Developing sub proteomic methods for large scale profiling

17 1.3 Aim of the project ...19 1.3.2 Differential expression profiling of serine hydrolases in normal and apoptotic cells ..... 33 Chapter 3: Results ...35 3.1 Labeling and identifying

DEVELOPING SUB-PROTEOMIC METHODS FOR LARGE

NATIONAL UNIVERSITY OF SINGAPORE

2003

ACKNOWLEDGMENTS

I wish to thank my supervisor, Dr Yao Shao Qin for giving me the opportunity

to do research in his laboratory, for his unwavering patience and encouragement during these years, and all the advice and helps during the writing of the thesis

To my lab mates for making the lab a warm, friendly and interesting place to work in

To my net group members and friends, especially Cheryl for her prayers, Raymond for his company and Jonathan for going through this Accelerated Master Program with me

To my family, especially my adorable siblings (David, Esther, John) for their support and encouragement

To God

TABLE OF CONTENTS

Acknowledgments i

Table of Contents ii

Summary vi

Publications ……….………viii

List of Figures……… ix

List of Abbreviations………xi

Chapter 1: Introduction 3

1.1 Post-genomic era and its challenges 3

1.2 Addressing the proteomic challenges 5

1.2.1 Differential in-gel electrophoresis (DIGE) 5

1.2.2 Sub-proteomic expression profiling of proteins using activity-based probes (ABPs) 8

1.2.2.1 Reactive unit 8

1.2.2.2 Linker/recognition unit 10

1.2.2.3 Tag unit 11

1.2.2.4 Application of ABP in gel-based proteomics 12

1.3.1 Apoptosis 15

1.3.1.1 Caspases 16

1.3.1.2 Serine proteases in apoptosis 17

1.3 Aim of the project 19

1.3.2 Differential expression profiling of serine hydrolases in normal and apoptotic cells 20

ii

1.3.3 In vivo labeling of caspases in apoptotic HeLa cells 21

Chapter 2: Materials and Methods 24

2.1 Materials 24

2.2 GatewayTM Technology 24

2.2.1 PCR amplification of a known serine hydrolase gene 24

2.2.2 BP reaction 25

2.2.3 LR reaction 26

2.2.4 Transformation and induction of protein expression in BL21 (AI) 26

2.4.4 Sequencing confirmation of clones 27

2.3 Preparation of bacterial and yeast cell lysate 28

2.4 Estimation of protein concentration 28

2.5 In vitro labeling 28

2.6 Affinity purification 28

2.7 Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) 29

2.8 Two-dimensional DIGE (2D DIGE) and imaging 29

2.9 Sypro ruby protein stain 30

2.10 Western blot 30

2.11 Elution of whole protein from SDS-PAGE gel 31

2.12 In-gel digestion of proteins 31

2.13 Matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (MS) 32

2.14 Preparation for apoptotic cells 32

2.14.1 Strains and culture condition 32

2.14.2 Apoptosis induction of HeLa cells by Ultraviolet (UV) irradiation 32

2.14.3 In vivo labeling 33

2.14.4 Evaluation of apoptosis 33

2.14.5 Preparation of HeLa cell extract 33

Chapter 3: Results 35

3.1 Labeling and identifying of a commercially available serine hydrolase .35

3.2 Optimization of sample preparation steps 37

3.3 Identification of serine hydrolases in yeast proteome 41

3.4Differential profiling of serine hydrolases 44

3.4.1 Proof of concept 44

3.4.2 Differential profiling of serine hydrolases in normal and apoptotic HeLa cells 48

Chapter 4: Discussion 59

4.1 Labeling and identification of a commercially available serine hydrolase……….59

4.2 Optimization of sample preparation step 60

4.3 Identification of serine hydrolases in a yeast proteome 62

4.4 Differential profiling of serine hydrolases 65

4.4.1 Proof of concept 65

iv

4.4.2 Differential profiling of serine hydrolases in normal and

apoptotic cells 68

4.5 In vivo labeling of caspases 71

4.5.1 In vivo labeling of caspases 71

4.5.2 In vivo labeling of caspase-associated substrates 73

4.6 Activity-based profiling 77

Chapter 5: Conclusion 80

Chapter 6: References 81

Chapter 7: Appendix 113

SUMMARY With the availability of complete genome sequence, emphasis has shifted towards the understanding of protein function Two-dimensional gel electrophoresis (2D-GE) is a state-of-the-art technique currently used for large-scale studies of proteomes from different organisms Due to its limited resolution of detection (detecting 1000-5000 proteins), this technique alone, however, is not sufficient to study complex proteomes, such as that of human (>40,000 proteins) Thus, there is an urgent need to develop a so-called sub-proteomic approach which is capable of analyzing, with higher-resolution, subsets of proteins in a proteome

We have developed a functional proteomic methodology that makes use of a combination of an emerging technology, DIGE and fluorescent probes The probes are made up of fluorophosphonates (FP) which are mechanism-based suicide inhibitors of serine hydrolases In DIGE, two pools

of proteins are labeled with 1-(5-carboxypentyl)-1-propylindocarbocyanine halide (Cy3) FP probe and 1-(5- carboxypentyl)-1-methylindodi-carbocyanine halide (Cy5) FP probe, respectively The labeled proteins are mixed and separated in the same 2D gel, allowing quantification of differential serine hydrolases expression Using Gateway Technology by Invitrogen, a known serine hydolase was introduced into the bacterial expression host Differences

vi

in expression of the serine hydrolase between induced and uninduced condition were quantified by 2D DIGE This methodology was further extended into quantifying differences in the expression of serine hydrolases between normal and apoptotic cells

In another part of the project, we report efficient labeling of caspases expressed inside apoptotic HeLa cells using fluorescently-labeled or biotinylated, fluoromethylketone (fmk)-containing probe Preliminary results with these probes indicated that they were highly cell-permeable, and caspase-

8 has been identified by biotinylated fmk-containing probe

In conclusion, 2D DIGE, in combination with fluorescent probes and

MS, will become a powerful tool for molecular characterization of cancer progression and identification of cancer-specific protein markers

PUBLICATIONS

1 Huang, X., Tan, E.L.P., Chen, G.Y.J., Yao, S.Q

“Enzyme-targeting small molecule probes for proteomics applications”,

Appl Genomics Proteomics, 2003, in press

2 Tan, E.L.P., Panicker, R.C., Tan, L.P., Chattopadhaya, S., Yao,

S.Q “Developing chemical biology tools for the study of functional proteomics”, Proceedings of the 1st International Symposium on Biomolecular Chemistry, Japan

3 Tan, E.L.P., Yao, S.Q “Activity-based, differential expression

profiling of serine hydrolases in different proteomes”, manuscript

in preparation

4 Tan, E.L.P, Panicker, R.C., Yao, S.Q “In vivo activity-based

profiling of caspases and their associated proteins”, manuscript in preparation

viii

LIST OF FIGURES

Number Page

Figure 1 Strategy of 2D difference gel electrophoresis (DIGE) 7

Figure 2 General structure of an activity-based probe 8

Figure 3 Strategy of gel-based activity profiling using ABP 13

Figure 4 Caspase activation cascade 17

Figure 5 Structure of FP-Cy3 20

Figure 6 Structure of FP-Cy5 21

Figure 7 Structure of FP-Biotin 21

Figure 8 Structure of caspase-biotin 22

Figure 9 Structure of caspase-fluorescein 23

Figure 10 SDS-PAGE of FP-Cy5-labeled chymotrypsin 35

Figure 11 MALDI-TOF-MS spectrum of undigested chemotrypsin 36

Figure 12 MALDI-TOF-MS spectrum of digested chymotrypsin 36

Figure 13 Chymotrypsin matched peptide alignment sequences 37

Figure 14 Optimization of lysis and labeling condition 38

Figure 15 2D fluorescent gel of FP-Cy5 labeled yeast proteins .38

Figure 16 2D gel showing FP-Cy5 labeled yeast fluorescent spots .40

Figure 17 2D gel showing FP-biotin labeled proteins from yeast lysate 42

Figure 18 Protein sequence alignment of sec17 and sec11 using ClustalW 43

Figure 19 : The GatewayTM system 45

Figure 20 PCR analysis to verify the presence of the gene, YSP3, in bacteria expression host BL21 (AI) 45

Figure 21 YSP3 Basepair alignment of the sequencing results 46

Figure 22 Over-expression of YSP3 by arabinose induction in bacteria expression host BL21 (AI) 47

Figure 23 DIGE of induced and non-induced transformed bacterial lysates 47

Figure 24 Normal (a) and apoptotic (b) HeLa cells 50 Figure 25 DIGE of normal and apoptotic cell lysates 50 Figure 26 HeLa cells labeled with caspase-fluorescein 53 Figure 27 Normal and apoptotic HeLa cells labeled with caspase-fluorescein 53 Figure 28 Fluorescent gel images of caspase-fluorescein labeled normal and apoptotic HeLa cells 54 Figure 29 Caspase-fluorescein labeling of apoptotic HeLa cells at different time point after UVB-irradiation 54 Figure 30 Comparison of fluorescence (a) and biotin-avidin (b) detection modes for analyzing caspases .55 Figure 31 MALDI-TOF-MS spectrum of caspase-8 56 Figure 32 Caspase-8 matched peptide alignment sequences 56 Figure 33 Comparing protein patterns between 2 Coomassie-stained images: (a) caspase-biotin-labeled HeLa cell lysate (b) unlabeled HeLa cell lysate 57 Figure 34 Schematic representation of the regulation based on araBAD 67 Figure 35 A possible scenario in the less stringent condition 76

x

LIST OF ABBREVIATION

2D-GE Two-dimensional gel electrophoresis

ABP Activity-based probe

BTEE N-benzoyl-L-tyrosine ethyl ester

MALDI-TOF matrix-assisted laser desorption ionization-time of flight

PBS phosphate-buffered saline

PCR Polymerase Chain Reaction

SDS-PAGE Sodium dodecyl sulphate polyacrylamide gel electrophoresis

TLCK N-tosyl-L-lysine chloromethyl ketone

TPCK N-tosyl-L-phenylalanine chloromethyl ketone

UV(B) Ultraviolet- (B)

PA28γ proteasome activator 28 subunit 3

ER endoplasmic reticulum

xii

C h a p t e r 1

INTRODUCTION 1.1 Post-genomic era and its challenges

Over the past two decades, there is an increasing number of prokaryotic and eukaryotic organisms being sequenced From a 49-kbp (base pair) bacteriophage lambda

genome (Sanger F et al., 1982) to a 2.91-billion bp human genome (Venter J.C et al.,

2001), these, nevertheless, provide an exciting blueprint to understand cell biology However, these huge sets of DNA sequences are insufficient, as it is known that proteins are ultimately responsible for most processes that take place within the cells Therefore, PROTEOMICS (the analysis of the entire PROTEin complement expressed by a genOME),

as defined by Marc Wilkins and Keith Williams (1996), grow increasingly important in the

post-genomics era (Fields S et al., 2001)

It has been estimated that the human proteome could contain from as few as 100,000 different proteins, to as many as a few million By studying the dynamic description of gene regulation, proteomics techniques offers a powerful tool to unravel gene functions, holding promises to significantly impact our understanding of the molecular composition and function of cells The changes in protein profiles, for example under the influence of different signals, bacterial/viral infection, or environmental factors, can be monitored and quantitated These will provide a powerful tool to identify proteins that can be potential targets for diagnostic or therapeutics

Despite its wide application, this strategy is not without limitations Firstly, when comparing protein differences between two samples, comparison of images from at least 2 different gels must be made But because of variations between gels, due to subtle changes

in experimental conditions, no two gel images are directly superimposable Very often, each sample must be run on several replicate gels to generate electronic ‘average’ gels which can then be compared As a result, high through-put analysis of many samples becomes tedious and impractical and subtle changes in the number of proteins and their expression levels between samples might be easily overlook Secondly, current detection methods of 2DE rely on staining methods like Coomassie blue and silver staining The latter, though being more sensitive and widely used, but it is still unsuitable for accurate quantitative analysis due to its limited dynamic range Therefore, 2D-GE can only be used

to analyze a small fraction of highly abundant proteins (<5000) in a proteome (Gygi S.P

et al., 2000) Membrane-bound and low abundance proteins, which are frequently the

protein interest of many researchers, have proven difficult to be analyze by 2D-GE, thereby rendering the technique insufficient for large-scale protein profiling in human

1.2 Addressing the proteomic challenges

In this section, two strategies that have been developed to address the two 2D-GE

limitations mentioned above will be presented

1.2.1 Differential in-gel electrophoresis (DIGE)

DIGE has been developed by Unlu M et al (1997) to address the issue on

difficulties of superimposing two 2D gels This technique relies on pre-electrophoretic labeling of each of the comparing samples with one of the three spectrally distinct fluorescent dyes, Cyanine-2 (Cy2), Cyanine-3 (Cy3), and Cyanine-5 (Cy5) The dyes have

an NHS-ester reactive group and are designed to covalently attach to epsilon amino group

of lysine of proteins via an amide linkage

The labeled samples are then quenched, mixed and run on a single gel Because the dyes are structurally similar, same proteins from different samples will co-migrate as same spots on the same gel Quantitation of differential protein expression is made easy by viewing the gel at different wavelengths and by comparing the intensity difference of the two different labels on the same spot (Figure 1) This positional reproducibility means image overlay rapidly shows any difference in protein levels between samples, and also accelerates image analysis, enabling higher sample throughput

In addition, there are other advantages for DIGE over conventional 2D-GE First, optimized DIGE technology is capable of detecting about half as many proteins as conventional silver staining, or four times as many proteins as colloidial Coomassie Blue

staining (Tonge R et al., 2001) It has been reported that DIGE is able to detect down to at least 500 pg of a protein (Orange P et al., 2000) Second, DIGE has a linear response to

variation in protein concentration over four orders of magnitude (104) (Tonge R et al.,

2001) This enables DIGE to provide more accurate quantitative data in less time than is achieved with traditional silver staining, which is only linear over 60-100 fold differences in

protein concentration (Patton W et al., 2000; Orange P et al., 2000)

To date, DIGE has been successfully used to profile changes in protein expression in

several toxicity studies (Ruepp S.U et al, 2002, Smith D.R et al, 2003) and in comparison between normal and cancer/diseased cells (Gharbi S et al, 2002, Zhou G et al 2002, Lee J.R et al, 2003, Yamanaka H et al, 2003) In one of the works by Zhou G et al (2002),

DIGE was applied to quantify the differences between laser capture procured esophageal carcinoma cells and normal epithelial cells Of the detected proteins,

microdissection-58 spots were up-regulated by >3-fold and 107 were down-regulated by >3-fold in cancer cells They have shown that 2D DIGE when combine with mass spectrometry can be a powerful tool for molecular characterization of cancer progression and identification of cancer-specific protein markers

In another work by Hu et al (2003), they reported the first case toexplore the throughput feature of DIGE, which detect changes in the yeast proteome under 15 different metal stresses About 50 yeast proteins which were either up-regulated or down-regulated during the metal stresses were identified

high-Protein extract 1 Protein extract 2

Label with fluorochome 1 Label with fluorochome 2

Mix labeled extracts

Separate by 2D electrophoresis

Image gel Excitation wavelength 1 Excitation wavelength 2

Image analysis and data quantification

Figure 1 Strategy of 2D DIGE

1.2.2 Sub-proteomic expression profiling of proteins using activity-based probes (ABPs)

As mentioned before, more than 100 000 proteins are expected from the genome and millions of protein-protein interactions are possible in protein complexes and networks Hence, fractionating the proteome into manageable and physiologically relevant subsets becomes relevant and critical This can be made possible with the use of activity-

(sub-proteome) Generally, ABP are made up of three units: reactive unit, linker/recognition unit and tag unit (Figure 2) The components and functions of each unit will be elaborated in following sections Each type of probe labels or isolates only a specific class of enzymes For example a fluorophosphonate (FP) probe synthesized by Cravatt’s group targets the

serine hydrolase family of enzyme (Liu Y et al 1999), while another probe synthesized by Bogyo’s group targets the papain family of cysteine proteases (Greenbaum D et al., 2002)

Reactive unit Linker/Recognition unit

as denaturing conditions during gel-based separation of probe-labeled proteins from unlabeled proteins This can only occur when an irreversible complex is formed between the probe and targeting enzyme, typically as a result of covalent reactions In cases where non-covalent complex is formed, the enzyme loses its 3-dimensional structure under the denaturing conditions, readily detach itself from the probe, leading to subsequent dissociation of the complex Consequently, the main function of the reactive units within the ABP is to ensure irreversible, covalent bonds are formed between the probe and enzymes it targets The reactive units in ABPs are made up of mechanism-based substrates

or suicide inhibitors, which label enzymes during the catalytic process

The chemical groups on different reactive units may be fine tuned to target different classes of enzymes based on their enzymatic activity under a complex background of other enzymes in a proteome Most ABPs use electrophilic chemical groups as their reactive units, which, upon binding to the enzymes, are able to selectively, as well as covalently,

modify the active sites of the enzymes (Jeffery D.A et al 2002) This strategy is quite

general, as most, if not all, enzymes invariably contain nucleophilic groups within their active sites, which provide good “electrophilic traps” for the reactive unit in the probe

The ABPs reported thus far can be divided into two types, based on mechanisms by which their reactive units covalently label the enzymes The first type is true mechanism-based probes which is designed by using actual enzyme mechanism as blueprint and is able

to covalently modify the enzyme catalytic center Examples of such ABP include: epoxysuccinyl derivatives, acyloxymethyl ketones, and peptidyl vinyl sulfone which target

cysteine protease (Thornberry N.A et al 1994, Greenbaum D et al.2000, Wang G et al 2003), fluorophosphonate (FP) group which target serine hydrolases (Liu Y et al 1999) and

sulfonate ester which target several different classes of enzymes (Thiolase, aldehyde dehydrogenase, NAD/NADP-dependent oxidoreductase, epoxide hydrolase, glutathione S-

transferase) (Adam G.C et al 2001, 2002a) The second type of reactive unit is based on

suicide inhibitors and requires enzyme to cleave the substrate to uncover highly active nucleophilic group which can eventually react with nearby protein residues Examples of

suicide inhibitor probes are p-hydroxymandelic acid derivatives and 2-difluoromethylphenyl phosphate targeted to phosphatases (Lo L.C et al.1996, Betley J.R et al 2002, Lo L et

al.2002, Zhu Q et al.2003)

1.2.2.2 Linker/recognition unit

In a typical ABP, the reactive and the tag units have different biological functions upon binding to the target enzyme The reactive unit typically contains certain chemical functionality that is recognized, and subsequently becomes reactive toward the active site of the enzyme The tag unit on the other hand, is typically made up of either a fluorescent

molecule (e.g FITC, Cy3 & Cy5 etc) or an affinity tag (e.g biotin), which in many cases is

detrimental to the binding between the probe and the enzyme Consequently, it is often necessary to have a linker unit, within the ABP, which serves not only to bridge the two units together, but also to minimize the overall binding interference of the tag unit toward the enzyme Additionally, the linker unit in many cases also serves as part of the enzyme-recognizing unit in facilitating the reactive unit-enzyme binding, as well as defining differential specificity of a probe toward different enzymes within the same class

There are generally two types of linker: sequence-specific and nonspecific linkers The first type is rarely used in broad-based ABP, because they are typically made up of defined peptide sequence and are highly specific to their targeting enzymes The nonspecific linkers are more commonly used, as they are well-suited for potential large-scale protein profiling experiments, which aim to simultaneously target many different enzymes within the same class These linkers are typically made up of long aliphatic chains of either hydrophobic (e.g alkyl chains) or hydrophilic molecules (e.g polyethylene glycols, polyglycines), which serve to modulate the probes’ activities (e.g reactivity, solubility etc)

(Kidd D et al 2001)

1.2.2.3 Tag unit

The tag unit is incorporated into an activity-based small molecule probe to facilitate sensitive detection, quantitation, isolation and the subsequent identification of enzymes upon their labeling by the probe Most commonly used tags are: biotin and its derivatives, fluorescent dyes, isotopes and others, just to name a few

Biotin is an affinity tag, its interaction with streptavidin is one of the strongest covalent interactions known and can stand a relatively harsh washing condition (Green

non-N.M et al 1973, Reznik G.O et al 2001) The biotin tagged-probe labeled protein can be

isolated, concentrated and purified by streptavadin-agarose bead from the crude cell lysate and subsequently identified by MS This strategy is valuable to detect low abundance enzyme The labeled protein can also be separated by gel based method and visualized by commercially available western blot detection kit (Bio-rad website) However, very often

100-fold better than biotin tag labeled probe (Patricelli M.P et al 2001)

1.2.2.4 Application of ABP in gel-based proteomics

ABP can provide valuable information (eg differential enzyme activities) only when

it is used with other proteomics techniques Currently, ABP have been successfully combined with two existing proteomics platforms, namely gel-based proteomics (Cravatt

B.F et al., 2000; Jessani N et al., 2002; Weihofen A et al., 2002) and microarray-based technologies (Chen et al., 2003) Due to the nature of this project, the latter will not be

discussed and only the former will be elaborated

ABP, when combined with 2DE-MS, simultaneously detects changes in the activity

of many proteins of the same class in the cells, tissues and fluid samples (Cravatt B.F et

al., 2000) Typically, proteins, which were extracted from cells or tissues, were incubated

with ABP, and followed by separation by gel-based techniques like sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE), and the tag signal of each bands were detected by a suitable techniques (e.g fluorescence imaging or western blots) The activities of the probe-labeled enzymes may be profiled and MS was carried out to identify the labeled proteins (For overall scheme refer to Figure 3)

Proteome (extracted cell lysate)

Incubation with fluorescent ABP

An example of a work carried out by Jessani N et al (2002) is the profiling of

serine hydrolases activity of a panel of breast and melanoma cell lines In this paper, two FP probes were used, one was rhodamine- coupled FP probe for easy profiling of serine hydrolases activity, and another was biotin coupled FP probe for rapid isolation and identification of labeled proteins By quantitatively comparing the profiling data of different cell lines, a serine hydrolase (KIAA1363) was found to be a molecular marker for cancer cell invasiveness This work aptly demonstrated the potential of ABP in enzyme activity profiling and their subsequent contribution to the diagnosis and possible treatment for cancers

Another application of enzyme-targeting probes is in the identification of new functions in unknown proteins from a proteome Because the probe is able to specifically label a class, or in some instances a few classes, of enzymes, it could be used to characterize previously uncharacterized enzymes in a proteome based on their activity An example of this type of work is in the identification of a novel signal peptidase by TBL4K probe

(Weihofen A et al., 2002) The authors used the probe to label the detergent-solubilized,

ER-membrane subproteome, followed by separation on a SDS-PAGE They identified a 42-kD protein, which was labeled by TBL4K probe in the presence of detergents Upon further analysis by MS, the protein was confirmed to be a previously uncharacterized protein Consequently, they were able to further classify the protein as a signal peptidase, belonging to a class of rare proteins called intramembrane proteases

1.3.1 Apoptosis

Apoptosis, or programmed cell death, is a cell-intrinsic process that is essential for animal development and tissue homeostasis It is needed for the formation of fingers and

toes of foetus and formation of synapses between neurons in the brain (Jacobson M.D et

al., 1997) But in human, both excessive and insufficient apoptosis can lead to severe

diseases For example, abnormal up-regulation of apoptosis contributes to neurodegenerative like Alzheimer's disease, Parkinson's disease, Huntington's disease and amyotrophic lateral (Yuan J and Yankner B.A., 2000) On the other hand, suppression of the apoptotic machinery causes autoimmune diseases and is a hallmark of cancer (Hanahan

D and Weinberg R.A., 2000; Thompson C.B 1995)

Apoptosis can be triggered by a variety of stimuli, these include cytokines, hormones, viruses, and toxins The initial sign of apoptosis was morphological: dying cells exhibit a characteristic pattern of changes, which include cytoplasmic shrinkage, active membrane blebbing, chromatin condensation, and, typically, fragmentation into

membrane-enclosed vesicles (Wyllie A.H et al., 1980) This visible transformation is

accompanied by a number of biochemical changes Changes at the cell surface include

externalization of phosphatidylserine (Martin S.J et al., 1995) and other alterations that

promote recognition by phagocytes (reviewed by Savill J., 1997) Intracellular changes include the degradation of chromosomal DNA into high-molecular-weight (Oberhammer

F et al., 1993) and oligonucleosomal fragments (Wyllie A.H., 1980), as well as cleavage

of a specific subset of cellular polypeptides (Kaufmann S.H 1989; Ucker D.S et al., 1992; Lazebnik Y.A et al., 1993) This cleavage is now known to be accomplished by a

specialized family of cysteine-dependent aspartate-directed proteases (Lazebnik Y.A et

al., 1994) termed caspases (Alnemri E.S et al., 1996)

1.3.1.1 Caspases

The first discovery of the involvement of protease in apoptosis was in nematode

Caenorhabditis elegans (Yuan J et al., 1993) The CED-3, a cysteine protease with unusual

substrate specificity for aspartate residues, soon led to the discovery of its mammalian relative (caspases) Since then, a large number of CED-3-related proteases have been identified At least 14 distinct mammalian caspases have been identified, but not all of them participate in apoptosis (Shi Y., 2002)

Caspases are synthesized as proenzymes that are composed of three domains: an terminal prodomain, a large subunit (about 20kDa) and a small subunit (about 10kDa)

N-(Earnshaw W.C et al., 1999) Caspases are generally divided into two groups: the initiators,

which initiate the apoptosis cascade, and the effector which carry out proteolytic events leading directly to cellular breakdown The initiators are activated by proteolytic cleavage and subsequent heterodimerise Once activated, they trans-activate other caspases (the effectors) by proteolytic cleavage This creates a proteolytic cascade that spreads outwardly and destroys key proteins, leading to apoptosis

Prodomain Large subunit Small subunit

Initiator procaspase (caspase-2, -8, -9, -10)

Active initiator caspase (heterotetramer)

ActivateExecutioner procaspase (caspase-3, -6, -7)

Active executioner caspase

Caspase substrate

Apoptosis Figure 4 Caspase activation cascade

1.3.1.2 Serine proteases in apoptosis

Serine hydrolases, which contain a serine at the active center that participates in the formation of an intermediate ester to transiently form an acyl-enzyme complex, represent one of the largest and most diverse classes of enzymes in higher eukaryotes, it makes up

~3% of the predicted Drosophila proteome (Rubin G.M et al., 2000) Subclasses of serine

hydrolases include proteases, lipases, esterases, amidases, and transacylases They play

important roles in numerous developmental and tissue-specific events in vivo, including

inflammation (Clark J.D et al., 1995), blood coagulation (Kalafatis M et al, 1997), angiogenesis (Mignatti P et al, 1996), peptide hormone processing (Seidah N.G et al, 1997), neural plasticity (Yoshida S et al, 1999), and T-lymphocyte- mediated cytotoxicity (Smyth M.J et al, 1996) In addition, several diseases are associated with dysfunctions in

serine proteases and/or their endogenous inhibitory proteins, including emphysema (Kato

G.J et al, 1999), hemorrhagic disorders (Kato G.J et al, 1999), and cancer (Declerck Y.A et al, 1997)

Compared to caspases, participation of other proteases in apoptosis is less understood (Johnson D.E., 2000), and this include serine protease Involvement of serine proteases in apoptosis (Serpases) has been studied mostly by observing whether particular apoptotic events can be prevented by the specific inhibitors of these enzymes In the early

studies, Gorczyca W et al (1992) have shown that fragmentation of DNA in HL-60 cells

treated with DNA topoisomerase inhibitors to induce apoptosis was prevented by

irreversible inhibitors of serine proteases such as N-tosyl-L-phenylalanine chloromethyl

ketone (TPCK), diisopropylfluoro-phosphate (DFP), and N-tosyl-L-lysine chloromethyl ketone (TLCK), as well as by excess of the substrates N-benzoyl-L-tyrosine ethyl ester

(BTEE) and N-tosyl-L-argininemethyl ester (TAME) Similarly, Bruno S et al (1992a, b)

observed that the same inhibitors and substrates inhibited nuclear fragmentation as well as fragmentation of DNA in other cell types, including thymocytes treated with the corticosteroid prednisolone It was also observed that these inhibitors prevented

destabilization of double-stranded DNA (Hara S et al., 1996) which during apoptosis

becomes sensitive to denaturing agents and can be detected as single-stranded DNA (Hotz

M.A et al., 1992) These initial observations were confirmed in many subsequent studies

and in other cell systems (Kim R et al., 2001; Mansat V el al., 2001; Gong B et al., 1999; Komatsu N et al., 1998)

One thing to note is, similar to caspases, the activities of many serine hydrolases are regulated in a posttranslational manner (e.g., zymogen cleavage for catalytic activation

(Khan A.R et al., 1998), inhibitor binding for catalytic inactivation (Whisstock J et al.,

1998 and Roberts R.M et al., 1995), indicating that standard, abundance-based genomics

and proteomics methods may be of limited use for the functional analysis of these enzymes To date, the number of mammalian serine hydrolases identified is impressive, with the endogenous functions of many members of this enzyme family remaining unknown Hence, there is a calling needs for alternative methods to study these enzymes One approach for the analysis of serine hydrolase function would be to characterize these enzymes collectively, rather than individually, using ABP

1.3 Aim of the project

In this project, we aimed to develop methods to profile a subset of a proteome, and it

is divided into two main parts In the first part, we aimed to combined DIGE technology and ABP to profile differential expression of serine hydrolases in normal and ultra violet (UV)-induced apoptotic HeLa cells And in the second part, using another ABP, we aimed to label

and identify caspases in vivo in apoptotic HeLa cells

1.3.2 Differential expression profiling of serine hydrolases in normal and apoptotic cells

It has been reported by Liu et al (1999) that a FP, when linked to a small molecule

reporter group, can serve as a potent and selective probe for simultaneous monitoring of all serine hydrolases This is because majority of the serine hydrolases are irreversibly inhibited by FP derivatives like diisopropyl fluorophosphates (Walsh C.T., 1979; Creighton T.E., 1993), whereas cysteine, aspartyl, and metallohydrolases are mostly inert

to such agents In addition, the reactivity of FPs with serine hydrolases requires that the enzymes be in a catalytically active state (Walsh C.T., 1979; Creighton T.E., 1993;

Pineiro-Sanchez M.L et al., 1997), hence making it an ideal ABP

In this part of the project, FP probe tagged with Cy3, Cy5 and biotin (Figure 5, Figure 6, Figure 7) were used to profile ONLY serine hydrolases The probes will be refered

to as FP-Cy3, FP-Cy5 and FP-biotin respectively from here after

N

N+

NH N

O

P F O O

Figure 5 Structure of FP-Cy3

N+

NH N

O

P F O

O

Figure 6 Structure of FP-Cy5

N H

H

P F

O O S

NH

NH

O

O

Figure 7 Structure of FP-Biotin

1.3.3 In vivo labeling of caspases in apoptotic HeLa cells

Currently, most applications of ABP are applied to proteins extracted from cells or

tissues lysates, (Jessani N et al., (2002); Weihofen A et al., 2002) which means that the

proteins are not in their native cellular environment Hence, “native” characters of enzymes, like endogenous activators/inhibitors effect and enzyme location, may be out of vision To

meet the challenge of in situ enzyme activity profiling in a dynamic cellular environment, a few ABP have been generated to label proteome in vivo (Greenbaum D et al., 2002; Speers A.E et al., 2003)

In our laboratory, a fellow colleague (Liau M.L et al 2003) had previously reported

a synthesis of a probe that labels caspases Liau showed that this probe labels commercially

available caspases over other non-caspase enzymes in vitro In the second part of this

project, we will present an extension of her work Similar probes were synthesized, but instead of using the reported Cy3 as tag unit, fluorescein and biotin were separately used to generate two other probes (the probe will be referred to as caspase-fluorescein and caspase-

biotin respectively) (Figure 8 and Figure 9) Using these probes, we aimed to show in vivo

labeling and identification of caspases in UV-induced apoptotic HeLa cells

F

H N

O

N H S

O O O

C h a p t e r 2

MATERIALS AND METHODS 2.1 Materials

All ABP were synthesized by my colleagues from the Department of Chemistry,

National University of Singapore Detailed synthesis of FP probe was described by Chen et

al.(2003), and the synthesis of caspase probe was described by Liau et al (2003)

Chymotrypsin were bought from Sigma (St Louis, Mo)

2.2 GatewayTM Technology

2.2.1 Polymerase Chain Reaction (PCR) amplification of a known serine hydrolase gene

A gene, subtilisin-like protease III (YSP3) from yeast ex-clone (Invitrogen, Ca) was extracted using Qiaprep Miniprep system (Qiagen) PCR of YSP3 was performed using the

forward primer BP-Exclone_F (5’- GGG GAC AAG TTT GTA CAA AAA AGC AGG

CTC CGT GGA ATT CCA GCT GAC CAC CAT G –3’), and the reverse primer

BP-Exclone_R (5’- GGG GAC CAC TTT GTA CAA GAA AGC TGG GTC GAT CCC

CGG GAA TTG CCA TG –3’) which contained both the attB sites (in bold) and the clone specific sequence (underlined) 50 µl reaction was performed using 100ng of YSP3 DNA, 25 pmole of each primers, Deep Vent polymerase buffer, 0.2mM of each dNTPs, 1 units of Deep Vent Polymerase and H2O Amplification was carried out using Peltier thermal cycler at 94oC for 4 mins, followed by 30 cycles of 94oC for 45 sec, 53oC for 45 sec, 72oC for 2 min, and a final extension at 72oC for 5mins

ex-Precipitation of the PCR products was done by adding 5 µl of 3 M Na Acetate pH 5.2, 100

µl of cold ethanol (100 %), and placed in -20oC for 10 hours Precipitated DNA was centrifuged (4000 rpm, 4oC for 1 hour) and pellet washed with 100 µl of 70 % ethanol The pellet was dried by placing on warm heat block, and 15 µl of H2O was used to re-suspend the DNA

2.2.2 BP reaction

The PCR products were used to perform 1/4 BP reaction with the pDONR201 The

BP reaction was performed by adding 100 fmol of the PCR products, 300 ng of supercoiled pDONR201, 1 µl of 5x BP Clonase reaction buffer (Invitrogen), 1 µl BP Clonase mix (Invitrogen) and topping up with TE Buffer pH 8.0 (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) to 6.25 µl The reaction mix was incubated at 25oC for 1 hour, before inactivation by adding 1 µg proteinase K (37 oC for 10 mins) Transformation was carried out by adding all

of the BP reaction mix to 100 µl of competent Top10 cells [(F-, mcrA,

∆(mrr-hsdRMS-mcrBC), φ80lacZ∆M15 ∆lacX74, deoR, recA1, araD139 ∆(ara-leu)7697, galK, rpsL(StrR), endA1, nupG], using heat shock method Cells were plated onto Luria-Bertoni

(LB) Km Agar (supplemented with 50µg/ml Kanamycin) Colonies from the plates were picked and inoculated in 5 ml LB Km Broth Plasmids were purified using Qiaprep Miniprep system (Qiagen) and ran on 1.2 % agarose gel PCR analysis was carried out to verify the present of the insert in the Entry clone using attB1_F (5’- ACA AGT TTG TAC AAA AAA GCA GGC T -3’) and attB2_R primers (5’- ACC ACT TTG TAC AAG AAA GCT GGG T -3’), which contained complementary sequences to the recombination sites

2.2.3 LR reaction

The correct clones were selected for 1/8 LR reaction (refer section 2.6) The LR reaction was performed by adding 37.5 ng of supercoiled Entry clone (genes of interest in pDONR), 37.5 ng of linearized Destination vector (pDEST17, pEGFP_DEST17 or pSTA_DEST17), 0.5 µl of 5x LR Clonase reaction buffer (Invitrogen), 0.5 µl LR Clonase mix (Invitrogen) and topping up with TE Buffer pH 8.0 to 3.12 µl The reaction mix was incubated at 25oC for 1 hour, before inactivation by adding 0.5 µg proteinase K (37 oC for

10 mins) Transformation was carried out by adding all of the LR reaction mix to 100 µl of competent Top10 cells, using heat shock method Cells were plated onto LB Amp Agar (supplemented with 100µg/ml Ampicillin) Colonies from the plates were picked and inoculated in 5 ml LB Amp Broth Plasmids were purified using Miniprep and ran on 1.2% agarose gel to check for correct size PCR analysis was carried out to verify the present of the insert using attB1_F and attB2_R primers which contained complementary sequences to the recombination sites Crude miniprep DNA screened by analytic PCR using attB1_F and attB2_R to confirm the authenticity of the clones

2.2.4 Transformation and induction of protein expression in BL21 (arabinose-induced) (AI)

Correct clones identified from LR reaction contained gene of interest inserted within the recombination site 1 µl of the plasmid was used to transform BL21 AI (Invitrogen) using the heat shock method 350 µl of the culture was plated on LB Amp Agar, with addition of 3.5 µl of 20% Arabinose One colony from the plate was picked to prepare 5 ml

seed culture in LB Amp 1 ml of the seed culture was used to inoculate 50 ml of LB Amp and culture incubated at 37 oC 280 rpm At 0.4 OD600, 500 µl of 20 % Arabinose was added

to induce protein expression from T7 promoter, and will be called “induced cell culture” In another flask of culture, no arabinose was added, and will be called “non-induced cell culture” The culture was allowed to incubate for additional 4 hours after the induction Cells were harvested by centrifugation at 4000 rpm, 4 oC for 10 mins

2.4.4 Sequencing confirmation of clones

Sequencing was performed using the ABI PRISM BigDye Terminator v3.0 Sequencing Kit (Applied Biosystem) 1/2 Reaction sequencing mix (20 µl) was prepared by mixing 4 µl Ready reaction mix, 2 µl 5x Sequence Buffer, 3.2 µl of (1 pmole/µl) Primers, 200-500 ng of BL21 (AI) plasmid DNA and H2O BP-Exclone_F was used as primers Reaction was carried out in the Peltier thermal cycler at 96 oC for 30 sec, 50 oC for 15 sec, and 60 oC for 4 min (25 cycles) The extension products were purified by adding 80 µl of ethanol/sodium acetate solution (3.0 µl of 3 M sodium acetate NaOAc pH 4.6, 62.5 µl of non-denatured 95 % ethanol EtOH, and 14.5 µl of H2O), and tubes allowed to stand at room temperature for 15 minutes to precipitate the extension DNA This was followed by centrifugation (15000 rpm for 20 mins) to pellet the DNA Washing was performed with 70

% EtOH, and the precipitant dried and stored in -20 oC Samples were sequenced using the ABI Prism 3100 sequencer (Applied Biosystem) at the Department of Biological Sciences, DNA Sequencing Laboratory (DSL), National University of Singapore

2.3 Preparation of bacterial and yeast cell lysate

Equal volume of ice-cold lysis buffer (5 mM Tris-HCl (pH 8.8) and 0.1 % Triton-X)

and acid washed glass beads (Sigma, 425-600 microns) were added to the cell pellet, and

the tube was placed in a mixer mill (Retsch, Haan, Germany) for 3 times of 6 minutes at 30 beat/sec at 4 oC Supernatants (protein sample) were collected after ultracentrifugation at 75

2.4 Estimation of protein concentration

Concentration of the protein samples were estimated using Bradford protein assay (Bio-rad, Richmond, Calif.) with standard protocols recommended by vendor (Bradford M., 1976)

The cell lysate was incubated with Streptavidin MagneSphere Paramagnetic

Particles (Promega, MA) for 1 h The resin was washed with 1% SDS and bound proteins

were eluted by boiling for 30 min in the same buffer The supernatant was kept for PAGE or 2D-GE

SDS-2.7 SDS-PAGE

Reactions were quenched with one volume of standard 2 X SDS-PAGE loading buffer, followed by heating the samples at 90oC for 5 min before analysis on 12 % or 15%

SDS-PAGE (Chen Y.J.G et al., 2003)

2.8 2D DIGE and imaging

Unless otherwise indicated, all 2D experiments were performed on an Ettan

Manual, Amersham) Fluoresecntly labeled sample or affinity purified sample in the rehydration buffer were applied to the immobilized pH gradient strip (pH 4–7, 18 cm; Amersham, USA), passively rehydrated for 1 h, followed by 12 h of active rehydration

the strips were equilibrated in equilibration buffer 1 (6 M urea, 2% SDS, 0.375 M Tris (pH 8.8), 20 % glycerol, 130 mM dithiothreitol; 15 min), then in equilibration buffer 2 (6

M urea, 2 % SDS, 0.375 M Tris (pH 8.8), 20% glycerol, 135 mM iodoacetamide; 15 min), before being applied directly to a 12% SDS-PAGE gel The SDS electrophoresis was run first at 2.5 W/gel (for 30 min), then 17 W/gel until the dye front was 1 mm from the