University of Groningen Evolution of enantioselectivity Boersma, Ykelien Lolkje IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2007 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Boersma, Y. L. (2007). Evolution of enantioselectivity: selection of improved hydrolase variants. s.n. Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date: 13-07-2020
Transcript
University of Groningen
Evolution of enantioselectivityBoersma, Ykelien Lolkje
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.
Document VersionPublisher's PDF, also known as Version of record
Publication date:2007
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):Boersma, Y. L. (2007). Evolution of enantioselectivity: selection of improved hydrolase variants. s.n.
CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.
CHAPTER 1: Introduction & aim of the thesis 9 CHAPTER 2: Selection strategies for improved biocatalysts 21 CHAPTER 3: Comparison and functional characterisation
of three homologous carboxylesterases of Bacillus subtilis 41
CHAPTER 4: Phage display of an intracellular carboxylesterase
of Bacillus subtilis: a comparison of the Sec and Tat pathway export capabilities 59
CHAPTER 5: Directed evolution of Bacillus subtilis
lipase A using enantiomeric phosphonate inhibitors: crystal structures and phage display 75
CHAPTER 6: A novel selection system for enantioselectivity
of Bacillus subtilis lipase A based on bacterial growth 93
CHAPTER 7: Rational design of Bacillus subtilis
lipase A loop hybrids: insertion of a lid structure inverts enantioselectivity 107
CHAPTER 8: A validated gas chromatographic method
for the evaluation of enantioselectivity in kinetic resolution experiments 123
CHAPTER 9: Summary, general discussion & perspectives 133 References 143 Nederlandse samenvatting 161 Dankwoord 167 List of publications 173
1
INTRODUCTION
&
AIM OF THE THESIS
INTRODUCTION & AIM OF THE THESIS
11
Chiral drugs The term optical activity is derived from the interaction of chiral materials with polarised light. A solution of the (-)-form of an optical isomer rotates the plane of polarisation of a beam of monochromatic light in a counter-clockwise direction, vice-versa for the (+) optical isomer. This property was first observed by Jean-Baptiste Biot in 1815. It gained considerable importance in the sugar industry and analytical chemistry, and later on in the pharmaceutical industry. Biot’s student Louis Pasteur discovered in 1848 that the phenomenon of optical activity had a molecular basis. Today, the definition of chiral molecules with a single chiral carbon atom states that they are so-called enantiomers whose structural mirror images are non-superimposable upon one another. Enantiomers have the same physicochemical properties, though behave differently under certain conditions 1.
Chirality in drug molecules is rapidly growing as an important aspect in drug design, research, and development. Enantiomers have different therapeutic activities and can interact differently with chiral biomolecules, such as receptors. Chirality thus introduces selectivity, and often specificity, in drug action 2,3. Recent advances in chiral technology and the ability to synthesise enantiomerically pure compounds, together with stricter regulations from the Food and Drug Administration (FDA) in the USA, have led the pharmaceutical industry to attempt to develop new chemical entities as single enantiomers 4. In parallel, there has been an interest in ‘chiral switches’, the replacement of an already approved racemic mixture of both enantiomers of a drug by a single enantiomer 5. Still, despite the knowledge that a single enantiomer can cause the therapeutic effect, many drugs are administered as a racemic mixture as there is no means of separating both enantiomers available. However, the rise of biotechnology has brought novel solutions for chiral separations.
This thesis focuses on the use of bacterial enzymes from Bacillus subtilis to separate two enantiomers from a racemic mixture. A survey of different drug areas reveals that chirality is relatively common among drugs in general with over 50% of the drugs being chiral 6 (figure 1).
Cardiovascular
OtherVaccines
Analgesics
Ophthalmic
RespiratoryAntiviral
Hormones/Endocrinology
Haematology
Dermatological
Gastrointestinal
Central nervous system
Cancer
Antibiotics/Antifungals
Figure 1: The distribution of chiral drugs over different classes of drugs 6.
CHAPTER 1
12
Chirality has also been recognised as an important factor in the cardiovascular field. Among the worldwide sales of all pharmaceutical agents, cardiovascular drugs possess great potential: e.g. total sales of cardiovascular drugs in 2000 were $46.6 billion, whereas single-enantiomer drugs sales were $26.9 billion of total revenues 3. Therefore, the substrates of this research, β-adrenergic receptor antagonists (β-blockers), represent a highly relevant class of chiral molecules (figure 2).
O NH
CH3
CH3
OH
NH
CH3
CH3OHNH
SCH3
O
O
NH
O NH
O
OCH3
OH
O
O
CH3
CH3
OH
*
Propranolol
*
Sotalol
*
Carvedilol
*
1,2-O-isopropylidene-sn-glycerol
Figure 2: Structural formulas of the racemic β-adrenergic receptor blockers propranolol, sotalol and carvedilol, and their racemic precursor 1,2-O-isopropylidene-sn-glycerol. These drugs are mostly used to treat cardiovascular disorders such as hypertension, cardiac arrhythmia, and ischemic heart disease. Each of these drugs contains at least one chiral center, and a high degree of enantioselectivity in binding to the β-adrenergic receptor. Amongst chiral drugs, the β-adrenergic receptor blocking drugs are one of the best-understood classes from the perspective of stereoselectivity in pharmacokinetics and pharmacodynamics. Generally, the cardiac β-blocking activity resides in the S-(-)-enantiomer 7. Sotalol is an exception: it has an R-(-) and S-(+) conformation, in which the R-(-)-enantiomer has the β-blocking activity. Both enantiomers of sotalol share an equivalent degree of class III antiarrhythmic potency 8. Carvedilol is a relatively new agent that is marketed as a racemate. Similar to other β-adrenergic antagonists, the S-(-)-enantiomer is more potent as an antagonist of the β-receptor 9. Industrial biocatalysis As stated previously, the regulatory authorities now impose a much stricter requirement for enantiopurity. Consequently, there is a great need for cost-effective processes for the synthesis of enantiopure (precursors of) pharmaceutical compounds. Biocatalysis has emerged as an important tool in the industrial synthesis of pharmaceuticals and
INTRODUCTION & AIM OF THE THESIS
13
pharmaceutical intermediates. Enzymes can perform intricate regioselective and/or enantioselective biochemical transformations and can accelerate reaction rates by enormous factors. High enzymatic selectivity also allows efficient reactions with few by-products, thereby making enzymes an environmentally friendly alternative to conventional chemical catalysts 10-13. However, the number and diversity of enzymatic applications are modest, perhaps in part because of perceived or real limitations of biocatalysts, such as limited enzyme availability, substrate scope, and operational stability 14. To date, successful applications of biocatalysts have been largely confined to hydrolytic enzymes such as lipases and esterases 15-17. Lipases The commercial use of microbial lipases, of both bacterial and fungal origin, comprises many different applications. These enzymes play an important role in the field of detergents, as well as in the production of food ingredients. Other applications are the removal of pitch from pulp in the paper industry, and as biocatalysts in the field of fine chemicals which are used in the pharmaceutical industry, in flavour development, and in synthetic organic chemistry 18-21. The great interest in lipases is mainly due to their favourable properties in terms of enantioselectivity, regioselectivity and broad substrate specificity.
Two criteria have been used to classify a bacterial lipolytic enzyme as a true lipase (E.C. 3.1.1.3): first, the enzyme should be activated in the presence of a water-lipid interface; the activity should increase in the presence of the emulsified triglyceride substrate. This phenomenon is termed interfacial activation 22. Second, the enzyme should contain a lid, a surface loop of the protein controlling access to the active site 23-25. However, these criteria appeared to be unsuitable for classification, mainly because of the existence of lipolytic enzymes without a lid structure, or not exhibiting interfacial activation, such as B. subtilis lipase A and Fusarium solani cutinase 26,27. Today, lipases are defined as carboxylesterases catalysing the hydrolysis of long-chain acyl glycerols. Glycerol esters with an acyl chain length of more than ten carbon atoms can be regarded as lipase substrates 28,29 (figure 3).
OH
O
O
O
O
O
O
OOH
OH
OH
Lipase+ 3 H2O + 3
Figure 3: Lipase catalysed hydrolysis of a triglyceride 30. Some 47 different bacterial lipases were identified based upon amino acid sequence homology. They were grouped into eight different families. Family I consists of the ‘true’ lipases, such as Pseudomonas lipases and lipases from Bacillus and Staphylococcus. The second group, family II, does not show the conventional pentapeptide Gly-Xaa-Ser-Xaa-Gly but rather displays a Gly-Asp-Ser-(Leu) motif. In this family, lipases from Streptomyces scabies, from Pseudomonas aeruginosa, and from Salmonella typhimurium
CHAPTER 1
14
are found. In family III, lipases from Streptomyces albus and Moraxalla are found. Family IV is comprised of enzymes that display a striking amino acid sequence similarity to mammalian hormone-sensitive lipase. In family V, enzymes originating form mesophilic bacteria as well as from cold-adapted bacteria are found, while family VI represents the smallest lipolytic enzymes known, with molecular masses in the range of 23-26 kDa. In contrast, family VII is comprised of a number of rather large bacterial esterases. Finally, in family VIII three enzymes that show a striking similarity to several class C β-lactamases are found 31.
Though bacterial lipases are classified in these eight families, they show high structural homology. Several 3D structures have been elucidated and they all show a characteristic folding pattern, the so-called α/β hydrolase fold 32 (figure 4). The lipase core is composed of a central β-sheet consisting of up to eight different β-strands, connected by up to six α-helices. The activity relies mainly on a catalytic triad usually formed by a serine, an aspartate (or glutamate), and a histidine residue in that order. The serine residue usually appears in the conserved pentapeptide Gly-Xaa-Ser-Xaa-Gly, forming a characteristic β-turn-α-motif, the so-called nucleophilic elbow.
Figure 4: Schematic overview of the canonical α/β hydrolase fold 32. Substrate hydrolysis starts with a nucleophilic attack by the serine residue on the carbonyl carbon atom of the ester bond, thereby forming a tetrahedral intermediate stabilised by hydrogen bonding to nitrogen atoms of the main chain. These residues form the so-called oxyanion hole, a spatial arrangement of hydrogen bond donors that stabilises the transition state of the catalysed reaction. An alcohol is liberated, leaving behind an acyl-lipase complex, which is finally hydrolysed with liberation of the fatty acid and regeneration of the enzyme 25,33 (figure 5).
INTRODUCTION & AIM OF THE THESIS
15
O
Ser
O
RO
O RO
H
HO
HHO
O
O
OAsp
NNH
His
O
OAsp
O
Ser
HNN
H
His
OR
O
N
H
N
H
O
OAsp
NN+
H
His
H O
Ser
O OH
N
H
N
H
O
OAsp
NN+
H
His
H O
Ser
Product ISubstrate
Acyl enzymeFree enzyme
Product II
Oxyanion hole
Figure 5: Hydrolysis of a butyrate ester by an α/β hydrolase fold enzyme. Most industrially applied lipases require specific chaperones for activation, which makes it difficult to express them in a heterologous host. The use of a homologous host is usually undesirable, as most strains are not regarded as ‘generally recognised as safe’ (GRAS) organisms. However, lipase A of Bacillus subtilis is an important exception: it does not require a specific chaperone and it can be overproduced in the GRAS organism Bacillus 27. Lipases of Bacillus subtilis will be discussed more in detail in the next paragraph. Lipases from Bacillus subtilis The Gram-positive bacterium Bacillus subtilis produces a number of extracellular enzymes, among which are at least two different lipolytic enzymes, lipase A (LipA) and lipase B (LipB) 34. The lipA gene was first cloned in the homologous host and sequenced by Colson et al in 1992. Purification of the enzyme from the culture supernatant yielded an enzyme of 181 amino acids with a molecular weight of 19.3 kDa 35. Remarkably, the first glycine in the conserved pentapeptide is replaced by an alanine. LipA revealed to have maximal stability at alkaline pH up to pH 12, with an optimum at pH 10. Its catalytic triad consisted of S77, D133 and H156. The enzyme was capable of hydrolysing p-nitrophenyl esters with a preference towards caprylate esters 27,35,36 (figure 6).
CHAPTER 1
16
A) B)
Figure 6: 3D structure of Bacillus subtilis lipase A 27. A) Side view of the structure, which shows six β-sheets (blue) surrounded by five α-helices (green). The active site residues S77, D133 and H156 are shown in black. B) Frontal view of the catalytic site of lipase A. After sequencing the complete genome of Bacillus subtilis, a homologue of LipA was found 37. The corresponding enzyme was classified as an esterase; this esterase, LipB, had a sequence identity of 74% to LipA. The lipB gene encodes an enzyme that consists of 182 amino acids with a calculated molecular mass of 19.5 kDa. Characterisation of the enzyme revealed that LipA and LipB had similar specific activities towards p-nitrophenyl esters and triacyl glycerides with a maximum activity towards tricaprylin. Although a poor substrate to LipA, LipB was completely unable to hydrolyse the typical lipase substrate triolein 34,38. Esterases Esterases (E.C. 3.1.1.1) represent a diverse group of hydrolases catalysing the cleavage and formation of ester bonds. In general, esterases show specificity for either the alcohol or the acid moiety, not for both 39. The 3D structures of carboxylesterases show the characteristic α/β hydrolase fold, with a catalytic triad composed of serine, aspartate and histidine. Usually, the conserved pentapeptide Gly-Xaa-Ser-Xaa-Gly is found as well. The reaction mechanism is the same as for lipases. Like lipases, esterases have been shown to be stable and active in organic solvents, though this feature is more pronounced with lipases. However, the main difference with lipases is the substrate specificity: esterases accept preferentially esters with an acyl side chain shorter than ten carbon atoms as a substrate; esterase activity is found to be highest towards more water soluble substrates. Alternatively, lipases are perfectly capable of hydrolysing these esterase substrates as well 29,40. Esterases and lipases can also be distinguished by a pH-dependent electrostatic ‘signature’. The active site of lipases displays a negative potential in the pH-range associated with their maximum activity (around pH 8); whereas esterases show a similar pattern, but around pH 6, which correlates with their usually lower pH-activity optimum 39.
INTRODUCTION & AIM OF THE THESIS
17
Esterases are widely distributed in animals, plants as well as microorganisms. Several esterase encoding genes have been cloned and overexpressed in suitable hosts. However, these enzymes scarcely meet an industrial application, mainly because of their limited commercial availability and frequently observed moderate enantioselectivity 41. The most promising microbial biocatalysts seem to originate from Bacillus and Pseudomonas species 40. Bacillus esterases will be discussed more in detail in the next paragraph. Esterases from Bacillus subtilis The best studied esterase is most likely carboxylesterase NP from Bacillus subtilis Thai I-8. This enzyme was characterised as a very efficient enantioselective biocatalyst for the kinetic resolution of non-steroidal anti-inflammatory drug (NSAID) esters 42 (figure 7).
OCH3
CH3
OCH3
O
OCH3
CH3
OCH3
OOCH3
CH3
O
O
Esterase
*
+
Naproxen methyl ester
S-naproxen R-naproxen methyl ester Figure 7: The kinetic resolution process of naproxen. The R-naproxen methyl ester can be recycled into the process. The esterase shows modest selectivity towards chiral alcohols, which is in sharp contrast to its enantioselectivity towards chiral carboxylic acids. Besides naproxen, various other 2-aryl propionic acids are produced with high enantioselectivity 43,44. Carboxylesterase NP has a molecular mass of 32 kDa and is produced as an intracellular protein; its structure is still unknown.
In Bacillus subtilis 168, nine genes have been characterised as encoding for intracellular or extracellular esterases. Furthermore, eight other genes can be regarded as potential esterases. Two of these esterases, carboxylesterase A (CesA, encoded by the cesA gene) and carboxylesterase B (CesB, encoded by the cesB gene), showed a high homology towards carboxylesterase NP of Bacillus subtilis Thai I-8 (98% and 64% identity on protein level, respectively). Both enzymes have been characterised; CesA seems to be more suitable for the enantioselective production of chiral carboxylic acids, such as NSAIDs,
CHAPTER 1
18
whereas CesB can find application in the enantioselective production of the chiral alcohol 1,2-O-isopropylidene-sn-glycerol 45. Directed evolution Mutagenesis Most applications of enzymes in biocatalysis do not rely on the natural reaction catalysed by them but rather use non-natural substrates. Furthermore, the reaction system can differ substantially as well from the environment in which the enzymes have evolved in nature. Thus, quite often activity, stability, substrate specificity and enantioselectivity need to be improved 46,47. Rational design, a technique which relies on extensive knowledge of structure-function relationships of enzymes, used to offer the only possibility for the creation of new enzyme properties and functions. Several successes have been reported 48-
51. Yet during the past decade, directed evolution has emerged as a powerful means to overcome the limitations of the use of wild type enzymes for industrial biocatalysis 52-54. The enzyme’s properties and functions can easily be engineered without any required knowledge of the structure. Directed evolution has been successfully applied to several distinct enzymes to alter their characteristics 55,56. The technique is essentially composed of two steps: first, mutagenesis of the gene(s) encoding the enzyme(s), and second, identification of the desired biocatalyst variants within these mutant libraries by either screening or selection 46. The gene(s) encoding for the improved variants are then used to parent the next round of directed evolution. Thus, the ultimate goal of directed evolution is to accumulate improvements through repetitive rounds of mutagenesis and identification (figure 8).
Figure 8: A schematic overview of a directed evolution experiment. To date, a great number of techniques are available to mutate the gene(s) of interest. These can be roughly divided into two categories: those based on point mutations and those making use of recombination. The first method based on point mutations was error-prone PCR (epPCR) 57. In this technique, a thermostable polymerase without proofreading ability is used, which incorporates wrong nucleotides during the extension of the new DNA strand 58,59. In addition to epPCR, the generation of focused mutant libraries also provides the user with a useful tool in directed evolution. Certain hot spots or hot regions, positions in the amino acid sequence believed to be important with respect to a given property, are subjected to randomisation by saturation mutagenesis or cassette mutagenesis 56. In principle, these methods should lead to a mutant library containing all 20 different mutants corresponding to the 20 different amino acids at a defined position.
INTRODUCTION & AIM OF THE THESIS
19
The second method, based on recombination, is based on the breaking and rejoining of DNA in new combinations. Many different versions of recombination exist, DNA shuffling being the most prominent one 60. In general, DNA shuffling allows beneficial mutations from multiple genes to be recombined. A number of methods, based on this principle, have been developed 61. Selection The second step in directed evolution experiments, identification of interesting variants by screening or selection, remains the most critical one. Suitable assay screening methods should enable a fast, accurate and targeted identification of desired biocatalysts out of libraries comprising 104-106 mutants. The main advantage of selection over screening is that many more variants in the library can be analysed simultaneously. Selection strategies exploit conditions favouring the exclusive survival of desired variants, thus uninteresting variants are never seen. Thus, the evolutionary character of the overall process is stimulated, which makes the directed evolution approach rational in a different sense. Consequently, selecting enzyme variants is much faster and can be carried out with higher throughput 59,62.
Several strategies can be applied to select for improved variants. Enzymes are analysed on the basis of the properties of interest, but as they cannot amplify themselves, the selection system must simultaneously select for the genes encoding them as well. These genes are then amplified and subjected to further rounds of mutagenesis, selection and amplification. Thus phenotype and genotype must be linked in a physical way 63. Selection systems such as phage display, cell surface display, ribosome display and growth selection establish this linkage. These will be discussed in more detail in chapter 2.
CHAPTER 1
20
Aim of the thesis This thesis describes the application of lipase A and carboxylesterase A from B. subtilis 168 for the production of enantiopure IPG out of its racemic ester. IPG is the chiral starting compound in the production of β-adrenergic receptor antagonists; to obtain an enantiopure product, synthesis must start out with enantiopure IPG. To improve the yield of enantiopure IPG, we have applied directed evolution to improve the enantioselectivity of the wild type enzymes to make them more suitable as (industrial) biocatalysts.
First, cassette mutagenesis was employed to mutate specific regions of the lipA gene. Mutants were selected with a novel phage display system based on dual selection using phosphonate suicide inhibitors. In an alternative approach, a dual selection system based on bacterial growth was developed, consisting of a mimic substrate as well as a phosphonate suicide substrate. This system provides a means to select not only on enantioselectivity, but on catalytic efficiency as well. Using both systems, novel enantioselective biocatalysts were selected and characterised.
In a second, rational approach, lid-like structures were designed and inserted in the lipase, as lipase A is one of the few lipases without a lid. It was therefore tempting to determine the enantioselectivity of lipase A with such a lid-like structure and improve these variants further by saturation mutagenesis on predefined hotspots and select mutants by making use of the growth selection system.
2
SELECTION STRATEGIES FOR
IMPROVED BIOCATALYSTS
YKELIEN L. BOERSMA, MELLONEY J. DRÖGE & WIM J. QUAX
FEBS J 2007; 274: 2181-2191
SELECTION STRATEGIES FOR IMPROVED BIOCATALYSTS
23
Selection Strategies for Improved Biocatalysts Enzymes have become an attractive alternative to conventional catalysts in numerous industrial processes. However, their properties do not always meet the criteria of the application of interest. Directed evolution is a powerful tool to adapt the enzyme’s characteristics; nevertheless, to select evolved variants remains a critical step. As a consequence, new selection strategies have been developed during the past decades enabling selection for the desired enzymatic activity. This review focuses on these novel strategies for the selection of enzymes from large libraries, in particular of those enzymes that are employed in the synthesis of pharmaceutical intermediates and pharmaceuticals.
Introduction Directed evolution Over the past decades, it has become clear that enzymes hold a great potential for industry. Enzymes are among the most remarkable biomolecules known because of their extraordinary specificity and catalytic power 58. The specificity and (enantio- and regio-) selectivity of certain enzymatic transformations makes them appealing for the production of fine chemicals and pharmaceutical intermediates. To date, more than 500 products over a wide spectrum of applications are manufactured by enzymes. Well-known examples are ephedrine, aspartame and amoxicillin 14,64. However, in many cases naturally occurring enzymes lack features necessary for the application of interest, since after all they have evolved in nature to serve a different purpose than the acceleration of industrial processes. To overcome these limitations, directed evolution has emerged as a powerful and versatile means to tailor enzymes in order to adapt their properties to process requirements. It mimics the process of Darwinian evolution in the test tube, combining mutagenesis and recombination with selection or screening for improved variants with the desired characteristics 46,47. The main advantage is that the enzyme’s properties and functions can easily be engineered even without any required knowledge of the structure.
Figure 1: Schematic overview of a directed evolution experiment. The technique is essentially comprised of two steps: first, mutagenesis of the gene(s) encoding the enzyme(s) 46. Enough diversity should be created in the starting gene, such that an improvement in the desired characteristic of the protein will be found in a library of variants. Second, the variants are analysed on the basis of the properties of interest by either
CHAPTER 2
24
screening or selection. The gene(s) encoding the improved variants are identified and then used to parent the next round of directed evolution 65. Thus, the ultimate goal of directed evolution is to accumulate improvements through repetitive rounds of mutagenesis and identification (figure 1). Directed evolution has been successfully applied to several distinct enzymes to alter their characteristics 55,56,58,61.
To date, a broad range of strategies to introduce mutations into the starting gene(s) are available. Moreover, methods for constructing diverse molecular libraries continue to accumulate. These can be roughly divided into two categories: those based on point mutations and those making use of recombination. Belonging to the first category are methods such as oligonucleotide-directed randomisation and error-prone polymerase chain reaction (epPCR). In principle, these methods lead to a mutant library possibly containing any of the 20 different amino acids at a defined position (for a review, see 56,66-68). The second method, based on recombination, is based on the breaking and rejoining of DNA in new combinations. Many different versions of recombination exist, DNA shuffling being the most prominent one 60. In general, DNA shuffling allows beneficial mutations from multiple genes to be recombined. A number of methods, based on this principle, have been developed 61,67.
The second step in directed evolution, identification of the improved variants by either screening or selection, remains the most critical one. This review will focus on some of the strategies to be applied in this step. The choice to apply either screening or selection is discussed in the next paragraph. The second step in directed evolution The advantage of selection over screening Assays capable of rapidly isolating rare valuable variants from a large mutant library are of key importance to the task of evolving proteins in the laboratory, and need to be tailored for every enzyme and reaction. Naturally, when more variants are analysed, the odds of finding a valuable variant within the library are increased. However, the typical library size is still many orders of magnitude larger than the number of protein variants that can be screened. Directed evolution experiments are therefore often limited by the availability of a suitable high-throughput screening or selection system 69,70.
At the basis of all screening and selection strategies is a linkage between the gene, the enzyme it encodes and the product of the activity of the enzyme. The main difference between screening and selection is that screening is performed on individual genes or clones, whereas selections are performed simultaneously on the entire pool of variants 69,71. Screening and selection methods should meet certain criteria. First, if possible they should be directly associated with the property of interest. After all, you get what you select for is the first rule of directed evolution. The substrate should be identical or at least as close as possible to the target substrate, and product detection should be under multiple turnover conditions to ensure selection of effective catalysts. Second, the assay should be sensitive over the desired dynamic range. In the first rounds of directed evolution experiments, all improved mutants should be recovered, while in more advanced rounds more stringent
SELECTION STRATEGIES FOR IMPROVED BIOCATALYSTS
25
conditions need to be applied to isolate only the best variants. Finally, the screening or selection procedure should be applicable in a high-throughput format 69.
The advantage of screening is that the difference between substrate and product of an enzymatic reaction can be determined directly or indirectly in almost every case. Most screening methods rely on the use of fluorogenic or chromogenic substrates, which are converted in spectroscopically different products 12,69,72. Recently, FACS analysis has emerged as an important high-throughput means for detection 73. However, the disadvantage of screening can be put down to the fact that every single mutant must be tested for the desired enzymatic reaction, even variants that might not be active or are incorrectly folded. In a directed evolution experiment using recombination, these non-active enzymes are typically 50-80% of the total library. Therefore, selection is preferred over screening 12.
In general, selection techniques are less labour-intensive and more efficient than screening techniques as they allow for the analysis of more mutants simultaneously. Screening limits the number of individual library members to be analysed to roughly 104 variants, whereas in selection strategies the library size is extended to 1010 up to 1013 variants 11,74. Selection strategies exploit conditions favouring the exclusive survival of desired variants, therefore uninteresting variants are never seen. Thus, the evolutionary character of the overall process is stimulated, which makes the directed evolution approach rational in a different sense. However, with every selection system the possibility always exists that viable but unanticipated variants will surface. If these false positives become too abundant, an efficient screening step or a redesign of the selection procedure may be necessary 62. One drawback of selections for enzymatic activity though is the required substantial up-front customisation or the availability of relevant auxotrophs in genetic complementation 62,70,75.
This review discusses some of the genotype-phenotype linkages possible for the selection of interesting enzyme variants. The focus will be on selection for pharmaceutically relevant substrates. In vivo and in vitro selection systems The genotype-phenotype linkage can be acquired in different ways. A cell-type linkage can be created by compartmentalising the protein and the gene together in cells. All in vivo selection systems have in common that the genetic library obtained in the first step of directed evolution must be transformed into cells, either bacteria or yeast. The selection is based on the presence of a catalytic activity which provides a growth advantage to microorganisms possessing that specific activity. Thus, cell-type linkage methods have been used mainly for the selection of catalytic enzymes. However, low transformation efficiency or cells circumventing selection pressure or lack of transport of the substrate are some of the limitations of in vivo selection 11,69,71.
These limitations can be overcome by in vitro selection techniques. With these systems, no transformation step is necessary. Transformation efficiency is therefore no longer a limiting factor. Thus, library sizes can be much larger, up to 1014 members, enabling the exploration of a larger fraction of sequence space. Furthermore, these selections can be performed under more stringent conditions, as selection is not dependent on viable cells. Thus, nonphysiological conditions such as elevated temperatures, extreme pH or even organic
CHAPTER 2
26
media can be applied. As in vitro selection does not require living cells membrane barriers between the enzyme and the substrate are non-existing. Examples of (partially) in vitro selection techniques are phage display, ribosome display, cell surface display and in vitro compartmentalisation (IVC) 76-78.
In this review, both in vivo and in vitro techniques for the selection of enzymatic activity and some examples of their applications will be discussed.
In vivo selection for enzymatic activity In vivo selection links cell survival to enzyme activity. This strategy is termed genetic or growth selection. Historically, genetic selection has been widely used to identify biosynthetic genes and pathways. The general strategy for genetic selection involves the introduction of a metabolic requirement for the desired activity into the host cells (figure 2A). Plasmids, encoding for a mutant library of the protein of interest, are introduced into a suitable host for selection, preferably a mutant strain of a well-characterised bacterium, such as Escherichia coli. Selective conditions for the target function of the protein encoded by the plasmid are imposed in such a way that only those cells expressing variants with the desired phenotype are viable. By means of the enzymatic activity that is used for selection, usually an essential nutrient is provided to compensate for a deficiency in the strain used, but the product can also enable cell survival from increasing concentrations of toxic compounds by neutralising them. These selected variants can be further characterised after selection 79.
Figure 2: A) General strategy for genetic selection. In the bacterium, the plasmid DNA is transcribed and translated to the enzyme, which can in its turn convert the substrate into the nutrient. B) Strategy for chemical complementation, as performed by van Sint Fiet et al 79. After enzymatic conversion, the product of the transcriptional activator NahR, which activates TetA, enables cell growth on selective minimal medium.
SELECTION STRATEGIES FOR IMPROVED BIOCATALYSTS
27
Prokaryotic selection systems offer a number of advantages over selection in eukaryotes. Transformation efficiencies are much higher in prokaryotes and cell division times are much shorter than for eukaryotes; besides that, prokaryotic genomes are less complex 80.
Genetic complementation Genetic selections have been rated extremely valuable for evolving enzymes with improved catalytic activity, improved stability, and altered specificity. In literature, many examples of successful genetic selections have been described 62,74,81-88. In this review, we will mainly focus on the experiments that have been developed for the selection of enzymes by growth. In particular, enzymes that can be used for the synthesis of pharmaceuticals and pharmaceutical intermediates will be discussed.
DeSantis et al broadened the substrate specificity of E. coli 2-deoxyribose-5-phosphate aldolase (DERA), which is a unique enzyme amongst the aldolase family as it catalyses the reversible condensation of two aldehydes. DERA is an attractive biocatalyst as it accepts a broad range of substrates, with a strong preference for phosphorylated substrates. An important application of this enzyme is found in the synthesis of the antitumour agent Epothilone A and the enzyme might also be used in the synthesis of precursors of statins 89,90. To expand and improve DERA’s activity towards nonphosphorylated substrates, the so-called E. coli SELECT was engineered auxotrophic for acetaldehyde. Afterwards, this E. coli strain was transformed with a DERA mutant library and grown on minimal medium supplemented with the non-phosphorylated unnatural substrate D-2-deoxyribose. This substrate can only be converted to acetaldehyde by DERA variants. The selection and identification of novel DERA variants using this system is currently in progress 90,91.
Besides selection for altered substrate specificity, genetic selection can also be applied for the selection of enzymes with improved catalytic activity. The ultimate goal of Otten et al was to convert the glutaryl acylase from Pseudomonas SY-77 into an adipyl acylase by directed evolution. This enzyme could be applied in the one-step bioconversion of adipyl-7-aminodesacetoxycephalosporanic acid (adipyl-7-ADCA) to 7-ADCA, which is of key importance for the synthesis of semi-synthetic cephalosporins. After constructing an epPCR mutant library, genetic selection was applied by using adipyl-leucine as a mimic substrate and a leucine auxotrophic E. coli strain as selection host. Consequently, only enzymatic hydrolysis of adipyl-leucine would allow for growth. Variants with an improved growth capability on the mimic substrate also showed an improved activity towards the real β-lactam substrate. Selected mutants demonstrated a nearly 10-fold improved ratio of adipyl-7-ADCA over glutaryl-7-ACA hydrolysis. Expansion of the acyl binding pocket of the enzymes formed the explanation for this improvement 92.
Improvement of enantioselectivity is difficult to select for. Hwang et al proposed a selection system for enantioselectivity based on differential cell growth. The principle is based on toxic product formation using chiral antibiotic esters. Upon enantioselective hydrolysis by commercially available esterases and lipases, the antibiotic was released and cell growth was inhibited. Thus, the difference in cell density is directly correlated with the enantioselectivity 93. Recently, we have developed a genetic selection system using an aspartate auxotroph E. coli strain to select for both enantioselectivity and enzymatic activity. The system was applied to the selection of hydrolase variants such as Bacillus subtilis lipase A. Using this
CHAPTER 2
28
system, we were able to select mutants with inverted and improved enantioselectivities towards the chiral synthon 1,2-O-isopropylidene-sn-glycerol (IPG), an important precursor in the synthesis of β-adrenergic receptor antagonists (unpublished data). Chemical complementation Genetic complementation, as described in the previous section, has proven to be a powerful approach to enzyme evolution. However, these complementations are often limited by the (natural) reactions that have to be used for selection. Therefore, a chemical complementation strategy was developed, in which complementation is solely dependent on the product that is formed and not on the reaction pathway itself 94. In this approach, enzymatic activity is linked to the transcription of an essential gene using a yeast three-hybrid system. The enzyme substrate acts as an inducer for dimerisation to reconstitute a DNA-binding and transcriptional activation domain of an artificial transcription factor. Thus, enzymatic activity can result in the activation or repression of expression of either an essential or a toxic reporter gene. The system enables selection for biocatalytically active cells from a background of inactive cells. Baker et al explored the cephalosporin hydrolysis by a β-lactam hydrolase from Enterobacter cloacae P99 as a model reaction. They linked the enzymatic activity of this enzyme to transcription of a lacZ reporter gene in vivo 94,95. The utility of this system has also been demonstrated for glycosynthase from Humicola insolens, yielding a variant with a five-fold increase in glycosynthase activity 96.
Witholt et al developed a selection system using an engineered E. coli strain which detects the production of benzoate and 2-hydroxybenzoate from their corresponding aldehydes catalysed by benzaldehyde dehydrogenase (XylC) of Pseudomonas putida. Detection was carried out by a mutant of the transcriptional activator protein NahR from P. putida. This mutant specifically recognises the products and not the aldehyde substrates. Furthermore, it activates transcription from its cognate salicylate promoter. Genes encoding either a fragment of β-galactosidase or the tetracycline antiporter TetA were cloned behind this promoter to enable the selection of biocatalytically active cells producing XylC. On selective minimal medium plates containing tetracycline E. coli could only grow after addition of the aldehyde substrates (figure 2B). As this system depends merely on the synthesis of the product and not on its reaction pathway, several different enzymatic reactions can be monitored. This system was successfully used to detect nitrilase, amidase, aldehyde oxidase, and aldehyde dehydrogenase activities yielding differently substituted benzoates 79,97. In vitro selection for enzymatic activity Similar to in vivo selection systems, in vitro selection systems also link genotype and phenotype. Examples of these selection strategies are systems which display variants on the surface, such as phage display, cell surface display and ribosome display. Surface display allows unhindered accessibility of the substrate as well as reaction conditions of choice. These systems will be discussed in the next sections, together with their applications.
SELECTION STRATEGIES FOR IMPROVED BIOCATALYSTS
29
Phage display Phage display technology provides a versatile tool for exploring interactions between proteins, peptides, and small molecule ligands. Derivatives of M13 filamentous phages are most commonly used for display of proteins on the surface of E. coli. This viral particle replicates and assembles without killing the host cell. The proteins and peptides to be displayed can be expressed on the surface of a phage by inserting the gene of interest into the gene of one of the phage coat proteins, such as g3p and g8p. During phage assembly in the periplasmic space, the fusion proteins are incorporated into the nascent phage particle 98-
100.
The successful display of many enzymes, such as amylases 101, ß-lactamases 102, lipases103, subtiligases 104, endoxylanases 105, and transferases 106 has been reported previously. As phage particles are assembled in the cell envelope of E. coli, translocation of the fusion protein across the inner membrane of E. coli to the periplasm of the cell is a prerequisite for proper phage display. Thus, mostly periplasmic instead of intracellular proteins are employed in phage display. However, even periplasmic proteins that fold well in the periplasmic space frequently show a poor display on filamentous phages. Recently, presentation of poorly displayed enzymes was improved by exchanging the Sec signal peptide for translocation for SRP-dependent signal sequences. resulting in a >1,000-fold enrichment per selection round 107. The display of enzymes that fold in the cytoplasm was recently demonstrated by exchanging the signal peptide, either by using a Sec-dependent signal peptide 108, or by making use of the Tat translocation pathway followed by association with the g3p coat protein in the periplasm 109. These developments open doors to new possibilities for the selection of enzymes. Indirect selection based upon affinity One of the first selection strategies for phage-displayed enzymes was based on binding to substrates, products or transition state analogues (figure 3). Phage-enzymes are added to an immobilised target: only those that do bind are isolated. These phages are amplified and undergo subsequent rounds of selection. In this way, enzymes have been selected on the basis of their stability since only stably folded phage-enzymes will have bound to the target 110. As the catalytic power of enzymes stems from their higher affinity for transition state analogues than for substrates, selection by transition state analogues should be more efficient. This has been shown for several catalytic antibodies 111,112.
Selection of enzymes using suicide inhibitors is based on this same affinity principle. Suicide substrates irreversibly bind and thus inhibit the selected enzyme 113,114. This principle has been demonstrated by Beliën et al using endoxylanases. Endoxylanases are important enzymes for industrial processes such as bread making and beer production. Endoxylanase inhibitors considerably affect the functionality of endoxylanases in biotechnological processes. Endoxylanase I from Aspergillus niger and endoxylanase A from B. subtilis were functionally displayed on the coat of M13 filamentous phages and incubated with immobilised endoxylanase inhibitors. The phage-enzymes bound with high specificity to the immobilised inhibitors. However, for industrial processes, low specificity binding of endoxylanases to the inhibitor is required. Thus, these results will find their application in the selection of inhibitor-insensitive endoxylanase variants 105.
CHAPTER 2
30
Figure 3: Affinity-capture of phage-displayed enzyme variants on an immobilised substrate. The enzyme is fused to the g3p coat protein. Some enzymes will bind to the immobilised substrate (♦), others will not (○). The affinity selection strategy was also employed in the selection of B. subtilis 168 lipase A variants with improved and inverted enantioselectivity towards the chiral substrate IPG. Dröge et al used a phosphonate ester of enantiopure IPG, coupled to SIRAN beads to facilitate recovery of selected variants. A library of variants was constructed on a region near the active site, and displayed on filamentous phages. Their selection strategy was dual: first, undesired mutants were removed by incubating phage-enzymes with the phosphonate inhibitor coupled to the unwanted enantiomer of IPG. Second, phage-enzymes that did not bind to the first inhibitor were incubated with a second inhibitor, coupled to the desired enantiomer of IPG, thus selecting for variants with desired enantioselectivity. After four rounds of selection, a variant with an inverted enantioselectivity towards the desired wanted enantiomer of IPG was found and characterised. The increase in enantioselectivity however, was modest, which can be explained by the fact that in the 3D model of the structure both enantiomers fit equally well 103,115.
To select B. cereus metallo-β-lactamase variants with improved catalytic activity towards benzylpenicillin, Ponsard et al used a slightly different affinity selection technique. An epPCR library of metallo-β-lactamase was constructed and displayed on phages. The phage-enzymes were inactivated by complexing the enzyme’s cofactor zinc(II) with EDTA. Then inactivated variants with affinity for the substrate benzylpenicillin were absorbed onto immobilised benzylpenicillin. Finally, the inactivated variants were catalytically eluted by adding a zinc(II) salt, thereby reactivating the selected variants. After two rounds of selection, the catalytic activity of the selected mutants was increased 60-fold. In general, this method is applicable for all enzymes necessitating the presence of a cofactor. Nevertheless, the apoenzyme still needs to possess affinity for the substrate 116.
However, affinity selections do have some limitations. The immobilisation of the target (substrate) can have a great influence on the outcome of the selection process 99. Besides that, a disadvantage of the use of suicide inhibitors is that the selection process is based on binding and not on product release and catalytic turnover. Consequently, binding does not necessarily correlate with the catalytic activity of the enzyme. Thus, affinity selection
SELECTION STRATEGIES FOR IMPROVED BIOCATALYSTS
31
should only be used if a change of substrate specificity is desired, not if an enhancement of rate acceleration or turnover is required. Direct selection for catalytic activity Direct selection of enzymes on the basis of catalytic activity is more difficult, as reaction products readily diffuse from the reaction site. In order to inhibit this diffusion, a physical link between the phage-enzyme and the substrate should be established 111. Several methods are available to achieve this link, a number of them will be discussed here.
One approach to procure the enzyme-substrate link is to couple the substrate to a reactive thiol or amine of the phage coat using a maleimide-based linker (figure 4A). This strategy was used in the directed evolution of Bordetella pertussis adenylate cyclase variants. The wild type enzyme catalyses the conversion of ATP into cAMP. Active phage-cyclases were able to convert a substrate analogue which was coupled to the phage into cAMP. The phage-enzymes were recovered by incubation on beads derivatised with a single-chain antibody fragment against cAMP. With these tools, they devised a selection scheme that permits the selection of active phage-cyclases with an enrichment factor of approximately 70-fold for each round of selection 117. However, the disadvantage of this system is that cross-reactivity is not excluded, as the phage-enzyme and the substrate are at some distance from each other.
Figure 4: A) Direct selection of phage-displayed enzyme variants by linking the substrate to the phage coat. B) Selection of polymerases; both an acidic peptide and the enzyme variants are fused to g3p. By incorporating a biotinylated nucleotide, the phage enzyme is captured using streptavidin. C) In vivo selection combined with in vitro display. The substrate, fused to g3p, is biotinylated in the cytoplasm, and the whole complex is then displayed on a phage and affinity-captured with avidin.
CHAPTER 2
32
In an alternative approach, both the enzyme and the substrate can be specifically co-localised on the g3p coat protein, resulting in a display of both the enzyme and the substrate (figure 4B). This would limit cross-reactivity, as the phage-enzyme and the substrate are in each other’s vicinity. To expand the substrate specificity of a DNA polymerase towards the synthesis of unnatural polymers from 2’-O-methyl ribonucleoside triphosphates, a combined display was achieved by reengineering the helper phage genome: it now contained a gIII gene fused to DNA encoding an ‘acidic’ peptide. This peptide is able to form a leucine zipper and a disulfide bond with a ‘basic’ peptide. The latter is conjugated to an oligonucleotide, the primer in the polymerisation reaction. In the selection, extension of the primer that was still bound to the polymerase ultimately resulted in the incorporation of a biotinylated nucleotide, which in turn could be used for affinity–capture of the complex on streptavidin. The method led to one mutant with modified substrate specificity, though it had the same fidelity with unnatural substrates as the wild type enzyme with natural substrates 118,119. This principle was applied to the evolution of thermostable reverse transcriptases as well 120, but an application apart from nucleic acid polymerases is unlikely.
Routenberg Love et al developed a substrate attachment-strategy for the phage display of glycosyltransferase. They focused on E. coli glycosyltransferase MurG, an important enzyme for antibiotic synthesis. Cultures containing a MurG encoding plasmid were infected with selenocysteine (Sec)-expressing helper phages. The outcome was a phage with both the enzyme and a Sec handle on the same end of the phage particle. The Sec handle could be used for binding to biotin and subsequent capture on streptavidin. The activity of the phage-enzyme was established by incubation with UDP-[14C]-GlcNAc and a biotin-labelled lipid I analogue. Phage-enzymes were found to be active, thereby opening doors for the phage-display evolution of related glycosyltransferases 121.
A relatively novel approach is the combination of in vitro selection with in vivo enzymatic activity. Here, substrate- and enzyme-encoding DNA are both introduced in E. coli and expressed in the cytoplasm (figure 4C). The substrate however, is expressed as a fusion protein to one of the coat proteins of M13. After a catalytic reaction in the cytoplasm, the product, likewise fused to the coat protein, is incorporated in the phage coat and displayed on its surface. Upon affinity-capture of the product in vitro, the gene encoding for the selective enzyme is automatically selected for as well. The power of this system was demonstrated with E. coli biotin protein ligase (BPL) as a model enzyme. BPL catalyses the highly specific formation of biotinyl-5’-adenylate from biotin and ADP, and transfers biotin to a specific lysine residue on the biotin carboxyl carrier protein, a subunit of acetyl-CoA carboxylase. The researchers used the biotin-tag-peptide (Btag) as the substrate to be fused to g3p and subsequently to be displayed on the phage coat. Phages displaying the reaction product could be captured with avidin coated beads. An advantage of this method is that the selected enzymes are stable in vivo, since the enzymatic reaction takes place in the (natural) cellular environment. This makes the system extremely suitable for the selection of enzymes that catalyse modifications of peptides or proteins, such as protein ligases, acetylases, kinases, phosphatases, ubiquitinases, and proteases. Furthermore, the enzyme does not need to be secreted to the periplasmic space; consequently, size, nature and folding problems do not propose a limiting factor to display 122. The researchers also demonstrated the utility of the g8p coat protein for substrate fusion. Due to the higher copy number of g8p, more product could be displayed. If the catalytic power of the enzyme of interest is
SELECTION STRATEGIES FOR IMPROVED BIOCATALYSTS
33
moderate or the affinity of the product of the enzymatic reaction for the capture molecule is relatively low, this method could be very effective 123. Bacterial cell surface display Cell surface display is a technique to present peptides or proteins on the surface of Gram-negative or Gram-positive bacteria by fusing them to surface anchoring motifs or otherwise named carrier proteins. This technique has had its impact on a wide range of biotechnological and industrial applications. It has led to substantial progress in whole cell biocatalysis, live-vaccine development, biosorbent and biosensor development, epitope mapping, antigen delivery, inhibitor design and protein library screening. In contrast to phage display, the size of the displayed protein is not a limiting factor in cell surface display. An additional advantage of this system is that bacterial cells are used: they are self-replicative and are sufficiently large to be examined by optical methods, including fluorescence microscopy or FACS analysis. The latter allows for high-throughput screening 124-127.
Selection of a host strain for surface display is an important consideration. A good host has to meet certain criteria, such as compatibility with the protein to be displayed and easy cultivation without cell lysis. Furthermore, the host cell should exhibit low activities of cell wall- and extracellular proteases. For Gram-negative bacteria, the fragility of the outer cell membrane can be a problem; nevertheless, E. coli remains an attractive host for cell surface display, because of its high transformation efficiency. Gram-positive bacteria seem more suitable for surface display purposes, as they have only one cell membrane and their cell walls are thicker and consequently more rigid. Bacillus and Staphylococcus strains are most commonly used for this purpose 125-127.
Not only is the host strain of importance, the anchoring motif to which the protein of interest is fused needs to meet some criteria as well. First, it should have an efficient signal sequence to transport a premature fusion protein across the inner membrane. Naturally, the characteristics of passenger proteins to be displayed on the surface have an influence on the transport as well. Second, it should have a strong anchoring structure to insert into the outer membrane and keep fusion proteins attached to the cellular surface. Again, this attachment is influenced by the passenger protein. Third, the anchoring motif should be compatible with the foreign sequences to be inserted or fused. Last, it should be resistant to any attack of proteases in the periplasm or in the medium. Each type of anchoring motif has its own characteristics and can therefore be useful for specific applications 126. A number of these applications will be discussed on the basis of their anchoring motif in the following paragraphs. Outer membrane proteins as an anchoring motif Outer membrane proteins (Omp) span the membrane several times, and are therefore attractive as an anchoring motif for cell surface display. They are mainly built up of anti-parallel β-strands, resulting in β-barrel structures. Sequences of the protein of interest are inserted into a permissive site of the anchoring motif. Examples of studied membrane proteins used as anchoring motifs are OmpA, OmpS, LamB, PhoE, and OmpC. The Lpp’OmpA system allows for C-terminal fusions, thus being more suitable as a carrier of larger inserts 125,128. Cell surface display on the basis of Omp has been applied previously
CHAPTER 2
34
for whole cell biocatalysis 129, in particular for the enzymatic resolution of chiral compounds. Displayed enzymes act as whole cell biocatalysts. Lee et al applied the cell surface display strategy in the enantioselective resolution of racemic compounds by Pseudomonas fluorescens lipase while anchoring the gene of interest to Salmonella typhimurium OmpC. 130. A second lipase, the thermostable Bacillus sp. strain TG43 lipase, was also applied in whole cell biocatalysis while displayed on the cellular surface. The lipase was fused to the ninth loop of FadL by C-terminal deletion-fusion. FadL, an Omp involved in long-chain fatty acid transport in E. coli, was shown to be useful as an anchoring motif for the display of lipases on the surface of E. coli 131. Autotransporter proteins as anchoring motif The so-called autodisplay system has been developed on the basis of the secretion mechanism of the autotransporter family of proteins. Autotransporters are synthesised as precursor proteins and the passenger is an integral part of the protein. Beside the passenger, the precursor protein contains a signal peptide necessary for transport across the inner membrane, a C-terminal β-barrel for transport across the outer membrane and a connecting peptide. The latter, a so-called linker, will ensure full surface access. Hydrolases, foldases, β-lactamases and oxidoreductases have been autodisplayed on the surface of E. coli using autodisplay 124,132. Some examples and their applications will be discussed below.
For the molecular evolution of Burkholderia gladioli esterase A, a system autodisplaying the enzyme was developed. The esterase was genetically fused to the autotransporter domains of the adhesin (AIDA-I) involved in diffuse adherence to HeLa cells. After transformation of E. coli with the plasmid, the fusion protein was transported to and expressed on the cell surface. The enzyme was displayed in its active form, as was demonstrated by the conversion of p-nitrophenyl acetate to p-nitrophenol. These results provide a basic selection strategy for the evolution of a biocatalyst such as esterase A 133. An antibody-independent detection method was developed for future selections.
Apart from the autotransporter AIDA-I, a membrane-anchored esterase (EstA) from Pseudomonas aeruginosa was also used in autodisplay. EstA is enzymatically active when anchored to the cell surface of P. aeruginosa, but also on the surface of the heterologous host E. coli. As a consequence, EstA can only be used as an autotransporter for hydrolytic enzymes when in an inactive form. Three hydrolytic enzymes, B. subtilis lipase A, Fusarium solani cutinase, and the large Serratia marcescens lipase, were fused to an inactive variant of EstA. All three enzymes were displayed as fusion proteins on the surface of E. coli, as was revealed by FACS analysis and more importantly, the three lipases retained enzymatic activity 134. EstA was also employed in the autotransportation of a lipase specific foldase (Lif), the protein LipH from P. aeruginosa, required in the folding of extracellular lipases from Pseudomonads and related strains to convert lipases in their active form. Surface displayed LipH was analysed by FACS, and proved to be functional by efficiently refolding chemically denatured lipase. This system can therefore be used in the selection of large libraries of foldase variants 135.
Although EstA itself is an excellent anchoring motif for autodisplay, the protein itself can also be used for surface display using its hydrolytic characteristics. To select for this property, the product should not diffuse from the reaction environment. Thus, Becker et al developed a cell surface display system termed ESCAPED (Enzyme Screening by Covalent
SELECTION STRATEGIES FOR IMPROVED BIOCATALYSTS
35
Attachment of Products via Enzyme Display) in which the product is linked to the outer membrane of E. coli by tyramide conjugation. In a model experiment, the researchers used an octanoic ester of biotin-tyramide to be hydrolysed by EstA. The resulting tyramide-conjugate was used as a substrate for horse radish peroxidase (HRP), the latter being surface-displayed as well. HRP reacts with hydrogen peroxide and biotin-tyramide to produce a quinone structure bearing a radical. This radical interacted with tyrosine residues in close vicinity to HRP on the cellular surface, resulting in biotin deposition and subsequent affinity-capture using streptavidin-coated magnetic beads. Thus, only cells displaying an active hydrolase were selected. After only two rounds of cell sorting, the enriched population of bacterial cells showed an increased esterase activity. This method can be applied in the selection of proteases and phosphatases (figure 5) 73,136.
Figure 5: The ESCAPED system. The substrate, a biotin-tyramide ester, is converted by the cell surface-displayed enzyme EstA (◊); then released tyramide (▲) reacts with cell surface-displayed HRP (○) to form radicals that will bind to the cell surface of E. coli. Detection is achieved with tagged streptavidin. Cell surface display of Gram positive bacteria In some applications, the use of genetically modified bacteria is less desirable, e.g. in food processing and vaccine development. Recently, a novel surface display system using non-genetically modified Gram-positive bacteria was developed. The system is based on the peptidoglycan-binding domain of the major autolysin AcmA of Lactococcus lactis and enables functional display of heterologous proteins on the surface of genetically unmodified Gram-positive bacteria. The cell wall-binding domain is designated the protein anchor (PA), which directs the protein to the cell membrane. Hybrid PA fusion proteins exhibit the same properties. To prove the principle of the display system, two biocatalysts, B. licheniformis α-amylase and E. coli β-lactamase, were functionally displayed on the surface of L. lactis by coupling them to PA. Activity assays showed that the enzymes were still active. This system can be used for the immobilisation of enzymes to be used in industrial processes. In conclusion, this surface display system provides a cheap, flexible and easy-to-handle alternative to display proteins on non-genetically modified bacteria 137. This may prove to be an advantage in developing vaccines.
CHAPTER 2
36
Ribosome display The previously discussed selection strategies, phage- and cell surface display, only work ‘partially’ in vitro. In the last years, cell-free selection strategies such as ribosome display and in vitro compartmentalisation have gained momentum, as a transformation step is circumvented. In ribosome display, the physical link between genotype and phenotype is achieved by mRNA-ribosome-protein complexes formed in the translation step which can be directly used for selection. In vitro translation occurs in a cell-free system, such as E. coli S30 or rabbit reticulocytes (figure 6). The system can either be coupled or uncoupled. Coupled systems make use of DNA; they are simpler, more efficient and avoid problems of mRNA degradation. In contrast, uncoupled systems require mRNA 138.
Figure 6: Schematic overview of selection by ribosomal display. In a cell-free system, dsDNA is transcribed to mRNA, which forms a complex with the riboxome. mRNA lacks a stop codon, thereby stalling translation. Selection is based on affinity-capture of the complex. After disassembly of the complex, mRNA is amplified by RT-PCR. During translation, ribosomal complexes are formed containing a functionally folded protein from the ribosomal tunnel. For the protein to fold, the fusion protein is constructed in such a way that the domain of interest is fused to a C-terminal spacer. This enables the protein to fold while the spacer is still in the ribosomal tunnel. The fusion lacks a stop codon at the mRNA level; as a consequence, the release of mRNA and the polypeptide is stalled. The complex is further stabilised by high concentrations of magnesium(II) and low temperatures. Following selection in an RNase-free environment, mRNA is released from the ternary complex by removing magnesium(II) and cDNA is prepared by RT-PCR. This step allows for additional mutations to be incorporated 76.
SELECTION STRATEGIES FOR IMPROVED BIOCATALYSTS
37
Ribosome display has been mostly applied to the selection of single-chain antibody fragments and peptides from designed libraries. Some enzymatic selections have been reported as well, which will be discussed in the next section.
The first selection of an enzyme based on catalytic activity was reported by Amstutz et al. As a model system, the researchers displayed RTEM β-lactamase on the surface of the ribosome in complex with its mRNA. The gene was fused to the C-terminus of E. coli TolA, which served as a spacer. The enzyme was correctly folded and displayed in an active form on the ribosome. Selection was based on affinity by making use of a mechanism based suicide inhibitor, biotinylated ampicillin sulfone. After translation, the complexes were incubated with the inhibitor and rescued with streptavidin-coated beads. Per round of selection, active β-lactamase could be enriched over an inactive mutant >100-fold 139. This same affinity principle was also employed in the selection of dehydrofolate reductase (DHFR) variants. Mutants were selected using a substrate analogue, methotrexate immobilised on agarose beads. Only a ribosome complex containing an active DHFR could bind to the methotrexate beads. Four mutants showing the same activity as the wild type enzyme were selected and further characterised 140.
To select for catalytic activity of ligases, T4 DNA ligase was used as a model enzyme. mRNA encoding for both T4 DNA ligase and a spacer was hybridised with double stranded DNA (dsDNA), a substrate of T4 DNA ligase. The resulting hybrid was translated in vitro, up to the point where the ribosome reached the site of RNA-DNA hybridisation. The enzyme was functionally displayed on the ribosome, which was shown by ligation of the dsDNA in the complex with biotinylated dsDNA probe. Thus, the complex displaying active T4 DNA ligase was labelled with biotin and selected by binding to streptavidin-coated beads. A 40-fold enrichment over an inactive mutant could be achieved. Using this method, it is possible to evolve new functions of T4 DNA ligase as well as of other ligases 141. Ribosome display has not been widely used, however.
Ribozymes, enzymes made of RNA, have been shown to be capable of biocatalysis as well (for a review, see 142). However, the focus is on their proteinaceous counterparts, thus ribozymes are beyond the scope of this review. In vitro compartmentalisation In vitro compartmentalisation (IVC) simulates cellular compartments in which only the reaction to be selected for is performed. The technique is based on water-in-oil emulsions, where the water phase is dispersed in the oil phase with the aid of surfactants to form microscopic aqueous compartments. Thus, artificial cells of approximately 5 fL are created. These droplets contain on average a single gene; here, transcription, translation and expression of the resulting proteins can all take place from in vitro active components. The oil phase remains mostly inert and limits the diffusion of DNA and proteins between compartments (figure 7) 71,143.
CHAPTER 2
38
Figure 7: General strategy for selection by IVC. Active enzyme-encoding genes can be isolated on the basis of the presence of the product, while inactive enzyme-encoding genes can be discarded on the basis of an unmodified substrate. This strategy was employed in the selection of the methyltransferases HaeIII (M.HaeIII) 144 and HhaI (M.HhaI). Selection was performed by extracting the genes from the emulsion and subjecting them to digestion using a restriction enzyme that cleaves the non-methylated DNA 145,146. By biotinylation, the active enzyme-encoding genes could be affinity-captured and used in subsequent rounds of selection. This strategy resulted in catalytically improved enzyme variants.
IVC was also used as the selection strategy of choice in the evolution of a bacterial phosphotriesterase (PTE), which is applied in the degradation of pesticides. Here, a so-called microbead-display library technique was used. Streptavidin-coated microbeads displaying a library of the gene and the enzyme PTE it encodes were compartmentalised in a first emulsion. Then, after breaking the first emulsion, the microbeads were recompartmentalised together with a soluble substrate coupled to caged-biotin. After the enzymatic reaction, the biotinylated product was captured on the streptavidin microbeads, allowing for selection with a fluorescent labelled anti-product antibody. The second emulsion was broken and product-coated beads could then be enriched and analysed by FACS. This method resulted in an enrichment of PTE activity, with one variant exhibiting a 63 times higher activity than the wild type enzyme 147.
As seen previously, the direct sorting by FACS may allow for versatile and powerful high-throughput systems. Using fluorogenic substrates, artificial cells compartmentalising genes encoding active enzymes would become fluorescent, making them applicable for selection by FACS. However, the continuous oil phase currently employed in IVC is not compatible with FACS analysis. Thus, double emulsions (water-oil-water) were developed. As a proof of principle, M.HaeIII encoding for DNA methyltransferase, and FolA encoding for E. coli dehydrofolate reductase were compartmentalised, FolA together with FITC-BSA. Both genes were tagged with biotin to allow affinity-capture on streptavidin after breakage of the emulsions. The genes isolated from the droplets sorted by FACS were amplified by PCR
SELECTION STRATEGIES FOR IMPROVED BIOCATALYSTS
39
and appeared at a 1:3 FolA:M.HaeIII ratio, indicating a 30-fold enrichment. This result demonstrated that there was no mixing of either DNA or FITC-BSA, thus the genotype-phenotype link is conserved in this system 148. The applicability of this system was demonstrated in the selection of novel β-galactosidase variants 149 and serum paraoxonase (PON1) with thiolactonase activity 143. Mastrobattista et al evolved Ebg, a protein with at present an unknown function, into an enzyme with considerable β-galactosidase activity, while Aharoni et al achieved a 100-fold improvement in thiolactonase activity. Concluding remarks Directed evolution has become an important means to improve an enzyme or alter its substrate specificity. To be able to analyse libraries of at least 1010 variants at a time, several selection strategies have been set up, always linking phenotype to genotype. Nevertheless, selection for catalysis remains a difficult task: enzymatic activity still needs to be specifically tailored for each enzyme, reaction and substrate. Each system has its own advantages and disadvantages. In vivo methods, though elegant, can be limited in their use because of the narrow range of reactions that can be used in selection. Partial in vitro techniques have been developed to overcome these limitations, although they still involve a transformation step. Cell–free systems combined with FACS analysis hold enormous potential, allowing a rapid analysis of enzyme variants. Thus, to our opinion these systems will prove the most promising for the future of directed enzyme evolution.
3
COMPARISON AND FUNCTIONAL CHARACTERISATION OF THREE
HOMOLOGOUS INTRACELLULAR CARBOXYLESTERASES OF
BACILLUS SUBTILIS
MELLONEY J. DRÖGE, REIN BOS,
YKELIEN L. BOERSMA & WIM J. QUAX
J MOL CATAL B: ENZYMATIC 2005; 32: 261-270
INTRACELLULAR CARBOXYLESTERASES OF BACILLUS SUBTILIS
43
Comparison and functional characterisation of three homologous
intracellular carboxylesterases of Bacillus subtilis Enzymatic hydrolysis of racemic mixtures may provide an attractive method for the enantiopure production of chiral pharmaceuticals. For example, the carboxylesterase NP of Bacillus subtilis Thai I-8 is an excellent biocatalyst in the kinetic resolution of NSAID esters, such as naproxen and ibuprofen methyl esters. Two homologues of this enzyme were identified when the genome sequence of B. subtilis 168 was revealed in 1997. We characterised one of the homologues, YbfK, as a very enantioselective 1,2-O-isopropylidene-sn-glycerol caprylate esterase, while only modest enantioselectivity towards the naproxen ester was observed. The other homologue, the carboxylesterase NA has not been characterised yet. The purpose of the present study was to fully characterise these three highly homologous esterases with respect to their applicability towards the enantiospecific hydrolysis of a wide range of compounds. The esterase genes were cloned and expressed in B. subtilis using a combination of two strong promotors in a multi-copy vector. After purification of the enzymes from the cytoplasm of B. subtilis, the biochemical and enantioselective properties of the enzymes were determined. Although all carboxylesterases have similar physico-chemical properties, comparison of their specific activities and enantioselectivities towards several compounds revealed rather different substrate specificities. We conclude that carboxylesterase NP and carboxylesterase NA are particularly suited for the enzymatic conversion of naproxen esters, while YbfK offers enantiopure (+)-IPG from its caprylate ester. Given the carboxylesterase activities of the esterases it has been proposed to rename the nap gene of B. subtilis 168 into cesA and the ybfK gene into cesB.
Introduction Carboxylesterases (E.C. 3.1.1.1.) represent a diverse group of hydrolytic enzymes catalysing the cleavage and formation of ester bonds. In spite of their distribution throughout humans, animals, plants and microorganisms, their physiological functions remain to be elucidated. Nevertheless, due to their high stability, their activity in organic solvents, and their high regio- and stereospecificity, carboxylesterases appear to be attractive biocatalysts for organic chemistry 41.
Many bacterial esterases have been cloned and overexpressed during the last decades in order to assess their enantioselective properties 42,150-153. The most attractive biocatalysts seem to originate from Bacillus and Pseudomonas species (for a review, see 40). For example, the naproxen esterase of B. subtilis Thai I-8 was characterised as a very efficient enantioselective biocatalysts for the kinetic resolution of non-steroidal anti-inflammatory drug (NSAID) esters, such as the naproxen and ibuprofen methyl ester. Its modest selectivity towards chiral alcohols such as the interesting chiral intermediate
CHAPTER 3
44
1,2-O-isopropylidene-sn-glycerol 43,44 is in sharp contrast with its enantioselective properties towards chiral carboxylic acids.
Although many well-established methods are available today for the directed evolution of the enantioselective properties of a potential biocatalyst 114,154-161, the search for homologous and paralogous genes in the sequence information derived from several genome projects, for example the genome project of B. subtilis 168, offers an attractive alternative approach for finding alternative biocatalysts for rational drug preparation.
In B. subtilis 168, nine genes have been functionally characterised as intracellular or extracellular esterases. In addition, eight genes can be classified as potential esterases 37,38,42,162-170. Two of these esterases of B. subtilis 168, carboxylesterase NA (encoded by the nap gene) and YbfK (ybfK gene) showed a high homology towards the carboxylesterase NP of B. subtilis Thai I-8 (98% and 64% identity on protein level, respectively). Given the carboxylesterase activities of the esterases it has been proposed to rename the nap gene of B. subtilis 168 into cesA and the ybfK gene into cesB. Recently, the CesB protein was characterised as a very enantioselective 1,2-O-isopropylidene-sn-glycerol esterase, while only modest enantioselectivity towards naproxen esters was observed 162. In contrast, CesA has not been characterised yet.
In this paper, we describe the production of these three highly homologous carboxylesterases in B. subtilis, and their biochemical characterisation. A pH and temperature dependency profile was established for all three enzymes. In particular, we have determined the specific activity of the three enzymes towards several chiral compounds, with chirality residing in both the carboxylic acid part as well as the alcohol part of the ester, in order to investigate the applicability of these enzymes for kinetic resolution experiments. Experimental procedures Plasmids, bacterial strains, and media The plasmids and bacterial strains that were used in the present study are listed in table I. The following media were used: 2xTY medium containing Bactotrypton (1.6% w/v), Bacto yeast extract (1% w/v) and sodium chloride (0.5% w/v); medium to prepare B. subtilis competent cells containing 100 mM potassium phosphate buffer, pH 7, 1% w/v glucose, 0.4% w/v potassium l-glutamate, 3 mM trisodium citrate, 3 mM MgSO4, 0.0022 % w/v ferric ammonium citrate, 0.1% w/v casein hydrolysate and 0.002% w/v L-tryptophane 171. Antibiotic agents were used in the following concentrations: ampicillin 100 µg.mL-1, kanamycin 20 µg.mL-1, and chloramphenicol 5 µg.mL-1. Chemicals The methyl ester of S-naproxen was provided by Prof. H.V. Wikström (Department of Medicinal Chemistry, University of Groningen, Groningen, The Netherlands). (-)-IPG acetate, (-)-IPG butyrate, (-)-IPG caprylate, (+)-IPG acetate, (+)-IPG butyrate and (+)-IPG caprylate were kindly provided by M.T. Reetz (Max-Planck Institut für Kohlenforschung, Mülheim, Germany). ß-naphtyl acetate, p-nitrophenyl esters, racemic ibuprofen methyl ester, α-methylbenzyl acetate, N-acetyl phenylalanine methyl ester, α-methoxyphenyl
INTRACELLULAR CARBOXYLESTERASES OF BACILLUS SUBTILIS
B. subtilis Thai I-8 CBS 679.85 B. subtilis 168 TrpC2 37 B. subtilis 1050 NprR2, nprE18, aprA3, est::Cm, ∆LipA 35,172 Bs1050(pMA) B. subtilis 1050 transformed with pMA, a
pMA5 derivative, containing the HpaII and cesA promotor
162
Bs1050(pMAybfK) B. subtilis 1050 transformed with pMAybfK 162 Bs1050(pMAnap) B. subtilis 1050 transformed with pMAnap This work Bs1050(pMAthai) B. subtilis 1050 transformed with pMAthai 162
168 cesB gene, downstream of the HpaII and cesA promoter
162
pMAnap pMA5 derivative, containing the B. subtilis 168 cesA gene, downstream of the HpaII and cesA promoter
This work
pMAthai pMA5 derivative, containing the B. subtilis Thai I-8 nap gene, downstream of the HpaII and cesA promoter
162
acetic acid methyl ester and ß-phenyl lactic acid methyl ester were all purchased from Sigma Chem. Co. (Axel, The Netherlands). Oligonucleotides To construct the plasmid mentioned in table I, the following primers were used (Life Technologies, UK): Pnapfor1: 5’-GTTACGGATCCCTCCATTGTGCTCG-3’ (BamHI); naprev1: 5’-GAGAAGCTTGAAGCAT ATTGCAGGACTTTAT-3’ (HindIII); napfor2: 5’-CTTATTTATGCTGGTACCCACATTCATTTAAA CAA-3’ (KpnI). Newly created restriction sites are indicated in bold italics. DNA techniques Recombinant DNA techniques were performed as described by Sambrook et al 176. Enzymes endonucleases were purchased from Life Technologies. Plasmid DNA was prepared as described by Birnboim & Doly 177. DNA purification was performed by using the Qiaquick Gel Extraction kit (Qiagen, Hilden, Germany).
CHAPTER 3
46
Construction of the plasmids The cesB gene was cloned as described previously by Dröge et al 162. The promotor of the cesA gene and the cesA gene itself (both originating from B. subtilis 168) were amplified using the primers Pnapfor1 and naprev1. All PCRs were performed using Pfu DNA polymerase (Stratagene, La Jolla, CA, USA). The PCR protocol was as follows: 4 min at 94°C, followed by 25 cycles of 1 min at 94°C, 1 min 50°C and 1 min at 72°C. At the end of the protocol, DNA production was finished with 10 min heating at 72°C. The amplified gene fragments were cloned into the KpnI and HindIII sites of pMA5, an Escherichia coli/B. subtilis shuttle vector. In this plasmid, pMAnap, the cesA gene became located downstream of the cesA promotor. BamHI digestion removed the E.coli replicon and positioned the gene and the cesA promotor downstream of the strong Gram positive HpaII promotor 35,174,175. The shortened plasmid was used to transform Bacillus strain BCL1050. Restriction analysis and DNA sequencing were used to verify the sequence of the construct. The nap gene of B. subtilis Thai I-8 was amplified from chromosomal DNA of B. subtilis Thai I-8 using primers napfor2 and naprev1. The gene was cloned in pMA5 by replacing the RsrII and HindIII fragment of pMAnap with the RsrII and HindIII digested PCR fragment resulting in pMAthai. Enzyme purification To produce enzymes, B. subtilis was grown in 2 L shake flasks, containing 500 mL 2xTY medium, at 37ºC at 300 rpm with good aeration for 16 h. After harvesting the cells, the cytoplasmic fraction was isolated as described by Dröge et al 162. CesA and carboxylesterase NP were purified by loading the cytoplasmic fraction on 3 coupled Hitrap Q columns (1.6 × 2.5 cm; Amersham Pharmacia Biotech, Uppsala, Sweden), equilibrated with 10 mM Tris HCl, pH 8, containing 1 mM EDTA, using a flow of 5 mL.min-1. Elution was performed with a linear gradient from 0 to 1 M of sodium chloride in 10 mM Tris HCl, pH 8, containing 1 mM EDTA. The collected fractions (10 mL) were screened for the presence of CesA or carboxylesterase NP using an SDS PAGE and the naproxen methyl ester assay. The fractions containing carboxylesterase were pooled and 0.5 mM ammonium sulphate was added. This solution was loaded on a MT20 column (15 × 110 mm; Bio-Rad, Hercules, CA, USA) packed with 20 mL phenyl sepharose HP (Amersham Pharmacia Biotech, Uppsala, Sweden), and equilibrated with 10 mM phosphate buffer, pH 8, containing 0.5 M ammonium sulphate. Elution was performed using a combination of a stepwise and linear gradient from 0.5 M to 0 M ammonium phosphate in 10 mM phosphate buffer, pH 8. Fractions (10 mL) were screened for activity and the carboxylesterase containing fractions were pooled and stored at -20ºC. The protein concentration was determined both by the Bradford and Lowry method (Pierce, Rockford, Illinios, USA). CesB was purified as described previously 162. Electrophoresis SDS PAGE was performed on a 12% separating and a 4% stacking gel 178. Molecular mass markers were purchased from Bio-Rad (Bio-Rad, Hercules, CA, USA). Proteins were stained by the silver staining procedure of Pierce (Pierce, Rockford, Illinios, USA) or by Coomassie Brilliant Blue R-250 staining (Pierce, Rockford, Illinios, USA).
INTRACELLULAR CARBOXYLESTERASES OF BACILLUS SUBTILIS
47
Esterase activity assays
Naproxen methyl ester assay Esterase activity was determined using the naproxen methyl ester assay. 13 mg (S)- or (R)-naproxen methyl ester was dissolved in 10 mL 14.3% w/v Tween 80 in 0.07 M MOPS buffer, pH 7.5, at 60°C in an ultrasonic bath (60 min). The solution was diluted to 50 mL with 0.07 M MOPS buffer, pH 7 42,179. Samples were diluted with 0.1 M MOPS buffer containing 0.2% w/v BSA to a volume of 250 µL. 10 mM phosphate buffer, pH 8, was diluted correspondingly and was used as a reference. The sample solution and the substrate solutions were preincubated at 32°C in a water bath. 750 µL substrate solution was added to the sample solution and the final solutions were incubated in a water bath for 4 hours at 32°C. The samples were analysed by HPLC. HPLC was performed using an Isco pump 2350, an Isco gradient mixer 2360 (ISCO Inc., Lincoln, NE, USA), a Kontron autosampler 360 (Kontron Instruments SpA, Milan, Italy), and a Shimadzu SPDM6A-Diode Array detector (Shimadzu Europe GmbH, Duisburg, Germany). The chromatographic conditions used were as follows: an analytical column (LiChrospher 100 RP-18, 5 µm; LiChrocart 250-4), a guard column (LiChrospher 100 RP-18, 5 µm; LiChrochart 4-4, Merck Darmstadt, Germany), an eluens consisting of methanol:10% acetic acid (90:10 v/v), an isocratic flow of 0.75 mL.min-1 with a pressure of 1500 psi, an injected volume of 20 µL, a DAD wave length of 239 nm, a band width of 2 nm, a spectrum absolute scale (mAbs) of –10-1000, and a normalisation threshold of 10 mABS. The capacity factor (k’) for naproxen and the methyl ester of naproxen was 1.43 and 1.82, respectively. The hydrolysis by the blanks was always zero. β-Naphtyl acetate assay and ibuprofen methyl ester assay Both substrates were dissolved in 10 mL 14.3% w/v Tween 80 in 0.07 M MOPS buffer, pH 7.5. Naphtyl acetate was dissolved at 60 ºC in an ultrasonic bath (60 min). The solutions were diluted to 50 mL with 0.07 M MOPS buffer, pH 7.5. The assay was performed as described above. The conversion of the ester was determined using HPLC analysis as described for the naproxen methyl ester assay above. UV absorption was detected at wavelengths of 274 nm (ß-naphtyl acetate) and 239 nm (ibuprofen methyl ester), respectively. 1,2-O-isopropylidene-sn-glycerol (IPG) ester assay The esters of IPG were dissolved in 10 mL 14.3% w/v Tween 80 in 0.07 M MOPS buffer, pH 7.5 and diluted to 50 mL with 0.07 M MOPS buffer, pH 7.5. Samples were diluted with 0.1 M MOPS buffer containing 0.2% w/v BSA to a volume of 150 µL. 10 mM phosphate buffer, pH 8, was diluted correspondingly and was used as a reference. The sample solution and the substrate solutions were preincubated at 32°C in a water bath. 500 µL substrate solution was added to the sample solution and the final solutions were incubated in a water bath at 32°C for 4 hours. After incubation, 500 µL saturated NaCl solution was added and the aqueous solution was extracted twice with 1 mL ethyl acetate. GC analysis was performed on a Hewlett Packard 5890 series II gas chromatograph equipped with a 7673 injector and a Hewllet Packard 3365 Chemstation. The chromatographic conditions used were as follows: a WCOT fused-silica CP-wax 52 CB column (10 m × 0,25 mm id, film thickness 0.25 µm, Chrompack International, Middelburg, The Netherlands), an oven
CHAPTER 3
48
temperature programme of 50-125°C at 3°C.min-1, an injector temperature of 250°C, a detector (FID) temperature of 300°C, the carrier gas was helium, an inlet pressure of 5 psi (1 psi = 6894.76 Pa), a linear gas velocity of 26 cm.s-1, a split ratio of 56:1 and an injected volume of 1 µL.
Chiral GC analysis was performed on the same Hewlett Packard 5890 series II gas chromatograph. The used chromatographic conditions were as follows 180: a WCOT fused silica heptakis(6-O-tButyldimethylsilyl-2,3-di-O-methyl)-ß-cyclodextrin column (50% in OV-1701 w/w, 25 m × 0.25 mm id, film thickness 0.25 µm, Prof. W.A. König, Institut für Organische Chemie, Universität Hamburg, Germany), an oven temperature programme of 10 min at 90°C, 90-125°C at 3°C.min-1, 5 min 125°C, an injector temperature of 250°C, a detector (FID) temperature of 300°C, the carrier gas was helium, an inlet pressure of 17 psi, a inear gas velocity of 40 cm.s-1, a split ratio of 11:1, and an injected volume of 1 µL. A different oven temperature programme was applied to separate the enantiomers of IPG-caprylate. The programme consisted of 10 min. at 90°C, 90-150°C at 3°C.min-1, 15 min 150°C. The capacity factor (k’) was calculated using the formula k’ = (Tr – T0)/T0. Tr confers to the retention time (min) of the compound and T0 to the retention time of methane (the void volume), which was 1.0505 min 180. The hydrolysis of butyrate and caprylate esters by the blanks was always zero, while the hydrolysis of acetate esters was negligible. α-Methylbenzyl acetate, N-acetyl phenylalanine methyl ester, α-methoxyphenyl acetic acid methyl ester and β-phenyl lactic acid methyl ester assay All substrates were diluted to 10 mM in 10 mM Tris HCl buffer, pH 7.5, according to the manufacturer’s instructions. Samples were diluted with 10 mM Tris HCl buffer, pH 7.5, to a volume of 100 µL. 1 mL of substrate solution was added and the final solutions were incubated in a water bath at 32°C for 4 hours. The assays were performed using the HPLC system as described above for the naproxen methyl ester assay. UV absorption was detected at a wavelength of 254 nm. p-Nitrophenyl ester assay 10 mM solutions of the different p-nitrophenyl esters in methanol were prepared. 0.5 mM p-nitrophenylcaprylate (50 µL) was added to 900 µL assay buffer, containing 50 mM phosphate buffer (pH 8), 0.36% v/v Triton X100 and 0.1% w/v gum arabic. Samples were diluted with assay buffer to a volume of 50 µL and added to the substrate solution. The absorbance was measured at a wavelength of 410 nm. Concentrations were calculated using a molar extinction coefficient of 15.000 M-1.cm-1. Corrections were made for spontaneous hydrolysis of the substrate. pH optimum The pH optimum was determined using the naproxen methyl ester assay as described above. The activity of the enzymes at different pH was assessed using S-naproxen methyl ester solutions dissolved in: 0.07 M glycine HCl buffer (pH 3), 0.07 M potassium phosphate buffer (pH 4), 0.07 M potassium phosphate buffer (pH 6), 0.07 M Tris HCl buffer (pH 7), 0.07 M MOPS buffer (pH 7.5), 0.07 M Tris HCl buffer (pH 9), and 0.07 M glycine buffer (pH 11), respectively.
INTRACELLULAR CARBOXYLESTERASES OF BACILLUS SUBTILIS
49
Temperature optimum The temperature optimum was determined using the naproxen methyl ester assay as described above. Enzymatic activity was determined by incubating the enzyme reaction at different temperatures of 4, 20, 30, 40, 50, and 60ºC, respectively. Data analysis One unit (U) is defined as the amount of enzyme that hydrolyses 1 µmol of substrate ester per minute. Enantiomeric ratios, E, were defined as the ability of the enzyme to distinguish between enantiomers 181,182. When the E value exceeded 100, the enantiomeric excess, ee, was calculated. All data were the results of three experiments. Results Cloning of the carboxylesterase genes After the elucidation of the genome sequence of B. subtilis 168, it became clear that this organism contained two genes with a high homology to the carboxylesterase NP of B. subtilis Thai I-8. These proteins, CesA (encoded by the cesA gene) and CesB (encoded by the cesB gene) were respectively 98% and 64% identical on protein level (figure 1). The characteristic pentapeptide for most lipases and esterases, Gly-Xaa-Ser-Xaa-Gly, was present in all three carboxylesterases.
cesB was cloned in pMA5, an E. coli/B. subtilis shuttle vector as described previously 162. The cesA gene and its promotor of B. subtilis 168 were cloned in the same plasmid. In short, the KpnI and HindIII digested PCR fragment was ligated in the pMA5. After ligation, E. coli DH5α was transformed. Digestion with BamHI and subsequent self-ligation of the resulting vector removed the E. coli replicon and positioned the cesA gene and its promotor region downstream of the strong Gram positive HpaII promotor 35,174,175. Both promotors are expressed constitutively in B. subtilis. The shortened plasmid was used to transform the lipase A and esterase A negative B. subtilis strain 1050 35,172, resulting in strain Bs1050(pMAnap). Then, the RsrII/HindIII fragment of the plasmid pMAnap was exchanged with the RsrII/HindIII fragment of the PCR amplified gene fragment of nap of B. subtilis Thai I-8. DNA sequencing confirmed that the sequences of all constructs were correct.
CHAPTER 3
50
Figure 1: Comparison of the sequences of carboxylesterase NP (B. subtilis Thai I-8, upper lane), CesA (B. subtilis 168, middle lane) and CesB (B. subtilis 168, lower lane). The matching amino acids are in black, the Gly-Xaa-Ser-Xaa-Gly motif is underlined Expression in B. subtilis and isolation of the enzyme Comparison of the cytoplasmic fractions of Bs1050(pMAnap) and Bs1050(pMAthai) revealed comparable overexpression levels of the esterases in the cytoplasm of B. subtilis 1050, while the expression of CesB of strain Bs1050(pMAybfK) was somewhat lower (data not shown). CesA and carboxylesterase NP were purified using a two-step purification protocol (table II).
INTRACELLULAR CARBOXYLESTERASES OF BACILLUS SUBTILIS
51
Table II: Purification of CesA, carboxylesterase NP and CesB from the cytoplasm of a 1 L culture of B. subtilis 1050. N.d.: not determined.
CesB (B. subtilis 168) Sonication 330 n.d. n.d. n.d. n.d. HiTrap Q 14 1.7 0.12 1 100 Phenyl sepharose 6 1.02 0.17 1.42 60 Firstly, anion exchange chromatography at pH 8 was performed. Although many proteins present in the cytoplasm of B. subtilis bound to a HiTrap Q column at pH 8, CesA and carboxylesterase NP were detected in the flow-through. Surprisingly, SDS PAGE and Coomassie staining revealed that the flow-through fraction contained almost purified carboxylesterase. Afterwards, a second chromatography step based on hydrophobic interaction was performed to remove the remaining contaminating proteins. CesA and carboxylesterase NP bound strongly to a phenylsepharose column and eluted during the isocratic flow with a phosphate buffer without ammonium sulphate. CesB was purified as described previously 162. SDS-PAGE and silver staining confirmed the purity of all samples (figure 2).
Figure 2: SDS PAGE, silver staining. Lane 1) Purified CesA of B. subtilis 168. Lane 2) Purified carboxylesterase NP of B. subtilis Thai I-8. Lane 3) Purified CesB of B. subtilis 168.
CHAPTER 3
52
Enzymatic activity of the esterases
Hydrolysis of the naproxen methyl ester The catalytic activity towards the S-methyl ester of naproxen was determined (table III) using a non chiral HPLC method. The specific activities of CesA and carboxylesterase NP were not significantly different, 4.2 and 4.6 U.mg-1 respectively. Comparison of the specific activities of CesB and CesA and carboxylesterase NP revealed an approximate 25-fold lower specific activity of CesB towards the naproxen ester (e.g. the specific activity was 0.17 U.mg-1 for CesB). It should be noted that the production of S-naproxen by CesB is still far above the detection limit of the HPLC system and for this, it can be attributed to enzymatic activity of CesB. Moreover, the hydrolysis of the substrate by the blanks was always below detection.
Then, the activity of the three esterases towards the methyl ester of R-naproxen was determined and compared with the hydrolysis of the S-naproxen methyl ester. Table III summarises these results. The three esterases displayed similar specific activities towards the R-naproxen methyl ester, ranging from a specific activity of 0.019 U.mg-1 (CesB) to 0.022 U.mg-1 (CesA). It should be mentioned that the production of R-naproxen by the blanks was always below detection. Therefore, the specific activities towards the R-naproxen methyl ester can be attributed to enzymatic activity.
Consequently, the enantioselectivity of these enzymes towards the racemic naproxen methyl ester can be calculated from the activities towards both substrates. This results in an apparent selectivity towards the S-naproxen methyl ester of at least 99% (carboxylesterase NP and CesA) and 85% (CesB) enantiomeric excess, ee, respectively. Ibuprofen methyl ester and β-napthyl acetate assay The specific activities of the esterases towards the ibuprofen methyl ester and ß-naphtyl acetate are summarised in table III. Comparison of the specific activities of CesA, carboxylesterase NP and CesB revealed an almost 40-fold lower specific activity of CesB towards the racemic ibuprofen ester. Furthermore, all three esterases have a low specific activity towards ß-naphtyl acetate. Table III: Specific activities (U.mg-1) of CesA, carboxylesterase NP and CesB.
INTRACELLULAR CARBOXYLESTERASES OF BACILLUS SUBTILIS
53
Hydrolysis of 1,2-O-isopropylidene-sn-glycerol esters The substrate specificity and enantioselectivity of the three esterases towards IPG esters with different aliphatic side chains were determined (table IV) using non chiral GC analysis. Both CesA and carboxylesterase NP showed affinity towards both stereoisomers of the IPG esters, ranging from an specific activity 0.051 U/mg-1 (IPG caprylate esters) to 0.002 U.mg-1 (IPG acetate esters). As reported previously, the highest activity of CesB was measured when (+)-IPG caprylate was used as a substrate (0.022 U.mg-1). Surprisingly, CesB was unable to hydrolyse (-)-IPG caprylate esters. In contrast to CesA and carboxylesterase NP, CesB was unable to hydrolyse (-)- and (+)-IPG butyrate. (-)- and (+)-IPG acetate were hydrolysed by CesB. Table IV: Enantioselective properties of CesA, carboxylesterase NP and CesB towards IPG esters.
Enzyme IPG acetate IPG butyrate IPG caprylate E-value ee (%) E-value ee (%) E-value ee (%) CesA (B. subtilis 168)
1.0 2.0 1.3 2.3 1.9 26.1
Carboxylesterase NP (B. subtilis Thai I-8)
1.1 2.6 1.1 3.6 1.9 24.1
CesB (B. subtilis 168)
1.3 10.1 1* 1* >200 >99.9
1*: below detection (0.6 ng IPG). The enantioselectivity towards IPG was determined in a kinetic resolution experiment using a chiral GC method. Table IV summarises these enantioselectivities. Both CesA and carboxylesterase NP showed only modest enantioselectivities towards the IPG esters, ranging from an E-value of 1.3 (CesA, IPG butyrate esters) to 1.0 (CesA, IPG acetate esters). As CesB was unable to hydrolyse (-)-IPG caprylate esters (conversion below detection), the estimated E-value of CesB towards IPG-caprylate was >200 (corresponding to an estimated enantioselectivity of at least 99.9%). Additionally, CesB was unable to hydrolyse (-)- and (+)-IPG butyrate, and (-)- and (+)-IPG acetate were hydrolysed but only modest enantioselectivities were observed (E-value is 1.9). Hydrolysis of α-methylbenzyl acetate, N-acetyl phenylalanine methyl ester, α-methoxyphenyl acetic acid methyl ester and β-phenyl lactic acid methyl ester CesB was able to hydrolyse the α-methylbenzyl acetate and N-acetyl phenylalanine methyl ester, whereas α-methoxyphenyl acetic acid methyl ester and β-phenyl lactic acid methyl ester were not hydrolysed at all. However, very low specific activities were observed (data not shown). CesA and carboxylesterase NP only hydrolysed the N-acetyl phenylalanine methyl ester with a comparable specific activity compared to CesB. Comparison of the specific activities and enantioselective properties The specific activities and enantioselective properties of CesA, carboxylesterase NP and CesB are summarised in table III and table IV. Comparison of the specific activities of CesB with CesA and carboxylesterase NP towards chiral carboxylic acid esters (e.g. S-naproxen methyl ester, R-naproxen methyl ester and ibuprofen methyl ester) revealed
CHAPTER 3
54
relatively lower specific activities for CesB. In contrast, similar specific activities were observed when chiral alcohol esters were used as a substrate ((+)-IPG caprylate and (-)- and (+)-IPG acetate). Most interestingly, CesA and carboxylesterase NP showed highest enantioselectivity towards the production of the chiral carboxylic acid naproxen, whereas CesB displayed excellent enantioselective properties towards the chiral alcohol IPG. Hydrolysis of p-nitrophenyl esters The hydrolytic activity of the three esterases was studied using various p-nitrophenyl esters, varying from a C2 to a C18 alkyl chain length. The three esterases showed activity only towards short chain length esters (up till C8). CesA and carboxylesterase NP showed maximal activity when C6 esters were used as a substrate whereas CesB had maximum activity towards the C8 ester. Interestingly, CesB was unable to hydrolyse the C6 ester of p-nitrophenol (figure 3).
Figure 3: Hydrolysis of p-nitrohenyl esters with different aliphatic side chain, ranging from C2 to C18 with C2 stepwise, by CesA, carboxylesterase NP and CesB. The assay conditions were described in the experimental procedures. Biochemical characteristics The hydrolysis of the S-naproxen methyl ester was studied at different temperatures in order to determine the influence of the temperature on the enzymatic reaction (figure 4). Maximal activities were observed at 30ºC for CesA and carboxylesterase NP, while maximal activity and stability of CesB was observed at 40ºC. Incubation at higher temperatures resulted in a rapid inactivation of esterase activity. Above 50ºC, no enzymatic activity could be observed. To exclude proteolytic degradation of the enzymes during the incubation at 40ºC, these samples were analysed by SDS PAGE and Coomassie staining. Comparison with the samples incubated at 30ºC revealed that no major additional protein bands in addition to the 34 kDa band could be observed on the stained gel.
INTRACELLULAR CARBOXYLESTERASES OF BACILLUS SUBTILIS
55
Figure 4: Effect of the temperature (4, 20, 30, 40, 50 and 60 ˚C) on the hydrolysis of the S-naproxen methyl ester by CesA, carboxylesterase NP and CesB. The assay conditions were described in the experimental procedures. Investigation of the influence of the pH on S-naproxen methyl ester hydrolysis was performed at pH 3 to 11 (figure 5). All three esterases displayed maximal activity at pH 7.5 (MOPS buffer), while the enzymes were inactive below pH 6. Compared to CesA and carboxylesterase NP, CesB showed a relatively high activity at pH 11 (53%; 28% and 5% of maximal activity for CesB, carboxylesterase NP, CesA, respectively). Analysis of the samples incubated at pH 7.5, 9 and 11 using SDS PAGE, revealed that the enzymes were not degraded.
Figure 5: Effect of pH on the hydrolysis of the S-naproxen methyl ester by CesA, carboxylesterase NP and CesB. The assay conditions were described in the experimental procedures.
CHAPTER 3
56
Discussion Today, more than 50% of the organic pharmaceuticals are chiral 6. Since chiral compounds represent more than 50% of the world-wide most frequently prescribed drugs, the interest in the preparation and isolation of chiral drugs has increased dramatically. However, although it is commonly accepted that the specific effect of a drug is caused by just one enantiomer out of a racemic mixture of compounds, both enantiomers are still frequently applied as it has been difficult to properly separate them up till now 5,6. In recent years, the use of enzymes for the preparation of optically enriched compounds has become an alternative to chemical synthesis. Bacterial enzymes, like lipases and esterases, are capable of enantioselective hydrolysis and esterification in an environmentally friendly and cheap process. The advantageous characteristics of esterases reside in their stability, their activity in organic solvents, the fact that they do not require cofactors, and their high regio- and stereospecificity. These properties make esterases attractive biocatalysts for organic chemistry 40.
Over the past years, many esterases have been cloned and overexpressed. A large number of bacterial esterases have been described as well 41,42,183-186. However, only a few of them are useful as biocatalysts in kinetic resolution experiments. In the present study, we have shown that genome analysis has provided an excellent tool for the isolation of two novel bacterial esterases with marked enantioselective properties. One of these esterases, CesA of B. subtilis 168 showed an identity of 98% to the carboxylesterase NP of B. subtilis Thai I-8. This Thai I-8 esterase was characterised in 1994 as an effective biocatalyst in the racemic resolution of propionate esters with an aromatic ring containing a 2-substituent, such as 2-aryl propionates, 2-(aryloxy)propionates and N-aryl alanine esters (> 99% ee) 42-44,187. Relatively slow hydrolysis and poor enantioselectivities were observed when methyl acetate, 2-substituted butyrates, 2-substituted pentanoates, 2-substituted 2-phenyl acetates or amino acid esters were used as a substrate 187. The activity and enantioselectivity towards substrates with chirality residing in the alcohol part of the ester were also investigated, but the majority of these substrates, such as IPG and 1-phenyl ethanol esters, showed rather poor enantioselectivities 187. In this study we have compared the activities and enantioselectivities of the homologous CesA of B. subtilis 168 with carboxylesterase NP. No marked differences were observed between these two enzymes and the highest enantioselectivities were obtained using substrates with chirality residing in the carboxylic acid part of the ester, such as the NSAIDs naproxen and ibuprofen. The obvious explanation for the observed similarities in specific activity is found mainly in the 98% identical sequence.
The other esterase CesB, was identified as a paralogue of carboxylesterase NP. We have previously reported the isolation and cloning of this intracellular esterase of B. subtilis 168 in order to determine whether the 36% difference in amino acid sequence resulted in altered stereospecific characteristics towards IPG esters. Comparison of the specific activities and enantioselective properties of CesB with CesA and carboxylesterase NP showed in fact some striking differences. First of all, the catalytic activity and enantioselectivity of CesB towards substrates with chirality in the carboxylic acid part of the esters was relatively low. For example, the specific activity towards the methyl ester of S-naproxen was approximately 25-fold lower. In line with these results, the catalytic activity of CesB towards another 2-aryl propionate ester, the ibuprofen methyl ester, was almost 40-fold
INTRACELLULAR CARBOXYLESTERASES OF BACILLUS SUBTILIS
57
lower compared to CesA and carboxylesterase NP. Secondly, CesB combined very narrow substrate specificity towards IPG esters with markedly enantioselective properties, whereas CesA and carboxylesterase NP could hydrolyse all IPG esters, although without enantioselectivity. Due to this enantioselectivity, CesB can be used for chiral resolution of IPG caprylate. Although the specific activity of CesB towards (+)-IPG caprylate is almost 200 times lower than the specific activity of carboxylesterase NP towards S-naproxen methyl ester, it could be sufficient for the development of a (+)-IPG production process. Compared to other studies, the amount of IPG produced is likely to be enough for large-scale production in a bioreactor using immobilised CesB 188. Interestingly, this narrow substrate specificity of CesB is also reflected in the hydrolysis of the p-nitrophenol esters since CesB only hydrolysed acetate (C2), butyrate (C4) and caprylate (C8) esters p-nitrophenol, while the caproate (C6) ester was not hydrolysed at all.
Although comparison of the specific activities and enantioselectivities of CesB with CesA and carboxylesterase NP towards several chiral ester substrates revealed some more marked differences, their biochemical characteristics appear to be strikingly similar. All three enzymes displayed maximal activities between 30 and 40ºC. Incubation at higher temperatures resulted in a rapid inactivation of enzymatic activity. Investigation of the pH dependency revealed a maximal activity at pH 7.5. A major difference, however, relative to CesA and carboxylesterase NP, was the high activity of CesB at pH 11.
In conclusion, we have identified three homologous intracellular carboxylesterases from B. subtilis. It should be noted that a classical enzyme screening programme would probably never have identified CesA nor CesB, since wild type cells show no activity towards (S)-naproxen and IPG esters. Comparison of the enantioselective properties towards several esters compounds revealed that rather similar physico-chemical properties (pH and temperature optimum) can be associated with rather different enantioselectivities. CesA and carboxylesterase NP seem to be more suitable for the enantioselective production of chiral carboxylic acid such as the NSAIDs ibuprofen and naproxen, whereas CesB can be applied for the enantioselective production of the chiral alcohol IPG.
Acknowledgement This project was funded by the European Commission under proposal numbers BIO4-98-0249 and QLK3-CT-2001-00519. We thank all partners for their discussions and contributions leading to the generation of this project. We thank ing. P.G. Tepper (Department of Medicinal Chemistry; University of Groningen, Groningen, The Netherlands) for the synthesis of the methyl ester of S-naproxen and Mr Rüggeberg (Max Planck Institut, Mülheim, Germany) for the synthesis of the IPG esters.
4
PHAGE DISPLAY OF AN INTRACELLULAR CARBOXYLESTERASE OF
BACILLUS SUBTILIS: A COMPARISON OF THE SEC AND TAT
PATHWAY EXPORT CAPABILITIES
MELLONEY J. DRÖGE*, YKELIEN L. BOERSMA*, PETER G. BRAUN,
ROBBERT JAN BUINING, MATTIJS K. JULSING, KARIN G.A. SELLES,
JAN MAARTEN VAN DIJL & WIM J. QUAX
* These authors contributed equally to this work
APPL ENVIRON MICROBIOL 2006; 72: 4589-4595
PHAGE DISPLAY OF CARBOXYLESTERASE A
61
Phage display of an intracellular carboxylesterase of Bacillus subtilis:
a comparison of the Sec and Tat pathway export capabilities Using the phage display technology proteins can be displayed at the surface of bacteriophages as a fusion to one of the phage coat proteins. Here, we describe the development of this method for the fusion of an intracellular carboxylesterase of Bacillus subtilis to the phage minor coat protein g3p. Carboxylesterase A was cloned in the g3p-based phagemid pCANTAB 5E upstream of the sequence encoding the phage g3p and downstream of a signal peptide-encoding sequence. The phage-bound carboxylesterase was correctly folded and fully enzymatically active as determined from the hydrolysis of the naproxen methyl ester with Km values of 0.15 mM and 0.22 mM for the soluble and phage displayed carboxylesterase, respectively. The signal peptide directs the encoded fusion protein to the cell membrane of Escherichia coli, where phage particles are assembled. In this study, we assessed the effects of several signal peptides, both Sec- and Tat- dependent, on the translocation of the carboxylesterase in order to optimise the phage display of this enzyme normally restricted to the cytoplasm. Functional display of Bacillus carboxylesterase A could be achieved when Sec-dependent signal peptides were used. Although a Tat-dependent signal peptide can direct carboxylesterase translocation across the inner membrane of E. coli, proper assembly into phage particles does not seem to occur.
Introduction In the past decade, the most remarkable successes from protein engineering have been the result of combining random mutagenesis and screening by means of a high-throughput assay 189. Unfortunately, for many enzymes, such as esterases and lipases, no high-throughput methods are available and, consequently, the evaluation of the enantioselectivity of lipases and esterases is dependent on time-consuming assays 40. Thus, it would be highly advantageous if the screening process could be combined with a rapid selection method that limits the amount of mutants to be assayed.
Phage display 190 is a well-defined technique that has lead to a breakthrough in selection methodology for enzymes with desirable properties from a pool of mutants. Derivatives of M13 filamentous phages, a phage particle with a single stranded genome encapsulated by the phage coat proteins 190, are most commonly used for display in Escherichia coli. Enzymes can be expressed as a fusion to one of the M13 phage coat proteins, such as the g3p protein 190. As phage particles are assembled in the cell envelope of E. coli, translocation of the g3p fusion protein across the inner membrane of E. coli is a prerequisite for proper phage display. The g3p protein is synthesised with an 18-residue amino-terminal signal peptide that targets this protein to the E. coli Sec machinery for membrane insertion 191. Theoretically, any protein fused to the amino-terminal region of the g3p protein that is efficiently translocated across the inner membrane and that is able to enter the phage assembly site can be presented as a fusion protein on M13 phages 190.
CHAPTER 4
62
Today, phage display is applicable for the selection of small peptides, antibody fragments and enzymes. For instance, many enzymes, such as amylases 101, ß-lactamases 102, lipases 103, and transferases 192, have successfully been displayed on bacteriophages. Most of these are extracellular enzymes exported from the cytoplasm of the homologous host in a signal peptide-dependent manner. An important exceptions is the glutathion-S-transferase from Schistosoma japonicum 192.
In the general procedure for phage display, a gene of interest is cloned in a phagemid vector downstream of the signal sequence of g3p or pelB in order to direct the corresponding protein through the cell membrane via the Sec-dependent translocation pathway. Apart from the Sec pathway, which transports unfolded proteins over the inner membrane of E. coli 193,194, the twin-arginine translocation (Tat) pathway can be distinguished. In sharp contrast to the Sec routing, the Tat pathway of E. coli seems to accept only folded proteins for membrane translocation 195,196. In this context, we were curious to know whether translocation of a g3p fusion protein via the Tat pathway of E. coli would result in productive phage display of a cytoplasmic protein. To verify this idea, we compared the capability and effectiveness of the Sec-specific signal peptides of E. coli TEM-ß-lactamase (SpBla) and g3p (SpG3p) and the Tat-specific signal peptide of the E. coli trimethylamine N-oxide (TMAO) reductase (TorA; the signal peptide is referred to as SpTor) in the export and phage display of B. subtilis carboxylesterase NA (CesA) and lipase A (LipA). SpBla was used, as its Sec-specificity is very well documented 197, whereas SpTor was selected for this purpose, because it was previously shown to direct the export of heterologous or truncated proteins 198,199. Furthermore, LipA of B. subtilis, a lipase with a twin arginine signal peptide yielding partial dependence on TatC 200, was used as a control protein for display, because functional phage display of this enzyme with the help of the g3p signal peptide was recently demonstrated 103. Importantly, the functional phage display of both CesA and LipA is of particular biotechnological interest as the corresponding displayed fusion proteins can be used for the selection of improved variants for the enantioselective conversion of several interesting pharmaceutical compounds as was recently shown for the selection of an S-(+)-1,2-O-isopropylidene-sn-glycerol specific LipA mutant 115,162. Experimental procedures Plasmids, bacterial strains and media The plasmids and bacterial strains that were used in the present study are listed in table I. E. coli HB2151∆tatC was constructed by P1 transduction of the tatC::Spec allele 201. Helper phage M13K07 and pCANTAB 5E were purchased from Pharmacia (Amersham Pharmacia Biotech, Uppsala, Sweden). Genencor International (Leiden, The Netherlands) provided a fermentor broth of the strain B. subtilis 1051, producing LipA 35. LB medium contained: Bactotrypton (1% w/v), Bacto yeast extract (0.5% w/v) and sodium chloride (0.5% w/v); 2xTY medium contained: Bactotrypton (1.6% w/v), Bacto yeast extract (1% w/v) and sodium chloride (0.5% w/v). Antibiotic agents (Duchefa Biochemie, Haarlem, The Netherlands) were used in the following concentrations: ampicillin 100 µg.mL-1, kanamycin 50 µg.mL-1.
PHAGE DISPLAY OF CARBOXYLESTERASE A
63
Table I: Bacterial strains and plasmids
Strain or plasmid Genotype and/or properties Source or reference E. coli strains TG-1 supE K-12, ∆(lac-pro), thi, hdsD5/F’,
traD36, proAB, laqIq, lacZ∆M15 Amersham Pharmacia Biotech, Uppsala, Sweden
Plasmids pCANTABSpBlaCesA pCANTAB 5E derivative containing the
B. subtilis 168 cesA gene downstream of the bla signal sequence
This work
pCANTABSpG3pCesA pCANTAB 5E derivative containing the B. subtilis 168 cesA gene downstream of the g3p signal sequence
This work
pCANTABSpTorACesA pCANTAB 5E derivative containing the B. subtilis 168 cesA gene downstream of the torA signal sequence
This work
pCANTABSpBlaLipA pCANTAB 5E derivative containing the B. subtilis 168 lipA gene downstream of the bla signal sequence
This work
pCANTABSpG3pLipA pCANTAB 5E derivative containing the B. subtilis 168 lipA gene downstream of the g3p signal sequence
This work
pCANTABSpTorALipA pCANTAB 5E derivative containing the B. subtilis 168 lipA gene downstream of the torA signal sequence
This work
Chemicals The methyl ester of S-naproxen was provided by Prof. H.V. Wikström (Department of Medicinal Chemistry, University of Groningen, Groningen, The Netherlands). Butyrate esters of both enantiomers of 1,2-O-ispropylidene-sn-glycerol (IPG) were provided by Prof. M.T. Reetz (Max-Planck Institut für Kohlenforschung, Mülheim, Germany). p-Nitrophenyl caprylate was purchased from Sigma Chem. Co. (Axel, The Netherlands). DNA techniques Recombinant DNA techniques were performed as described by Sambrook et al 176. Plasmid DNA was prepared as described by Birnboim & Doly 177. DNA purification was performed using the Qiaquick Gel Extraction kit (Qiagen, Hilden, Germany). Construction of the phagemids The LipA encoding gene (lipA) was cloned in the phagemid pCANTAB 5E, downstream of a modified g3p signal sequence, as described previously 103. The cesA gene (cesA) sequence
CHAPTER 4
64
was amplified from the chromosomal DNA of B. subtilis 168 using napfor1 (5’-GCATGAATCATAGGCCCAGCCGGCCATGGCACAAAACCATTCATCTAGTATTCC-3’) and naprev1 (5’-GATCGTTAGAATGCGGCCGCCCGTGAAATGCCTGTT-3’) primers. PCR was performed using Pfu DNA polymerase (Stratagene, La Jolla, CA, USA). The amplified gene fragment was cloned in E. coli TG-1 into the SfiI and Eco52I sites of a modified pCANTAB vector 202. To exchange the g3p signal sequence (VKKLLFAIPLVVPFYAAQPAMA) for the signal sequences of E. coli SpBla (MSIQHFRVALIPFFAAFCLPAMA) or E. coli SpTorA (MNNNDLFQASRRRFLAQLGGLTVAGMLGPSLLTPRRATAAMA), a PCR was performed using primers 5’-CCCAAGCTTGGTACCGTTGGAGCCTTTTTTTTGGAGATTTTCAACATGAAATTTGTAAAAAGAAGG-3’ in combination with 5’-GCGGCCATGGCAGGAAGGCAAAATGC-3’, and primers 5’-CCCAAGCTTGGTACCGTTGGAGCCTTTTTTTTGGAGATTTTCAACATGAAATTTGTAAAAAGAAGG-3’ in combination with 5’-GCGGCCATGGCCGCAGTCGCACGTCGCGGCG-3’ for the amplification of SpBla from template pUC18 203 and SpTorA from pJDT1 204, respectively. The PCR programme consisted of two steps. In the first step 4 min at 94°C, followed by 8 cycles of 1 min at 94°C, 2 min 50°C and 1 min at 72°C were adopted. Then, 5 pmol of the primer commonsmall (5’-CCCAAGCTTGGTACCGTTGG-3’), was added and 22 cycles of 1 min at 94°C, 2 min 50°C and 1 min at 72°C were applied. At the end, DNA synthesis was finished for 10 min at 72°C. The amplified gene fragments cloned in E. coli TG-1 into the HindIII and NcoI sites of pCANTABSpG3pCesA and pCANTABSpG3pLipA, respectively. Isolation of the periplasmic fraction, spheroplasts and whole cell extracts E. coli HB2151 was grown in 50 mL tubes containing 10 mL 2xTY medium, ampicillin and 1 mM of IPTG. The cultures were incubated at 37ºC at 250 rpm for 16 h. The OD600 was recorded and the cells were harvested and resuspended in 10 mM Tris HCl, pH 7.4. After centrifugation, the cells were resuspended in 200 µL buffer containing 10 mM Tris HCl, pH 8.0, 25% sucrose, 2 mM ethylene diamine tetra acetic acid (EDTA) and 0.5 mg.mL-1 lysozyme. After incubation on ice for 20 min, 50 µL buffer containing 10 mM Tris HCl, pH 8.0, 20% sucrose and 125 mM MgCl2 was added. The suspension was centrifuged at 12,000 x g for 10 min and the supernatant, representing the periplasmic fraction, was isolated and used as enzyme solution in the activity assay. The pellet was resuspended in 200 µL buffer containing 50 mM Tris HCl, pH 8.4, and 2 mM EDTA. The suspension, representing the spheroplasts, was used as enzyme solution in the activity assay. The protein content of each fraction was determined by a Bradford assay in duplicate using bovine serum albumin (BSA) as a standard (Pierce, Rockford, Illinois, USA). To obtain whole cell extracts, cells were centrifuged at 12,000 x g and taken up in 200 µL buffer containing 50 mM Tris HCl, pH 8.4, and 2 mM EDTA. Phage rescue 1010 helper phages were added to exponential phase growing E. coli TG-1 cells transformed with the plasmids mentioned in table I (phage-to-bacterium ratio 30:1), followed by 16 h of
PHAGE DISPLAY OF CARBOXYLESTERASE A
65
growth at 28ºC in a glucose depleted 2xTY medium containing ampicillin and kanamycin. Phages were precipitated by the addition of 5% w/v polyethylene glycol (PEG4000) in 2.5 M NaCl. After centrifugation, the phages were resuspended in 2 mL 10 mM Tris HCl buffer, pH 7.4, containing 1 mM EDTA, and filtered through a 0.45 µm filter (Millipore, Bedford, MA, USA). The number of phage particles in the suspension was determined by absorption spectroscopy according to 205 using a NanoDrop ND-1000 (NanoDrop Technologies, Wilmington, USA). The protein content was also determined by performing a Bradford assay using BSA as a standard. Electrophoresis Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS PAGE) was performed on 11% separating and 4% stacking gels for LipA 178, and 12.5% separating and 3% stacking gels for CesA. Molecular mass markers were purchased from Bio-Rad. After electrophoresis, proteins were blotted to nitrocellulose and immunostained with a rabbit antiserum against LipA or CesA or with mouse monoclonal antibodies against g3p (PSKAN3; MoBiTec, Göttingen, Germany). Detection of the antibody was performed with alkaline phosphatase-conjugated antibodies against rabbits (LipA and CesA antiserum) or mice (g3p antibody). Enzyme kinetics Enzymatic activity of LipA was determined spectrophotometrically by the p-nitrophenyl caprylate assay as described previously 36. Esterase activity was determined using a naproxen methyl ester assay 162. The Michaelis-Menten constant (Km), specific activities and turnover numbers of the soluble and phage-bound CesA (phages were produced using E. coli TG-1 transformed with pCANTABSpBlaCesA) were determined with S-naproxen methyl ester substrate concentrations between 0.25 mM and 0.75 mM. All data were expressed as mean ± SEM (n = 3). The statistical significance of differences was tested at a significance level of p < 0.05 using a two-tailed Student's t-test. For the analysis of IPG butyrate ester hydrolysis, assays and GC analyses as described by Dröge et al were performed 180. Results Construction of the phagemids With the ultimate aim to display LipA and CesA of B. subtilis on M13 phages, the corresponding genes were cloned in the phagemid pCANTAB 5E, downstream of a modified g3p signal sequence and upstream of a collagenase cleavage site, six consecutive histidine residues (His-tag), an amber stop codon, and the sequence encoding residues 3-406 of the g3p coat protein 103. The original g3p maturation site in this phagemid, SHS, was modified to AAQPAMA in order to better resemble the consensus for signal peptidase I cleavage 202. The His-tag can be used for the purification of wild-type and mutant enzymes. The collagenase cleavage site can be used for phage rescue during phage display selection 103.
CHAPTER 4
66
In order to compare the effect of different Sec- and Tat-specific signal peptides in phage display, the sequences encoding the signal peptides SpBla and SpTor were used to replace the g3p signal sequence in pCANTABSpG3pLipA. The resulting constructs were stably maintained in E. coli HB2151 and E. coli HB2151∆tatC. Export and phage display of Bacillus LipA
Processing of LipA The periplasmic fraction, spheroplasts and whole cell extracts of E. coli HB2151 and E. coli HB2151∆tatC transformed with the plasmids pCANTABSpBlaLipA, pCANTABSpG3pLipA, or pCANTABSpTorALipA were isolated to determine the processing of LipA and the activity towards p-nitrophenyl caprylate. As the TAG stop codon is not suppressed in these strains, the different forms of LipA encoded by the three plasmids will not be fused to g3p. SDS PAGE under reducing conditions and Western blot analysis with a rabbit antiserum against LipA detected mature LipA at approximately 21 kDa, corresponding to the molecular mass of the His-tagged enzyme (figure 1A, 1C and 1E).
Figure 1: Detection of LipA in the periplasmic fraction, spheroplasts, and whole-cell extract (10 µg of protein per lane). SDS PAGE (11% gel), Western blotting, and immunostaining with a rabbit antiserum against LipA were performed with the periplasmic fractions (A), spheroplasts (C) and whole-cell extracts (E) of E. coli HB2151 and E. coli HB2151∆tatC transformed with pCANTABSpBlaLipA (lanes 1 and 2), pCANTABSpG3pLipA (lanes 3 and 4), or pCANTABSpTorALipA (lanes 5 and 6). Lanes 1, 3, and 5 contained samples from E. coli HB2151, and lanes 2, 4, and 6 contained samples from E. coli HB2151∆tatC. Activities of LipA in the periplasmic fraction (B) and in the lysed spheroplasts (D) were determined using p-nitrophenyl caprylate as the substrate. +, enzyme activity; –, no enzyme activity.
PHAGE DISPLAY OF CARBOXYLESTERASE A
67
The different precursor proteins were detectable at apparent molecular masses of 23 to 27 kDA, depending on the signal peptide used (SpBlaLipA 24 kDa; SpG3pLipA 23 kDa; SpTorALipA 27 kDa, respectively). Notably, for some signal peptide-LipA fusions, the SpTorALipA fusion in particular, several distinct precursor forms were observed. During isolation of the periplasmic fractions, some cell lysis may have occurred as precursor proteins were detectable in these fractions. This effect was most prominent in strains containing pCANTABSpTorALipA, suggesting that the production of the SpTorALipA fusion protein makes the cells more prone to cell lysis, or that the corresponding precursor forms are not efficiently retained in the inner membrane. Conversely, mature LipA was observed in the spheroplasts, which is obviously explained by the presence of membranes in these fractions and the lipophylic behaviour of lipases in general. However, comparison of the periplasmic and spheroplast fractions revealed that, with the exception of the SpTorALipA-producing E. coli strains, the relative amounts of the different precursor proteins were lower in the periplasmic fractions than in the spheroplasts.
Most interestingly, LipA was present exclusively in precursor forms in the periplasmic and spheroplast fractions of the ∆tatC mutant transformed with plasmid pCANTABSpTorALipA. As the precise fusion between SpTorA and LipA, encoded by pCANTABSpTorALipA, resulted in the accumulation of relatively high levels of SpTorALipA precursor forms in the wild type E. coli strain, an alternative SpTorALipA fusion was made containing the original signal peptidase recognition site of SpTorA and the first three residues of the mature TorA protein (pCANTABSpTorA-AQAATD-LipA) (figure 2). This, however, did not result in reduced levels of SpTorALipA precursor accumulation. Importantly, none of the two SpTorALipA constructs was processed to mature LipA in the ∆tatC strain, indicating that they require a functional Tat machinery for membrane translocation and subsequent processing by signal peptidase.
Figure 2: TatC-dependent processing of SpTorALipA fusions. SDS-PAGE (11% gel), Western blotting, and immunostaining with a rabbit antiserum against LipA were performed with cell lysates of E. coli HB2151 and E. coli HB2151∆tatC transformed with pCANTABSpTorA-AQAATD-LipA (lanes 1 and 2, respectively) and cell lysates of E. coli HB2151 and E. coli HB2151∆tatC transformed with pCANTABSpTorALipA (lanes 3 and 4, respectively). To investigate the enzymatic activity of LipA in the periplasmic fraction and lysed spheroplasts of E. coli HB2151 and the ∆tatC mutant, the specific activity towards p-nitrophenyl caprylate was determined (figure 1B and 1D). All fractions containing mature LipA were able to hydrolyse the caprylate ester of p-nitrophenol (indicated by a +). Note
CHAPTER 4
68
that no LipA activity was detectable in fractions derived from the ∆tatC mutant producing SpTorALipA (indicated by a -), and that the same was true for fractions of E. coli strains not producing LipA (data not shown). Together, these results show that LipA precursors are translocated across the membrane and processed to the mature form in E. coli HB2151 both if using a Sec-dependent or a Tat-dependent signal peptide. Phage display of LipA E. coli TG-1 cells, transformed with pCANTABSpBlaLipA, pCANTABSpG3pLipA, or pCANTABSpTorALipA, respectively, were infected with M13K07 helper phages to produce phage particles containing the phagemid genome and a mixture of wild-type g3p and LipA-g3p fusion proteins. The LipA fusion to g3p results from a partial suppression of the TAG stop codon in the E. coli TG-1 host cells. To visualise the presence of phage-bound LipA, phage particles were analysed by SDS PAGE under reducing conditions and immunoblotting with a rabbit antiserum against LipA (figure 3A).
Figure 3: Detection of phage-bound LipA. A) LipA-g3p fusion proteins in 0.25-µg phage suspensions were visualized by SDS PAGE, Western blotting, and immunostaining with rabbit antisera against LipA (αLipA) (left panel) or mouse monoclonal antibodies against g3p (αg3p) (right panel). Phages were isolated from the growth media of E. coli TG-1 cells transformed with pCANTABSpBlaLipA (SpBla), pCANTABSpG3pLipA (SpG3p), or pCANTABSpTorALipA (SpTorA). B) Hydrolysis of p-nitrophenyl caprylate by phage-bound LipA. +, hydrolysis; –, no hydrolysis. A LipA-g3p fusion protein was detectable in phages isolated from the growth media of E. coli TG-1 cells transformed with pCANTABSpBlaLipA or pCANTABSpG3pLipA. This fusion protein had an apparent molecular mass of approximately 85 kDa, which corresponds to the apparent molecular mass of LipA plus a g3p protein on SDS PAGE gels 206. Notably, the LipA-g3p fusion protein was absent from the phage suspension derived from E. coli TG-1 cells transformed with pCANTABSpTorALipA. Previously, it has been demonstrated by the determination of the Michaelis-Menten constants, and the turnover numbers of the soluble and phage-bound LipA (phages were produced using E. coli TG-1 transformed with pCANTABSpG3pLipA) that phage-bound LipA was correctly folded and fully enzymatically active 103. To assess whether the specific lipase activity in the different phage suspensions corresponded with the amount of fusion protein, the activity towards p-nitrophenyl caprylate was determined. Importantly, figure
PHAGE DISPLAY OF CARBOXYLESTERASE A
69
3B demonstrates indeed that the presence of fusion protein correlates with LipA activity. Overall, the highest lipase activity and highest amount of fusion protein were obtained when Sec-specific signal peptides were used. Although the Tat-specific signal peptide of TorA can be used to direct export of LipA of B. subtilis to the periplasmic space of E. coli, proper phage display of the TatC-dependently exported enzyme seems to be impaired. Export and phage display of Bacillus CesA
Processing of CesA To investigate the export and processing of CesA, the periplasmic fraction, spheroplasts and whole cell extracts of E. coli HB2151 and E. coli HB2151∆tatC transformed with plasmids pCANTABSpBlaCesA, pCANTABSpG3pCesA, or pCANTABSpTorACesA were isolated, and SDS PAGE under reducing conditions and Western blot analyses were performed with a rabbit antiserum against CesA (figure 4A, 4C and 4E). Mature CesA with the predicted molecular mass of 36 kDa was detectable in both strains. Furthermore, CesA precursor forms were detectable at apparent molecular masses ranging from 38 to 40 kDa (SpBlaCesA, 38 kDa; SpG3pCesA, 38 kDa; SpTorACesA, 40 kDa, respectively) in most periplasmic fractions and spheroplasts. Again, precursor forms of CesA were detectable in the periplasmic fractions, indicating that the E. coli strains were either subject to lysis upon spheroplasting, or that the precursors are not effectively retained in the inner membrane. Importantly, figure 4 shows that the cytoplasmic protein CesA of B. subtilis can be exported to the periplasm of E. coli using Sec-specific signal peptides. Mature CesA could be detected in the periplasmic fraction of E. coli HB2151 cells producing SpTorACesA as well, though most of the mature CesA was present in the cytoplasmic fraction of these cells. In contrast, the SpTorACesA produced in the ∆tatC mutant remained mainly in the precursor form, which is consistent with the fact that SpTorA is a Tat-specific signal peptide. These observations suggest that CesA can be translocated across the membrane via the Tat machinery.
Determination of the IPG hydrolysing activity of CesA in the periplasmic fraction and lysed spheroplasts of E. coli HB2151 and the ∆tatC mutant revealed that an endogenous IPG hydrolase is present in the cytoplasm of the E. coli. This complicated the determination of the specific CesA activity in the lysed spheroplast fractions. In contrast, no endogenous IPG hydrolase activity was detectable in the periplasmic fractions of E. coli HB2151 and the ∆tatC mutant (figure 4B). Although the level of IPG conversion was very low, enzymatic activities could be demonstrated in most of the periplasmic fractions of E. coli HB2151 and the ∆tatC mutant (enzymatic activity indicated by a +, higher activity indicated by ++). No detectable CesA activity was, however, observed in the periplasmic fraction of E. coli HB2151 ∆tatC transformed with plasmid pCANTABSpTorACesA (indicated by a -). In a further attempt to determine the activity of CesA in lysed spheroplasts, the specific activity towards methyl ester of S-naproxen was explored (figure 4D). Although the rate of conversion to S-naproxen was low, CesA activity could be demonstrated in most of the lysed spheroplast fractions. However, no S-naproxen methyl ester hydrolysis was detectable in the lysed spheroplast fraction of the ∆tatC mutant transformed with pCANTABSpTorACesA. These findings imply that the CesA precursor forms detectable in these spheroplasts and periplasmic cell fractions are not enzymatically active.
CHAPTER 4
70
Figure 4: Detection of CesA in the periplasmic fraction, spheroplasts, and whole-cell extract (20 µg of protein per lane): SDS-PAGE (12.5% gel), Western blotting, and immunostaining with a rabbit antiserum against CesA of the periplasmic fractions (A), spheroplasts (C), and whole-cell extracts (E) of E. coli HB2151 and E. coli HB2151∆tatC transformed with pCANTABSpBlaCesA (lanes 1 and 2), pCANTABSpG3pCesA (lanes 3 and 4), or pCANTABSpTorACesA (lanes 5 and 6). Lanes 1, 3, and 5 contained samples from E. coli HB2151, and lanes 2, 4, and 6 contained samples from E. coli HB2151∆tatC. (B) Hydrolysis of racemic esters of 1,2-O-ispropylidene-sn-glycerol butyrate. (D) Hydrolysis of the methyl ester of S-naproxen. +, enzymatic activity; ++, higher enzymatic activity; –, no enzymatic activity. Phage display of CesA Phage particles were produced using E. coli TG-1 cells transformed with pCANTABSpBlaCesA, pCANTABSpG3pCesA, or pCANTABSpTorACesA. SDS-PAGE under reducing conditions and Western blot analysis with mouse monoclonal antibodies with a rabbit antiserum against CesA and against g3p were performed. The apparent molecular mass of the CesA-g3p fusion protein was approximately 95 kDa, which corresponds to the masses of CesA plus a g3p protein. This fusion protein was detected with the CesA antibody at 95 kDa (figure 5A). Two specific protein bands reacting with the antibodies against the g3p coat protein were detectable (figure 5B). These bands corresponded to g3p (apparent molecular mass of 65 kDa 206) and to a CesA-g3p fusion protein. Notably, the CesA-g3p fusion protein was not detectable in the phage suspension derived from cells producing SpTorACesA. The reaction of αCesA with bands around 65 kDa and 35 kDa reflects, most likely, the presence of degradation products of the CesA-g3p fusion protein. Despite the fact that only very low levels of (S)-naproxen methyl ester hydrolysis were detectable, CesA activity could be demonstrated for both phage suspensions derived from cells producing CesA-g3p with Sec-specific signal peptides.
PHAGE DISPLAY OF CARBOXYLESTERASE A
71
Figure 5: Detection of phage-bound CesA. A) CesA-g3p fusion proteins in 0.25-µg phage suspensions were visualized by SDS PAGE, Western blotting, and immunostaining with rabbit antisera against CesA (αCesA) (left panel) or mouse monoclonal antibodies against g3p (αg3P) (right panel). Phages were isolated from the growth media of E. coli TG-1 cells transformed with pCANTABSpBlaCesA, pCANTABSpG3pCesA, or pCANTABSpTorACesA. (B) Hydrolysis of the methyl ester of (S)-naproxen by phage-bound CesA. +, enzymatic activity; –, no enzymatic activity. Enzymatic activity of soluble and phage-bound Bacillus CesA To investigate whether the kinetic properties of the phage-bound CesA were unaltered, the Km and the specific activities were measured of both soluble and phage-bound CesA (table 2). The steady state hydrolysis of S-naproxen methyl ester showed that the Km of the enzyme remained unchanged (being not significantly different (p > 0.05)), suggesting that the protein is correctly folded and fully enzymatically active. In contrast, the specific activity of the phage-bound CesA was reduced. This observed difference is likely to result from the different individual weights of the soluble CesA and a phage particle (5.98 x 10-17 and 2.36 x 10-14 mg, respectively). Taken together, our present findings show that both LipA and CesA can be effectively exported from the cytoplasm of E. coli, and displayed on M13 phages with the help of Sec-specific signal peptides, but not with the Tat-specific signal peptide of TorA. Table II: Enzyme kinetics of CesA from Bacillus subtilis. (n=3)*
Discussion The present studies report for the first time the functional phage display of the cytoplasmic protein CesA of B. subtilis as a fusion to the phage M13 minor coat protein g3p. It was tempting to speculate that phage display of heterologous cytoplasmic proteins, such as Bacillus CesA, in E. coli might be enhanced if the respective fusion proteins were exported from the cytoplasm via the Tat pathway, as the Tat pathway exists for the export of intracellularly folded proteins 207,208. Remarkably, our results show that functional phage display of CesA-g3p and LipA-g3p fusion proteins could only be achieved if Sec-specific signal peptides (SpBla and SpG3p) were used for translocation of the fusion protein across the inner membrane of E. coli. In marked contrast, the use of the Tat-specific signal peptide SpTorA did not result in functional phage display of these g3p fusion proteins. These results are in accordance with the recently published results of Paschke & Höhne. These authors demonstrated that fusion proteins of mutated GFP with the C-terminal domain of g3p, using a TorA or PelB signal sequence, could not sufficiently be displayed on phages. However, phage display was ultimately achieved by transporting g3p and GFP to the periplasm independently, followed by combination using a coiled coil/disulfide strategy. The authors suggest that the unfolded g3p domain is not suitable for Tat-dependent export 109.
At present, the exact reason why Tat-specific export of the tested g3p fusion proteins did not result in their incorporation into phages remains unclear. The assembly of M13 phages occurs at sites in the cell envelope where the inner and outer membranes are in close contact 209. Prior to incorporation into the phage particle, all phage proteins are assembled in the inner membrane 210,211. Specifically, the g3p protein requires the Sec pathway for inner membrane assembly 191. Thus, at least two possible explanations for the ineffectiveness of SpTorA in phage display are conceivable. Firstly, the bacterial Tat machinery seems to accept only folded proteins for translocation 195, which may have a negative impact on the assembly of g3p fusion proteins into phages. Possibly, the CesA-g3p and LipA-g3p fusion proteins are only competent for assembly into phages if they are translocated via the Sec machinery in an unfolded state. Translocation in a folded state via the Tat machinery might render them incompetent for phage assembly. Secondly, the Tat system may not be able to sort proteins to the specific sites where phage assembly takes place. For example, the Tat pathway may export the g3p fusion proteins to the periplasm. This would hamper the assembly of these fusion proteins into phages, because they need to remain attached to the inner membrane for this purpose. However, mis-sorting of g3p fusion proteins to the periplasm seems somewhat unlikely as it has been demonstrated recently that integral membrane proteins with a carboxyl-terminal membrane anchor (like g3p) can be inserted into the membrane by a Tat-dependent mechanism 212. At least in the case of the CesA-g3p fusion, a mis-sorting event seems nevertheless a plausible explanation for the lack of phage incorporation upon Tat-dependent membrane translocation, because the mature CesA (not fused to g3p) that resulted from SpTorACesA processing was in part released into the periplasm. However, most of the mature protein was detected in the spheroplasts. In contrast, the mature forms of CesA that resulted from SpBlaCesA or SpG3pCesA processing were detected abundantly in the periplasmic cell fraction.
Remarkably, cell fractionation experiments revealed that a significant proportion of all hybrid CesA and LipA precursor proteins analysed in these studies was readily released
PHAGE DISPLAY OF CARBOXYLESTERASE A
73
from spheroplasts into the spheroplast supernatant (i.e. the periplasmic fraction). This suggests that spheroplasts of cells producing these precursor proteins are either sensitive to cell lysis or that the precursor proteins are not effectively retained in the inner membrane of E. coli HB2151 and its ∆tatC mutant derivative. Although it is presently difficult to distinguish between these two possibilities, we favour the idea that the precursors are poorly retained. This preference relates to the observation that the endogenous IPG hydrolase activity of E. coli HB2151 is detectable only in cytoplasmic cell fractions, but not in periplasmic cell fractions. However, we cannot rule out the possibility that this hydrolase activity is absent from the periplasmic fractions due to the presence of an, as yet, unidentified periplasmic inhibitor. Conversely, the presence of mature LipA and CesA in the cytoplasmic fractions can be explained by the presence of membranes in these fractions and the lipophylic behaviour of these lipases and esterases. Finally, the CesA and LipA activity assays on periplasmic and cytoplasmic fractions of the E. coli HB2151 ∆tatC mutant revealed that the SpTor precursor forms of these proteins are enzymatically inactive. Thus, it seems that the fusion of CesA and LipA to the SpTor signal peptide does not only preclude the display of these proteins on M13 phages, but also affects their ability to fold into an enzymatic active form. This could imply that these proteins do not reach the relevant folding catalysts when they are targeted into the Tat pathway. In fact, this may not only be true for mature CesA and LipA, but also for the CesA-g3p and LipA-g3p fusion proteins. If so, the inability of SpTorA to direct functional phage display may relate both to targeting and folding problems.
In conclusion, functional display of the cytoplasmic protein CesA of B. subtilis can be achieved when Sec-dependent signal peptides are used for this purpose. Although the use of a Tat-dependent signal peptide, SpTorA, can result in CesA precursor processing, the mature form of this protein remains membrane bound. Proper phage display using SpTorA seems to be impossible for this substrate. It will be a major challenge for future phage display research to elucidate the molecular mechanisms underlying these observations. It is anticipated that such studies will provide novel insights concerning the mechanism of g3p assembly into M13 phages. Acknowledgements This project was funded by the European Commission under proposal numbers QLK3-CT-2001-00519 and QLK3-CT-1999-00917. We thank all partners for their discussions and contributions leading to the generation of this project.
5
DIRECTED EVOLUTION OF BACILLUS SUBTILIS LIPASE A BY USE OF
MELLONEY J. DRÖGE*, YKELIEN L. BOERSMA*, GERTIE VAN
POUDEROYEN, TITIA E. VRENKEN, CARSTEN J. RÜGGEBERG,
MANFRED T. REETZ, BAUKE W. DIJKSTRA & WIM J. QUAX
* These authors contributed equally to this work
CHEMBIOCHEM 2006; 7: 149-157
DIRECTED EVOLUTION OF LIPASE A
77
Directed evolution of Bacillus subtilis Lipase A by enantiomeric
phosphonate inhibitors: crystal structures and phage display selection Phage display can be used as a protein engineering tool for the selection of proteins with desirable binding properties from a library of mutants. Here, we describe the application of this method for the directed evolution of Bacillus subtilis lipase A, an enzyme that has marked properties for the preparation of the pharmaceutically relevant chiral compound 1,2-O-isopropylidene-sn-glycerol (IPG). PCR mutagenesis using spiked oligonucleotides was employed for saturation mutagenesis of a stretch of amino acids near the active site. After expression of these mutants on bacteriophages, dual selection on S-(+)- and R-(-)-IPG stereoisomers covalently coupled to enantiomeric phosphonate suicide inhibitors (SIRAN Sc and Rc inhibitors, respectively) was used for the isolation of variants with an inverted enantioselectivity. The mutants were further characterised by determination of their Michaelis-Menten parameters. The 3D structures of the Sc and Rc inhibitor – lipase complexes were determined to provide structural insight into the mechanism of enantioselectivity of the enzyme. In conclusion, we have used phage display as a fast and reproducible method for the selection of Bacillus lipase A mutant enzymes with an inverted enantioselectivity.
Introduction In pharmacotherapy, the use of enantiopure drugs is often associated with improved potency and selectivity of the compound. In addition, adverse side effects resulting from the ‘wrong’ enantiomer are eliminated. As such, the production and isolation of pure enantiomers has become an important process in pharmaceutical and chemical industries. In particular, bacterial enzymes, such as lipases, have been used for the production of enantiopure intermediates and drugs, since they are capable of enantioselective hydrolysis and esterification 40,40,213-215. Moreover, in the past decades, protein engineering has opened the possibility to tailor enzymes with specific properties and interesting practical applications 216.
We chose to study the chiral intermediate 1,2-O-isopropylidene-sn-glycerol (IPG), since IPG is a starting compound in the chemical synthesis of ß-adrenoceptor antagonists. Lipase A of Bacillus subtilis 168 (LipA) is of particular interest for the production of enantiopure IPG as it shows some advantageous characteristics compared to other lipases. It has a relatively low molecular mass (181 amino acids, 19 kDa) and it can be well expressed and engineered in Escherichia coli and B. subtilis, since it does not require specific chaperones for proper folding and secretion. Furthermore, this enzyme represents one of the few examples of a lipase that does not show interfacial activation in the presence of oil-water interfaces 35,36,172. Finally, the crystal structure of B. subtilis LipA has been elucidated 27. Like other lipases, the fold of B. subtilis LipA resembles the core of enzymes of the α/ß hydrolase fold. Due to its small size and the absence of a separate lid domain, which is
CHAPTER 5
78
Table I: Hydrolysis of different racemic IPG esters by B. subtilis LipA (n=4).
* E values are defined as the ability of the enzyme to distinguish between enantiomers (E < 1 enantiopreference for (+)-IPG; E > 1 enantiopreference for (-)-IPG) present in the larger lipases, the B. subtilis LipA can be regarded as a minimal α/ß hydrolase fold enzyme.
We have already established that purified LipA can hydrolyse acetate, butyrate and caprylate esters of IPG, although hardly any enantioselective preference was observed (ee ranging from 20.3% to 73.2% for (-)-IPG, table I). Therefore, our aim was to invert and improve the enantioselective hydrolysis by B. subtilis LipA of the enantiomer of interest, S-(+)-IPG. We mutated residues 16 to 20 of the mature LipA with localised saturation PCR mutagenesis, as this region might be of particular interest for the enantioselective characteristics of this enzyme 27. Recently, it was demonstrated that LipA of B. subtilis can be functionally expressed as a fusion to the phage coat protein g3p 103. In this study, we evaluated for the first time the use of enantiomeric phosphonate suicide inhibitors for the selection of phage displayed LipA mutants with an inverted enantioselectivity. These phosphonates bind covalently to the active site serine of a lipase molecule, thus mimicking the conformation of the first transition state in ester hydrolysis (scheme 1) 103,217. Finally, the 3D structures of both enantiomers of the IPG phosphonate inhibitor bound in the lipase active site were determined to provide a structural explanation for the enantioselectivity of B. subtilis LipA.
O
O Ser77O
O O
P
O
O Ser77O
O O
Transition state Structure of mimic Scheme 1: Schematic comparison of the structure of covalently bound IPG phosphonate inhibitor mimicking the tetrahedral transition state in the ester hydrolysis.
DIRECTED EVOLUTION OF LIPASE A
79
Experimental procedures Plasmids, bacterial strains and media E. coli TG-1 (supE, K12 ∆(lac-pro), thi, hsdD5/F’, traD36, proAB, laqIq, lacZ∆-M15), E. coli HB2151 (K12 ∆(lac-pro), ara, nalr, thi/F’, proAB, laqIq, lacZ∆-M15), helperphage M13KO7 and pCANTAB 5E were purchased from Pharmacia (Amersham Pharmacia Biotech, Uppsala, Sweden). 2xTY medium contained: Bactotrypton (1.6% w/v), Bacto yeast extract (1% w/v) and sodium chloride (0.5% w/v). As antibiotic agents ampicillin (100 µg.mL-1) and kanamycin (50 µg.mL-1) (Duchefa Biochemie, Haarlem, The Netherlands) were used. Mutagenesis of the lipA gene The LipA encoding gene was cloned in the phagemid pCANTAB 5E, downstream of a modified g3p signal sequence, as described previously 103. To introduce an ApaLI restriction site in the lipA gene, a silent mutation at base pair position 27 of the lipA gene was made using the oligonucleotides lipapaLI (Life Technologies, UK): 5’-CAGTCGTTATGGTGCACGGTATTGGAGGGGC-3’ (ApaLI) and lipArev: 5’-TCAAGGTTTTGTTGCGGCCGCCTTCGTATTCTGGCC -3’ (Eco52I). Restriction sites are indicated in bold italics. To construct the mutant library, the following primer was used (Eurosequence, Groningen, The Netherlands): 00B468a: 5’-ATGGTGCACGGTATTGGAGGGGCA576557665555878GGAATTAAGAGC-3’ (ApaLI; 5 = 80% T; 6 = 80% A; 7 = 80% C; 8 = 80% G; the remaining 20% is an equal mixture of the other three bases). Recombinant DNA procedures were carried out as described by Sambrook et al 176. Plasmid DNA was prepared according to Birnboim & Doly 177. DNA purification was performed using the Qiaquick Gel Extraction kit (Qiagen, Hilden, Germany). The library was amplified from the chromosomal DNA of B. subtilis 168 using Pfu DNA polymerase (Stratagene, La Jolla, CA, USA). The gene fragment was digested with ApaLI and Eco52I (Pharmacia), and cloned in E. coli TG-1 into the ApaLI and Eco52I sites of the phagemid pCANTABLip-CH 103. Prior to phage display, the constructs were sequenced to assess the mutation ratio. The LipA F17DF19W mutant was constructed using the oligonucleotide F17DF19W 5’-TCCAGTCGTTATGGTGCACGGTATTGGAGGGGCATCAGACAATTGGGCGGGAATTAAGAG-3’ (ApaLI) and lipArev. The gene fragment was digested with ApaLI and Eco52I and cloned in E. coli TG-1 into the ApaLI and Eco52I sites of the phagemid pCANTABLip-CH. Phage rescue Phages were rescued as described by Dröge et al 103. In short, helperphages were added to exponential phase growing E. coli TG-1 cells (i.e. phage-to-bacterium ratio 30:1), followed by 16 h of growth at 28ºC in 2xTY medium. Phages were precipitated by addition of polyethylene glycol (40% w/v) in NaCl (0.6 M). After centrifugation, the phages were resuspended in Tris HCl buffer (10 mM, pH 7.4) containing ethylene diamine tetraacetic acid (EDTA, 1 mM), and passed through a 0.45 µm filter (Millipore, Bedford, MA, USA.). The number of phage particles was determined by mixing 150 µL of a diluted phage suspension (102 – 108-fold) with 150 µL exponential-phase growing E. coli TG-1 cells. After plating and overnight growth, the number of infective phages was counted.
CHAPTER 5
80
Dual selection using enantiomeric phosphonate inhibitors Enantiomeric phosphonate inhibitors (SIRAN Sc and Rc inhibitors) were synthesised according to Reetz et al 217. All selections were in assay buffer (50 mM phosphate buffer, pH 8, 0.36% v/v Triton X100 and 0.1% w/v gum arabic) at room temperature. Triton X100 was used to reduce the aspecific interaction. For selection round 1, 10 mg of the SIRAN Sc inhibitor (loading 10 mmol inhibitor.g-1 SIRAN) was washed twice for 10 minutes using 500 µL assay buffer. Note that the inhibitor does not dissolve in this solution. The reaction mixture was vertically rotated during incubation. Subsequently, lipase expressing phages were added and allowed to react for 15 minutes. Then, the supernatant was removed and the inhibitor was washed 8 times for 10 minutes to remove all non-bound and aspecifically bound phages. Afterwards, collagenase A (0.1% w/v, Roche, Mannheim, Germany) was added (1 hour incubation at 37ºC) to elute the bound phages. The total number of phages in each fraction was determined by phage counting. The phagemids were propagated by mixing the phages with exponential phase growing E. coli TG-1 cells (i.e. phage-to-bacterium ratio 30:1). After overnight growth, phages expressing lipase can be produced as described above. Subsequently, 10 mg of the SIRAN Rc and Sc inhibitors were washed separately. Lipase expressing phages of round 1 were added to the SIRAN Rc inhibitor and allowed to react for 15 minutes (selection round 2). Then, the supernatant was isolated and incubated on the SIRAN Sc inhibitor. After 5 minutes, the supernatant was removed and the inhibitor was washed 8 times for 10 minutes. Afterwards, collagenase A (0.1% w/v) was added to elute the bound phages. The total number of phages in each fraction was determined by phage counting. The phagemids were propagated using E. coli TG-1 cells and, after overnight growth, lipase expressing phages were produced as described above. This procedure was repeated another two times (selection round 3 and 4). After the fourth round of selection, the eluted phages were incubated with exponential phase growing E. coli HB2151 for isolation of lipase from the periplasm. Isolation of the periplasmic fraction E. coli HB2151 was grown in 50 mL tubes containing 10 mL 2xTY medium, ampicillin and of isopropyl-β-D-galactopyranoside (IPTG, 1 mM). The tubes were incubated at 37ºC at 250 rpm for 16 h. The OD600 was recorded and the cells were harvested and resuspended in Tris HCl buffer (10 mM, pH 7.4). After centrifugation, the cells were resuspended in 200 µL buffer containing Tris HCl, (10 mM, pH 8.0), sucrose (25% w/v), EDTA (2 mM) and lysozyme (0.5 mg.mL-1, Sigma-Aldrich, Steinheim, Germany). After incubation on ice for 20 min, 50 µL buffer containing Tris HCl (10 mM, pH 8.0), sucrose (20% w/v) and MgCl2 (125 mM) was added. The suspension was centrifuged and the supernatant, containing the periplasmic fraction, was isolated and used as enzyme solution in the IPG ester assay. The protein content of this fraction was determined by performing a Bradford assay in triplicate using bovine serum albumin (BSA) as a standard (Pierce, Rockford, Illinois, USA). 1,2-O-isopropylidene-sn-glycerol ester assay Periplasmic fractions were diluted with MOPS buffer (0.07 M, pH 7.5), containing BSA (0.2% w/v), to a final volume of 100 µL. References were diluted correspondingly but contained no enzyme solution. The esters of IPG (1 mM) were dissolved in 10 mL MOPS buffer (0.07 M, pH 7.5), containing Tween 80 (14.3% w/v), and diluted to 50 mL with
DIRECTED EVOLUTION OF LIPASE A
81
MOPS buffer (0.07 M, pH 7.5). 500 µL substrate solution was added to the enzyme solution and the final mixture was incubated in a water bath at 32°C. After incubation, 500 µL saturated NaCl solution and 10 µL internal standard solution (racemic 3-hexene-1-ol, 5 mg.mL-1 assay buffer) was added to the sample solution and the aqueous solution was extracted twice with 1 mL ethyl acetate. GC analysis was performed as described by Dröge et al 180. One unit (U) is defined as the amount of enzyme that hydrolyses 1 µmol IPG ester per minute. Enantiomeric excesses, ee, were calculated according to Chen et al and were defined as the ability of the enzyme to distinguish between enantiomers 181. All data were expressed as mean ± SEM. The statistical significance of differences was tested at a significance level of p<0.05 using a two-tailed Student's t-test. Enzyme purification and characterisation E. coli HB2151 was grown for 16 h at 37 °C in 1 L of 2xTY medium supplemented with ampicillin (100 µg.mL-1) and IPTG (1 mM). The periplasmic fraction was isolated as described above. Histidine tagged LipA was purified using an immobilised metal affinity chromatography (IMAC) resin (Talon, Clontech, Palo Alto, CA, USA). The protein was eluted with a linear gradient of imidazole (0.5 M in 10 mM Tris HCl, pH 7, containing 0.2 M NaCl). LipA N18I was purified as described by Lesuisse et al 36. Kinetic parameters were obtained by fitting the experimental data from Eadie-Hofstee plots. The mean and SEM values of four measurements were calculated. The kcat was calculated using the theoretical molecular mass of the his-tagged enzyme, 21 kDa. Binding of Rc or Sc inhibitors to crystallised lipase A Crystals were grown by vapour diffusion in hanging drops, which contained 3 µL B. subtilis LipA solution (6 mg.mL-1) and 3 µL reservoir solution (35% PEG4000, 0.1 M ethanolamine, pH 10.0, 20 mM Na2SO4 and 3 mM CdCl2) 27. The crystals were transferred to a 10 µL sitting drop with PEG4000 (45%), ethanol amine (0.1 M, pH 10.0), and Na2SO4 (20 mM) in equilibrium with the same solution. The pH was lowered stepwise per 15 minutes to pH 7.5 by removing most of the liquid surrounding the crystal followed by the addition of a similar solution of 0.5 unit lower pH. At pH 8.5 the buffer was replaced by triethanol amine (100 mM) as an additional step. Finally, the washing at pH 7.5 was done twice. To the final 10 µL drop, 1 µL of the phosphonate inhibitor (either Rc or Sc) was added. These phosphonate inhibitors form small oil drops on the surface of the drops containing the crystals. After 65 hours the crystals were flash frozen in liquid nitrogen and stored until data collection. X-ray data collection and refinement The data collection statistics are given in table II. The data were processed and scaled using DENZO and SCALEPACK 218. Reduction to structure factor amplitudes was performed with TRUNCATE 219. The B. subtilis LipA structure (PDB code 1I6W) was rigid body refined with CNS with both molecules in the asymmetric unit as separate bodies, without water molecules and Cd2+ ion 220. One round of simulated annealing was performed and the electron density at the active site was inspected. In both cases (Rc and Sc inhibitor), both molecules in the asymmetric unit clearly showed additional density extending from S77 (PDB code 1R4Z and 1R5O, for the Rc and Sc inhibitor, respectively). The electron density
CHAPTER 5
82
also showed a change of conformation of residues 10 to 15 in molecule B. The phosphonate inhibitors were built in the electron density and the loop of residues 10 to 15 in molecule B was rebuilt to fit the electron density. The structure was refined with CNS and water molecules were added. The refinement statistics are listed in table II. The solvent accessible surfaces were calculated using AREAIMOL from the CCP4 package 220,221. Table II: Data collection and refinement statistics.
Data collection Rc inhibitor Sc inhibitor Beamline ESRF, ID14-2 ESRF, ID14-2 Resolution (Å) 40-1.80 100-1.45 Space group P212121 P212121 a (Å), b (Å), c (Å) 39.49, 83.32, 94.64 39.59, 83.59, 95.08 Number of reflections 166133 347481 Number of unique reflections 26465 48857 Completeness (%) 88.4 85.6 Completeness last shell (%) 56.2 35.4 Rmerge (%) 5.5 4.2 Rmerge last shell (%) 21.1 30.5 I/sigma 19.4 27.3 I/sigma last shell 3.8 2.2
R.m.s. deviation from ideality Bond lengths (Å) 0.0047 0.0047 Bond angles (˚) 1.19 1.23 Results and discussion
Construction of the mutant library The LipA encoding gene (Genbank accession number M74010) was cloned in phagemid pCANTAB 5E as described previously 103. An ApaLI restriction site was introduced at bp 27 by a silent mutation. A spiked oligonucleotide was used to construct a saturated mutant library directed towards amino acids 16 to 20 of mature LipA (figure 1). These oligonucleotides resembled the lipA sequence for approximately 80%. Theoretically, this would yield a mutant library, which is composed mostly of single and double
DIRECTED EVOLUTION OF LIPASE A
83
Figure 1: Schematic representation of the strcture of B. subtilis LipA. The structure shows a single compact domain that consists of six β-strands in a parallel β-sheet, surrounded by five α-helices. Two α-helices are located at one side of the β-sheet, whereas the other three are found at the opposite side (including a very small α-helix consisting of ony four amino acid residues). The catalytic triad residues serine, histidine and aspartate are labelled with S, H and D, respectively. The catalytic S77 is positioned at a very sharp turn between a β-strand and an α-helix, the so-called nucleophile elbow. The letters N and C indicate the N- and C-termini, respectively. The amino acids 16-20, which have been subjected to saturation mutagenesis, are indicated in black. mutations. Sequence analysis of 24 clones revealed that 8% of the mutants had no amino acid mutation while the percentage of single, double, triple, quadruple mutated amino acids was 25%, 42%, 13%, 8%, respectively. Only 4% of the mutants contained five mutated amino acids. After transformation of E. coli TG-1, a mutant library consisting of 5 × 104 colonies was obtained. This should be more than sufficient to saturate all possible single mutations at the amino acid level. E. coli TG-1, containing the mutant library, was infected with M13KO7 helperphages to produce phage particles containing the phagemid genome and a mixture of wild type and fusion g3p. The lipase fusion to g3p results from a partial suppression of the TAG stop codon in the host cells, E. coli TG-1. The number of phages with a possible lipase fusion protein was 8.3 × 1010, as determined by phage counting. Dual selection In order to develop a selection system for phage-bound LipA mutants, two transition state analogues were synthesised 217. These phosphonates bind covalently to the active site serine, thus mimicking the conformation of the first transition state of a substrate in the active site of a lipase molecule (scheme 1) 103,217. These two immobilised inhibitors contained the substrate analogue, 1,2-O-isopropylidene-sn-glycerol, respectively (+)-IPG or (-)-IPG (scheme 2A), a leaving group, p-nitrophenol, and a glass-bead matrix connected to the phosphonate moiety through a linker (scheme 2D). The dual selection procedure applied is schematically represented in figure 2. For the initial selection of all phages with a lipase fusion, selection round 1 was designed to
CHAPTER 5
84
O
OOH
O
OOH
O
OO
O
O
OO
O
O
O
P
O
OO
NO2
O
O
P
O
OO
NO2
O
O
P
O
OO
NO2
O
O
NH
Si
O
O
OSIRAN
O
O
P
O
OO
NO2
O
O
NH
Si
O
O
OSIRAN
A) B)
C)
D)
9 9
S-(+)-IPG R-(-)-IPG Butyrate ester of S-(+)-IPG= R-(+)-IPG butyrate
SIRAN coupled phosphonate inhibitor of S-(+)-IPG= SIRAN Sc inhibitor
SIRAN coupled phosphonate inhibitor of R-(-)-IPG= SIRAN Rc inhibitor
Scheme 2: A) Chemical structure of 1,2-O-isopropylidene-sn-glycerol (IPG). B) Chemical structure of the butyrate ester of IPG. C) Chemical structure of the soluble lipase inhibitors. D) Chemical structure of the SIRAN immobilised lipase suicide inhibitors. Note that the inhibitor is enantiopure at the chiral atom of the IPG molecule (Sc or Rc inhibitor) and racemic at the phosphorous atom (mixture of Rp and Sp inhibitor). The absolute configuration of IPG changes upon attachment to the phosphorous atom or upon ester bond formation. specifically isolate the phages that were able to bind to the SIRAN Sc inhibitor. For this purpose, 8.3 × 1010 phages were incubated with the SIRAN Sc phosphonate inhibitor and allowed to react for 15 minutes. After collagenase digestion, 1.4 × 106 phages containing the phagemid genome (as indicated by the presence of an ampicillin resistance marker) were obtained. Those phages were propagated by infection of exponential phase growing E. coli TG-1 cells. After overnight incubation, helperphages were added to produce a new phage pool containing the phagemid genome and a mixture of wild type and fusion g3p. Then, selection rounds 2, 3 and 4 were designed for the selection of enzymes with inverted enantioselectivity. We have previously established that phage-bound LipA reacts covalently with the SIRAN Rc and Sc inhibitors, revealing a half life of inactivation (t½ values) of 4.3 ± 0.2 (n=3) and 5.0 ± 0.1 (n=3) minutes for the SIRAN Rc and Sc inhibitor, respectively 103. 1.0 × 1011 phages were incubated on the SIRAN Rc inhibitor and allowed to react for 15 minutes (selection round 2).
The supernatant, which contained the non-bound phages, was isolated and re-incubated on the SIRAN Sc inhibitor. After five minutes, the non-bound phages were discarded by removing the supernatant and the bound phages (1.8 × 106) were isolated by a collagenase
DIRECTED EVOLUTION OF LIPASE A
85
Table III: Dual selection experiment. The number of phase in each fraction was determined by phage counting. Note that the detection level of this method is 4 × 102 phages.ml-1. Amp: ampicillin resistance (phagemid genome); kana: kanamycin resistance (helperphages); n.d.: not determined.
Number of phages Amp Kana Total Selection round 1
8.0 × 1010 3.2 × 109 8.3 × 1010
Selection round 1 (collagenase elution)
1.4 × 106 9.0 × 103 1.4 × 106
Selection round 2
1.0 × 1011 8.0 × 107 1.0 × 1011
Selection round 2 (collagenase elution)
1.8 × 106 7.2 × 103 1.8 × 106
Selection round 3
1.0 × 1012 4.4 × 108 1.0 × 1012
Selection round 3 (collagenase elution)
1.2 × 106 n.d. 1.2 × 106
Selection round 4
1.2 × 1012 8.0 × 109 1.2 × 1012
Selection round 4 (collagenase elution)
2.0 × 106 4.0 × 104 2.0 × 106
digestion. The phagemids were propagated using E. coli TG-1 cells and a new phage pool was produced. All results concerning the dual selection method are summarised in Table III. Additionally, we have determined the initial activity of each phage pool before each round of phage selection using the p-nitrophenyl caprylate assay as described previously 103,217. This assay revealed rather similar activities for each phage pool (e.g. approximately 0.4 U.mg-1). After two additional rounds of selection (selection round 3 and 4), the eluted phages were allowed to infect E. coli HB2151. As the TAG stop codon is not suppressed in this strain, soluble lipase will be secreted into the periplasm. Enzymatic activity The specific activity and enantioselectivity of 15 randomly chosen LipA mutants from selection round 4 towards IPG butyrate esters (scheme 2B) were determined and compared with wild type LipA (table III and figure 3). The periplasmic fraction was isolated to determine the enantioselectivity towards IPG butyrate, and the DNA was isolated for sequencing purposes. 11 mutants had enantioselectivities comparable to the wild type LipA, whereas one mutant was not enantioselective at all. Sequence analysis revealed that 8 of these 11 mutants contained a wild type amino acid sequence. In this respect, it should be mentioned that prior to the selection sequence analysis of 24 randomly chosen clones revealed only 2 wild type sequences. Therefore, the selection procedure has obviously enriched the phage pool with wild type sequence. Yet, 4 of the selected clones with a wild type amino acid sequence contained mutations on the DNA level, indicating that selection
CHAPTER 5
86
Figure 2: Schematic representation of the applied dual selection method. had in fact occurred. Most interestingly, 2 mutants had an inverted enantioselectivity (ee values of +32.9% and +35.3%). Sequence analysis revealed that both inverted mutants had the same sequence, namely LipA N18I. Therefore, the procedure selects in favour of the wild type LipA or LipA N18I sequence. Table IV: Enantioselective hydrolysis of racemic esters of IPG butyrate by periplasmic fractions of LipA mutant strains (n=3).
E values are defined as the ability of the enzyme to distinguish between enantiomers (E < 1 enantiopreference for (-)-IPG; E > 1 enantiopreference for (+)-IPG). Statistical significance of differences: ** p < 0.01, *** p < 0.001.
DIRECTED EVOLUTION OF LIPASE A
87
Mutant
WT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
ee (%
)
-60
-40
-20
0
20
40
60
WT
F17L
WT
WT
WT
F17IF19N
WT WT
N18I
***
F17DF19.
***
N18I
***
**
F19L
F17YA20S
WT
WT
Figure 3: Enantioselective hydrolsys of racemic butyrate esters of IPG by LipA mutants. The assay conditions were described in the experimental procedures. Statistical significance of differences: ** p < 0.01; *** p < 0.001. Characterisation of the LipA N18I variant LipA and LipA N18I were purified and the Km and kcat values for the butyrate ester of IPG were determined (table V). Interestingly, the kcat value of LipA N18I on the butyrate ester of (-)-IPG was reduced (p < 0.05), but its Km value was not significantly lowered. As a consequence, LipA and LipA N18I do not have significant different catalytic efficiencies for butyrate ester of (-)-IPG (p > 0.05). In contrast, the Km value of LipA N18I towards the butyrate ester of (+)-IPG was significantly lowered (p < 0.01), which was reflected in a higher catalytic efficiency. Additionally, comparison of the enantiomeric excess ee towards (+)-IPG revealed an inversion of the enantioselectivity of LipA N18I compared to LipA (ee value for LipA N18I = +32.9% and +35.3%; ee value for LipA = -33.3%). Table V: Enzyme characteristics of the LipA N18I mutant (n=3).
Statistical significance of differences: * p < 0.05; ** p < 0.01.
CHAPTER 5
88
Crystal structures of LipA in complex with the Rc and Sc inhibitors The active site of B. subtilis LipA is solvent exposed, because no lid is present. It is located at the bottom of a small cleft between two loops consisting of residues 10 to 15 and 131 to 137. In the crystal the active sites are exposed to a solvent channel as well. From the electron density maps it was clear that the IPG phosphonate inhibitors (scheme 2C) bound to both LipA molecules in the asymmetric unit (molecules A and B). After refinement, the electron density for the inhibitors bound in the active site is well defined in one molecule of the asymmetric unit but slightly less defined in the other molecule (figure 4). Residues 12-16 in molecule B change conformation upon binding of the Sc or Rc inhibitor. In the unbound form, this loop partly closed the active site, but when the inhibitors are bound, it has the same conformation as in molecule A to make space for the inhibitor 27.
Figure 4: Final 2Fo-Fc electron density (contoured at 1 σ) of covalently bound inhibitors at S77. A) Rc inhibitor in molecule A. B) Rc inhibitor in molecule B. C) Sc inhibitor in molecule A. D) Sc inhibitor in molecule B. The Rc and Sc inhibitors are enantiopure at the chiral carbon atom of the IPG group, but they are racemic at the phosphorous atom (Rp or Sp). However, the structures of the complexes show that in both cases only one phosphorous enantiomer (Sp) is observed. It is generally assumed that the mechanism of phosphonate inhibition by serine hydrolases occurs via an in-line displacement reaction, resulting in an inversion of the configuration of the phosphorous atom. Since the priorities of the substituents at the phosphorous atom, as defined by Cahn et al, change during this nucleophilic displacement and the enzyme-inhibitor complex has the Sp conformation, the fast-reacting inhibitor must have been the Sp enantiomer 222.
The Sc and Rc inhibitors are bound in a similar way in a hydrophobic groove on the LipA surface (figure 5). As anticipated, one oxygen atom of the phosphonate group is hydrogen bonded to the two backbone amides in the oxyanion hole. No hydrogen bond interactions exist between the rest of the inhibitor and the protein. It is remarkable that in both enantiomers the oxygen atom between the IPG group and the phosphonate does not form a
DIRECTED EVOLUTION OF LIPASE A
89
hydrogen bond with H156. This hydrogen bond would be expected to be present in a true tetrahedral reaction intermediate of fatty acid ester hydrolysis 29. Nevertheless, the broader electron densities in this part of the inhibitors indicate some flexibility of the oxygen atom, suggesting such a hydrogen bond with H156 is possible part of the time. Both enantiomers have good Van-der-Waals contacts with the protein, with mainly hydrophobic residues. The solvent accessible surface of LipA decreases with very similar values upon binding of the Sc (197 Å2) or Rc inhibitor (204 Å2). The lack of clear enantioselectivity of B. subtilis LipA is understandable from the structures, since there are no specific interactions of the protein with the covalently bound transition state mimics and both enantiomers fit equally well in the hydrophobic groove.
Figure 5: Stereoview of the environment of the covalently bound Rc inhibitor (A) and Sc inhibitor (B) to LipA. Surface representation of LipA with the Rc inhibitor (C) and Sc inhibitor (D) in the hydrophobic groove of LipA.
CHAPTER 5
90
Conclusion In this study, we have described a novel method for the directed evolution of enantioselective lipases. This principle is based on the combination of phage display technology and a selection step using enantiomeric phosphonate inhibitors. The relationship between the 3D structure and the enantioselectivity towards the substrate of interest, in our case IPG, can be studied in detail, since we have successfully elucidated the crystal structure of B. subtilis LipA with covalently bound enantiomeric phosphonate inhibitors. In the past years, several methods for the directed evolution of hydrolases have been introduced, but most of them were based on high-throughput screening and not on selection 189,223-229.
For probing large libraries, selection of improved variants is a prerequisite and phage display has been proposed to select catalysts with improved or novel activities. The use of transition state analogues 230, substrates anchored to the phage 104,231-233 or suicide inhibitors 234-237 has been reported. To select variants with an inverted enantioselectivity for the production of enantiopure IPG, we employed the phage display technique with phosphonate suicide inhibitors, since phosphonates mimic the first transition state in ester hydrolysis 238-
241. We showed previously that a SIRAN immobilised phosphonate inhibitor is capable of an irreversible inactivation of B. subtilis LipA 103,217. Separate coupling of this racemic phosphonate inhibitor to both enantiomers of IPG resulted in (+)- and (-)-forms of the IPG phosphonate inhibitor (SIRAN Sc and Rc inhibitor, respectively). Thus, the selection method presented in this study is based on covalent attachment of a lipase substrate analogue to the active site of the displayed variants of B. subtilis LipA.
Here, we introduce a novel sequential incubation method using both enantiomers of this suicide inhibitor, a so-called dual selection method, to firstly counterselect phages on a phosphonate inhibitor coupled to the undesired enantiomer of IPG (SIRAN Rc inhibitor) followed by a positive selection on the phosphonate linked to the desired enantiomer (SIRAN Sc inhibitor). We have investigated the feasibility of this phage display selection by constructing a B. subtilis LipA library randomised in a region flanking the active site (amino acid 16-20 of the mature LipA). One of the residues within this region, N18, has been suggested to be of particular interest for the enantioselective properties of this enzyme 27,242,243.
After 4 rounds of selection 15 mutants were characterised in detail. Strikingly, 11 mutants with enantioselectivity ratios comparable to the wild type LipA were retrieved. DNA sequencing revealed that a strong selection for wild type amino acids was observed. Interestingly, in 1 out of the 15 mutants a TGA termination codon was observed while a fully enzymatically active enzyme could still be isolated from the periplasm of those mutants. Indeed western blot analysis revealed the presence of a fully intact lipase (data not shown). A possible explanation for this phenomenon was provided by the fact that the TGA stop codon is not fully suppressed in this strain. Correspondingly, it is known that tryptophan is inserted at the UGA (opal) stop codon in some E. coli strains, resulting in an intact protein 244,245. To test this hypothesis, we have constructed the LipA F17DF19W mutant, which contained the UGG codon (tryptophan) instead of UGA codon (stop codon). After isolation of the periplasmic fraction of this mutant, the enantioselectivity towards IPG
DIRECTED EVOLUTION OF LIPASE A
91
butyrate was determined. In fact, this mutant had an enantiomeric excess ee of +2.13% at 31% conversion.
Besides the selection of variants with wild type like activities, we could demonstrate for 2 mutants that an N18I mutation accounted for inversion of the enantioselectivity towards IPG. Interestingly, comparison of the Michaelis-Menten kinetics of the LipA N18I mutant with wild type LipA revealed that the kcat value of LipA N18I on the butyrate ester of (-)-IPG was significantly reduced, while the Km value of LipA N18I towards the (+)-IPG ester was significantly lowered. As a consequence, LipA N18I possesses an improved catalytic efficiency towards the (+)-IPG ester. Upon hydrolysis of this ester, the enantiomer of interest (+)-IPG is formed.
Comparison of the enantiomeric excess ee towards the IPG butyrate ester revealed an inversion of the enantioselectivity of LipA N18I as compared to LipA (ee value for LipA N18I = +32.9% and +35.3%; ee value for LipA = -33.3%). Although the changes in ee value are significant and encouraging, for practical applications the observed ee’s are still rather low. Several explanations can be given for the modest increase of the enantioselectivity: A) The library size spanning residues 16-20 is rather limited and mutation N18I simply is the best single mutant in this library. B) Catalytic activity does not solely depend on binding of the enzyme to the substrate, but also on completion of the catalytic cycle. C) The chiral carbon atom in IPG is relatively far from the actual site of catalysis in the enzyme. Moreover the (-)-IPG and (+)-IPG used for selection are stereochemically not very different and therefore the creation of a highly enantioselective enzyme may require more rigorous modification of the enzyme such as insertion of amino acids.
With respect to B) one may wonder whether selection for strong binding could lead to the selection of inactive enzymes with impaired dissociation of the substrate. These mutants would still pass the selection process since the phages are rescued from the SIRAN beads by cleavage of the collagenase site leaving the immobilised enzyme behind. Within the group of selected mutants there was, however, no accumulation of inactive enzymes, which implies that this is not a major hurdle in the procedure. An important prerequisite for selection on phage is the formation of correctly folded enzymes. In fact, Danielsen et al have recently shown that the lipase from T. lanuginosa could be displayed on phages as a fusion to the phage coat protein g3p, but subsequent selection of improved lipase variants on suicide inhibitors failed 235. They suggest that the complex process of folding and secretion of lipases complicates the expression in a heterologous host. In addition, the activity of T. lanuginosa lipase is enhanced in the presence of lipid-water interfaces, complicating the design of a selection protocol. LipA of B. subtilis appears to be an important exception, as this enzyme does not require a chaperone for functional phage display. Another important advantage of this lipase is the lack of a lid, enabling the characterisation of the enzyme by Michaelis-Menten kinetics. The results obtained in this study clearly demonstrate that the selected LipA mutants were fully catalytically active and therefore correctly folded.
With respect to C) an explanation is found in the 3D structure of the covalent B. subtilis lipase-inhibitor complex, which illustrates that the active site cleft of LipA can accommodate both IPG enantiomers equally well, and that both enantiomers have very similar binding modes. This explains that only a slight enantiopreference is observed. By
CHAPTER 5
92
replacing the side chain of N18 by an isoleucine side chain several rotamers of isoleucine are possible. Each of these rotamers will have a too close Van-der-Waals contact with neighbouring amino acids (L160 (~2.3 Å) and A15 (~2.8 Å) or with H10 (~2.7 Å) and F19 (~3.2 Å) depending on the rotamer conformation), since an isoleucine has an additional branch on the Cβ atom compared to asparagine. Therefore, small local rearrangements have to occur to make space for I18, but the exact outcome is hard to predict. Thus, more rigorous mutations may have to be constructed to evolve LipA of B. subtilis as a highly enantioselective enzyme for the kinetic resolution of IPG esters. Our outlook is therefore to focus on extending the amino acid region flanking the active site (amino acids 16-20 of mature LipA) using small stretches of amino acids. These stretches will be selected using the 3D structures of homologues of Bacillus LipA. Ideally, a lid-like structure will thus be formed covering the active site. Possibly, this could have a larger impact on the enantioselectivity of LipA. It is also thinkable that highly enantioselective LipA mutants can be obtained when a secondary alcohol instead of IPG (a primary alcohol) is present in the phosphonate inhibitor.
In conclusion we have explored the feasibility to use phage display for the selection of an enantiospecific lipase. On the basis of the 3D structure of Bacillus LipA with bound IPG inhibitors the effect of the mutation has been explained. For further improvements of the lipase more extensive alterations to the enzyme structure are required. Acknowledgement This project was funded by the European Commission under project number QLK3-CT-2001-00519. We thank all partners for their discussions and contributions leading to the generation of this project.
6
A NOVEL SELECTION SYSTEM FOR ENANTIOSELECTIVITY OF
BACILLUS SUBTILIS LIPASE A BASED ON BACTERIAL GROWTH
YKELIEN L. BOERSMA, MELLONEY J. DRÖGE,
ALMER M. VAN DER SLOOT, TJAARD PIJNING,
BAUKE W. DIJKSTRA & WIM J. QUAX
SUBMITTED
A GROWTH SELECTION SYSTEM FOR B. SUBTILIS LIPASE A
95
A novel selection system for enantioselectivity of
Bacillus subtilis Lipase A based on bacterial growth In directed evolution experiments, success strongly depends on the availability of screening or selection methods. Genetic selections have been rated extremely valuable for evolving enzymes with improved catalytic activity, improved stability, and altered substrate specificity. However, enantioselectivity is a difficult parameter to select for. In this study, we present a novel selection strategy which not only selects for catalytic activity, but for the first time for enantioselectivity as well. An Escherichia coli aspartate auxotroph was transformed with a mutant library of Bacillus subtilis lipase A and plated onto selective minimal medium, supplemented with an ester of aspartate coupled to the desired enantiomer of interest S-(+)-1,2-O-isopropylidene-sn-glycerol (IPG). To develop a dual selection tool, a phosphonate ester of the opposite IPG enantiomer was added to the selective minimal medium as well to inhibit growth of less enantioselective variants. Three rounds of selection were imposed, in which selection pressure was increased by raising the phosphonate concentration. Mutants with an inverted and improved enantioselectivity towards S-(+)-IPG were selected. One variant, D133AV136D, was of particular interest, as the catalytic acid had migrated to a position further along the loop connecting β7 and αE. This variant was further characterised.
Introduction Biocatalysis has emerged as a powerful tool for the industrial synthesis of pharmaceuticals and their intermediates. 246 Enzymes are capable of performing complex regioselective and/or enantioselective reactions and can accelerate reaction rates by enormous factors. High enzymatic selectivity also results in efficient reactions with few by-products, thus making enzymes an environmentally friendly alternative to conventional chemical catalysis 10,13. To date, successful applications of biocatalysts in industrial processes have been largely confined to hydrolytic enzymes such as lipases and esterases 15,16.
Most applications of enzymes do not rely on the natural reaction catalysed by them, but rather concern non-natural substrates. However, wild type enzymes often show poorly compatible substrate specificity, poor stability or insufficient (enantio-) selectivity for the cost-effective production of a particular product. In this respect, directed evolution has emerged as an important means for the improvement of nature’s catalysts to make them more suitable as industrial biocatalysts. This technique is essentially composed of two steps: first, mutagenesis of the gene(s) encoding the enzyme(s) and, second, identification of desired biocatalyst variants within these mutant libraries by screening or selection 246. Thus, the success of directed evolution experiments often depends on the choice of diversity-generation methods and the availability of screening or selection methods 247.
The identification of improved variants by high-throughput screening or selection requires high-quality substrates and assays as the selectivity obtained with a surrogate substrate can
CHAPTER 6
96
differ significantly from that towards the real substrate, leading to false-positive variants 248. For many enzymes however, such as lipases and esterases, no high-throughput screening methods are available. Especially the probing of the enantioselectivity of lipases and carboxyl esterases depends on time-consuming assays 249. A major improvement can be obtained if selection instead of screening can be introduced. The main advantage of selection over screening is that many more variants in the library can be analysed simultaneously. Selection strategies exploit conditions favouring the exclusive survival of desired variants; consequently, uninteresting variants are never seen. Thus, selecting enzyme variants is much faster and can be carried out with higher throughput: these strategies allow for the evaluation of larger libraries of mutants, even as large as 1010 11,250.
A powerful method for selection is based on catalytic potency reflected in bacterial growth. In vivo selection links cell survival to enzymatic activity. The general strategy for this genetic selection involves the introduction of a metabolic requirement for the desired activity into the host cells. Plasmids, encoding for a mutant library of the protein of interest, are introduced into a suitable host for selection, preferably a mutant strain of a well-characterised bacterium, such as Escherichia coli. Selective conditions for the target function of the protein encoded by the plasmid are imposed in such a way that only those cells expressing variants with the desired phenotype are viable (figure 1) 74,246,250.
Figure 1: General strategy for selection based on bacterial growth. A compound is taken up by the bacterium and converted by the expressed enzyme to the essential nutrient. Often, mimic substrates supplemented to the minimal medium are required to allow an in vivo selection to be designed for a specific reaction. This approach might be especially useful to generalise the use of in vivo selections to most chemical reactions and to the reactions yielding non-natural products 251. These substrates should be soluble in aqueous solutions as cell growth is generally not possible in mixtures of organic solvents with aqueous solutions. Interference of substrates and products with the cellular environment should also be avoided as much as possible. Many examples of genetic selections have been described previously 74,81,92,250,252,253.
A GROWTH SELECTION SYSTEM FOR B. SUBTILIS LIPASE A
97
Our aim was to develop a novel growth selection system in which variants were not only selected on the basis of their catalytic potency, but for the first time on their enantioselectivity as well. We chose to study lipase A (LipA) of Bacillus subtilis 168, a hydrolytic enzyme that was characterised in detail 27. LipA can be expressed in the periplasm of the auxotroph E. coli K-12 PA340/T6, a bacterium in which both pathways for aspartate synthesis have been knocked out 254. It was previously established that LipA is capable of hydrolysing acetate, butyrate and caprylate esters of IPG, a precursor in the synthesis of β-adrenoceptor antagonists, although hardly any enantioselective preference was observed (ee ranging from 20.3% to 73.2% for (-)-IPG) 115. The ultimate goal was to improve and moreover invert the enantioselectivity towards the wanted enantiomer S-(+)-IPG. Therefore, we mutated residues 132 to 136; D133, one of the amino acids of the catalytic triad, is within this region and therefore this stretch of amino acids might be of particular interest for the catalytic activity and enantioselective properties of the enzyme. We transformed the mutant library to the E. coli aspartate knock-out and plated it onto selective minimal medium plates, supplemented with an ester of enantiopure IPG coupled to aspartate (scheme 1). The expressed LipA variant had to hydrolyse this aspartate ester to release aspartate for uptake and subsequent bacterial growth. To avoid growth of bacteria expressing less enantioselective variants, a dual selection step was introduced, in which a phosphonate ester of the undesired IPG enantiomer was added to the minimal medium. These phosphonates bind covalently to the active site serine of the lipase, thus mimicking the transition state in ester hydrolysis. By using the crystal structure of wild type LipA, models were made to provide insight in the structural rearrangements explaining the altered enantioselectivity of LipA mutants.
O
OOH
O
OOH
O
OO
O
OH
O
NH2
O
OO
O
OH
O
NH2
O
O
P
O
OO
NO2
O
O
P
O
OO
NO2
A) B)
C)
S-(+)-IPG R-(-)-IPG Aspartate ester of S-(+)-IPG= R-(+)-IPG aspartate
Asparate ester of R-(-)-IPG= S-(-)-IPG butyrate
((S)-2,2-dimethyl-1,3-dioxan-4-yl)methyl 4-nitrophenyl butylphosphonate= phosphonate inhibitor of S-(+)-IPG = Sc inhibitor
((R)-2,2-dimethyl-1,3-dioxan-4-yl)methyl 4-nitrophenyl butylphosphonate= phosphonate inhibitor of R-(-)-IPG = Rc inhibitor
Scheme 1: A) Chemical structure of 1,2-O-isopropylidene-sn-glycerol (IPG). B) Chemical structure of the aspartate ester of IPG. C) Chemical structure of the soluble lipase inhibitors. Note that the inhibitor is enantiopure at the chiral atom of the IPG molecule (Sc or Rc inhibitor), and racemic at the phosphorous atom (mixture of Rp and Sp inhibitor). The absolute configuration of IPG changes upon attachment to the pshophorous atom or upon ester bond formation.
CHAPTER 6
98
Experimental procedures Plasmids, bacterial strains and media E. coli K-12 PA340/T6 (thr-1, leuB6(Am), fhuA2, lacY1, glnV44(AS), gal-6, λ-, gdhA1, hisG1(Fs), rfbD1, galP63, ∆(gltB-gltF)500, rpsL9, malT1(λR), xylA7, mtlA2, ∆argH1, thi-1) was kindly provided by the E. coli Genetic Stock Center (Yale University, New Haven, USA). pCANTAB 5E was purchased from Pharmacia (Amersham Pharmacia Biotech, Uppsala, Sweden). 2xTY medium contained: Bactotrypton (1.6% w/v), Bacto yeast extract (1% w/v) and sodium chloride (0.5% w/v). As antibiotic agents ampicillin (100 µg.mL-1) and streptomycin (100 µg.mL-1) (Duchefa Biochemie, Haarlem, The Netherlands) were used. M9 minimal medium contained Na2HPO4.7H2O (4 g. L-1), KH2PO4 (15 g.L-1) and sodium chloride (2.5 g.L-1). Chemicals Both enantiomers of the aspartate esters of IPG as well as both enantiomers of the butyl phosphonate esters of IPG were synthesised by Syncom BV (Groningen, The Netherlands). Butyrate esters of both enantiomers of 1,2-O-ispropylidene-sn-glycerol (IPG) were kindly provided by Prof. M.T. Reetz (Max-Planck Institut für Kohlenforschung, Mülheim, Germany). p-Nitrophenyl caprylate was purchased from Sigma Chem. Co. (Axel, The Netherlands). Supplemental amino acids (Thr, Arg, Leu, His, Asp), thiamine, MgSO4 and CaCl2 were purchased from Sigma-Aldrich (Steinheim, Germany). Construction of the mutant library The LipA encoding gene was cloned in the phagemid pCANTAB 5E, downstream of a modified g3p signal sequence, as described previously 103. To introduce an XbaI restriction site in the lipA gene, a silent mutation at base pair position 423 of the lipA gene was introduced using primers LipXbaIFor (Life Technologies, UK): 5’ – TTACTTATCTAGATTAGATGGTGCTGA – 3’ and LipXbaIRev: 5’ – CTAGCACCATCTAATCTAGATAAGTAA – 3’. The XbaI restriction site is indicated in bold italics. To construct the mutant library on the region of amino acid 132 to 136, the following primers were used: CanHindFor (Life Technologies, UK) 5’ – CCATGATTACGCCAAGCTTTGGAGCC – 3’ (HindIII restriction site in bold italics) and 00B470Rev (Eurogentec, Groningen, The Netherlands) 5’ – TAATCTAGATAAGTAATTCAT867665765657887ACTGCTGTAAATGGA – 3’ (XbaI restriction site in bold italics; 5 = 80% T; 6 = 80% A; 7 = 80% C; 8 = 80% G; the remaining 20% is an equal mixture of the other three bases). Recombinant DNA procedures were carried out as described by Sambrook et al 176. Plasmid DNA was prepared using the Qiaspin miniprep kit (Qiagen, Hilden, Germany) and DNA purification was performed using the Qiaquick gel extraction kit (Qiagen, Hilden, Germany). The library was amplified from chromosomal DNA of B. subtilis 168 using Pfu polymerase (Stratagene, La Jolla, CA, USA). The gene fragment was digested with HindIII and XbaI (New England Biolabs, Ipswich, MA, USA), and cloned in E. coli TG-1 into the HindIII and XbaI sites of the plasmid pCANTABLip-CH 103. The obtained mutant library was sequenced to assess the mutation ratio. The LipA variant D133AV136A was constructed using Quikchange® PCR (Stratagene, La Jolla, USA) with the following primers (Operon, Cologne, Germany): LipDAVAFor
A GROWTH SELECTION SYSTEM FOR B. SUBTILIS LIPASE A
99
5’ – CATTTACAGCAGTGCCGCAATGATTGCCATGAATTACTTATCCAGA – 3’ and LipDAVARev 5’ – TCTGGATAAGTAATTCATGGCAATCATTGCGGCACTGCTGTAAATG -3’. The obtained constructs were sequenced to verify the base pair order. Selection on selective minimal medium E. coli K-12 PA340/T6 cells were made chemically competent and transformed with 50 ng of mutant plasmid DNA 176. They were starved by incubation in 0.9% w/v NaCl for 2 h at 37°C and plated onto selective M9 minimal medium agar (1.6% w/v) plates. The medium was supplemented with MgSO4 (1 mM), CaCl2 (0.1 mM), essential amino acids (Thr, His, Arg, Leu; 20 µg.mL-1) and thiamine (1 µg.mL-1). As a sole aspartate source, an aspartate ester of R-(-)-IPG or S-(+)-IPG (1.5 mM) was added to the agar plates. To select for enantioselective variants, a butyl phosphonate ester of the undesired IPG enantiomer (1.5 mM) was added to the medium, to eliminate non-selective variants or variants selective for the opposite enantiomer. Approximately 2500 viable E. coli K-12 PA340/T6 containing the mutant plasmids were plated onto each plate. As control plates were used: LB, aspartate (20 µg.mL-1; Sigma) and Min (no aspartate or esters present). Plates were incubated at 30ºC. Upon appearance colonies were picked, the periplasmic fraction was isolated and used in the 1,2-O-isopropylidene-sn-glycol ester assay as described below. Remaining colonies were harvested and plated again onto selective minimal medium with an increased concentration of phosphonate ester (3 mM) for a second round of selection. For a third round of selection, colonies were plated onto selective minimal medium with a concentration of 7.5 mM of phosphonate ester, selected and the enantioselectivity was determined using the 1,2-O-isopropylidene-sn-glycerol ester assay. From bacteria expressing a lipase variant with an inverted enantioselectivity the plasmid DNA was isolated and sequenced to examine the mutation(s). Isolation of the periplasmic fraction E. coli K-12 PA340/T6 was grown in 50 mL tubes containing 10 mL 2xTY medium, ampicillin and isopropyl-β-D-galactopyranoside (IPTG, 1 mM). The tubes were incubated at 37ºC at 250 rpm for 16 h. The OD600 was recorded and the cells were harvested and resuspended in Tris HCl buffer (10 mM, pH 7.4). After centrifugation, the cells were resuspended in 200 µL buffer containing Tris HCl (10 mM, pH 8.0), sucrose (25% w/v), EDTA (2 mM) and lysozyme (0.5 mg.mL-1, Sigma-Aldrich, Steinheim, Germany). After incubation on ice for 20 min, 50 µL buffer containing Tris HCl (10 mM, pH 8.0), sucrose (20% w/v) and MgCl2 (125 mM) was added. The suspension was centrifuged and the supernatant, containing the periplasmic fraction, was isolated and used as enzyme solution in the IPG ester assay. The protein content of this fraction was determined by performing a Bradford assay in triplicate using bovine serum albumin (BSA) as a standard (Pierce, Rockford, Illinois, USA). 1,2-O-isopropylidene-sn-glycerol ester assay Periplasmic fractions were diluted with MOPS buffer (0.07 M, pH 7.5), containing BSA (0.2% w/v), to a final volume of 100 µL. References were diluted correspondingly but contained no enzyme solution. The esters of IPG (1 mM) were dissolved in 10 mL MOPS
CHAPTER 6
100
buffer (0.07 M, pH 7.5), containing Tween 80 (14.3% w/v), and diluted to 50 mL with MOPS buffer (0.07 M, pH 7.5). 500 µL substrate solution was added to the enzyme solution and the final mixture was incubated in a water bath at 32°C. After incubation, 400 µL of a saturated NaCl solution and 10 µL internal standard solution (racemic 3-hexene-1-ol, 5 mg.mL-1 assay buffer) were added to the sample solution and the aqueous solution was extracted twice with 1 mL ethyl acetate. GC analysis was performed as described by Dröge et al 180. One unit (U) is defined as the amount of enzyme that hydrolyses 1 µmol IPG ester per minute. Enantiomeric excesses, ee, were calculated according to Chen et al and were defined as the ability of the enzyme to distinguish between enantiomers 181. All data were expressed as mean ± SEM. The statistical significance of differences was tested at a significance level of p < 0.05 using a two-tailed Student's t-test. Modelling of variant D133AV136D In variant D133AV136D, the position of the catalytic acid was shifted towards the end of the loop connecting β7 and αE. A model of the selected variant was constructed with Swiss-PDBViewer version 3.7 255 using wild type LipA (PDB code 1I6W) while mutating positions 133 and 136. To examine whether the repositioning of the acid was a unique feature within the family of α/β hydrolases or this same topology was shared, a one-2-all fit was made using TOPOFIT 256. This method analyses the similarity in protein structure by making use of 3D Delaunay triangulation patterns derived from backbone representation 257. Results Construction of the mutant library The LipA encoding gene (Genbank accession number M74010) was cloned in phagemid pCANTAB 5E as described previously 103. An XbaI site was introduced at base pair 423 by a silent mutation. Cassette mutagenesis was applied to construct a saturated mutant library directed towards amino acids 132 to 136 of mature LipA. These oligonucleotides resembled the lipA sequence for approximately 80%. In theory, this would yield a mutant library composed mostly of single and double mutations. Sequence analysis of 13 clones revealed that 15% of the mutants had no amino acid mutation while the percentage of single and double mutated amino acids was for both 31% each. Two clones showed deletions. After transformation of E. coli TG-1, a mutant library consisting of 3.5 × 104 colonies was obtained. This should be more than sufficient to saturate all possible single mutations at the amino acid level. This library was used for the transformation of the aspartate auxotroph E. coli K-12 PA340/T6. Dual selection In order to develop a growth selection system based on bacterial growth for LipA variants with improved enantioselectivity, aspartate esters of enantiopure IPG were synthesised. To examine the functional character of the mimic substrate IPG aspartate, the hydrolysis of the mimic substrate by purified WT LipA was determined in the IPG assay and compared with the hydrolysis of the real substrate IPG butyrate. The enantiomeric excess ee of the hydrolysis of IPG butyrate was 12.9% ± 1.6 towards the (-) enantiomer. The enantiomeric
A GROWTH SELECTION SYSTEM FOR B. SUBTILIS LIPASE A
101
excess ee of the hydrolysis of IPG aspartate was 5.5% ± 1.1 towards the (-) enantiomer. No statistically significant difference was observed in catalytic activity towards the substrates (p < 0.001).
To introduce a dual selection step, soluble IPG phosphonate inhibitors showing similarity to the aspartate esters were synthesised. This type of inhibitor was previously used in the selection of phagebound lipase, then coupled to SIRAN beads 103,115. The soluble inhibitors contained the enantiopure substrate analogue IPG, a leaving group, p-nitrophenol and a buturyl side chain connected to the racemic phosphonate. To assess the inhibitory effect of the IPG phosphonate esters, an inhibition assay was performed using purified wild type LipA according to 103. The t50 of inhibition was 10 minutes for both enantiomers. Furthermore, cells expressing wild type LipA were plated on minimal medium supplemented with the IPG phosphonate ester and incubated at 30 degrees. LB agar plates, taken as a control, showed approximately 2500 colonies after overnight incubation, while on the minimal medium plates supplemented only with the phosphonate a few small colonies appeared only after 5 days of incubation. These results indicate that the used phosphonate inhibitor limits rather than abolishes growth. Selection on selective minimal medium E. coli K-12 PA340/T6 competent cells were transformed with the mutant library and, as a control, with wild type LipA. A negative control (untransformed E. coli) was taken into account as well. All transformation mixtures were plated on selective minimal medium, for which M9 was prepared without ammonium chloride to exclude a nitrogen source for the auxotroph, thereby limiting growth. As a positive control for growth, all transformation mixtures were plated on LB agar plates; approximately 2500 colonies appeared after overnight incubation. On all selective minimal medium plates supplemented with either aspartate or the aspartate ester, approximately 1500 colonies appeared after two days. On plates supplemented with the phosphonate inhibitor in addition to the aspartate ester, colonies also appeared after two days, though their number was reduced to approximately 1000 colonies. Negative control plates showed small colonies after 5 to 10 days of incubation. After the second round of selection, approximately 750 colonies would appear on plates supplemented with both the inhibitor and the aspartate ester, while after the third round approximately 400 colonies were found.
After two days of growth, 50 colonies were randomly chosen from the plates supplemented with both aspartate ester and the phosphonate inhibitor to assess their activity and enantioselectivity compared to wild type LipA. As (+)-IPG is the enantiomer of interest, colonies selected from plates supplemented with the R-(+)-IPG aspartate ester and the S-(-)-IPG phosphonate inhibitor were of greater interest. The remaining colonies were harvested and plated onto new selective plates. The periplasm of the selected colonies was isolated and analysed using the IPG assay. Only the best LipA variants were sequenced to verify the base pair order. The catalytic activities, enantioselectivities and the mutations are shown in table I.
Most selected colonies from the S-(-)-IPG aspartate ester had activities and enantioselectivities comparable to those of wild type LipA (conversion approximately 20%, ee approximately 25% towards R-(-)-IPG). However, mutants selected from plates supplemented with the R-(+)-IPG aspartate ester showed an inverted and improved
CHAPTER 6
102
Table I: Enantioselective hydrolysis of racemic esters of IPG butyrate by periplasmic fractions of selected LipA variants (n=3). R = R-(+)-IPG aspartate; S = S-(-)-IPG aspartate.
3rd round R8 28.9 ± 0.5 +73.1 ± 0.7 (+)-IPG D133A, V136D enantioselectivity. The catalytic activity remained comparable to that of the wild type. The enantioselectivity increased in subsequent rounds of selection with higher selection pressure, from an ee value of 34% in the first round of selection, to an ee value of 73% in the third round of selection. Sequence analysis revealed that mutations were distributed throughout the whole region of amino acids 132 to 136, where mutation of position M134 was most frequently found. However, a mechanistic explanation for this inversion and increase in enantioselectivity is hard to find, since the active site cleft of B. subtilis LipA is relatively open and accommodates both IPG enantiomers equally well, and both enantiomers have equal binding modes 27. Thus, small local rearrangements will most probably take place, the exact outcome of which is hard to predict.
Interestingly enough, position D133 was prone to mutations as well. This aspartate residue is part of the catalytic triad of the enzyme, and therefore, special attention was paid to mutations at this position. The mutant found in the third round of selection, D133AV136D, was subjected to further analysis. Characterisation of the D133AV136D mutant After three rounds of selection, mutant D133AV136D was obtained. This mutant showed an inverted and improved enantioselectivity compared to wild type LipA. Its specific activity was comparable to the wild type (conversion of 23.4% for wild type LipA versus 28.9% for mutant D133AV136D). The most interesting feature of this mutant lies in the fact that the D133 residue of the catalytic triad is replaced by the catalytically inactive alanine residue. As the 3D structure of LipA has been solved 27, the crystal structure was used to examine this mutation in more detail. It was speculated that the mutated aspartate residue at position 136 might take over the role of D133 as the catalytic acid (figure 2).
A GROWTH SELECTION SYSTEM FOR B. SUBTILIS LIPASE A
103
Figure 2: A) Catalytic site of LipA wild type, with residues S77, D133 and H156. B) Catalytic site of selected variant D133AV136D, in which the acid is repositioned. During substrate hydrolysis, the hydroxyl group of the serine at position 77 is activated by transferring its proton to the H156 side chain, which becomes positively charged. The O- ion of S77 performs a nucleophilic attack on the substrate, which results in the formation of a negatively charged tetrahedral intermediate. This intermediate is stabilised by two neighbouring peptide NH-groups. The proton transfer from the S77 hydroxyl group to H156 is facilitated by D133, which properly orients the imidazole ring of H156 and ensures that H156 is in the right tautomeric form to accept a proton from S77. In addition, the positive charge on the histidine residue is stabilised by the negatively charged carboxyl moiety of D133 (figure 2A) 25,27,29. The ester bond is cleaved and, upon protonation by H156 the alcohol moiety of the substrate is released, but the acid part of the substrate remains covalently bound to the enzyme (the acyl-enzyme). In the second step of the reaction the acyl enzyme is hydrolysed by a water molecule which is activated by the uncharged H156. Finally, H156 donates the proton, acquired from the hydrolytic water molecule, to S77 and the enzyme is ready for a new round of catalysis. In the case of mutant D133AV136D however, the alanine at position 133 is unable to stabilise the positive charge of H156. Since the double mutant does show appreciable activity, it is likely that, the aspartate residue introduced at position 136 takes over the role of D133. Modelling shows that indeed an aspartate residue at position 136 can interact with H156 in such a way that the H156 side chain can also still interact with the S77 side chain (figure 2B).
To test the hypothesis that D136 takes over the role of D133 the double mutant D133AV136A was constructed. D133AV136A was expressed in the periplasm of E. coli HB2151, which was then isolated and used in the IPG assay. Compared to the blanks, there was no conversion of IPG butyrate by the variant. Thus, in the D133AV136D double mutant D136 is essential for activity, demonstrating that D136 can take over the role of D133.
Mutant D133AV136D shows an inverted and improved enantioselectivity compared to wild type LipA. Since the active site cleft of B. subtilis LipA is relatively open and accommodates both IPG enantiomers equally well, it is difficult to explain the inverted and
CHAPTER 6
104
improved enantioselectivity. Most likely, a somewhat altered orientation of H156 as a result of its interaction with D136 instead of D133 causes subtle differences in the interaction between the enzyme and the substrate, which may lead to the observed differences in enantioselectivity. Discussion In the past, genetic selections such as presented in this study have had their impact on the evolution of enzymes with improved catalytic activity, improved stability, and altered substrate specificity. However, enantioselectivity is a difficult property to select for, though some indirect screening methods based on bacterial growth have been described 93,258. In these methods, the enantiopure substrate of interest was coupled to a compound, which, upon release, was toxic to the bacterium. Individual variants could be identified by measuring growth rates of cells in liquid media as cells showing hydrolytic activity were unable to grow due to released toxic compound. A high-throughput screening method can thus be established for enhanced enantioselectivity of hydrolytic enzymes as these methods are based on growth rate. However, screening involves the individual examination of variants, and is therefore cumbersome. By applying a selection strategy, multiple variants can be examined at the same time. Here, we have reported a novel genetic selection system which selects for enantioselectivity in addition to catalytic activity.
An aspartate auxotrophic E. coli strain was used for selection based on bacterial growth. The bacterium was transformed with a library directed towards a small loop region around the active site of LipA and grown on selective minimal medium agar plates. By supplementing an aspartate ester of enantiopure IPG to the minimal medium, lipase variants expressed in the periplasm could be selected not only for catalytic activity, but also for enantioselectivity. The enantioselectivity of wild type LipA towards this mimic substrate IPG aspartate and towards its real substrate IPG butyrate did not differ in a statistically significant manner, implying that the mimic compound was a suitable substrate in our selection system. After each round of selection, colonies were picked, the periplasmic fraction was isolated and assayed, and the more enantioselective variants were sequenced in order to verify the base pair order. Ultimately, three rounds of selection were performed, in which the imposed selection pressure was increased by raising the concentration of the phosphonate inhibitor. After the third round, a further improvement of the enantioselectivity was not observed. Most likely, this can be explained by the small library size, which is only comprised of five randomised amino acids. Upon using bigger libraries, it would be expedient to apply DNA shuffling 60 or CASTing 269 in order to select variants with even higher enantioselectivities.
A dual selection strategy using phosphonate suicide substrates was introduced here to decrease the total number of colonies and reduce the frequency of selecting less enantioselective variants. This inhibitor reduces rather than completely inhibits growth, which might be due to the poor enantioselectivity of the wild type LipA towards the substrate of interest, IPG. The ultimate goal was to select only the more enantioselective variants. By using the phosphonate suicide inhibitor, one could argue that a discrepancy between kcat and Km is introduced: the expressed variant could show a high value of kcat towards the phosphonate while the Km value towards the aspartate ester can be equally low.
A GROWTH SELECTION SYSTEM FOR B. SUBTILIS LIPASE A
105
Thus, due to binding to the phosphonate inhibitor, interesting variants could be missed. Nevertheless, in our case the selection strategy yielded lipase variants with an inverted and improved enantioselectivity.
In the past, genetic selections have been useful in the elucidation of structure-function relationships in enzymes. Hence, important and often overlooked roles played by multiple subtle interactions between active-site residues could be examined 88. In our work, variants were selected from a library around the catalytic acid D133. It was tempting to speculate on the mechanism of action of variants with a mutated D133 residue. As demonstrated by the sequences of the selected variants, the aspartate residue was not conserved but prone to mutation as well. Similar to other α/β hydrolase fold enzymes, the catalytic triad of B. subtilis LipA has a preserved arrangement. The topological positions of the nucleophile and the general base histidine after β-strand 5 and 8, respectively, are fully conserved within this family, while the location of the catalytic acid is not 259. In most α/β hydrolase family members, the catalytic acid residue aspartate or glutamate is located following strand β7 (figure 3A); in mutant D133AV136D, the catalytic acid has slightly migrated to a position near the end of the loop connecting β7 and αE (figure 3B).
Figure 3: A) Canonical α/β hydrolase fold, with the nucleophile after β5, the acid after β7, and the histidine residue after β8. B) Topology of variant D133AV136D, in which the acid has shifted towards the end of the loop connecting β7 and αE. C) Topology of HPL, which contains a catalytic acid after β6. An aspartate residue after β7 is still present, though it is not active. To investigate the existence of other α/β hydrolase fold enzymes with a repositioned catalytic acid similar to mutant D133AV136D, a one-2-all fit with D133AV136D as a search model was made 256. No other enzymes with a similar topology as in D133AV136D were found. In contrast, the enzymatically active human pancreatic lipase shows an equivalent catalytic acid which has been repositioned in the opposite direction compared to the acid in variant D133AV136D to a position following strand β6 259,260 (figure 3C).
CHAPTER 6
106
Double mutants of Geotrichum candidum lipase 261 and of haloalkane dehalogenase from Xanthobacter auxotrophicus 262, both constructed to resemble the topology of human pancreatic lipase, were found to be enzymatically active as well.
Mutant D133AV136D showed an inversion and improvement of the enantioselectivity as well. From the 3D structure, it is clear that the active site cleft of B. subtilis LipA is relatively open. Thus, both IPG enantiomers are accommodated equally well; it is therefore difficult to explain the inverted and improved enantioselectivity. The side chain of D133 is replaced with a shorter alanine side chain, thereby creating space for the side chain of D136. Small local rearrangements will most probably take place due to a different orientation of H156; this is likely to affect the interaction with the substrate. However, the exact effect on the binding situation of the desired IPG enantiomer remains hard to predict.
The main prerequisite of a growth-based selection system such as presented here is that the mutants must be expressed in the periplasm of the E. coli auxotroph in order to convert the mimic substrate, and, theoretically, this might limit a broad application of this method. We have, however, recently demonstrated that even an intracellular enzyme can be translocated to the periplasm solely by inserting a specific signal sequence. The intracellular carboxylesterase A (CesA) from B. subtilis 168 was succesfullty translocated to the periplasm using a Sec-dependent signal sequence 108 and preliminary results show that wild type CesA transformed to the E. coli aspartate auxotroph and plated on selective minimal medium plates, displayed more growth on plates supplemented with S-(-)-IPG aspartate. Upon hydrolysis, the absolute conformation of IPG changes to R-(-)-IPG. The enantiopreference of CesA towards S-(-)-IPG is therefore in accordance with its enantioselectivity towards the real substrate IPG butyrate 45, which is directed towards R-(-)-IPG. The developed selection system can therefore in the future be applied in the selection of CesA mutants. In summary, the presented growth selection system can be generally applied in the selection of hydrolase variants, as long as they are translocated to the periplasmic space. Conclusion In this study, a novel bacterial growth system has been developed that is suitable for the enantioselective selection of hydrolytic enzymes. The applicability of this strategy was demonstrated by the selection of lipase variants with an inverted and improved enantioselectivity. Since this system can be applied for the enantioselective selection of other hydrolase variants as well, we believe we have established a strategy that is generally applicable and will provide new perspectives in the evolution of enzyme enantioselectivity. Acknowledgement This project was funded by the European Commission under project number QLK3-CT-2001-00519. We thank all partners for their discussions and contributions leading to the generation of this project. The authors thank Ms M. Berlyn of the E. coli Genetic Stock Center for providing the auxotrophic strain. Y.L.B. would like to thank Dr. R.H. Cool for fruitful discussions.
7
RATIONAL DESIGN OF BACILLUS SUBTILIS LIPASE A LOOP HYBRIDS:
INSERTION OF A LID STRUCTURE INVERTS ENANTIOSELECTIVITY
YKELIEN L. BOERSMA, MELLONEY J. DRÖGE,
TJAARD PIJNING, MARGRIET S. BOSMA, REMKO T. WINTER,
GERTIE VAN POUDEROYEN, BAUKE W. DIJKSTRA & WIM J. QUAX
RATIONAL DESIGN OF LIPASE A LOOP HYBRIDS
109
Rational design of Bacillus subtilis Lipase A loop hybrids:
insertion of a lid structure inverts enantioselectivity Lipases are successfully applied in enantioselective biocatalysis. All lipases show an α/β hydrolase fold as well as a nucleophile elbow where the catalytic serine is located. Most lipases show a lid structure controlling access to the active site, however, the Bacillus subtilis lipase A (LipA) is an exception. Its active site is solvent exposed, which might explain its modest enantiopreference in the kinetic resolution of 1,2-O-isopropylidene-sn-glycerol (IPG) esters. Thus, our aim was to improve and invert its enantioselectivity by inserting longer loops. These longer loops, originating from the homologous enzymes cutinase (Fusarium solani) and acetylxylan esterase (Penicillium purpurogenum), replaced a loop near the active site, thereby narrowing the active site cleft. The resulting loop hybrids showed an activity comparable to the wild-type enzyme; however the enantioselectivity was inverted towards the desired enantiomer of IPG. The enantioselectivity of the cutinase variant was further improved by directed evolution.
Introduction The introduction of enantiopure active substances is enforced through stricter regulations of the US Food and Drug Administration (FDA). Targeted synthesis of one enantiomer is demanded 4,263. Selective catalysis is therefore becoming a requirement for the chemical industry. The application of enzymes offers high selectivity and efficient reactions with few by-products 13,247. Lipases are examples of successful application of enzymes in enantioselective biocatalysis. They are active in both aqueous and low-water media, and various (synthetic) molecules can serve as their substrates. The most common application of lipases is the preparation of chiral building blocks, especially by kinetic resolution of racemic mixtures 264. During the last decade, 3D structures of 26 lipases have been elucidated. These structures show an α/β hydrolase fold as well as a nucleophile elbow where the catalytic serine is located 25,264. Most lipases show a lid structure controlling access to the active site 265. The interaction of the enzyme with insoluble substrates induces the displacement of the lid, making the active site available for the substrate. In the presence of lipid aggregates, the catalytic activity will increase. This phenomenon is known as interfacial activation. The sequence of the lid structure can affect both specificity and enantioselectivity of lipases 22,264,266.
The Bacillus subtilis 168 Lipase A (LipA) is of particular interest, as it does not contain a lid domain; as a consequence, the active site is solvent-exposed. Furthermore, this enzyme represents one of the few examples of a lipase that does not show interfacial activation. Like other lipases, the fold of this enzyme resembles the core of enzymes of the α/β hydrolase fold. This class of enzymes has a catalytic machinery composed by a nucleophile, an acid and a histidine; the nucleophilic serine is invariably located at the center of an extremely sharp turn between a β-strand and an α-helix. The occurrence of an oxyanion
CHAPTER 7
110
hole is shared by all members of the family as well 267. Due to its small size and relatively low molecular mass (181 amino acid residues, 19 kDa), the B. subtilis LipA can be regarded as a minimal α/β hydrolase fold enzyme 27,35,36,115,172.
Other small α/β hydrolase folded enzymes without lids include cutinase from Fusarium solani (abbreviated as 1CEX) 268,269 and acetylxylan esterase from Penicillium purpurogenum (abbreviated as 2AXE) 270. The 3D structures of these enzymes have been elucidated and appear similar to LipA: five β-strands are common in all three structures, as well as four α-helices. LipA is the smallest of these three enzymes. 1CEX and 2AXE show an accessible active site as well, with an intact preformed oxyanion hole which stabilises the negatively charged reaction intermediates. However, compared to the active sites of 1CEX and 2AXE, the cleft of LipA is shallower and wider: the loops lining the cleft are shorter and do not extend as much from the core of the protein 27. Moreover, 1CEX (197 amino acid residues) and 2AXE (207 amino acid residues) have generally longer lid-like loops lining the active site cleft.
In this study, our aim was to investigate the effect both on the activity and the enantioselectivity of the introduction of loops from 1CEX and 2AXE in LipA, thereby extending a loop lining the active site with three amino acids and one amino acid, respectively. This would lead to an artificial lid-like structure, narrowing the cleft. Contrary to a previous study 265, this exchanged loop from LipA is a loop near the active site, comprised of amino acid residues 11-20. It was shown previously that this loop is of key importance to the enantioselectivity of the enzyme 115. To insert the lid-like structures, a structural alignment of LipA, and 1CEX and 2AXE was made. Stretches of amino acids from loops near the active sites of 1CEX and 2AXE were chosen as these would most likely not disturb the secondary structure of LipA. The enantioselectivity of the 1CEX loop mutant was further improved by directed evolution on hotspots within the loop and the lipase parent molecule. Experimental procedures Computational design of the loop mutants A DALI search 271 with the B. subtilis LipA as a search model identified the two closely related α/β hydrolase fold enzymes without lid-like domains in the Protein Data Bank: A) cutinase from Fusarium solani (PDB code: 1CEX) 268 and B) acetylxylan esterase from Penicillium purpurogenum (PDB code: 1BS9) 270. It was our goal to replace the loop of residues 11-20 with the structurally corresponding (longer) loops of either cutinase or acetylxylan esterase. Therefore, structural alignments with these enzymes were made (figure 1). The following proposals for loop exchange were made, based on a "rigid" exchange, i.e. the new loops were pasted into LipA as they are observed in their respective crystal structure.
For 1CEX, the alignment showed that the replacing loop was three residues longer than the original LipA loop, and comprised residues 41-53 (figure 2). Although the backbone of the 1CEX loop would nowhere clash with surrounding LipA residues, some of the side chains might cause steric problems. In particular, T50 and L51 (1CEX) would be too close to L160 (LipA).
RATIONAL DESIGN OF LIPASE A LOOP HYBRIDS
111
The structural alignment of 2AXE with LipA showed that the replacing loop was two residues longer than the original LipA loop, it comprised residues 12-23 of 2AXE (figure 3). Consequently, a conformational rearrangement would be needed for the hybrid structure to fold properly. Plasmids, bacterial strains and media E. coli HB2151 (K12 ∆(lac-pro), ara, nalr, thi/F’, proAB, laqIq, lacZ∆-M15) and pCANTAB 5E were purchased from Pharmacia (Amersham Pharmacia Biotech, Uppsala, Sweden). E. coli K-12 PA340/T6 (thr-1, leuB6(Am), fhuA2, lacY1, glnV44(AS), gal-6, λ-, gdhA1, hisG1(Fs), rfbD1, galP63, ∆(gltB-gltF)500, rpsL9, malT1(λR), xylA7, mtlA2, ∆argH1, thi-1) was kindly provided by the E. coli Genetic Stock Center (Yale University, New Haven, USA). 2xTY medium contained: Bactotrypton (1.6% w/v), Bacto yeast extract (1% w/v) and sodium chloride (0.5% w/v). As antibiotic agents ampicillin (100 µg.mL-1) and streptomycin (100 µg.mL-1) (Duchefa Biochemie, Haarlem, The Netherlands) were used. M9 minimal medium contained Na2HPO4.7H2O (4 g. L-1), KH2PO4 (15 g.L-1) and sodium chloride (2.5 g.L-1), Chemicals Both enantiomers of aspartate esters of IPG as well as both enantiomers of butylphosphonate esters of IPG were synthesised by Syncom BV (Groningen, The Netherlands). Butyrate esters of both enantiomers of 1,2-O-ispropylidene-sn-glycerol (IPG) were kindly provided by Prof. M.T. Reetz (Max-Planck Institut für Kohlenforschung, Mülheim, Germany). p-Nitrophenyl caprylate was purchased from Sigma Chem. Co. (Axel, The Netherlands). Supplemental amino acids (Thr, Arg, Leu, His (10 mg.L-1)), thiamine, MgSO4 and CaCl2 were purchased from Sigma-Aldrich (Steinheim, Germany). Construction of the loop mutants The LipA encoding gene (Genbank accession number M74010) was cloned in the phagemid pCANTAB 5E, downstream of a modified g3p signal sequence 103. To introduce the 1CEX lid structure, an overlapping PCR strategy was used. All primers used in the reactions were purchased at Invitrogen (Groningen, The Netherlands) (for sequences, see table I). In the first step, the lid structure was constructed using primers 1CEXfor in combination with LipAlooprev, and 1CEXrev in combination with LipAloopfor. The AvaI restriction site, indicated in bold italics, was inserted to make restriction analysis possible. The two overlapping PCR ‘half fragments’ were purified and used as template in a second PCR step with the flanking primers LipAloopfor and LipAlooprev.
To construct the 2AXE mutant, the same PCR strategy was followed using primers 2AXEfor in combination with LipAlooprev, and 2AXErev in combination with LipAloopfor. The BamHI restriction site, indicated in bold italics, was inserted to make restriction analysis possible. In a second PCR step, the obtained overlapping PCR fragments were combined using LipAloopfor and LipAlooprev primers.
Mutant libraries on position 51 in the 1CEX loop were constructed using primer 1CEXN51BamHIFor in combination with LipAlooprev and primer 1CEXN51BamHIRev in combination with LipAloopfor. The BamHI restriction site, indicated in bold italics, was
CHAPTER 7
112
Table I: Primer sequences Restriction sites are in bold italics, L51G (mutation to GGG) and L51Q (mutation to CAA), as well as L160A (mutation to GCG) and L160V (mutation to GTG) are underlined.
L160AlaFor GTTGGACACATCGGCCTTGCGTACAGCAGCCAAGTA L160AlaRev TACTTGGCTGCTGTACGCAAGGCCGATGTGTCCAAC L160ValFor GTTGGACACATCGGCCTTGTGTACAGCAGCCAAGTA L160ValRev TACTTGGCTGCTGTACACAAGGCCGATGTGTCCAAC inserted to make restriction analysis possible. In a second PCR step, the obtained overlapping PCR fragments were combined using LipAloopfor and LipAlooprev primers. Mutant libraries on position 50 in the 1CEX loop with either a glutamine (CAA) or a glycine (GGG) at position L51 (underlined) were constructed using primers 1CEX5051XBamFor in combination with LipAlooprev and 1CEX5051XBamRev in combination with LipAloopfor. The BamHI restriction site, indicated in bold italics, was put in to make restriction analysis possible. In a second PCR step, the obtained overlapping PCR fragments were combined using LipAloopfor and LipAlooprev primers. Recombinant DNA procedures were carried out as described by Sambrook et al 176. Plasmid DNA was prepared using the Qiaprep Spin Miniprep kit (Qiagen, Hilden, Germany). DNA purification was performed using the Qiaquick Gel Extraction kit (Qiagen, Hilden, Germany). The gene fragment was digested with AgeI and HindIII (New England Biolabs), and cloned in E. coli HB2151 into the AgeI and HindIII sites of the phagemid pCANTABLip-CH 103. The constructs were sequenced to verify the correct base pair order.
To construct the L160A and L160V variants, QuikChange® PCR (Stratagene, La Jolla, CA, USA) was applied. The obtained constructs were sequenced to verify the base pair order.
RATIONAL DESIGN OF LIPASE A LOOP HYBRIDS
113
Isolation of the periplasmic fraction E. coli HB2151 or E. coli K-12 PA340/T6 were grown in 50 mL tubes containing 10 mL 2xTY medium, ampicillin and isopropyl-β-D-galactopyranoside (IPTG, 1 mM). The tubes were incubated at 28ºC at 250 rpm for 16 h. The OD600 was recorded and the cells were harvested and resuspended in Tris HCl buffer (10 mM, pH 7.4). After centrifugation, the cells were resuspended in 200 µL buffer containing Tris HCl (10 mM, pH 8.0), sucrose (25% w/v), EDTA (2 mM) and lysozyme (0.5 mg.mL-1). After incubation on ice for 20 min, 50 µL buffer containing Tris HCl (10 mM, pH 8.0), sucrose (20% w/v) and MgCl2 (125 mM) was added. The suspension was centrifuged and the supernatant, containing the periplasmic fraction, was isolated and used as enzyme solution in the IPG ester assay. The protein content of this fraction was determined by performing a Bradford assay in triplicate using bovine serum albumin (BSA) as a standard (Pierce, Rockford, Illinois, USA). 1,2-O-isopropylidene-sn-glycerol ester assay Periplasmic fractions were diluted with MOPS buffer (0.07 M, pH 7.5), containing BSA (0.2% w/v), to a final volume of 100 µL. References were diluted correspondingly but contained no enzyme solution. The esters of IPG (1 mM) were dissolved in 10 mL MOPS buffer (0.07 M, pH 7.5), containing Tween 80 (14.3% w/v), and diluted to 50 mL with MOPS buffer (0.07 M, pH 7.5). 500 µL substrate solution was added to the enzyme solution and the final mixture was incubated in a water bath at 32°C. After incubation, 400 µL saturated NaCl solution and 10 µL internal standard solution (racemic 3-hexene-1-ol, 5 mg.mL-1 assay buffer) was added to the sample solution and the aqueous solution was extracted twice with 1 mL ethyl acetate. GC analysis was performed as described by Dröge et al 180. One unit (U) is defined as the amount of enzyme that hydrolyses 1 µmol IPG ester per minute. Enantiomeric excesses, ee, were calculated according to Chen et al and were defined as the ability of the enzyme to distinguish between enantiomers 181. All data were expressed as mean ± SEM. The statistical significance of differences was tested at a significance level of p < 0.05 using a two-tailed Student's t-test. Selection on selective minimal medium The aspartate auxotroph E. coli K-12 PA340/T6 was made chemically competent and transformed with 50 ng of mutant plasmid DNA of the library at position L51 in the loop 176. Growth selection was performed as described elsewhere. Cells were starved by incubation in 0.9% w/v NaCl for 2 h at 37°C and plated onto selective minimal medium agar plates supplemented with either aspartate or an ester of aspartate coupled to the desired enantiomer of IPG. To increase selection pressure, a phosphonate inhibitor, i.e. a phosphonate ester of the undesired enantiomer of IPG, was added to the medium. After two days of growth, 16 colonies were randomly selected and sequenced. The periplasm was isolated and activity and enantioselectivity was tested towards the real substrate, a racemic IPG butyrate ester.
As a second generation of LipA mutants, mutants from a library at position T50 in the loop while L51 was mutated in either a glutamine or a glycine residue, were selected using the growth selection procedure as described above. Colonies were selected, the periplasm was isolated and activity was tested on racemic IPG butyrate. Based upon their activity and enantioselectivity, the plasmid DNA was sequenced to verify the base pair order.
CHAPTER 7
114
Results Design and construction of the loop mutants As a result of the DALI search using B. subtilis LipA as a search model, Fusarium solani cutinase and Penicllium purpurogenum acetylxylan esterase were identified to be structurally related to LipA, being α/β hydrolase fold enzymes without lid-like domains as well. A structural alignment of the three proteins gave rmsd values of 1.9 Å for 1CEX (145 Cα's) and 2.9 Å for 2AXE (148 Cα's) 27 (figure 1).
Figure 1: Alignment of LipA with cutinase and acetylxylan esterase in the region of residues 11-20, the region of loop replacement. Residues 11-20 of LipA follow the first β-strand (β3); residues 11-15 form a loop including one of the oxyanion-hole residues (I12), while residues 16-19 form a short 310-helix preceding helix αA, which starts with the alanine residue at position 20. Several residues in this peptide line a hydrophobic pocket, which holds the IPG moiety of the IPG phosphonate inhibitors covalently bound to LipA 115. Among these residues, N18 has been suggested to be important for the enantioselective properties of LipA. Thus, this loop of residues 11-20 was chosen for exchange with loops from cutinase and acetylxylan esterase, thereby narrowing the active site cleft.
For both hybrid enzymes it was difficult to predict the effects of the inserted residues as the "real" structure can be different from the rigid replacements due to different residue and side chain characteristics, and because of the loop extension. In the cutinase loop, the residues closest to the IPG phosphonate inhibitor would be G41, S42, E44 and T50 (figure 2).
RATIONAL DESIGN OF LIPASE A LOOP HYBRIDS
115
Figure 2: A) Replacement of LipA residues 11-20 (light blue) with cutinase residues 41-53 (yellow); the R-IPG inhibitor bound to LipA is shown in stick representation. B) Stick representation of the cutinase loop. Especially the side chains of residues E44 and T50 could influence the binding of substrates by the hybrid enzyme. However, the polar nature of the E44 side chain makes it unlikely to point to the interior of the protein; instead, it would be more favourable to point outwards into the solvent.
The replacing loop of acetylxylan esterase comprises residues 12-23, two residues more than the LipA loop (figure 3). Steric problems are observed only for residues Y19 (too close to LipA A38 and K23) and S21 (too close to LipA L160). The hybrid protein would have to be in a different conformation than just the "cut and paste" structure, to overcome steric clashes. Residues E12, T13, A15 and S21 would mostly determine the nature of the binding pocket (figure 3B), and therefore it may be suggested that they might influence the substrate specificity of the hybrid enzyme.
Figure 2: A) Replacement of LipA residues 11-20 (light blue) with acetylxylan esterase residues 12-23 (green); the R-IPG inhibitor bound to LipA is shown in stick representation. B) Stick representation of the acetylxylan esterase loop.
CHAPTER 7
116
The overlapping PCR technique was applied to exchange the lipase loop of amino acids 11-20 with the cutinase loop of amino acids 42-53, and with the acetylxylan esterase loop of amino acids 12-23. After two rounds of PCR, the resulting product was cloned into pCANTABLip-CH, and the constructs were stably maintained in E. coli HB2151. Sequencing of the construct revealed a correct base pair order.
For cutinase, it was speculated that position T50 and in particular L51 might cause steric hindrance to L160 in the lipase molecule due to their respective side chains. First, a mutant library was constructed on position L51 in the 1CEX loop using saturation mutagenesis. After transformation to E. coli HB2151, a mutant library consisting of 1250 colonies was obtained. Sequence analysis of 10 clones revealed that only one mutant did not have an amino acid mutation, two clones had a mutation to serine, while the other seven clones all had a different amino acid mutation. As a consequence to its size, the library was found to be satisfactory. A second round of mutagenesis was applied, in which libraries on position T50 in the loop were constructed while position L51 was mutated to either a glycine or a glutamine. After transformation of E. coli HB2151, mutant libraries consisting of 1000 colonies were obtained. Sequence analysis of 10 clones revealed that all variants showed a different mutation. Thus, the libraries were found to be satisfactory. For the construction of the L160A and L160V mutants in the lipase molecule, QuikChange® was applied. The mutants were found to have the correct base pair order after sequencing. By applying the same method, the L160A and L160V mutations were incorporated in the gene of the L51G and L51Q variants as well. Sequencing results showed that the mutations had been incorporated correctly. Enzymatic activity of the loop mutants Plasmids encoding for the 1CEX and the 2AXE hybrids were transformed to E. coli HB2151, as well as wild type LipA. Colonies were selected and the periplasm was isolated. Translocation to the periplasm was verified on a 12% SDS PAGE gel (data not shown). The activity of the mutants compared to the wild type LipA on IPG butyrate esters was tested and found to be inverted compared to the wild type (table II). As the 1CEX hybrid showed to be the more promising variant in terms of enantioselectivity, saturation mutagenesis to the previously defined hotspots T50 and L51 in the loop was applied to further improve its activity and enantioselectivity. Table II: Activity and enantioselectivity of the constructed loop mutants in the hydrolysis of IPG butyrate (n=3).
Variant Conversion (%)
ee (%)
Enantiomer formed in excess
WT LipA 28.2 ± 1.0 -12.9 ± 1.2 (-)-IPG 1CEX 29.7 ± 1.0 +26.48 ± 1.6 (+)-IPG 2AXE 20.9 ± 1.9 +6.0 ± 0.7 (+)-IPG Selection of variants in the 1CEX loop From the computational design, it was speculated that steric hindrance might be caused by the side chains of T50 and in particular of L51 in the 1CEX loop. To improve the folding of the constructed variant, a saturated mutant library was initially constructed on position L51.
RATIONAL DESIGN OF LIPASE A LOOP HYBRIDS
117
Mutants were selected using the growth selection system based on bacterial growth as described previously. After two days of incubation, colonies appeared. Sixteen colonies were selected based upon their appearance, the periplasm was isolated and the activity and enantioselectivity towards racemic IPG butyrate was determined. All selected mutants showed an improvement of the inverted enantioselectivity (figure 4). Selection was in favour of mutation L51G, which was retrieved at six consecutive times. Remarkably, the parent 1CEX sequence was not retrieved. Together with the L51Q mutant, which was retrieved three times, this mutant showed a conversion of the IPG ester comparable to the parent loop mutant. As these mutations also proved to have an advantageous effect on the enantioselectivity, reflected in an improved ee value, the L51G and L51Q mutants were further characterised.
Figure 4: Activity and enantioselectivity of L51 variants in the hydrolysis of IPG butyrate. Assay conditions were described in the experimental section. Variants L51G and L51Q were used to parent a second mutant library on position T50. After two days of growth, colonies were again selected based upon their appearance, the periplasm was isolated and the activity and enantioselectivity towards racemic IPG butyrate was determined (table III). \
Based on either their activity or their enantioselectivity, variants were sequenced to determine the mutation. In contrast to the selection of L51 variants, in this selection round the parent sequence was selected for as well in combination with mutation L51G. However, mutation of position T50 did not have a favourable additive effect on the enantioselectivity: all variants showed a loss of enantioselectivity.
CHAPTER 7
118
Table III: Activity and enantioselectivity of variants at position T50 in the 1CEX loop after a second round of selection (n=3).
L51Q L51Q 37.7 ± 1.9 43.2 ± 2.1 (+)-IPG 6 12.5 ± 1.3 25.0 ± 1.4 (+)-IPG T50V 7 14.4 ± 1.7 36.8 ± 1.2 (+)-IPG T50W 8 27.7 ± 1.4 15.7 ± 0.9 (+)-IPG T50A 9 11.4 ± 1.8 27.4 ± 1.2 (+)-IPG T50V 10 27.1 ± 0.9 14.9 ± 0.8 (+)-IPG T50C 11 17.9 ± 1.0 51.2 ± 0.7 (+)-IPG T50P Enzymatic activity of L160 variants in the LipA molecule It was seen that the side chains of residues L51 and T50 might cause steric hindrance to the side chain of L160 in the LipA molecule. To improve the enzyme’s folding, shorter side chains for L160 were introduced by replacing the leucine residue by either an alanine or a valine residue. The L160A and L160V mutants were combined with the L51G and L51Q mutations in the 1CEX loop as well. Then, E. coli HB2151 cells were transformed with the mutants, colonies were selected and the periplasm was isolated to be used in the IPG assay. The resulting activities and enantioselectivities are shown in figure 5. No significant change in the activity of the enzyme was observed. However, there was a remarkable loss of enantioselectivity, which is especially seen in variants with mutations at both position L51 and L160. By shortening the side chains at both positions, more space might be available in the binding pocket to accommodate (-)-IPG, thus leading to a loss of enantioselectivity.
RATIONAL DESIGN OF LIPASE A LOOP HYBRIDS
119
Figure 5: Activity and enantioselectivity of variants at position L51 (1CEX loop) and L160 (LipA), and combination thereof (n =3). Assay conditions were described in the experimental procedures. Discussion All lipases share the α/β hydrolase fold, whereas most lipases though not all have an α-helix acting as a mobile lid, thus determining the conformation of the enzyme. From a structural point of view, the region of the lid is involved in the modulation of the activity and moreover the selectivity of lipases 24,272-276. Mutations in the lid region of Pseudomonas fragi lipase affected its chain length specificity from short-chain triglyceride substrates towards C8 substrates. Furthermore, its thermostability was increased 277. Modification of the lid structure can even alter enantioselectivity, as was seen for Geotrichum candidum lipase isozymes, where after rational recombination of selected portions of the two isozymes, a novel lipase with an enantioselectivity superior to that of the best wild-type parent isozyme was created 278. Deletion of the whole lid domain of human gastric lipase was even shown to abolish the pH-stat measurable activity on triglycerides as substrates 279. These examples all show that the role of the lid is subtler than simply controlling access to the enzyme active site.
In this work, we wanted to assess the effect of the insertion of somewhat longer loops near the active site on both the activity and the enantioselectivity in an enzyme without a lid-like structure. B. subtilis LipA naturally does not possess a lid-like structure. Based on a DALI search with this enzyme as a lead model, Fusarium solani cutinase and Penicillium purpurogenum acetylxylan esterase were identified as structural homologues to LipA. These enzymes feature the α/β hydrolase fold as well, which enables approaches of rational mutagenesis and domain swapping. A loop near the active site was chosen for the exchange, as this loop has been indicated to be important for the enantioselectivity of the
CHAPTER 7
120
enzyme. Previously, position N18 within this loop was shown to affect enantioselectivity in such a way, that enantiopreference was inverted towards the desired enantiomer of IPG, possibly by a better accommodation of the S-enantiomer in the active site pocket of LipA 115. In contrast to the present study, Eggert et al used a loop consisting of residues 39-51 of B .subtilis LipA, which is slightly more remote from the active site, to exchange it with loops from cutinase, acetylxylan esterase, and human pancreatic lipase 265. Here, more amino acids were inserted: nine for cutinase, ten for acetylxylan esterase, and 25 for human pancreatic lipase. Thus, instead of narrowing the active site cleft as presented in our work, it was hypothesised that the active site was now more shielded from the solvent and the substrate. As with the hybrids described in this work, their variants proved to be enzymatically active on either of their substrates of interest, p-nitrophenyl palmitate or tributyrin, although showing a lower activity than the wild type. In organic solvent, the activity of the variants containing the artificial lid was even lower. A change in enantioselectivity of the variants containing the lid-like structure was not determined. The overall decrease in activity of the variants compared to the wild type enzyme could be explained by the fact that the artificial lid-like structures were not as adapted to the core-protein of LipA as they were to their host-protein structure 280. It would however be challenging to combine both effects of narrowing the active site cleft and inserting a lid-like structure controlling access to the active site, and examine the effect these dramatic changes in protein structure have on the enantioselectivity.
In our work, the constructed variants showed an activity comparable to wild-type LipA, which can be concluded from the conversion rate. An inversion of the enantioselectivity was observed for both variants. From the models alone it might be said that the desired enantiomer S-IPG is better accommodated in the binding pocket than the R-IPG enantiomer, since the active site cleft is now narrower than in the wild type enzyme. However, other drastic replacements within the molecule are not unlikely to occur, as approximately seven per cent of the total protein structure is altered. As a consequence, it is difficult to predict how the substrate will exactly be accommodated in the binding pocket. For the variant with the inserted acetylxylan esterase loop, the change in enantioselectivity was modest; which might be explained by the fact that in total only one extra amino acid was inserted. Consequently, the active site is somewhat narrower, but can still accommodate both enantiomers reasonably well. Furthermore, the inserted stretch of amino acids might not be the optimal sequence. From the models, it is known which residues determine the nature of the binding pocket and which residues will most likely interact with IPG. Thus, it would be worthwhile to optimise the enantioselectivity of this loop hybrid by randomising the inserted loop region on these positions by e.g. cassette mutagenesis and assess which sequence is optimal.
The cutinase variant proved to be more promising as a starting point for further optimisation of the enantioselectivity. This variant has three extra amino acid residues compared to the wild type LipA, which might form an explanation for the further increased and improved enantioselectivity. The active site is in this case even narrower, thus more able to accommodate the S-enantiomer than the R-enantiomer. However, enantioselectivity being far from optimal, site-directed mutagenesis was applied to the specific hotspots T50 and L51 in the loop region as well as to L160 in the LipA molecule to reduce Van-der-Waals clashes. Selection of variants at position L51 was in favour of a glycine residue, as
RATIONAL DESIGN OF LIPASE A LOOP HYBRIDS
121
well as a glutamine residue. Moreover, an improvement of enantioselectivity was observed for all selected variants. Yet residue 51 is not directly interacting with the substrate. Hence, the improved enantioselectivity at this position is most likely due to a more compact folding of the enzyme, since the Van-der-Waals clashes are reduced by these mutations. This observation might be substantiated by the fact that a similar improvement of the enantioselectivity was not observed when position L160 in the LipA molecule was mutated to an alanine or a valine residue. Though these mutations resulted in active variants, the enantioselectivity was equal to or less than the enantioselectivity of the parent loop mutant. Combination of either L51G or L51Q with variants on position T50 in the cutinase loop resulted in a decrease in enantioselectivity as well, possibly due to the fact that the R-enantiomer is now accommodated better in comparison with the single mutant on position L51. The stretch of amino acids used in the cutinase loop might not be the optimal sequence with regard to enantioselectivity. Thus, to further optimise the enantioselectivity, the loop residues directly interacting with the substrate could be subjected to mutagenesis.
In conclusion, we have designed and engineered two lipase hybrids by exchanging stretches of amino acids. Changing approximately seven per cent of the protein structure, lipase hybrids were created with inverted enantioselectivities compared to the wild type parent enzyme. We anticipate that the knowledge gained from this study will serve as a lead for future protein engineering to create lipases that are more enantioselective than those offered by nature. Conclusion Bacillus subtilis lipase A (LipA) has excellent properties with regard to its kinetics for application in industrial biocatalysis of enantiopure β-adrenergic receptor antagonists and in particular of their chiral building block 1,2-O-isopropylidene-sn-glycerol (IPG). Nevertheless, its enantioselectivity towards these substrates is modest, and moreover directed towards the unwanted enantiomer of IPG. From its crystal structure it was clear that LipA has a relatively open active site cleft. It was previously demonstrated that a rigourous change in amino acid sequence and active-site architecture can give rise to altered enzyme function. This implies the replacement of several of the enzyme’s surface loop structures. Thus, we hypothesised that extending a loop lining the active site cleft and consequently narrowing it, might improve and invert the enzyme’s enantioselectivity. Homologues of LipA without a lid-like structure were identified, and the LipA hybrids were rationally designed. Exchange of loops from Fusarium solani cutinase and Penicillium purpurogenum acetylxylan esterase resulted in active LipA hybrids, which showed an inverted enantioselectivity. The enantioselectivity of the cutinase hybrid, being the more promising variant of the two, was further evolved by directed evolution. Thus, variants with improved enantioselectivity were selected by growth selection and characterised.
From an engineering point of view, this work extends the use of enzyme engineering by grafting loops near the active site. The combination of rational design of the hybrids and directed evolution on specific hotspots has proven successful. Our work, therefore, provides
CHAPTER 7
122
novel perspectives in the evolution of enzyme enantioselectivity, and can find an application in the improved synthesis of enantiopure β-adrenergic receptor antagonists. Acknowledgement This project was funded by the European Commission under project number QLK3-CT-2001-00519. We thank all the partners for their discussions and contributions leading to the generation of this project. Y.L.B. would like to thank Dr. R.H. Cool and Dr. A.M. van der Sloot for fruitful discussions.
8
A VALIDATED GAS CHROMATOGRAPHIC METHOD FOR
THE EVALUATION OF ENZYMATIC ENANTIOSELECTIVITY IN KINETIC
RESOLUTION EXPERIMENTS
YKELIEN L. BOERSMA, PAULA C. SCHELTINGA,
MELLONEY J. DRÖGE, REIN BOS & WIM J. QUAX
J SEP SCI 2005; 28: 501-505
APPLICATION OF LIPASE A
125
A validated gas chromatographic method for the evaluation of
enzymatic enantioselectivity in kinetic resolution experiments An enantioselective gas chromatography (GC) method has been developed and validated for the determination of the enantiomers of citronellol in kinetic resolution experiments. S-(-)-β-citronellol is a precursor for rose oxide. After solid-phase extraction (SPE) with ethyl acetate, the enantiomers of R-(+)-β-citronellol and S-(-)-β-citronellol and their corresponding acetate- and butyrate esters were separated through enantioselective GC respectively. The method was validated and found to be reproducible, specific, accurate and precise. Analyte recoveries and detection limits were also determined. The applicability of this method was shown in a kinetic resolution experiment using lipase A of Bacillus subtilis.
Introduction The terpenoid citronellol is of great interest to industrial companies as a flavor and fragrance compound 281,282. Citronellol can be used for the production of L-cis-oxide or rose oxide. The latter has a powerful fruity odour and an odour threshold of 0.5 ppb, whereas d-cis-oxide has a herbal odour and an odour threshold of 50 ppb. In a chemical reaction L-cis-oxide can be produced directly from the S-enantiomer of citronellol 283,284. The R-enantiomer is then considered isomeric ballast (scheme 1).
OH O
O2, hν
Na2SO3H+
S-(-)-β-citronellol (4S,2R)-(+)-cis-rose oxide
Scheme 1: Synthesis of rose oxide from enantiopure S-(-)-β-citronellol.
Traditionally, a single enantiomer is obtained through chemical separation of the enantiomers. However, this method is expensive for commercial production, not to mention unfriendly to the environment. Enzymes have become an attractive alternative for chemical production: enzymes are capable of enantioselective hydrolysis and esterification in an environmentally friendly and cost-effective process 13,247,285. The flavour industry is therefore interested in the use of enzymes to synthesise natural aromas. The use of hydrolases such as lipases and esterases in these syntheses has been described before for
CHAPTER 8
126
many different substrates 13,20,21,40,286. Several methods to assess enzymatic activity in this kind of experiments have been described. 264,287,288.
The main reaction for the production of the S-enantiomer of citronellol involves the enzymatic hydrolytic kinetic resolution of esters derived from racemic citronellol. One of the major problems regarding kinetic resolution concerns the development of reliable enantioselective assays, particularly when many enzymes need to be screened. Therefore, the purpose of this study is to develop a sensitive, specific, reproducible and validated enantioselective gas chromatography method for the determination and quantification of the enantiomers of citronellol in kinetic resolution experiments. Solid-phase extraction (SPE), a technique mostly applied in clinical and environmental studies, is used to isolate both substrate and product. The application of this method allows future automation of the assay. 289,290 Experimental procedures Materials Rac-β-citronellyl acetate and rac-β-citronellyl butyrate esters were synthesised as described in 291. R-(+)-β-citronellol and S-(-)-β-citronellol were purchased from Extrasynthese S.A. (Genay, France). Tween 80 and 4-morpholino propanesulfonic acid (MOPS) were purchased from Duchefa (Haarlem, The Netherlands). The following solvents were used: methanol HPLC grade (Biosolve, Valkenswaard, The Netherlands); ethyl acetate p.a. (Merck, Darmstadt, Germany); deionised water was freshly prepared by a Milli-Q system (Millipore Waters, Eschborn, Germany). A fermentor broth of B. subtilis 1051 was provided by Genencor International (Leiden, The Netherlands). This strain contained the plasmid pMA5lip 35, producing lipase A. This enzyme was purified as described previously 36. GC conditions GC analysis was performed on a Hewlett Packard 5890 series II gas chromatograph equipped with a 7673 split/splitless injector and a Hewlett Packard 3365 Chemstation under the following conditions: column BETA DEX 225 (25% 2,3-di-O-acetyl-6-O-TBDMS-β-cyclodextrin in SPB-20 poly(20% phenyl/80% dimethylsiloxane)), film thickness 0.25 µm; 30 m x 0.25 mm ID (SupelcoTM, Zwijndrecht, The Netherlands); oven temperature programme 35 min at 90 °C, 90-170 °C at 4 °C min-1; injector temperature 250 °C; detector (FID) temperature 300 °C; carrier gas helium; inlet pressure 125 kPa; linear gas velocity 40 cm s-1; split ratio 100:1; injected volume 1 µL. Sample preparation Stock solutions (2 mg/mL) and working solutions of R-(+)-β-citronellol and S-(-)-β-citronellol and their esters were freshly prepared in 0.07 M MOPS buffer pH 7.4, containing 14.3% w/v Tween 80, and diluted five times with 0.07 M MOPS buffer pH 7.4 (assay buffer). The aqueous solution was extracted using SPE. The C18 columns (Bakerbond 1 mL solid phase extraction column, #7020-01 100 mg per column, J.T. Baker, Deventer, The
APPLICATION OF LIPASE A
127
Netherlands) conditioned using respectively 2 mL of methanol and 2 mL of deionised water-methanol (1:1 v/v). 500 µL sample solution containing one of the enantiomers of citronellol was applied to the column. Next, the column was washed with 2 mL deionised water. Then, the column was dried under vacuum for 10 minutes. After that, R-(+)-β-citronellol and S-(-)-β-citronellol were eluted using 2 mL of ethyl acetate. GC analysis was performed as described above. Validation of the assay The assay was validated according to 292. Linalool was chosen as an internal standard. However, as linalool gave no reproducible results, a calibration curve was used instead. Selectivity To examine the selectivity in the determination of R-(+)-β-citronellol and S-(-)-β-citronellol and their esters, a sample solely containing assay buffer was extracted and analysed as described above. Linearity Eight samples in the range of 60 to 400 µg of R-(+)-β-citronellol or S-(-)-β-citronellol were diluted to 500 µL using assay buffer and extracted with SPE as described above in 2 mL ethyl acetate. Thus, a range of 30 to 200 ng of citronellol was obtained. To assess linearity, chromatographic signals were fitted to linear graphs using least-squares regression. Precision and accuracy The analytical precision was assessed at three mass levels. The samples were prepared by addition of 80, 200, and 400 µg of R-(+)-β-citronellol and S-(-)-β-citronellol to 500 µL assay buffer, and extracted as described above. The peak area was recorded and all samples were analysed once on six different days to assess reproducibility. Intra-day variation representing repeatability in peak area measurement was determined from six consecutive injections of the extracted samples. To obtain the intra-day and inter-day coefficients of variation, mean and standard deviations were calculated for each series of analyses. The precision around the mean value should not exceed a CV value of 15%. The accuracy of the method was assessed at the three mass levels mentioned above and determined by expressing the mean of the assayed concentrations as a percentage of the weighed-in concentration. These standards were prepared in ethyl acetate and directly injected into the chromatographic system. Recovery Standard solutions were prepared in ethyl acetate and directly injected into the chromatographic system. Recoveries were calculated by adding exact quantities (40, 100, 200 ng) of R-(+)-β-citronellol and S-(-)-β-citronellol to 500 µL assay buffer. The samples were extracted as described above. This procedure was performed ten times for both enantiomers and the average peak of the citronellol enantiomer in the sample was compared with the corresponding peak of the standard solution.
CHAPTER 8
128
Detection limit and lower limit of quantitation The detection limit of the method was calculated using a signal-to-noise ratio of ≥ 3. For that, the noise signal was obtained as the magnitude of the area of the chromatogram proceeding the R-(+)-β-citronellol and S-(-)-β-citronellol peaks. The lower limit of quantitation (LLOQ) was defined as the concentration that could be quantified with an accuracy of 90-110% and a precision of <15% using a calibration curve. The precision and accuracy were determined as described above. Citronellyl ester assay The esters of citronellol were dissolved in 5 mL 0.07 M MOPS buffer pH 7.4, containing 14.3% w/v Tween 80, and diluted to 25 mL with 0.07 M MOPS buffer pH 7.4. The final solution contained 1 mM of each enantiomer. The lipase A of B. subtilis was diluted with 0.1 M MOPS buffer to a final concentration of 1.5 µg of lipase in 150 µL assay buffer. 10 mM phosphate buffer, pH 8, was diluted correspondingly and used as a reference. 500 µL of substrate solution was added to the enzyme solution and the final mixture was incubated in a water bath at 32 °C for 75 min. After incubation, the aqueous solution was extracted using SPE. GC analysis was performed as described above. The amounts of R-(+)-β-citronellol and S-(-)-β-citronellol were calculated using the calibration curves. The enantiomeric ratio E, defined as the ability of the enzyme to distinguish between enantiomers, was calculated according to Chen et al 181. Results and Discussion Chromatographic conditions In this study, a temperature gradient programme was applied to combine a good separation of the enantiomers of citronellol and its acetate and butyrate esters, with a reasonable time of analysis. All peaks were symmetrical and separated as can be seen in figure 1; therefore, both citronellyl acetate and butyrate can serve as a substrate in hydrolysis experiments.
This was shown in the application of the method. From the peak area in the recorded chromatograms, the specific activity and enantioselectivity of lipase A of B. subtilis towards esters of citronellol with two different aliphatic side chains was calculated. Lipase A showed affinity for both stereoisomers of citronellyl esters, with an enantioselectivity of E = 1.2 towards R-(+)-β-citronellol (citronellyl acetate); and E = 1.2 (citronellyl butyrate) towards R-(+)-β-citronellol. The hydrolysis of the acetate and butyrate esters by the blanks was always below detection. Validation of the assay Linearity The curves for the extraction of R-(+)-β-citronellol and S-(-)-β-citronellol were linear in the range of 30 to 200 ng. Least square regression data of the calibration curves are shown in table I.
APPLICATION OF LIPASE A
129
Figure 2: A) Chromatogram of R-(+)-β-citronellol (1), S-(-)-β-citronellol (2), and rac-β-citronellyl acetate (3). B) Chromatogram of R-(+)-β-citronellol (1), S-(-)-β-citronellol (2), and rac-β-citronellyl butyrate (3). The capacity factors (k’) were: 10.37 for R-(+)-β-citronellol; 10.28 for S-(-)-β-citronellol; 12.63 for rac-citronellyl acetate; 15.19 for rac-citronellyl butyrate. The void volume (retention time of methane) was 3.324 min. Table I: The mean ± SD (n = 10) calibration curve for R-(+)-β-citronellol and S-(-)-β-citronellol expressed by a coefficient of correlation and a regression equation.
Precision and accuracy The intra-day coefficients of variation ranged between 1.6 and 2.0%, and between 0.8 and 1.1%, for R-(+)-β-citronellol and S-(-)-β-citronellol respectively; the inter-day coefficient was in the range of 0.7 to 1.1% for R-(+)-β-citronellol and of 0.3 to 1.7% for S-(-)-β-citronellol. The intra-day and inter-day coefficients of variation were low and less than 15%. The analysis can therefore be considered precise. The proposed method can be considered accurate in the range of 40 to 200 ng of citronellol as the coefficients of variation are less than 10% (table II). Table II: Coefficients of variation (C.V.) (%) (n = 6) for the precision by intra-day variation (repeatability) and inter-day variation (reproducibility) and accuracy.
Amount (ng) R-(+)-β-citronellol S-(-)-β-citronellol Repeatability (%) 40 1.6 0.8 100 1.6 1.1 200 2.0 0.7 Reproducibility (%) 40 1.1 1.7 100 0.7 0.3 200 0.8 0.6 Accuracy (%) 40 2.9 1.8 100 1.6 2.3 200 3.2 2.2 Recovery The recoveries of R-(+)-β-citronellol and S-(-)-β-citronellol using SPE are shown in table III. Both enantiomers show recoveries above 76%. The recovery was acceptable and considered to be reproducible. Table III: Recoveries (%) ± SD of R-(+)-β-citronellol and S-(-)-β-citronellol (n = 10).
Amount (ng) R-(+)-β-citronellol S-(-)-β-citronellol 40 83.5 ± 8.8 85.5 ± 10.0 100 81.1 ± 10.7 83.0 ± 4.6 200 76.2 ± 8.7 82.9 ± 5.1 Detection limit and lower limit of quantitation Limits of detection (LOD) were established as the amount of analyte that provides a signal-to-noise ratio ≥ 3. The LOD had a value of 5 ng for citronellol and of 4 ng for citronellyl esters. The lower limit of quantitation (LLOQ) was defined as the lowest calibration standard which could be quantified with an accuracy of 90-110% and a precision of < 15%. For this method, the LLOQ was 40 ng for both R-(+)-β-citronellol and S-(-)-β-citronellol.
APPLICATION OF LIPASE A
131
Conclusion In this study, we have demonstrated that racemic mixtures of citronellol and its acetate and butyrate esters can be separated on a chiral stationary phase by gas chromatography. Until now, no validated enantioselective assays have been described for the determination of citronellol enantiomers, although assays using a chiral column have been reported. However, kinetic resolution experiments as described in our study were not performed 293,294.
Previous studies by gas chromatography have identified lipases that are capable to catalyse esterification reactions in organic solvents. Moreover, lipases have been employed in preparing many flavour and fragrance esters 295-297. However, the stereospecificity of the enzymatic hydrolysis of racemic citronellyl esters in kinetic resolution experiments has not been explored before by gas chromatography. In this respect, we present a novel, convenient and direct method for analysis of citronellol and its esters in kinetic resolution experiments, in which derivatisation of the analyte is not required. The method has been validated, and is found to be reproducible, specific, accurate and precise. Furthermore, the application of SPE allows automation of the assay. In future research, this method will be applied in the selection of highly enantioselective enzymes, such as lipases and esterases, towards S-(-)-β-citronellol by directed evolution. Acknowledgement The authors would like to thank ing. P.G. Tepper (Faculty of Mathematics and Natural Sciences, Pharmacy, Department of Medicinal Chemistry, University of Groningen, The Netherlands) for his help in the synthesis of the citronellyl esters and dr. H.J. Woerdenbag for critically reading this manuscript.
9
SUMMARY, DISCUSSION
& GENERAL PERSPECTIVES
SUMMARY, GENERAL DISCUSSION & PERSPECTIVES
135
Summary Chirality in drug molecules is an important aspect in drug design, research, and development. Enantiomers can exhibit different therapeutic activities and can interact differently with chiral biomolecules, such as receptors. Chirality thus introduces selectivity, and often specificity, in drug action. Still, despite the knowledge that a single enantiomer can cause the therapeutic effect, many drugs are administered as a racemic mixture as there is no means of separating both enantiomers available. Biocatalysis has emerged as an important tool in the industrial synthesis of pharmaceuticals and pharmaceutical intermediates, giving rise to novel solutions for the chiral separation of enantiomers. Enzymes can perform intricate regioselective and/or enantioselective biochemical transformations and can accelerate reaction rates by enormous factors. High enzymatic selectivity also allows efficient reactions with few by-products, thereby making enzymes an environmentally friendly alternative to conventional chemical catalysts. However, the number and diversity of enzymatic applications are modest, perhaps in part because of perceived or real limitations of biocatalysts, such as limited enzyme availability, substrate scope, and operational stability. To date, successful applications of biocatalysts have been largely confined to hydrolytic enzymes such as lipases and esterases. The great interest in these enzymes is mainly due to their favourable properties in terms of enantioselectivity, regioselectivity and broad substrate specificity (Chapter 1). This thesis focuses on the use of hydrolytic enzymes from Bacillus subtilis to separate two enantiomers from a racemic mixture. Carboxylesterase A (CesA) and B (CesB) are intracellular enzymes of B. subtilis 168. These enzymes belong to the α/β hydrolase fold family of enzymes, with a catalytic triad composed of a serine, a histidine and an aspartate residue. Lipase A (LipA) is an extracellular enzyme secreted by B. subtilis 168 and consists of only 181 amino acids (19 kD). Its crystal structure shows the α/β hydrolase fold, but in contrast to other lipases LipA does not contain a lid structure controlling access to the active site.
Most applications of enzymes in biocatalysis do not rely on the natural reaction catalysed by them but rather use non-natural substrates. The hydrolases described above can be applied in the enantiopure production of 1,2-O-isopropylidene-sn-glycerol (IPG), a non-natural substrate. This chiral synthon is a precursor in the synthesis of β-adrenoceptor antagonists. To increase the yield of the desired enantiomer S-(+)-IPG, the wild type enzymes need to be optimised. In this respect, activity, stability, substrate specificity and enantioselectivity need to be improved. Two strategies can be applied: directed evolution and rational design. Directed evolution has emerged as a powerful tool to overcome the limitations of the use of wild type enzymes for industrial biocatalysis. The enzyme’s properties and functions can easily be engineered without any required knowledge of the structure, in contrast to rational design. The technique is essentially composed of two steps: first, mutagenesis of the gene(s) encoding the enzyme(s), and second, identification of the desired biocatalyst variants within these mutant libraries by either screening or selection. The gene(s) encoding the improved variants are then used to parent the next round of directed evolution. Thus, the
CHAPTER 9
136
ultimate goal of directed evolution is to accumulate improvements through repetitive rounds of mutagenesis and identification.
The second step in directed evolution experiments, identification of interesting variants by screening or selection, remains the most critical one. The main advantage of selection over screening is that many more variants in the library can be analysed simultaneously. Selection strategies exploit conditions favouring the exclusive survival of desired variants; consequently uninteresting variants are ignored. As enzymes are unable to amplify themselves, selection systems must simultaneously select for the genes encoding them as well. Thus, there is always a physical linkage between the gene, the enzyme it encodes, and the product of the enzyme’s activity. Selection techniques such as genetic selection, phage display, cell surface display and ribosome display establish this linkage. An overview of the currently available strategies to improve on nature’s catalysts is given in Chapter 2. The possible applicability of the homologous carboxylesterases CesA and CesB in industrial biocatalysis is described in Chapter 3. Using genome analysis, these enzymes were found to be highly homologous to the previously characterised enzyme carboxylesterase NP from B. subtilis Thai I-8, a very efficient enantioselective biocatalyst for the kinetic resolution of non-steroidal anti-inflammatory drug (NSAID) esters. The cesA and cesB genes were isolated, cloned, expressed, and biochemically characterised. All three homologous enzymes were applied in the enantioselective hydrolysis of a wide range of substrates. A marked difference in substrate specificity and enantioselectivity was observed for the three carboxylesterases, while their physico-chemical characteristics remained similar. CesA and carboxylesterase NP seem more suited to be applied in the kinetic resolution of chiral carboxylic acids such as NSAIDs, while CesB can be applied in the enantioselective production of IPG. To improve the enantioselectivity of CesA towards the desired enantiomer S-(+)-IPG, directed evolution can be applied. In order to be able to select for improved variants, a phage display method was developed. A method to present CesA as a fusion protein to the phage coat protein g3p was established (Chapter 4). The cesA gene was cloned upstream of the sequence encoding for the g3p phage coat protein. To optimise the display of the intracellular enzyme CesA, the effect of several different signal peptide sequences was assessed. Both the Sec and Tat translocation pathway were taken into account. Functional display of CesA could only be achieved when a Sec-dependent signal peptide was directing its translocation. In contrast, the use of a Tat-dependent signal peptide did not result in functional phage display; however it did result in carboxylesterase precursor processing. Thus, this phage display strategy provides a means for future cloning and selection of (intracellularly produced) heterologous proteins. Phage display technology was also applied in the directed evolution of Bacillus LipA, as described in Chapter 5. The lipA gene was cloned upstream of a phage g3p encoding sequence, and downstream of a modified g3p signal sequence in the phagemid pCANTAB 5E. Cassette mutagenesis was applied to a confined region of the lipase; thus, a library on a stretch of amino acids near the active site (residues 16-20) was constructed. This region was thought to be of influence in the binding and conversion of the substrate. A dual selection
SUMMARY, GENERAL DISCUSSION & PERSPECTIVES
137
system was developed for the isolation of enzyme variants, in which immobilised enantiomeric phosphonate suicide inhibitors were employed. Variants expressed as a fusion protein to the g3p phage coat protein were incubated on (+)- and (-)-IPG phosphonate suicide inhibitors. Mutants with an inverted enantioselectivity towards (+)-IPG were selected. These variants were further characterised and the 3D structures of the Sc and Rc inhibitor-LipA complexes were determined to clarify the enantioselective characteristics of the enzyme. In conclusion, a novel selection method for the evolution of enantioselectivity of B. subtilis LipA has become available. Additionally, this strategy could be generally applied in the selection of other enantioselective hydrolases towards chiral substrates. The major drawback of the phage display system however, is that selection is solely based on binding of the enzyme to the inhibitor and not on product release and catalytic turnover. Furthermore, binding does not necessarily correlate with the enzyme’s catalytic activity. Therefore, to select for both enantioselectivity and catalytic activity of LipA, a growth selection system was developed. Chapter 6 describes the use of an E. coli aspartate auxotrophic strain transformed with a mutant library to select improved variants on minimal medium supplemented with an ester of enantiopure IPG coupled to aspartate. The mutant library was constructed on residues 132-136, which comprises the aspartate residue at position 133, a member of the catalytic triad. To impose selection pressure, a phosphonate suicide inhibitor coupled to the undesired enantiomer of IPG was added to the minimal medium as well. Thus, a dual selection system was established. Variants with an enantioselectivity towards S-(+)-IPG were selected and further characterised. One variant was found having a repositioned catalytic acid residue. In conclusion, the developed system provides a novel tool in the selection of Bacillus LipA variants, in which both catalytic activity and enantioselectivity are taken into account. Moreover, this system is generally applicable in the selection of enantioselective hydrolases with any chiral substrate. Directed evolution was used to evolve the lipase’s enantioselectivity. However, as the crystal structure of LipA was solved previously, rational design could be applied as well (Chapter 7). Our hypothesis was that the wild type’s poor enantioselectivity towards IPG esters might be due to the fact that this enzyme is one of the few lipases that does not contain a lid structure controlling access to the active site. Therefore, longer loops from two enzymes showing a high structural homology to B. subtilis LipA, Fusarium solani cutinase and Penicillium purpurogenum acetylxylan esterase, were introduced in LipA,. Thus, a loop lining the active site cleft was extended, an artificial lid-like structure was created and the active site cleft might become narrower. The hybrid enzymes proved to be enzymatically active with an inverted enantioselectivity towards S-(+)-IPG. The enantioselectivity of the cutinase hybrid was further optimised by applying saturation mutagenesis to hotspots in the molecule. The combination of rational design of the hybrids and directed evolution on specific hotspots proved to be successful. This work, therefore, provides novel perspectives in the evolution of enzyme enantioselectivity. In Chapter 8, the application of Bacillus LipA in the kinetic resolution of citronellyl esters is described. A specific, sensitive, and reproducible gas chromatographic method was developed for the determination and quantification of citronellol enantiomers. This
CHAPTER 9
138
terpenoid is used in the flavour and fragrance industry. Thus, an assay is available for the determination of enantioselectivity of lipase variants in kinetic resolution experiments. General discussion and perspectives Enzymes have become an attractive alternative to conventional catalysts in numerous industrial processes. However, their properties do not always meet the criteria of the application of interest. Directed evolution has emerged as a powerful means to overcome these limitations and alter the enzyme’s characteristics by applying mutagenesis techniques and subsequently screen or select for improved variants 46. Many examples of the successful application of directed enzyme evolution to improve enzymatic characteristics have been described in literature 15,55,92,115,286.
It is generally accepted that improvement of an enzyme’s enantioselectivity is a difficult parameter to probe. For many enzymes, such as lipases and esterases, no high throughput methods are available; those assays that have been developed, are usually time-consuming 40. In this thesis, the focus was on the selection of Bacillus subtilis lipase A (LipA) variants and carboxylesterase A (CesA) as enantioselective biocatalysts for the production of enantiopure 1,2-O-isopropylidene-sn-glycerol (IPG). To invert and improve the enantioselectivity of LipA towards the desired enantiomer S-(+)-IPG, cassette mutagenesis was applied and five libraries on regions near or at the active site were constructed. The second step in directed evolution, the identification of variants, is always critical. In general, selection is preferred over screening techniques, since selecting variants is less labour-intensive. Moreover, selection is more efficient than screening as it allows for the examination of variants simultaneously. Thus, we chose to select rather than screen for improved enantioselectivity.
In a first attempt, we employed phage display as a method of choice to link genotype to phenotype in order to select for variants with improved enantioselectivity (Chapter 5). It was previously established that LipA, an enzyme transported to the periplasmic space due to fusion to the g3p signal sequence, can be displayed as a fusion protein to the g3p coat protein while retaining its activity 103. For an intracellular enzyme such as CesA, display is usually less easily achieved. By inserting Sec-dependent signal sequences, the enzyme was transported to the periplasm where phage assembly could take place. Thus, CesA variants can in future also be selected by applying phage-display technology (Chapter 4). Several strategies can be applied to select phage-enzyme variants. Here, the principle of indirect selection by affinity-capture was applied. A fast and reproducible system based on enantiomeric suicide phosphonate inhibitors was developed. The applicability of the system was demonstrated by the selection of a mutant with inverted enantioselectivity (Chapter 5). The combination of phage display with affinity selection of mutants has been mostly applied to improve substrate specificity 102,105,116,235; however, the novelty of the developed dual selection strategy lies in the fact that it was applied for the first time to select for enantioselectivity.
Unfortunately, affinity selections do have some limitations. A disadvantage of the use of suicide inhibitors is that the selection process is based on binding and not on product release and catalytic turnover(s). Consequently, binding does not necessarily correlate with
SUMMARY, GENERAL DISCUSSION & PERSPECTIVES
139
the catalytic activity of the enzyme. Thus, affinity selections ideally find their application in changing the enzyme’s substrate specificity, and not when an enhancement of rate acceleration or turnover is required. Direct selection of enzymes on the basis of catalytic activity is more difficult, as reaction products readily diffuse from the reaction site. In order to inhibit this diffusion, a physical link between the phage-enzyme and the substrate should be established 110. A relatively novel and promising approach is the combination of in vitro selection with in vivo enzymatic activity. Here, substrate- and enzyme-encoding DNA are both introduced in E. coli and expressed in the cytoplasm, the substrate being fused to one of the coat proteins of M13. After a catalytic reaction in the cytoplasm, the product, likewise fused to the coat protein, is incorporated in the phage coat and displayed on its surface. Upon affinity-capture of the product in vitro, the gene encoding for the selective enzyme is automatically selected for as well 122,123.
Selection systems are based on a link between phenotype and genotype. The simplest link is achieved by using a cellular compartment. All in vivo selection systems have in common that the genetic library obtained in the first step of directed evolution must be transformed into cells. Selection is based on the introduction of a specific catalytic activity which provides a growth advantage to microorganisms possessing that particular activity. In our work, we developed an in vivo selection system in which an aspartate auxotrophic bacterium was dependent on the hydrolysing capability of the expressed lipase variant (Chapter 6). By supplementing an enantiopure ester of IPG and aspartate to the minimal medium, lipase variants could be selected not only for catalytic activity, but also for enantioselectivity. The dual selection strategy which was applied in phage display was introduced here as well to select only the more enantioselective variants. Thus, less enantioselective mutants were picked less frequently. Three rounds of selection were performed, in which the imposed selection pressure was increased. However, one can argue that there is a discrepancy between kcat and Km: kcat towards the phosphonate can be high while the Km towards the aspartate ester can be equally high. Thus, interesting variants will be missed. Nevertheless, in our case the selection strategy yielded lipase variants with an inverted and improved enantioselectivity. The strategy described in this thesis is generally applicable for hydrolases, as long as they are translocated to the periplasmic space. As it was established that the intracellular enzyme CesA can also be translocated to the periplasm upon insertion of a Sec-dependent signal sequence, this system can also be applied in the selection of improved variants of this enzyme. To invert and improve the enantioselectivity of LipA, we have made use of two different mutant libraries constructed by cassette mutagenesis in order to prove the principle of the developed selection systems. For further enhancement of the enantioselectivity towards chiral esters, a number of suggestions can be made.
First, several strategies to mutate the gene can be applied.
- Saturation mutagenesis Due to the small size of LipA in comparison with other lipases, it is relatively easy to mutate every codon in the gene. This approach would yield 181 mutant libraries. Funke et al constructed saturated mutant libraries using NNS (N = A, C, G, or T; S = C or G) oligonucleotides at each possible codon. The libraries were screened for enantioselectivity
CHAPTER 9
140
towards their model substrate 1-(2-naphthyl)ethyl acetate and several hotspots were identified 48.
- Combinatorial active site saturation test (CAST) In a different approach, the so-called combinatorial active site saturation test (CAST) could be applied to obtain improved variants. Reetz et al originally developed this method to expand the substrate acceptance of enzymes, but this technique can also be applied to improve the enantioselectivity of biocatalysts. The method constitutes a compromise between saturation mutagenesis at a single site and randomisation at multiple sites by cassette mutagenesis and combines both rational design and directed evolution. First, regions are rationally selected, then they are randomised 298.
- Recombination of beneficial mutations DNA shuffling was first introduced by Stemmer 60, and today it is one of the most important recombination strategy, though many variations have been developed 66. All techniques allow further improvement of the enantioselectivity by combining mutations beneficial for the enantioselectivity of the enzyme. The genes are fragmented and subsequently assembled into a library of full-length genes by repeated cycles of PCR. However, the main disadvantage of DNA shuffling for recombination is that large quantities of DNA are required, while during the procedure, large amounts of DNA are lost. Staggered extension PCR (StEP) is an alternative method which generates full-length recombined genes in the presence of template(s). It is possible, during only one round of PCR, to create a library of recombined DNA sequences 299. Our mutants from the five libraries obtained with cassette mutagenesis and selected in a first round of growth selection will undergo this procedure.
- Rational design of loop variants This strategy involves a combination of rational design and directed evolution. When the 3D structure of LipA was elucidated, it became clear that due to the absence of a lid-like structure, this lipase could be regarded as a minimal α/β hydrolase fold enzyme. Thus, it was tempting to speculate on the influence of inserting a lid-like structure on the enantioselectivity of the lipase. After structural alignments with Fusarium solani cutinase (1CEX) and Penicillium purpurogenum acetylxylan esterase (2AXE) were made, stretches of amino acids near the active site of LipA were exchanged for stretches of amino acids (i.e.. loop regions) from 1CEX and 2AXE (Chapter 7). Thus, a loop near the active site was extended, thereby narrowing the active site cleft. Indeed, an effect on the enantioselectivity was observed: the selectivity was now directed towards the desired enantiomer S-(+)-IPG. The enantioselectivity was further improved by saturation mutagenesis at predefined hotspots. Thus, this example clearly demonstrates the success of the combination of rational design and directed evolution.
- Incorporation of unnatural amino acids To date, over 30 unnatural amino acids have been incorporated into proteins with high fidelity and efficiency. This has changed the approach to protein use and study. An important question in this case is how the incorporation of unnatural amino acids would alter an enzyme’s characteristics beyond the 20 known amino acids. Unnatural amino acids have been incorporated into enzymes before, though alterations in an enzyme’s activity are rarely reported. 300 Recently, Jackson et al incorporated unnatural amino acids in the active
SUMMARY, GENERAL DISCUSSION & PERSPECTIVES
141
site of nitroreductase from E. coli and successfully demonstrated that unnatural amino acids can indeed tune activity for different substrates and improve the efficiency of this enzyme > 30-fold beyond that accessible with the set of natural amino acids 301. In our case, it would be compelling to incorporate an unnatural amino acid at position N18 in the lipase molecule, as this position has proven to be important in enantioselectivity. In the phage-display selected variant, the asparagine side chain was replaced by an isoleucine side chain, which has an additional branch on the Cβ atom in relation to the asparagine. Around position N18 is a relatively large amount of space, even though small local rearrangements need to be made in order for the isoleucine side chain to fit in the molecule. Unnatural amino acids tend to have even bigger side chains in comparison with the isoleucine residue, thus requiring bigger rearrangements. Possibly the active site cleft is simultaneously narrowed, thus enabling only one enantiomer to be accommodated. For a more mechanistic study, the H156 from the catalytic triad could be taken into consideration for the incorporation of an unnatural amino acid. In the performed growth selections, a mutant with a lysine instead of a histidine at this position was found. To further determine the effect of the H156 in catalysis, histidine analogues could be incorporated.
Another approach would be to use a different substrate. In this work, we described the selection of variants enantioselectively hydrolysing an acyl ester of the primary chiral alcohol IPG. The enantioselectivity of the lipase towards this enzyme is quite poor. The same holds true for the enantioselective hydrolysis of esters of citronellol, a primary alcohol used in fragrances 302 (Chapter 8). Possibly enantioselectivity is improved when a more bulky substrate is used for selection, e.g. an ester of a secondary alcohol such as phenyl ethanol. LipA showed an enantiopreference of > 99% towards the R-enantiomer (unpublished data). However, the synthesis of phosphonates and aspartate esters of phenyl ethanol proved to be rather difficult, the reaction products being liable to decomposition. One could also think of the application of fluorogenic substrates. Upon hydrolysis of a fluorogenic substrate in a growth selection system, cells compartmentalising genes encoding for active enzymes would become fluorescent. However, the major limitation of growth selections is that they are hampered by transformation efficiency. To overcome this limitation, in vitro compartmentalisation could be an attractive alternative. This technique is based on water-in-oil emulsions, where the water phase is dispersed in the oil phase with the aid of surfactants to form microscopic aqueous compartments. Thus, artificial cells of approximately 5 fL are created. These droplets contain on average a single gene; here, transcription, translation and expression of the resulting proteins can all take place. The oil phase remains mostly inert and limits the diffusion of genes and proteins between compartments. Active enzyme-encoding genes can be isolated on the basis of the presence of the product, while inactive enzyme-encoding genes can be discarded on the basis of an unmodified substrate. Using fluorogenic substrates, artificial cells compartmentalising genes encoding for active enzymes would become fluorescent. Consequently, these compartments can be analysed by FACS, enabling a fast high-throughput screening of variants 69,148,303.
As an alternative, carboxylesterases instead of lipases could be used as biocatalysts. We have shown that carboxylesterases of B. subtilis are capable of enantioselective hydrolysis of IPG esters. However, it has been demonstrated in a previous study that elucidation of the crystal structure of the enzyme CesA is a difficult task. Although a wild type
CHAPTER 9
142
carboxylesterase data set was obtained, no heavy atom derivatives could be obtained to solve the phase problem. Thus, a combined approach of rational design and directed evolution is not yet feasible. However, we did succeed in translocation of the intracellular enzyme of Bacillus to the periplasmic space of E. coli, thereby enabling both phage assembly and growth selection. Thus, it would be interesting to identify hotspots important for enantioselectivity in the gene by introducing random mutations and select improved variants with either selection system.
2. Tucker,G.T. Chiral switches. The Lancet 355, 1085-1087 (2000).
3. Ranade,V.V. & Somberg,J.C. Chiral Cardiovascular Drugs: An Overview. Am J Ther 12, 439-459 (2005).
4. Breuer,M., Ditrich,K., Habicher,T., Hauer,B., Kesseler,M., Stürmer,R., & Zelinski,T. Industrial methods for the production of optically active intermediates. Angew Chem Int Ed Engl 43, 788-824 (2005).
5. Agranat,I., Caner,H., & Caldwell,J. Putting chirality to work: the strategy of chiral switches. Nature reviews 1, 753-768 (2002).
6. Stinson,S.C. Chiral drugs. Chem Eng News 78, 55-78 (2000).
7. Mehvar,R. & Brocks,D.R. Stereospecific pharmacokinetics and pharmacodynamics of beta-adrenergic blockers in humans. J Pharm Pharm Sci 4, 185-200 (2001).
8. Kato,R., Ikeda,N., Yabek,S.M., Kannan,R., & Singh,B.N. Electrophysiologic effects of the levo- and dextrorotatory isomers of sotalol in isolated cardiac muscle and their in vivo pharmacokinetics. J Am Coll Cardiol 7, 116-125 (1986).
9. Bartsch,W., Sponer,G., Strein,K., Muller-Beckmann,B., Kling,L., Bohm,E., Martin,U., & Borbe,H.O. Pharmacological characteristics of the stereoisomers of carvedilol. Eur J Clin Pharmacol 38, S104-S107 (1990).
10. Bornscheuer,U.T. & Buchholz,K. Highlights in biocatalysis - Historical landmarks and current trends. Eng Life Sci 5, 309-323 (2005).
11. Cipolla,L. Combinatorial libraries of biocatalysts: application and screening. Comb Chem High Throughput Screen 7, 101-114 (2004).
24. Dugi,K.A., Dichek,H.L., & Santamarina-Fojo,S. Human Hepatic and Lipoprotein Lipase: The Loop Covering the Catalytic Site Mediates Lipase Substrate Specificity. J. Biol. Chem. 270, 25396-25401 (1995).
26. Martinez,C., Nicolas A., Tilbeurgh,H.v., Egloff,M.P., Cudrey,C., Verger,R., & Cambillau,C. Cutinase, a lipolytic enzyme with a preformed oxyanion hole. Biochemistry 33, 83-89 (1994).
27. Pouderoyen,G.v., Eggert,T., Jaeger,K.-E., & Dijkstra,B.W. The crystal structure of Bacillus subtilis Lipase: a minimal α/β hydrolase fold enzyme. J Mol Biol 309, 215-226 (2001).
28. Ferrato,F., Carrière,F., Sarda,L., & Verger,R. A critical reevaluation of the phenomenon of interfacial activation. Methods Enzymol 286, 327-347 (1997).
29. Jaeger,K.-E., Dijkstra,B.W., & Reetz,M.T. Bacterial biocatalysts: molecular biology, three-dimensional structures, and biotechnological applications of lipases. Annu Rev Microbiol 53, 315-351 (1999).
30. Jaeger,K.-E. & Reetz,M.T. Directed evolution of enantioselective enzymes for organic chemistry. Curr Opin Chem Biol 4, 68-73 (2000).
32. Ollis,D.L., Cheah,E., Cygler,M., Dijkstra,B.W., Frolow,F., Franken,S.M., Harel,M., Remington,S.J., Silman,I., Schrag,J.D., & et al The alpha/beta hydrolase fold. Protein Eng 5, 197-211 (1992).
33. Jaeger,K.-E. & Reetz,M.T. Microbial lipases form versatile tools for biotechnology. Tibtech 16, 396-403 (1998).
34. Eggert,T., Pouderoyen,G.v., Dijkstra,B.W., & Jaeger,K.-E. Lipolytic enzymes LipA and LipB from Bacillus subtilis differ in regulation of gene expression, biochemical properties and three-dimensional structure. FEBS letters 502, 89-92 (2001).
35. Dartois,V., Coppée,J.Y., Colson,C., & Baulard,A. Genetic analysis and overexpression of lipolytic activity in Bacillus subtilis. Appl Environ Microbiol 60, 1670-1673 (1994).
36. Lesuisse,E., Schanck,K., & Colson,C. Purification and preliminary characterization of the extracellular lipase of Bacillus subtilis 168, an extremely basic pH-tolerant enzyme. Eur J Biochem 216, 155-160 (1993).
37. Kunst,F., Ogasawara,N., Moszer,I., Albertini,A.M., Alloni,G., Azevedo,V., Bertero,M.G., Bessieres,P., Bolotin,A., Borchert,S., Borriss,R., Boursier,L., Brans,A., Braun,M., Brignell,S.C., Bron,S., Brouillet,S., Bruschi,C.V., Caldwell,B., Capuano,V., Carter,N.M., Choi,S.K., Codani,J.J., Connerton,I.F., Danchin,A., & et al The complete genome sequence of the gram-positive bacterium Bacillus subtilis. Nature 390, 249-256 (1997).
38. Eggert,T., Pancreac'h,G., Douchet,I., Verger,R., & Jaeger,K.-E. A novel extracellular esterase from Bacillus subtilis and its conversion to a monoacylglycerol hydrolase. Eur J Biochem 267, 6459-6469 (2000).
39. Fojan,P., Jonson,P.H., Petersen,M.T.N., & Petersen,S.B. What distinguishes an esterasee from a lipase: A novel structural approach. Biochimie 82, 1033-1041 (2000).
40. Bornscheuer,U.T. & Kazlauskas,R.J. Hydrolases in organic chemistry. Regio- and stereoselective biotransformations. (Wiley-VCH Verlag GmbH, Weinhein, Germany; 1999).
41. Bornscheuer,U.T. Microbial carboxyl esterases: classification, properties and application in biocatalysis. FEMS Microbiol Rev 26, 73-81 (2002).
42. Quax,W.J. & Broekhuizen,C.P. Development of a new Bacillus carboxyl esterase for use in the resolution of chiral drugs. Applied Microbiology and Biotechnology 41, 425-431 (1994).
43. Azzolina,O., Collina,S., & Vercesi,D. Stereoselective hydrolysis by an esterase: a strategy for resolving 2-(R,R'-phenoxy)propionyl ester racemates. Il Farmaco 50, 725-733 (1995).
REFERENCES
147
44. Azzolina,O., Vercesi,D., Collina,S., & Ghislandi,V. Chiral resoltuion of methyl 2-aryloxypropionates by biocatalytic stereospecific hydrolysis. Il Farmaco 50, 221-226 (1995).
45. Dröge,M.J., Bos,R., Boersma,Y.L., & Quax,W.J. Comparison and functional characterisation of three homologous intracellular carboxylesterases of Bacillus subtilis. J Mol Catal B: Enzymatic 32, 261-270 (2005).
46. Bornscheuer,U.T. Trends and challenges in enzyme technology. Adv Biochem Engin Biotechnol 100, 181-203 (2005).
47. Dalby,P.A. Optimising enzyme function by directed evolution. Curr Opin Struc Biol 13, 500-505 (2003).
48. Funke,S.A., Otte,N., Eggert,T., Bocola,M., Jaeger,K.E., & Thiel,W. Combination of computational prescreening and experimental library construction can accelerate enzyme optimization by directed evolution. Protein Engineering, Design and Selection 18, 509-514 (2005).
49. Hiroki,K., Hirotaka,M., & Jiro,S. Rational Design of Thermostable Lactate Oxidase by Analyzing Quaternary Structure and Prevention of Deamidation. Biotechnology Letters V27, 1777-1784 (2005).
50. Oberg,B. Rational design of polymerase inhibitors as antiviral drugs. Antiviral Research 71, 90-95 (2006).
51. Pletnev,V.Z., Thomas,J.L., Rhaney,F.L., Holt,L.S., Scaccia,L.A., Umland,T.C., & Duax,W.L. Rational proteomics V: Structure-based mutagenesis has revealed key residues responsible for substrate recognition and catalysis by the dehydrogenase and isomerase activities in human 3[beta]-hydroxysteroid dehydrogenase/isomerase type 1. The Journal of Steroid Biochemistry and Molecular Biology 101, 50-60 (2006).
52. Farinas,E.T., Bulter,T., & Arnold,F.H. Directed enzyme evolution. Current Opinion in Biotechnology 12, 545-551 (2001).
53. Lutz,S. & Patrick,W.M. Novel methods for directed evolution of enzymes: quality, not quantity. Curr Opin Biotech 15, 291-297 (2004).
54. Arnold,F.H. & Moore,J.C. Optimizing industrial enzymes by directed evolution. Adv Biochem Eng Biotechnol 58, 1-14 (1997).
55. Johannes,T.W. & Zhao,H. Directed evolution of enzymes and biosynthetic pathways. Curr Opin Microbiol 9, 261-267 (2006).
56. Wong,T.S., Zhurina,D., & Schwaneberg,U. The diversity challenge in directed protein evolution. Comb Chem High Throughput Screen 9, 271-288 (2006).
57. Leung,D.W., Chen,E., & Goeddel,D.V. A method for random mutagenesis of a defined DNA segment using a modified polymerase chain reaction. Technique 1, 11-15 (1989).
58. Kaur,J. & Sharma,R. Directed evolution: an approach to engineer enzymes. Crit Rev Biotechnol 26, 165-199 (2006).
59. Reetz,M.T. in Advances in Catalysis, Edn. Volume 49. ed. Bruce,C.G.a.H. 1-69 (Academic Press,2006).
60. Stemmer,W.P. Rapid evolution of a protein in vitro by DNA shuffling. Nature 370, 389-391 (1994).
62. Taylor,S.V., Kast,P., & Hilvert,D. Investigating and engineering enzymes by genetic selection. Angew Chem Int Ed Engl 40, 3310-3335 (2001).
63. Matsuura,T. & Yomo,T. In vitro evolution of proteins. Journal of Bioscience and Bioengineering 101, 449-456 (2006).
64. Rubin-Pitel,S.B. & Zhao,H. Recent advances in biocatalysis by directed enzyme evolution. Comb Chem High Throughput Screen 9, 247-257 (2006).
REFERENCES
148
65. Williams,G.J., Nelson,A.S., & Berry,A. Directed evolution of enzymes for biocatalysis and the life sciences. Cell Mol Life Sci 61, 3034-3046 (2004).
66. Wang,T.W., Zhu,H., Ma,X.Y., Zhang,T., Ma,Y.S., & Wei,D.Z. Mutant library construction in directed molecular evolution: casting a wider net. Mol Biotechnol 34, 55-68 (2006).
67. Evolutionary methods in biotechnology - Clever tricks for directed evolution. (Wiley-VCH Verlag, Weinheim, Germany; 2004).
68. Sylvestre,J., Chautard,H., Cedrone,F., & Delcourt,M. Directed evolution of biocatalysts. Org Proc Res Develop 10, 562-571 (2006).
69. Aharoni,A., Griffiths,A.D., & Tawfik,D.S. High-throughput screens and selections of enzyme-encoding genes. Curr Opin Chem Biol 9, 210-216 (2005).
70. Delagrave,S. & Murphy,D.J. In vitro evolution of proteins for drug development. Assay Drug Develop Technol 1, 187-198 (2003).
71. Leemhuis,H., Stein,V., Griffiths,A.D., & Hollfelder,F. New genotype–phenotype linkages for directed evolution of functional proteins. Curr Opin Struc Biol 15, 472-478 (2005).
73. Farinas,E.T. Fluorescence activated cell sorting for enzymatic activity. Comb Chem High Throughput Screen 9, 321-328 (2006).
74. Griffiths,J.S., Cheriyan,M., Corbell,J.B., Pocivavsek,L., Fierke,C.A., & Toone,E.J. A bacterial selection for the directed evolution of pyruvate aldolases. Bioorg Med Chem Letters 12, 4067-4074 (2004).
75. Firestine,S.M., Salinas,F., Nixon,A.E., Baker,S.J., & Benkovic,S.J. Using an AraC-based three-hybrid system to detect biocatalysts in vivo. Nat Biotechnol 18, 544-547 (2000).
76. Amstutz,P., Forrer,P., Zahnd,C., & Plückthun,A. In vitro display technologies: novel developments and applications. Curr Opin Biotechnol 12, 400-405 (2001).
77. Dower,W.J. & Mattheakis,L.C. In vitro selection as a powerful tool for the applied evolution of proteins and peptides. Curr Opin Chem Biol 6, 390-398 (2002).
81. Evnin,L.B., Vasquez,J.R., & Craik,C.S. Substrate specificity of trypsin investigated by using a genetic selection. Proc Natl Acad Sci 87, 6659-6663 (1990).
82. Henning,H., Leggewie,C., Pohl,M., Muller,M., Eggert,T., & Jaeger,K.E. Identification of novel benzoylformate decarboxylases by growth selection. Appl Environ Microbiol 72, 7510-7517 (2006).
83. McLoughlin,S.Y., Jackson,C., Liu,J.W., & Ollis,D.L. Growth of Escherichia coli coexpressing phosphotriesterase and glycerophosphodiester phosphodiesterase, using paraoxon as the sole phosphorus source. Appl Environ Microbiol 70, 404-412 (2004).
84. van der Veen,B.A., Potocki-Veronese,G., Albenne,C., Joucla,G., Monsan,P., & Remaud-Simeon,M. Combinatorial engineering to enhance amylosucrase performance: construction, selection, and screening of variant libraries for increased activity. FEBS Lett 560, 91-97 (2004).
85. Yano,T., Oue,S., & Kagamiyama,H. Directed evolution of an aspartate aminotransferase with new substrate specificities. Proc Natl Acad Sci 95, 5511-5515 (1998).
86. Parikh,M.R., Greene,D.N., Woods,K.K., & Matsumura,I. Directed evolution of RuBisCO hypermorphs through genetic selection in engineered E. coli. Protein Eng Des Sel 19, 113-119 (2006).
92. Otten,L.G., Sio,C.F., Vrielink,J., Cool,R.H., & Quax,W.J. Altering the substrate specificity of cephalosporin acylase by directed evolution of the β-subunit. J Biol Chem 277, 42121-42127 (2002).
93. Hwang,B.Y., Oh,J.M., Kim,J., & Kim,B.G. Pro-antibiotic substrates for the identification of enantioselective hydrolases. Biotechnol Lett 28, 1181-1185 (2006).
94. Baker,K., Bleczinski,C., Lin,H., Salazar-Jimenez,G., Sengupta,D., Krane,S., & Cornish,V.W. Chemical complementation: A reaction-independent genetic assay for enzyme catalysis. Proc Natl Acad Sci 99, 16537-16542 (2002).
95. Sengupta,D., Lin,H., Goldberg,S.D., Mahal,J.J., & Cornish,V.W. Correlation between catalytic efficiency and the transcription read-out in chemical complementation: A general assay for enzyme catalysis. Biochem 43, 3570-3581 (2004).
96. Lin,H., Tao,H., & Cornish,V.W. Directed evolution of a glycosynthase via chemical complementation. J Am Chem Soc 126, 15051-15059 (2004).
97. Konarzycka-Bessler,M. & Jaeger,K.E. Select the best: novel biocatalysts for industrial applications. Trends Biotechnol 24, 248-250 (2006).
98. Smith,G.P. Filamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface. Science 228, 1315-1317 (1985).
99. Kehoe,J.W. & Kay,B.K. Filamentous phage display in the new millennium. Chem Rev 105, 4056-4072 (2005).
100. Paschke,M. Phage display systems and their applications. Appl Microbiol Biotechnol 1-10 (2005).
102. Vanwetswinkel,S., Marchand-Brynaert,J., & Fastrez,J. Selection of the most active enzymes from a mixture of phage-displayed β-lactamase mutants. Bioorg Med Chem Lett. 6, 789-792 (1996).
103. Dröge,M.J., Rüggeberg,C.J., van der Sloot,A.M., Schimmel,J., Dijkstra,D.S., Verhaert,R.M.D., Reetz,M.T., & Quax,W.J. Binding of phage displayed Bacillus subtilis lipase A to a phosphonate suicide inhibitor. J Biotechnol 101, 19-28 (2003).
104. Atwell,S. & Wells,J.A. Selection for improved subtiligases by phage display. Proc Natl Acad Sci 96, 9497-9502 (1999).
105. Beliën,T., Hertveldt,K., Brande,K.v.d., Robben,J., Campenhout,S.v., & Volckaert,G. Functional display of family 11 endoxylanases on the surface of phage M13. J Biotechnol 115, 249-260 (2005).
106. Hansson,L.O., Widersten,M., & Mannervik,B. Mechanism based phage display and selection of active-site mutants of human glutathione transferase A1-1 SNAr reactions. Biochem 36, 11252-11260 (1997).
107. Steiner,D., Forrer,P., Stumpp,M.T., & Pluckthun,A. Signal sequences directing cotranslational translocation expand the range of proteins amenable to phage display. Nat Biotechnol 24, 823-831 (2006).
REFERENCES
150
108. Dröge,M.J., Boersma,Y.L., Braun,P.G., Buining,R.J., Julsing,M.K., Selles,K.G., van Dijl,J.M., & Quax,W.J. Phage display of an intracellular carboxylesterase of Bacillus subtilis: A comparison of the Sec and Tat pathway capabilities. Appl Environ Microbiol 72, 4589-4594 (2006).
110. Bai,Y. & Feng,H. Selection of stably folded proteins by phage-display with proteolysis. Eur J Biochem 271, 1609-1614 (2004).
111. Bertschinger,J., Heinis,C., & Neri,D. in Phage display in biotechnology and drug discovery. ed. Sidhu,S.S. 461-491 (Taylor & Francis Group, Boca Raton, Florida, USA; 2005).
112. Fujii,I., Fukuyama,S., Iwabuchi,Y., & Tanimura,R. Evolving catalytic antibodies in a phage-displayed combinatorial library. Nat Biotechnol 16, 463-467 (1998).
113. Fernandez-Gacio,A., Uguen,M., & Fastrez,J. Phage display as a tool for the directed evolution of enzymes. Trends Biotechnol 21, 408-414 (2003).
114. Lin,H. & Cornish,V.W. Screening and selection methods for large-scale analysis of protein function. Angew Chem Int Ed Engl 41, 4402-4425 (2002).
115. Dröge,M.J., Boersma,Y.L., Pouderoyen,G.v., Vrenken,T.E., Rüggeberg,C.J., Reetz,M.T., Dijkstra,B.W., & Quax,W.J. Directed evolution of Bacillus subtilis Lipase A by use of enantiomeric phosphonate inhibitors: crystal structures and phage display selection. Chembiochem 7, 149-157 (2006).
116. Ponsard,I., Galleni,M., Soumillion,P., & Fastrez,J. Selection of metalloenzymes by catalytic activity using phage display and catalytic elution. Chembiochem 2, 253-259 (2001).
117. Strobel,H., Ladant,D., & Jestin,J.L. In vitro selection for enzymatic activity: A model study using adenylate cyclase. J Mol Biol 332, 1-7 (2003).
118. Fa,M., Radeghieri,A., Henry,A.A., & Romesberg,F.E. Expanding the substrate repertoire of a DNA polymerase by directed evolution. J Am Chem Soc 126, 1748-1754 (2004).
119. Holmberg,R.C., Henry,A.A., & Romesberg,F.E. Directed evolution of novel polymerases. Biomol Eng 22, 39-49 (2005).
120. Vichier-Guerre,S., Ferris,S., Auberger,N., Mahiddine,K., & Jestin,J.L. A population of thermostable reverse transcriptases evolved from Thermus aquaticus DNA polymerase I by phage display. Angew Chem Int Ed Engl 45, 1-5 (2006).
121. Routenberg Love,K., Swoboda,J.G., Noren,C.J., & Walker,S. Enabling glycosyltransferase evolution: A facile substrate-attachment strategy for phage-display enzyme evolution. Chembiochem 7, 753-756 (2006).
122. Fujita,S., Taki,T., & Taira,K. Selection of an active enzyme by phage display on the basis of the enzyme's catalytic activity in vivo. Chembiochem 6, 315-321 (2005).
123. Taki,T., Fujita,S., & Taira,K. Selection of a catalytically active enzyme in vivo via phage display of its product. Lett Drug Des Discov 2, 475-478 (2005).
128. Benhar,I. Biotechnological applications of phage and cell display. Biotechnol Adv 19, 1-33 (2001).
REFERENCES
151
129. Narita,J., Okano,K., Tateno,T., Tanino,T., Sewaki,T., Sung,M.-H., Fukuda,H., & Kondo,A. Display of active enzymes on the cell surface of Escherichia coli using PgsA anchor protein and their application to bioconversion. Appl Microbiol Biotechnol 70, 564-572 (2006).
131. Lee,S.H., Choi,J., Park,S.J., Lee,S.Y., & Park,B.C. Display of bacterial lipase on the Escherichia coli cell surface by using FadL as an anchoring motif and use of the enzyme in enantioselective biocatalysis. Appl Environ Microbiol 70, 5074-5080 (2004).
132. Maurer,J., Jose,J., & Meyer,T.F. Autodisplay: one-component system for efficient surface display and release of soluble recombinant proteins from Escherichia coli. J Bacteriol 179, 794-804 (1997).
133. Schultheiss,E., Paar,C., Schwab,H., & Jose,J. Functional esterase surface display by the autotransporter pathway in Escherichia coli. J Mol Catal B: Enzymatic 18, 89-97 (2002).
134. Becker,S., Theile,S., Heppeler,N., Michalczyka,A., Wentzela,A., Wilhelm,S., Jaeger,K.-E., & Kolmara,H. A generic system for the Escherichia coli cell-surface display of lipolytic enzymes. FEBS Lett 579, 1177-1182 (2005).
135. Wilhelm,S., Rosenau,F., Becker,S., Buest,S., Hausmann,S., Kolmar,H., & Jaeger,K.-E. Functional cell-surface display of a lipase-specific chaperone. Chembiochem 8, 55-60 (2007).
136. Becker,S., Schmoldt,H.U., Adams,T.M., Wilhelm,S., & Kolmar,H. Ultra-high-throughput screening based on cell-surface display and fluorescence-activated cell sorting for the identification of novel biocatalysts. Curr Opin Biotechnol 15, 323-329 (2004).
137. Bosma,T., Kanninga,R., Neef,J., Audouy,S.A.L., van Roosmalen,M.L., Steen,A., Buist,G., Kok,J., Kuipers,O.P., Robillard,G., & Leenhouts,K. Novel surface display system for proteins on non-genetically modified Gram-positive bacteria. Appl Environ Microbiol 72, 880-889 (2006).
139. Amstutz,P., Pelletier,J.N., Guggisberg,A., Jermutus,L., Cesaro-Tadic,S., Zahnd,C., & Pluckthun,A. In vitro selection for catalytic activity with ribosome display. J Am Chem Soc 124, 9396-9403 (2002).
140. Takahashi,F., Ebihara,T., Mie,M., Yanagida,Y., Endo,Y., Kobatake,E., & Aizawa,M. Ribosome display for selection of active dihydrofolate reductase mutants using immobilized methotrexate on agarose beads. FEBS Lett 514, 106-110 (2002).
141. Takahashi,F., Funabashi,H., Mie,M., Endo,Y., Sawasaki,T., Aizawa,M., & Kobatake,E. Activity-based in vitro selection of T4 DNA ligase. Biochem Biophys Res Comm 336, 987-993 (2005).
142. Fiammengo,R. & Jaschke,A. Nucleic acid enzymes. Current Opinion in Biotechnology 16, 614-621 (2005).
143. Aharoni,A., Amitai,G., Bernath,K., Magdassi,S., & Tawfik,D.S. High-throughput screening of enzyme libraries: Thiolactonases evolved by fluorescence-activated sorting of single cells in emulsion compartments. Chem Biol 12, 1281-1289 (2005).
144. Cohen,H.M., Tawfik,D.S., & Griffiths,A.D. Altering the sequence specificity of HaeIII methyltransferase by directed evolution using in vitro compartmentalization. Protein Eng Des Sel 17, 3-11 (2004).
146. Lee,Y.F., Tawfik,D.S., & Griffiths,A.D. Investigating the target recognition of DNA cytosine-5 methyltransferase HhaI by library selection using in vitro compartmentalisation. Nucl Acids Res 30, 4937-4944 (2002).
147. Griffiths,A.D. & Tawfik,D.S. Directed evolution of an extremely fast phosphotriesterase by in vitro compartmentalization. EMBO J 22, 24-35 (2003).
REFERENCES
152
148. Bernath,K., Hai,M., Mastrobattista,E., Griffiths,A.D., Magdassi,S., & Tawfik,D.S. In vitro compartmentalization by double emulsions: sorting and gene enrichment by fluorescence activated cell sorting. Anal Biochem 325, 151-157 (2004).
149. Mastrobattista,E., Taly,V., Chanudet,E., Treacy,P., Kelly,B.T., & Griffiths,A.D. High-Throughput screening of enzyme libraries: In vitro evolution of a [beta]-galactosidase by fluorescence-activated sorting of double emulsions. Chem Biol 12, 1291-1300 (2005).
150. Kim,J.Y., Choi,G.S., Jung,I.S., Ryu,Y.W., & Kim,G.J. A systematic approach for yielding a potential pool of enzymes: practical case for chiral resolution of (R,S)-ketoprofen ethyl ester. Protein Engineering, Design and Selection 16, 357-364 (2003).
151. Krebsfanger,N., Schierholz,K., & Bornscheuer,U.T. Enantioselectivity of a recombinant esterase from Pseudomonas fluorescens towards alcohols and carboxylic acids. Journal of Biotechnology 60, 105-111 (1998).
152. Petersen,E.I., Valinger,G., Solkner,B., Stubenrauch,G., & Schwab,H. A novel esterase from Burkholderia gladioli which shows high deacetylation activity on cephalosporins is related to [beta]-lactamases and -peptidases. Journal of Biotechnology 89, 11-25 (2001).
153. Talker-Huiber,D., Jose,J., Glieder,A., Pressnig,M., Stubenrauch,G., & Schwab,H. Esterase EstE from Xanthomonas vesicatoria (Xv_EstE) is an outer membrane protein capable of hydrolyzing long-chain polar esters. Applied Microbiology and Biotechnology 61, 479-487 (2003).
154. Cohen,N., Abramov,S., Dror,Y., & Freeman,A. In vitro enzyme evolution: the screening challenge of isolating the one in a million. Trends in Biotechnology 19, 507-510 (2001).
155. Jaeger,K.-E., Eggert,T., Eipper,A., & Reetz,M.T. Directed evolution and the creation of enantioselective biocatalysts. Appl Microbiol Biotechnol 55, 519-530 (2001).
157. Reetz,M.T. Controlling the enantioselectivity of enzymes by directed evolution: Practical and theoretical ramifications. Proc Nat Acad Sci 101, 5716-5722 (2004).
158. Reetz,M.T. & Jaeger,K.-E. Enantioselective enzymes for organic synthesis created by directed evolution. Chem Eur J 6, 407-412 (2000).
159. Reetz,M.T., Torre,C., Eipper,A., Lohmer,R., Hermes,M., Brunner,B., Maichele,A., Bocola,M., Arand,M., Cronin,A., Genzel,Y., Archelas,A., & Furstoss,R. Enhancing the enantioselectivity of an epoxide hydrolase by directed evolution. Org Lett 6, 177-180 (2004).
160. Pelletier,J. & Sidhu,S. Mapping protein-protein interactions with combinatorial biology methods. Current Opinion in Biotechnology 12, 340-347 (2001).
161. Henke,E. & Bornscheuer,U. Esterases from Bacillus subtilis and B. stearothermophilus share high sequence homology but differ substantially in their properties. Applied Microbiology and Biotechnology 60, 320-326 (2002).
163. Bischoff,D.S. & Ordal,G.W. Sequence and characterization of Bacillus subtilis CheB, a homolog of Escherichia coli CheY, and its role in a different mechanism of chemotaxis. J. Biol. Chem. 266, 12301-12305 (1991).
164. Chen,Y.R., Usui,S., Queener,S.W., & Yu,C.A. Purification and properties of ap-nitrobenzyl esterase from Bacillus subtilis. Journal of Industrial Microbiology and Biotechnology 15, 10-18 (1995).
165. Eder,S., Shi,L., Jensen,K., Yamane,K., & Hulett,F.M. A Bacillus subtilis secreted phosphodiesterase/alkaline phosphatase is the product of a Pho regulon gene, phoD. Microbiology 142, 2041-2047 (1996).
166. Higerd,T.B. & Spizizen,J. Isolation of Two Acetyl Esterases from Extracts of Bacillus subtilis. J. Bacteriol. 114, 1184-1192 (1973).
REFERENCES
153
167. Kneusel,R.E., Schiltz,E., & Matern,U. Molecular characterization and cloning of an esterase which inactivates the macrolide toxin brefeldin A. J. Biol. Chem. 269, 3449-3456 (1994).
168. Moore,J.C. & Arnold,F.H. Directed evolution of a para-nitrobenzyl esterase for aqueous-organic solvents. Nat Biotechnol 14, 458-467 (1996).
169. Nilsson,R.P., Beijer,L., & Rutberg,B. The glpT and glpQ genes of the glycerol regulation in Bacillus subtilis. Microbiology 140, 723-730 (1994).
170. Riefler,J.F.3. & Higerd,T.B. Characterization of intracellular esterase A from Bacillus subtilis. Biochim Biophys Acta 429, 191-197 (1976).
171. Kunst,F. & Rapoport,G. Salt stress is an environmental signal affecting degradative enzyme synthesis in Bacillus subtilis. J. Bacteriol. 177, 2403-2407 (1995).
172. Dartois,V., Baulard,A., Schanck,K., & Colson,C. Cloning, nucleotide sequence and expression in Escherichia coli of a lipase gene from Bacillus subtilis 168. Biochim. Biophys. Acta 1131, 253-260 (1992).
173. Norrander,J., Kempe,T., & Messing,J. Construction of improved M13 vectors using oligodeoxynucleotide-directed mutagenesis. Gene 26, 101-106 (1983).
174. Bruckner,R., Zyprian,E., & Matzura,H. Expression of a chloramphenicol-resistance determinant carried on hybrid plasmids in Gram-positive and Gram-negative bacteria. Gene 32, 151-160 (1984).
175. Zyprian,E. & Matzura,H. Characterization of signals promoting gene expression on the Staphylococcus aureus plasmid pUB110 and development of a gram-positive expression vector system. DNA 5, 219-225 (1986).
176. Sambrook,J., Fritsch,E.F., & Maniatis,T. Molecular cloning: A laboratory manual. (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, USA; 1989).
177. Birnboim,H. & Doly,J. A rapid alkaline extraction procedure for screening recmbinant plasmid DNA. Nucleic Acids Res. 7, 1513-1523 (1979).
178. Laemmli,U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685 (1970).
179. Mutsaers,J.H.G.M. & Kooreman,H.J. Preparation of optically pure 2-aryl- and 2-aryloxy-propionates by selective enzymatic hydrolysis. Recl. Trav. Chim. Pays-Bas 110, 185-188 (1991).
180. Dröge,M.J., Bos,R., Woerdenbag,H.J., & Quax,W.J. Chiral gas chromatography for the determination of 1,2-O-isopropylidene-sn-glycerol stereoisomers. J Sep Sci 26, 1-6 (2003).
181. Chen,C.S., Fujimoto,Y., Girdaukas,G., & Sih,C.J. Quantitative analyses of biochemical kinetic resolutions of enantiomers. J. Am. Chem. Soc. 109, 7294-7299 (1982).
182. Chen,C.S., Wu,S.H., Girdaukas,G., & Sih,C.J. Quantitative analyses of biochemical kinetic resolution of enantiomers. 2. Enzyme-catalyzed esterifications in water-organic solvent biphasic systems. J Am Chem Soc 109, 2812-2817 (1987).
183. Choi,G.S., Kim,J.Y., Kim,J.H., Ryu,Y.W., & Kim,G.J. Construction and characterization of a recombinant esterase with high activity and enantioselectivity to (S)-ketoprofen ethyl ester. Protein Expression and Purification 29, 85-93 (2003).
184. Jackson,M.A., Labeda,D.P., & Becker,L.A. Enantioselective hydrolysis of ethyl 2-hydroxyalkanoates by an extracellular esterase from a Bacillus sphaericus strain. Enzyme and Microbial Technology 17, 175-179 (1995).
185. Simoes,D.d., McNeill,D., Kristiansen,B., & Mattey,M. Purification and partial characterisation of a 1.57 kDa thermostable esterase from Bacillus stearothermophilus. FEMS Microbiology Letters 147, 151-156 (1997).
190. Smith,G.P. Filamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface. Science 228, 1315-1317 (1985).
191. Rapoza,M.P. & Webster,R.E. The filamentous bacteriophage assembly proteins require the bacterial Sec protein for correct localization to the membrane. J. Bacteriol. 175, 1856-1859 (1993).
192. Hansson,L.O., Widersten,M., & Mannervik,B. Mechanism based phage display and selection of active-site mutants of human glutathione transferase A1-1 SNAr reactions. Biochem. 36, 11252-11260 (1997).
193. Manting,E.H. & Driessen,A.J. Escherichia coli translocase: the unravelling of a molecular machine. Mol. Microbiol. 37, 226-238 (2000).
194. Stephenson,K. Sec-dependent protein translocation across biological membranes: evolutionary conservation of an essential protein transport pathway. Mol. Membr. Biol. 22, 17-28 (2005).
195. DeLisa,M.P., Tullman,D., & Georgiou,G. Folding quality control in the export of proteins by the bacterial twin-arginine translocation pathway. Proc. Natl. Acad. Sci. USA 100, 6115-6120 (2003).
196. Muller,M. Twin-arginine-specific protein export in Escherichia coli. Research in Microbiology 156, 131-136 (2005).
197. Beha,D., Deitermann,S., Müller,M., & Koch,H.G. Export of beta-lactamase is independent of the signal recognition particle. J. Biol. Chem. 278, 22161-22167 (2003).
198. Bogsch,E.G., Sargent,F., Stanley,N.R., Berks,B.C., Robinson,C., & Palmer,T. An essential component of a novel bacterial protein export system with homologues in plastids and mitochondria. J. Biol. Chem. 273, 18003-18006 (1998).
199. Santini,C.-L., Bernadac,A., Zhang,M., Chanal,A., Ize,B., Blanco,C., & Wu,L.-F. Translocation of jellyfish green fluorescent protein via the Tat system of Escherichia coli and its periplasmic localization in response to osmotic up-shock. J. Biol. Chem. 276, 8159-8164 (2001).
200. Jongbloed,J.D., Antelmann,H., Hecker,M., Nijland,R., Bron,S., Airaksinen,U., Pries,F., Quax,W.J., Dijl,J.M.v., & Braun,P.G. Selective contribution of the twin-arginine translocation pathway to protein secretion in Bacillus subtilis. J. Biol. Chem. 277, 44068-44078 (2002).
201. Stanley,N.R., Palmer,T., & Berks,B.C. The twin arginine consensus motif of Tat signal peptides is involved in Sec-independent protein targeting in Escherichia coli. J. Biol. Chem. 275, 11591-11596 (2000).
202. Verhaert,R.M.D., Duin,J.v., & Quax,W.J. Processing and functional display of the 86 kDa heterodimeric penicillin G acylase on the surface of phage fd. Biochem. J. 342, 415-422 (1999).
203. Norrander,J., Kempe,T., & Messing,J. Construction of improved M13 vectors using oligodeoxynucleotide-directed mutagenesis. Gene 26, 101-106 (1983).
204. Thomas,J.D., Daniel,R.A., Errington,J., & Robinson,C. Export of active green fluorescent protein to the periplasm by the twin-arginine translocase (Tat) pathway in Escherichia coli. Mol. Microbiol. 39, 47-53 (2001).
205. Bonnycastle,L.L.C., Menendez,A., & Scott,J.K. in Phage Display: A Laboratory Manual. eds. Barbas III,C.F., Burton,D.R., Scott,J.K., & Silverman,G.J. 15.17-15.18 (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, N.Y.; 2001).
206. Tesar,M., Beckmann,C., Röttgen,P., Haase,B., Faude,U., & Timmis,K.N. Monoclonal antibody against pIII of filamentous phage: an immunological tool to study fusion protein expression in phage display systems. Immunotechnology 1, 53-64 (1995).
REFERENCES
155
207. Berks,B.C., Palmer,T., & Sargent,F. The Tat protein translocation pathway and its role in microbial physiology. Adv. Microbiol. Physiol. 47, 187-254 (2003).
208. Robinson,C. & Bolhuis,A. Protein targeting by the twin-arginine translocation pathway. Nat. Rev. Mol. Cell Biol. 2, 350-356 (2001).
209. Lopez,J. & Webster,R.E. Assembly site of bacteriophage f1 corresponds to adhesion zones between the innter and outer membranes of the host cell. J. Bacteriol. 163, 1270-1274 (1985).
210. Feng,J., Model,P., & Russel,M. A trans-envelope protein complex needed for filamentous phage assembly and export. Mol. Microbiol. 34, 745-755 (1999).
211. Rakonjac,J., Feng,J., & Model,P. Filamentous phage are released from the bacterial membrane by a two step mechanism involving a short C-terminal fragment of pIII. J. Mol. Biol. 289, 1253-1265 (1999).
212. Hatzixanthis,K., Palmer,T., & Sargent,F. A subset of bacterial inner membrane proteins integrated by the twin-arginine translocase. Mol. Microbiol. 49, 1377-1390 (2003).
213. Matsumae,H., Furui,M., & Shibatani,T. Lipase-catalyzed asymmetric hydrolysis of 3-phenylglycidic acid ester, the key intermediate in the synthesis of diltiazem hydrochloride. Ferment. Bioeng. 75, 93-98 (1993).
223. Arnold,F.H. Combinatorial and computational challenges for biocatalyst design. Nature 409, 253-257 (2001).
224. Bornscheuer,U.T. & Pohl,M. Improved biocatalysts by directed evolution and rational protein design. Current Opinion in Chemical Biology 5, 137-143 (2001).
226. Henke,E., Bornscheuer,U.T., Schmid,R.D., & Pleiss,J. A molecular mechanism of enantiorecognition of tertiary alcohols by carboxylesterases. Chembiochem 4, 485-493 (2003).
228. Kazlauskas,R.J. Molecular modeling and biocatalysis: explanations, predictions, limitations and opportunities. Curr. Opin. Chem. Biol. 4, 81-88 (2000).
229. Reetz,M.T. New methods for the high-throughput screening of enantioselective catalysts and biocatalysts. Angew. Chem. Int. Ed. 41, 1335-1338 (2002).
230. Widersten,M., Hansson,L.O., Tronstad,L., & Mannervik,B. Use of phage display and transition-state analogs to select enzyme variants with altered catalytic properties: glutathione transferase as an example. Methods Enzymol. 328, 389-404 (2000).
231. Demartis,S., Huber,A., Viti,F., Lozzi,L., Giovannoni,L., Neri,P., Winter,G., & Neri,D. A strategy for the isolation of catalytic activities from repertoires of enzymes displayed on phage. J. Mol. Biol. 286, 617-633 (1999).
232. Jestin,J.-L., Kristensen,P., & Winter,G. A method for the selection of catalytic activity using phage display and proximity coupling. Angew. Chem. Int. Ed. 38, 1124-1127 (1999).
233. Soumillion,P., Jespers,L., Bouchet,M., Marchand-Brynaert,J., Winter,G., & Fastrez,J. Selection of beta-lactamase on filamentous bacteriophage by catalytic activity. J. Mol. Biol. 237, 415-422 (1994).
234. Avalle,B., Vanwetswinkel,S., & Fastrez,J. In vitro selection for catalytic turnover from a library of β-lactamase mutants and penicillin-binding proteins. Bioorg. Med. Chem. Lett. 7, 479-484 (1997).
235. Danielsen,S., Eklund,M., Deussen,H.J., Graslund,T., Nygren,P.A., & Borchert,T.V. In vitro selection of enzymatically active lipase variants from phage libraries using a mechanism-based inhibitor. Gene 272, 267-274 (2001).
237. Vanwetswinkel,S., Marchand-Brynaert,J., & Fastrez,J. Selection of the most active enzymes from a mixture of phage-displayed β-lactamase mutants. Bioorg. Med. Chem. Lett. 6, 789-792 (1996).
238. Deussen,H.J., Danielsen,S., Breinholt,J., & Borchert,T.V. Design and synthesis of triglyceride analogue biotinylated suicide inhibitors for directed molecular evolution of lipolytic enzymes. Bioorg. Med. Chem. 10, 2027-2031 (2000).
239. Deussen,H.J., Danielsen,S., Breinholt,J., & Borchert,T.V. A novel biotinylated suicide inhibitor for directed molecular evolution of lipolytic enzymes. Bioorg. Med. Chem. 8, 507-513 (2000).
241. Mannesse,M.L.M., Boots,J.-W.P., Dijkman,R., Slotboom,A.J., Hijden,H.T.W.M.v., Egmond,M.R., Verheij,H.M., & Haas,G.H.d. Phosphonate analogues of triacylglycerols are potent inhibitors of lipase. Biochim. Biophys. Acta 1259, 56-64 (1995).
242. Eipper, A., Eggert, T., Jaeger, K.-E., Pouderoyen, G. van, Dijkstra, B. W., Dröge, M. J., Quax, W. J., and Reetz, M. T. Directed evolution of the enantioselectivity of Bacillus subtilis Lipase A. 2001.
243. Eggert,T., Funke,S.A., Rao,N.M., Acharya,P., Krumm,H., Reetz,M.T., & Jaeger,K.-E. Multiplex-PCR-based recombination as a novel high-fidelity method for directed evolution. Chembiochem 6, 1062-1067 (2005).
244. Lu,K.V., Rohde,M.F., Thomason,A.R., Kenney,W.C., & Lu,H.S. Mistranslation of a TGA termination codon as tryptophan in recombinant platelet-derived growth factor expressed in Escherichia coli. Biochem. J. 15, 411-417 (1995).
245. MacBeath,G. & Kast,P. UGA read-through artefacts - when popular gene expression systems need a pATCH. Biotechniques 24, 789-794 (1998).
246. Bornscheuer,U.T. Trends and challenges in enzyme technology. Biotechnology for the Future 100, 181-203 (2005).
REFERENCES
157
247. Zhao,H., Chockalingam,K., & Chen,Z. Directed evolution of enzymes and pathways for industrial biocatalysis. Curr Opin Biotech 13, 104-110 (2002).
248. Schmidt,M., Hasenpusch,D., Kahler,M., Kirchner,U., Wiggenhorn,K., Langel,W., & Bornscheuer,U.T. Directed Evolution of an Esterase from Pseudomonas fluorescens Yields a Mutant with Excellent Enantioselectivity and Activity for the Kinetic Resolution of a Chiral Building Block. Chembiochem 7, 805-809 (2006).
249. Bornscheuer,U.T., Altenbuchner,J., & Meyer,H.H. Directed evolution of an esterase: screening of enzyme libraries based on pH-indicators and a growth assay. Bioorg Med Chem 7, 2169-2173 (1999).
250. Jestin,J.L. & Kaminski,P.A. Directed enzyme evolution and selections for catalysis based on product formation. J Biotechnol 113, 85-103 (2004).
251. Sio,C.F., Riemens,A.M., Laan,J.M.v.d., Verhaert,R.M.D., & Quax,W.J. Directed evolution of a glutaryl acylase into an adipyl acylase. Eur J Biochem 269, 4495-4504 (2002).
252. McLoughlin,S.Y., Jackson,C., Liu,J.W., & Ollis,D.L. Increased expression of a bacterial phosphotriesterase in Escherichia coli through directed evolution. Prot Expres Purif 41, 433-440 (2005).
254. Pahel,G., Zelenetz,A.D., & Tyler,B.M. gltB Gene and Regulation of Nitrogen Metabolism by Glutamine Synthetase in Escherichia coli. J Bacteriol 133, 139-148 (1978).
255. Guex,N. & Peitsch,M.C. SWISS-MODEL and the Swiss-PdbViewer: An environment for comparative protein modeling. Electrophoresis 18, 2714-2723 (1997).
256. Ilyin,V.A., Abyzov,A., & Leslin,C.M. Structural alignment of proteins by a novel TOPOFIT method, as a superimposition of common volumes at a topomax point. Protein Sci 13, 1865-1874 (2004).
258. Reetz,M.T. & Wang,L.W. High-throughput selection system for assessing the activity of epoxide hydrolases. Comb Chem High Throughput Screen 9, 295-299 (2006).
259. Schrag,J.D., Winkler,F.K., & Cygler,M. Pancreatic lipases: evolutionary intermediates in a positional change of catalytic carboxylates? J. Biol. Chem. 267, 4300-4303 (1992).
260. Todd,A.E., Orengo,C.A., & Thornton,J.M. Evolution of function in protein superfamilies, from a structural perspective. Journal of Molecular Biology 307, 1113-1143 (2001).
261. Schrag,J.D., Vernet,T., Laramée,L., Thomas,D.Y., Recktenwald,A., Okoniewska,M., Zionek,E., & Cygler,M. Redesigning the active site of Geotrichum candidum lipase. Protein Eng 8, 835-842 (1995).
262. Krooshof,G.H., Kwant,E.M., Damborsky,J., Koca,J., & Janssen,D.B. Repositioning the Catalytic Triad Aspartic Acid of Haloalkane Dehalogenase: Effects on Stability, Kinetics, and Structure. Biochemistry 36, 9571-9580 (1997).
263. Blaser, H.-U., Pugin, B., and Spindler, F. Progress in enantioselective catalysis assessed from an industrial point of view. J Mol Catal A: Chemical . 2005.
264. Beisson,F., Tiss,A., Rivière,C., & Verger,R. Methods for lipase detection and assay: a critical review. Eur J Lipid Sci Technol 102, 133-153 (2000).
265. Eggert,T., Leggewie,C., Puls,M., Streit,W., Van Pouderoyen,G., Dijkstra,B.W., & Jaeger,K.-E. Novel Biocatalysts by Identification and Design. Biocatal Biotransform 22, 139-144 (2004).
266. Secundo,F., Carrea,G., Tarabiono,C., Brocca,S., & Lotti,M. Activity and enantioselectivity of wildtype and lid mutated Candida rugosa lipase isoform 1 in organic solvents. Biotechnol bioeng 86, 236-240 (2005).
REFERENCES
158
267. Longhi,S., MANNESSE,M., VERHEIJ,H.M., DE HAAS,G.H., EGMOND,M., KNOOPS-MOUTHUY,E., & Cambillau,C. Crystal structure of cutinase covalently inhibited by a triglyceride analogue. Protein Sci 6, 275-286 (1997).
268. Longhi,S., Czjzek,M., Lamzin,V., Nicolas,A., & Cambillau,C. Atomic resolution (1.0 A) crystal structure of Fusarium solani cutinase: stereochemical analysis. Journal of Molecular Biology 268, 779-799 (1997).
269. Martinez,C., Geus,P.d., Lauwereys,M., Matthyssens,M., & Cambillau,C. Fusarium solani cutinase is a lipolytic enzyme with a catalytic serine accessible to solvent. Nature 356, 615-618 (1992).
270. Ghosh,D., Erman,M., Sawicki,M., Lala,P., Weeks,D.R., Li,N., Pangborn,W., Thiel,D.J., Jörnvall,H., Gutierrez,R., & Eyzaguirre,J. Determination of a protein structure by iodination: the structure of iodinated acetylxylan esterase. Acta Crystallog Sect D 55, 779-784 (1999).
271. Holm,L. & Sander,C. Protein Structure Comparison by Alignment of Distance Matrices. Journal of Molecular Biology 233, 123-138 (1993).
272. Akoh,C.C., Lee,G.-C., & Shaw,J.-F. Protein engineering and applications of Candida rugosa lipase isoforms. Lipids 39, 513-526 (2004).
273. Brocca,S., Secundo,F., Ossola,M., Alberghina,L., Carrea,G., & Lotti,M. Sequence of the lid affects activity and specificity of Candida rugosa lipase isoenzymes. Prot Sci 12, 2312-2319 (2003).
275. Secundo,F., Carrea,G., Tarabiono,C., Brocca,S., & Lotti,M. Activity and enantioselectivity of wildtype and lid mutated Candida rugosa lipase isoform 1 in organic solvents. Biotechnol bioeng 86, 236-240 (2004).
276. Seizaburo,S., Masaji,I., Harukazu,F., Masahiro,N., & Mitsuyoshi,U. Creation of Rhizopus oryzae lipase having a unique oxyanion hole by combinatorial mutagenesis in the lid domain. Applied Microbiology and Biotechnology 68, 779-785 (2005).
277. Santarossa,G., Lafranconi,P.G., Alquati,C., DeGioia,L., Alberghina,L., Fantucci,P., & Lotti,M. Mutations in the "lid" region affect chain length specificity and thermostability of a Pseudomonas fragi lipase. FEBS letters 579, 2383-2386 (2005).
278. Holmquist,M. & Berglund,P. Creation of a synthetically useful lipase with higher than wild-type enantioselectivity and maintained catalytic activity. Org Lett 1, 763-765 (1999).
279. Miled,N., Bussetta,C., De caro,A., Riviere,M., Berti,L., & Canaan,S. Importance of the lid and cap domains for the catalytic activity of gastric lipases. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology 136, 131-138 (2003).
280. Secundo,F., Carrea,G., Tarabiono,C., Gatti-Lafranconi,P., Brocca,S., Lotti,M., Jaeger,K.E., Puls,M., & Eggert,T. The lid is a structural and functional determinant of lipase activity and selectivity. Journal of Molecular Catalysis B: Enzymatic 39, 166-170 (2006).
281. Yadav,G.D. & Lathi,P.S. Synthesis of citronellol laurate in organic media catalyzed by immobilized lipases: kinetic studies. J Mol Catal B: Enzymatic 27, 113-119 (2004).
282. Wang,Y. & Linko,Y.Y. Lipase-catalyzed enantiomeric synthesis of citronellyl butyrate. J Ferm Bioeng 80, 473-477 (1995).
283. Yamamoto,T., Matsuda,H., Utsumi,Y., Hagiwari,T., & Kanisawa,T. Synthesis and odor of optically active rose oxide. Tetrahedron Lett 43, 9077-9088 (2002).
284. Yamamoto,T., Shimada,A., Ohmoto,T., Matsuda,H., Ogura,M., & Kanisawa,T. Olfactory study on optically active citronellyl derivatives. Flavour Fragr J 19, 121-133 (2004).
287. Gupta,R., Rathi,P., Gupta,N., & Bradoo,S. Lipase assays for conventional and molecular screening: an overview. Biotechnol Appl Biochem 37, 63-71 (2003).
288. Reetz,M.T. Combinatorial and evolution-based methods in the creation of enantioselective catalysts. Angew Chem Int Ed Engl 40 , 248-310 (2001).
289. Hyötyläinen,T. & Riekkola,M.-L. Approaches for on-line coupling of extraction and chromatography. Anal Bioanal Chem 378, 1962-1981 (2004).
290. Smith,R.M. Before the injection - modern methods of sample preparation for separation techniques. J Chromatogr A 1000, 3-27 (2003).
291. Furniss,B.S., Hannaford,A.J., Smith,P.W.G., & Tatchell,A.R. Vogel's textbook of practical organic chemistry. (Longman, Essex, United Kingdom; 1989).
292. Hoogkamer,P. & Lammers,N. in Vloeistof chromatografie. eds. T.Remmelts & C.Lipschitz (tenHagenStam Uitgevers, Den Haag, The Netherlands; 2001).
293. Demyttenaere,J.C.R., Vanoverschelde,J., & Kimpe,N.d. Biotransformation of (R)-(+)- and (S)-(-)-citronellol by Aspergillus sp. and Penicillium sp., and the use of solid-phase microextraction for screening. J Chromatogr A 1027, 137-146 (2004).
294. Luxova,A., Urbanova,K., Velterova,I., Terzo,M., & Borg-Karlson,A.K. Absolute configuration of chiral terpenes in marking pheromones of bumblebees and cuckoo bumblebees. Chirality 16, 228-233 (2004).
295. Larios,A., Garcia,H.S., Oliart,R.M., & Valerio-Alfaro,G. Synthesis of flavor and fragrance esters using Candida antarctica lipase. Appl Microbiol Biotechnol 65, 373-376 (2004).
296. Dosanjh,N.S. & Kaur,J. Immobilization, stability and esterification studies of a lipase from a Bacillus sp. Biotechnol Appl Biochem 36, 7-12 (2002).
298. Reetz,M.T., Wang,L.W., & Bocola,M. Directed Evolution of Enantioselective Enzymes: Iterative Cycles of CASTing for Probing Protein-Sequence Space. Angew Chem Int Ed Engl 45, 2494 (2006).
299. Zhao,H., Giver,L., Shao,Z., Affholter,J.A., & Arnold,F.H. Molecular evolution by staggered extension process (StEP) in vitro recombination. Nat Biotechnol 16, 258-261 (1998).
300. Wang,L., Xie,J., & Schultz,P.G. Expanding the genetic code. Annual Review of Biophysics and Biomolecular Structure 35, 225-249 (2006).
301. Jackson,J.C., Duffy,S.P., Hess,K.R., & Mehl,R.A. Improving Nature's Enzyme Active Site with Genetically Encoded Unnatural Amino Acids. J. Am. Chem. Soc. 128, 11124-11127 (2006).
302. Boersma,Y.L., Scheltinga,P.C., Dröge,M.J., Bos,R., & Quax,W.J. A validated gas chromatographic method for the evaluation of enzymatic enantioselectivity in kinetic resolution applications. J Sep Sci 28, 501-505 (2005).
303. Griffiths,A.D. & Tawfik,D.S. Miniaturising the laboratory in emulsion droplets. Trends in Biotechnology 24, 395-402 (2006).
NEDERLANDSE SAMENVATTING
NEDERLANDSE SAMENVATTING
163
Evolutie van Enantioselectiviteit:
Selectie van verbeterde Hydrolase Varianten
Het belang van chiraliteit bij de ontwikkeling van nieuwe geneesmiddelen wordt steeds beter onderkend. Met dit scheikundige begrip worden verbindingen aangeduid die dezelfde chemische structuurformule hebben, maar driedimensionaal elkaars spiegelbeeld zijn. Deze spiegelbeelden, zogenaamde enantiomeren, bezitten dezelfde fysisch-chemische eigenschappen. Ze verschillen echter in de wijze waarop gepolariseerd licht wordt gedraaid: de ene enantiomeer draait het licht linksom, de andere rechtsom. Enantiomeren van een geneesmiddel gaan in het lichaam interacties aan met chirale biomoleculen, zoals receptoren en enzymen; verschillende enantiomeren van een geneesmiddel zullen dan ook een verschillende interactie aangaan met het aangrijpingspunt in het lichaam, en derhalve een verschillend werkingsprofiel hebben. Zo kan de ene enantiomeer een therapeutisch effect bewerkstelligen, terwijl de andere enantiomeer inactief is, of nog erger bijwerkingen veroorzaakt. Een tragisch voorbeeld van dit laatste is thalidomide (Softenon®), een slaapmiddel dat voorgeschreven werd aan zwangere vrouwen eind jaren ’60 van de vorige eeuw. Dit geneesmiddel werd toegediend als een mengsel van beide enantiomeren, een zogenaamd racemisch mengsel. De R-enantiomeer van thalidomide was verantwoordelijk voor het rustgevende effect, terwijl de S-enantiomeer allesbehalve inactief was: deze veroorzaakte de bekende teratogene effecten. Sindsdien zijn geneesmiddelfabrikanten verplicht om beide enantiomeren in een racemisch mengsel op hun werkingsprofiel te onderzoeken. Steeds vaker worden enantiozuivere geneesmiddelen, die alleen de werkzame enantiomeer bevatten, op de markt gebracht. Veelal worden de enantiomeren op chemische wijze van elkaar gescheiden, maar dit is voor lang niet alle verbindingen mogelijk. Bovendien is een chemische scheiding relatief duur en milieu-onvriendelijk. Biokatalyse, het gebruik van enzymen, vormt een milieuvriendelijk en goedkoop alternatief voor de chemische scheiding. Voor dit proces zijn geschikte enzymen essentieel: dezen zijn in staat om specifiek één enantiomeer te produceren en daarnaast kunnen ze de reactie ook nog eens versnellen. De meest toegepaste enzymen in de industriële biokatalyse zijn lipases en esterases, enzymen die esterverbindingen zowel vormen als hydrolyseren. Ze hebben gunstige eigenschappen met betrekking tot enantioselectiviteit, regioselectiviteit en accepteren een groot scala aan substraten. In dit onderzoek lag de focus op het gebruik van hydrolytische enzymen uit het micro-organisme Bacillus subtilis 168 om enantiomeren uit een racemisch mengsel van elkaar te scheiden (Hoofdstuk 1). Zowel carboxylesterase A (CesA) en carboxylesterase B (CesB) als lipase A (LipA) werden toegepast voor de enantiozuivere productie van S-(+)-1,2-O-isopropylideen-sn-glycerol, kortweg IPG, uit een racemisch mengsel van IPG esters. IPG is een bouwsteen van onder andere chirale β-blokkers. Deze klasse van geneesmiddelen wordt toegepast bij hart- en vaatziekten.
De productie van S-(+)-IPG door de bovengenoemde enzymen is echter nog niet optimaal. De activiteit, substraatspecificiteit en enantioselectiviteit dienen verbeterd te worden. Hiervoor kunnen twee strategieën toegepast worden: zogenaamde “directed evolution”, evolutie in een reageerbuis, en “rational design”, het ontwerpen van gunstige mutaties in een molecuul waarvan de driedimensionale structuur bekend is.
NEDERLANDSE SAMENVATTING
164
Met behulp van directed evolution kunnen de eigenschappen van een enzym gemakkelijk aangepast worden zonder dat er voorkennis van de structuur van het enzym nodig is. De techniek beslaat twee stappen: eerst worden mutaties in het gen dat codeert voor het enzym aangebracht; vervolgens worden varianten met de gewenste eigenschappen geselecteerd en geïdentificeerd. Deze cyclus kan een aantal maal herhaald worden totdat de beste mutant gevonden is.
De tweede stap, het selecteren van verbeterde mutanten, blijft een kritische stap. Er wordt gebruik gemaakt van condities waarbij alleen de gewenste, verbeterde varianten geïsoleerd worden; er is sprake van ‘survival of the fittest’. Minder interessante varianten zullen in dat geval niet geselecteerd worden. Omdat enzymen niet in staat zijn zichzelf te vermenigvuldigen, moet er tegelijkertijd ook geselecteerd worden voor de genen die voor een specifiek enzym coderen. Er moet altijd een fysieke link zijn tussen het coderende gen, het enzym, en het product van de enzymatische activiteit. Verschillende selectietechnieken, zoals groeiselectie, faag display, ribosoom display en in vitro compartmentalisatie maken deze link mogelijk. Deze methoden worden besproken in Hoofdstuk 2. De mogelijke toepassing van de carboxylesterases CesA en CesB in een industrieel proces wordt besproken in Hoofdstuk 3. Deze enzymen bleken sterk homoloog te zijn aan een ander esterase, carboxylesterase NP uit B. subtilis Thai I-8. Dit enzym is een efficiënte, zeer enantioselectieve biokatalysator in de productie van niet-steroïde, anti-inflammatoire geneesmiddelen (ontstekingsremmende pijnstillers, NSAIDs) zoals ibuprofen en naproxen. De genen van cesA en cesB werden geïsoleerd en in een expressievector gezet. Na overexpressie en zuivering van deze enzymen werden de biochemische eigenschappen onderzocht. Hun substraatspecificiteit en enantioselectiviteit werden voor een groot aantal substraten vergeleken met het carboxylesterase NP. Opvallend was dat, hoewel de drie carboxylesterases dezelfde biochemische eigenschappen vertoonden, de substraatspecificiteit onderling sterk verschilde. CesA en carboxylesterase NP zijn zeer geschikt voor de enantiozuivere productie van chirale carboxylzuren zoals naproxen en ibuprofen. CesB daarentegen lijkt juist meer geschikt voor de enantiozuivere productie van chirale alcoholen, zoals IPG. CesA lijkt minder geschikt voor de enantiozuivere productie van IPG. Om de substraatspecificiteit van CesA te vergroten, kan directed evolution toegepast worden. Voor de tweede stap in directed evolution, selectie van een verbeterd enzym, is een faag display selectiemethode opgezet (Hoofdstuk 4). Een faag, een zogenaamd bacterieel virus, kan met zijn eigen circulair DNA een bacteriële gastheer infecteren. Het faag DNA bevat de genetisch informatie voor verschillende virale manteleiwitten, die na infectie van de bacteriële gastheer in grote hoeveelheden geproduceerd worden. Uiteindelijk komen de nieuw geproduceerde fagen vrij uit de bacterie. De faag display techniek maakt gebruik van één van de manteleiwitten van de faag (in dit geval het eiwit g3p): het DNA coderend voor een enzymvariant wordt ingebracht in het DNA coderend voor het manteleiwit. Dit resulteert in de presentatie van de enzymvariant op het faagoppervlak, als fusie met het g3p manteleiwit. In deze methode wordt de fysieke link tussen genotype en fenotype tijdens selectie behouden: de enzymvariant (fenotype) is op het faagoppervlak aanwezig, terwijl in de faag het DNA (genotype) aanwezig is.
NEDERLANDSE SAMENVATTING
165
CesA is een intracellulair enzym. Om de presentatie van dit enzym op de faag te verbeteren, werd het effect van verschillende signaalpeptiden onderzocht. Deze signaalpeptiden zorgen ervoor dat het enzym via één van de twee bestaande transport routes als fusie-eiwit op het faagoppervlak terecht kan komen. Hieruit bleek dat CesA alleen na transport via de Sec-route als fusie-eiwit op de faagmantel gepresenteerd kon worden. Een volgende stap is het aanbrengen van mutaties in het cesA gen, waardoor ook op eiwitniveau mutaties kunnen optreden. Op die manier ontstaat een ander enzym met eventueel andere eigenschappen, dat met behulp van faag display geselecteerd kan worden. In Hoofdstuk 5 wordt beschreven hoe de faag display methode ook werd toegepast voor het omkeren en verbeteren van de enantioselectiviteit van B. subtilis lipase A (LipA). In een groep aminozuren dichtbij het katalytische centrum van LipA werd een mutantenbank gemaakt. Deze mutantenbank werd als fusie-eiwit met g3p door de faag gepresenteerd. Om een meer enantioselectieve mutant te vinden, werd de interactie van de verschillende mutant-presenterende fagen met IPG bestudeerd. Enantiozuiver IPG werd hiervoor gekoppeld aan een geïmmobiliseerde fosfonaat. De onstane verbinding, IPG fosfonaat, lijkt op een IPG ester; er vindt een nucleofiele aanval van het katalytische serine residu van LipA plaats op de fosfonaat, waarbij serine irreversibel bindt aan de fosfonaat. De fosfonaat inactiveert dus het enzym. Alleen actieve lipasevarianten zullen gebonden worden, en daarmee worden dezen gescheiden van inactieve mutanten.
De geproduceerde fagen werden geïncubeerd met beide enantiomeren van de IPG-fosfonaat remmers. Om te selecteren voor varianten met een omgekeerde enantioselectiviteit, werd een tweeledige selectiemethode opgezet: de mutanten werden eerst geïncubeerd op de fosfonaat gekoppeld aan het ongewenste enantiomeer van IPG. Mutanten met een ongewenste enantioselectiviteit werden op deze manier weggevangen door binding aan de IPG-fosfonaat. Vervolgens werden de ongebonden fagen geïncubeerd op de fosfonaat gekoppeld aan het gewenste IPG-enantiomeer. Met deze methode werden mutanten geselecteerd met een omgekeerde enantioselectiviteit ten opzichte van het wildtype LipA. Hiermee is aangetoond dat faag display gebruikt kan worden om varianten met een verbeterde enantioselectiviteit te selecteren. Deze methode is niet alleen geschikt voor lipase, maar kan ook toegepast worden bij het verbeteren van andere enzymen. Het grote bezwaar van faag display is dat de selectie voornamelijk gebaseerd is op binding van het gepresenteerde enzym aan de remmer. De belangrijkste eigenschap van een enzym is juist dat deze een substraat kan omzetten in een product; daar wordt bij deze faag display selectiemethode geen rekening mee gehouden. Er kan dus een zeer enantioselectieve variant geïsoleerd worden, die niet enzymatisch actief is. Om voor zowel enantioselectiviteit als katalytische activiteit te kunnen selecteren, is een groeiselectiesysteem ontwikkeld. Hoofdstuk 6 beschrijft hoe in dit systeem activiteit en enantioselectiviteit gekoppeld worden aan de overlevingskansen van een bacterie. Er wordt gebruik gemaakt van een aspartaat auxotrofe E. coli stam: deze bacterie is niet in staat om zelf het aminozuur aspartaat te produceren en is daarom voor zijn groei afhankelijk van opname uit het milieu. De E. coli stam werd getransformeerd met een mutantenbank geconstrueerd rond het katalytische centrum van LipA. Aan het medium werd niet aspartaat toegevoegd, maar een ester van aspartaat en het gewenste IPG-enantiomeer. Wil de bacterie kunnen groeien, dan moet deze esterverbinding gehydrolyseerd worden door LipA
NEDERLANDSE SAMENVATTING
166
varianten die tot expressie komen in het periplasma van de cel. Dit is de selectie op activiteit; omdat enantiozuiver IPG gebruikt wordt, vindt er tevens selectie op enantioselectiviteit plaats. Alleen na omzetting van de IPG-aspartaat ester kan de bacterie aspartaat opnemen en groeien. Om de selectiedruk op te voeren, werd een fosfonaat gekoppeld aan de ongewenste enantiomeer van IPG aan het medium toegevoegd. Hierdoor zullen minder actieve varianten met een ongewenste enantioselectiviteit weggevangen worden: er vindt een irreversibele binding plaats van het enzym aan de fosfonaat, waardoor de IPG-aspartaat ester niet meer gehydrolyseerd kan worden. De bacterie zal geen aspartaat kunnen opnemen en gaat dood. De toepasbaarheid van deze methode werd aangetoond door de selectie van mutanten met een omgekeerde enantioselectiviteit ten opzichte van wild type lipase. Eén van de geselecteerde varianten bleek een licht verschoven katalytische aspartaat te hebben. Deze mutant werd verder gekarakteriseerd en er werd gekeken naar de structuur om het enigszins veranderde reactiemechanisme te kunnen verklaren. Deze elegante groeiselectiemethode is niet alleen geschikt voor de selectie van lipase varianten, ook andere enzymen kunnen op vergelijkbare wijze geselecteerd worden zolang ze getransporteerd worden naar het periplasma van de cel. Voor directed evolution is het niet noodzakelijk dat de structuur van een enzym bekend is. Voor rational design is dat wel het geval: aan de hand van de structuur worden zogenaamde “hotspots” aangewezen, gebieden die belangrijk kunnen zijn voor bepaalde eigenschappen van het enzym. Door specifieke mutaties in deze hotspots zou een verbeterd enzym kunnen ontstaan. In Hoofdstuk 7 is rational design op LipA toegepast. Kijkend naar de kristalstructuur van LipA, valt op dat het enzym geen structuuronderdeel heeft dat het katalytische centrum afschermt als een deksel. Dit zou een verklaring kunnen zijn voor de lage enantioselectiviteit van het enzym ten opzichte van bepaalde substraten, zoals IPG. Door het verlengen van loops rond het katalytische centrum van LipA zou een soortgelijke structuur geconstrueerd kunnen worden. Hierbij zou het katalytische centrum enigszins vernauwd worden, wat de enantioselectiviteit wellicht ten goede zou komen. Om dit te bewerkstelligen, werd een loop uit LipA uitgewisseld tegen langere loops uit Fusarium solani cutinase en Penicillium purpurogenum acetylxylaanesterase, twee enzymen die, gezien hun structuur, homoloog zijn aan LipA. De resulterende hybride-enzymen bleken enzymatisch actief en vertoonden een omgekeerde enantioselectiviteit ten opzichte van wildtype LipA. De enantioselectiviteit van de cutinase-hybride werd vervolgens verder verbeterd met behulp van directed evolution. Hiermee werd aangetoond dat de combinatie van rational design en directed evolution succesvol kan zijn in het optimaliseren van de enantioselectiviteit van een enzym. Er worden nieuwe perspectieven in de evolutie van enantioselectiviteit geboden. Tenslotten komt in Hoofdstuk 8 de toepassing van Bacillus LipA in de kinetische resolutie van citronellyl esters aan bod. Citronellol is van groot belang voor de geur- en smaakstofindustrie. Om de activiteit en enantioselectiviteit van mutanten voor dit substraat te bepalen, is een specifieke, gevoelige, en reproduceerbare chirale analysemethode voor de beide enantiomeren van citronellol ontwikkeld en gevalideerd. Wanneer grote aantallen varianten geanalyseerd moeten worden, kan deze methode geautomatiseerd worden.
DANKWOORD
DANKWOORD
169
Op 20 januari 2003, alweer vier jaar geleden, stapte ik op de negende uit de lift voor een sollicitatiegesprek als AiO bij de afdeling Farmaceutische Biologie. Na afloop werd ik door Melloney rondgeleid over de afdeling, waar ik mezelf eigenlijk al zag werken. Twee dagen later hoorde ik dat ik mocht komen. Toch bleef ik twijfelen: kon ik dit wel? Mijn Heit vond: Het is niet nodig te hopen om te ondernemen, noch te slagen om te volharden. Ik ben hem dankbaar voor dat motto. Ik heb ondernomen, ik heb volhard, en dat laatste hoefde ik gelukkig niet alleen te doen! Handig dus dat er in het boekje altijd nog het allerbelangrijkste laatste hoofdstuk is, het hoofdstuk dat als eerste gelezen wordt en wat misschien wel het moeilijkste is om te schrijven (want: vergeet ik niemand?): Het Dankwoord. Wim, op 22 januari belde je om te zeggen dat ik als AiO op jouw afdeling aan de slag mocht. Met heel veel enthousiasme heb ik aan de Little Lipase gewerkt, hoewel dat ook wel opgevat werd als ongeduld... Ik wil je bedanken voor de mogelijkheid om mijn enthousiasme te kunnen botvieren op LipA en vooral ook voor de vrijheid om mijn eigen onderzoek te plannen en uitvoeren. I would like to thank the members of the reading committee, Prof. Dijkstra, Prof. Jaeger, and Prof. Verpoorte for critically reading the manuscript. Het project EvoCatal waar ik op ging werken was een EU-project. I would like to thank all partners from Germany, France and The Netherlands for fruitful discussions and contributions leading to results shown in this thesis. Tjaard, jou wil ik in het bijzonder bedanken voor de plezierige samenwerking. Je leerde me vooral voorzichtig te formuleren. Bedankt voor de figuren die dit boekje mooier maken! Lieve Melloney, je liet een geweldige basis achter voor mijn onderzoek. Toen ik begon, wist ik nog weinig van moleculaire biologie, maar jij hielp me om die achterstand in te halen. Helaas stonden we te kort samen op het lab; of, zoals we nu zeggen: als we nog één jaar extra samen op het lab hadden gestaan, hadden we er nog wel drie artikelen uitgehaald... Ondanks dat je niet meer op de plek naast me in het lab stond, was je nog steeds betrokken bij alles wat ik deed. Ik kan je niet genoeg bedanken voor alle hulp, overleg, suggesties en vooral ook het kritisch doorlezen van alle manuscripten. Het was een eer jouw werk voort te zetten! Little Lipase, uiteraard moet ik ook jou bedanken. Zonder jou geen boekje! Je had het soms zwaar in 957, opboksen tegen het TRAIL homotrimeer... Maar je hebt je kranig geweerd als (bijna) TRAIL monomeer (althans qua grootte)! DAViD, je was een welkome verrassing in een groeiselectie die ik niet heel erg leuk vond, dankjewel! Rein, zittend op het hoekje van je bureau met een kop koffie en even praten was een goed begin van de dag. Bedankt voor je hulp bij de GC, maar met name ook bij de HPLC (waar ik een hekel aan had!) en het runnen van samples! Lab, de beurt is aan jullie. Janita, Oliver, PapaRon, Sieb, Freeke, Torsten, Albert, Elfahmi, Mattijs, Remco, Anna-Margaretha, Oktavia, Mohammed, Charles, Linda, Mariette, Johanna, Michiel, Luís, Marieke, Almer, Robbert, Gudrun, Carlos, Mags, Lidia, Geeske, Mariana, Ilse, Vinod, Jean-Yves, RoDo, Helga, Asia, Agata, Pol, Evelina, Jan Maarten, Gerrit: thank you (as the majority of you is foreign, I will adjust) for your help, your advice,
DANKWOORD
170
but most of all for all the fun we had in the lab! Sieb, bedankt dat ik altijd ‘stiekem’ flessen en erlenmeyers van je mocht pakken, maar vooral bedankt voor alle media en buffers die je altijd maakt, het maakt het werk zoveel gemakkelijker! Linda, o captain my captain. Samen op congres naar het Verre Duitsland vorig jaar was een Ervaring, bedankt voor alle gezelligheid! Veel succes in Delft, ik hoop dat Groep Otten Groep Boersma zal verslaan in grootte! Gerrit, het was goed dat er weer een ‘enzym-man’ op het lab kwam. Bedankt voor alle adviezen en goede gesprekken! Lidia, je deed veel voor het lab, wat ik zeer gewaardeerd heb. Bedankt voor al je praktische tips! Asia, we started at the lab around the same time. I hope your experiments will work out well and I wish you all the best! Luís, you are The Heir, which by the way should make a great movie title I think. Anyway, you started at a time when I was really stressed, I am sorry for giving you a hard time by being a tough cookie (without capital C of course). I think our work relationship has changed for the better, and I will always be available as the Little Lipase Consultant (I was told a consultant is paid better than a personal assistant). I hope you will find The Mutant with an ee of more than 99%! Misschien triest maar toch waar: ik woonde vooral op 957 in plaats van in de Hamburgerstraat de afgelopen vier jaar. Daarom wil ik heel graag mijn kamergenoten bedanken voor alle gezelligheid. My dear Mags, you left pretty soon after I moved to 957. Was it coincidence…? Anyway, I would like to thank you for your wisdom and wit, but most of all for bestowing on me your Irish accent. At least I can get away as a native speaker nowadays! Almer (Sloot), ik verhuisde naar Mags’ plek achter jou en ik vraag me nu af of dat eigenlijk iets uitgemaakt heeft...? Hoe dan ook, het ‘half-zes-benen-in-de-vensterbank-Senseo-moment’ (vanaf nu een woord) was een goed moment van de dag. We probeerden altijd Hoogwaardige Gesprekken te voeren, hoewel we uiteindelijk nooit besloten hebben wie de beste Nederlandse schrijver van de 20e eeuw was. Dankjewel voor al je wetenschappelijke inzichten en adviezen, niemand heeft zo’n kennis als jij! Robbert, Bohtjo (dit moet voor het nageslacht natuurlijk), minnaar... Het wordt lastig hier nog overheen te gaan! Jij bekeek dingen sceptischer dan ik, wat maakte dat het erg prettig was met je te overleggen. We hebben vooral de laatste tijd veel serieuze gesprekken gevoerd over Little Lipase die bindingsangst had aan welke kolom dan ook. Ik hoop dat we nog een manier vinden om Little Lipase hier overheen te helpen! Maar, mede namens Little Lipase, ontzettend bedankt voor al je hulp! Carlos, my Cookie, mi Galleta. What can I say? Ça va? It was so nice having Skype sessions while other conversations were going on in the office, but also to look up and start choking. Thanks for being peculiar (yes, you too)! Sadness… Despair… I will miss you! While talking about Carlos: Ulrike, thanks very much for your attempt in designing the cover of the thesis! Gudrun, take care of the boys, and I think you have the potential to be a good Mini-Me, just improve a bit on the evil part! Ook belangrijk: de leden van Groep Boersma ofwel de studenten. Lotte, Paula (Jut of Jul?), Jurre, Wu, Margriet (Bloem), Remko (Mike) en Riaaz: bedankt voor jullie hulp, inzet, en enthousiasme! Zonder jullie was dit proefschrift minder dik geweest! Uli, jij uiteraard ook bedankt. Waarvoor? Bovenal voor de YBASUD! Helaas werkte het niet zoals we wilden, als in dat het meteen uit elkaar viel, maar het idee was leuk. Dank voor de borrels en de Spaans en de gezelligheid!
DANKWOORD
171
Mijn Mannen mogen in dit rijtje niet ontbreken. Charles, Almer, Mattijs: ik heb jullie op mijn laatste werkdag een pre-dankwoord gestuurd, jullie waren Heel Belangrijk en ik geloof niet dat ik hier nog overheen kom. Mattijs, we hadden altijd goede, persoonlijke gesprekken samen, dankjewel daarvoor. Oogcontact was veelzeggend! Het was een eer om bij jou te paranimfen. Charles, samen een zaterdagmiddag-witbiertje, na een uitgebreide shopactie bijvoorbeeld, fijn! Tenks voor alle klets, gezelligheid en ook de peptalks! Lieve Almer, allereerst natuurlijk bedankt voor de geuzennaam, het leek me verstandig Evil-Lien maar gewoon te adopteren. Inmiddels ben ik stiekem een beetje trots op die naam! Uiteraard ben ik ook blij met de werkervaring die ik opgedaan heb als jouw PA/PC (kan ik dat eigenlijk gewoon op mijn CV zetten?), en ook met alle vergeten PA-dagen die je briljant goedgemaakt hebt met het Toetjesrestaurant. Bedankt voor je geduld, de waardering, de gezelligheid, alle klets (welke kant het dan ook maar op gaat). Eigenlijk kan ik niet anders zeggen dan: gracias para TODO! Lieve Maaike, mijn secundus. Ik leerde je kennen via Spaans, wat we inmiddels allebei wat ‘minder’ beoefenen. Je had ook farmacie gestudeerd, was in ‘96 begonnen en bleek een jaargenote te zijn van Mariette. Dat je uit de opernbare apotheek weg wilde, vond ik vooral Heel Goed! Inmiddels zit je voor Organon zelfs in Parijs, en toch kun je nog steeds voor mij paranimfen! Ik vind het echt geweldig dat je mijn paranimf wilt zijn, dankjewel! Lieve Jet, je wilde graag een pagina-lang dankwoord. Pff... Ik denk dat ik te veel pagina’s kan vullen aan dankwoord, dus misschien kan ik beter wat bondiger zijn. Ik ga onze Relatie beschrijven als één van de Slechte Woensdagavond Films (maar dan versneld) die we soms kijken, goed? We leerden elkaar kennen toen jij je bijvak deed bij Sloot, en hoewel we eerst om elkaar heen draaiden, klikte het stiekem eigenlijk wel goed (ik zei al dat het een Slechte Woensdagavond Film ging worden). Dus gingen we samen eten, samen drinken (beter!), samen hangen., samen picknicken, samen beamen, samen haardvuur... Welk een goed huwelijk (is de divorce echt nog steeds eminent?)! Je ging bij FaBio werken en dat was qua koffie, luns en thee een verhoging van de feestvreugde! Ik hoop dat je in je onderzoek zowel een agonist als een antagonist zult vinden (wat jij wilt), maar vooral dat we samen een of andere samenwerking kunen opzetten (niet op de TIP natuurlijk)! Maar eigenlijk wil ik gewoon heel erg dankjewel zeggen voor de vriendschap, het vertrouwen en de gezelligheid. En uiteraard dat je mijn paranimf bent natuurlijk. Tenks luv! Lieve Mem, hier sluit ik mijn dankwoord af. Er kwamen vaak enthousiaste verhalen over de nieuwe hobby vanuit Groningen, hoewel ik betwijfel of Mem (en ook Heit) mijn werk altijd helemaal begrepen. Lieve Mem, bedankt ‘dat Mem er altyd steat’, bedankt voor het vertrouwen en bovenal voor het goede contact!
LIST OF PUBLICATIONS
173
M.J. Dröge, R. Bos, Y.L. Boersma, W.J. Quax. Comparison and functional characterization of three homologous intracellular carboxylesterases of Bacillus subtilis. J Mol Catal B – Enzymatic 2005; 32: 261-270. Y.L. Boersma, P.C. Scheltinga, M.J. Dröge, R. Bos, W.J. Quax. A validated gas chromatographic method for the evaluation of enzymatic enantioselectivity in kinetic resolution applications. J Sep Sci 2005; 28: 501-505. Y.L. Boersma, M.J. Dröge, G. van Pouderoyen, T.E. Vrenken, C.J. Rüggeberg, M.T. Reetz, B.W. Dijkstra, W.J. Quax. Directed evolution of Bacillus subtilis lipase A by use of enantiomeric phosphonate inhibitors: crystal structures and phage display. Chembiochem 2006; 7: 149-157. Y.L. Boersma, M.J. Dröge, P.G. Braun, R.J. Buining, M.K. Julsing, K.G.A. Selles, J.M. van Dijl, W.J. Quax. Phage display of an intracellular carboxylesterase of Bacillus subtilis: a comparison of the Sec and Tat pathway export capabilities. Appl Environ Microbiol 2006; 72: 4589-4595. R. Bos, T. Windono, H.J. Woerdenbag, Y.L. Boersma, A. Koulman, O. Kayser. HPLC-Photo array detection analysis of curcuminoids in Curcuma species indigenous to Indonesia. Phytochemical Analysis 2007; 18: 118-122. Y.L. Boersma, M.J. Dröge, W.J. Quax. Selection strategies for improved biocatalysts. FEBS J 2007; 274: 2181-2195 Y.L. Boersma, M.J.Dröge, T. Pijning, M.S. Bosma, R.T. Winter, G. van Pouderoyen, B.W. Dijkstra, W.J. Quax. Rational design of Bacillus subtilis lipase A loop hybrids: insertion of a lid structure improves enantioselectivity. (in preparation) Y.L. Boersma, M.J. Dröge, A.M. van der Sloot, T. Pijning, B.W. Dijkstra, W.J. Quax. A novel selection system for enantioselectivity of Bacillus subtilis lipase A based on bacterial growth. (submitted)
Stellingen
behorende bij het proefschrift
Evolution of enantioselecitivity Selection of improved hydrolase variants
1. De combinatie van rational design en directed evolution biedt nieuwe perspectieven voor de
succesvolle verbetering van de enzymatische enantioselectiviteit. (dit proefschrift)
2. Insertie van een aantal aminozuren in het molecuul van lipase A doet de enantioselectiviteit omkeren. (dit proefschrift)
3. Loop 11-20 van lipase A is door zijn positie dichtbij het katalytische centrum van het enzym van groot belang voor de enantioselectiviteit van het enzym en de substraatspecificiteit. (dit proefschrift)
4. Aspartaat 133 van de zogenaamde “catalytic triad” van lipase A is geen strikt geconserveerde positie: bij verplaatsing van dit aminozuur blijft het enzym actief, maar keert de enantioselectiviteit om. (dit proefschrift)
5. Het grote voordeel van een bacteriële groeiselectie is, dat zowel op enzymatische activiteit als op enantioselectiviteit geselecteerd kan worden. (dit proefschrift)
6. De ontdekking van het “after dinner”-eiwit STAMP2 en de mogelijke ontwikkeling hiervan tot lifestyle drug zal paradoxaal genoeg de strijd tegen obesitas weinig goed doen. (Cell 2007; 129: 539-548)
7. Tips om te overleven in een zwart gat zijn feitelijk nutteloos. (Nature News, 18-05-2007)
8. Serendipiteit is rechtevenredig met creativiteit en fouten, maar omgekeerd evenredig met planning en organisatie.
9. Moleculair koken zou volgens fysisch-chemicus Hervé This voor een betere smaak zorgen (Nat Mater 2005; 4: 5-7); proefondervindelijk valt de consistentie van een ei anderhalf uur verwarmd bij 60 ˚C echter in het niet bij de consistentie van een gekookt ei. (proefondervindelijk vastgesteld op 04-10-2006)
10. De nu oprukkende Bimbocultuur vraagt om een derde emancipatiegolf. (NRC Handelsblad, 14-04-2007)
11. Onlangs werd vastgesteld dat een zieke Drosophila vlieg slecht slaapt; vertaald naar de mens geeft deze bevinding het aloude doktersadvies “bedrust” meer waarde. (Curr Biol 2007;17: R353-R355)
12. Het ontbreken van een term die het midden houdt tussen kennis en vriend leidt tot een devaluatie van de laatstgenoemde.
