Analysis of p13k independent survival pathways in the prostate cancer cell line LN cap 1

...22 1.4.2 Aberrant Growth Factors Signaling in Cancer Cells...24 1.4.3 Growth Factors-Regulated Survival Signaling Pathways: PI3K-Akt Pathway.. ...158 4.1 EGF AND SERUM ACTIVATE PI3K-A

ANALYSIS OF PI3K-INDEPENDENT SURVIVAL

PATHWAYS IN THE PROSTATE CANCER CELL LINE

my abilities to accomplish the work involved I also wish to thank my co-supervisor Associate Professor Ren Ee Chee, Department of Microbiology, for opening the door

to the field of scientific research and his support through the years I am also indebted

to Professor Shazib Pervaiz and his lab in Department of Physiology for their helpful collaboration in knowledge and expertise

My warmest thanks to my lab colleagues and friends for making the sometimes long days in the lab more bearable I want to thank them for being my teachers, my extra pair of hands when two weren’t enough, my ‘cheerleaders’ when the going gets tough and most of all for being my friends

Finally, my deepest gratitude to my family for their support and encouragement throughout the years, and a special thanks to Seng, who has stood by me through the good and bad times I would not have been able to do this without his unfailing love and faith in me

Contents

Acknowledgements ii

Contents iii

Summary vi

Abbreviations xi

1.1.1 Overview of Apoptosis 1

1.1.2 Molecular mechanisms of apoptosis: Caspases as the central executioner of Apoptosis 2

Activation of caspases .4

Extrinsic and Intrinsic Apoptotic Pathway .5

Substrates of caspases 7

1.2 BCL-2 FAMILY 8 1.2.1 Role of Mitochondria in Apoptosis 8

1.2.2 Bcl-2 family proteins 10

The Multidomain Pro-survival proteins 12

The Multidomain Pro-apoptotic proteins 13

The BH3-only Pro-apoptotic proteins 15

1.2.3 Interactions Among Bcl-2 family members .17

1.3 DEFECTS IN APOPTOSIS AND CANCER 18 1.3.1 Mutations that Confer Apoptosis Resistance .19

1.4 ANTI-APOPTOTIC MECHANISMS: GROWTH FACTOR SIGNALING 22 1.4.1 Growth Factor Signaling .22

1.4.2 Aberrant Growth Factors Signaling in Cancer Cells 24

1.4.3 Growth Factors-Regulated Survival Signaling Pathways: PI3K-Akt Pathway .27

Aberrations of PI3K-Akt Signaling in Cancer .27

PI3K-Akt Signaling .28

Akt-mediated Survival Signaling 29

1.4.4 Growth Factors-Regulated Survival Signaling Pathways: Ras-Raf-MEK-ERK Pathway 32

Aberrations of Ras-Raf-MEK-ERK Signaling in Cancer 32

Ras-Raf-MEK-ERK signaling 33

p90 Ribosomal S6 Kinase (RSK) 38

1.5 ANTI-APOPTOTIC MECHANISMS: REDOX REGULATION OF

1.5.1 Sources of Reactive Oxygen Species (ROS) and Redox Balance .45

1.5.2 ROS in Cell Signaling .48

1.5.3 ROS in Tumorigenesis .54

1.5.4 The Pro-Survival Role of Superoxide anion in Cancer 55

1.6 PROSTATE CANCER 59 1.6.1 Prostate Cancer Development and Progression .59

1.6.2 Survival Signals in Prostate Cancer Development and Progression .63

1.6.3 Growth Factor Signaling and Survival in Prostate Cancer 65

1.7 AIM OF STUDY 66 CHAPTER 2: MATERIALS AND METHODS 68 2.1 MATERIALS 68 2.1.1 Chemicals 68

2.1.2 Antibodies 69

2.1.3 Plasmids used in the study 70

2.1.4 Cell lines and cell culture 71

2.2 METHODS 71 2.2.1 Treatment of Cells 71

2.2.2 Cell Viability Assay (Crystal Violet Assay) .72

2.2.3 DNA Fragmentation Assay 73

2.2.4 Caspase Activity Assay 74

2.2.5 Transient Transfection 75

2.2.6 β-galactosidase Survival Assay 76

2.2.7 SDS-PAGE and Western Immunoblotting 77

2.2.8 Intracellular Superoxide Measurement .78

2.2.9 RNA Interference (RNAi) Assay .79

2.2.10 Subcellular Fractionation .79

2.2.11 In vitro Akt Kinase Assay .81

2.2.12 Immunofluorescence Assay for Bax Activation .82

2.2.13 Statistical Analysis .83

CHAPTER 3: RESULTS 84 3.1 GROWTH-FACTOR REGULATION OF CELL SURVIVAL 84 3.1.1 PI3K-Akt pathway is the major survival pathway in serum-deprived LNCaP cells 84

3.1.2 Serum and Epidermal Growth Factor activate an alternative survival mechanism that is PI3K-independent 93

3.1.3 EGF but not serum-mediated survival is MEK-dependent .99

3.1.4 EGF inhibits LY-induced apoptosis via inactivation of Bad .106

3.1.5 Serum-mediated inhibition of LY-induced cell death is independent of Bad inactivation 114

3.1.6 LY-mediated apoptosis is Bad-dependent 116

3.1.7 RSK1 is the Bad kinase activated by EGF in LNCaP cells .121

3.1.8 Role of ErbB receptors in EGF- and serum-mediated survival of

LNCaP cells 130

3.1.9 Bax is required for LY-mediated apoptosis in LNCaP cells 138

3.1.10 Serum promotes cell survival in LY-treated LNCaP cells via

inhibition of Bad and Bax translocation 141

3.2 REACTIVE OXYGEN SPECIES REGULATION OF CELL

SURVIVAL 147 3.2.1 Intracellular superoxide anions modulate serum’s inhibition of

LY-induced cell death .147

3.2.2 Intracellular superoxide and activation of MEK-ERK-RSK

pathway .153

3.2.3 Bad phosphorylation is regulated by intracellular O2·− level .158

4.1 EGF AND SERUM ACTIVATE PI3K-AKT-INDEPENDENT

4.2 EGF-MEDIATED SURVIVAL IS DEPENDENT ON MEK-ERK

4.3 EGF-MEDIATED SURVIVAL REQUIRES EGFR’S TYROSINE

4.4 PHOSPHORYLATION AND INACTIVATION OF BAD IS AN

IMPORTANT MECHANISM OF GROWTH FACTOR MEDIATED

4.5 BAD EXPRESSION REGULATES CANCER CELLS

4.6 SERUM-MEDIATED SURVIVAL IS INDEPENDENT OF

MEK-ERK- AND PI3K-AKT-INDEPENDENT PATHWAY IN LNCAP

CELLS 174 4.7 CROSSTALK BETWEEN PI3K-AKT PATHWAY AND MEK-ERK

4.8 PRESENCE OF SERUM INDUCES A NON-CONDUCIVE

ENVIRONMENT FOR TRANSLOCATION OF BAD AND BAX TO

4.9 INCREASED LEVEL OF SUPEROXIDE ANION PROMOTES

References 184

Summary

Understanding the mechanisms behind tumor cells ability to evade cell death when confronted with multiple apoptotic signals during the course of cancer development and progression as well as during treatment with anti-cancer drugs, is of key importance towards development of efficient targeted therapy for different types

of cancer In prostate cancer (PCa), one of the most common mutations found is inactivation mutation of PTEN (phosphatase and tensin homologue deleted on chromosome 10), resulting in constitutive activation of PI3K-Akt signaling, recognized as a major survival pathway in PCa cells However, there is increasing evidence supporting the existence of PI3K-Akt-independent survival pathways in PCa Deregulation of growth factor signaling is often observed during the course of PCa, and is proposed to gain importance as the tumor progresses towards androgen-independence In this study, we provide evidence for the role of growth factors- and serum-mediated activation of PI3K-Akt-independent survival signaling in PCa Using LNCaP prostate cancer cell line which harbors a PTEN frameshift mutation, we showed that EGF activated the MEK-ERK signaling pathway to promote LNCaP survival independently of PI3K-signaling Inhibition of apoptosis by EGF was mediated mainly through EGF’s ability to phosphorylate the pro-apoptotic BH3-only Bcl-2 protein, Bad, at Ser75, which has been shown to sequester the protein in the cytosol, preventing Bad from antagonizing pro-survival Bcl-2 functions Moreover we demonstrated that RSK1 as the kinase activated downstream of MEK-ERK signaling responsible for phosphorylating Bad Using siRNA strategy, we demonstrated that silencing Bad inhibited apoptosis similar to the level afforded by EGF, whereas silencing of Bax, a multidomain pro-apoptotic Bcl-2 protein, completely inhibited

apoptosis, supporting the role of Bad as an “enabler” and Bax as an “effector” of

apoptosis in LNCaP cells

Moreover, we show that serum-mediated survival, unlike EGF, was

independent of MEK-ERK signaling Although serum also phosphorylated Bad on

Ser75, it was sensitive to inhibition of PI3K signaling and not MEK signaling,

implying that serum phosphorylation of Bad was not the mechanism behind

serum-mediated survival under those conditions We proceeded to demonstrate that serum

inhibited translocation of both Bad and Bax to the mitochondria in a

PI3K-independent manner, which likely accounts for serum-mediated survival Additionally

we show that while EGF transmits it survival signals through EGFR tyrosine kinase

activation (not ErbB2), serum-mediated survival signaling did not require EGFR or

ErbB2 tyrosine activity

Previous studies in our lab demonstrated the role of increased O2·− in

inhibition of apoptosis by diverse apoptotic triggers in tumor cells While others have

shown growth factors and serum increase production of O2·−via NADPH oxidase, we

found that serum did not induce significant increase in O2·−levels in LNCaP cells

However, when LNCaP’s steady-state level of O2·−was decreased using an inhibitor

of NADPH oxidase, DPI, serum-mediated survival was abrogated, while increasing

O2·− levels using DDC an inhibitor of superoxide dismutase, protected the cells

Interestingly, we also show that phosphorylation of Bad and ERK1/2 was sensitive to

regulation by O2·− levels However, further studies are required to elucidate the

molecular targets of O2·−in promoting survival as well as their regulation of Bad and

ERK1/2

List of Figures

Figure I: Bcl-2 family members 11

Figure II: Schematic diagram of Growth Factor-mediated survival signaling via PI3K-Akt pathway .30

Figure III: Domain structure and regulatory phosphorylations sites of RSK1 .39

Figure IV: Schematic diagram of Growth Factor-mediated survival signaling via Ras-Raf-MEK-ERK pathway .44

Figure V: Production of ROS in cells .48

Figure 1: PI3K-Akt pathway is constitutively activated in LNCaP in the absence of growth factors .86

Figure 2: Restoring a functional PTEN sensitizes LNCaP cells to cell death in the absence of growth factors .88

Figure 3: LY294002, a specific inhibitor of PI3K, sensitizes LNCaP cells to cell death in the absence of growth factor 91

Figure 4: LY-mediated cell death is caspase-3- and caspase-9-dependent but not caspase-8 92

Figure 5: Serum and EGF increased viability of LY-treated LNCaP cells 94

Figure 6: Serum and EGF decrease LY-mediated DNA fragmentation 95

Figure 7: Serum and EGF decrease LY-mediated caspase-3 and caspase-9 activation 96

Figure 8: Serum and EGF does not activate PI3K-Akt pathway in the presence of LY294002 .97

Figure 9: EGF induces robust and sustained ERK phosphorylation in LNCaP cells.100 Figure 10: EGF- and serum induced-ERK phosphorylation in LNCaP cells is MEK-dependent .102

Figure 11: EGF but not serum-mediated decrease in DNA fragmentation is inhibited by U0126, a specific inhibitor of MEK1/2 .104

Figure 12: EGF but not serum-mediated decrease in caspase-3 activation is inhibited by U0126 105

Figure 13: Serum and EGF does not alter Bcl-2 or Bcl-xL protein expression 108

Figure 14: Serum induces phosphorylation of endogenous Bad at Ser75 .109

Figure 15: LY294002 induces total dephosphorylation of Bad 110

Figure 16: EGF activation of the MEK-ERK pathway leads to strong phosphorylation of Bad 112

Figure 17: Ser75 is the major phosphorylation site of Bad by EGF 113

Figure 18: Serum-mediated phosphorylation of Bad is dependent on PI3K but not MEK activity 115 Figure 19: Bad is required for LY-mediated cell death execution in LNCaP cells 117 Figure 20: EGF prevents LY-mediated translocation of Bad to the mitochondria 119 Figure 21: RSK is strongly phosphorylated by EGF via MEK-dependent pathway 122 Figure 22: RSK1 not RSK2 is the major Bad kinase activated by EGF in LNCaP cells 124 Figure 23: Silencing RSK1 not RSK 2 attenuates EGF-mediated inhibition of

apoptosis .125 Figure 24: Silencing RSK1 or RSK2 does not attenuate serum-mediated inhibition of apoptosis .126 Figure 25: Transfection of dominant-negative RSK1 decreases phosphorylation of Bad by EGF 128 Figure 26: Transfection of dominant-negative RSK1 attenuates EGF-mediated

inhibition of apoptosis 129 Figure 27: Dose-response of LNCaP cell viability and caspase-3 activation to ErbB receptor kinase inhibitors 133 Figure 28: Effects of AG1478 and AG879 on EGFR and ErbB2 phosphorylation by EGF and serum .134 Figure 29: Inhibition of EGFR and ErbB2 activity do not prevent serum-mediated inhibition of LY-induced death in LNCaP cells .137 Figure 30: Bax is required for induction of apoptosis in LY-treated LNCaP cells 140 Figure 31: LY-induced initial Bax translocation to the mitochondria is inhibited by serum but not EGF .143 Figure 32: LY-induced initial Bax activation is inhibited by serum but EGF 145 Figure 33: DPI decrease O2·− level and abrogates serum’s protection in LY-induced apoptosis .149 Figure 34: DDC increase intracellular O2·− concentration in LNCaP cells .151 Figure 35: DDC-mediated increase intracellular O2·− reverts sensitization to apoptosis induced by DPI .152 Figure 36: Activation of MEK-ERK pathway is regulated by intracellular superoxide 156 Figure 37: DDC increases activation of MEK-ERK-RSK signaling cascade in a dose-dependent manner .157 Figure 38: DPI induces Bad dephosphorylation .159 Figure 39: DPI-induced caspase-3 activation in LY-treated LNCaP cells requires Bad 160 Figure 40: DDC induces Bad phosphorylation 160 Figure 41: PMA-induced O2·− production induces Bad phosphorylation and cell

List of Tables

Table 1: Properties of the members of the caspase family 3

Abbreviations

AFC 7-amino-4-trifluoromethyl coumarin

Bad Bcl-2-antagonist of cell death

Bax Bcl-2-associated X protein

BCECF-AM

2’,7’-bis-(2-carboxyethyl)-5-(and-6-)-carboxyfluorescein-acetoxymethyl Bcl-2 B-cell CLL/lymphoma 2

Bcl-xL Bcl-x long form

Bid BH3-interacting domain agonist

Bim Bcl-2 interacting mediator of cell death

BSA Bovine serum albumin

CTKD C-terminal kinase domain

DDC Diethyldithiocarbamate

DTT Dithiothreitol

EDTA Ethylenediamine tetraacetic acid

EGF Epidermal growth factor

EGFR Epidermal growth factor receptor

EGTA Ethyleneglycol tetraacetic acid

ERK Extracellular signal-regulated kinase

FBS Fetal bovine serum

HBSS Hank’s balanced salt solution

HEPES N-(2-hydroxylethyl)piperazine-N’-(2-ethanesulfonic acid)

IAP Inhibitor of apoptosis protein

LNCaP Lymph node metastasis of prostate cancer

LY LY294002

MAPK Mitogen-activated protein kinase

Mcl-1 Myeloid cell leukemia 1

MEK Mitogen/extracellular-signal regulated kinase kinase

MOM Mitochondria outer membrane

MOMP Mitochondria outer membrane permeabilization

NADPH Nicotinamide adenine dinucleotide phosphate (reduced)

NF-κB nuclear factor immunoglobulin κ chain enhancer-B cell

NTKD N-terminal kinase domain

PCa Prostate cancer

PCD Programmed cell death

PIPES Piperazine-N,N′-bis(2-ethanesulfonic acid)

PMA Phorbol 12-myristate 13-acetate

RFU Relative fluorescence unit

RLU Relative luminescence unit

ROS Reactive oxygen species

siRNA Small interfering ribonucleic acid

“apoptosis” was coined by Kerr et al to describe the morphological processes

observed in controlled cell suicide (Kerr et al 1972)

Apoptosis can be distinguished from other forms of cell death by a characteristic pattern of morphological and molecular changes Some of the key morphological features include cell shrinkage, chromatin condensation and margination to the nuclear periphery, plasma membrane blebbing and fragmentation

of the cell into multiple, compact, membrane-bound ‘apoptotic bodies’ (Kerr et al 1972; Skalka et al 1976; Arends et al 1990; Eastman et al 1994) During the early stages there appears to be preservation of structure of most cellular organelles with the exception of the mitochondria (Vander Heiden et al 1997) On the molecular level, there is internucleosomal DNA cleavage (Wyllie 1980) and exposure of

phagocytosis (Fadok and Henson 1998) Techniques based on detecting some of these changes are now standard tools for demonstrating apoptosis

In contrast to the tightly regulated apoptotic cell death which is an active process requiring energy, cell death by necrosis is an accidental passive process in response to gross injury The early events of necrosis features swelling of the cell and organelles followed by rupture of the plasma membrane with little DNA degradation Unlike apoptosis, where apoptotic bodies are engulfed by macrophages and rapidly removed, in necrosis, cellular contents are released into the cell’s environment triggering an inflammatory response and damaging surrounding cells (Fiers et al 1999) It should be mentioned that not all of these morphological criteria in the strictest sense are seen in all cell types during cell death While apoptosis is possibly the most commonly occurring PCD, many forms of ‘apoptosis-like’ and ‘necrosis-like’ PCD have been reported over the years (Leist and Jaattela 2001) Many of these other forms of PCDs are less well-defined but undoubtedly they hold biological significance and likely clinical significance in treatment of disease

1.1.2 Molecular mechanisms of apoptosis: Caspases as the central executioner

of Apoptosis

Of greater interest is that all the typical signs of apoptosis are the consequence

of a complex biochemical cascade of events Once apoptosis is triggered, regardless

of the origin of the death signal, a common cell death machinery is eventually activated where a family of cysteine proteases called caspases takes center stage (Nicholson and Thornberry 1997) Caspases are evolutionarily conserved and can be found in mammals, all the way down to insects and nematodes (Yuan et al 1993; White et al 1994; Nicholson and Thornberry 1997) In fact it was the identification of

a set of genes dedicated to regulation of apoptosis in the nematode Caenorhabditis

elegans (Ellis and Horvitz 1986), followed by discovery of their counterparts in

humans (Yuan et al 1993; Hengartner and Horvitz 1994) that launched the field of

apoptosis Studies on C elegans provided important insights into how the core

apoptotic machinery executes cell death Two genes identified, ced-3 and ced-4, were

found to be absolutely necessary for apoptosis to occur ced-3 was found to encode

for a protein similar to the mammalian interleukin 1β converting enzyme (ICE) (Yuan

et al 1993) Soon more members of the CED-3/ICE family of proteases were

identified, and up to date there are at least 14 caspases identified in mammals (see

Table 1), with half of them having roles in apoptosis The term ‘caspase’ was later

adopted for all members of this family (Alnemri et al 1996)

m= murine b= bovine

Table 1: Properties of the members of the caspase family

(Adapted from Vermeulen et al 2005 Ann Hematol 84:627-639 and Philchenkov et al

2004 Exp Oncol 26,2:82-97)

Caspase Other names Tetrapeptide

preferance Prodomain size Prodomain motifs Function

initiator Caspase-8 MACH/FLICE/

initiator

initiator Caspase-3 CPP32/Apopain/

TX/ICH-2/ICE REL -II

LEVD/(W/L)EHD Long CARD Inflammation

Activation of caspases

All known caspases possess a cysteine residue at their active-sites which is crucial for their catalytic activity; and they cleave their substrate after an aspartic acid residue (Asp-X sites) within a tetrapeptide recognition motif (see Table 1), hence the name ‘caspase’ (cysteine aspartate-specific proteases) (Thornberry and Lazebnik 1998) The recognition motifs differ significantly among the caspases and in part confer substrate specificity It also explains their ability to perform diverse biological functions Caspases-1, -4, -5, -11 and -12 seems to be mainly involved in the regulation of inflammatory response and little in apoptosis execution (Vermeulen et

al 2005) In contrast, gene-knockout studies have shown that caspase-3, -8, -9, -2, -6, -7 and -10 have been shown to play an important role in apoptosis signaling and execution (Kuida et al 1996; Kuida et al 1998; Thornberry and Lazebnik 1998; Varfolomeev et al 1998)

Like most proteases, caspases are synthesized in the cell as inactive zymogens called procaspase, which consists of a prodomain at the N-terminal, followed by a large (~20 kDa), p20 and small (~10 kDa), p10 subunit They can be activated rapidly

by proteolytic cleavage of the region between the p20 and p10 domains and also removal of prodomain, to form the mature and active caspase which is usually a heterotetramer consisting of two p20/p10 heterodimers (Earnshaw et al 1999) The cleavage sites on the procaspases are themselves Asp-X sites, indicating that caspases are capable of being activated by autoproteolysis Indeed caspases involved in apoptosis can be divided into two groups: the ‘initiator’ caspases which includes caspase-2, -8, -9 and -10 with long prodomains (more than 90 amino acid residues); and the ‘effector’ caspases, including caspase-3, -6, and -7 with short prodomains (20-

30 amino acid residues) The long prodomain in initiator caspases contains

structurally related motifs; the DED (death effector domain) in caspase-8 and -10, and the CARD (caspase recruitment domain) in caspase-2 and -9 During apoptosis, interactions between DED or CARD motifs on the initiator caspase prodomain with similar motifs on adaptor proteins lead to activation of the procaspase Activated initiator caspases then proteolytically cleaves and activates downstream effector caspases (caspase-3, -6, -7) which proceed to execute the apoptosis program by cleaving vital cellular proteins This ‘caspase signaling cascade’, beginning with the activation of initiator caspases, not only integrates upstream apoptotic signals but also allows for amplification of caspase activity (Thornberry et al 1997; Hirata et al 1998; Slee et al 1999)

Extrinsic and Intrinsic Apoptotic Pathway

Two major pathways of caspase activation have been characterized, namely the extrinsic pathway and the intrinsic pathway Apoptosis induced by aggregation of cell surface death receptors (extrinsic pathway) like the CD95/Fas, tumor necrosis factor receptor-1 (TNFR1) and TRAIL receptors DR4 and DR5, is initiated mainly by caspase-8 In the case of CD95/Fas receptor, upon binding of the CD95/Fas ligand, the death receptor oligomerizes and recruits an adaptor protein, the Fas-associated protein with death domain (FADD) via interaction of the death domain (DD) located

on both FADD and the cytoplasmic tail of the death receptor FADD also contain another important domain, the death effector domain (DED), through which it recruits procaspase-8 via homologous interaction with another DED in procaspase-8’s prodomain region (Boldin et al 1996) Together they form the death-inducing signaling complex (DISC) which serves as a platform to bring procaspase-8 in close

procaspase-8 dimerization and activation due to their low intrinsic proteolytic activity (Muzio et al 1998; Boatright et al 2003) Active caspase-8 in turn cleaves effector caspases like caspase-3, activating the downstream caspase signaling cascade (Stennicke et al 1998) Alternatively, cellular stress or DNA damage caused by various stimuli including cytotoxic agents, UV irradiation, oxidative stress, growth factor withdrawal and aberrant oncogene expression (Fearnhead et al 1998; Soengas

et al 1999; Kaufmann and Earnshaw 2000; Wang 2001) mediate caspase activation via the intrinsic pathway When the cell senses these cellular stress or damage, it activates caspases from within the cell to eliminate itself The mitochondria play a central role in the integration and propagation of the intrinsic apoptotic signals leading

to caspase activation In most cases, these cellular stress eventually lead to mitochondrial dysfunction like loss of mitochondrial membrane potential and changes

in mitochondrial membrane permeability which causes the release of pro-apoptotic factors such as cytochrome c from the mitochondria intermembrane space into the cytoplasm (Bernardi et al 1999; Loeffler and Kroemer 2000) Once in the cytoplasm,

cytochrome c activates Apaf-1 (apoptosis protease–activating factor 1), the C elegans death gene ced-4 homologue In the presence of dATP/ATP, Apaf-1 undergoes

conformational change and forms a heptameric complex allowing for the recruitment

of procaspase-9 via interaction of their respective caspase recruitment domain (CARD) (Li et al 1997; Srinivasula et al 1998; Acehan et al 2002) This mulitcomponent complex of approximately 700 -1400 kDa, called the apoptosome (Cain et al 2002), provides a high concentration and proper protein conformation suitable for activation of procaspase-9 Activated caspase-9 goes on to cleave and activate effector caspase like caspase-3 and caspase-7

Substrates of caspases

Once activated, effector caspases cleaves specific proteins to begin the degradation phase that gives rise to the typical apoptotic morphology Caspases contribute to disassembly of the cellular structure for example by cleaving nuclear lamins leading to nuclear shrinking and chromatin condensation (Orth et al 1996; Rao et al 1996) Caspases also cleave several proteins involved in cytoskeleton organization, including gelsolin (Kothakota et al 1997), p21-activated kinase 2 (PAK2) (Rudel and Bokoch 1997) and focal adhesion kinase (FAK) (Wen et al 1997), leading to membrane blebbing and changes in cell shape Cleavage of the inhibitor of caspase-activated DNAse (CAD), iCAD, by caspase, frees the active endonuclease to translocate to the nucleus and degrade nuclear DNA, producing the characteristic internucleosomal DNA fragmentation observed in apoptosis (Sakahira

et al 1998) Caspase activation also inactivates or deregulates proteins involved in DNA repair like poly-(ADP-ribose) polymerase (PARP) (Duriez and Shah 1997) Other proteins involved in cell cycle regulation, transcription and cell signaling have also been reported (Vermeulen et al 2005)

1.2 BCL-2 FAMILY

1.2.1 Role of Mitochondria in Apoptosis

The mitochondria play an essential role in apoptotic cell death in mammalian cells Besides mediating apoptosis in the intrinsic pathway, it also amplifies the extrinsic apoptotic pathways in certain cell types, making it the point of convergence for both pathways When the mitochondrion senses the proper apoptotic signals, it undergoes several structural and morphological changes that culminate in the release

of apoptogenic factors like cytochrome c from the mitochondrial intermembrane space into the cytosol to trigger caspase activation (Martinou and Green 2001; Zamzami and Kroemer 2001)

The mitochondrial event that appears vital to ensure cell death is the permeabilization of the mitochondrial outer membrane (MOMP) (Kroemer 2002) which precedes the release of the apoptogenic factors The precise mechanisms of MOMP are still much debated on; however two prominent models have been proposed The first model for MOMP involves the mitochondrial permeability transition which refers to an abrupt transition in permeability of the inner mitochondrial membrane to solutes up to 1500 Da, through formation of the permeability transition pore (PTP) (Haworth and Hunter 1979; Hunter and Haworth 1979a; Hunter and Haworth 1979b) The PTP multiprotein complex is believed to comprise of the voltage-dependent anion channel (VDAC) in the outer membrane, the soluble matrix protein cyclophilin D (CyD), and the adenine nucleotide translocase (ANT) in the inner membrane, forming a “megachannel” that spans the contact sites between the inner and outer mitochondrial membranes (Zoratti and Szabo 1995; Crompton 1999; Kuwana and Newmeyer 2003) Certain pro-apoptotic stimuli such as increased Ca2+ levels or oxidative stress induces the opening of PTP, allowing influx

of water and ions into the mitochondria matrix, causing the loss of mitochondrial

membrane potential (∆Ψm), uncoupling of oxidative phosphorylation and matrix

swelling This leads to mechanical disruption of the mitochondrial outer membrane and subsequent release of apoptogenic factors into the cytosol (Green and Reed 1998; Kuwana and Newmeyer 2003) However, there is increasing evidence questioning the general view that permeability transition is fundamental for MOMP Some of the classic features of permeability transition such as mitochondria matrix swelling do not always occur in apoptosis (Jurgensmeier et al 1998; De Giorgi et al 2002), moreover,

in some cases, cytochrome c release and caspase activation occurs before any detectable loss of ∆Ψm (Bossy-Wetzel et al 1998; von Ahsen et al 2000; Waterhouse

et al 2001), implying that permeability transition is not absolutely necessary for caspase activation and apoptosis to occur

The second model of MOMP is highly dependent on the Bcl-2 family proteins The pro-apoptotic members of the Bcl-2 family, Bax and Bak, have been shown to be essential for apoptosis since cells doubly-deficient in these two proteins do not undergo MOMP and are resistant to cytochrome c release induced by multiple apoptotic stimuli (Wei et al 2001; Zong et al 2001; Degli Esposti and Dive 2003) It

is proposed that the BH3-only members of the Bcl-2 family relay apoptotic signals from various sources to activate Bax/Bak, thus inducing their homo-oligomerization

to form pores in the mitochondrial outer membrane large enough for the release of apoptogenic factors The hypothesis was borne from the observation that Bcl-2 family protein share structural similarities with the pore-forming domains of bacterial toxins (Muchmore et al 1996; Sattler et al 1997; Suzuki et al 2000) Consistent with this, are the findings that Bax oligomers are capable of pore formation in artificial membranes (Antonsson et al 1997; Antonsson et al 2000; Saito et al 2000) More

recently, Kuwana et al demonstrated using vesicles reconstituted from isolated

mitochondrial membrane or defined liposomes, that activated or oligomerized Bax alone is capable of forming openings in these membranes that can allow passing of large molecules (up to 2 MDa), without the need for other mitochondrial proteins (Kuwana et al 2002)

However it is possible that Bcl-2 family proteins regulate MOMP by interacting with the proteins involved in permeability transition Interaction between Bax and ANT (Marzo et al 1998) and also VDAC (Shimizu et al 1999; Adachi et al 2004) to mediate apoptosis have been reported Moreover mitochondria isolated from Bcl-2 transfected cells have been shown to be resistant to mitochondria permeability transition (Susin et al 1996) It is certainly conceivable that more than one mechanism may collaborate simultaneously or sequentially in permeabilizing the mitochondrial outer membrane While the mechanism remains controversial, nevertheless, Bcl-2 family proteins seem to play a prominent role in regulation of release of apoptogenic factors from mitochondria and consequently apoptosis regardless of the mechanism involved

1.2.2 Bcl-2 family proteins

The Bcl-2 family proteins are key players in the initiation of the apoptosis machinery as they regulate the release of apoptogenic factors from the mitochondria Presently, there are at least 20 members in the Bcl-2 family of proteins in mammalian cells All members of the Bcl-2 family contain one to four regions of high amino acid sequence homology to the Bcl-2 protein, known as the Bcl-2 homology (BH) domains (BH1-4) (see Figure I) They can be divided into two main groups – the anti-apoptotic

or pro-survival members comprising of Bcl-2, Bcl-xL (Boise et al 1993), Bcl-w

(Gibson et al 1996), Mcl-1 (Kozopas et al 1993) and A1 (Choi et al 1995), which has three to four BH domains; and the pro-apoptotic members, which can be further divided into 2 subgroups The first pro-apoptotic subgroup, usually referred to as the multidomain pro-apoptotic members, comprise of Bax, Bak and Bok They are structurally similar to Bcl-2 and possess BH1, BH2 and BH3 domains The second subgroup named the BH3-only members as it bears only the BH3 domain includes Bad, Bid, Bim, Bik, Noxa, Puma, Bmf and Hrk (Bouillet and Strasser 2002) The BH domains are important for the Bcl-2 family proteins’ function as well as for heterodimerization among different members of the family (Yin et al 1994; Cheng et

al 1996a; Adams and Cory 1998; Gross et al 1999) Additionally, most members also feature a hydrophobic transmembrane (TM) domain at the C-terminal, which most likely enables membrane localization (Nguyen et al 1993) Proteins in the different groups are capable of forming either homo-oligomers or hetero-oligomers with one another Upon apoptotic stimulation, the pro-survival and pro-apoptotic members of the Bcl-2 family interact with each other to determine the cell’s fate

Figure I: Bcl-2 family members Members of the Bcl-2 family are divided into three

groups- the pro-survival Bcl-2 group, and the pro-apoptotic Bax-like group and only group They are characterized by the existence of one to four of the conserved Bcl-2 homology domains (BH1-BH4) Most of the members also possess a

BH3-hydrophobic region (TM) for membrane localization at the C-terminal

The Multidomain Pro-survival proteins

The first member of this family, Bcl-2, is a proto-oncogene that was identified

at the chromosomal breakpoint between chromosomes 14 and 18, t(14;18) in human follicular B-cell lymphomas (Tsujimoto et al 1984a; Bakhshi et al 1985; Tsujimoto

et al 1985; Cleary et al 1986; Tsujimoto and Croce 1986) Unlike other oncogenes at that time, Bcl-2 expression did not promote tumorigenesis by inducing cell proliferation Instead its primary mode of action seemed to be the prevention of cell death when challenged with various cytotoxic stimuli (Vaux et al 1988; McDonnell

et al 1989) In fact, these findings pushed the field of apoptosis center stage in cancer research as it became clear that tumorigenesis is not merely due to excessive cell proliferation but also impairment of apoptosis

Members of the pro-survival group have been found localized mostly on the membranes of cellular organelles like the mitochondrial membrane, the endoplasmic reticulum (ER) and the nuclear envelope (Krajewski et al 1993; de Jong et al 1994) Membrane targeting is most likely enabled by a highly hydrophobic transmembrane (TM) region located at the carboxy-terminal, anchoring the protein to the membrane

on the cytoplasmic face (Nguyen et al 1993; Kaufmann et al 2003) Some members

of this group like Bcl-xL and Bcl-w, are found both in the cytosol and mitochondria membrane They reportedly associate tightly with the mitochondrial membrane upon receiving apoptotic signal which triggers a conformational change (Hsu et al 1997b; Wilson-Annan et al 2003) This localization of the pro-survival Bcl-2 family members to mitochondria membrane following an apoptotic signal is consistence with their main function which appears to be protecting the integrity of the outer mitochondrial membrane thus preventing the release of cytochrome c and caspase activation (Green and Reed 1998; Gross et al 1999) The ability of pro-survival Bcl-2

members to heterodimerize with pro-apoptotic members appears to be important for their anti-apoptotic properties (Yin et al 1994; Sedlak et al 1995) Crystallography studies of some pro-survival Bcl-2 members have revealed similar core three-dimensional structure among them (Muchmore et al 1996; Petros et al 2001; Petros

et al 2004) that is important for their heterodimerization with other members of Bcl-2 family Notably, the hydrophobic groove formed by residues from BH1, BH2 and BH3 domain has been shown to be capable of binding an exposed BH3 α-helix of pro-apoptotic Bcl-2 family members (Sattler et al 1997) Heterodimerization of pro-survival Bcl-2 members with pro-apoptotic members neutralizes them and prevents their aggregation at the mitochondrial membrane; however the exact mechanism is still contentious

The Multidomain Pro-apoptotic proteins

The multidomain pro-apoptotic members of Bcl-2 family, also sometimes referred to as the “Bax-like” pro-apoptotic proteins, consists currently of Bax, Bak and Bok While Bax and Bak are widely distributed (Krajewski et al 1994; Krajewski

et al 1996); the less-known Bok expression is more limited to reproductive tissues (Hsu et al 1997a) Knockout studies in mice revealed that while Bax and Bak show partial functional redundancy in many cell types, they are absolutely required for inducing apoptosis triggered by various death stimuli via the mitochondrial pathway (Lindsten et al 2000; Wei et al 2001; Zong et al 2001) In healthy cells, Bax resides

in the cytosol as soluble monomeric protein while Bak is found on the outer membrane of the mitochondria and endoplasmic reticulum in an inactive state (Wei et

al 2000) Upon receiving apoptotic signals, Bax undergoes a conformational change

the mitochondrial membrane eventually leading to permeabilization of the outer

mitochondrial membrane (Hsu and Youle 1997; Wolter et al 1997; Antonsson et al

2001) In its inactive state, Bax’s C-terminal transmembrane domain required for

membrane-docking, is folded into its hydrophobic groove formed by BH1, BH2 and

BH3 domains, similar to that of Bcl-xL and Bcl-2 This may explain its cytosolic

localization (Suzuki et al 2000; Schinzel et al 2004) In addition, the N-terminal of Bax, in particular the first 20 amino acids, has been proposed to work together with the C-terminal to regulate mitochondria targeting (Goping et al 1998; Nechushtan et

al 1999; Cartron et al 2002)

The exact trigger mechanism leading to Bax activation is still unclear although several proteins have been implicated in regulating Bax activation Among these are the Bax inhibitors - Ku70 (Sawada et al 2003), protein 14-3-3 (Nomura et al 2003a) and Humanin (Guo et al 2003), which prevent Bax activation by directly binding thus sequestering the protein in the cytosol Also certain BH3-only protein like t-Bid (the active truncated form of Bid) and possibly Bim, have been reported to have the ability

to directly induce Bax activation (Desagher et al 1999; Korsmeyer et al 2000; Letai

et al 2002; Marani et al 2002; Kuwana et al 2005) More controversially are the proposed regulation of Bax via phosphorylation by cellular kinase like GSK-3β and Akt (Gardai et al 2004; Linseman et al 2004); and alterations in intracellular pH (Khaled et al 1999; Tafani et al 2002; Ahmad et al 2004)

Activation of Bax involves a conformational change in the protein that exposes cryptic epitopes at the N-terminus (Hsu and Youle 1998; Nechushtan et al 1999) The conformational change in activated Bax can be detected using antibodies against epitope 6A7, consisting of amino acids 13 to 19, a method commonly used to differentiate between the inactive and active state As this event appears to coincide

with Bax translocation to the mitochondria membrane, it has been postulated that the change in conformation may play a role in translocation by uncovering mitochondrial targeting sequences (Schinzel et al 2004; Cartron et al 2005) Once at the outer mitochondrial membrane, Bax oligomerizes into large complexes that can be detected

by confocal and electron microscopy (Nechushtan et al 2001; De Giorgi et al 2002) Bax oligomerization is believed to be required for mitochondrial membrane permeabilization by forming pores on the membrane, as they have structurally similar pore-forming domains as those found in bacterial toxins (Schendel et al 1998) Although the exact nature and size of these pores are still unknown, Bax oligomers are able to form pores in liposomes and isolated mitochondria large enough for cytochrome c release (Antonsson et al 2000; Saito et al 2000; Kuwana et al 2002) leading to caspase activation and apoptosis

The BH3-only Pro-apoptotic proteins

The BH3-only members of the Bcl-2 family are named so because they share only a short region of homology (9-16 amino acids), constituting the BH3 domain, with the other members of the family Their BH3 domain is important for their pro-apoptotic function and is required for binding with the multidomain members of the Bcl-2 family (Huang and Strasser 2000) Structural studies have demonstrated that the amphipathic α-helix BH3 domain of the BH3-only protein is inserted into a hydrophobic pocket on the surface of pro-survival Bcl-2 proteins (Sattler et al 1997) There are at least 8 members known currently in mammals, and they are believed to act as cellular sensors that relay diverse stress signal to initiate the apoptotic process (Huang and Strasser 2000) In healthy mammalian cells, BH3-only proteins are

mechanisms, which includes transcriptional and post-translational modifications Regulation of Puma, Noxa, Hrk5 and Bim appear to be subjected to transcriptional control whereby upon encountering a specific stress trigger, BH3-only protein expression is increased Other BH3-only proteins like Bad, Bid, Bik and Bmf are more readily detected in healthy cells and their activity are subjected to post-translational regulation Full-length Bid is inactive and needs to undergo proteolytic cleavage for example by caspase-8, resulting in an active truncated protein, tBid Proteolytic cleavage exposes its BH3 domain for interaction with other Bcl-2 relatives and also allows it to be myristoylated, facilitating its movement to the mitochondria(Li et al 1998; Luo et al 1998; Zha et al 2000) Bmf and Bim are sequestered through binding to components of the cytoskeleton, the myosin V motors and the dynein motor complex respectively (Puthalakath et al 1999; Puthalakath et al 2001) and are released upon receiving the proper stress cues Bad and Bik are regulated through phosphorylation In healthy cells, growth factors signaling leads to phosphorylation of Bad by various kinases which allows the 14-3-3 scaffold protein to bind and sequester Bad in the cytoplasm Growth factor withdrawal leads to Bad dephosphorylation, allowing it to detach from 14-3-3 and migrate to the mitochondria

to antagonize the pro-survival Bcl-2 members (Zha et al 1996) At the mitochondria, binding of Bad to Bcl-2 or Bcl-xL via its BH3 domain interferes with their pro-survival function and sensitizes the cell to apoptosis (Yang et al 1995)

It is possible that these diverse modes of activation of the multiple BH3-only proteins allows for some signaling specificity Knock-out studies of Bid and Bim (Bouillet et al 1999; Yin et al 1999) and expression analysis studies (Oda et al 2000; Villunger et al 2000; Nakano and Vousden 2001; Puthalakath et al 2001) indicate that there is a degree of selectivity in the activation of BH3-only whereby different

BH3-only proteins are more essential than others in inducing apoptosis in different cell types, and different BH3-only proteins are activated in response to different apoptotic stimuli For example Bim seems to be required for initiating apoptosis in lymphocytes expose to cytokine deprivation or calcium flux but not to γ-irradiation (Bouillet et al 1999); Bmf is activated in response to anoikis (Puthalakath et al 2001); and Puma and Noxa protein expression is induced by p53 in response to DNA damage (Oda et al 2000; Nakano and Vousden 2001) However, activation of BH3-only proteins cannot trigger apoptosis in the absence of multidomain pro-apoptotic members as shown in knockout studies of Bax and Bak (Cheng et al 2001a; Zong et

al 2001), placing them upstream of Bax/Bak in apoptotic signal transduction

1.2.3 Interactions Among Bcl-2 family members

Once the BH3-only proteins are activated by apoptotic trigger, how they interact with their multidomain Bcl-2 relatives to initiate the apoptosis process remains a subject of intense study Direct interaction between BH3-only proteins and the multidomain pro-survival Bcl-2 members like Bcl-2 and Bcl-xL has been well established (Ottilie et al 1997; Petros et al 2000; Liu et al 2003) On the other hand, although there are many reports on the effects of BH3-only proteins on the multidomain pro-apoptotic Bcl-2 members, their direct interaction with each other has not been clearly demonstrated till recently (Cartron et al 2004; Harada et al 2004) Together with recent studies using BH3 peptides, several groups have suggested that the BH3-only proteins can be categorized into two types: the ‘sensitizers’ and the

‘activators’ (Letai et al 2002; Kuwana et al 2005) The activator BH3-only proteins which include Bid and Bim so far, can bind to both the pro-survival and the pro-

interact with Bax/Bak leading to their activation In contrast the BH3 domains of the sensitizer BH3-only proteins like Bad and Bik, are unable to induce Bax/Bak activation and can only bind to the pro-survival members like Bcl-2 and Bcl-xL to repress their function (Letai et al 2002; Kuwana et al 2005) These findings have led

to the current model of how interactions between members of the Bcl-2 family protein control the induction of apoptosis Upon encountering a variety of stress signals, the BH3-only proteins undergo various modifications and become activated before transducing the stress signal to other Bcl-2 members The activator BH3-only proteins will proceed to bind to and activate Bax/Bak leading to cytochrome c release and eventually caspase activation and cell death However, in the presence of pro-survival Bcl-2 members like Bcl-2 or Bcl-xL, activated BH3-only protein’s higher affinity for Bcl-2/Bcl-xL compared to Bax/Bak, enables the pro-survival Bcl-2 members to sequester the activator BH3-only proteins away from Bax/Bak and apoptosis does not proceed On the other hand, cell death will occur if there are suitable sensitizer BH3-only proteins present that are able to displace the activator BH3-only proteins from the pro-survival Bcl-2 members, freeing them to trigger Bax/Bak activation (Letai et

al 2002; Kuwana et al 2005; Certo et al 2006)

1.3 DEFECTS IN APOPTOSIS AND CANCER

Apoptosis is an important physiological process in the regulation of tissue homeostasis in adult tissues, in part by eliminating cells with potentially harmful mutations or deregulated cell cycle control Defects in the apoptosis program can disrupt this balance of cell proliferation and cell death and lead to development of multiple diseases Often, the growth-promoting oncoproteins like Myc and E2F1 that

promote malignant transformation are also powerful inducers of apoptosis The dual function of these oncoproteins - cell proliferation and cell death, is believed to operate

as an in-built mechanism to curb their oncogenic potential (Askew et al 1991; Harrington et al 1994; Shan and Lee 1994; Wu and Levine 1994) In order to survive malignant transformation, a developing tumor must acquire mechanisms to suppress apoptosis Later on, the acquired ability to suppress apoptosis also provides a growth advantage to developing tumor cells by allowing them to overcome threats to their survival as they progress to an invasive and metastatic stage These include nutrient deprivation, hypoxia and loss of matrix support, all of which can trigger stress-induced apoptosis in cells Moreover cancer cells have to overcome apoptosis triggered by DNA damage as a result of defective repair, telomere erosion, oncogene deregulation and therapy (Hoeijmakers 2001; Harris 2002) Therefore, mutations that confer resistance to apoptosis allow subpopulation of oncogenic transformed, genetically unstable or damaged cells to escape elimination and continue to proliferate

In fact, resistance to cell death, in particular apoptotic cell death, is now understood to

be central to tumorigenesis and is a hallmark of many types of cancer (Hanahan and Weinberg 2000)

1.3.1 Mutations that Confer Apoptosis Resistance

Cancer cells acquire the ability to evade apoptosis through several strategies One of the most common is the functional inactivation of the p53 tumor suppressor,

where mutation or deletion of p53 is the most frequent genetic abnormality, occuring

in more than 50% of human cancers (Levine 1997) p53 is activated upon sensing DNA damage and other stress signals including hypoxia and oncogene

hyperactivation and proceeds to induce cell cycle arrest and apoptosis p53, as a transcription factor, indirectly induce apoptosis by regulating the expression of several genes involved in apoptosis induction like Bax, Apaf-1, caspase-9, CD95/Fas, DR5 and Noxa (Miyashita and Reed 1995; Owen-Schaub et al 1995; Sheikh et al 1998; Soengas et al 1999) However, recently p53 has been reported to induce apoptosis directly by activating the pro-apoptotic Bax (Chipuk and Green 2004)

Alterations in expression of genes regulating the apoptotic signaling pathway can also lead to tumorigenesis and tumor cells’ resistance to apoptosis Among these include overexpression of anti-apoptotic proteins such as those from the Bcl-2 family

The anti-apoptotic gene Bcl-2 (B-cell lymphoma 2) which is overexpressed in B-cell

lymphomas due to chromosomal translocation to an immunoglobulin locus (Tsujimoto et al 1984a; Tsujimoto et al 1984b) was found to drive the survival of lymphoma cells by inhibiting cell death rather than promoting proliferation (Vaux et

al 1988; McDonnell et al 1989; Hockenbery et al 1990) Bcl-2 has also been found

to be overexpressed in other cancers such as prostate (Matsushima et al 1997; Keshgegian et al 1998) ovarian (Mano et al 1999), breast (Lipponen et al 1995; Le

et al 1999; Nakopoulou et al 1999), small cell lung carcinoma (SCLC) (Ben-Ezra et

al 1994; Higashiyama et al 1995), lymphocytic leukemias (Schena et al 1992) and is usually associated with poor prognosis Overexpression of other pro-survival Bcl-2 family of proteins like Bcl-xL and Mcl-1 have also been reported in several types of cancer (Amarante-Mendes et al 1998; Kaufmann et al 1998; Puthier et al 1999; Michels et al 2005) Overexpression of Bcl-2 and Bcl-xL is also associated with resistance to chemotherapeutic agents (Ohmori et al 1993; Minn et al 1995) and radiotherapy (Harima et al 1998; Mackey et al 1998) Conversely, pro-apoptotic members of the Bcl-2 family of proteins like Bax and Bak have been reported to be

mutated or downregulated in gastrointestinal cancer and leukemia (Rampino et al 1997; Yamamoto et al 1997; Kondo et al 2000)

Mutations of critical components in survival signaling pathways have also been employed by cancer cell as a mechanism for suppressing apoptosis These mutations include those that alter the expression of survival factors like growth factors and cytokines, mutations of upstream receptors that transmit the survival signals to the intracellular circuitry, and upregulation of activators or downregulation of suppressors within the survival pathway As a result, these mutations often lead to hyperactivation or constitutive activation of survival pathways that help to suppress apoptosis The mechanism and mutations involved, especially pertaining to survival signaling regulated by growth factors, will be further discussed in the following section (1.4)

The prevalent therapeutic approach in cancer treatment has been the induction

of apoptosis in cancer cells via various means like chemotherapy, γ-irradiation and immunotherapy (Fulda and Debatin 2004) Their success at the initial stage is largely due to the increased sensitivity of cancer cells to apoptosis as they have to sustain oncogenic lesions and also unfavorable growth conditions compared to their normal counterparts (Evan and Littlewood 1998; Schmitt and Lowe 1999) However, in many cases, after a period of time these tumor cells develop resistance to conventional therapies Frequently, the resistance can be attributed to deregulated anti-apoptotic mechanisms (such as those mentioned above), originally set in place to prevent inadvertent initiation of the apoptosis process in healthy cells (Green and Evan 2002; Johnstone et al 2002; Kaufmann and Vaux 2003) However, it is believed that the resistant tumors are highly dependent on these deregulations in the mechanisms that control apoptosis for their continued survival Therefore much effort is being invested

in devising strategies to overcome this resistance to apoptosis in the hope of restoring the sensitivity of tumor cells to anti-cancer treatment (Cummings et al 2004; Fesik 2005) Hence our understanding of the intricate apoptotic and survival signaling pathways of tumor cells and other factors that influence the apoptosis process is not only key to the discovery of potential new targets for cancer therapy but it may also help us to increase the efficiency of current treatments.

1.4 ANTI-APOPTOTIC MECHANISMS: GROWTH FACTOR SIGNALING

1.4.1 Growth Factor Signaling

Growth factors represent a large group of polypeptides that mediate diverse cellular functions, from stimulating proliferation to regulating cell differentiation, motility and survival They are synthesized by various cell types and tissues and usually act locally within the tissues where they are produced Several families of growth factors and their cognate receptors have been classified, and while some have

a restricted range of target cells, others can target many different types of cell It would appear that growth factors exhibit broad-ranging actions on diverse cell types

in different tissues, however, their activity in vivo is subjected to complex control

with regards to their distribution, availability and delivery to corresponding cells Moreover, the final outcome for the interaction of a particular growth factor and its target cell is dependent on the cell type and its cellular context (Favoni and de Cupis 2000) Growth factors exert their effects by binding to their specific receptors on the cell surface, triggering receptor oligomerization and activation (Heldin 1995) These receptors form homodimers or, alternatively, heterodimerize with other members

within the family, thereby diversifying their ligand binding capacity and the effective downstream signaling pathways (Lemmon and Schlessinger 1994)

The receptor tyrosine kinases (RTKs) represent a large family of receptors that bind growth factors from several subfamilies including epidermal growth factors (EGF), platelet-derived growth factor (PDGF), vascular endothelial growth factor (VEGF), fibroblast growth factor (FGF) and insulin-like growth factors (IGF) hepatocyte growth factor (HGF) There are at least 58 genes that encode for receptor tyrosine kinases known to date, and they can be divided to 20 subfamilies based on sequence homology and their extracellular ligand-binding domains (Robinson et al 2000; Adjei and Hidalgo 2005) All RTKs share certain structural similarities; they all possess an extracellular ligand-binding portion, a single transmembrane region, and a cytoplasmic portion which consist of a highly conserved catalytic tyrosine kinase domain and a more divergent carboxyl-terminal tail Binding of growth factors to appropriate RTKs leads to receptor dimerization and activation of the receptors’ intrinsic tyrosine kinase activity; and concomitant autophosphorylation of conserved tyrosine residues within its cytoplasmic region The phosphorylated tyrosine residues act as docking sites, allowing other intracellular proteins with phosphotyrosine-binding domains to physically associate with the RTKs Association with RTKs, not only leads to phosphorylation and activation of these intracellular proteins, but in effect, translocates these proteins to the plasma membrane where their substrates are located In addition, activated RTKs also engage adaptor proteins like Grb2 (Growth factor receptor-bound protein 2) with multiple protein interaction domains, thereby diversifying the range of downstream effector proteins From here, the signal from an activated RTK is rapidly propagated to multiple intracellular signaling pathways that,

independently or in combination, control critical cellular process as including cell

proliferation, differentiation, inhibition of apoptosis and cell migration

1.4.2 Aberrant Growth Factors Signaling in Cancer Cells

Therefore it is not surprising that deregulation of growth factor signaling is one of the fundamental factors that promote cancer development and progression (Hanahan and Weinberg 2000) In fact a large number of known oncogenes behave in

a way that imitates normal growth factor signaling enabling cells to bypass the requirements for extracellular growth signals thus leading to autonomous growth Among the common aberrations that lead to deregulated growth factor signaling found in tumors include those that affect the extracellular growth factors, the growth factor receptors or the downstream intracellular signaling pathways involved Limitation in growth factors availability can trigger apoptotic signals in affected cells This especially affects cancer cells located in the core of a developing tumor mass where there is limited supply of nutrients However, in many cases, these cancer cells develop the ability to produce the required growth factors, thereby reducing their dependence on exogenous growth factors from surrounding tissue This switch from paracrine to autocrine growth factor signaling often seen during tumor progression, not only allow tumor cells to establish autonomous growth but also increases its invasiveness by promoting metastasis and cell survival in new environments (Mercurio et al 2004)

Another common aberration in growth factor signaling arises from mutations

of RTKs There are at least 30 RTKs implicated in human cancers with several types

of mutations including receptor overexpression due to gene amplification, genomic

rearrangements such as chromosomal translocation leading to formation of chimeric proteins and gain-of-function or deletion mutations (Blume-Jensen and Hunter 2001) Receptor overexpression causes the receptor to be hyperresponsive to baseline levels

of growth factor, or in extreme cases lead to growth factor-independent signaling (Di

Fiore et al 1987) An important example is the erbB-2 (HER2/neu) gene, a member of

the EGF receptor family, is frequently amplified in human breast and ovarian carcinomas (Kraus et al 1987; Natali et al 1990; Parkes et al 1990), and detection of high levels of ErbB-2 in breast carcinomas is indicative of poor prognostics (Slamon

et al 1987) Growth factor-independent signaling is also observed in a deletion

mutation of EGFR gene that produces a truncated receptor lacking part of its

extracellular ligand binding domain (Humphrey et al 1990; Wong et al 1992) This EGFRvIII mutant is unable to bind ligands but has a constitutive kinase activity (Wong et al 1992) and has been shown to be present in human tumors including malignant gliomas, breast and non-small cell lung carcinomas (Humphrey et al 1988; Yamazaki et al 1988; Sugawa et al 1990; Garcia de Palazzo et al 1993; Wikstrand et

al 1995) On the whole, mutations of RTKs frequently lead to upregulated or constitutive activation of the receptor kinase activity and consequently downstream signaling

Extracellular signals received by growth factor activated RTKs are transmitted downstream to a complex network of signaling pathways that translate these signals into cellular response Alterations in the key signaling components of these pathways have also been shown to promote malignant transformation and tumor progression Central to this is the Ras-Raf-MAPK pathway, which control diverse cellular functions including proliferation, differentiation and survival Point mutations of the

ras genes that result in constitutively active Ras are the most common dominant

oncogene mutation in human cancer The point mutations usually target three codons,

12, 13 and 61, which interferes with GTP hydrolysis and locks Ras in its GTP-bound (active) state (Bos 1989) In addition, discovery of new effectors of this pathway, and also cross-talk between Ras-Raf-MAPK and other pathways like the PI3K-Akt pathway (Hunter 1997; Rommel and Hafen 1998), means that any mutation of its signaling components could potentially have multiple biological effects

These aberrations in growth factor signaling allow the cells to acquire important capabilities that promote cancer development and progression This includes the ability to proliferate autonomously as they loose their dependence on extracellular growth signals from surrounding tissue As a result cancer cells are not subjected to the same proliferative constraints by their environment as normal cells In addition, deregulated growth factor signaling confers the ability to evade apoptosis in cancer cells Several important survival pathways have been shown to be activated by

a wide array of growth factors Growth factors including IGF-I and IGF-II, PDGF and EGF have been reported to prevent apoptosis in various cell types by activating the PI3K-Akt and Ras-Raf-ERK pathways (Talapatra and Thompson 2001; Henson and Gibson 2006) Moreover certain growth factors like VEGF have the ability to induce growth of new blood vessels allowing cancer cells to spread through solid tissues (Ferrara 1999; Jackson et al 2002) The ability of growth factors to promote angiogenesis is an important step towards the development of a more invasive and metastatic tumor Overall, aberrations in growth factor signaling give rise to uncontrolled and excessive activation of the intracellular signal transduction pathways that regulate important cellular processes, ultimately promoting malignant transformation and tumor progression

1.4.3 Growth Factors-Regulated Survival Signaling Pathways: PI3K-Akt

Pathway

Aberrations of PI3K-Akt Signaling in Cancer

Extracellular signaling molecules like growth factors promote cell survival by activating survival signaling pathways that consequently suppress apoptosis One well-known survival pathway is the signaling pathway of the phosphoinositide-3-kinase (PI3K) through its downstream effector, Akt/PKB, a serine/threonine protein kinase (Yao and Cooper 1996; Franke et al 1997; Crowder and Freeman 1998; Zheng

et al 2002) The PI3K-Akt signaling pathway is frequently constitutively activated or amplified in many types of cancers (Nicholson and Anderson 2002) Amplification of the gene that encodes for the p110 catalytic subunit of PI3K have been reported in ovarian and squamous cell cancer (Shayesteh et al 1999; Woenckhaus et al 2002)

and the AKT2 gene in ovarian, breast and pancreatic cancer (Bellacosa et al 1995;

Cheng et al 1996b; Ruggeri et al 1998), whereas deletion mutation of the p85 regulatory subunit of PI3K have been reported in human colon and ovarian cancer (Philp et al 2001) In these cases, the PI3K-Akt signaling pathway is usually constitutively activated due to overexpression of the normal protein or mutations that allow the signaling proteins to bypass negative regulation The most compelling evidence of PI3K-Akt pathway role in human cancer, however, comes from studies of its antagonist, the phosphatase PTEN (phosphatase and tensin homolog deleted on chromosome 10) PTEN induces apoptosis by antagonizing the PI3K-Akt pathway, specifically by inhibiting the phosphorylation and activation of Akt (Stambolic et al 1998) Studies have shown that loss of PTEN occurs in a wide variety of human cancer including glioblastoma (Wang et al 1997), prostate (McMenamin et al 1999),

2000; Zhou et al 2000), underscoring its role as a powerful tumor suppressor (Ali et

al 1999; Yamada and Araki 2001; Vivanco and Sawyers 2002)

al 1998) but also brings PI3K in contact with its lipid substrate on the plasma membrane PI3K are also activated by RTKs in an indirect manner through Ras which binds and activates the p110 subunit (Kodaki et al 1994; Rodriguez-Viciana et al 1994; Pacold et al 2000)

Activated PI3K catalyzes the transfer of a phosphate group from ATP to the D3 position of the phosphatidylinositols (PtdIns) generating 3’-phosphatidylinositol phosphates (PIPs), of which primary importance is PtdIns(3,4,5)P3 (PIP3) generated from PtdIns(4,5)P2 (PIP2) (Wymann and Pirola 1998) PIP3 serves as a lipid second