was used instead of LB agar, and was prepared according to manufacturer’s recommendation. Difco™ Penase was added to the culture medium to a final concentration of 100,000 IU/mL to remove the unconverted penicillin substrates. The test bacteria, listed in Table 3.1, were streaked onto the MHA plates according to the format illustrated in Figure 3.1. Then, 180 µL of enzyme reaction mixture, prepared as described in Chapter 3.2.6.1., were added to the well in the centre of the agar plate.

Same volume of equimolar concentration of penicillin substrates was prepared and added to another set of bioassay plates without Difco™ Penase added. This is to determine the effect of penicillin on the test bacteria and allow comparison with that 109

of the respective bioconverted cephalosporin product. The bioassay plates were incubated at 37°C for 24 h. This allows the diffusion of the penicillin substrates and the bioconverted cephalosporin products into the agar medium, while the test bacteria grow. The unconverted penicillin substrates in the enzyme reaction mixture will be degraded by Difco™ Penase and the growth of the test bacteria along the respective streak lines will reflect the effect of the cephalosporin products. The zone of growth inhibition was determined by measuring the distance of the streak line from the well to the start-point of bacterial growth and expressed in millimetre. The antimicrobial spectra of the expanded cephalosporin products were determined from at least three independent experiments.

8 2 well

7 3

6 4

Figure 3.1. Inoculation format of test bacteria for hole-plate bioassay. Each arrow indicates the inoculation of a bacterial culture in an outward direction from the edge of the well. (1. M. luteus ATCC 381; 2. E. coli Ess; 3. S. aureus ATCC 9144; 4. E. coli K12; 5. E. aerogenes; 6. S. cholerasuis ATCC 51741; 7. E. faecalis ATCC 29212)

110

3.2.8.2. Determination of the efficacies of penicillin substrates and their respective expanded cephalosporin moieties

In order to compare the efficacies of the expanded cephalosporin products with their respective penicillin substrates more accurately, a colony-forming unit

(CFU) assay method was used (Lin et al., 2008; Yenugu et al., 2003; Porter et al.,

1997). This method employed a standardised number of CFUs used for testing the inhibitory effect of each antibiotic in a smaller scale compared to hole-plate bioassay.

The test organisms used were E. coli BL21 and E. coli BL21-AmpR, which has been transformed with a plasmid, pGEX-6P-1 (GE Healthcare Life Sciences). pGEX-6P-1 plasmid carries an ampicillin-resistance gene, which is constitutively expressed in E. coli. Approximately 100 CFU of test organism suspended in 2X Mueller-Hinton (II) broth were mixed with equal volume of antibiotic solution and incubated at 37°C. For experiments with E. coli BL21, the incubation time was one hour, whereas for those with E. coli BL21-AmpR, it was three hours. At the end of incubation, the samples were plated using pour-plate method and Luria-Bertani molten agar. The solidified agar plates were then incubated at 37°C for 24 hours before the number of CFUs was counted. For those samples that were incubated for three hours, 2-10x dilutions were done after incubation. Then, 100 µL aliquots of neat and diluted samples were plated to ensure countable number of colonies in the agar plates. For diluted samples, only plates with 30-300 colonies were counted and used for the calculation of the original viable bacterial count.

111

3.2.8.3. Assessing additional properties conferred on expanded cephalosporin products: elastase inhibition assay

Calbiochem® InnoZyme™ Human Neutrophil Elastase Immunocapture

Activity Assay kit (Merck Pte Ltd) was used to assess the potential protease inhibitory effect of cephalosporin on human neutrophil elastase (HNE). The inhibition assay was set up based on manufacturer’s instruction, with slight modification. A Costar® 96- well half-volume black microplate (Corning Incorporated Life Sciences, USA) was used. The reaction mixture contained a final concentration of 2 ng HNE and 1X methoxy-succinyl-Ala-Ala-Pro-Val-7-amino-4-methylcoumarin (MeOSuc-AAPV-

AMC) fluorogenic substrate in the assay buffer provided in the kit. The test sample was prepared as described in Chapter 3.2.6.1., except that methanol was not added to stop the reaction. Instead, the reaction mixture was filtered through a Microcon® centrifugal filter (Millpore™) with a molecular weight cut-off of 30 kDa to remove any protein, which may act as a substrate for HNE. Then 10 µL of filtrate were added to each reaction mixture and the elastase activity was detected over an hour.

Phenylmethanesulfonyl fluoride (PMSF) is a serine protease inhibitor and has been reported to inhibit HNE (Kim and Kang, 2000). A final concentration of 5mM PMSF was used as a positive control to inhibit the proteolytic activity of HNE. The proteolytic activity of HNE is expressed in nanomolar per minute.

3.2.9. Biocomputational analysis

Comparative analyses of gene and protein sequences as well as protein structures were carried out using biocomputational programs. The nucleotide sequences of known genes and the amino acid sequences of known proteins were retrieved from GenBank Database (http://www.ncbi.nlm.nih.gov/genbank/). The 112

amino acid sequences of annotated DAOCSs and DACSs from different organisms were retrieved from the GenBank database, and the accession numbers were as follows: S. clavuligerus DAOCS, P18548; A. chrysogenum DAOC/DACS, P11935; S. chartreusis DAOCS, AY318743; S. ambofaciens DAOCS, AY318742; S. jumonjinensis DAOCS, AF317908; N. lactamdurans DAOCS, Q03047; L. lactamgenus DAOCS, CAA39984; S. clavuligerus DACS, P42220; Streptomyces sp. isolate 65PH1 DACS, AY318744; N. Lactamdurans DACS, P42219; and L. lactamgenus DACS, CAA39985. The nucleotide and amino acid sequences obtained were aligned using CLUSTAL W Multiple Sequence Alignment Program (version

1.7) (Thompson et al., 1994) based on segment of similarity between sequences to compare its conservedness. MatGAT programme was used to generate the similarity/identity matrices of annotated DAOCSs and DACSs (Campanella et al.,

2003). Self-optimised prediction program (SOPMA) (http://au.expasy.org/tools/)

(Geourjon and Deléage, 1995) was employed to predict secondary structures of proteins after mutations. This program affirmed an accurate prediction of 69% of amino acids for a three-state description (α-helix, β-sheet and coil) of the secondary structure.

Viewing and manipulation of three-dimensional structures were using the

SwissPdb Viewer program version 3.7 (Guex et al., 1999). The protein structures can be presented as wire frame, space-filled or ribbon models. The program also provided different features, which facilitated the analyses of protein structure such as superimposition of proteins and virtual mutation. SwissPdb Viewer and the user manual are available at the ExPASy Molecular Biology server (http://spdbv.vital- it.ch/). Since the tertiary structure of DACS has yet to be resolved, the tertiary structure of scDACS was modelled using Swiss-Model, which is an automated 113

knowledge-based protein modelling server (http://swissmodel.expasy.org/) (Kiefer et al., 2009).

Amino acid residues interacting with the penicillin substrate were determined using the LIGPLOT program (Wallace et al., 1995). This program automatically generates a schematic diagram of protein-ligand interactions mediated by hydrogen bonds and hydrophobic contacts. Mutational models were generated using SwissPdb

Viewer and were further assessed using LIGPLOT to validate the changes in the protein-ligand interactions.

3.2.10. Reproducibility of results

All the experiments carried out in this project were repeated at least thrice to ensure the reproducibility of the results, unless stated otherwise. The cloned and mutated cefE and cefF gene sequences were fully sequenced on both strands to ensure authenticities of the sequences. All enzyme assays and calculation of enzyme activities were determined from three independent experiments, unless stated otherwise.

114

Chapter 4: Improving the ring-expansion activity of scDAOCS

DAOCS is an industrially important enzyme for the production of cephalosporin precursor, 7-ADCA. As the chemical ring-expansion of penicillin G to yield 7-ADCA is expensive and polluting, enzymatic ring-expansion method using

DAOCS is deemed as an environmentally safe alternative (Velasco et al., 2000).

Although deacylation of cephalosporin C is also employed to yield the other cephalosporin precursor 7-ACA, the cost of production is considered high due to the instability of cephalosporin C in the fermentation culture that results in lower yield

(Elander, 2003).

The limitation of DAOCS for industrial application is its substrate specificity.

Although the substrate specificity of DAOCS is considered wide, its ring-expansion activity for penicillin G and other analogues is too low for industrial application.

Hence, DAOCS needs to be structurally modified to improve its catalytic activity for penicillin G. Many earlier studies were done to characterise the enzyme and certain mutations were found to enhance its ring-expansion activity for hydrophobic penicillins, particularly at the C-terminus of the enzyme (Wu et al., 2005; Chin et al.,

2004; Chin and Sim, 2002; Lee et al., 2002; Chin et al., 2001).

The C-terminus of DAOCS has often been implicated to play a crucial role in selecting, binding and orienting the penicillin substrate during catalysis (Chin et al.,

2004, 2001; Valegard et al., 2004, 1998; Chin and Sim, 2002; Wei et al., 2003; Lee et al., 2002; Lipscomb et al., 2002; Lloyd et al., 1999). However, earlier studies of the

C-terminus of scDAOCS were performed without any structural information of 115

scDAOCS-penicillin complex (Chin et al., 2004, 2001; Chin and Sim, 2002; Wei et al., 2003; Lee et al., 2002). Thereafter, the crystal structure of scDAOCS complexed with penicillin G and ampicillin were resolved, and our understandings of the structural mechanisms that influence the properties of this enzyme were increased

(Valegard et al., 2004). The structural data have also facilitated investigation and interpretation of the effects of enzyme modifications with precise spatial consideration. However, the mechanisms of the improvements achieved from the mutations of the C-terminal residues of scDAOCS, i.e., N304 and R306, were still not clear.

Approaches such as homeologous recombination, chemical mutagenesis, error-prone PCR and gene shuffling have been used to improve the catalytic properties of scDAOCS for penicillin G conversion (Adrio et al., 2002; Gao et al.,

2003; Hsu et al., 2004; Wei et al., 2003). However, it is difficult to determine the exact mechanism that contributes to such enhancement using these methods. Hence, further studies were done in this project to further understand the principles that govern the structure-function relationship in scDAOCS to yield a more superior enzyme for industrial application and even produce new cephalosporins. Two main regions were selected. The first region is the C-terminus of scDAOCS and the second region is the small sub-domain where mutations at residues V275 and C281 have resulted in higher ring-expansion activity (Wei et al., 2003). Figure 4.1 shows the spatial location of amino acid residues selected for this study. A rational approach using SDM was adopted to improve the substrate recognition of scDAOCS for penicillin G and other penicillins such as ampicillin, carbenicillin and phenethicillin, which differ slightly in their side chains.

116

Figure 4.1. Spatial location of amino acid residues V275, C281, N304, I305, R306 and R307 in scDAOCS (1UNB). The iron-binding motif H183XD185XnH243 is coloured green while the co-substrate-binding motif R258XS260 is coloured light green. The amino acid residues studied is coloured orange. The substrate ampicillin and co-substrate α-ketoglutarate (α-KG) are coloured brown and purple, respectively.

4.1. Identification of potential sites for the improvement of scDAOCS in hydrophobic penicillin conversions

Development of X-ray crystallography and NMR has allowed resolution of the molecular structure of a protein, which is invisible to the naked eye. Biocomputational tools have provided a means to rapidly collate the various structural levels of a protein for comparative analysis and prediction. Combining these two technologies, protein

117

molecular structures could be virtually conceptualised, examined and categorised based on their structural similarities. This has also permitted a more rational approach to improve the functionalities of a protein and to further understand the principles that govern the structure-function relationship of proteins.

Survey on mutational studies of enzymes with known tertiary structures revealed that mutations closer to the active site or substrate binding site of an enzyme often improved the enantioselectivity, substrate selectivity and alternate catalytic activity (Morley and Kazlauskas, 2005). Hence, to alter or improve the substrate preference of scDAOCS, it will be more productive to focus on its substrate binding pocket. Since the difference between penicillin N and hydrophobic penicillins, e.g. penicillin G, lies in their side-chains, it is only logical to focus on the region of the substrate binding pocket of scDAOCS that interacts with the side-chain of the penicillin substrates (Figure 4.2). The amino acid residues in the substrate binding pocket of scDAOCS complexed with penicillin G (1UOB) and ampicillin (1UNB) were examined using SwissPdb Viewer and only residues T72, M73, S102, L158,

R160, R162, F264, N304 and I305 surround the side chain of the penicillin substrates

(Figure 4.3). Then LIGPLOT (Wallace et al., 1995) (Figure 4.4) and Ligand-Protein

Contacts (Sobolev et al., 1999; http://bip.weizmann.ac.il/oca-bin/lpccsu) (Table 4.1) programs were used to predict the amino acid residues that interact with the penicillin substrates and their modes of interaction. Since the prediction criteria differ slightly between each program, the results from each analysis were compared and the common amino acid residues predicted by these programs in the two structures were identified (Table 4.2). Only residues M73, L158, R162 and F264 were consistently identified by the three programs to interact with the penicillin substrates.

118

Interestingly, many earlier studies have identified the C-terminal amino acid residues to be potential sites for enzyme improvement (Wei et al., 2003; Chin and

Sim, 2002; Chin et al., 2001; Lee et al., 2002, 2001), which were not consistently reflected in the in silico predictions. For example, single mutations of C-terminal residues N304, R306 and R307 to a leucine were demonstrated to improve the conversion activity of scDAOCS for various hydrophobic penicillin substrates (Chin and Sim, 2002). In order to gain a better understanding of the mechanisms behind these improvements, residue N304 was selected for further analysis since it lies closely to the side chain of the penicillin substrate. Using a rational approach, a complete library of amino acid replacements was done on this position and mutations to alanine, lysine, methionine and arginine were found to also improve the conversion activity of scDAOCS to different extents (Chin et al., 2004). However, only the enzyme specific activities of these mutants were determined and reported using a bioassay, which is a non-continuous product detection method. Hence, an enzyme kinetic assay developed by Dubus and co-workers (2001) was employed in this study to determine the kinetic properties of these improved mutant enzymes, i.e., N304A,

N304K, N304L, N304M and N304R. Upon understanding the basis of the improvement in N304 modification, this approach was also adopted to study the role of residue R306 in influencing the catalytic activity of scDAOCS.

119

(a) (b)

H H H2N N S N S CH3 CH3

COOH O N CH3 O N CH3 O O COOH COOH

Figure 4.2. Schematic diagrams of scDAOCS substrate binding pocket bound with (a) penicillin N and (b) penicillin G. Both blue and red arcs represent the substrate binding pocket of scDAOCS. The blue arc represents region of the substrate binding pocket to be preserved and the red arc represents region to be modified.

(a) (b)

Figure 4.3. Close-up view of substrate binding residues in scDAOCS crystal structures complexed with (a) penicillin G (1UOB) and (b) ampicillin (1UNB). Only the residues surrounding the side chains of both substrates are presented. Residues T72, M73, S102, L158, R160, R162, F264, N304 and I305 are located in the vicinity of the side-chain of both substrates. The backbones and side chains of amino acid residues are in white and orange colours, respectively. Both ampicillin and penicillin G are presented in CPK colours. Dotted green lines represent hydrogen bonds.

120

(a) (b)

Figure 4.4. LIGPLOT analyses of scDAOCS crystal structures complexed with (a) penicillin G (1UOB) and (b) ampicillin (1UNB).

121

Table 4.1. Ligand-Protein Contacts analyses of scDAOCS complexed with (a) penicillin G (1UOB) and (b) ampicillin (1UNB).

(a) Specific contacts Residue Distance Surface HB Arom Phob DC M73 2.7 31.9 - - + - S102 3.3 3.8 - - - - L158 3.6 14.8 - - + - R160 0.4 489.5 - - + + R162 1.8 50.2 + - - + F164 4.3 3.5 - - - - M180 3.0 45.9 - - + + T190 4.2 0.2 - - + - I192 3.0 34.5 - - + + L204 4.4 2.5 - - + - F225 4.2 0.7 - - + - V245 4.3 1.9 - - + + V262 3.6 28.3 - - + + F264 3.3 36.4 - + + - N304 4.2 8.5 - - + + I305 4.0 15.7 - - + +

(b) Specific contacts Residue Distance Surface HB Arom Phob DC T72 3.6 7.9 - - - - M73 2.5 47.8 - - + - R74 4.5 1.1 - - - - L158 2.6 34.5 - - + - R160 0.7 368.0 + - + + R162 1.8 69.6 + - - + F164 6.2 0.4 - + - M180 3.1 28.4 - - + + T190 3.8 1.1 - - + - I192 2.9 47.2 - - + + L204 4.5 0.7 - - + - F225 4.1 0.7 - - + - V245 3.5 4.5 - - + + V262 3.3 37.7 - - + + F264 3.1 31.9 - + - - N304 3.4 19.9 + - - + I305 4.2 12.7 - - - + Legend: Distance- nearest distance (Å) between atoms of the ligand and the residue Surface- contact surface area (Å2) between the ligand and the residue HB - hydrophilic-hydrophilic contact (hydrogen bond) Arom - aromatic-aromatic contact Phob - hydrophobic-hydrophobic contact DC - hydrophobic-hydrophilic contact (destabilising contact) +/- - indicates presence/absence of a specific contact

122

Table 4.2. Comparison between the predicted substrate binding residues in scDAOCS crystal structures complexed with penicillin G (1UOB) and ampicillin (1UNB). The amino acid residues that are identified by all three programmes are highlighted in grey. The amino acid residues located in the vicinity of the side-chain of both penicillin substrates illustrated in Figure 4.3 are underlined.

SwissPdb Viewer LPC analyses LIGPLOT analyses (within 5Å) Residue 1UOB 1UNB 1UOB 1UNB 1UOB 1UNB No. T72 - ü - - ü ü M73 ü ü ü ü ü ü R74 - ü - - ü ü S102 ü - - - ü ü L158 ü ü ü ü ü ü R160 ü ü - - ü ü R162 ü ü ü ü ü ü F164 ü ü - - ü - M180 ü ü ü ü ü ü T190 ü ü - ü ü ü I192 ü ü ü ü ü ü L204 ü ü - - ü ü F225 ü ü - - ü ü V245 ü ü - - ü ü V262 ü ü ü ü ü ü F264 ü ü ü ü ü ü N304 ü ü - - ü ü I305 ü ü - - ü ü

123

4.2. In silico structural analyses of scDAOCS mutants

When the primary sequence of a protein is changed, the secondary and tertiary structures of the protein may be affected. Consequently, the function of the protein will be altered. With the lack of equipment to determine the actual secondary and tertiary structures of a protein after mutations, prediction of these structures will be useful in understanding the functional changes observed in mutant scDAOCSs.

Analysis of the mutated protein sequences using the improved SOPMA program (Combet et al., 2000; Geourjon and Geleage, 1995) suggested no significant alteration in their secondary structures (Appendix III). Alterations in the secondary structures of the mutant enzymes with improved catalytic activity were also observed in those with reduced activity. This suggests that the alterations in the catalytic properties of the mutant scDAOCSs are not just a function of changes in the secondary structures. The spatial orientation of an amino acid residue and its stereo- chemical effect on its neighbouring residues and substrates may collectively alter the function of the protein.

Three-dimensional modelling of the mutant enzymes using Swiss-Model program (Arnold et al., 2006; Schwede et al., 2003; Guex and Peitsch, 1997; http://swissmodel.expasy.org/) also did not reveal any significant changes in their tertiary structures (data not shown). The in silico mutational results obtained using

SwissPdb Viewer will be illustrated together with the discussion of the assay results in the respective chapters, in order to facilitate the understanding of the effects of the mutations on the enzymatic properties of scDAOCS.

124

4.3. Generation and verification of mutant scDAOCS sequences

The cefE gene from S. clavuligerus NRRL 3585 was previously cloned into pGK expression vector (pscEXP-GST) (Sim and Sim, 2001). SDM was performed using the QuikChangeTM SDM kit (Stratagene) with modifications as previously reported (Loke and Sim, 2001). The recombinant expression vector pscEXP-GST was used as the mutagenesis template. The mutagenic oligonucleotide primers were synthesised by 1st Base (Singapore) or Sigma-Proligo (Singapore) (Table 3.3). The mutagenised plasmids were transformed into Escherichia coli BL21 (DE3) strain. The site-specific nucleotide modifications were verified using the ABI PRISM™ dye terminator cycle sequencing kit and ABI PRISM™ 3100 Genetic Analyzer as described in Chapter 3.2.4.2. The mutated genes were fully sequenced to ascertain that no other mutation has been incorporated during the PCR amplification process.

4.4. Heterologous expression and purification of wild-type and mutant scDAOCSs

After the mutation sites were confirmed via DNA sequencing, recombinant E. coli BL21 (DE3) clones harbouring the wild-type and mutant scDAOCS constructs were induced for protein expression. High-level soluble expression of wild-type and mutant GST-scDAOCS fusion proteins was obtained as revealed by intense protein bands of approximately 60 kDa in SDS-PAGE analyses (Figure 4.5). The cell lysates containing GST-scDAOCS fusion proteins were then subjected to affinity chromatography purification using the MicroSpin™ GST purification module. The molecular masses of the purified fusion proteins were approximately 60 kDa, and the proteins were at least 90% pure as estimated by the Quantity One program. These results showed that the mutations did not significantly affect the soluble expression of 125

the enzyme, which allowed subsequent determination of the properties of the mutant enzymes.

kDa M 1 2 3 4 5 6 7 8 116 92

GST-scDAOCS 50

Figure 4.5. SDS-PAGE analysis of cell-free extracts and purified fractions of E. coli BL21 (DE3) clone expressing recombinant wild-type and mutant GST-scDAOCSs. Lane M: Precision Protein Standards™ Bio-Rad 161-0318; Lane 1: cell-free extract of E. coli clone expressing recombinant wild-type scDAOCS; Lane 2: cell-free extract of E. coli clone expressing recombinant N304R mutant scDAOCS; Lane 3: cell-free extract of E. coli clone expressing recombinant R306L mutant scDAOCS; Lane 4: cell-free extract of E. coli clone expressing recombinant N304RR306L mutant scDAOCS; Lane 5: purified recombinant wild-type scDAOCS; Lane 6: purified recombinant N304R mutant scDAOCS; Lane 7: purified recombinant R306L mutant scDAOCS; Lane 8: purified recombinant N304RR306L mutant scDAOCS. Similar profiles were also obtained for other scDAOCS mutants generated in this project.

126

Chapter 5: A complete library of amino acid alterations at R306 in scDAOCS demonstrates its structural role in the ring-expansion activity

5.1. A complete library of amino acid alterations R306 of scDAOCS

The conversions of hydrophobic penicillin substrates namely ampicillin, penicillin G, phenethicillin, and carbenicillin by the wild-type and mutant scDAOCSs were tested via hole-plate bioassay with E. coli Ess strain as the test microorganism.

All experimental data were confirmed by at least three independent measurements.

The specific activity of wild-type scDAOCS for penicillin G, ampicillin, phenethicillin, and carbenicillin conversions were 9.51 ± 0.47 U/mg, 3.03 ± 0.17

U/mg, 7.33 ± 0.29 U/mg, and 1.10 ± 0.20 U/mg, respectively. The specific activities of the mutant enzymes were compared with that of the wild-type enzyme and expressed as relative activities (Table 5.1).

As demonstrated in the previous study (Chin and Sim, 2002), the replacement of R306 to leucine enhanced its ring-expansion activity. It showed similar improvements in the ampicillin, penicillin G, phenethicillin, and carbenicillin ring- expansion activities at the levels of 186, 105, 173, and 283% that of the wild-type scDAOCS, respectively (Table 5.1). Upon mutation to a proline, the conversion activities of the mutant enzyme for all four substrates were not detectable using the current assay condition. Other mutants possessed varying levels of activities when compared to the wild-type enzyme. Generally, replacing arginine with other charged residues reduced the conversion activity of the enzyme drastically except for lysine,

127

which retained about half the activity of the wild-type enzyme. The replacement with polar residues reduced the conversion activity of scDAOCS to 16–62% that of the wild-type enzyme for ampicillin conversion, whereas the substitution to an amino acid with aromatic side chain retained the activity at the levels of 55–90%. The trend of the relative activities of the mutant enzymes observed in ampicillin conversions was similarly observed in penicillin G, phenethicillin, and carbenicillin conversions. It is noteworthy that two additional mutations, R306I and R306M, managed to improve the activity of scDAOCS at the levels of 102–283% for all four substrates.

To validate the ring-expansion activity of the improved mutants, i.e., R306L,

R306I, and R306M, the respective reaction mixtures were subjected to HPLC analyses. The substrate (ampicillin) and product (cephalexin) could be discriminated at resolved retention times of approximately 19 and 18 min, respectively (Figure 5.1).

Since the retention time of the converted product coincided with that of the cephalexin standard, this authenticated the identity of the product in the enzyme reactions. Furthermore, the chromatographic results demonstrated that only cephalexin was produced by the mutant scDAOCSs at higher concentrations than that of the wild-type enzyme. The specific activities of wild-type, R306I, R306L, and

R306M mutant scDAOCSs for ampicillin conversion were 6.41, 9.11, 10.90, and 7.23

U/mg, respectively (Table 5.2). The R306L, R306I, and R306M mutant scDAOCSs showed improvement in the conversion of ampicillin at the levels of 170, 142, and

113% of the wild-type enzyme, respectively. These results were comparable to those obtained via bioassays (Table 5.1).

The enzyme kinetic assay reported by Dubus and co-workers (2001) was similarly adopted to determine the changes in the kinetic parameters of scDAOCS after mutations. The Km values of R306I (2.64 mM), R306L (1.09 mM) and R306M 128

(5.68 mM) mutant scDAOCSs for ampicillin conversion were significantly reduced compared to that of the wild-type enzyme (8.33 mM) (Table 5.3), indicating a significant improvement in the substrate binding affinity of the enzyme.

Consequently, the kcat/Km values of R306I, R306L, and R306M mutant scDAOCSs for ampicillin conversion were approximately 2.6-, 5.1-, and 1.2-folds higher than that of the wild-type enzyme. Among these three mutants, R306L mutant scDAOCS has the highest catalytic efficiency.

129

Table 5.1. Relative activities of the wild-type and mutant scDAOCSs determined by hole-plate bioassay. H H N S N S R CH3 DAOCS R

N CH3 N α-KG, Succinate, CH3 O O CO , H O O COOH 2 2 2 COOH 6-APA 7-ADCA

a Relative activity (%) NH2 COOH

R-group of substrate O O O Mutations Amino acid side chain Ampicillin Penicillin G Phenethicillin Carbenicillin NH Wild-type 100 100 100 100 CH2 CH2 CH2 NH C NH2

R306K CH2 CH2 CH2 CH2 NH2 50 53 45 32

b R306G H 21 27 19 ND

R306A CH3 46 37 30 27

R306S CH2 OH 39 37 27 30

R306C CH2 SH 18 13 11 ND

OH R306T 40 52 32 34 CH CH3 CH R306V 3 34 28 36 32 CH CH3 O R306D 8 4 4 ND CH2 C OH O R306N 16 14 12 16 CH2 C NH2 O R306E 8 7 6 ND CH2 CH2 C OH O R306Q 62 67 54 62 CH2 CH2 C NH2

R306M CH2 CH2 S CH3 113 102 115 122

CH R306L 3 186 105 173 283 CH2 CH CH3

CH R306I 3 168 113 149 167 CH CH2 CH3 H N R306H CH2 16 10 8 ND N

R306F CH2 55 41 44 39

R306Y CH2 OH 70 59 61 60

CH2 NH R306W 90 61 82 83

R306P H C ND ND ND ND C OOH aThe relative activity of each mutant enzyme is expressed as the percentage of the specific activity of the mutant enzyme relative to that of the wild-type enzyme, which was set at 100%. bND: not detected; 6-APA: 6-aminopenicillanic acid; 7-ADCA: 7-aminodeacetoxycephalosporanic acid

130

20uLampconversion2pH2001:10_UV1_260nm1 20uLampconversion2pH2002:10_UV1_260nm 20uLampconversion2pH2001:10_UV1_260nm@01,BASEM1 20uLampconversion2pH2002:10_UV1_260nm@02,BASEM

mAU

18.19

NH2 H N S NH2 60 H O N N S O CH3 CH3 COOH O N CH3 40 O COOH

20 19.01 18.99 21.52 21.53 0 19.73 22.01 19.70 21.08

17.69 17.81 -20

15.0 16.0 17.0 18.0 19.0 20.0 21.0 min Figure 5.1. HPLC chromatogram of wild-type scDAOCS reaction (red) vs time 0 (blue) for ampicillin conversion. Appearance of new peak (arrow) in the chromatogram indicates the cephalosporin product. The retention time of the product from ampicillin conversion corresponds to that of cephalexin standard (Rt = 18.20 ± 0.10). The products of various mutant scDAOCSs showed the same retention times as that of wild-type enzyme suggesting no further modification to the products.

131

Table 5.2. Specific activities of the wild-type and mutant scDAOCSs (R306I, R306L and R306M) for ampicillin conversion determined by HPLC analyses.

NH2 NH2 H H N S N S CH3 DAOCS

O N CH3 O N O α-KG, Succinate, O CH3 C O2 CO2, H2O OOH COOH Ampicillin Cephalexin

Enzyme type Specific activitya (U/mg) Relative activityb (%) Wild-type 6.41 ± 0.60 100

R306I 9.11 ± 1.65 142

R306L 10.90 ± 1.29 170

R306M 7.23 ± 0.49 113 aOne unit of activity is the amount of DAOCS required to form the equivalent of 1µg of cephalexin per minute. The specific activities are shown as the mean ± standard deviation of at least three independent experiments with duplicates. bThe relative activity of each mutant enzyme is expressed as the percentage of the specific activity of the mutant enzyme relative to that of the wild-type enzyme at 100%.

Table 5.3. Kinetic parameters of the wild-type and mutant scDAOCSs (R306I, R306L and R306M) for ampicillin conversion.

a -1 a -1 -1 Enzyme Km (mM) kcat (s ) kcat/Km (M s ) Relative kcat/Km Wild-type 8.33 ± 0.92 0.056 ± 0.002 6.76 1.0

R306I 2.64 ± 0.64 0.046 ± 0.003 17.50 2.6

R306L 1.09 ± 0.11 0.037 ± 0.003 34.22 5.1

R306M 5.68 ± 0.58 0.047 ± 0.001 8.27 1.2 aThe values are mean ± standard deviations of at least three independent experiments with triplicates.

132

5.2. The role of amino acid residue R306 in the activity of scDAOCS

An earlier study on the modification of the C-terminus of scDAOCS showed that R306L mutation improved the ring-expansion activity of scDAOCS in penicillin

G, ampicillin, carbenicillin and phenethicillin conversions (Chin and Sim, 2002).

However, due to the lack of a crystal structure complexed with a substrate during the earlier study, the position of 306 relative to the substrate binding site was not clear.

The recently elucidated crystal structure of scDAOCS complexed with penicillin G and ampicillin revealed that R306 residue is located away from the catalytic site

(Figure 5.2). It is surrounded by five hydrophobic residues Y184, L186, W297, I298 and V303, and is partially exposed to the aqueous environment. The side chain of residue R306, being surrounded by these five hydrophobic residues, may help to preserve the interaction between the β7 (Y184 and L186, which forms part of the co- factor binding motif), α10 (W297 and I298), and β15 (V303) strands (Figure 5.2).

This interaction is necessary to stabilise the integrity of the H183XD185XnH243 motif to ensure proper binding and activation of the iron (II) ion. Furthermore, these six positions are well-conserved in all annotated DAOCSs (Figure 5.3), indicating the importance of this hydrophobic cleft. Therefore, any mutations at position 306 of scDAOCS would exert an indirect effect on the catalytic property of the enzyme.

Generally, there are two factors that influence the dynamics of residue R306 in its contribution to the function of scDAOCS. Firstly, amino acid residues adjacent to residue R306 (residues I305 and R307) will affect the spatial orientation of residue

R306. Secondly, interaction of the side chain of residue R306 with the external environment and its surrounding residues will indirectly affect the polypeptide backbone of the C-terminus. Since the earlier study on the C-terminus of scDAOCS has revealed that the replacement of arginine with hydrophobic leucine at this site 133

improved its specific activity (Chin and Sim, 2002), this improvement was likely due to the side chain of leucine interacting with its surrounding residues and indirectly, affecting the C-terminus in interacting with the penicillin substrate. Hence, it could be postulated that the replacement of R306 to amino acid residues with aliphatic side chains similar to leucine will yield a comparable effect, for example R306I and

R306M mutations. Conversely, substitutions to polar and charged residues will have adverse effects on the activity of scDAOCS.

The structural role of R306 in the catalytic property of scDAOCS was thus deduced from observing the effect of replacing it with the other 19 proteinogenic amino acids. From the hole-plate bioassay results, the trend observed was that both the length and hydrophobicity of the side chain of the amino acid residue are important in influencing the function of the enzyme. Firstly, residue R306 was replaced with proline, which confers rigidity to the protein structure as its side chain is joined to its α-amino group. This rigidity tends to confer a kink into the protein structure and often disrupts the function, and sometimes solubility, of the protein when mutated at critical positions (Reiersen and Rees, 2001). Since the activity of

R306P mutant was lost and its solubility was unaffected, this implied that the conformation of the C-terminal end of scDAOCS is crucial for its catalytic property.

Secondly, replacement of R306 with glycine resulted in 80% reduction in the specific activities of the enzyme for all four substrates. This suggests the need for a longer side chain at this position to control the dynamics of the C-terminal tail or to shield the hydrophobic cleft from the aqueous environment. Evidently, a slight augmentation in the enzyme activity was gained when R306 was replaced with alanine.

Thirdly, replacement with charged residues resulted in tremendous decrease in the conversion activity of scDAOCS, retaining less than 20% of the wild-type activity 134

for all substrates. This supported the postulation that charged residues are detrimental to this site, probably by perturbing the integrity of the hydrophobic cleft. However, this was not observed for both arginine and lysine residues. It has been reported that the pKa of the side chain of arginine or lysine would be lowered if they were to be buried in a hydrophobic environment (Guillen Schlippe and Hedstrom, 2005). This would render the amino group in the side chains of both residues to be unprotonated and uncharged. Consequently, the integrity at this region might be preserved.

However, the relative activity of R306K mutant was approximately half of that of the wild-type enzyme. This suggests that the structural behaviour of lysine in a protein is different from that of arginine. It could be possible that the lysine side chain, being aliphatic, is rotated away from the hydrophobic region, hence exposing the hydrophobic region to the aqueous environment. In contrast, the arginine side chain, which has a planar conformation, will be able to shield the hydrophobic region.

Fourthly, when the carboxyl group of the acidic residues (glutamate and aspartate) were substituted to an amide (to become glutamine and asparagine respectively), there were small increments in the conversion activities, further supporting the postulated detrimental effect of charged residues. Interestingly, by comparing the length of structurally similar residues of various R306 mutants, it was observed that the activity of the enzyme improves in tandem with an increase in carbon atoms in the side chain of the substituted amino acid. For example, in the asparagine and glutamine replacements, where both residues are similar in their structures except that glutamine is longer in its side chain by an alkyl carbon, the activity of R306Q mutant enzyme was much higher than that of the R306N mutant enzyme. A similar trend was also observed when R306S mutant was compared with

R306T mutant, or R306V mutant with R306I and R306L mutants. 135

Lastly, when the polar side chain was replaced with a non-polar side chain, for example methionine, leucine, or isoleucine, the conversion activities of the mutant enzymes were higher than that of the wild-type enzyme. These results supported the hypothesis that replacement of R306 with amino acids structurally similar to leucine was able to improve the enzyme activity via hydrophobic interaction with the surrounding residues. As reflected by their Km values for ampicillin conversion, the improvements obtained were due to the enhancement in the substrate binding affinity of the enzyme upon these mutations. Hsu and co-workers (2004) have also reported a similar observation where A246V mutation in the shuffled DAOCS FF8 clone might have resulted in better hydrophobic packing with the surrounding residues, namely

L178, F201, V202, L204, H244, and A247. Since both H244 and A247 residues are close to the iron ligand H243, the improved hydrophobic packing might exert an effect on the function of the enzyme similar to R306L mutation, leading to enhanced conversion activity. Although non-polar amino acids are expected to interact well with the hydrophobic region formed by residues Y184, L186, W297, I298 and V303, replacements of R306 with aromatic amino acid residues, i.e., phenylalanine, tyrosine, and tryptophan, have showed otherwise; possibly due to the bulkiness of the side chains or the torsion angles of the side chains towards the external environment instead of the hydrophobic cleft.

Together, these results corroborate the mechanism and importance of the length and hydrophobicity of the amino acid side chain at position 306 in maintaining the integrity of the hydrophobic cleft, which in turn, is capable of affecting the ferrous iron-binding motif. These results also support the notion of engineering the position

306 of scDAOCS to accept hydrophobic penicillin substrates.

136

Figure 5.2. Close-up view of residue R306 (purple) surrounded by five hydrophobic residues Y184, L186, W297, I298, and V303 (green). This arrangement strengthens the β7 (where residues H183 and D185 of the iron-binding motif reside), α10 and β15 (C-terminal) conformations. (Amp: ampicillin; α-helices: red; β-sheets: blue).

Figure 5.3. Part of the multiple alignments of annotated DAOCS primary sequences. The five hydrophobic residues Y184, L186, W297, I298 and V303 (highlighted in green, scDAOCS numbering), surrounding residue R306 (highlighted in purple) in scDAOCS, are highly conserved. Scl, S. clavuligerus; Sch, S. chartreusis; Sam, S. ambofaciens; Sju, S. jumonjinensis; Nla, N. lactamdurans; Lla, L. lactamgenus; Ach, A. chrysogenum (DAOC/DACS). (* indicates a fully conserved position; : indicates a highly conserved position; . indicates a weakly conserved position)

137

Chapter 6: Relevance of single mutations at amino acid positions 275, 281, 304 and 305 of scDAOCS leading to improved penicillin conversion activities

6.1. Kinetic properties of N304 mutant scDAOCSs

In an earlier study on residue N304 of scDAOCS, mutation to a basic amino acid, i.e. lysine or arginine, gave the highest improvement in the enzyme’s specific activity among the five mutants N304A, N304K, N304L, N304M and N304R in the four hydrophobic penicillin conversions (Chin et al., 2004). The specific activities of scDAOCS mutants for ampicillin conversion determined by HPLC are comparable to those determined by bioassay (Chin et al., 2004).

In order to better understand the mechanism of these improvements, kinetic analyses of all improved mutants, i.e., N304A, N304K, N304L, N304M and N304R single mutants, were done using the enzyme kinetic assay reported by Dubus and co- workers (2001). The results showed that all the five single mutants have lower Km values than the wild-type enzyme (Table 6.1). This indicates that these mutations improved the substrate binding affinity of scDAOCS for ampicillin. N304K and

N304R mutations yielded the highest improvement in the substrate binding affinity of scDAOCS with approximately 13- (0.54 mM) and 26-fold (0.26 mM) reduction in Km values, respectively, compared to the wild-type enzyme (6.94 mM). These led to higher catalytic efficiencies, as reflected by higher kcat/Km values than the wild-type enzyme (Table 6.1).

138

Since lysine and arginine residues are basic amino acids, it is likely that the positively-charged side chains of these two amino acids enhanced the interaction of the enzyme with the penicillin substrates via ionic interaction. However, in silico structural and mutational analyses using SwissPdb Viewer and LIGPLOT suggested otherwise. Based on the crystal structures of scDAOCS complexed with ampicillin and penicillin G, the amino acid residue at position 304 of scDAOCS is likely to interact with the side chain of the penicillin substrate (Valegard et al., 2004). Hence, it is unlikely that there will be any ionic interaction between lysine or arginine residues with the hydrophobic side chain of ampicillin. Furthermore, LIGPLOT results indicate that both lysine and arginine are able to bind the penicillin substrate via hydrophobic interaction (Figure 6.1).

Table 6.1. Kinetic parameters of wild-type and N304 mutant scDAOCSs for ampicillin conversion.

a -1 a -1 -1 Enzyme Km (mM) kcat (s ) kcat/Km (M s ) Relative kcat/Km Wild-type 6.94 ± 0.70 0.059 ± 0.004 8.6 1.0

N304A 1.09 ± 0.08 0.040 ± 0.002 36.5 4.2

N304K 0.54 ± 0.03 0.038 ± 0.003 70.1 8.2

N304L 1.65 ± 0.30 0.041 ± 0.001 24.8 2.9

N304M 5.12 ± 0.52 0.054 ± 0.003 10.5 1.2

N304R 0.26 ± 0.01 0.052 ± 0.001 197.1 22.9 aThe values are mean ± standard deviations of at least three independent experiments with triplicates.

139

Figure 6.1. Schematic diagrams of predicted residues interacting with ampicillin substrate in scDAOCS (1UNB) using SwissPdb Viewer (left column) and LIGPLOT program (right column). (A) The predicted substrate-interacting residues in wild-type scDAOCS did not include residue N304. After in silico mutation using SwissPdb Viewer, the mode of interaction was assessed by using LIGPLOT program and the mutated residues were circled: (B) N304R mutant and, (C) N304K mutant.

140

6.2. The importance of amino acid position 304 of scDAOCS

The involvement of residue N304 in substrate interaction was not clear.

Although it resides close to the side chain of the penicillin substrate, it was predicted not to form any non-covalent bonds with the substrate based on the crystal structures of scDAOCS complexed with penicillin G and ampicillin (Valegard et al., 2004).

However, many studies have suggested otherwise (Chin et al., 2004; Chin and Sim,

2002; Lee et al., 2002; Chin et al., 2001). Bearing in mind that penicillin N is the native substrate of scDAOCS and not penicillin G or ampicillin, the structural conformation observed in the crystal structure of scDAOCS-penicillin G or scDAOCS-ampicillin complexes does not reflect the ideal or optimal interaction between the enzyme and the substrate. These structures probably reflect the minimal interaction of the enzyme with a penicillin substrate in order to be resolved together.

In a previous study, the replacement of residue N304 with a non-polar residue

(leucine, methionine or alanine) or a basic charged residue (lysine or arginine) could improve the catalytic activity of the enzyme (Chin et al., 2004). Although both of the mutant enzymes with a basic residue substitution had higher relative activities than the mutant enzymes with a non-polar residue substitution, structural analyses did not suggest that there was any ionic interaction or hydrogen bonding between the basic residues and the penicillin substrate. Studies have shown that perturbation of its pKa might occur when the arginine residue is buried in the hydrophobic catalytic core of an enzyme, leading to deprotonation of the guanidinium group at a lower pH (Guillen

Schlippe and Hedstrom, 2005). Hence, it could be the aliphatic region of the side chains of both arginine and lysine residues having hydrophobic interaction with the hydrophobic side chains of the penicillin substrates that leads to improved substrate binding affinity of the enzyme. This hypothesis was further supported by the 141

LIGPLOT prediction results, which suggested that there are hydrophobic interactions between the designated residues and the penicillin substrate. Therefore, a decrease in the Km value of scDAOCS after N304X mutation will corroborate this plausible interaction between these amino acid residues and the penicillin substrates.

In this study, the kinetic parameters of these mutants were determined. The Km values of the mutants were considerably lower than that of the wild-type enzyme. This indicated substantial improvement in the substrate binding affinity of the enzyme.

This also supported the notion that a mutation of asparagine at position 304 to a hydrophobic amino acid residue improves the substrate binding affinity of scDAOCS for hydrophobic penicillins via better hydrophobic interaction. Interestingly, Wei and co-workers (2003) also demonstrated that N304K mutation was able to improve substrate binding affinity of scDAOCS for penicillin N by 3.5-fold and increased the catalytic efficiency by about 4-fold. This further supported the involvement of residue

N304 in interacting with the native substrate. However, the exact mechanism for this improvement is not structurally clear and will only be elucidated upon crystallisation of scDAOCS with penicillin N.

This study has provided an important insight for modifying an enzyme’s substrate specificity, facilitated by the availability of an enzyme-substrate crystal structure. Although a substrate may be bound by an enzyme, the interaction mode may not always be optimum, especially for a non-native substrate. Being able to recognise the plausible site for modification is critical in enhancing the substrate interaction to improve enzymatic activity. Subsequently, the question of substitution to the right amino acid residue becomes easier since there are only 19 proteinogenic amino acids left to consider. By considering the properties of each amino acid, empirical testing could be narrowed down. 142

6.3. Relevance of V275I, C281Y and I305M single mutations leading to improved penicillin conversion activities

Wei and co-workers (2003) demonstrated that the V275I, C281Y and I305M mutations improved the conversion activity of scDAOCS only with penicillin G. It is not clear whether the effects of these mutations would be the same for the conversion of other penicillin substrates. In this study, the substrate specificity of these mutant enzymes was determined using the hole-plate bioassays and was found to include ampicillin, phenethicillin, and carbenicillin (Table 6.2). HPLC analyses of ampicillin conversions by these mutants were done to affirm the hole-plate bioassay results

(Table 6.3). The bases of improvement observed in these mutants were also explored using enzyme kinetic assays (Table 6.4). These results were also compared with other improved single mutants studied in this project.

The conversion activities of V275I, C281Y and I305M mutants were higher than that of the wild-type enzyme for the ring-expansion of penicillin G, ampicillin, carbenicillin and phenethicillin (Table 6.2). The improvement observed for the

C281Y and I305M single mutants for the conversion of different penicillins was greater than that observed for the V275I mutant. Compared to the relative activities of other single mutants, the relative activity of C281Y showed the greatest improvement for carbenicillin conversion (572%), followed by the I305M mutant (450%). For ampicillin conversion and phenethicillin conversion, the activities of the N304R mutant were dominant. However, this mutant was only as efficient as the C281F mutant for penicillin G conversion. These results were similarly observed in the

HPLC analyses (Table 6.3).

143

Structural analyses of these three residues using SwissPdb Viewer revealed that I305 is approximately 4.5 Å from the β-lactam ring of the penicillin substrate

(Figure 6.2). After mutation to methionine, the side chain, which is longer than that of isoleucine, is closer to the β-lactam ring of the penicillin substrate. Thus, the methionine residue could potentially stabilise or orient the penicillin substrate in a more receptive manner for ring-expansion. Interestingly, I305L mutation also showed improvement in the conversion activity of scDAOCS using penicillin G although at a level lower than I305M mutant (Wei et al., 2003). This suggests that the branching point at Cγ of the side chain in leucine is more effective in interacting with the penicillin substrate compared to isoleucine, which is branched at Cβ of its side chain.

The molecular mechanisms of V275I and C281Y single mutants that contribute to the increased catalytic activity of the enzyme are still unclear. Since the two residues are not located at the active site of scDAOCS, it is reasonable to postulate that their effects are structure-based. Residue V275 is located in a small sub- domain on the β13 strand and surrounded by residues F273, V285, and L287 in the polypeptide chain extending from strand β13 to strand β14 (Figure 6.3). The interaction of these residues is important in maintaining the integrity of this small sub- domain, which extends to the C-terminal end of the enzyme. When V275 was replaced with leucine, which is structurally similar to isoleucine, comparable relative activity was observed. Since V275 residue is strictly conserved among the annotated

DAOCSs, it is likely that an aliphatic residue at this position is important for the interaction with the surrounding hydrophobic residues to maintain the integrity of this sub-domain. However, the relative activities of the V275I mutant for conversion of all penicillins are about two-fold higher than those of V275L. This suggests that the

144

branching variation in the side chains of isoleucine and leucine at this position is important for influencing the enzyme activity.

Residue C281 is also located within the small sub-domain between the α9 helix and the β14 strand (Figure 6.3). In silico mutational analyses revealed that a hydrogen bond was formed between the hydroxyl group on the side chain of tyrosine and the backbone oxygen of V238 when C281 was mutated to a tyrosine. This hydrogen bond could possibly help to stabilise the polypeptide chain leading to residue H243, which is critical for iron (II) coordination. This is consistent with the results reported by Wei and co-workers (2003). To further validate the contribution of the hydrogen bond, C281 was changed to phenylalanine, which resembles tyrosine structurally except that the hydroxyl group on the aromatic ring is absent.

Interestingly, the C281F mutant exhibited specific activity similar to that of the

C281Y mutant. This implied that the improved activity obtained was not due to the predicted hydrogen bond formation. Since phenylalanine is non-polar, the interactions of either tyrosine or phenylalanine at position 281 with its surrounding residues are likely to be due to hydrophobic and van der Waals forces.

145

Table 6.2. Relative activities of wild-type and mutant scDAOCSs determined by bioassay.

Relative activitya (%)

NH2 H CH3 COOH H H N S O CH N H N S CH3 S N CH CH3 S 3 CH3 O N CH3 O N CH O N CH3 3 O N CH O O O 3 COOH COOH COOH O COOH Enzyme Ampicillin Penicillin G Phenethicillin Carbenicillin Wild-type 100 100 100 100 V275I 129 157 150 227 V275L 85 130 99 113 C281Y 166 251 280 572 C281F 191 264 300 518 N304Ab 192 163 163 242 N304Kb 270 237 211 404 N304Lb 163 129 173 235 N304Mb 120 92 127 161 N304Rb 346 263 360 398 I305M 313 234 318 450 R306Lc 186 105 173 283 R307Lc 133 124 122 108 N304AR306L 214 114 319 393 N304KR306L 245 80 380 482 N304LR306L 145 40 185 218 N304MR306L 194 72 245 232 N304RR306L 256 92 344 514 V275IN304A 265 233 330 721 C281YN304A 294 383 714 911 V275IN304K 334 294 635 867 C281YN304K 376 428 948 1180 V275IN304L 225 141 245 524 C281YN304L 330 211 471 1001 V275IN304M 160 174 261 230 C281YN304M 225 277 428 485 V275IN304R 360 381 698 814 C281YN304R 430 441 1041 1309 V275II305M 406 420 685 1057 C281YI305M 491 495 1109 1347 V275IR306L 343 168 379 665 C281YR306L 394 191 555 1010 V275IR307L 172 169 238 333 C281YR307L 274 334 520 786 aThe relative activity of each mutant enzyme is the mean specific activity of the mutant enzyme relative to that of the wild-type enzyme, which was defined as 100%. The mean specific activity of each enzyme was determined using at least three independent experiments with duplicates. bscDAOCS mutant obtained from a previous study (Chin et al., 2004). cscDAOCS mutant obtained from a previous study (Chin and Sim, 2002).

146

Table 6.3. Specific and relative activities of wild-type and mutant scDAOCSs for ampicillin conversion determined by HPLC.

NH2 NH2 H H N S N S CH3 DAOCS

O N CH3 O N O α-KG, Succinate, O CH3 C O2 CO2, H2O OOH COOH Ampicillin Cephalexin

Enzyme Specific activitya (Umg-1) Relative activityb (%) Wild-type 5.67 ± 0.59 100 V275I 7.76 ± 0.19 137 V275L 4.42 ± 0.25 78 C281Y 12.30 ± 0.42 217 C281F 12.21 ± 0.63 215 c N304A 9.84 ± 0.21 173 c N304K 13.07 ± 0.37 230 c N304L 10.24 ± 0.34 180 c N304M 6.34 ± 0.18 112 c N304R 12.92 ± 0.32 228 I305M 17.24 ± 0.68 304 d R306L 9.83 ± 0.28 173 d R307L 6.08 ± 0.17 107 N304AR306L 15.33 ± 2.07 270 N304KR306L 17.82 ± 1.31 314 N304LR306L 9.51 ± 0.42 168 N304MR306L 12.70 ± 0.96 224 N304RR306L 17.67 ± 2.34 311 V275IN304A 11.42 ± 2.80 201 C281YN304A 13.38 ± 3.52 236 V275IN304K 13.19 ± 2.38 233 C281YN304K 14.61 ± 2.13 257 V275IN304L 10.26 ± 1.05 181 C281YN304L 12.20 ± 3.05 215 V275IN304M 8.26 ± 0.93 146 C281YN304M 13.26 ± 1.51 234 V275IN304R 18.06 ± 1.77 318 C281YN304R 20.39 ± 2.80 359 V275II305M 25.48 ± 1.38 449 C281YI305M 28.21 ± 0.77 497 V275IR306L 19.99 ± 1.38 352 C281YR306L 21.72 ± 1.34 383 V275IR307L 12.39 ± 1.67 218 C281YR307L 18.24 ± 1.45 322 aOne unit of activity was defined as the amount of DAOCS required to form the equivalent of 1 µg of cephalexin per min at 30°C. The specific activities are the means ± standard deviations of at least three independent experiments with duplicates. bThe relative specific activity of each mutant enzyme is expressed as the percentage of the specific activity of the mutant enzyme relative to that of the wild-type enzyme, which was defined as 100%. cscDAOCS mutant obtained from a previous study (Chin et al., 2004). dscDAOCS mutant obtained from a previous study (Chin and Sim, 2002).

147

Table 6.4. Kinetic parameters of wild-type and mutant scDAOCSs for ampicillin conversion.

a -1 a -1 -1 Enzyme Km (mM) kcat (s ) kcat/Km (M s ) Relative kcat/Km Wild-type 6.94 ± 0.70 0.059 ± 0.004 8.6 1.0 V275I 3.41 ± 0.17 0.044 ± 0.001 12.8 1.5 V275L 4.26 ± 0.32 0.032 ± 0.001 7.5 0.9 C281Y 1.02 ± 0.07 0.044 ± 0.001 43.5 5.1 C281F 1.35 ± 0.10 0.042 ± 0.003 31.2 3.6 N304A 1.09 ± 0.08 0.040 ± 0.002 36.5 4.2 N304K 0.54 ± 0.03 0.038 ± 0.003 70.1 8.2 N304L 1.65 ± 0.30 0.041 ± 0.001 24.8 2.9 N304M 5.12 ± 0.52 0.054 ± 0.003 10.5 1.2 N304R 0.26 ± 0.01 0.052 ± 0.001 197.1 22.9 I305M 1.00 ± 0.08 0.104 ± 0.011 104.1 12.1 R306L 1.01 ± 0.05 0.050 ± 0.002 49.2 5.7 R307L 5.96 ± 1.48 0.050 ± 0.003 8.4 1.0 N304AR306L 0.20 ± 0.10 0.034 ± 0.008 170.8 19.9 N304KR306L 0.12 ± 0.03 0.033 ± 0.005 285.1 33.2 N304LR306L 0.40 ± 0.09 0.029 ± 0.001 71.1 8.3 N304MR306L 0.36 ± 0.06 0.026 ± 0.001 72.3 8.4 N304RR306L 0.19 ± 0.03 0.030 ± 0.005 160.1 18.6 V275IN304A 0.68 ± 0.10 0.042 ± 0.002 61.7 7.2 C281YN304A 0.38 ± 0.03 0.052 ± 0.003 137.0 15.9 V275IN304K 0.38 ± 0.07 0.040 ± 0.002 107.8 12.5 C281YN304K 0.13 ± 0.003 0.046 ± 0.004 361.9 42.1 V275IN304L 1.04 ± 0.13 0.043 ± 0.002 41.0 4.8 C281YN304L 0.34 ± 0.03 0.043 ± 0.001 125.4 14.6 V275IN304M 3.60 ± 0.40 0.048 ± 0.002 13.5 1.6 C281YN304M 0.80 ± 0.03 0.047 ± 0.001 59.3 6.9 V275IN304R 0.11 ± 0.03 0.039 ± 0.001 366.9 42.7 C281YN304R 0.05 ± 0.01 0.043 ± 0.002 865.2 100.6 V275II305M 0.70 ± 0.06 0.105 ± 0.003 151.3 17.6 C281YI305M 0.21 ± 0.09 0.122 ± 0.010 569.9 66.3 V275IR306L 0.51 ± 0.15 0.037 ± 0.004 72.3 8.4 C281YR306L 0.15 ± 0.01 0.037 ± 0.001 254.3 29.6 V275IR307L 4.27 ± 0.31 0.049 ± 0.003 11.4 1.3 C281YR307L 1.20 ± 0.07 0.046 ± 0.002 38.3 4.5 aThe values are means ± standard deviations of at least three independent experiments with triplicates.

148

Figure 6.2. Schematic diagrams of predicted residues interacting with ampicillin substrate in scDAOCS (PDB accession number: 1UNB) using SwissPdb Viewer (left column) and LIGPLOT program (right column). (A) The predicted substrate- interacting residues in wild-type scDAOCS did not include residue I305. After in silico mutation using SwissPdb Viewer, the mode of interaction was assessed by using LIGPLOT program and the mutated residues were circled: (B) I305M mutant.

149

Figure 6.3. Spatial orientation of residues V275 and C281 (orange) in scDAOCS (PDB accession number: 1UNB). Top panel: close-up view of residues V275 and C281. Residues V275 and C281 are located at a small sub-domain away from the active site. Residue V275 is located on β13 strand and surrounded by F273, V285 and L287 residues. Residue C281 is located along the polypeptide chain extending between α9 helix and β14 strand. Bottom panel: in silico mutations of (A) V275I, (B) V275L, (C) C281Y and (D) C281F mutations using SwissPdb Viewer (dotted green line: hydrogen bond).

150

6.4. The importance of the small sub-domain where V275 and C281 reside

Earlier studies by Wei and co-workers (2003) have identified V275I and

C281Y single mutations to be beneficial in improving the conversion activity of scDAOCS for penicillin G. However, it was not clear how these two mutations led to higher conversion activities in scDAOCS.

The replacement of valine at amino acid position 275 to a more hydrophobic isoleucine probably improved its interaction with surrounding residues F273, V285 and L287. In this project, V275L mutation also yielded slight improvement in the ring-expansion activity of penicillin G and carbenicillin, and comparable ring- expansion activity of ampicillin and phenethicillin as the wild-type enzyme. These two conserved mutations likely further stabilised the sub-domain itself and indirectly affect the catalytic activity of scDAOCS (Figure 6.4). Hence, the improvements obtained by these two mutations were not as great as that obtained for C281Y mutation.

An earlier postulation for C281Y mutation was that upon mutating residue

C281 to a tyrosine, a hydrogen bond could be formed with the backbone of V238, which leads to β11 where residue H243 resides (Figure 6.4). This could possibly help to stabilise the polypeptide chain, which forms part of the co-factor binding motif critical for iron (II) coordination. In this study, a mutation to phenylalanine showed that the plausible hydrogen bonding is not the leading factor in enhancing the conversion activity of scDAOCS; the aromatic ring in the side chain of the amino acid residue is sufficient to enhance the ring-expansion activity of scDAOCS.

Interestingly, another mutation L277Q was also reported to elevate the activity of scDAOCS in penicillin G conversion (Table 6.5) (Wei et al., 2005). This residue resides near the polypeptide chain that leads to β11 (Figure 6.4), which probably has a 151

similar effect as C281Y and C281F mutations. Furthermore, based on the kinetic studies of V275I, L277Q and C281Y mutants, only L277Q and C281Y mutants demonstrated higher catalytic efficiency than the wild-type enzyme for both penicillin

N and penicillin G conversions. This suggests that there is a certain amount of flexibility in this sub-domain, which is further stabilised by interacting with the main domain upon either L277Q or C281Y mutation. This stabilisation could indirectly influence the catalytic center and lead to a higher catalytic activity. Therefore, it is likely that this small sub-domain plays a very critical role in the functionality of scDAOCS.

Table 6.5. Kinetic studies of wild-type and mutant scDAOCSs in penicillin N and penicillin G conversions (Wei et al., 2005, 2003).

Penicillin N Penicillin G -1 -1 -1 -1 -1 -1 Enzyme Km (mM) kcat (s ) kcat/Km (M s ) Km (mM) kcat (s ) kcat/Km (M s ) Wild-type 0.014 0.307 22,000 2.58 0.0453 18

V275I 0.012 0.252 20,000 1.68 0.0502 30

L277Q 0.011 0.297 27,000 1.02 0.1036 102

C281Y 0.006 0.273 47,000 0.68 0.0744 109

152

Figure 6.4. Spatial location of small sub-domain where residues V275 and C281 reside. Residue H243 reside on β11 and residues N304, I305 and R306 reside on β15.

153

Chapter 7: Relevant double mutations of scDAOCS result in higher binding specificities which improve penicillin bioconversion

7.1. Double mutation analyses of scDAOCS

Multiple mutations are often necessary to obtain a desired property or function in a protein. As reviewed by Wells (1990) and Mildvan and co-workers (1992), the effect of introducing a second mutation into a mutant enzyme could be synergistic, additive, partially additive, antagonistic or could have no additional effect, and therefore, the interactions between two or more mutations warrant experimental elucidation.

The effects of selected single mutations on the activity of scDAOCS have been demonstrated in various studies, with two observations made. The C-terminal residues (i.e., N304, I305, R306, and R307) are close to the penicillin substrate. Thus, mutations at these sites would have a direct effect on enzyme catalysis through interactions with the substrate. In contrast, mutations at the V275 and C281 residues, which are not located at the catalytic site, would have indirect consequences. Since

N304 of scDAOCS has been thoroughly studied for beneficial amino acid substitutions, it is a good candidate for investigating the effects of combining such mutations on the catalytic activity of the enzyme and the molecular effects of the mutations on each other. Therefore, two types of mutation combinations were tested;

N304X mutations (where X is alanine, leucine, methionine, lysine, or arginine) were

154

combined with distal site mutations (V275I and C281Y), as well as a mutation at a site close to position 304 (R306L).

7.1.1. Comparative analyses of V275IN304X, C281YN304X, and N304XR306L double mutants

A comparative analysis of the V275IN304X and C281YN304X double mutants with the corresponding single mutants revealed that the specific activities of the double mutants were higher than those of the corresponding single mutants for all the substrates tested (Table 6.2). The C281N304X double mutants were consistently more active than the V275IN304X double mutants. The trend observed was that the greatest increase in the conversion activities resulting from double mutations involved the single mutation that resulted in the greatest improvement (i.e., combination of

C281Y with N304R or N304K). This suggests that combining relevant improved single mutants can have additive or synergistic effects on the resulting enzymatic capability.

In contrast, some of the N304XR306L double mutants differed from the improved V275IN304X and C281YN304X double mutants (Table 6.2). For example, for ampicillin conversion, the relative activities of the N304KR306L (245%) and

N304RR306L (256%) double mutants were lower than those of the corresponding single mutants, the N304K (270%) and N304R (346%) mutants. Only slight increases in the specific activity were observed for the N304AR306L (214%) and

N304MR306L (194%) double mutants compared to the corresponding single mutants, the N304A (192%), N304M (120%), and R306L (186%) single mutants. For phenethicillin conversion, the N304AR306L, N304KR306L, and N304MR306L double mutants had higher relative activities than the corresponding single mutants. 155

As for carbenicillin conversion, the N304AR306L, N304KR306L, and N304RR306L double mutants had higher relative activities than the corresponding single mutants.

These results suggest that the N304LR306L double mutation was not an ideal combination as the activity of the N304LR306L double mutant for any of the three conversions was lower than that of the corresponding single mutants. Intriguingly, all

N304XR306L double mutations were counter effective for penicillin G conversion, and the relative activities were similar to or lower than that of the wild-type enzyme.

These results suggest the possibility that one mutation affects the environment of another mutation in close proximity, leading to undesirable effects.

7.1.2. Effects of V275I or C281Y with other C-terminal mutations

Since improvement was obtained by combining the V275I or C281Y mutation with N304X mutations, it can be extrapolated that when V275I and C281Y are combined with other C terminal mutations (namely, I305M, R306L, and R307L), similar outcomes should be obtained. As expected, V275I-C terminus double mutants exhibited further improvement compared to the corresponding single mutants (Table

6.2). Although the trend for most C281Y-C terminus double mutants was similar to the trend for the V275I-C terminus double mutants, the C281YR306L double mutant did not show further improvement in the relative penicillin G conversion activity

(191%) compared to the C281Y single mutant (251%). Nevertheless, these results supported the conclusion that the V275I and C281Y mutations increased the scDAOCS activity when they were combined with other beneficial C-terminal mutations.

156

7.1.3. Determination of specific activities of wild-type and mutant scDAOCSs by

HPLC

HPLC was also used to confirm the results obtained from bioassays for ampicillin conversion. These results were comparable to those obtained with bioassays (Table 6.3).

7.1.4. Kinetic parameters of wild-type and mutant scDAOCSs for ampicillin conversion

The Km values of all mutant enzymes were lower than that of the wild-type scDAOCS (6.94 mM) for ampicillin conversion, indicating that the substrate binding affinities were higher (Table 6.4). In the single mutants, the N304R (0.26 mM),

N304K (0.54 mM), and I305M (1.00 mM) mutations increased the substrate binding affinity of the enzyme most effectively. This property was further enhanced when these mutations were combined with the V275I, C281Y, or R306L mutations.

Interestingly, the I305M mutation doubled the turnover rate of the enzyme

(0.104 s-1) compared to that of the wild-type scDAOCS (0.059 s-1). This is the only mutation that significantly increased the turnover rate of scDAOCS’s ring-expansion activity. Similar kcat values were obtained regardless of the type of mutation with

-1 which I305M was combined; the kcat for V275II305M was 0.105 s , and the kcat for

C281YI305M was 0.122 s-1. However, the other mutations affected the turnover rate of scDAOCS to various degrees. It was observed that the double mutants containing

-1 the R306L mutation had the lowest kcat values (0.026 s for the N304MR306L double mutant and 0.037 s-1 for both the V275IR306L and C281YR306L double mutants). In contrast, the R306L single mutation (0.050 s-1) did not affect the turnover rate of the ring-expansion activity as much as the double mutations. 157

In general, the kcat/Km values of all double-mutant scDAOCSs were higher than that of the wild-type enzyme for ampicillin conversion. Four double mutants,

C281YN304R, C281YI305M, V275IN304R, and C281YN304K, showed the greatest improvement in the enzyme’s catalytic efficiency (101-, 68-, 42-, and 42-fold increments, respectively). These kinetic data supported the hypothesis that the ring- expansion activity of these double mutants is superior.

7.2. Combinatory mutagenesis as an approach in understanding the substrate binding mechanism of scDAOCS

More than one mutation in an enzyme is generally required to alter its properties in a desirable way. When two mutations are combined in a single protein, there can be various effects (Mildvan et al., 1992; Wells, 1990). The effects can be cumulative if the mutations are at distal sites (Skinner and Terwilliger, 1996; Lo et al.,

1995). For example, in crystallographic studies of double mutations in the gene V protein of bacteriophage f1, it was found that the coordinate shifts were greatest near the site of mutation and decreased with increasing distance (Skinner and Terwilliger,

1996). Very few changes were observed more than 10Å from each mutation site.

Thus, the influence of each mutation on the structural changes in an enzyme seems to be distance dependent. When two mutation sites are sufficiently distal, the effects of the mutations are independent of each other. However, the precise effect of combining two mutations that are close together is still not well understood. Hence, the effect of one mutation on another was investigated in this study using combinatory mutagenesis.

The bioassay results showed that N304XR306L double mutant, which had two mutations that were one amino acid apart (approximately 7.1Å), exhibited similar or 158

higher specific activities compared to the corresponding single mutants for ampicillin, phenethicillin, and carbenicillin conversions. Surprisingly, the penicillin G conversion activity of N304XR306L double mutants was similar to or lower than that of the wild- type enzyme. Since the side chains of residues N304 and R306 of scDAOCS are in opposite directions, it is very likely that they have an opposing effect on each other.

The presence of a functional group in the side chain of ampicillin (amino group), carbenicillin (carboxyl group) and phenethicillin (methyl group) probably stabilised or enhanced the interaction of the penicillin substrate with the amino acid residue at position 304 and opposed the effect at position 306 after N304XR306L double mutations.

When N304X mutations were combined with V275I and C281Y mutations, which are 28 and 22 amino acids from position 304, respectively, further improvements compared to the corresponding single mutants were observed. The bioassay results showed that approximately 10-fold increment in the carbenicillin conversion activity of scDAOCS was observed when the C281Y mutation was combined with the N304K and N304R mutations. Increases were also observed when the C281Y mutation was combined with other C-terminal single mutations (I305M,

R306L, and R307L). These results showed that the effects of the mutations are independent of each other and the overall effect could be additive or synergistic.

Therefore, this distance-dependent effect should be considered for rational engineering of DAOCS with desired properties. The results further supported the notion that the C281Y, N304K, N304R, and I305M mutations are strategic sites for engineering scDAOCS with improved substrate specificity and catalytic activity.

159

Chapter 8: Characterisation of scDACS

Native DACS from S. clavuligerus ATCC 27064 (synonymous to strain DSM

738 and NRRL 3585) has been purified and shown to be active as a monomer. It requires α-ketoglutarate, ferrous iron and molecular oxygen for its hydroxylation activity. It is stimulated by DTT or reduced glutathione, and ATP. However, a large volume of culture is required to purify the native enzyme for characterisation.

Therefore, high level heterologous expression would be beneficial. This has been attempted by another group where the cefF gene has been cloned and expressed in E. coli but the highly-expressed DACS occurred as inclusion bodies (Kovacevic and

Miller, 1991).

Recombinant DNA technology has offered the opportunity to express the desired protein with a fusion partner to improve its expression and solubility. GST as a fusion partner has successfully overcome many cases where the protein-of-interest was insolubly expressed. Furthermore, it allows purification of the protein-of-interest using affinity chromatography via GST tag. Since recombinant GST-scDAOCS has been shown to be active, a similar strategy was adopted for the study of scDACS.

This also allows a direct comparison of the functional activities of scDAOCS and scDACS. Hence, cefF gene of S. clavuligerus NRRL 3585 strain was cloned into the same expression vector pGK and expressed as recombinant GST-scDACS in E. coli

BL21 (DE3) strain. Subsequently, expression and solubility of recombinant GST- scDACS was determined using SDS-PAGE analysis. Upon obtaining solubly- expressed fusion protein, functional characterisation of scDACS was also carried out.

160

8.1. Cloning of scDACS

Both forward (OL963) and reverse (OL964) cloning primers were designed based on the sequence deposited in GenBank, with an accession number M63809. The forward and reverse cloning primers were designed with the inclusion of BamHI and

EcoRI restriction endonuclease recognition sequences at their 5’-ends, respectively.

This was to facilitate cloning of cefF gene into pGK expression vector. Upon PCR amplification of cefF from genomic DNA of S. clavuligerus NRRL 3858, an approximately 1 kbp PCR product was obtained and detected using agarose gel electrophoresis (Figure 8.1). This product was excised and purified from the agarose gel using QIAquick® Gel Extraction kit (Qiagen). The purified product was subsequently ligated into pCR®-Blunt II TOPO® vector using Zero Blunt® TOPO®

PCR Cloning kit (Invitrogen). The nucleotide sequence of the ligated PCR product was then validated by DNA sequencing. Upon validating the identity of the ligated product to be cefF gene, it was sub-cloned into pGK expression vector and fully sequenced again to ensure authenticity.

Interestingly, the nucleotide sequence of the cloned cefF gene possessed a nucleotide different from the deposited one. This cloned sequence has a transversion mutation at nucleotide position 126. This T126G transversion mutation resulted in a conserved amino acid mutation of an aspartic acid to a glutamic acid at amino acid position 42 (Figure 8.2). Repeated PCR amplification using another genomic DNA extract of S. clavuligerus and DNA sequencing showed the same sequence, hence, suggesting a sequence variant of scDACS from the earlier report (Kovacevic and

Miller, 1991). This recombinant plasmid is designated as pGK-scDACS.

161

M 1 2

2kbp

1kbp cefF: 957bp 750bp

Figure 8.1. Agarose gel electrophoresis of PCR-amplified product of cefF gene from S. clavuligerus genomic DNA. Lane M: GeneRuler™ 1kb DNA ladder (Fermentas); Lane 1: PCR amplification of cefF; Lane 2: negative control (without S. clavuligerus genomic DNA added in the PCR reaction).

A Forward sequencing Reverse sequencing

B N: T126G DACS_F GGGGAGAAGGACCACCGGCTGGCCACGGACACGGCGATGGACTTCTTCGCGAACGGCACC 157 M63809 GGGGATAAGGACCACCGGCTGGCCACGGACACGGCGATGGACTTCTTCGCGAACGGCACC 180 ***** ******************************************************

DACS1F MADTPVPIFNLAALREGADQEKFRECVTGMGVFYLTGYGAGEKDHRLATDTAMDFFANGT 60 M63809 MADTPVPIFNLAALREGADQEKFRECVTGMGVFYLTGYGAGDKDHRLATDTAMDFFANGT 60 *****************************************:****************** aa: D42E

Figure 8.2. Verification of PCR-amplified cefF gene of S. clavuligerus NRRL 3858. A. Electropherograms of forward and reverse DNA sequencing of PCR amplified cefF gene. B. Top panel shows the partial nucleotide sequence alignment of PCR amplified (DACS_F) and GenBank M63809 cefF genes, illustrating the nucleotide difference at position 126. Bottom panel shows the partial protein sequence alignment of translated amino acid sequence of PCR-amplified (DACS1F) and GenBank M63809 cefF genes, illustrating the amino acid difference at position 42.

162

8.2. Heterologous expression of recombinant scDACS

After verifying the nucleotide sequence of scDACS, the E. coli BL21 (DE3) clone harbouring pGK-scDACS was cultured and induced using the same conditions for scDAOCS to express the recombinant protein for purification and functional analysis. SDS-PAGE analysis of the cell-free extract showed that recombinant GST- scDACS was highly expressed in E. coli BL21 (DE3) at 25ºC. The recombinant GST- scDACS fusion protein was purified using Glutathione Sepharose 4B chromatography column. Cleavage by PreScission™ protease released the scDACS with an estimated molecular mass of approximately 35 kDa. A representative SDS-PAGE analysis of the cell free extract and purified fraction of the E. coli clone harbouring pGK- scDACS is shown in Figure 8.3.

Figure 8.3. SDS-PAGE analysis of cell-free extracts and purified fractions of E. coli BL21 (DE3) clone expressing recombinant GST-scDACS. Lane M: Precision Protein Standards™ Bio-Rad 161-0318; Lane 1: cell-free extract; Lane 2: insoluble fraction; Lane 3: purified fraction; Lane 4: PreScission™ Protease-cleaved scDACS; Lane 5: eluted fraction after protease cleavage; Lane 6: cell-free extract of E. coli BL21 (DE3) harbouring pGK; Lane 7: insoluble fraction of E. coli BL21 (DE3) harbouring pGK; Lane 8: purified GST; Lane 9: purified GST-scDAOCS.

163

8.3. Functional and mutational studies of recombinant scDACS

8.3.1. Identification of functional amino acid residues of scDACS

The amino acid sequences of all annotated DACSs deposited in GenBank were retrieved and aligned with all annotated DAOCSs and AchDAOC/DACS using

ClustalW multiple sequence alignment program (Figure 8.4). The conserved

HXDXnH co-factor binding motif and RXS co-substrate binding motif are highlighted in green and turquoise colours, respectively. The substrate binding residues of scDAOCS reported by Hsu et al. (2004) are highlighted in grey colour. These residues are highly conserved among DAOCSs, DACSs and AchDAOC/DACS, except amino acid position 73 (scDAOCS amino acid numbering) where it could be methionine, isoleucine, valine, alanine or threonine. Although position 158 is mainly leucine in

DAOCSs and it is a valine in all annotated DACSs, it was reported to be a valine in S. chartreusis DAOCS. Since leucine and valine are aliphatic non-polar amino acids, they are probably not involved in substrate selectivity, i.e., they do not play a role in differentiating the substrate, leading to ring-expansion or hydroxylation activity.

Hence, the amino acid residues involved in differentiating the substrates have to be other amino acid residues in the β-barrel that defines the substrate binding pocket of the enzyme.

In silico analysis of scDACS structure revealed 57% and 73% sequence identity and similarity to scDAOCS (Table 8.1). Structural modelling of scDACS using Swiss-Model revealed highly homologous tertiary structure with scDAOCS

(Figure 8.5). Since scDACS bears the same catalytic platform and is likely to possess the same tertiary structure as scDAOCS, structural and biochemical information of scDAOCS could be useful in identifying the amino acid residues involved in differentiating substrate recognition. The C-terminus of scDAOCS has been reported 164

to be important for substrate recognition. Similarly, the C-terminal residues of

AchDAOC/DACS have been reported to be important for ring-expansion and hydroxylation activities (Wu et al., 2005; Lloyd et al., 2004). Hence, it is likely that the C-terminal residues of scDACS are crucial. Interestingly, residues N304 and R306 of scDAOCS are conserved among the DAOCSs and AchDAOC/DACS (Figure 8.4).

The corresponding residues in DACSs are also conserved but are instead threonine and histidine, respectively (Figure 8.4). Since the resolved crystal structure of scDAOCS showed that residue N304 is oriented close to penicillin, it is likely that the corresponding residue T308 in scDACS will interact with the cephalosporin substrate for hydroxylation (Figure 8.6). Similarly, residue T310 of scDACS, which is the corresponding residue of R306 of scDAOCS, is likely to play a role in substrate recognition (Figure 8.6).

165

166

Figure 8.4. Multiple sequence alignment of annotated DAOCSs and DACSs of various microorganisms depicted with the secondary structure of scDAOCS. The secondary structure of scDAOCS was resolved from its crystal structure (Valegard et al., 1998). The cylinders refer to α-helices, the block arrows refer to β-sheets and the black lines refer to random coils of the tertiary structure of scDAOCS. The iron and co-substrate binding motifs of scDAOCS and scDACS are highlighted in green and turquoise colours, respectively. The substrate binding residues of scDAOCS reported by Hsu and co-workers (2004) and the corresponding residues in scDACS are highlighted in grey colour. Position 42 of scDACS and the corresponding positions in other DACS homologues are highlighted in a red box. Positions 304 and 306 of scDAOCS and the corresponding positions in other DAOCSs and DACSs are highlighted in blue boxes. (Scl: S. clavuligerus NRRL3585; Sch: S. chartreusis 102SH3; Sam: S. ambofaciens 29SA4; Sju: S. jumonjinensis NRRL5741; Nla: N. lactamdurans; Lla: L. lactamgenus YK90; Ach: A. chrysogenum; S65PH1: Streptomyces sp. isolate 65PH1) (GenBank accession numbers: Scl_DAOCS – P18548, Sch_DAOCS – AY318743, Sam_DAOCS – AY318742, Sju_DAOCS – AF317908, Nla_DAOCS – Z13974, Lla_DAOCS – CAA39984, Ach_DAOC/DACS – P11935, Scl_DACS – P42220, S65PH1_DACS – AY318744, Nla_DACS – P42219, Lla_DACS – CAA39985) (* indicates a fully conserved position; : indicates a highly conserved position; . indicates a weakly conserved position)

Table 8.1. Sequence homologies between various annotated DAOCSs and DACSs.

Sch Sam Sju Nla Lla Scl Ach Scl S65PH1 Nla Lla Similarity\Identitya DAOCS DAOCS DAOCS DAOCS DAOCS DAOCS DAOC/DACS DACS DACS DACS DACS Sch DAOCS 79.7 77.2 66.9 54.2 76.5 54.1 54.4 60.3 58 52.9 Sam DAOCS 88.7 85.9 70.4 54.5 81.4 54.1 56.3 60.3 58.3 52.2 Sju DAOCS 87.1 91.3 68.2 53.3 81.7 52.9 57.2 60.3 57.4 51.6 Nla DAOCS 78.7 84.1 82.2 56.6 70.1 57.3 57.7 61.5 60.8 54.1 Lla DAOCS 70.8 71.2 70.5 72.4 56.1 52 53.7 52.8 53.9 53 Scl DAOCS 86.8 92.3 91 83.4 72.4 53.8 57.2 60.3 60.3 52.7 Ach DAOC/DACSb 68.7 67.5 67.5 70.8 69 67.8 51.8 51.4 52 48.6 Scl DACS 70.1 72.3 72 71.7 69.3 73.3 66 75.8 76.1 50.9 S65PH1 DACS 75 76.6 76.6 74.2 69 76.9 67.2 86.8 76.6 53.7 Nla DACS 73 74 74 72.6 67.4 75.2 65.4 84.3 85.3 53 Lla DACS 69.6 69.6 69.6 70.4 72.4 69 62.3 65.1 68.1 68.1 aSequence homology values were determined using MatGAT programme (Campanella et al., 2003). (GenBank accession numbers: Scl_DAOCS – P18548, Sch_DAOCS – AY318743, Sam_DAOCS – AY318742, Sju_DAOCS – AF317908, Nla_DAOCS – Z13974, Lla_DAOCS – CAA39984, Ach_DAOC/DACS – P11935, Scl_DACS – P42220, S65PH1_DACS – AY318744, Nla_DACS – P42219, Lla_DACS – CAA39985) bA bifunctional enzyme that possesses both deacetoxycephalosporin C synthase and deacetylcephalosporin C synthase activities. (Scl: S. clavuligerus NRRL3585; Sch: S. chartreusis 102SH3; Sam: S. ambofaciens 29SA4; Sju: S. jumonjinensis NRRL5741; Nla: N. lactamdurans; Lla: L. lactamgenus YK90; Ach: A. chrysogenum; S65PH1: Streptomyces sp. isolate 65PH1)

167

Figure 8.5. Crystal structure of scDAOCS (1UOB) and homology model of scDACS. Homology model of scDACS is generated using Swiss-Model programme. The structures were viewed and manipulated using SwissPdb Viewer. The green-coloured regions in scDACS model are the constructed regions of the corresponding disordered regions in the crystal structure of scDAOCS. The loop coloured red carries the three additional amino acid residues unique to scDACS.

scDAOCS scDACS

Figure 8.6. C-termini of scDAOCS crystal structure (1UOB) and scDACS in silico model. scDAOCS C-terminal residues N304, I305 and R306 have been shown to be important sites for manipulation to upgrade the enzyme’s capabilities. The corresponding residues in scDACS are residues T308, M309 and H310, respectively.

168

8.3.2. Hydroxylation activity of scDACS

As the reaction mechanism of DACS is very similar to that of DAOCS, except for the difference in their substrate preferences, the ring-expansion assay conditions were often adopted, with the penicillin substrate replaced by DAOC (Lloyd et al.

2004; Baker et al. 1991). Therefore, in this study, the optimised reaction mixture reported by Sim and Sim (2001) for the detection of the ring-expansion activity of

GST-scDAOCS was adopted. Since the native hydroxylation substrate DAOC is not easily available, an analogue of DAOC, cephalexin, was used as a replacement. Under the reverse-phase HPLC conditions reported by Lloyd and co-workers (2004), the retention time of DAC (7.5 min) was lower than that of DAOC (13 min). This is expected since the addition of a hydroxyl group to DAOC will decrease its hydrophobicity and result in a lower retention time than the unhydroxylated substrate.

Hence, the HPLC analysis described in Chapter 3.2.6.2 was used to detect for the hydroxylated product formed in the hydroxylation assay.

The hydroxylation assay was done using purified GST-scDACS recombinant protein. The reaction was stopped by adding equal volume of absolute methanol to the reaction mixture. Subsequently, the test samples were analysed using reverse-phase

HPLC. The HPLC chromatograms of scDACS before and after the reaction were compared. The presence of a new peak will indicate the formation of the hydroxylated product of cephalexin. However, there was no sign of a new peak in the HPLC chromatograms of scDACS after both 30 min and 1 hour reactions.

Earlier studies on AchDAOC/DACS suggested its substrate specificity to be very narrow. It has been reported to show very low or negligible hydroxylation of

DAOC G, which is the ring-expanded product of penicillin G, and cephalexin (Lloyd et al., 2004). Hence, it is likely that cephalexin is not a suitable substrate for scDACS. 169

Although this recombinant scDACS is a D42E variant of the reported scDACS, it is unlikely that the failure to detect hydroxylation activity was due to this sequence variation since the corresponding position in other DACS homologues was reported to be either glutamic acid or serine residues (Figure 8.4). Since scDACS is structurally homologous to scDAOCs, it can be postulated that scDACS could be engineered to possess ring-expansion activity. This will indirectly suggest that the mutated residues are important for substrate differentiation.

8.3.3. Ring-expansion activity of scDACS

Earlier studies on scDACS have reported approximately 3% of ring-expansion activity compared to scDAOCS using the native substrate of DAOCS, penicillin N

(Baker et al. 1991). This suggests that scDACS is highly specific in its substrate recognition and hence, defines its enzymatic activity as hydroxylation. The standard ring-expansion assay reported by Sim and Sim (2001) was used on recombinant GST- scDACS. The penicillin substrates used were penicillin G and ampicillin, since penicillin N is not easily available. The bioassays revealed no zone of growth inhibition of test bacterium E. coli Ess for both substrates. These results suggest that scDACS does not possess ring-expansion activity using penicillin G and ampicillin as substrates, despite having high structural homology with scDAOCS.

8.3.4. C-terminal mutational studies of scDACS

Since the corresponding residues of N304 and R306 of scDAOCS are T308 and H310 in scDACS, respectively, these two residues were selected for mutational studies. To engineer scDACS with ring-expansion activity, T308 and H310 were mutated to asparagine and arginine, respectively, to mimic the C-terminus of wild- 170

type scDAOCS. Since N304R and R306L single mutants of scDAOCS showed improved ring-expansion activity using penicillin G as a substrate (Chin et al., 2004;

Chin and Sim, 2002), T308 and H310 of scDACS were also replaced with arginine and leucine, respectively, to determine possible enhancement in ring-expansion activity.

A similar mutagenesis strategy described for generating scDAOCS mutants was used to generate these scDACS mutants. The SDM template used was pGK- scDACS and the respective mutagenic oligonucleotides used are listed in Table 3.4.

SDS-PAGE analysis of the mutant cell lysates revealed similar profile as that of the wild-type enzyme (Figure 8.3). This suggests that the various mutations did not affect the soluble expression of the mutant sequences in the expression host E. coli BL21

(DE3).

The scDACS mutants were subsequently tested for ring-expansion activity.

Table 8.2 shows the results of the hole-plate bioassays. T308N mutant did not show any detectable ring-expansion activity, whereas H310R mutant showed approximately

2% of wild-type scDAOCS’s ring-expansion activity. Interestingly, T308R and

H310L single mutants, which are mimics of scDAOCS N304R and R306L single mutants with improved activities, showed 2% and 18% ring-expansion activity, respectively. This supports the possibility of conferring ring-expansion activity by mutating critical residues in the putative substrate binding pocket of scDACS. To further validate this notion, T308NH310R and T308RH310R double mutants, which are closer mimics of wild-type and N304R mutant of scDAOCS, respectively, were generated and evaluated (Table 8.2). T308NH310R and T308RH310R double mutants showed 3% and 8% ring-expansion activities, respectively. These are higher than the respective single mutants of scDACS. Together, these results suggest that the C- 171

terminal residues T308 and H310 of scDACS, which are the corresponding residues of the C-terminal residues N304 and R306 of scDAOCS, are important in influencing the substrate recognition capability of the enzyme, leading to the type of enzymatic reaction.

Table 8.2. Specific and relative activities of scDAOCS, wild-type and mutant scDACSs in penicillin G conversion determined using a hole-plate bioassay.

Enzyme Specific activity (U/mg) a Relative activity (%)b Wild-type scDAOCS 6.58 ± 0.243 100

Wild-type scDACS NDc -

T308N ND -

T308R 0.11 ± 0.003 2

H310R 0.11 ± 0.002 2

H310L 1.21 ± 0.006 18

T308NH310R 0.23 ± 0.034 3

T308RH310R 0.53 ± 0.034 8 aOne unit of enzyme activity is the amount of enzyme required to form the equivalent of 1 µg of cephalosporin C per minute at 30°C. The specific activities are shown as the mean ± standard deviation of two independent experiments with duplicates. bRelative activity of each enzyme is expressed as the percentage of the specific activity of the enzyme relative to that of scDAOCS at 100%. cNot detected.

172

8.4. Proposed differential substrate recognition of scDAOCS and scDACS

It has long been recognised that scDACS resembles scDAOCS in various aspects. Although both enzymes use the same catalytic platform and reactants for enzymatic reaction, they recognise different substrates for respective enzymatic reactions. Since the catalytic reaction is essentially the same, i.e., oxidation, the difference lies in the substrate selectivity. Hence, structural differences in the substrate binding pocket should account for this reactivity difference. However, the challenge in this study is that the native substrates of DAOCS and DACS, i.e., penicillin N and DAOC, respectively, are not easily available. Providentially, other penicillins could be used as the substrate for probing the functionality of scDAOCS, albeit at a lower enzymatic activity (Chin and Sim, 2002; Cho et al., 1998). This has permitted various mutational studies of scDAOCS to be done, resulting in a better understanding of its structure-function relationship.

Intriguingly, A. chrysogenum has a bifunctional enzyme that possesses both ring-expansion and hydroxylation activity, i.e., AchDAOC/DACS. In silico structural modelling of AchDAOC/DACS yielded similar structural folds as scDAOCS except for the additional 20 amino acid residues at its C-terminus, which could not be modelled (Chin et al., 2003). Truncation of this additional tail resulted in 50% loss of hydroxylation activity using DAOC as a substrate and doubled the ring-expansion activity using penicillin G as a substrate (Lloyd et al., 2004). This shows that the additional C-terminal tail has a role in the hydroxylation activity of

AchDAOC/DACS. Importantly, M306I mutation of AchDAOC/DACS demonstrated the absolute importance of this residue at this position for hydroxylation activity

(Table 8.3) (Lloyd et al., 2004). This accounts for the low or negligible hydroxylation activity of scDAOCS, which has an isoleucine at the corresponding position 305. 173

However, this does not account for the inability of scDACS to ring-expand penicillin

G since it also possesses a methionine at the corresponding position 309.

Retrospectively, scDACS does not have the additional amino acid residues as

AchDAOC/DACS. Hence, the determinants for specific reactivity in the monofunctional scDAOC and scDACS have yet to be fully comprehended. Since scDACS could not hydroxylate cephalexin, a reverse engineering approach was adopted in this study to obtain more clues to this functional specificity.

Although scDAOCS has negligible hydroxylation activity on DAOC, a crystal structure of scDAOCS bound with DAOC has been resolved (1UOG). The spatial orientation of DAOC in this structure was very similar to that of penicillin G (1UOB) and ampicillin (1UNB) in scDAOCS (Figure 8.7). Assuming that scDACS was to bind DAOC in a similar manner for hydroxylation, it would be very likely that scDAOCS will be able to hydroxylate DAOC. However, this position adopted by

DAOC in the crystal structure of scDAOCS is not productive for hydroxylation

(Valegard et al., 2004). One possible reason is that hydroxylation of DAOC occurs at the methyl group in the C3 position of the dihydrothiazine ring of DAOC. This position is too distant from the catalytic centre for hydroxylation to occur (Figure

8.8). It is likely that DAOC has to be re-oriented to position the C3 methyl group to the iron centre.

Another important clue lies in the C-termini of DAOCSs, DACSs and

DAOC/DACS. Multiple sequence alignment of all annotated sequences revealed conservation at the corresponding position 304 of scDAOCS, i.e., asparagine in all

DAOCSs and AchDAOC/DACS, and threonine in all DACSs. Since the side chain of penicillin N and DAOC is aminoadipate, which is zwitterionic, and residue N304 of scDAOCS is likely to interact with the side chain of the penicillin substrate during 174

ring-expansion, the distinct conservedness of the amino acid residue at this corresponding position in all annotated DAOCSs and DACSs suggests that asparagine is required for ring-expansion and threonine is required for hydroxylation. It is very likely that the type of amino acid residue at this position will guide the substrate in the correct orientation for reaction. This will also explain the inability of scDACS to hydroxylate cephalexin since the side chain of cephalexin is hydrophobic and not zwitterionic. As for AchDAOC/DACS, the presence of the additional amino acid residues at its C-terminus probably compensates for the inability of residue N305 to interact with DAOC for hydroxylation.

In this study, scDACS was shown not to exhibit any ring-expansion and hydroxylation activity using penicillin G and ampicillin as respective substrates. Upon

T308R mutation, 2% relative ring-expansion activity of wild-type scDAOCS was obtained. This indicates that amino acid position 308 of scDACS is indeed the corresponding position of 304 of scDAOCS and is important for substrate recognition.

However, the undetectable ring-expansion activity of T308N mutant scDACS, which is a mimic of wild-type scDAOCS, and the low ring-expansion activity of T308R mutant also suggests that there are other residues involved in substrate recognition.

Mutation of residue H310 of scDACS, which is the corresponding residue of R306 of scDAOCS, to either arginine or leucine validated this notion. By combining both mutations at T308 and H310, i.e., T308NH310R and T308RH310R double mutants, further enhancement in the ring-expansion activity of scDACS was achieved, indicating the importance of the type of amino acid residues in these two positions.

However, combining T308N and T308R mutations with H310R mutation were not sufficient to confer a level of ring-expansion activity similar to that of scDAOCS.

Since H310L mutation resulted in a much higher ring-expansion activity than H310R 175

mutation, i.e., 18% vs 2% relative ring-expansion activity, respectively, it will be worthwhile to investigate the outcomes of combining this mutation with T308N and

T308R mutations. Since residue W82 of AchDAOC/DACS has been reported to be important for substrate selectivity, it is very likely that the corresponding residue A83 in scDACS is also involved in substrate selectivity. This is further supported by distinct serine and alanine conservedness in the annotated DAOCSs and DACSs, respectively (Lloyd et al., 2004). However, residues E81-N90 were highly disordered and could not be resolved in the crystal structure of scDAOCS (Valegard et al., 2004).

Hence, the involvement of residue S82, which is the corresponding residue of W82 in

AchDAOC/DACS and that of A83 in scDACS, in the ring-expansion of scDAOCS could not be postulated.

Table 8.3. Relative ring-expansion and hydroxylation activities of wild-type and mutant AchDAOC/DACSs using penicillin G and DAOC as respective substrates (Lloyd et al., 2004).

Relative ring-expansion Relative hydroxylation Enzyme activity (%) activity (%) Wild-type 100 100 W82A 5.5 71 W82S 44 18 N305L 107 85 M306I 59 <2 ∆310 203 50 ∆328 64 45 ∆310/M306I 185 3 ∆310/N305L/M306I 370 2

176

Figure 8.7. Superimposition of three crystal structures of scDAOCS complexed with ampicillin (brown) (1UNB), penicillin G (orange) (1UOB) and DAOC (gold) (1UOG).

Figure 8.8. Spatial orientation of the methyl group (circled) at the C7 position of dihydrothiazine ring of DAOC. It is approximately 4.5Å away from the iron centre of scDAOCS.

177

Chapter 9: Bioactivities of expanded cephalosporin products

In this project, various scDAOCS mutants were demonstrated to convert penicillin G, ampicillin, carbenicillin and phenethicillin to penicillinase-resistant cephalosporin products. They were also demonstrated to produce higher yields of the respective cephalosporin products than the wild-type enzyme. This allows further assessment of these cephalosporins for their bioactivities against a wider range of bacteria and for other potential uses. The maximum concentration of cephalosporin products formed by selected C281YI305M double mutant scDAOCS in the enzyme reaction was determined using HPLC. Since the dihydrothiazine ring of cephalosporins absorbed maximally at 260 nm of ultra-violet light (Dubus et al.,

2001; Baldwin and Crabbe, 1987), quantification of the cephalosporin products was based on the detection of their dihydrothiazine rings. The appearance of a new peak in the HPLC chromatogram of an assay sample compared to a control sample indicated the presence of the cephalosporin product (Figure 5.1). The area under the peak was used to determine the concentration of the cephalosporin product. A standard curve was generated using a range of concentration of cephalexin. The concentrations of cephalosporin products yielded from the enzymatic conversions of penicillin G, ampicillin, carbenicillin and phenethicillin were calculated using this standard curve.

The maximum concentrations of cephalosporin products produced in the ring- expansion assays containing penicillin G, ampicillin, carbenicillin and phenethicillin were approximately 1 mM, 1 mM, 0.4 mM and 0.8 mM, respectively.

178

9.1. Antimicrobial spectra of expanded cephalosporin products

The hole-plate bioassay used for the determination of the enzyme activity of

DAOCS was employed to evaluate the inhibition profile of the cephalosporin products from the DAOCS ring-expansion assays. However, a modification was made where the test microorganisms were streaked onto the agar plate instead of seeding into it (Figure 3.1). This allows a higher number of test microorganisms to be evaluated per plate. Since DIFCO™ Penase was added to the bioassay plates to specifically degrade residual penicillin substrates, the zone of growth inhibition will reflect the bioactivity of the cephalosporin products. In order to compare the efficacies of the penicillins and their respective expanded cephalosporin products, equimolar concentrations of the respective penicillin substrates were also tested but

DIFCO™ Penase was omitted in those bioassay plates. Since cephalexin is the cephalosporin moiety of ampicillin, the commercial cephalexin was used as a control.

The test microorganisms used in the bioassay were Micrococcus luteus ATCC 381,

Staphylococcus aureus ATCC 9144, Enterococcus faecalis ATCC 29212, Escherichia coli Ess strain, E. coli K12, Enterobacter aerogenes and Salmonella cholerasius

ATCC 51741.

Figure 9.1 shows the average zone of growth inhibition against the test microorganisms for ampicillin, its cephalosporin product and cephalexin. The inhibition profile of the cephalosporin product upon ampicillin conversion was identical to that of commercial cephalexin. However, both were not as bioactive as ampicillin against M. luteus ATCC 381, E. coli Ess strain, S. aureus ATCC 9144, E. coli K12 and S. cholerasius ATCC 51741, i.e., they showed smaller zones of inhibition than ampicillin. Conversely, both were able to inhibit E. aerogenes but

179

ampicillin was not able to. Interestingly, only ampicillin was able to inhibit E. faecalis

ATCC 29212.

The inhibitory profile of penicillin G included M. luteus ATCC 381, E. coli

Ess strain, S. aureus ATCC 9144, E. coli K12, S. cholerasius ATCC 51741 and E. faecalis ATCC 29212. In contrast, its cephalosporin product was only able to inhibit

M. luteus ATCC 381, E. coli Ess strain and S. aureus ATCC 9144, and at a lower level than penicillin G. This was similarly observed for carbenicillin, phenethicillin and their respective cephalosporin products. The inhibitory profile of carbenicillin included five bacteria, namely M. luteus ATCC 381, E. coli Ess strain, S. aureus

ATCC 9144, E. coli K12 and S. cholerasius ATCC 51741, while that of its cephalosporin product only included the first three bacterial species. The inhibitory profile of phenethicillin was the narrowest among the penicillins, i.e., it included only

M. luteus ATCC 381, E. coli Ess strain, S. aureus ATCC 9144, and E. faecalis ATCC

29212. The inhibitory profile of its cephalosporin product was similar to that of penicillin G- and carbenicillin-converted cephalosporin products, which included only the first three bacterial species.

180

(A) (B)

25.0 30.0

18.8 25.0 22.2 20.0 0.5mM ampicillin 16.3 0.5mM penicillin G 15.5 0.5mM ampicillin converted product 16.0 20.0 19.0 17.8 0.5mM penicillin G converted product 14.3 0.5mM cephalexin 15.0 11.8 11.0 15.0 12.0 11.2 12.0 10.8 10.0 10.2 7.7 10.0 7.5 7.7 6.3 6.0 6.2 6.3 5.8 5.3 5.0 5.0 5.0 3.3 Average zone of inhibition (mm)

Average zone of inhibition (mm) 2.7

0.0 0.0 M. luteus E. coli ESS S. aureus ATCC E. coli K12 E. aerogenes S. cholerasius E. faecalis M. luteus E. coli ESS S. aureus ATCC E. coli K12 E. aerogenes S. cholerasius E. faecalis ATCC 9144 ATCC 51741 ATCC 29212 9144 ATCC 51741 29212

(C) (D) 18.0 20.0 15.0 18.0 16.0 13.3 15.5 16.0 14.0 12.8 12.8 0.2mM carbenicillin 14.0 13.0 12.0 0.2mM carbenicillin converted product 0.4mM phenethicillin 12.0 11.0 0.4mM phenethicillin converted product 10.0 8.5 10.0 7.7 7.2 8.0 6.3 7.7 7.3 8.0 6.0 5.3 6.0 4.0 2.7 4.0 Average zone of inhibition (mm) Average zone of inhibition (mm) 2.0 2.0

0.0 0.0 M. luteus E. coli ESS S. aureus ATCC E. coli K12 E. aerogenes S. cholerasius E. faecalis ATCC M. luteus E. coli ESS S. aureus ATCC E. coli K12 E. aerogenes S. cholerasius E. faecalis ATCC 9144 ATCC 51741 29212 9144 ATCC 51741 29212

Figure 9.1. Antimicrobial spectra of (A) ampicillin and cephalexin, (B) penicillin G, (C) carbenicillin, (D) phenethicillin and their respective ring-expanded cephalosporin products. These compounds were tested against Micrococcus luteus ATCC 381, Staphylococcus aureus ATCC 9144, Enterococcus faecalis ATCC 29212, Escherichia coli Ess strain, E. coli K12, Enterobacter aerogenes and Salmonella cholerasius ATCC 51741.

181

9.2. Efficacies of penicillin substrates and their respective expanded cephalosporin moieties

The hole-plate bioassay results suggest that the antimicrobial spectra of the cephalosporin products are similar or slightly smaller than those of their respective penicillin substrates. However, cephalosporins are generally penicillinase-resistant and more efficacious than the penicillins. The results obtained from the hole-plate bioassays might be due to the relatively small range of microorganisms tested and the unstandardised inoculum size of each test microorganism. In order to illustrate the efficacies of the cephalosporin products against a penicillin-resistant bacterium, a colony-forming unit (CFU) assay, which employs a standardised inoculum size, was used to assess the efficacies of the penicillins and their respective cephalosporin products in inhibiting a penicillinase-producing E. coli.

The pGEX-6P-1 plasmid contains an ampicillin-resistance factor, which is constitutively expressed. This ampicillin-resistance factor encodes a TEM-1 β- lactamase, which has been demonstrated to be a penicillinase (Bush et al., 1995). The pGEX-6P-1 plasmid was transformed into E. coli BL21 (DE3) to mimic the gain of penicillin-resistant phenotype. The transformed ampicillin-resistant E. coli BL21

(DE3) strain (E. coli BL21-AmpR) was cultivated using LB broth containing 100

µg/mL ampicillin for 24 hours. Thereafter, enumeration of the bacterial culture was done to estimate the viable bacterial count using dilution plating method before the

CFU assay was performed. The bacterial culture was stored at 4°C prior to CFU assay to retard further proliferation. Upon obtaining the viable bacterial count of the transformed and untransformed E. coli BL21 (DE3) cultures, CFU assay was done according to the procedure described in Chapter 3.2.8.2, where the bacterial cultures were diluted in sterile Mueller-Hinton (II) broth to obtain approximately 100 CFU per 182

assay condition. The efficacies of the antibiotics were compared based on the reaction mixture before and after conversion by scDAOCS mutant enzyme. To ensure that the reagents for ring-expansion assay do not contribute to the results of the CFU assay, they were included as a control without the addition of any antibiotics. Thereafter, the viable bacterial count of this control was used to determine the percentage of survival upon exposure to respective β-lactam antibiotics.

Figure 9.2A shows the CFU assay results of penicillin G, ampicillin, carbenicillin, phenenthicillin and their respective cephalosporin products against untransformed E. coli BL21 (DE3) cells. These results were comparable to those obtained for E. coli K12 in hole-plate bioassays, i.e., penicillin G, ampicillin and carbenicillin were able to inhibit the growth of E. coli K12 to various extents but not phenethicillin. This is expected since E. coli BL21 (DE3) strain is phylogenetically related to E. coli K12 (Jeong et al., 2009; Studier et al., 2009). Similarly, the results also showed that the cephalosporin products of penicillin G, ampicillin and carbenicillin were able to inhibit the growth of E. coli BL21 (DE3) cells but not that of phenethicillin in the CFU assays.

Figure 9.2B shows the CFU assay results of penicillin G, ampicillin, carbenicillin, phenethicillin and their respective cephalosporin products against E. coli

BL21-AmpR cells. The results showed that the cephalosporin products of ampicillin and carbenicillin were able to inhibit the growth of E. coli BL21-AmpR cells but not their respective penicillin moieties. Interestingly, penicillin G was able to inhibit approximately 80% of the E. coli BL21-AmpR population, while its cephalosporin product showed 99.9% inhibition. Similarly, phenethicillin and its cephalosporin product did not show any growth inhibition of E. coli BL21-AmpR cells.

183

Figure 9.2. Efficacies of penicillin G, ampicillin, carbenicillin, phenethicillin and their respective ring-expanded cephalosporin products against (A) untransformed E. coli BL21 (DE3) and (B) transformed E. coli BL21 (DE3) expressing ampicillin- resistance gene. The number of CFUs for each antibiotic sample was compared to the negative control, which contained the test organism and the DAOCS ring-expansion assay reagents except the penicillin substrates, and was expressed in percentage.

184

9.3. Assessing additional properties conferred on expanded cephalosporin products: inhibition of elastase

Calbiochem® InnoZyme™ Human Neutrophil Elastase Immunocapture activity assay kit was used to assess the potential inhibitory activity of the cephalosporin products. The plausible elastase inhibitory effect of the cephalosporin products were compared based on the reaction mixture before and after conversion by scDAOCS mutant enzyme, assuming that the amount of cephalosporin product formed is sufficiently potent to demonstrate some level of elastase inhibition. Due to the nature of the assay set-up, there was a 10X dilution to the DAOCS assay samples.

Hence, the final concentration of the penicillins in the elastase assay was approximately 1 mM. The final concentrations of the cephalosporin products of penicillin G, ampicillin, carbenicillin and phenethicillin in the elastase assays were

100 µM, 100 µM, 40 µM and 80 µM, respectively.

Figure 9.3 shows the results of the elastase inhibitory assay of PMSF, DAOCS ring-expansion assay reagents, penicillin G, ampicillin, carbenicillin, phenethicillin and their respective cephalosporin products. The uninhibited elastase activity was determined to be 98.3 nM/min. As expected, PMSF almost inhibited the activity of elastase at 5 mM concentration. However, none of the penicillins and their respective cephalosporin products showed any inhibitory activity against elastase.

185

Figure 9.3. Elastase inhibitory assay of penicillins and their respective expanded cephalosporin moieties.

186

9.4. Potential applications of expanded cephalosporin products

A plethora of semi-synthetic penicillins and cephalosporins has been generated over the last six decades. Studies on these penicillins have validated the efficacies of the aminopenicillins, e.g. ampicillin and amoxicillin, as antibacterials.

This is, in part, due to the presence of the amino group in the side chain of the aminopenicillins. This importance was conserved in the structure of many latter semi- synthetic cephalosporins, e.g. cephalexin, cefadroxil, cefuroxime, cefotaxime and ceftazidime, which have been classified into generations instead of the earlier structural classification of penicillin (Marshall and Blair, 1999; Wright, 1999; Hansch et al., 1990). Interestingly, the cephalosporin moieties of other penicillins, e.g. the carboxypenicillins (carbenicillin and ticarcillin), were not generated or reported.

Hence, it is vital to validate the bioactivities of these cephalosporins as they could be potential chemotherapeutics.

Although the wild-type scDAOCS is able to accept a range of penicillins as substrates (Chin and Sim, 2002; Cho et al., 1998), its limitation is the low activity in the ring-expansion of hydrophobic penicillin substrates, resulting in a low amount of product formed. This restricted further analysis of these products for their antimicrobial spectra. Although earlier studies on the structure-function relationship of β-lactam antibiotics have illustrated the importance of the amino group in the α- position of the side chain of either a penicillin or a cephalosporin in broadening its antimicrobial spectrum (Wright, 1999; Hansch et al., 1990), other functional groups, e.g., a carboxyl group, have been shown to be effective against other bacteria

(Rolinson, 1998; Tan and File, 1995). Furthermore, there are various factors affecting the efficacy of an antibiotic. For example, in a bacterium, there are many different types of PBPs and β-lactamases, which have different affinities for the same antibiotic 187

(Sauvage et al., 2008). Hence, it is difficult, if not impossible, to predict the efficacy of an antibiotic. It is still inevitable that the antimicrobial spectra of these cephalosporin products need to be determined empirically.

In this project, several mutations have significantly improved the ring- expansion activity of scDAOCS in converting penicillin G, ampicillin, carbenicillin and phenethicillin to their respective cephalosporin products. HPLC analysis of ampicillin conversions has also verified the production of cephalexin, which is the generic name of the cephalosporin moiety of ampicillin, and suggested that no further modification was done to the product even when the enzyme has been modified.

Although the cephalosporin moieties of penicillin G, carbenicillin and phenethicillin were not available for authenticating the in vitro ring-expanded products of these three penicillin substrates, the same retention time of the products formed by the mutant and wild-type enzymes suggested the same conclusion. Therefore, only

C281YI305M double mutant scDAOCS was selected to produce the cephalosporin products of penicillin G, ampicillin, carbenicillin and phenethicillin. Since cephalexin is available commercially, the results of the enzyme-converted cephalexin were used to compare with that of the standard cephalexin to assess the precision of HPLC in cephalosporin product quantification and the hole-plate bioassay method for screening of antimicrobial spectrum. As reflected by the hole-plate bioassay results, the antimicrobial spectra of both cephalosporins were the same, which supports the reliability of the antimicrobial screening results of other cephalosporin products.

The antimicrobial spectra of the cephalosporin moieties of penicillin G, carbenicillin and phenethicillin were narrower than that of cephalexin. They were also smaller than that of the respective penicillins. This is expected since a cephalosporin is apparently not as active as its penicillin moiety against the same organism probably 188

due to its bulkier dihydrothiazine ring, which affects its binding with the PBPs

(Hansch et al., 1990). For example, it has been reported that although cephalexin was inhibitory to E. coli, the affinities of various E. coli PBPs for cephalexin were also lower than ampicillin, resulting in higher MIC (Table 9.1) (Curtis et al., 1979). Other modification has to be done to cephalosporin to render it more efficacious to overcome the limitation of its dihydrothiazine ring. Hence, it is likely that the cephalosporin moieties of penicillin G, carbenicillin and phenethicillin are not ideal antibiotics and need further structural modification in order to be efficacious against a wide range of bacterial pathogens.

Nevertheless, a cephalosporin is generally considered more efficacious than penicillin because of its natural resistance to penicillinase. Hence, it is often preferred over penicillin as a candidate for synthesising newer and more efficacious β-lactam antibiotic to overcome the emergence of resistant pathogens. In order to verify that the cephalosporin products are indeed penicillinase-resistant, a CFU assay using penicillin-susceptible and penicillin-resistant E. coli was employed. The results showed that while both the penicillins and their respective cephalosporin products were effective against susceptible E. coli, the cephalosporin products were able to inhibit the bacterium, but not the penicillins, when the organism expressed a penicillinase. This supports the notion that the cephalosporins are more efficacious than the penicillin moieties.

For the past 60 years, the applications of β-lactams are mainly as chemotherapeutics or for research purposes. However, in the recent decades, new applications for β-lactams were discovered (Eisenstein, 2005; Rothstein et al., 2005).

For example, in the recent effort of screening for an inducer of glutamate transporter expression in mice, a number of semi-synthetic β-lactams was found to induce the 189

expression of this transporter and prevent neurotoxicity (Rothstein et al., 2005).

However, the molecular mechanism involved in inducing the expression of glutamate transporter is still not clear.

Since β-lactams are acylating inhibitors of bacterial PBPs, which are classified as serine proteases, it has been suggested that they could also inhibit mammalian serine proteases (Veinberg et al., 2003; Davies et al., 1991; Doherty et al., 1986). By modifying the C7α position of the cephalosporin nucleus, a cephalosporin could be made potent to inhibit human leukocyte elastase (Alpegiani et al., 1994; Knight et al.,

1992; Doherty et al., 1986). Since the cephalosporin moieties of penicillin G, ampicillin, carbenicillin and phenethicillin are C7β isomers, which have been reported to be weak inhibitors or inactive (Knight et al., 1992; Doherty et al., 1986), it is expected that the elastase inhibition assays showed no potency in any of the cephalosporin products tested. Nevertheless, these cephalosporins may have applications other than as antibacterials, which remain to be uncovered.

Table 9.1. Affinities of various penicillins and cephalosporins for the PBPs of E. coli DC0 (Curtis et al., 1979).

I50 binding concentration (µg/mL) for: MIC (µg/mL) β-Lactams PBP1a PBP1b PBP2 PBP3 PBP4 PBP5 PBP6 DC0 DC2* Penicillin G 0.5 3.0 0.8 0.9 1.0 24 19 16 1.6 Penicillin V 1.4 9.5 7.7 10.5 14 50 17 50 12.5 Carbenicillin 2.1 5.0 4.0 2.1 3.5 130 120 2 1 Amoxicillin 0.7 2.2 0.9 4.1 2.5 110 9 3.2 1.6 Ampicillin 1.4 3.9 0.7 0.9 2.0 140 9 3.2 0.5 Cephalexin 4 240 >250 8 30 >250 >250 8 4 Cefotaxime 0.05 0.7 5 <0.05 30 >50 >50 0.08 0.01 Cephalothin <0.25 16 37 1 60 125 130 1.6 0.8 Cephaloridine 0.25 2.5 50 8 17 >250 >250 2 2 *Strain DC2 is a permeability mutant of E. coli DC0. The penicillin and cephalosporin that share the same C7 side chain are underlined, i.e., ampicillin and cephalexin, respectively.

190

Chapter 10: Concluding remarks and future directions

The C-terminal residues of scDAOCS have been repeatedly demonstrated to be amenable to modification in order to improve the ring-expansion activity of scDAOCS. This study has provided a further understanding of the mechanisms behind these improvements. Interestingly, the effect of combining the single mutations at the

C-terminus is less predictable when the two mutation sites are too close to each other.

N304XR306L double mutants of scDAOCS did not show additive or synergistic effect when compared to the respective single mutants for ampicillin, carbenicillin and phenethicillin conversions. For penicillin G conversion, the enzymatic activities of the double mutants were even lower than that of the wild-type enzyme. It is very likely that this is due to the opposing effect of the residues at these two positions, i.e., the interaction forces at position 304 are towards the substrate binding site, while those at position 306 are away from the substrate binding site. Since residue I305 is also interacting with the penicillin substrate, by combining I305M single mutation with any of the N304X single mutations, further enhancement in the conversion activity of scDAOCS could be attained.

scDAOCS has been modified and achieved approximately 4- to 10-fold increments in the specific activity of various penicillin conversions and almost 100- fold improvement in its catalytic efficiency in ampicillin conversion. However, it is still at least a magnitude lower than the conversion of the native substrate penicillin N

(Table 10.1). Since the effect of a mutation is independent of another when they are at least 10Å apart, by combining those improved mutations that conferred at least 2-fold 191

improvement in the catalytic efficiency of the enzyme, it is likely that the catalytic efficiency of scDAOCS can be further improved. In addition, since L277Q and

C281Y single mutations in the small sub-domain could enhance the catalytic efficiency of the enzyme, it could be postulated that by combining L277Q and C281Y mutations, a synergistic effect could be achieved. Eventually, the ring-expansion activity of scDAOCS for industrially preferred penicillin substrates could be significantly enhanced by strategic combination of all the C-terminal and sub-domain mutations using a rational approach.

Although the mechanism of substrate recognition of scDAOCS for hydrophobic penicillin substrates has been illuminated, studies on scDACS have elicited more questions with respect to the differences in their substrate selectivity, leading to different enzymatic reactions. Upon obtaining the crystal structures of the enzyme-substrate complexes of the various scDAOCS and scDACS mutants generated in this study, the exact mechanism could be elucidated. By understanding the mechanism of substrate selectivity of DAOCS, DACS and DAOC/DACS, the maximum potential or limitation of these enzymes could be realised, which is important for industrial application.

In this study, some of the mutant scDAOCSs are more inclined to accept ampicillin than penicillin G, and thus, they could be exploited for efficient and cheaper industrial production of cephalexin. Since amoxicillin is an analogue of ampicillin, it is also appealing to investigate the potential of these mutant scDAOCSs in ring-expanding amoxicillin to cefadroxil for industrial production.

192

Table 10.1. Kinetic parameters of wild-type and mutant scDAOCSs in penicillin N, penicillin G and ampicillin conversions.

-1 -1 -1 Enzyme Substrate Km (mM) kcat (s ) kcat/Km (M s ) References

H H2N N S CH3

COOH O N CH3 Wild-type O 0.014 0.307 22,000 Wei et al., 2005 COOH Penicillin N H N S Wei et al., 2005; Wild-type CH3 2.58 0.0453 18 O N CH3 Hsu et al., 2004 O COOH FF8* Penicillin G 0.014 0.0297 2,121 Hsu et al., 2004

NH2 H Wild-type N S 6.94 0.059 8.6

O N O C281YN304R COOH 0.05 0.043 865.2 This study Ampicillin C281YI305M 0.21 0.122 569.9 *Recombinant DAOCS derived from gene shuffling

193

References

Abraham EP, Newton GGF. 1961. The structure of cephalosporin C. Biochem J 79: 377-393.

Adrio JL, Demain AL. 2002. Improvements in the formation of cephalosporins from penicillin G and other penicillins by bioconversion. Org Process Res Dev. 6: 427-433.

Adrio JL, Hintermann GA, Demain AL, Piret JM. 2002. Construction of hybrid bacterial deacetoxycephalosporin C synthases (expandases) by in vivo homeologous recombination. Enzyme Microb Technol. 31: 932-940.

Aharonowitz Y, Demain AL. 1978. Carbon catabolite regulation of cephalosporin production in Streptomyces clavuligerus. Antimicrob Agents Chemother. 14: 159- 164.

Allen MJ, Fang TY, Li Y, Liu HL, Chen HM, Coutinho P, Honzatko R, Ford C. 2003. Protein engineering of glucoamylase to increase pH optimum, substrate specificity and thermostability. Patent no.: 6,537,792 (USPTO).

Almog O, Gallagher DT, Ladner JE, Strausberg S, Alexander P, Bryan P, Gilliland GL. 2002. Structural basis of thermostability. Analysis of stabilizing mutations in subtilisin BPN'. J Biol Chem. 277(30): 27553-27558.

Almog O, Gallagher T, Tordova M, Hoskins J, Bryan P, Gilliland GL. 1998. Crystal structure of calcium-independent subtilisin BPN' with restored thermal stability folded without the prodomain. Proteins. 31(1): 21-32.

Alpegiani M, Bissolino P, Corigli R, Del Nero S, Perrone E, Rizzo V, Sacchi N, Cassinelli G, Franceschi G, Baici A. 1994. Cephem sulfones as inactivators of human leukocyte elastase. 5. 7 alpha-Methoxy- and 7 alpha-chloro-1,1-dioxocephem 4- ketones. J Med Chem. 37(23): 4003-4019.

Alvarez E, Meesschaert B, Montenegro E, Gutierrez S, Diez B, Barredo JL, Martin JF. 1993. The isopenicillin-N acyltransferase of Penicillium chrysogenum has isopenicillin-N amidohydrolase, 6-aminopenicillanic acid acyltransferase and penicillin amidase activities, all of which are encoded by the single penDE gene. Eur J Biochem. 215: 323-332.

Andersson I, van Scheltinga ACT, Valegard K. 2001. Towards new beta-lactam antibiotics. Cell Mol Life Sci. 58: 1897-1906.

Aoki H, Okuhara M. 1980. Natural beta-lactam antibiotics. Annu Rev Microbiol. 34: 159-181.

Appelbaum PC. 2006. MRSA--the tip of the iceberg. Clin Microbiol Infect. 12(Suppl 2): 3-10.

194

Arnold K, Bordoli L, Kopp J, and Schwede T. 2006. The SWISS-MODEL Workspace: A web-based environment for protein structure homology modelling. Bioinformatics. 22: 195-201.

Arroyo M, de la Mata I, Acebal C, Castillón MP. 2003. Biotechnological applications of penicillin acylases: state-of-the-art. Appl Microbiol Biotechnol. 60(5): 507-514.

Baker BJ, Dotzlaf JE, Yeh WK. 1991. Deacetoxycephalosporin C hydroxylase of Streptomyces clavuligerus. Purification, characterization, bifunctionality, and evolutionary implication. J Biol Chem. 266(8): 5087-5093.

Baldwin JE, Crabbe MJ. 1987. A spectrophotometric assay for deacetoxycephalosporin C synthase. FEBS Lett. 214(2): 357-361.

Barber MS, Giesecke U, Reichert A, Minas W. 2004. Industrial enzymatic production of cephalosporin-based β-lactams. Adv Biochem Eng Biotechnol. 88: 179-215.

Barbour A, Schmidt S, Rand KH, Derendorf H. 2009. Ceftobiprole: a novel cephalosporin with activity against Gram-positive and Gram-negative pathogens, including methicillin-resistant Staphylococcus aureus (MRSA). Int J Antimicrob Agents. 34: 1-7.

Barends TR, Yoshida H, Dijkstra BW. 2004. Three-dimensional structures of enzymes useful for beta-lactam antibiotic production. Curr Opin Biotechnol. 15: 356- 363.

Barredo JL, van Solingen P, Diez B, Alvarez E, Cantoral JM, Kattevilder A, Smaal EB, Groenen MA, Veenstra AE, Martin JF. 1989. Cloning and characterization of the acyl-coenzyme A: 6-aminopenicillanic-acid-acyltransferase gene of Penicillium chrysogenum. Gene. 83: 291-300.

BeBoer C, Dietz A. 1976. The description and antibiotic production of Streptomyces hygroscopicus var. Geldanus. J Antibiot (Tokyo). 29(11): 1182-1188.

Beghi E, Bendotti C, Mennini T. 2005. Merits of a new drug trial for ALS? Science. 308: 632-633.

Bentley SD, Chater KF, Cerdeño-Tárraga AM, Challis GL, Thomson NR, James KD, Harris DE, Quail MA, Kieser H, Harper D, Bateman A, Brown S, Chandra G, Chen CW, Collins M, Cronin A, Fraser A, Goble A, Hidalgo J, Hornsby T, Howarth S, Huang CH, Kieser T, Larke L, Murphy L, Oliver K, O'Neil S, Rabbinowitsch E, Rajandream MA, Rutherford K, Rutter S, Seeger K, Saunders D, Sharp S, Squares R, Squares S, Taylor K, Warren T, Wietzorrek A, Woodward J, Barrell BG, Parkhill J, Hopwood DA. 2002. Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature. 417(6885): 141-147.

Berdy J. 2005. Bioactive microbial metabolites. J Antibiot (Tokyo). 58(1): 1-26.

Bojsen K, Svendsen A, Fuglsang CC, Anant Patkar S, Borch K, Vind J, Petri A, Glad SS, Budolfsen G. 2007. Lipolytic enzyme variants. Patent no.: 7,312,062 (USPTO).

195

Bornscheuer UT, Pohl M. 2001. Improved biocatalysts by directed evolution and rational protein design. Curr Opin Chem Biol. 5(2): 137-143.

Bottcher D, Bornscheuer UT. 2010. Protein engineering of microbial enzymes. Curr Opin Microbiol. 13(3): 274-282.

Boucher HW, Talbot GH, Bradley JS, Edwards JE, Gilbert D, Rice LB, Scheld M, Spellberg B, Bartlett J. 2009. Bad bugs, no drugs: no ESKAPE! An update from the Infectious Diseases Society of America. Clin Infect Dis. 48(1): 1-12.

Britsch L. 1990. Purification and characterization of flavone synthase I, a 2- oxoglutarate-dependent desaturase. Arch Biochem Biophys. 282(1): 152-160.

Bruggink A, Straathof AJ, van der Wielen LA. 2003. A ‘fine’ chemical industry for life science products: green solutions to chemical challenges. Adv Biochem Eng Biotechnol. 80: 69-113.

Bryan P, Alexander P, Strausberg S, Schwarz F, Lan W, Gilliland G, Gallagher DT. 1992. Energetics of folding subtilisin BPN'. Biochemistry. 31(21): 4937-4945.

Bryan PN. 2000. Protein engineering of subtilisin. Biochim Biophys Acta. 1543(2): 203-222.

Bryan PN. 2003. Calcium free subtilisin mutants. Patent no.: 6,541,234 (USPTO).

Bugg TD, Braddick D, Dowson CG, Roper DI. 2011. Bacterial cell wall assembly: still an attractive antibacterial target. Trends Biotechnol. 29(4): 167-173.

Bush K, Jacoby GA, Medeiros AA. 1995. A functional classification scheme for beta- lactamases and its correlation with molecular structure. Antimicrob Agents Chemother. 39(6): 1211-1233.

Bush K. 2010. Alarming β-lactamase-mediated resistance in multidrug-resistant Enterobacteriaceae. Curr Opin Microbiol. 13(5): 558-564.

Bush LM, Johnson CC. 2000. Ureidopenicillins and beta-lactam/beta-lactamase inhibitor combinations. Infect Dis Clin North Am. 14: 409-433, ix.

Calabrese EJ, Blain R. 2005. The occurrence of hormetic dose responses in the toxicological literature, the hormesis database: an overview. Toxicol Appl Pharmacol. 202(3): 289-301.

Campanella JJ, Bitincka L, Smalley J. 2003. MatGAT: an application that generates similarity/identity matrices using protein or DNA sequences. BMC Bioinformatics. 4: 29.

Campos Muñiz C, Zelaya TEC, Esquivel GR, Fernández FJ. 2007. Penicillin and cephalosporin production: a historical perspective. Rev Latinoam Microbiol. 49: 88- 98.

196

Carballido-Lopez R, Formstone A. 2007. Shape determination in Bacillus subtilis. Curr Opin Microbiol. 10: 611-616.

Castro JM, Liras P, Laiz L, Cortes J, Martin JF. 1988. Purification and characterization of the isopenicillin N synthase of Streptomyces lactamdurans. J Gen Microbiol. 134: 133-141.

Chain E. 1972. Thirty years of penicillin therapy. J R Coll Physicians Lond. 6: 103- 131.

Chaput JC, Woodbury NW, Stearns LA, Williams BA. 2008. Creating protein biocatalysts as tools for future industrial applications. Expert Opin Biol Ther. 8(8): 1087-1098.

Chauvette RR, Spencer JL, Mueller RA, Roeske RW, Morin RB, Ryan CW, Jackson BG, Pioch RP, Flynn EH, Vanheyningen E, Lavagnino ER. 1962. Chemistry of cephalosporin antibiotics. 2. Preparation of a new class of antibiotics and relation of structure to activity. J Am Chem Soc. 84: 3401.

Chen L, Coutinho PM, Nikolov Z, Ford C. 1996. Effect of replacing helical glycine residues with alanines on reversible and irreversible stability and production of Aspergillus awamori glucoamylase. Protein Eng. 9: 499-505.

Chen Y, Zhang W, Shi Q, Hesek D, Lee M, Mobashery S, Shoichet BK. 2009. Crystal structures of penicillin-binding protein 6 from Escherichia coli. J Am Chem Soc. 131: 14345-14354.

Chesnel L, Zapun A, Mouz N, Dideberg O, Vernet T. 2002. Increase of the deacylation rate of PBP2x from Streptococcus pneumoniae by single point mutations mimicking the class A beta-lactamases. Eur J Biochem. 269: 1678-1683.

Chin HS, Goo KS, Sim TS. 2004. A complete library of amino acid alterations at N304 in Streptomyces clavuligerus deacetoxycephalosporin C synthase elucidates the basis for enhanced penicillin analogue conversion. Appl Environ Microbiol. 70(1): 607-609.

Chin HS, Sim J, Seah KI, Sim TS. 2003. Deacetoxycephalosporin C synthase isozymes exhibit diverse catalytic activity and substrate specificity. FEMS Microbiol Lett. 218(2): 251-257.

Chin HS, Sim J, Sim TS. 2001. Mutation of N304 to leucine in Streptomyces clavuligerus deacetoxycephalosporin C synthase creates an enzyme with increased penicillin analogue conversion. Biochem Biophys Res Commun. 287(2):507-513.

Chin HS, Sim TS. 2002. C-terminus modification of Streptomyces clavuligerus deacetoxycephalosporin C synthase improves catalysis with an expanded substrate specificity. Biochem Biophys Res Commun. 295(1): 55-61.

Chin YW, Balunas MJ, Chai HB, Kinghorn AD. 2006. Drug discovery from natural sources. AAPS J. 8(2): E239-253.

197

Cho H, Adrio JL, Luengo JM, Wolfe S, Ocran S, Hintermann G, Piret JM, Demain AL. 1998. Elucidation of conditions allowing conversion of penicillin G and other penicillins to deacetoxycephalosporins by resting cells and extracts of Streptomyces clavuligerus NP1. Proc Natl Acad Sci USA. 95: 11544-11548.

Chua CS, Biermann D, Goo KS, Sim TS. 2008. Elucidation of active site residues of Arabidopsis thaliana flavonol synthase provides a molecular platform for engineering flavonols. Phytochemistry. 69(1): 66-75.

Cicero G, Carbonera C, Valegard K, Hajdu J, Andersson I, Ranghino G. 2007. Study of the oxidative half-reaction catalyzed by a non-heme ferrous catalytic center by means of structural and computational methodologies. Int J Quantum Chem. 107: 1514-1522.

Cirino PC, Mayer KM, Umeno D. 2003. Generating mutant libraries using error- prone PCR. Methods Mol Biol. 231: 3-9.

Clifton IJ, Doan LX, Sleeman MC, Topf M, Suzuki H, Wilmouth RC, Schofield CJ. 2003. Crystal structure of carbapenem synthase (CarC). J Biol Chem. 278: 20843- 20850.

Coates A, Hu Y, Bax R, Page C. 2002. The future challenges facing the development of new antimicrobial drugs. Nat Rev Drug Discov. 1(11): 895-910.

Collignon P, Powers JH, Chiller TM, Aidara-Kane A, Aarestrup FM. 2009. World Health Organization ranking of antimicrobials according to their importance in human medicine: A critical step for developing risk management strategies for the use of antimicrobials in food production animals. Clin Infect Dis. 49(1): 132-141.

Combet C, Blanchet C, Geourjon C and Deléage G. 2000. NPS@: Network Protein Sequence Analysis. TIBS. 25(3): 147-150.

Contreras-Martel C, Dahout-Gonzalez C, Martins Ados S, Kotnik M, Dessen A. 2009. PBP active site flexibility as the key mechanism for beta-lactam resistance in pneumococci. J Mol Biol. 387: 899-909.

Coque JJ, Enguita FJ, Cardoza RE, Martín JF, Liras P. 1996. Characterization of the cefF gene of Nocardia lactamdurans encoding a 3'-methylcephem hydroxylase different from the 7-cephem hydroxylase. Appl Microbiol Biotechnol. 44(5): 605-609.

Coque JJ, Enguita FJ, Martin JF, Liras P. 1995a. A two-protein component 7 alpha- cephem-methoxylase encoded by two genes of the cephamycin C cluster converts cephalosporin C to 7-methoxycephalosporin C. J Bacteriol. 177: 2230-2235.

Coque JJ, Martin JF, Calzada JG, Liras P. 1991. The cephamycin biosynthetic genes pcbAB, encoding a large multidomain peptide synthetase, and pcbC of Nocardia lactamdurans are clustered together in an organization different from the same genes in Acremonium chrysogenum and Penicillium chrysogenum. Mol Microbiol. 5: 1125- 1133.

198

Coque JJ, Martín JF, Liras P. 1993. Characterization and expression in Streptomyces lividans of cefD and cefE genes from Nocardia lactamdurans: the organization of the cephamycin gene cluster differs from that in Streptomyces clavuligerus. Mol Gen Genet. 236: 453-458.

Coque JJR, Perezllarena FJ, Enguita FJ, Fuente JL, Martin JF, Liras P. 1995b. Characterization of the cmcH genes of Nocardia lactamdurans and Streptomyces clavuligerus encoding a functional 3'-hydroxymethylcephem O-carbamoyltransferase for cephamycin biosynthesis. Gene. 162: 21-27.

Cortes J, Martin JF, Castro JM, Laiz L, Liras P. 1987. Purification and characterization of a 2-oxoglutarate-linked ATP-independent deacetoxycephalosporin C synthase of Streptomyces lactamdurans. J Gen Microbiol. 133: 3165-3174.

Curtis NA, Orr D, Ross GW, Boulton MG. 1979. Affinities of penicillins and cephalosporins for the penicillin-binding proteins of Escherichia coli K-12 and their antibacterial activity. Antimicrob Agents Chemother. 16(5): 533-539.

Davies C, White SW, Nicholas RA. 2001. Crystal structure of a deacylation-defective mutant of penicillin-binding protein 5 at 2.3-A resolution. J Biol Chem. 276: 616-623.

Davies J, Spiegelman GB, Yim G. 2006. The world of subinhibitory antibiotic concentrations. Curr Opin Microbiol. 9(5): 445-453.

Davies J. 1990. What are antibiotics? Archaic functions for modern activities. Mol Microbiol. 4(8): 1227-1232.

Davies J. 2006. Are antibiotics naturally antibiotics? J Ind Microbiol Biotechnol. 33(7): 496-499.

Davies P, Ashe BM, Bonney RJ, Dorn C, Finke P, Fletcher D, Hanlon WA, Humes JL, Maycock A, Mumford RA, Navia MA, Opas EE, Pachlock S, Shah S, Zimmerman M, Doherty JB. 1991. The discovery and biologic properties of cephalosporin-based inhibitors of PMN elastase. Ann NY Acad Sci. 624: 219-229.

Day AG. 2000. Mutant alpha-amylase having introduced therein a disulfide bond. Patent no.: 6,008,026 (USPTO).

Declerck N, Joyet P, Trosset JY, Garnier J, Gaillardin C. 1995. Hyperthermostable mutants of Bacillus licheniformis alpha-amylase: multiple amino acid replacements and molecular modelling. Protein Eng. 8(10): 1029-1037.

Declerck N, Machius M, Chambert R, Wiegand G, Huber R, Gaillardin C. 1997. Hyperthermostable mutants of Bacillus licheniformis alpha-amylase: thermodynamic studies and structural interpretation. Protein Eng. 10(5): 541-549.

Declerck N, Machius M, Wiegand G, Huber R, Gaillardin C. 2000. Probing structural determinants specifying high thermostability in Bacillus licheniformis alpha-amylase. J Mol Biol. 301(4): 1041-1057.

199

Demain AL, Elander RP. 1999. The beta-lactam antibiotics: past, present, and future. Antonie Van Leeuwenhoek. 75: 5-19.

Demain AL, Sanchez S. 2009. Microbial drug discovery: 80 years of progress. J Antibiot (Tokyo). 62(1): 5-16.

Demain AL. 1999. Pharmaceutically active secondary metabolites of microorganisms. Appl Microbiol Biotechnol. 52(4): 455-463.

Demain, AL, Cho H, Piret JM, Adrio JL, Fernandez M-JE, Baez-Vasquez MA, Hintermann G. 2002. Penicillin conversion. Patent no.: 6,383,773 (USPTO).

Demirev AV, Lee CH, Jaishy BP, Nam DH, Ryu DD. 2006. Substrate specificity of nonribosomal peptide synthetase modules responsible for the biosynthesis of the oligopeptide moiety of cephabacin in Lysobacter lactamgenus. FEMS Microbiol Lett. 255: 121-128.

Den Blaauwen T, Aarsman ME, Vischer NO, Nanninga N. 2003. Penicillin-binding protein PBP2 of Escherichia coli localizes preferentially in the lateral wall and at mid-cell in comparison with the old cell pole. Mol Microbiol. 47: 539-547.

Denome SA, Elf PK, Henderson TA, Nelson DE, Young KD. 1999. Escherichia coli mutants lacking all possible combinations of eight penicillin binding proteins: viability, characteristics, and implications for peptidoglycan synthesis. J Bacteriol. 181: 3981-3993.

Deresinski SC. 2008. Ceftobiprole: breaking therapeutic dogmas of the beta-lactam class. Diagn Microbiol Infect Dis. 61: 82-85.

Di Lallo G, Fagioli M, Barionovi D, Ghelardini P, Paolozzi L. 2003. Use of a two- hybrid assay to study the assembly of a complex multicomponent protein machinery: bacterial septosome differentiation. Microbiology. 149: 3353-3359.

Diez B, Barredo JL, Alvarez E, Cantoral JM, van Solingen P, Groenen MA, Veenstra AE, Martin JF. 1989. Two genes involved in penicillin biosynthesis are linked in a 5.1 kb SalI fragment in the genome of Penicillium chrysogenum. Mol Gen Genet. 218: 572-576.

Diez B, Gutierrez S, Barredo JL, van Solingen P, van der Voort LH, Martin JF. 1990. The cluster of penicillin biosynthetic genes. Identification and characterization of the pcbAB gene encoding the alpha-aminoadipyl-cysteinyl-valine synthetase and linkage to the pcbC and penDE genes. J Biol Chem. 265: 16358-16365.

Doherty JB, Ashe BM, Argenbright LW, Barker PL, Bonney RJ, Chandler GO, Dahlgren ME, Dom CP Jr, Finke P, Firestone RA, Fletcher D, Hagmann WK, Mumford R, OGrady L, Maycock AL, Pisano JM, Shah S, Thompson KR, Zimmerman M. 1986. Cephalosporin antibiotics can be modified to inhibit human leukocyte elastase. Nature. 322: 192-194.

200

Dotzlaf JE, Yeh WK. 1987. Copurification and characterization of deacetoxycephalosporin C synthetase/hydroxylase from Cephalosporium acremonium. J Bacteriol. 169(4): 1611-1618.

Dotzlaf JE, Yeh WK. 1989. Purification and properties of deacetoxycephalosporin C synthase from recombinant Escherichia coli and its comparison with the native enzyme purified from Streptomyces clavuligerus. J Biol Chem. 264: 10219-10227.

Dubus A, Lloyd MD, Lee HJ, Schofield CJ, Baldwin JE, Frere JM. 2001. Probing the penicillin sidechain selectivity of recombinant deacetoxycephalosporin C synthase. Cell Mol Life Sci. 58(5-6): 835-843.

Eisenstein M. 2005. New career possibilities for certain antibiotics? Lab Anim (NY). 34: 10.

Elander RP. 2003. Industrial production of β-lactam antibiotics. Appl Microbiol Biotechnol. 61: 385-392.

Enguita FJ, Liras P, Leitao AL, Martin JF. 1996. Interaction of the two proteins of the methoxylation system involved in cephamycin C biosynthesis. Immunoaffinity, protein cross-linking, and fluorescence spectroscopy studies. J Biol Chem. 271: 33225-33230.

Essack SY. 2001. The development of β-lactam antibiotics in response to the evolution of β-lactamase. Pharm Res. 18: 1391-1399.

Fernandez FJ, Cardoza RE, Montenegro E, Velasco J, Gutierrez S, Martin JF. 2003. The isopenicillin N acyltransferases of Aspergillus nidulans and Penicillium chrysogenum differ in their ability to maintain the 40-kDa alphabeta heterodimer in an undissociated form. Eur J Biochem. 270: 1958-1968.

Fernandez MJ, Adrio JL, Piret JM, Wolfe S, Ro S, Demain AL. 1999. Stimulatory effect of growth in the presence of alcohols on biotransformation of penicillin G into cephalosporin-type antibiotics by resting cells of Streptomyces clavuligerus NP1. Appl Microbiol Biotechnol. 52: 484-488.

French GL. 2005. Clinical impact and relevance of antibiotic resistance. Adv Drug Deliv Rev. 57: 1514-1527.

Frere JM, Ghuysen JM, Iwatsubo M. 1975. Kinetics of interaction between the exocellular DD-carboxypeptidase-transpeptidase from Streptomyces R61 and beta- lactam antibiotics. A choice of models. Eur J Biochem. 57: 343-351.

Fuda C, Hesek D, Lee M, Heilmayer W, Novak R, Vakulenko SB, Mobashery S. 2006. Mechanistic basis for the action of new cephalosporin antibiotics effective against methicillin- and vancomycin-resistant Staphylococcus aureus. J Biol Chem. 281: 10035-10041.

Gabor EM, Janssen DB, De Vries EJ. 2004. Penicillin acylase. Patent no.: WO/2004/111241 (WIPO).

201

Gabor EM, Janssen DB. 2004. Increasing the synthetic performance of penicillin acylase PAS2 by structure-inspired semi-random mutagenesis. Protein Eng Des Sel. 17(7): 571-579.

Gao Q, Piret JM, Adrio JL, Demain AL. 2003. Performance of a recombinant strain of Streptomyces lividans for bioconversion of penicillin G to deacetoxycephalosporin G. J Ind Microbiol Biotechnol. 30: 190-194.

Geourjon C, Deleage G. 1995. SOPMA: significant improvements in protein secondary structure prediction by consensus prediction from multiple alignments. Comput Appl Biosci. 11(6): 681-684.

Ghuysen JM. 1991. Serine β-lactamase and penicillin-binding proteins. Annu Rev Microbiol. 45: 37-67.

Giske CG, Ge J, Nordmann P. 2009. Activity of cephalosporin CXA-101 (FR264205) and comparators against extended-spectrum-betalactamase-producing Pseudomonas aeruginosa. J Antimicrob Chemother. 64: 430-431.

Gordon E, Mouz N, Duee E, Dideberg O. 2000. The crystal structure of the penicillin- binding protein 2x from Streptococcus pneumoniae and its acyl-enzyme form: implication in drug resistance. J Mol Biol. 299: 477-485.

Gotor V. 2002. Biocatalysis applied to the preparation of pharmaceuticals. Org Process Res Dev. 6(4): 420-426.

Guex N, Diemand A and Peitsch MC. 1999. Protein modeling for all. Trends Biochem Sci. 24: 364-367.

Guex N, Peitsch MC. 1997. SWISS-MODEL and the Swiss-PdbViewer: An environment for comparative protein modelling. Electrophoresis. 18: 2714-2723.

Guillen Schlippe YV, Hedstrom L. 2005. A twisted base? The role of arginine in enzyme-catalyzed proton abstractions. Arch Biochem Biophys. 433: 266-278.

Gunsior M, Breazeale SD, Lind AJ, Ravel J, Janc JW, Townsend CA. 2004. The biosynthetic gene cluster for a monocyclic beta-lactam antibiotic, nocardicin A. Chem Biol. 11(7): 927-938.

Gupta R, Beg QK, Lorenz P. 2002. Bacterial alkaline proteases: molecular approaches and industrial applications. Appl Microbiol Biotechnol. 59(1): 15-32.

Gutierrez S, Diez B, Montenegro E, Martin JF. 1991. Characterization of the Cephalosporium acremonium pcbAB gene encoding alpha-aminoadipyl-cysteinyl- valine synthetase, a large multidomain peptide synthetase: linkage to the pcbC gene as a cluster of early cephalosporin biosynthetic genes and evidence of multiple functional domains. J Bacteriol. 173: 2354-2365.

Gutierrez S, Velasco J, Fernandez FJ, Martin JF. 1992. The cefG gene of Cephalosporium acremonium is linked to the cefEF gene and encodes a

202

deacetylcephalosporin C acetyltransferase closely related to homoserine O- acetyltransferase. J Bacteriol. 174: 3056-3064.

Hamad B. 2010. The antibiotics market. Nat Rev Drug Discov. 9(9): 675-676.

Hamamatsu N, Nomiya Y, Aita T, Nakajima M, Husimi Y, Shibanaka Y. 2006. Directed evolution by accumulating tailored mutations: thermostabilization of lactate oxidase with less trade-off with catalytic activity. Protein Eng Des Sel. 19(11): 483- 489.

Hamilton-Miller JM. 1999. Beta-lactams: variations on a chemical theme, with some surprising biological results. J Antimicrob Chemother. 44(6): 729-734.

Hamilton-Miller JM. 2000. Sir Edward Abraham’s contribution to the development of the cephalosporins: a reassessment. Int J Antimicrob Agents. 15: 179-184.

Hamilton-Miller JM. 2008. Development of the semi-synthetic penicillins and cephalosporins. Int J Antimicrob Agents. 31: 189-192.

Hansch C, Sammes PG, Taylor JB. 1990. Comprehensive medicinal chemistry: the rational design, mechanistic study & therapeutic application of chemical compounds. 1st ed. Pergamon Press, Oxford, New York.

Hausinger RP. 2004. Fe(II)/alpha-ketoglutarate-dependent hydroxylases and related enzymes. Crit Rev Biochem Mol Biol. 39: 21-68.

Higgens CE, Kastner RE. 1971. Streptomyces clavuligerus sp. nov., a beta-lactam antibiotic producer. Int J Syst Bacteriol. 21: 326-31.

Hranueli D, Cullum J, Basrak B, Goldstein P, Long PF. 2005. Plasticity of the streptomyces genome-evolution and engineering of new antibiotics. Curr Med Chem. 12(14): 1697-1704.

Hsu JS, Yang YB, Deng CH, Wei CL, Liaw SH, Tsai YC. 2004. Family shuffling of expandase genes to enhance substrate specificity for penicillin G. Appl Environ Microbiol. 70(10): 6257-6263.

Hsu LY, Tan TY, Tam VH, Kwa A, Fisher DA, Koh TH; Network for Antimicrobial Resistance Surveillance (Singapore). 2010. Surveillance and correlation of antibiotic prescription and resistance of Gram-negative bacteria in Singaporean hospitals. Antimicrob Agents Chemother. 54(3): 1173-1178.

Hutchison CA, 3rd, Phillips S, Edgell MH, Gillam S, Jahnke P, Smith M. 1978. Mutagenesis at a specific position in a DNA sequence. J Biol Chem. 253: 6551-6560.

Ikeda H, Ishikawa J, Hanamoto A, Shinose M, Kikuchi H, Shiba T, Sakaki Y, Hattori M, Omura S. 2003. Complete genome sequence and comparative analysis of the industrial microorganism Streptomyces avermitilis. Nat Biotechnol. 21(5): 526-531.

203

Jensen SE, Westlake DW, Wolfe S. 1983. Analysis of penicillin N ring expansion activity from Streptomyces clavuligerus by ion-pair high-pressure liquid chromatography. Antimicrob Agents Chemother. 24: 307-312.

Jensen SE, Westlake DW, Wolfe S. 1985. Deacetoxycephalosporin C synthetase and deacetoxycephalosporin C hydroxylase are two separate enzymes in Streptomyces clavuligerus. J Antibiot (Tokyo). 38(2): 263-265.

Jeong H, Barbe V, Lee CH, Vallenet D, Yu DS, Choi SH, Couloux A, Lee SW, Yoon SH, Cattolico L, Hur CG, Park HS, Ségurens B, Kim SC, Oh TK, Lenski RE, Studier FW, Daegelen P, Kim JF. 2009. Genome sequences of Escherichia coli B strains REL606 and BL21(DE3). J Mol Biol. 394(4): 644-652.

Jestin JL, Vichier-Guerre S. 2005. How to broaden enzyme substrate specificity: strategies, implications and applications. Res Microbiol. 156: 961-966.

Jones RN, Pfaller MA. 1998. Bacterial resistance: a worldwide problem. Diagn Microbiol Infect Dis. 31: 379-388.

Kahan JS, Kahan FM, Goegelman R, Currie SA, Jackson M, Stapley EO, Miller TW, Miller AK, Hendlin D, Mochales S, Hernandez S, Woodruff HB, Birnbaum J. 1979. Thienamycin, a new beta-lactam antibiotic. I. Discovery, taxonomy, isolation and physical properties. J Antibiot (Tokyo). 1979 Jan;32(1): 1-12.

Kaneko H, Minagawa H, Shimada J. 2005. Rational design of thermostable lactate oxidase by analyzing quaternary structure and prevention of deamidation. Biotechnol Lett. 27(22): 1777-1784.

Karimova G, Dautin N, Ladant D. 2005. Interaction network among Escherichia coli membrane proteins involved in cell division as revealed by bacterial two-hybrid analysis. J Bacteriol. 187: 2233-2243.

Katz L, Khosla C. 2007. Antibiotic production from the ground up. Nat Biotechnol. 25(4): 428-429.

Kazlauskas RJ, Bornscheuer UT. 2009. Finding better protein engineering strategies. Nat Chem Biol. 5(8): 526-529.

Khaoua S, Lebrihi A, Germain P, Lefebvre G. 1991. Cephamycin C biosynthesis in Streptomyces cattleya: nitrogen source regulation. Appl Microbiol Biotechnol. 35: 253-257.

Kiefer F, Arnold K, Künzli M, Bordoli L,Schwede T. 2009. The SWISS-MODEL Repository and associated resources. Nucleic Acids Research. 37: D387-D392.

Kim WM, Kang K. 2000. Enzymatic and molecular biochemical characterizations of human neutrophil elastases and a cathepsin G-like enzyme. Mol Cells. 10(5): 498- 504.

204

Kimura H, Izawa M, Sumino Y. 1996. Molecular analysis of the gene cluster involved in cephalosporin biosynthesis from Lysobacter lactamgenus YK90. Appl Microbiol Biotechnol. 44: 589-596.

Kirby WM. 1944. Extraction of a highly potent penicillin inactivator from penicillin resistant staphylococci. Science. 99: 452-453.

Kirst HA. 2010. Antibiotics versus resistant bacteria: a continual challenge. Expert Opin Ther Pat. 20(10): 1271-1272.

Knight WB, Maycock AL, Green BG, Ashe BM, Gale P, Weston H, Finke PE, Hagmann WK, Shah SK, Doherty JB. 1992. Mechanism of inhibition of human leukocyte elastase by two cephalosporin derivatives. Biochemistry. 31(21): 4980- 4986.

Kovacevic S, Miller JR. 1991. Cloning and sequencing of the beta-lactam hydroxylase gene (cefF) from Streptomyces clavuligerus: gene duplication may have led to separate hydroxylase and expandase activities in the actinomycetes. J Bacteriol. 173: 398-400.

Kovacevic S, Weigel BJ, Tobin MB, Ingolia TD, Miller JR. 1989. Cloning, characterization, and expression in Escherichia coli of the Streptomyces clavuligerus gene encoding deacetoxycephalosporin C synthetase. J Bacteriol. 171: 754-760.

Laemmli UK. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227: 680-685.

Laudano JB. 2011. Ceftaroline fosamil: a new broad-spectrum cephalosporin. J Antimicrob Chemother. 66(Suppl 3): iii11-18.

Lee HJ, Lloyd MD, Clifton IJ, Harlos K, Dubus A, Baldwin JE, Frere JM, Schofield CJ. 2001b. Alteration of the co-substrate selectivity of deacetoxycephalosporin C synthase. The role of arginine 258. J Biol Chem. 276: 18290-18295.

Lee HJ, Lloyd MD, Harlos K, Clifton IJ, Baldwin JE, Schofield CJ. 2001a. Kinetic and crystallographic studies on deacetoxycephalosporin C synthase (DAOCS). J Mol Biol. 308: 937-948.

Lee HJ, Schofield CJ, Lloyd MD. 2002. Active site mutations of recombinant deacetoxycephalosporin C synthase. Biochem Biophys Res Commun. 292: 66-70.

Levine DP. 2006. Vancomycin: a history. Clin Infect Dis. 42(Suppl 1): S5-12.

Levinovitz AW, Ringertz N. 2001. The Nobel Prize: the first 100 years. Imperial College Press. London, UK.

Lin YQ, Li JY, Wang HY, Liu J, Zhang CL, Wang WT, Liu J, Li N and Jin SH. 2008. Cloning and identification of a novel sperm binding protein, HEL-75, with antibacterial activity and expressed in the human epididymis. Hum Reprod. 23(9): 2086-2094.

205

Lipscomb SJ, Lee HJ, Mukherji M, Baldwin JE, Schofield CJ, Lloyd MD. 2002. The role of arginine residues in substrate binding and catalysis by deacetoxycephalosporin C synthase. Eur J Biochem. 269: 2735-2739.

Liras P, Demain AL. 2009. Chapter 16. Enzymology of beta-lactam compounds with cephem structure produced by actinomycete. Methods Enzymol. 458: 401-429.

Liras P. 1999. Biosynthesis and molecular genetics of cephamycins - Cephamycins produced by Actinomycetes. Antonie Van Leeuwenhoek. 75: 109-124.

Livermore DM, Mushtaq S, Ge Y, Warner M. 2009. Activity of cephalosporin CXA- 101 (FR264205) against Pseudomonas aeruginosa and Burkholderia cepacia group strains and isolates. Int J Antimicrob Agents. 34: 402-406.

Llarrull LI, Fisher JF, Mobashery S. 2009. Molecular basis and phenotype of methicillin resistance in Staphylococcus aureus and insights into new beta-lactams that meet the challenge. Antimicrob Agents Chemother. 53: 4051-4063.

Lloyd MD, Lee HJ, Harlos K, Zhang ZH, Baldwin JE, Schofield CJ, Charnock JM, Garner CD, Hara T, Terwisscha van Scheltinga AC, Valegard K, Viklund JA, Hajdu J, Andersson I, Danielsson A, Bhikhabhai R. 1999b. Studies on the active site of deacetoxycephalosporin C synthase. J Mol Biol. 287: 943-960.

Lloyd MD, Lipscomb SJ, Hewitson KS, Hensgens CM, Baldwin JE, Schofield CJ. 2004. Controlling the substrate selectivity of deacetoxycephalosporin/ deacetylcephalosporin C synthase. J Biol Chem. 279(15): 15420-15426.

Lloyd MD, Merritt KD, Lee V, Sewell TJ, Wha-Son B, Baldwin JE, Schofield CJ, Elson SW, Baggaley KH, Nicholson NH. 1999a. Product-substrate engineering by bacteria: Studies on clavaminate synthase, a trifunctional dioxygenase. Tetrahedron. 55: 10201-10220.

Lloyd MD, Schofield CJ, Hewitson KS, Lipscomb S. 2005. Modified DAOC/DACS enzymes. Patent no.: WO 2005/019468 A1 (UKIPO).

Lo TP, Komar-Panicucci S, Sherman F, McLendon G, Brayer GD. 1995. Structural and functional effects of multiple mutations at distal sites in cytochrome c. Biochemistry. 34: 5259-5268.

Loder PB, Abraham EP. 1971. Biosynthesis of peptides containing -aminoadipic acid and cysteine in extracts of a Cephalosporium sp. Biochem J. 123: 477-482.

Loke P, Sim TS. 2000. Mutational analysis of tyrosine-191 in the catalysis of Cephalosporium acremonium isopenicillin N synthase. J Biochem (Tokyo). 127: 585- 589.

Loke P, Sim TS. 2001. A comparison of three site-directed mutagenesis kits. Z Naturforsch C. 56(9-10): 810-813.

206

Macheboeuf P, Contreras-Martel C, Job V, Dideberg O, Dessen A. 2006. Penicillin binding proteins: key players in bacterial cell cycle and drug resistance processes. FEMS Microbiol Rev. 30: 673-691.

Macleod AJ, Ross HB, Ozere RL, Digout G, van Rooyenc. 1964. Lincomycin: a new antibiotic active against staphylococci and other Gram-positive cocci: clinical and laboratory studies. Can Med Assoc J. 91: 1056-1060.

Madzak C, Mimmi MC, Caminade E, Brault A, Baumberger S, Briozzo P, Mougin C, Jolivalt C. 2006. Shifting the optimal pH of activity for a laccase from the fungus Trametes versicolor by structure-based mutagenesis. Protein Eng Des Sel. 19(2): 77- 84.

Maeda K, Luengo JM, Ferrero O, Wolfe S, Lebedev MY, Fang A, Demain AL. 1995. The substrate-specificity of deacetoxycephalosporin C synthase (expandase) of Streptomyces clavuligerus is extremely narrow. Enzyme Microb Technol. 17: 231- 234.

Mann J. 2004. Fighting bacteria. In: Life Saving Drugs: The Elusive Magic Bullet. p11-84.

Mao J. 2005. Glutamate transporter: an unexpected target for some antibiotics. Mol Pain. 1: 5.

Marshall WF, Blair JE. 1999. The cephalosporins. Mayo Clin Proc. 74: 187-195.

Martin JF, Gutierrez S. 1995. Genes for beta-lactam antibiotic biosynthesis. Antonie Van Leeuwenhoek. 67: 181-200.

Martin JF, Ullan RV, Casqueiro J. 2004. Novel genes involved in cephalosporin biosynthesis: The three-component isopenicillin N epimerase system. Adv Biochem Eng Biotechnol. 88: 91-109.

Martin JF, Ullan RV, Garcia-Estrada C. 2010. Regulation and compartmentalization of beta-lactam biosynthesis. Microb Biotechnol. 3: 285-299.

Martinelle M, Holmquiest M, Clausen IG, Patkar S, Svensen A, Hult K. 1996. The role of Glu87 and Trp87 in the lid of Humicola lanuginosa lipase. Protein Eng. 9: 519-524.

Martinez CA, Hu S, Dumond Y, Tao J, Kelleher P, Tully L. 2008. Development of a chemoenzymatic manufacturing process for pregabalin. Org Proc Res Dev. 12: 392- 398.

Massova I, Kollman PA. 2002. pKa, MM, and QM studies of mechanisms of beta- lactamases and penicillin-binding proteins: acylation step. J Comput Chem. 23: 1559- 1576.

Massova I, Mobashery S. 1998. Kinship and diversification of bacterial penicillin- binding proteins and beta-lactamases. Antimicrob Agents Chemother. 42: 1-17.

207

Matsuhashi M, Song MD, Ishino F, Wachi M, Doi M, Inoue M, Ubukata K, Yamashita N, Konno M. 1986. Molecular cloning of the gene of a penicillin-binding protein supposed to cause high resistance to beta-lactam antibiotics in Staphylococcus aureus. J Bacteriol. 167: 975-980.

Maurer KH. 2004. Detergent proteases. Curr Opin Biotechnol. 15(4): 330-334.

McHenney MA, Hosted TJ, Dehoff BS, Rosteck PR Jr., Baltz RH. 1998. Molecular cloning and physical mapping of the daptomycin gene cluster from Streptomyces roseosporus. J Bacteriol. 180(1): 143-151.

McPherson DC, Popham DL. 2003. Peptidoglycan synthesis in the absence of class A penicillin-binding proteins in Bacillus subtilis. J Bacteriol. 185: 1423-1431.

Mesaros N, Nordmann P, Plesiat P, Roussel-Delvallez M, Van Eldere J, Glupczynski Y, Van Laethem Y, Jacobs F, Lebecque P, Malfroot A, Tulkens PM, Van Bambeke F. 2007. Pseudomonas aeruginosa: resistance and therapeutic options at the turn of the new millennium. Clin Microbiol Infect. 13: 560-578.

Mildvan AS, Weber DJ, Kuliopulos A. 1992. Quantitative interpretations of double mutations of enzymes. Arch Biochem Biophys. 294: 327-340.

Miller LA, Ratnam K, Payne DJ. 2001. Beta-lactamase-inhibitor combinations in the 21st century: current agents and new developments. Curr Opin Pharmacol. 1: 451-458.

Miller TM, Cleveland DW. 2005. Medicine. Treating neurodegenerative diseases with antibiotics. Science. 307: 361-362.

Moellering RC Jr. 2010. NDM-1--a cause for worldwide concern. N Engl J Med. 363(25): 2377-2379.

Moellering RC Jr. 2011. Discovering new antimicrobial agents. Int J Antimicrob Agents. 37(1): 2-9.

Morin RB, Roeske RW, Flynn EH, Jackson BG. 1962. Chemistry of cephalosporin antibiotics. 1. 7-Aminocephalosporanic acid from cephalosporin C. J Am Chem Soc. 84: 3400-3401.

Morley KL, Kazlauskas RJ. 2005. Improving enzyme properties: when are closer mutations better? Trends Biotechnol. 23(5): 231-237.

Morlot C, Noirclerc-Savoye M, Zapun A, Dideberg O, Vernet T. 2004. The D,D- carboxypeptidase PBP3 organizes the division process of Streptococcus pneumoniae. Mol Microbiol. 51: 1641-1648.

Morlot C, Pernot L, Le Gouellec A, Di Guilmi AM, Vernet T, Dideberg O, Dessen A. 2005. Crystal structure of a peptidoglycan synthesis regulatory factor (PBP3) from Streptococcus pneumoniae. J Biol Chem. 280: 15984-15991.

208

Mulvey MR, Grant JM, Plewes K, Roscoe D, Boyd DA. 2011. New Delhi metallo-β- lactamase in Klebsiella pneumoniae and Escherichia coli, Canada. Emerg Infect Dis. 17(1): 103-106.

Murano K, Yamanaka T, Toda A, Ohki H, Okuda S, Kawabata K, Hatano K, Takeda S, Akamatsu H, Itoh K, Misumi K, Inoue S, Takagi T. 2008. Structural requirements for the stability of novel cephalosporins to AmpC beta-lactamase based on 3D- structure. Bioorg Med Chem. 16: 2261-2275.

Nagarajan R, Boeck LD, Gorman M, Hamill RL, Higgens CE, Hoehn MM, Stark WM, Whitney JG. 1971. Beta-lactam antibiotics from Streptomyces. J Am Chem Soc. 93: 2308-2310.

Newman DJ, Cragg GM. 2009. Natural products as drugs and leads to drugs: The historical perspective. In: Natural Product Chemistry for Drug Discovery. Buss AD, Butler MS, editors. p3-27. RSC Publishing. UK.

Newman DJ. 2008. Natural Products as Leads to Potential Drugs: An Old Process or the New Hope for Drug Discovery? J. Med. Chem. 51: 2589-2599.

Nicholas RA, Krings S, Tomberg J, Nicola G, Davies C. 2003. Crystal structure of wild-type penicillin-binding protein 5 from Escherichia coli: implications for deacylation of the acyl-enzyme complex. J Biol Chem. 278: 52826-52833.

Nicola G, Peddi S, Stefanova M, Nicholas RA, Gutheil WG, Davies C. 2005. Crystal structure of Escherichia coli penicillin-binding protein 5 bound to a tripeptide boronic acid inhibitor: a role for Ser-110 in deacylation. Biochemistry. 44: 8207-8217.

Nieboer M, Bovenberg RAL. 2006. Method to localize expandase in the cytosol. Patent no.: 6,995,003 (USPTO).

Nielsen, BR, Svendsen A, Pedersen H, Vind J, Hendriksen HV, Frandsen TP. 2002. Glucoamylase variants. Patent no.: 6,352,851 (USPTO).

Nozaki Y, Katayama N, Tsubotani S, Ono H, Okazaki H. 1985. Cephabacin M1-6, new 7-methoxycephem antibiotics of bacterial origin. I. A producing organism, fermentation, biological activities, and mode of action. J Antibiot (Tokyo). 38: 1141- 1151.

Nozaki Y, Okonogi K, Katayama N, Ono H, Harada S, Kondo M, Okazaki H. 1984. Cephabacins, new cephem antibiotics of bacterial origin. IV. Antibacterial activities, stability to beta-lactamases and mode of action. J Antibiot (Tokyo). 37: 1555-1565.

Ogawa J, Shimizu S. 2002. Industrial microbial enzymes: their discovery by screening and use in large-scale production of useful chemicals in Japan. Curr Opin Biotechnol. 13(4): 367-375.

Oh B, Kim M, Yoon J, Chung K, Shin Y, Lee D, Kim Y. 2003. Deacylation activity of cephalosporin acylase to cephalosporin C is improved by changing the side-chain conformations of active-site residues. Biochem Biophys Res Commun. 310(1): 19-27.

209

Ohnishi Y, Ishikawa J, Hara H, Suzuki H, Ikenoya M, Ikeda H, Yamashita A, Hattori M, Horinouchi S. 2008. Genome sequence of the streptomycin-producing microorganism Streptomyces griseus IFO 13350. J Bacteriol. 190(11): 4050-4060.

Okada Y, Yoshigi N, Sahar H, Koshino S. 1995. Increase in thermostability of recombinant barley α-amylase by random mutagenesis. Biosci Biotechnol Biochem. 59: 1152-1153.

Okami Y, Tazaki T, Katumata S, Honda K, Suzuki M, Umezawa H. 1959. Studies on Streptomyces kanamyceticus, producer of kanamycin. J Antibiot (Tokyo). 12: 252- 256.

Oliva M, Dideberg O, Field MJ. 2003. Understanding the acylation mechanisms of active-site serine penicillin-recognizing proteins: a molecular dynamics simulation study. Proteins. 53: 88-100.

Omura S, Ikeda H, Ishikawa J, Hanamoto A, Takahashi C, Shinose M, Takahashi Y, Horikawa H, Nakazawa H, Osonoe T, Kikuchi H, Shiba T, Sakaki Y, Hattori M. 2001. Genome sequence of an industrial microorganism Streptomyces avermitilis: deducing the ability of producing secondary metabolites. Proc Natl Acad Sci U S A. 98(21): 12215-12220.

Ono H, Nozaki Y, Katayama N, Okazaki H. 1984. Cephabacins, new cephem antibiotics of bacterial origin. I. Discovery and taxonomy of the producing organisms and fermentation. J Antibiot (Tokyo). 37: 1528-1535.

Oster LM, van Scheltinga AC, Valegard K, Hose AM, Dubus A, Hajdu J, Andersson I. 2004. Conformational flexibility of the C terminus with implications for substrate binding and catalysis revealed in a new crystal form of deacetoxycephalosporin C synthase. J Mol Biol. 343: 157-171.

O'Sullivan J, Aplin RT, Stevens CM, Abraham EP. 1979. Biosynthesis of a 7-alpha- methoxycephalosporin. Incorporation of molecular oxygen. Biochem J. 179: 47-52.

Otten LG, Sio CF, van der Sloot AM, Cool RH, Quax WJ. 2004. Mutational analysis of a key residue in the substrate specificity of a cephalosporin acylase. Chembiochem. 5(6): 820-825.

Otten LG, Sio CF, Vrielink J, Cool RH, Quax WJ. 2002. Altering the substrate specificity of cephalosporin acylase by directed evolution of the beta-subunit. J Biol Chem. 277: 42121-42127.

Owens DK, Crosby KC, Runac J, Howard BA, Winkel BS. 2008. Biochemical and genetic characterization of Arabidopsis flavanone 3beta-hydroxylase. Plant Physiol Biochem. 46(10): 833-843.

Palissa H, von Dohren H, Kleinkauf H, Ting HH, Baldwin JE. 1989. Beta-lactam biosynthesis in a gram-negative eubacterium: purification and characterization of isopenicillin N synthase from Flavobacterium sp. strain SC 12.154. J Bacteriol. 171: 5720-5728.

210

Pandey A, Nigam P, Soccol CR, Soccol VT, Singh D, Mohan R. 2000. Advances in microbial amylases. Biotechnol Appl Biochem. 31(Pt 2):135-52.

Park Y, Choi H, Lee DS, Kim Y. 1997. Improvement of the 3'-5' exonuclease activity of Taq DNA polymerase by protein engineering in the active site. Mol Cells. 7(3): 419-424.

Peracchi A. 2001. Enzyme catalysis: removing chemically 'essential' residues by site- directed mutagenesis. Trends Biochem Sci. 26(8): 497-503.

Peterson WH, Tornqvist EG. 1956. Penicillin production by high-yielding strains of Penicillium chrysogenum. Appl Microbiol. 4: 277-283.

Pfeifer Y, Witte W, Holfelder M, Busch J, Nordmann P, Poirel L. 2011. NDM-1- producing Escherichia coli in Germany. Antimicrob Agents Chemother. 55(3): 1318- 1319.

Poirel L, Fortineau N, Nordmann P. 2011. International transfer of NDM-1-producing Klebsiella pneumoniae from Iraq to France. Antimicrob Agents Chemother. 55(4):1821-1822.

Pollegioni L, Lorenzi S, Rosini E, Marcone GL, Molla G, Verga R, Cabri W, Pilone MS. 2005. Evolution of an acylase active on cephalosporin C. Protein Sci. 14: 3064- 3076.

Pollegioni L, Molla G, Sacchi S, Rosini E, Verga R, Pilone MS. 2008. Properties and applications of microbial D-amino acid oxidases: current state and perspectives. Appl Microbiol Biotechnol. 78(1): 1-16.

Popham DL, Setlow P. 1995. Cloning, nucleotide sequence, and mutagenesis of the Bacillus subtilis ponA operon, which codes for penicillin-binding protein (PBP) 1 and a PBP-related factor. J Bacteriol. 177: 326-335.

Popham DL, Setlow P. 1996. Phenotypes of Bacillus subtilis mutants lacking multiple class A high-molecular-weight penicillin-binding proteins. J Bacteriol. 178: 2079- 2085.

Popham DL, Young KD. 2003. Role of penicillin-binding proteins in bacterial cell morphogenesis. Curr Opin Microbiol. 6: 594-599.

Porter EM, van Dam E, Valore EV and Ganz T. 1997. Broad-spectrum antimicrobial activity of human intestinal defensin 5. Infect Immun. 65(6): 2396-2401.

Projan SJ. 2003. Why is big Pharma getting out of antibacterial drug discovery? Curr Opin Microbiol. 6: 427-430.

Reiersen H, Rees AR. 2001. The hunchback and its neighbours: proline as an environmental modulator. Trends Biochem Sci. 26: 679-684.

211

Remy I, Ghaddar G, Michnick SW. 2007. Using the beta-lactamase protein-fragment complementation assay to probe dynamic protein-protein interactions. Nat Protoc. 2: 2302-2306.

Rice LB. 2008. Federal funding for the study of antimicrobial resistance in nosocomial pathogens: no ESKAPE. J Infect Dis. 197(8): 1079-1081.

Riva S. 2001. Biocatalytic modification of natural products. Curr Opin Chem Biol. 5(2): 106-111.

Roach PL, Clifton IJ, Fulop V, Harlos K, Barton GJ, Hajdu J, Andersson I, Schofield CJ, Baldwin JE. 1995. Crystal structure of isopenicillin N synthase is the first from a new structural family of enzymes. Nature. 375: 700-704.

Roach PL, Clifton IJ, Hensgens CM, Shibata N, Schofield CJ, Hajdu J, Baldwin JE. 1997. Structure of isopenicillin N synthase complexed with substrate and the mechanism of penicillin formation. Nature. 387: 827-830.

Rolinson GN, Geddes AM. 2007. The 50th anniversary of the discovery of 6- aminopenicillanic acid (6-APA). Int J Antimicrob Agents. 29: 3-8.

Rolinson GN. 1998. Forty years of beta-lactam research. J Antimicrob Chemother. 41: 589-603.

Rothstein JD, Patel S, Regan MR, Haenggeli C, Huang YH, Bergles DE, Jin L, Dykes Hoberg M, Vidensky S, Chung DS, Toan SV, Bruijn LI, Su ZZ, Gupta P, Fisher PB. 2005. Beta-lactam antibiotics offer neuroprotection by increasing glutamate transporter expression. Nature. 433: 73-77.

Roy D, Sharma A, Bhowmick G, Roy MK, Ghosh AC. 1997. Characterization of Streptomyces sp. strain DRS-1 and its ampicillin transformation product. Folia Microbiologica. 42: 333-336.

Samson SM, Dotzlaf JE, Slisz ML, Becker GW, van Frank RM, Veal LE, Yeh WK, Miller JR, Queener SW, Ingolia TD. 1987. Cloning and expression of the fungal expandase/hydroxylase gene involved in cephalosporin biosynthesis. Biotechnology. 5: 1207-1214.

Samuelsen O, Thilesen CM, Heggelund L, Vada AN, Kümmel A, Sundsfjord A. 2011. Identification of NDM-1-producing Enterobacteriaceae in Norway. J Antimicrob Chemother. 66(3): 670-672.

Sanyal AK, Chowdhury B, Banerjee AB. 1992. Generation of high antimycotic activity during degradation of beta-lactam antibiotics. Lett Appl Microbiol. 14(5): 221-223.

Sauvage E, Duez C, Herman R, Kerff F, Petrella S, Anderson JW, Adediran SA, Pratt RF, Frere JM, Charlier P. 2007. Crystal structure of the Bacillus subtilis penicillin- binding protein 4a, and its complex with a peptidoglycan mimetic peptide. J Mol Biol. 371: 528-539.

212

Sauvage E, Kerff F, Terrak M, Ayala JA, Charlier P. 2008a. The penicillin-binding proteins: structure and role in peptidoglycan biosynthesis. FEMS Microbiol Rev. 32: 234-258.

Sauvage E, Powell AJ, Heilemann J, Josephine HR, Charlier P, Davies C, Pratt RF. 2008b. Crystal structures of complexes of bacterial DD-peptidases with peptidoglycan-mimetic ligands: the substrate specificity puzzle. J Mol Biol. 381: 383- 393.

Scheidegger A, Küenzi MT, Nüesch J. 1984. Partial purification and catalytic properties of a bifunctional enzyme in the biosynthetic pathway of beta-lactams in Cephalosporium acremonium. J Antibiot (Tokyo). 37(5): 522-531.

Schmidt FR. 2010. The β-lactam antibiotics: current situation and future prospects in manufacture and therapy. The Mycota. 10(2): 101-121.

Schneider T, Sahl HG. 2010. An oldie but a goodie - cell wall biosynthesis as antibiotic target pathway. Int J Med Microbiol. 300: 161-169.

Schwaber MJ, Carmeli Y. 2008. Carbapenem-resistant Enterobacteriaceae: a potential threat. JAMA. 300(24): 2911-2913.

Schwede T, Kopp J, Guex N, and Peitsch MC. 2003. SWISS-MODEL: an automated protein homology-modeling server. Nucleic Acids Research. 31: 3381-3385.

Secko D. 2005. Antibiotics that protect the brain. CMAJ. 172: 467-468.

Shampo MA, Kyle RA. 2003. Michael Smith--Canadian biochemist wins 1993 Nobel Prize. Mayo Clin Proc. 78: 804.

Shen YQ, Wolfe S, Demain AL. 1984. Deacetoxycephalosporin-C synthetase - importance of order of cofactor/reactant addition. Enzyme Microb Technol. 6: 402- 404.

Shin YC, Jeon JYJ, Jung KH, Park MR, Kim Y. Cephalosporin C acylase mutant and method for preparing 7-ACA. Patent no.: WO/2005/014821 (WIPO).

Silver LL. 2007. Multi-targeting by monotherapeutic antibacterials. Nat Rev Drug Discov. 6: 41-55.

Sim J and Sim TS. 2001. In vitro conversion of penicillin G and ampicillin by recombinant Streptomyces clavuligerus NRRL 3585 deacetoxycephalosporin C synthase. Enzyme Microb Tech. 29: 240-245.

Sim J, Sim TS. 2000. Mutational evidence supporting the involvement of tripartite residues His183, Asp185, and His243 in Streptomyces clavuligerus deacetoxycephalosporin C synthase for catalysis. Biosci Biotechnol Biochem. 64: 828-832.

213

Sim J, Wong E, Chin HS, Sim TS. 2003. Conserved structural modules and bonding networks in isopenicillin N synthase related non-haem iron-dependent oxygenases and oxidases. J Mol Catal B Enzym. 23: 17-27.

Singh R, Sharma R, Tewari N; Geetanjali, Rawat DS. 2006. Nitrilase and its application as a 'green' catalyst. Chem Biodivers. 3(12): 1279-1287.

Sio CF, Quax WJ. 2004. Improved beta-lactam acylases and their use as industrial biocatalysts. Curr Opin Biotechnol. 15(4): 349-355.

Skinner MM, Terwilliger TC. 1996. Potential use of additivity of mutational effects in simplifying protein engineering. Proc Natl Acad Sci USA. 93: 10753-10757.

Smith B, Warren SC, Newton GG, Abraham EP. 1967. Biosynthesis of penicillin N and cephalosporin C. Antibiotic production and other features of the metabolism of Cephalosporium sp. Biochem J. 103: 877-890.

So A, Furlong M, Heddini A. 2010. Globalisation and antibiotic resistance. BMJ. 341: c5116.

Sobolev V, Sorokine A, Prilusky J, Abola EE and Edelman M. 1999. Automated analysis of interatomic contacts in proteins. Bioinformatics. 15: 327-332.

Sonawane VC. 2006. Enzymatic modifications of cephalosporins by cephalosporin acylase and other enzymes. Crit Rev Biotechnol. 26: 95-120.

Song JH, Chang HH, Suh JY, Ko KS, Jung SI, Oh WS, Peck KR, Lee NY, Yang Y, Chongthaleong A, Aswapokee N, Chiu CH, Lalitha MK, Perera J, Yee TT, Kumararasinghe G, Jamal F, Kamarulazaman A, Parasakthi N, Van PH, So T, Ng TK; ANSORP Study Group. 2004. Macrolide resistance and genotypic characterization of Streptococcus pneumoniae in Asian countries: a study of the Asian Network for Surveillance of Resistant Pathogens (ANSORP). J Antimicrob Chemother. 53(3): 457-463.

Song JY, Jeong H, Yu DS, Fischbach MA, Park HS, Kim JJ, Seo JS, Jensen SE, Oh TK, Lee KJ, Kim JF. 2010. Draft genome sequence of Streptomyces clavuligerus NRRL 3585, a producer of diverse secondary metabolites. J Bacteriol. 192(23): 6317- 6318.

Spratt BG, Pardee AB. 1975. Penicillin-binding proteins and cell shape in E. coli. Nature. 254: 516-517.

Spratt BG, Strominger JL. 1976. Identification of the major penicillin-binding proteins of Escherichia coli as D-alanine carboxypeptidase IA. J Bacteriol. 127: 660- 663.

Spratt BG. 1975. Distinct penicillin binding proteins involved in the division, elongation, and shape of Escherichia coli K12. Proc Natl Acad Sci U S A. 72: 2999- 3003.

214

Stapley EO, Jackson M, Hernandez S, Zimmerman SB, Currie SA, Mochales S, Mata JM, Woodruff HB, Hendlin D. 1972. Cephamycins, a new family of beta-lactam antibiotics. I. Production by actinomycetes, including Streptomyces lactamdurans sp. n. Antimicrob Agents Chemother. 2(3): 122-131.

Stein RA, Goetz RM, Ganea GM. 2009. Ceftobiprole: a new beta-lactam antibiotic. Int J Clin Pract. 63: 930-943.

Strateva T, Yordanov D. 2009. Pseudomonas aeruginosa - a phenomenon of bacterial resistance. J Med Microbiol. 58: 1133-1148.

Strominger JL, Tipper DJ. 1965. Bacterial cell wall synthesis and structure in relation to the mechanism of action of penicillins and other antibacterial agents. Am J Med. 39: 708-721.

Studier FW, Moffatt BA. 1986. Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J Mol Biol. 189(1): 113-130.

Studier FW, Daegelen P, Lenski RE, Maslov S, Kim JF. 2009. Understanding the differences between genome sequences of Escherichia coli B strains REL606 and BL21(DE3) and comparison of the E. coli B and K-12 genomes. J Mol Biol. 394(4): 653-680.

Svendsen, A. 1996. Protein engineering of microbial lipases with industrial interest. In: Enzyme Technology for Industrial Applications.; Savage L, Southborough MA, editors. p98-123. IBI Biomedical Library.

Takagi H, Hirai K, Maeda Y, Matsuzawa H, Nakamori S. 2000. Engineering subtilisin E for enhanced stability and activity in polar organic solvents. J Biochem. 127(4): 617-625.

Tan JS, File TM Jr. 1995. Antipseudomonal penicillins. Med Clin North Am. 79: 679- 693.

Tessman I, Ishiwa H, Kumar S. 1965. Mutagenic effects of hydroxylamine in vivo. Science. 148: 507-508.

Thompson JD, Higgins DG, Gibson TJ. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22: 4673-4680.

Tijet N, Alexander DC, Richardson D, Lastovetska O, Low DE, Patel SN, Melano RG. 2011. New Delhi metallo-beta-lactamase, Ontario, Canada. Emerg Infect Dis. 17(2): 306-307.

Titelman E, Karlsson IM, Ge Y, Giske CG. 2011. In vitro activity of CXA-101 plus tazobactam (CXA-201) against CTX-M-14- and CTX-M-15-producing Escherichia coli and Klebsiella pneumoniae. Diagn Microbiol Infect Dis. 70: 137-141.

215

Tobin MB, Kovacevic S, Madduri K, Hoskins JA, Skatrud PL, Vining LC, Stuttard C, Miller JR. 1991. Localization of the lysine epsilon-aminotransferase (lat) and delta- (L-alpha-aminoadipyl)-L-cysteinyl-D-valine synthetase (pcbAB) genes from Streptomyces clavuligerus and production of lysine epsilon-aminotransferase activity in Escherichia coli. J Bacteriol. 173: 6223-6229.

Toda A, Ohki H, Yamanaka T, Murano K, Okuda S, Kawabata K, Hatano K, Matsuda K, Misumi K, Itoh K, Satoh K, Inoue S. 2008. Synthesis and SAR of novel parenteral antipseudomonal cephalosporins: discovery of FR264205. Bioorg Med Chem Lett. 18: 4849-4852.

Tucker JL. 2006. Green chemistry, a pharmaceutical perspective. Org Process Res Dev. 10(2): 315-319.

Turnbull JJ, Nakajima J, Welford RW, Yamazaki M, Saito K, Schofield CJ. 2004. Mechanistic studies on three 2-oxoglutarate-dependent oxygenases of flavonoid biosynthesis: anthocyanidin synthase, flavonol synthase, and flavanone 3beta- hydroxylase. J Biol Chem. 279(2): 1206-1216.

Ubukata K, Nonoguchi R, Matsuhashi M, Konno M. 1989. Expression and inducibility in Staphylococcus aureus of the mecA gene, which encodes a methicillin- resistant S. aureus-specific penicillin-binding protein. J Bacteriol. 171: 2882-2885.

Ubukata K, Yamashita N, Konno M. 1985. Occurrence of a beta-lactam-inducible penicillin-binding protein in methicillin-resistant staphylococci. Antimicrob Agents Chemother. 27: 851-857.

Ullan RV, Casqueiro J, Banuelos O, Fernandez FJ, Gutierrez S, Martin JF. 2002. A novel epimerization system in fungal secondary metabolism involved in the conversion of isopenicillin N into penicillin N in Acremonium chrysogenum. J Biol Chem. 277: 46216-46225.

Usher JJ, Lewis MA, Hughes DW, Compton BJ. 1988. Development of the cephalosporin C fermentation taking into account the instability of cephalosporin C. Biotechnol Lett. 10: 543-548.

Usui S, Yu CA. 1989. Purification and properties of isopenicillin N epimerase from Streptomyces clavuligerus. Biochim Biophys Acta. 999: 78-85.

Valegard K, Terwisscha van Scheltinga AC, Dubus A, Ranghino G, Oster LM, Hajdu J, Andersson I. 2004. The structural basis of cephalosporin formation in a mononuclear ferrous enzyme. Nat Struct Mol Biol. 11(1): 95-101.

Valegard K, Terwisscha van Scheltinga AC, Lloyd MD, Hara T, Ramaswamy S, Perrakis A, Thompson A, Lee HJ, Baldwin JE, Schofield CJ, Hajdu J, Andersson I. 1998. Structure of a cephalosporin synthase. Nature. 394: 805-809.

Vanden Boom TJ. 2000. Recent developments in the molecular genetics of the erythromycin-producing organism Saccharopolyspora erythraea. Adv Appl Microbiol. 47: 79-111.

216

Veinberg G, Vorona M, Shestakova I, Kanepe I, Lukevics E. 2003. Design of beta- lactams with mechanism based non-antibacterial activities. Curr Med Chem. 10(17): 1741-1757.

Velasco J, Luis Adrio J, Angel Moreno M, Díez B, Soler G, Barredo JL. 2000. Environmentally safe production of 7-aminodeacetoxycephalosporanic acid (7- ADCA) using recombinant strains of Acremonium chrysogenum. Nat Biotechnol. 18(8): 857-861.

Vicente M, Hodgson J, Massidda O, Tonjum T, Henriques-Normark B, Ron EZ. 2006. The fallacies of hope: will we discover new antibiotics to combat pathogenic bacteria in time? FEMS Microbiol Rev. 30: 841-852.

Villegas-Estrada A, Lee M, Hesek D, Vakulenko SB, Mobashery S. 2008. Co-opting the cell wall in fighting methicillin-resistant Staphylococcus aureus: potent inhibition of PBP 2a by two anti-MRSA beta-lactam antibiotics. J Am Chem Soc. 130: 9212- 9213.

Vollmer W, Holtje JV. 2004. The architecture of the murein (peptidoglycan) in gram- negative bacteria: vertical scaffold or horizontal layer(s)? J Bacteriol. 186: 5978-5987.

Vollmer W, Joris B, Charlier P, Foster S. 2008. Bacterial peptidoglycan (murein) hydrolases. FEMS Microbiol Rev. 32: 259-286.

Waksman SA. 1947. What is an antibiotic or an antibiotic substance? Mycologia. 39(5): 565-569.

Walkty A, Decorby M, Nichol K, Karlowsky JA, Hoban DJ, Zhanel GG. 2008. In vitro activity of ceftobiprole against clinical isolates of Pseudomonas aeruginosa obtained from Canadian intensive care unit (ICU) patients as part of the CAN-ICU Study. J Antimicrob Chemother. 62: 206-208.

Wallace AC, Laskowski RA and Thornton JM. 1995. LIGPLOT: a program to generate schematic diagrams of protein-ligand interactions. Prot Eng. 8: 127-134.

Ward JM, Hodgson JE. 1993. The biosynthetic genes for clavulanic acid and cephamycin production occur as a 'super-cluster' in three Streptomyces. FEMS Microbiol Lett. 110: 239-242.

Watve MG, Tickoo R, Jog MM, Bhole BD. 2001. How many antibiotics are produced by the genus Streptomyces? Arch Microbiol. 176(5): 386-390.

Wegman MA, Janssen MHA, van Rantwijk F, Sheldon RA. 2001. Towards biocatalytic synthesis of β-lactam antibiotics. Adv Synth Catal. 343: 559-576.

Wei CL, Yang YB, Deng CH, Liu WC, Hsu JS, Lin YC, Liaw SH, Tsai YC. 2005. Directed evolution of Streptomyces clavuligerus deacetoxycephalosporin C synthase for enhancement of penicillin G expansion. Appl Environ Microbiol. 71(12): 8873- 8880.

217

Wei CL, Yang YB, Wang WC, Liu WC, Hsu JS, Tsai YC. 2003a. Engineering Streptomyces clavuligerus deacetoxycephalosporin C synthase for optimal ring expansion activity toward penicillin G. Appl Environ Microbiol. 69(4): 2306-2312.

Wei Y, Havasy T, McPherson DC, Popham DL. 2003b. Rod shape determination by the Bacillus subtilis class B penicillin-binding proteins encoded by pbpA and pbpH. J Bacteriol. 185: 4717-4726.

Weinstein MJ, Luedemann GM, Oden EM, Wagman GH, Rosselet JP, Marquez JA, Coniglio CT, Charney W, Herzog HL, Black J. 1963. Gentamicin, a new antibiotic complex from Micromonospora. J Med Chem. 6: 463-464.

Wells JA. 1990. Additivity of mutational effects in proteins. Biochemistry. 29: 8509- 8517.

Whiteman PA, Abraham EP, Baldwin JE, Fleming MD, Schofield CJ, Sutherland JD, Willis AC. 1990. Acyl coenzyme A: 6-aminopenicillanic acid acyltransferase from Penicillium chrysogenum and Aspergillus nidulans. FEBS Lett. 262: 342-344.

Whitney JG, Brannon DR, Mabe JA, Wicker KJ. 1972. Incorporation of labeled precursors into A16886B, a novel β-lactam antibiotic produced by Streptomyces clavuligerus. Antimicrob Agents Chemother. 1: 247-251.

Wilcox MH. 2009. The tide of antimicrobial resistance and selection. Int J Antimicrob Agents. 34(S3): S6-10.

Williamson JM, Inamine E, Wilson KE, Douglas AW, Liesch JM, Albers-Schonberg G. 1985. Biosynthesis of the beta-lactam antibiotic, thienamycin, by Streptomyces cattleya. J Biol Chem. 260: 4637-4647.

Wilmouth RC, Turnbull JJ, Welford RW, Clifton IJ, Prescott AG, Schofield CJ. 2002. Structure and mechanism of anthocyanidin synthase from Arabidopsis thaliana. Structure. 10(1): 93-103.

Wohlgemuth R. 2010. Biocatalysis--key to sustainable industrial chemistry. Curr Opin Biotechnol. 21(6): 713-724.

Woodley JM. 2008. New opportunities for biocatalysis: making pharmaceutical processes greener. Trends Biotechnol. 26(6): 321-327.

Wright AJ. 1999. The penicillins. Mayo Clin Proc. 74: 290-307.

Wu XB, Fan KQ, Wang QH, Yang KQ. 2005. C-terminus mutations of Acremonium chrysogenum deacetoxy/deacetylcephalosporin C synthase with improved activity toward penicillin analogs. FEMS Microbiol Lett. 246(1): 103-110.

Wu XB, Tian XY, Ji JJ, Wu WB, Fan KQ, Yang KQ. 2011. Saturation mutagenesis of Acremonium chrysogenum deacetoxy/deacetylcephalosporin C synthase R308 site confirms its role in controlling substrate specificity. Biotechnol Lett. 33: 805-812.

218

Xie X, Tang Y. 2007. Efficient synthesis of simvastatin by use of whole-cell biocatalysis. Appl Environ Microbiol. 73(7): 2054-2060.

Yang YB, Wei CL, Hsu JS, Tsai YC. 2005. Mutated penicillin expandase and process for preparing 7-ADCA. Patent no.: 6,905,854 (USPTO).

Yeh WK. 1997. Evolving enzyme technology for pharmaceutical applications: case studies. J Ind Microbiol Biotechnol. 19(5-6): 334-343.

Yenugu S, Hamil KG, Birse CE, Ruben SM, French FS and Hall SH. 2003. Antibacterial properties of the sperm-binding proteins and peptides of human epididymis 2 (HE2) family: salt sensitivity, structural dependence and their interaction with outer and cytoplasmic membranes of Escherichia coli. Biochem J. 372: 473-483.

Yim G, Wang HH, Davies J. 2006. The truth about antibiotics. Int J Med Microbiol. 296(2-3): 163-170.

Yong D, Toleman MA, Giske CG, Cho HS, Sundman K, Lee K, Walsh TR. 2009. Characterization of a new metallo-beta-lactamase gene, blaNDM-1, and a novel erythromycin esterase gene carried on a unique genetic structure in Klebsiella pneumoniae sequence type 14 from India. Antimicrob Agents Chemother. 53(12): 5046-5054.

Yoshigi N, Okada Y. 1998. Barley beta-amylase structural gene. Patent no.: 5,726,057 (USPTO).

Zamorano L, Juan C, Fernandez-Olmos A, Ge Y, Canton R, Oliver A. 2010. Activity of the new cephalosporin CXA-101 (FR264205) against Pseudomonas aeruginosa isolates from chronically infected cystic fibrosis patients. Clin Microbiol Infect. 16(9): 1482-1487.

Zapun A, Contreras-Martel C, Vernet T. 2008a. Penicillin-binding proteins and β- lactam resistance. FEMS Microbiol Rev. 32: 361-385.

Zapun A, Vernet T, Pinho MG. 2008b. The different shapes of cocci. FEMS Microbiol Rev. 32: 345-360.

Zhanel GG, Lam A, Schweizer F, Thomson K, Walkty A, Rubinstein E, Gin AS. Hoban DJ, Noreddin AM, Karlowsky JA. 2008. Ceftobiprole: a review of a broad- spectrum and anti-MRSA cephalosporin. Am J Clin Dermatol. 9: 245-254.

Zhang W, Shi Q, Meroueh SO, Vakulenko SB, Mobashery S. 2007. Catalytic mechanism of penicillin-binding protein 5 of Escherichia coli. Biochemistry. 46: 10113-10121.

Zhou J, Kelly WL, Bachmann BO, Gunsior M, Townsend CA, Solomon EI. 2001. Spectroscopic studies of substrate interactions with clavaminate synthase 2, a multifunctional alpha-KG-dependent non-heme iron enzyme: correlation with mechanisms and reactivities. J Am Chem Soc. 123: 7388-7398.

219

Appendices

Appendix I

1. Detailed map of pCR®-Blunt plasmid vector (Invitrogen)

220

2. Detailed map of pGEX-6P-1 plasmid vector (GE Healthcare)

3. Detailed map of pGK plasmid vector (Loke and Sim, 2000)

221

Appendix II

Reagents used for SDS-PAGE analysis (Laemmli, 1970)

1. 10% SDS-polyacrylamide resolving gel mix

The gel mix was prepared as follows:

30% acrylamide mix (BioRad) 33.3mL 1.5M Tris-base (pH 8.8) 20.0mL 10% (w/v) SDS 0.8mL Deionised water 25.6mL 10% (w/v) APS (freshly prepared) 0.4mL TEMED 0.04mL

2. 4% SDS-polyacrylamide stacking gel mix

The gel mix was prepared as follows:

30% acrylamide mix (BioRad) 5.34mL 1M Tris-base (pH 6.8) 10.0mL 10% (w/v) SDS 0.4mL Deionised water 24.0mL 10% (w/v) APS (freshly prepared) 0.2mL TEMED 0.02mL

3. Staining solution

Methanol 50% (v/v) Acetic acid 10% (v/v) Coomasie blue (R250) 0.1% (w/v)

4. Destaining solution

Methanol 10% (v/v) Acetic acid 10% (v/v)

222

Appendix III

Alignment of predicted secondary structures of scDAOCS R306 mutants with that of the wild-type enzyme. The mutants with improved ring-expansion activity were highlighted in yellow. R306A, R306F, R306Y and R306W mutants were predicted to have the same secondary structure as R306M mutant. R306L and R306I mutants were predicted to share the same secondary structure.

223

Alignment of predicted secondary structures of scDAOCS single and double mutants with that of the wild-type enzyme.

224

List of Publications

Journals

(Achieved in this study)

1. Goo KS, Sim TS. 2011. Designing new β-lactams: implications from their targets, resistance factors and synthesizing enzymes. Curr Comput Aided Drug Des. 7(1): 53-80.

2. Goo KS, Chua CS, Sim TS. 2009. Directed evolution and rational approaches in improving Streptomyces clavuligerus deacetoxycephalosporin C synthase for cephalosporin production. J Ind Microbiol Biotechnol. 36: 619-633.

3. Goo KS, Chua CS, Sim TS. 2008. Relevant double mutations in bioengineered Streptomyces clavuligerus deacetoxycephalosporin C synthase result in higher binding specificities which improve penicillin bioconversion. Appl Environ Microbiol. 74: 1167-1175.

4. Goo KS, Chua CS, Sim TS. 2008. A complete library of amino acid alterations at R306 in Streptomyces clavuligerus deacetoxycephalosporin C synthase demonstrates its structural role in the ring-expansion activity. Proteins. 70: 739-747.

5. Chua CS, Biermann D, Goo KS, Sim TS. 2008. Elucidation of active site residues of Arabidopsis thaliana flavonol synthase provides a molecular platform for engineering flavonols. Phytochemistry. 69: 66-75.

(Biocomputational contribution)

6. Chua CS, Low H, Goo KS, Sim TS. 2010. Characterization of Plasmodium falciparum co-chaperone p23: its intrinsic chaperone activity and interaction with Hsp90. Cell Mol Life Sci. 67(10): 1675-1686.

Book

A book chapter titled, “Molecular modification of biosynthetic genes for improved production of penicillins and cephalosporins” has been accepted by Nova Science Publishers to be included in a book titled, “Penicillin: Biosynthesis, Applications and Adverse Effects” on 2nd September 2011.

Posters

1. Goo KS, Seah KI, Tan YH, Tan SHD, Sim TS. 2010. Evaluation of bioconverted cephalosporins produced by a bioengineered streptomycete 225

deacetoxycephalosporin C synthase. In: 11th International Symposium on the Genetics of Industrial Microorganisms, Australia. Abstract 128.

2. Tan YH, Goo KS, Seah KI, Sim TS. 2010. Biotransformation of a Streptomyces clavuligerus hydroxylase (DACS) to a bi-functional enzyme with ring-expansion activity. In: 11th International Symposium on the Genetics of Industrial Microorganisms, Australia. Abstract 80.

3. Goo KS, Tan YH, Sim TS. 2009. Alteration of two C-terminal residues threonine 308 and histidine 310 to arginine confers a hydroxylase Streptomyces clavuligerus deacetylcephalosporin C synthase with ring- expansion activity. In: VIII European Symposium of the Protein Society, Switzerland. Abstract 0550.

4. Goo KS, Sim TS. 2009. Modification of deacetoxycephalosporin C synthase to accept a plethora of hydrophobic penicillins for their bioconversions to expanded cephalosporins. In: BIT's 2nd World Congress of Industrial Biotechnology 2009, Korea. Abstract 43.

5. Goo KS, Sim TS. 2009. Mutagenic augmentation of a Streptomyces clavuligerus hydroxylase, deacetylcephalosporin C synthase, with a new ring- expansion activity. In: BioSysBio Conference 2009, United Kingdom. Abstract 0166.

6. Goo KS, Sim TS. 2008. Alteration of Streptomyces clavuligerus deacetoxycephalosporin C synthase substrate binding pocket to accommodate penicillins with aromatic side chains. In: 17th Meeting of Methods in Protein Structure Analysis, Japan. Abstract P-054. (selected for Young Scientist Talks Y-06)

7. Chua CS, Goo KS, Tan DSH, Sim TS. 2007. Identification of potential candidates for chimeric enzyme construction using Dragon Streptomyces Explorer. In: 6th International Symposium on Industrial Microbiology & Biotechnology, USA. Abstract pg 38.

8. Goo KS, Chua CS, Sim TS. 2007. Understanding the molecular basis of enhanced catalytic ring-expansion of ampicillin by Streptomyces clavuligerus deacetoxycephalosporin C synthase. In: 6th International Symposium on Industrial Microbiology & Biotechnology, USA. Abstract pg 39.

9. Chua CS, Biermann D, Goo KS, Sim TS. 2006. Potential exploitation of computational tools to elucidate the functional active sites of flavonol synthase from Arabidopsis thaliana for quercetin production. In: The 18th Annual Meeting of the Thai Society for Biotechnology, Thailand. Abstract VII-P-9.

226

10. Biermann D, Goo KS, Chua CS, Tan DSH, Sim TS. 2006. Mutational evidence for Citrus unshiu (Satsuma mandarin) flavonol synthase as a 2- oxoglutarate dependent dioxygenase. In: The 18th Annual Meeting of the Thai Society for Biotechnology, Thailand. Abstract X-P-2.

11. Goo KS, Sim TS. 2005. Combining C-terminus N304R and R306L mutations in Streptomyces clavuligerus deacetoxycephalosporin C synthase improves its substrate binding affinity leading to higher catalytic efficiency. In: Combined Scientific Meeting, Singapore. Abstract P26.

12. Goo KS, Chin HS, Seah KI, Sim TS. 2005. Rational engineering of Streptomyces clavuligerus deacetoxycephalosporin C synthase for higher efficiency and expanded substrate specificity. In: 5th International Symposium on Industrial Microbiology & Biotechnology, China. Abstract pg 114.

227