Towards targeted treatment for osteosarcoma · 2015. 1. 12. · provides an overview of (pre)clinical research efforts that were endeavoured upon in the past decade to exploit specific

Towards targeted treatment for osteosarcoma

Preclinical target identification studies

Jantine Posthuma de Boer

The investigations described in this dissertation were conducted at the laboratories of the VUmc Cancer Center Amsterdam (CCA) on behalf of the department of Orthopaedic Surgery, VU University Medical Center, Amsterdam, the Netherlands

Financial support for printing this dissertation was obtained from: de Nederlandse Orthopedische Vereniging, Medisch Centrum Alkmaar, Stichting Anna Fonds I NOREF, Implantcast Benelux, LINK Nederland, Takeda Nederland, ABN AMRO, Chipsoft, De Puy Synthes

© 2014 Jantine Posthuma de Boer. All rights reserved. No parts of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, without prior written permission of the author. ISBN: 978-94-91487-14-9

Lay-out by: Ridderprint BV, Ridderkerk, the Netherlands, cover design by Judith van der VeldenPrinted by: Ridderprint BV, Ridderkerk, the Netherlands Publisher: Medix Publishers BV, Keizersgracht 317A, 1016 EE Amsterdam

De digitale versie van dit proefschrift is te vinden in de YourThesis app en kan gelezen worden op een tablet of smartphone. De app kunt U downloaden in de App Store en de Google Play store of middels het scannen van onderstaande QR-code.

http://www.ridderprint.nl

http://www.ridderprint.nl

http://www.medixpublishers.nl

VRIJE UNIVERSITEIT

Towards targeted treatment for osteosarcoma

Preclinical target identification studies

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam,

op gezag van de rector magnificus prof.dr. F.A. van der Duyn Schouten,

in het openbaar te verdedigen ten overstaan van de promotiecommissie

van de Faculteit der Geneeskundeop donderdag 3 juli 2014 om 13.45

In de aula van de universiteit,De Boelelaan 1105

door

Jantine Posthuma de Boer

geboren te Blackburn, Verenigd Koninkrijk

promotoren: prof.dr. BJ van Royen prof.dr. GJL Kaspers

copromoteren: dr. MN Helder dr. VW van Beusechem

Ever triedEver failedNo matterTry againFail againFail better

-Samuel Beckett-

Aan mijn ouders en tweelingbroer

Table of contents

Chapter 1 General introduction 9

Chapter 2 Molecular alterations as targets for therapy in 15 metastatic osteosarcoma: a review of literature

Chapter 3 Mechanisms of therapy resistance in osteosarcoma: a review 33

Chapter 4 WEE1 inhibition sensitises osteosarcoma to radiotherapy 55

Chapter 5 Targeting JNK-interacting protein 1 (JIP1) sensitises 69 osteosarcoma to doxorubicin

Chapter 6 Surface proteomic analysis of osteosarcoma 87 identifies EPHA2 as receptor for targeted drug delivery

Chapter 7 Summary and discussion 109

Chapter 8 Nederlandse samenvatting 123

Reference list 133

List of abbreviations 145

Appendix A 149

Appendix B 153

Acknowledgement 163

Curriculun Vitae / List of publications 167

Chapter 1General introduction

10

Chapter 1

I N T R O D U C T I O NOsteosarcoma is the most common primary malignant bone tumour in children and adolescents. The estimated incidence rate worldwide is 4/million/year with a peak incidence at the age of 15-19 years and a second, though small, peak in the elderly population. The male to female ratio is 1.4 : 1. Conventional osteosarcoma most commonly arises in the metaphyses of the long bones; preferential sites for osteosarcoma are the distal femur or proximal tibia (70%), followed by the proximal humerus (10%) and pelvis (7%). At older age, osteosarcoma follows a different pattern and occurs more often in the axial skeleton. Osteosarcoma in the elderly may develop as a sequel to irradiation for a prior malignancy or in skeletal areas of pre-existent bone disease. [1-4] The clinical presentation of osteosarcoma is highly unspecific; pain and swelling are common symptoms and are generally mild. Given the insidious course of symptoms, both patient-delay and doctor-delay can detain the diagnosis of osteosarcoma. Conventional radiography generally displays a bone mass with accompanying periosteal reaction, often seen with a surrounding soft tissue mass containing calcifications. An additional MRI can provide valuable information regarding the dimension of the tumour and the degree of infiltration into surrounding tissues. Ultimately, the diagnosis is formulated after biopsy and histopathologic evaluation of the tumour cells. The formation of new, immature bone (osteoid), both on X-ray and histologic examination, is pathognomonic for osteosarcoma. [4,5]

Osteosarcomas can be classified according to histologic subtypes and tumour location. One can distinguish between central (medullary) and surface (cortical or juxtacortical) tumours. The term “conventional osteosarcoma” refers to central, high-grade, osteoblastic osteosarcomas. Central osteosarcomas can also be of the chondroblastic, fibroblastic or teleangiectatic subtype. Small cell osteosarcomas and intra-osseous well-differentiated osteosarcomas also belong to the central osteosarcomas. Surface osteosarcomas include parosteal (juxtacortical) well-differentiated osteosarcomas, periosteal osteosarcomas and high-grade surface osteosarcomas. The histologic subtype is of influence on the behaviour of the tumour; small cell osteosarcomas have a more aggressive behaviour, fibroblastic and teleangiectatic tumours tend to show a better response to chemotherapy than do osteoblastic and chondroblastic osteosarcomas. [3,6-8]

Osteosarcoma typically is a complex tumour with numerous genetic and chromosomal aberrancies, however, there is no characteristic mutation or immunohistochemical marker that defines osteosarcoma. [5,7] The aetiology of osteosarcoma remains largely unknown, although there are several factors that are implicated in the development of osteosarcoma. First, there are a few common chromosomal abnormalities found in osteosarcoma that can consist of both partial or complete gains or losses. These include: gain of chromosome 1, loss of chromosome(s) 9, 10, 13 and/or 17, amplifications of chromosomes 8 and 12, partial or complete loss of chromosome 6. Rearrangements of chromosomes 11, 19 and 20

11

General introduction

1are also frequently encountered. [3,9-11] Second, certain rare syndromes are associated with an increased risk of osteosarcoma development, such as Bloom syndrome, Rothmund-Thomson syndrome and Li-Fraumeni syndrome. Germ line mutations of the most well known tumour suppressor genes, p53 and retinoblastoma (Rb), are reported to be involved in the pathogenesis of osteosarcoma. Combined mutations or inactivations of both genes are often encountered. Patients with hereditary retinoblastoma very regularly develop osteosarcoma as secondary tumours and have an increased risk of 500-fold compared to the general population. [11-14] Third, metabolic bone disease can predispose to osteosarcoma development. Patients with Paget’s disease have a risk of 1% to develop osteosarcoma, a risk that is 2500-fold higher than in the general population, the cause of which has not been elucidated. [12] Finally, there are external and environmental factors that can increase the risk of osteosarcoma development, including irradiation therapy for previous malignancies and exposure to the chemical compound of beryllium. [12]

At present, the standard treatment for high-grade osteosarcoma is a multidisciplinary effort by (paediatric) oncologists, orthopaedic surgeons, pathologists, rehabilitation physicians and specialised nurses and includes neoadjuvant multi-agent chemotherapy (including doxorubicin, cisplatinum, ifosfamide and methotrexate), radical surgery and postoperative chemotherapy. [3,5,15-17] Local control of the primary tumour is essential to obtain cure for osteosarcoma and therefore, in patients with axial and/or unresectable osteosarcoma, in whom local control is difficult to achieve, there is a high risk of progression, relapse and/or metastasis. In selected cases, radiotherapy could offer improved chances for survival. [1,18,19]

With the advent of chemotherapy, 5-year survival rates for osteosarcoma have improved over the past three decades from 12% in the 1980’s to 65% at the present moment for patients with localised disease. However, in the case of metastatic or recurrent disease, 5-years survival rates are reduced to approximately 20%, with a median survival of around 1 year. Currently, the only reliable prognostic factor for disease-free survival is the response to induction chemotherapy, assessed by immunohistochemical examination of the excision specimen of the primary tumour. Response rate is defined as the percentage of necrosis in the primary tumour; >90% necrosis is considered a good response. [1]

Metastatic disease is a major issue in the course of osteosarcoma. Osteosarcoma has a high tendency to metastatic spread and the absolute majority (80%) of metastases are pulmonary metastases. Most commonly the pulmonary metastases develop in the periphery of the lungs. Approximately 20-30% of patients present with metastasis at initial diagnosis and, additionally, in 40% of patients metastases occur at a later stage of their disease. Apart from lung metastases, metastasis to other skeletal locations is also common (20%). These bone metastases are generally not lethal, however, they account for considerable morbidity. Furthermore, the presence of bone metastases correlates to inferior survival outcomes.

12

Chapter 1

Treating metastatic osteosarcoma remains a challenge and most deaths associated with osteosarcoma are the result of metastatic disease. In the case of metastatic disease, metastatectomy does yield improved survival and should therefore always be performed when feasible. [1,5,19-25]

A I M S A N D O U T L I N E S O F T H I S T H E S I SAltogether, survival outcomes remain unsatisfactory for patients with osteosarcoma. A variety of combination therapy regimens and dose escalation of several therapeutics have not improved survival outcomes, [5,15-17] implying that the treatments we currently provide lack efficacy. This lack of treatment efficacy could be attributed to therapy resistance of the osteosarcoma cells.

The objective of this thesis is to define strategies to subvert or circumvent this relative resistance to therapy, thereby ultimately improving the efficacy of existing therapies.

Chapter 2 provides an introduction to molecular alterations in metastatic osteosarcoma (cells) that could potentially serve as therapeutic targets. Understanding the biology of osteosarcoma metastasis may uncover essential molecules or mechanisms for the survival of metastatic cells and interfering with these might elicit an anti-tumour effect. Therefore, a detailed study of the biological characteristics and behaviour of metastatic osteosarcoma cells may provide a rational basis for innovative treatment strategies. This chapter also provides an overview of (pre)clinical research efforts that were endeavoured upon in the past decade to exploit specific molecular pathways in metastatic osteosarcoma, in order to discover novel targets for treatment and to test their efficacy.

As stated above, therapy resistance of osteosarcoma cells can attribute to inferior treatment efficacy. Chapter 3 enlightens mechanisms of this therapy resistance within osteosarcoma cells and provides a framework to the following chapters in which we investigate various methods of targeted therapy to improve the efficacy of radiotherapy, doxorubicin chemotherapy and the delivery of therapeutics to osteosarcoma cells. Targeted therapy can be achieved either by selectively targeting intracellular proteins essential for tumours to survive, or by a targeted delivery of therapeutics to the tumour by directing therapy selectively to extracellular surface proteins or receptors on tumour cells.

In Chapter 4 we address the issue of resistance to radiotherapy in osteosarcoma. One strategy to enhance radiotherapy efficacy in patients with osteosarcoma is to push irradiated cells forward through the cell cycle using a small-molecule inhibitor drug. In doing this, DNA repair prior to cell division is hampered and the osteosarcoma cells are pushed into mitotic catastrophe. This results in a sensitisation of osteosarcoma cells to radiation therapy.

The issue of chemoresistance of osteosarcoma is addressed in Chapter 5. We hypothesise that to increase chemotherapy efficacy, essential survival pathways should be targeted

13

General introduction

1concomitantly with administering chemotherapy, thus tilting the intracellular balance to cell death rather than cell survival. Here, we describe the use of functional genomics to discover kinases that, upon silencing, enhance the sensitivity of osteosarcoma to doxorubicin treatment. After candidate selection, pharmacological studies are performed to confirm the initial findings.

Apart from shifting molecular balances within osteosarcoma cells to favour cell death after radiation and chemotherapy, another strategy to improve therapeutic efficacy could be to enhance the delivery of currently applied drugs to the tumour cells. In Chapter 6 we describe the identification of a suitable surface receptor on osteosarcoma cells that can be used as a receptor for targeted drug delivery. In this chapter, we select a proteomics approach to study the surface proteome of osteosarcoma cells compared to healthy bone cells. After candidate selection and confirmation of surface expression, we used specifically targeted adenoviral vectors to study the selective intracellular uptake of these moieties into osteosarcoma cells. Ultimately, the targeted delivery of drugs to osteosarcoma cells can lead to higher active doses at the site of the tumour and thus higher treatment efficacy, while sparing healthy tissues.

Chapter 7 summarises the preceding chapters and forms a general discussion to the topics covered in this thesis, addressing unexplored areas of research, implementation of our obtained knowledge into new treatment modalities for osteosarcoma, and perspectives for the feasibility of developing future cancer treatments in general, and, osteosarcoma in particular. Finally, future perspectives are formulated to conclude this work.

Chapter 8 provides a Dutch summary.

Chapter 2Molecular alterations as targets

for therapy in metastatic osteosarcoma:

a review of literature

J. PosthumaDeBoer, M.A. Witlox, G.J.L. Kaspers, B.J. van Royen

Clinical & experimental metastasis. 2011 Jun; 28(5): 493-503

16

Chapter 2

A B S T R A C TTreating metastatic osteosarcoma (OS) remains a challenge in oncology. Current treatment strategies target the primary tumour rather than metastases and have a limited efficacy in the treatment of metastatic disease. Metastatic cells have specific features that render them less sensitive to therapy and targeting these features might enhance the efficacy of current treatment. A detailed study of the biological characteristics and behaviour of metastatic OS cells may provide a rational basis for innovative treatment strategies. The aim of this review is to give an overview of the biological changes in metastatic OS cells, the preclinical and clinical efforts targeting the different steps in OS metastases and how these contribute to designing a metastasis directed treatment for OS.

17

Molecular alterations as targets for therapy in metastatic osteosarcoma: a review of literature

2

I N T R O D U C T I O NOsteosarcoma (OS) is the most common primary malignant bone tumour in children and adolescents. The estimated incidence rate worldwide is 4/million/year, with a peak incidence at the age of 15-19 years. [2] In OS there is a high tendency to metastatic spread. Approximately 20% percent of patients present with lung metastases at initial diagnosis and, additionally, in 40% of patients metastases occur at a later stage. Eighty percent of all metastases arise in the lungs, most commonly in the periphery of the lungs, and exhibit resistance to conventional chemotherapy. [5,19-23] The 5-year survival rate for OS patients with metastases is 20% compared to 65% for patients with localised disease and most deaths associated with OS are the result of metastatic disease. [1,5,26-28]

For patients with pulmonary metastasis, especially those who have metastasis at initial diagnosis, the combination of radical metastasectomy and chemotherapy offers the best outcome and even potential cure. Nevertheless, recurrent development of pulmonary metastases after initial radical metastasectomy is reported to be high and repeated metastasectomies are sometimes performed. As metastasectomy does yield an improved survival in most patients it should therefore always be performed when feasible. [20,21,24,25,29]

In order to improve survival, the ultimate questions to be answered are: Why does osteosarcoma metastasise, particularly to the lungs? And, more importantly: Why does therapy fail in metastatic disease? In this regard, we hypothesise that drug resistance is a key issue in the failure to control metastatic disease. It has been shown that OS lung metastases display a biological behaviour different from the primary tumours. [20,25,30-42] Metastases are comprised of cell clones that differ from primary tumours with respect to ploidy, enzyme profile, karyotype and chemosensitivity. [7,20,43-46] Therapeutic regimens that target primary tumours are therefore unlikely to be successful in the treatment of metastatic disease.

Metastasis is considered to be the final though most critical step in tumorigenesis of malignant tumours. [47] The metastatic cancer cells subsequently complete the following steps: (1) Invasion through the extracellular host matrix and entrance into the circulation, (2) survival in the circulation and (3) evasion of the host immune system, (4) arrest and extravasation at a target organ site, (5) adherence and (6) survival in the target organ microenvironment and finally (7) formation of neovasculature to allow growth at the target organ site. (VII) [25,47-50] Each step is of equal importance and must be fully completed by the tumour cell to achieve successful metastasis. The altered biological behaviour in metastatic cells is the result of specific molecular changes. We will discuss each of these specific steps with special attention to the molecules involved in OS metastasis (Table 2.1) and the potential implications for therapy. Over the last decade, much research has been performed to try to unravel the biology of OS metastasis and many (pre)clinical studies have attempted to discover new treatment options for metastatic OS. For example, gene

18

Chapter 2

expression profiling of metastatic cells using a cDNA microarray approach has identified genes responsible for metastasis. [42,51-53] Also, expression levels of specific proteins in OS lung metastases have been analysed. In these studies, expression levels of proteins involved in metastases link the molecular aberrancies to clinical outcome in terms of survival rates. These alterations may also provide novel drug targets. [36,38,40,50,54-56] Table 2.2 summarises (pre)clinical studies for treating OS metastases.

The aim of this review is to give an overview of the biology of metastatic OS cells and of (pre)clinical efforts targeting the different steps in OS metastases, and, how these may contribute to designing a metastasis directed treatment for OS.

O S T E O S A R C O M A M E TA S TA S I S(1) Migration and InvasionMigration of cells away from the primary tumour and invasion through the extracellular matrix (ECM) towards the bloodstream is considered the first step contributing to metastasis. In OS, it has been described that MetalloProteinases (MMPs) 2 and 9 and m-Calpain play a role in degradation of the ECM. [25,27,31,33,34,42,48,49,57-60] Also, the Wnt/β-catenin pathway and Src-kinase are implicated as inducers of migration. [26,49,60,61] In OS, Notch is a relatively recently identified pathway and has been identified as a promoter of invasion. In highly metastatic OS cell lines there is an upregulation of the Notch1 and Notch2 receptor, as well as the Notch induced gene Hes1. In patient samples expression of the Hes1 gene inversely correlated with survival. [55,62-64](Pre)clinical studies: Notch In a preclinical setting, downregulation of the Notch signalling pathway has been shown to impair the invasiveness of cell lines, however, no effect on cell proliferation or in vitro tumorigenesis was found. Notch-inhibited cell lines had less potential to form lung metastases in an orthotopic mouse model when compared to untreated cell lines. The exact mechanism through which inhibition of the Notch pathway and its target gene Hes1 leads to reduced invasion still remains unknown. [55,64]

(2a) Survival in the bloodstream Dissemination of cancer cells through the body requires the cells to survive in the circulation. Non-malignant cells, as well as non-metastatic tumour cells, become apoptotic after loss of cell-cell adhesions or interaction with the ECM in their original tissue. This specific mode of apoptosis is called anoikis. One reason for this type of apoptosis to occur is that Integrin signalling ceases to exist in solitary cells. Under adherent conditions, survival is often mediated through Integrin signalling pathways in which Focal Adhesion Kinase (FAK) is a central player. FAK activates the important PI3K/Akt survival pathway. Metastatic tumour

19


2

cells evade anoikis by intrinsically activated survival pathways via for example PI3K or Akt signalling, independent of extrinsic Integrin and FAK signals. [25,26,31,65] Src kinase can also activate the PI3K/Akt and the Ras/MAPK survival pathways independent of FAK signalling and thus stimulate ainoikis resistance. [25,26,31,33,49,58,65] Other survival mechanisms are also of importance in the evasion of ainoikis, such as activation of the Nuclear Factor-kappa B (NF-κB) pathway [42,65] and Wnt-mediated upregulation of the β-catenin activity. High levels of β-catenin expression have been shown to be associated with a metastatic phenotype in OS. [25,34,35,60] Finally, overexpression of anti-apoptotic genes such as Bcl-2, Bcl-XL or FAK is exploited by solitary metastatic cells to obtain a survival benefit. [26,49,65]

(2b) Apoptosis resistanceThe survival of tumour cells through all stages of metastasis (not only in the bloodstream) is paramount to successful metastasis. Mechanisms involved in apoptosis resistance throughout metastasis include activation of the Src and NF-κB pathways and the overexpression of anti-apoptotic genes. [26,58,60,65,66] Wnt-signalling is also involved in resistance to apoptosis throughout other steps of metastasis and is considered to be important for tumour progression in general. Upon binding of Wnt to one of its receptors, β-catenin degradation in the cytoplasm is prevented. After β-catenin is stabilised, it translocates to the nucleus where it co-regulates oncogene transcription and cell cycle progression and hence promotes survival and proliferation. [25,32,60,61,67,68]

In established metastases, the tumour cells are confronted with receptor-mediated cell death. Binding of Fas on the surface of metastatic cells to its Fas-ligand (FasL) expressed constitutively on lung tissue, activates the Fas-apoptosis pathway and leads to cell death. Cross-linkage of Fas with FasL on one cell results in apoptosis as well. [50,69] Resistance to death-receptor-induced apoptosis is commonly seen and is highly important for the successful maintenance of metastases. The Fas/Fas-ligand pathway is a death receptor pathway that is often down-regulated in metastatic cell populations, rather than in primary tumours. [7,38,50] The Fas pathway is also of influence on chemotherapy-induced apoptosis and thus on its therapeutic efficacy. [70] Much (pre) clinical research has been performed concerning this pathway in metastatic OS and promising results have been obtained. These results are discussed following the section on immune evasion, since the Fas pathway plays a role in that as well.

Thus, apoptosis resistance is very much exploited by the metastatic cell and this feature is likely to contribute to resistance to therapy in metastatic OS. [65,66] The failure to induce apoptosis upon treatment is thought to be the result of a misbalance between pro- and anti-apoptotic signalling. Restoration of this balance, thereby creating an environment in favour of pro-apoptotic signalling could theoretically enhance treatment with cytotoxic agents. [26,49,59,65,70-72]

20

Chapter 2

(Pre)clinical studies: Wnt and SrcThe Wnt-pathway is a putative therapeutic target because a majority of OS samples show aberrant activation of this pathway, leading to the transcription of oncogenes and cell cycle progression. This in turn leads to proliferation and enhanced survival. [60,67,73] When targeting the Wnt-pathway, activating mutations in downstream molecules, for example, β-catenin can be of negative influence as it may bypass Wnt inhibition and preserve the invasive phenotype of the metastatic cell. [34] A preclinical study by Leow et. al [66] has shown that inhibition of the Wnt/β-catenin pathway resulted in lower levels of nuclear β-catenin, resulting in a decreased expression of β-catenin target genes. This led to an inhibition of migratory potential through downregulation of MMP-9, and a decrease in expression of Cyclin-D, c-myc and Survivin. The latter was responsible for an anti-proliferative effect and an increase in cell death. These results were recently confirmed by Rubin et. al, [60] who showed that re-expression of Wnt Inhibitory Factor 1 (WIF-1), a secreted Wnt-antagonist, inhibited Wnt signalling and reduced tumour growth and metastasis in OS mouse models. These results show a possible therapeutic benefit of Wnt-pathway disruption in the treatment of metastatic OS.

Src-kinase is a kinase that is involved in almost all steps of cancer metastasis, namely in proliferation, adhesion, migration, survival and angiogenesis. [58] Based on its multi-step involvement in metastasis, it could be an interesting therapeutic target in OS metastasis. Pre-clinical work shows that Src inhibition with Dasatinib effectively inhibits Src phosphorylation in primary tumours; however, it did not impair the development of pulmonary metastases. Histopathological analysis of both OS primary tumours and lung nodules showed minimal Src-kinase phosphorylation after treatment with Dasatinib. However, Src-kinase phosphorylation was low in untreated lung metastasis as well. This suggests that Dasatinib was effective in inhibiting Src-pathway activation in OS cells, but it is not clear what the phosphorylation status is during the stages of OS metastasis and how this influences the process. [33] The use of Dasatinib in patients with advanced (osteo)sarcomas was examined recently by the SARC (Sarcoma Alliance for Research through Collaboration) in a phase II clinical trial. Disappointingly, preliminary results show no treatment effect of Dasatinib as a single agent in patients with overt lung metastases. [74] The same group is looking into the effectiveness of Src inhibition with a more specific Src-kinase inhibitor (Saracatinib) to obtain progression free survival among patients with resectable OS lung metastases (clinicaltrials.gov/NCT00752206).

(3) Evasion of the Immune SystemAnother important precondition for the survival of metastatic cells is the evasion of the host immune surveillance throughout all the steps of metastasis. Tumour cells, either circulating or at the site of metastases, can modulate the immune system of the host in order to

21


2

achieve a survival advantage. Down-regulation of cell surface receptor HLA class 1 is one of such mechanisms. This impairs the recognition of tumour cells by the host cytotoxic T-lymphocytes. Tumour cells can also induce the production of immunosuppressive molecules such as IL-10. [25,75] Modulation of the immune system such that it recognizes and destroys (circulating) tumour cells would be a successful anti-metastatic treatment. Interferons are cytokines that can affect the recognition of tumour cells by the immune system by influencing the (re)expression of HLA molecules on the cell surface. Interferons also exert an anti-proliferative effect on OS cells through pathways that are yet unknown. [75] The balance between the intrinsic downregulation of HLA molecules of the tumour cells and the effect of Interferon stimulation will eventually determine whether the circulating tumour cell is cleared by the immune system or not.

Fas also plays a role in immune evasion. Fas expression leads to recognition by, and activation of cytotoxic natural killer (NK) cells and promotes elimination from the circulation by the host immune system. Successful down-regulation of the Fas molecule on the cell surface or corruption of downstream elements in the Fas pathway provides metastatic tumour cells with a survival advantage in the circulation and leads to an increase in metastatic potential. Patient samples from pulmonary OS metastases have been shown to be Fas-negative. [54,76] Indeed, absence of Fas expression correlates with disease progression and poor survival outcome. [38,50,69,76](Pre)clinical studies: FasAs the Fas receptor pathway is so important in the survival of metastatic cells, it is an attractive therapeutic target. Restoration of the Fas death pathway has been tried with success in preclinical models. Interleukin-12 (cytokine) therapy can achieve a dose-dependent upregulation of Fas on the surface of OS cells as well as a stimulation of cytotoxic T-cells and NK-cells. This renders the metastatic cells more sensitive to Fas-induced cell death in the microenvironment of the lung and enhances clearage of the cells from the circulation by the host immune system. [50] A drawback is the potent immunostimulatory effect of Interleukin-12 that can induce severe adverse effects after systemic administration in patients. [70,76]

In an in vivo experiment, intranasal administration of IL-12 resulted in Fas overexpression on OS lung metastases, leading to a decrease in tumour burden. Combination therapy with Ifosfamide, which induces the expression of FasL on the tumours, could further augment anti-tumour effect. [7,70]

IL-18 was reported to have similar effects on the activation of T-cells and NK-cells, as well as induction of the expression of FasL on already Fas-expressing tumours. This compound did not, however, exert an anti-tumour effect in mice bearing OS lung metastases. [77]

Gemcitabine is an agent that upregulates Fas-expression when administered as an aerosol therapy in mice bearing OS lung metastases. Gemcitabine aerosol therapy has been shown

22

Chapter 2

to effectively reduce size and number of pulmonary metastases. [5,69]Liposomal MTP-PE (muramyl tripeptide phosphatidyl ethanolamine) is a promising agent

for clinical use as it can induce endogenous IL-12 production and thus provide an up-regulation of Fas on OS cells but without the systemic toxicity encountered when exogenous IL-12 is administered to patients. [7] MTP-PE is a synthetic analogue of a component of bacterial cell walls. As an immunomodulatory agent it can also stimulate monocytes and macrophages to exert anti-tumour activity. The Children’s Oncology Group performed a prospective randomised phase III clinical trial with this compound in patients with high-grade conventional OS with metastases at diagnosis. Treatment with liposomal MTP-PE improved overall survival, irrespective of the chemotherapy regimen. These data are promising and suggest that there is a critical role for the Fas death pathway in chemotherapy response which can be exploited in clinical practice to enhance the efficacy of chemotherapy in OS. [54,78,79](Pre)clinical studies: Immune modulationModulation of the immune system to exert anti-tumour activity by the addition of Interferon-α (IFN-α) as a maintenance treatment after standard chemotherapeutic treatment is currently under investigation in the EURAMOS-1 trial, which is an initiative of the European and American Osteosarcoma Study Group. (www.ctu.mrc.ac.uk/euramos) IFN-α is immunomodulatory and able to stimulate a host-anti-tumour immune reaction and induce anti-proliferative signalling via the JAK/STAT1 pathway. [72] Accrual of patients for this worldwide trial was completed in July 2011. [75]

(4) Arrest and extravasation The mechanism of arrest of metastatic tumour cells at the distant organ sites remains controversial. One hypothesis is that metastatic cells are larger than ordinary cells in the circulation and that they become trapped in the microcirculation of a capillary bed. When trapped they form micro-embolisms and start interaction with the local environment. It is striking, however, that different tumour types have an organ specific preference for metastasis. The metastatic behaviour of OS is very distinct as over 80% of all metastases arise in the lungs and other organs usually remain unaffected. This suggests that the circulating tumour cell specifically ‘homes’ to distinct molecules that are expressed on the endothelium of the organ of preference. Although it might be trapped in different capillary beds throughout the body, it will interact with the surface molecules on the endothelium of the organ of interest rather than with endothelium at other sites. [49] There is evidence of endothelium-specific tropism in OS. [25,27,30,37,56,80]

23


2

Table 2.1 Steps of metastasis in OS and molecular alterations that contribute to each process.

Steps of Metastasis Molecular involvement References

I Migration and Invasion MMPs [25] [27] [31] [33] [34] [42] [48] [49] [58] [59] [60]m-Calpain [49] [57]Wnt [26] [49] [60] [61]Src [49] [58] Notch [55] [62] [63] [64]

II (a) Anoikis resistance PI3K/Akt [25] [26] [31] [65]Src/PI3k/Akt [25] [26] [31] [58]Src/Ras/MAPK [33] [49]NF-κB [42]Wnt/β-catenin [25]BcL family [26] [49] [65]

(b) Apoptosis resistance Src [26] [31] [33] [49] [58]NF-κB [42] [49] [65] [80]Wnt/β-catenin [25] [32] [34] [60] [61] [67] [68] Fas/FasL [5] [7] [38] [50] [69] [70]

III Evasion of immune system HLA-1 [25] [75]IL-10 [25]Fas [76]

IV Arrest and extravasation CXCR4-CXCL12, [25] [26] [27] [30] [37] [56] [82]CXCR3-CXCL9-11 [27]CXCR4/MMPs [26] [27] [30]CXCR3-4/Erk/NF-κB [27] [80]

V Adherence Ezrin/MAPK/Akt [25] [36] [40] [49]Ezrin/β4-Integrin/PI3K [83]CD44/Akt/mTOR [25] [34] [36]

VI Dormancy Integrin-α5β1 [86] [87]Integrin-α5β1/Erk/p38 [25] [49]Bcl-XL [25] [49]IGF/PI3K [88]ECM [86] [87]

VII Angiogenesis and Proliferation EGFR. PDGFR, VEGF, IGFR, TGF-β

[25] [26] [49] [54] [81] [84] [90] [91]

MMPs [26] [58] [91]VEGF/Erk/NF-κB [49] [81]VEGF/PI3K [49] [54] EGFR/Src/Ras/MAPK/STAT3 [26] [33] Src [25] [49] [58]Integrin/PI3K/Erk1-2 [26] [49] [80] [88]

Wnt/β-catenin/CyclinD-Survivin

[60] [66]

24

Chapter 2

Table 2.2 Preclinical and clinical studies targeting specific molecules in OS metastasis.

Steps of Metastasis Target Drug ReferencesI Migration and Invasion preclinical

Notch [55] [64]II (a) Anoikis resistance (b) Apoptosis resistance preclinical Wnt [60] [66] Src [33] clincial Src Dasatinib [74] www.clinicaltrials.gov/NCT00752206

III Evasion of immune system preclinical Fas IL12 [50] [70] [76] IL18 [77] Gemcitabine [69] clinical Fas Liposomal MTP-PE [78] IFN-α www.ctu.mrc.ac.uk/euramos [75]

IV Arrest and extravasation preclinical CXCR4 [37] CXCR3 [27]

V Adherence preclinical Ezrin [36] [40] [59] [83]VI Dormancy VII Proliferation and Angiogenesis preclinical Endostatin [96] IGF-1R [92] [97] clinical IGF-1R OncoLar [91] R1507 www.clinicaltrials.gov/NCT00642941 SCH717454 www.clinicaltrials.gov/NCT00617890

The processes of exit of the circulation and invasion at the distant organ site are mediated by chemokines and proteinases. Proteinases are responsible for extravasation whereas chemokines determine the site at which circulating tumour cells adhere. [26,30] Chemokines were initially thought to regulate leukocyte trafficking and homing, but recently they are also known as important components in the regulation of site-specific metastasis as they

http://www.clinicaltrials.gov/NCT00752206

http://www.ctu.mrc.ac.uk/euramos



25


2

bind to G-protein coupled receptors on the plasma membrane of specific cells, in the case of OS to receptors in the lung. [25,56,80-82] CXCR-4, a commonly expressed chemokine in OS, is involved in site-specific metastasis. Its sole ligand is CXCL12 which is expressed abundantly in the lung. Binding of CXCR-4 to CXCL12 allows adherence and extravasation of OS cells in the lung. [25-27,30,37,56,82] Laverdiere et al. [56] found that CXCR-4 expression levels in patient samples inversely correlated to event-free and overall survival. There was a positive correlation between CXCR-4 expression in primary tumours and the presence of metastases at initial diagnosis. Interestingly, expression levels of CXCR-4 were similar in primary tumours and lung metastases. This suggests that CXCR-4 expression is not regulated during metastasis, but is simply present. It could be of predictive value for the formation of metastasis.

CXCR-3, another chemokine, is expressed by OS as well as other malignancies. Its ligands are CXCL9, -10 and -11, all of which are expressed by lung, and this molecule is thought to co-operate with CXCR-4. Apart from mediating adherence, the interactions of CXCR-3 and -4 with their respective ligands also trigger pathways involved in other necessary events in metastasis, namely in invasion, survival and proliferation in the secondary tissue. [27,37,80] For example, hypoxia upregulates CXCR-3 and -4 expression, which in turn induce the expression of MMP-2 and -9 on the cell surface and modulate the microenvironment into an inflammation-like condition, abundant with growth factors and stimulation of angiogenesis. Furthermore, binding of CXCR-4 to CXCL12 can activate the NF-κB survival pathway via ERK (Extracellular-signal-Regulating-Kinase) -signalling and stimulate proliferation through MAPK signalling. Thus, apart from facilitating seeding at the distant organs site, chemokines play a very important role in the modulation of the microenvironment into a place permissive for the tumour cells to proliferate. [26,27,30,80](Pre)clinical studies: ChemokinesCXCR-4 is the most important chemokine-player in OS. Kim et al. [37] have demonstrated a reduction in metastatic tumour burden in an orthotopic mouse model in which cells were treated with a CXCR-4 inhibitor prior to injection of tumour cells into the mice. However, reduction of metastatic tumour burden without pre-treatment could not been shown consistently. The authors argue that the critical event, namely binding of CXCR-4 to CXCL12 with consecutive activation of signalling pathways, granting survival and proliferation, occurs too early in the establishment of metastases for inhibitory therapy of CXCR-4 to be beneficial for the patient with already existing metastasis. To what extent CXCR-4 inhibition could be beneficial in a preventive setting requires additional studies.

CXCR-3 inhibition was tested in an animal model for human OS lung metastases and showed a significant decrease in the development and progression of pulmonary lesions compared to the non-treated group. [27]

26

Chapter 2

(5) AdherenceEstablishment at a distant organ requires the metastatic cell to connect to its new environment and re-establish cell-cell adhesions. Ezrin is a membrane-cytoskeleton linker protein that plays an important role in cell - microenvironment interaction. It is thought to facilitate anchorage of OS cells to lung tissue, as well as to enhance survival mechanisms in the new environment through Integrin mediated activation of Akt and MAPK survival pathways. [25,36,40,49] The exact mechanism through which Ezrin mediates metastasis is not entirely clear, however, recently Wan et al. [83] discovered that β4-Integrin is an important mediator. β4-Integrin can bind Ezrin and Ezrin is required for the maintenance of this protein. β4-Integrin can activate the PI3K pathway and thus stimulate survival and proliferation in the newly arrived cells in the lung. β4-Integrin is found to be highly expressed in OS tumour samples from both primary and metastatic lesions. Furthermore, it was shown that β4-Integrin knockdown inhibits the formation of OS lung metastasis in vivo, and leads to prolonged survival.

High expression of Ezrin correlated with a higher risk of metastatic relapse and poor survival in OS patients. [36,40] Furthermore, it was found to be 3-fold overexpressed in lung metastases in a murine model for OS lung metastases. [52] CD44 is another surface molecule that can form a complex with Ezrin and correlates with metastasis and poor prognosis. Apart from influence on the cytoskeleton and cell shape, CD44 controls proliferation, growth arrest and survival via the Akt/mTOR pathway. [25,34,36](Pre)clinical studies: EzrinSuppression of Ezrin with a full-length anti-Ezrin construct did not inhibit primary tumour growth in a mouse model of OS, but it effectively inhibited the formation of metastases. It was speculated that metastatic OS cells express phosphorylated Ezrin only early after arrival in the lung, and this causes limited efficacy of suppression of Ezrin in readily established metastases, since its essential function in metastasis, namely connecting with the target organ site had already been fulfilled. [36] Recently however, Ren et al. [40] suggested that Ezrin phosphorylation is not only present in the early stage of metastasis, but also late in tumour progression, at the leading edge of large metastasic lesions. This finding was verified on sections of patient OS metastases.

Pignochino et al. [59] reported that Sorafenib inhibited invasion via reduction in MMP-2 production and inhibited survival via downregulation of Ezrin-activated MAPK/Akt signalling. Furthermore, Sorafenib could also induce apoptosis in OS cells through downregulation of members of the anti-apoptotic Bcl-2 family. Wan et al. [83] showed that inhibition of Ezrin-related β4-Integrin can reduce metastasis in a mouse model. Taken together, targeting Ezrin seems promising in the management of OS lung metastases.

27


2

(6) DormancyDormancy refers to a prolonged period of survival of single cells or small micrometastases. OS patients can progress with metastases after a disease free interval of many years. [84,85] This is most likely explained by the presence of micrometastases in a dormant state, which at some point are triggered develop into gross metastases.

Little is known concerning the biological processes regulating dormancy in OS. The anti-apoptotic gene Bcl-XL is thought to be involved in the survival of dormant cells, as well as α5β1-Integrin mediated activation of NF-κB. Furthermore the dormant state is thought to be regulated by the ratio between the ERK and p38-MAPK proteins, also steered by Integrin-α5β1 signalling. [25,49,86,87]

The mechanisms by which dormant tumour cells are at one point triggered to start proliferating are yet unaccounted for, however, the microenvironment is thought to play a regulatory role. Tumour outgrowth is dependent on vascularisation, and it has been suggested that endothelial cells in the microenvironment can both activate dormant tumour cells through cell-to-cell signalling and induce angiogenesis for nutrition. [87] The ECM is also thought to be involved in activation of dormant cells, as it serves as a source of growth and survival signals. It has been postulated that micrometastases that fail to properly connect to the ECM remain in the dormant state because they remain deprived of growth- and angiogenic signalling and go into quiescence as a means to survive. Anchorage to the ECM would stimulate cells to convert to a proliferative state via β1-Integrin signalling. [86] The microenvironment can be regulated by the tumour cells themselves, but also by host stromal cells. Leucocytes and macrophages can modulate the ECM to either form a pro- or anti-angiogenic microenvironment. Apart from this, other mediating factors can also be influenced by stromal cells. For example, Wnt can be secreted from macrophages, and cytokines secreted by stromal cells can upregulate the intracellular Wnt/β-catenin signalling pathway and hence induce survival and proliferation in a late stage in the process of metastasis. [26,49,68,86] Also, bone marrow derived progenitor cells (creating a ‘pre-metastatic niche’) can modulate the microenvironment and thus influence whether solitary cells or micrometastases remain dormant or are allowed to progress. [81,86] In an effort to elucidate the cellular mechanisms that establish the switch of dormant to rapidly growing cells, Almog et al. [88] designed an in vivo model for dormancy of various cancers, including OS, and performed gene-expression analysis of cells in the dormant state versus cells in a proliferative state. They found that during dormancy, there is an upregulation of anti-angiogenic proteins. In this pre-angiogenic situation, the tumour cells would lack the nutrition and oxygen needed to proliferate. The cells that had switched to the proliferative phenotype had elevated RNA levels of common cancer pathways such as PI3K- and IGF-pathways. They also found that Endocan was upregulated in rapidly proliferating cells, a protein that is also expressed on tumour endothelial cells. This might indicate that endothelial changes support the switch of cells from dormancy into the proliferative state.

28

Chapter 2

Dormancy could have a role in therapy resistance in OS metastases, however, whether this applies to OS and to what extent remains unknown. In general, dormancy can bring about drug resistance because non-proliferating cells are not so susceptible to conventional treatment. Most treatment modalities induce DNA damage which is usually more lethal to rapidly proliferating cells. [25,86,89] To intervene in this step of metastasis seems difficult. Angiogenesis seems to be an important factor. Elucidation of mechanisms that steer the switch from dormant to proliferative state may give some options. If it would be possible to keep the cells locked in the dormant state, it may grant the patient stable metastatic disease with prolonged survival.

(7) Angiogenesis and Proliferation Tumour growth and progression is often restricted by vascularisation and thus nutrition. Hypoxia leads to the upregulation of growth factor receptors, angiogenic cytokines and proteolytic enzymes, among which EGFR, PDGF-R, VEGF, IGF-1R, TGF-β, IL-8 and MMPs, all of these providing neo-angiogenesis and allowing proliferation. These molecules can be overexpressed by the tumour cell population itself, but can also be provided by host endothelial (progenitor) cells during neo-angiogenesis. [25,26,49,54,58,81,90] Apart from induction of neo-angiogenesis, VEGF also provides the tumour cells with a survival benefit via activation of ERK-1/2 /NF-κB and PI3K pathways. [49]

Players involved in provision of vasculature and nourishment are often encountered in other processes within OS metastasis as well. For example, Src-kinase activity is regulated through various growth factor receptors, such as EGFR and Integrin receptors. Src activation leads to Ras/MAPK signalling and activation of the transcription factor STAT3, allowing cell cycle progression and production of angiogenic factors such as fibroblast growth factor, VEGF and IL-8. Src phosphorylation by EGFR especially is considered to stimulate the onset of hyper-proliferation of tumour cells and induction of vascular permeability and neovascularisation. [33,58]

Proliferation of OS cells at a distant organ site is often mediated by Receptor-Tyrosine-Kinase or Integrin induced activation of PI3K and ERK1/2 pathways. [26,49,80] Alterations in cell cycle regulation can also promote proliferation by facilitating progression through the cell cycle checkpoints and speeding up the cycle. For example, the Wnt/β-catenin pathway is of influence on both G1/S and G2/M progression via activation of Cyclin-D by c-myc and activation of Survivin respectively. [60,66]

The Insulin-Like-Growth-Factor 1 (IGF-1) Receptor axis is also implicated in the development of OS. It is striking that most OS arise during or shortly after puberty. The influence of GH and IGF-1 on bone growth steer the longitudinal growth during the adolescent growth spurt and contribute to approximately 50% of bone cell proliferation. As there is a peak incidence of OS during the adolescent growth spurt, it is conceivable that there could be GH/IGF-1

29


2

axis involvement in tumour development. IGF-1R signalling can activate the PI3K/Akt/mTOR pathway and stimulate survival and proliferation in tumour cells. [54,90-92] (Pre)clinical studies: AngiogenesisAs OS is a highly vascularised tumour, a rationale exists to use this feature as a therapeutic target. High serum-VEGF levels correlate with metastatic relapse, tumour progression, poor response to chemotherapy and a decrease in survival. [54,90,93] Endostatin is an endogenous angiogenesis inhibitor, produced by tumours itself, involved in repression of neo-angiogenesis and is commonly expressed in human OS samples. It can also induce apoptosis in endothelial cells. Given its important role in angiogenesis, it was hypothesised that Endostatin could impair OS tumour growth and metastasis. [33,90,94,95] However, in a murine model of OS lung metastasis, Endostatin failed to induce tumour shrinkage in the lungs, although, it did retard growth of lung nodules. Treatment with this drug will not cure OS patients, but it may result in stable metastatic disease with prolonged survival. [96](Pre)clinical studies: ProliferationPharmacologic inhibition of the GH/IGF-1 axis and thus IGF-1R pathways has been explored. However, whereas there is evidence that IGF-1R signalling is important to primary OS growth, the extent to which IGF-1R (inhibition) could regulate OS metastases is not clear. In 2002, Mansky et al. [91] performed a phase I study in OS patients with metastatic and/or recurrent disease testing the clinical efficacy of the Somatostatin analog OncoLar. OncoLar was shown to significantly reduce circulating IGF-1 in patients. However, all patients enrolled showed disease progression.

More recently, fully humanised monoclonal antibodies (mABs) directed against the IGF-1R were tested in preclinical and clinical setting. In vivo IGF-1R inhibition with monoclonal antibodies induced growth retardation in subcutaneous models of OS. [92,97] Whether this growth delay will also be shown in OS metastasis is unknown. The SARC-011 clinical trial is evaluating the treatment effect of R1507, a mAB targeting the IGF-1R in patients with recurrent sarcomas, including OS (clinicaltrials.gov/NCT00642941). In another clinical trial the efficacy of SCH717454, also targeting the IGF-1R, is evaluated in relapsed OS patients. In this trial, both inoperable patients and patients in whom metastasectomy is feasible are included. The latter group will be treated pre- and post-metastasectomy and might, apart from tumour response rate, give information about progression-free survival (clinicaltrials.gov/NCT00617890).

C O N C L U S I O NIn conclusion, this review summarises potential molecular alterations that contribute to metastasis in OS and gives an overview of (pre)clinical efforts to develop new therapeutic targets for the treatment of metastatic OS. In spite of these efforts, OS metastasis is not yet

30

Chapter 2

well understood and there have been few breakthroughs in the treatment of this disease over the last decade. We hypothesise that certain molecular alterations seen in metastatic cells can also contribute to resistance to chemotherapy making such alteration interesting targets for treatment. Further unravelling the biology of OS metastasis will hopefully provide new insights to be used as a rational basis for innovative metastasis directed treatments for OS.

31


2

Chapter 3 Mechanisms of therapy resistance

in osteosarcoma: a review

J. PosthumaDeBoer, B.J. van Royen, M.N. Helder

Oncology Discovery. 2013 Dec 31; 1:8.

34

Chapter 3

A B S T R A C TTherapy resistance remains a challenge in the treatment for osteosarcoma (OS). By studying therapy resistance and gaining biological understanding of resistance in OS, novel treatment targets to sensitise OS could potentially be discovered. The aim of this review is to give an overview of the mechanisms of resistance employed by OS cells after cytotoxic treatment, the key molecules involved in therapy resistance combined with the (pre)clinical research performed on this subject in a search to discover means to overcome therapy resistance and thus improve treatment efficacy for OS.

35

Mechanisms of therapy resistance in osteosarcoma: a review

3

I N T R O D U C T I O NOsteosarcoma (OS) is the most common primary malignant bone tumour in children and adolescents. The gold standard for therapy consists of a combination of multi-agent neoadjuvant chemotherapy, followed by radical surgery and adjuvant chemotherapy. With this aggressive regimen, 5-year survival rates of approximately 65% are obtained in patients with localised disease. However, in the case of metastatic or recurrent disease, 5-year survival rates are reduced to only 20%. [1,5,6,19,20] In patients with axial, head-and-neck and/or unresectable OS, local control is difficult to achieve and these patients run a high risk of local relapse and the development of metastatic disease. Therefore the prognosis for these patients is worse compared to patients with OS of the extremity and 5-year survival is approximately 25%. [1,18,19,98] Attempts to improve therapy efficacy by dose escalation, alterations in combinations of chemotherapy and irradiation therapy (particularly for patients with unresectable OS), either combined with chemotherapy or not, have not improved survival outcomes. [5,15-17,99-101] Resistance to therapy, both intrinsic and acquired, can be accountable for essential treatment failure as well as for recurrence after a disease-free interval. Methods to subvert therapy resistance and re-establish sensitivity of OS for existing treatments potentially increases therapy efficacy and thus survival. This review summarises the literature published in the past decade on various mechanisms of therapy resistance encountered in OS and how they could be meaningful in the design of targeted therapies to improve the sensitivity of OS to existing treatments.

R E S I S TA N C EBasic principles of anticancer therapy are proliferation inhibition and (re)activation of apoptosis. Cancer cells can utilise various mechanisms to circumvent or counteract the cytotoxic stimuli induced by anticancer therapy, including: (1) non-proliferative state (i.e., maintaining a stem-cell like phenotype and/or dormancy, (2) cell cycle alterations (3) enhanced drug efflux, (4) increased detoxification, (5) increased DNA repair (combined with cell cycle alterations) and (6) apoptosis resistance. [39,71,90,102] Table 3.1 provides an overview of these mechanisms, including key molecules involved in the particular mechanism. Resistance of cancer cells to therapy can be either intrinsic to the cancer cell, or acquired as the result of genomic instability. Acquired resistance would lead to the selection of resistant clones after a treatment that could, in a later stage of the disease, lead to relapse or metastatic disease progression. Intrinsic therapy resistance would lead to an essential poor response to therapy and account for the so-called ‘poor responders’. [7,26,47,51,103,104] Both types of therapy resistance are, in general, the result of specific molecules or molecular signalling in the OS cells. In the following section, each of the abovementioned mechanisms of therapy resistance will be elucidated and the contributing molecules in OS cells will be discussed, followed by a summary of previous (pre)clinical efforts to target the specific resistance mechanisms in attempts to improve treatment sensitivity in OS.

36

Chapter 3

Mec

hani

sms

of R

esis

tanc

eM

olec

ular

invo

lvem

ent

Dru

g Ta

rget

Dru

gRe

fere

nces

IN

on-p

rolif

erati

ve s

tate

(a)

Stem

-like

cel

lsP-

glyc

opro

tein

[107

] TG

F-β1

[106

]ST

AT3

[108

](b

)D

orm

ancy

PI3K

/Akt

[25]

[31]

[65]

[113

] In

tegr

in-α

5β1/

NF-

κB/E

RK/p

38-M

APK

[25]

[86]

miR

-190

(EPH

A5,

Ang

iom

otin)

miR

-190

miR

-190

ove

rexp

ress

ion

[110

]II

Cell

cycl

e al

tera

tions

G1

chec

kpoi

ntp5

3[[

43] 4

4] [4

5] [6

3] [1

14] [

118]

Cdk2

[102

] [12

7]S-

phas

e ch

eckp

oint

ATR/

Chk1

/Cdc

25A

Zole

dron

ate

[119

] [12

0]Ch

k1[4

5] [1

26]

Chk1

Caffe

ine

[118

]G

2/M

che

ckpo

int

ATM

/Chk

2/Cd

c25C

/Cdc

2-Cy

clin

B[4

3] [4

4] [4

5] [1

18] [

123]

Cdc2

(Cdk

1)[1

21] [

122]

[123

] [12

5] [1

26]

STAT

3[4

4] [9

5]N

F-κB

/Sur

vivi

n[6

5] [6

6]G

AD

D45

α (M

APK

3K-J

NK/

p39-

MA

PK)

[42]

[71]

[124

]Cd

c2 (C

dk1)

Flav

opir

idol

[45]

[102

] [12

6]Cd

c2-C

yclin

BW

EE1

inhi

bito

r[1

22]

Chk2

Caffe

ine

[118

]III

Redu

ced

drug

acc

umul

ation

P-gl

ycop

rote

in[6

5] [9

0] [1

02] [

129]

[130

] [13

1]P-

glyc

opro

tein

RNA

i: si

ABC

B1[1

35]

Ezri

nst

able

Ezr

in m

utan

t (sh

RNA

)[1

31]

Redu

ced

Fola

te C

arri

er[1

32] [

133]

[134

]IV

Incr

ease

d de

toxi

ficati

onG

luta

thio

ne S

-tra

nsfe

rase

P1

(GST

P1)

[39]

[124

]G

STP1

/JN

K- c

-Jun

/p38

-MA

PK/E

RK1/

2[1

36] [

137]

GST

P1-1

-TRA

F1N

BDH

EX[1

37]

GST

NBD

HEX

[136

] [13

7]V

Incr

ease

d D

NA

repa

ir(A

TM/D

NA-

PK) γ

H2A

X[1

38]

APE

-1[1

39]

ERCC

2 or

XPD

gen

e[1

40]

γH2A

Xsi

-γH

2AX

[138

]γH

2AX

miR

-138

ove

rexp

ress

ion

[138

]A

PE-1

si-A

PE1

[139

]

Mec

hani

sms

of R

esis

tanc

eM

olec

ular

invo

lvem

ent

Dru

g Ta

rget

Dru

gRe

fere

nces

VIA

popt

osis

resi

stan

cep5

3/BA

X/N

OXA

/PU

MA

/p53

AIP

1[6

5] [1

18]

p53/

MD

M2

Nut

lin(-

3a)

[145

] [14

6]Fa

s [3

8] [5

0] [6

9] [7

0] [7

6]

Fas

IL-1

2[5

0] [7

0] [7

6]

Fas

Gem

cita

bine

[7] [

69]

Fas

Lipo

som

al M

TP-P

E[5

4] [7

5] [7

8] [7

9] [1

47] [

148]

[149

] [1

50]

FasL

Ifosf

amid

e[6

9] [7

0]N

F-κB

/Sur

vivi

n[6

6]Cd

k7,C

dk9/

NF-

κBFl

avop

irid

ol[1

02] [

127]

JN

K/c-

Jun/

AP-

1[1

24] [

142]

JNK

JIP1

inhi

bito

r[1

42]

Wnt

/β-c

aten

in[3

2] [3

5] [6

6]W

nt- β

-cat

enin

/ Sy

ndec

an-2

[103

(Wnt

/β-c

aten

in) T

CF[1

03]

Synd

ecan

-2/e

ndot

helin

-1/

ERK1

/3/

-> P

IK3/

Akt

/Cal

pain

-6Ca

lpai

n-6

RNA

i si-c

alpa

in-6

[104

]Sy

ndec

an-2

/end

othe

lin-1

/NF-

κB[1

04]

mTO

R/PI

3K/A

kt[1

41]

mTO

RC1/

AM

PK[1

41]

CXCR

-3/-

4 /

MM

P-2/

-9[2

7] [3

3] [6

6] [8

0] [1

03] [

104]

[106

] [1

32] [

141]

CXCR

-4/C

XCL1

2/N

F-κB

[27]

[37]

[56]

[80]

Ever

olim

us[1

41]

EMT

[34]

[49]

[58]

TG

F-β

[26]

[61]

[64]

E-ca

dher

in/β

-cat

enin

(Wnt

/β-c

aten

in)

[35]

[60]

[61]

Aut

opha

gyH

MG

B1[1

44]

HM

GB1

[144

]

Tabl

e 3.

1 M

echa

nism

s of

resi

stan

ce in

OS

and

key

mol

ecul

es in

volv

ed p

er m

echa

nism

37


3

Mec

hani

sms

of R

esis

tanc

eM

olec

ular

invo

lvem

ent

Dru

g Ta

rget

Dru

gRe

fere

nces

IN

on-p

rolif

erati

ve s

tate

(a)

Stem

-like

cel

lsP-

glyc

opro

tein

[107

] TG

F-β1

[106

]ST

AT3

[108

](b

)D

orm

ancy

PI3K

/Akt

[25]

[31]

[65]

[113

] In

tegr

in-α

5β1/

NF-

κB/E

RK/p

38-M

APK

[25]

[86]

miR

-190

(EPH

A5,

Ang

iom

otin)

miR

-190

miR

-190

ove

rexp

ress

ion

[110

]II

Cell

cycl

e al

tera

tions

G1

chec

kpoi

ntp5

3[[

43] 4

4] [4

5] [6

3] [1

14] [

118]

Cdk2

[102

] [12

7]S-

phas

e ch

eckp

oint

ATR/

Chk1

/Cdc

25A

Zole

dron

ate

[119

] [12

0]Ch

k1[4

5] [1

26]

Chk1

Caffe

ine

[118

]G

2/M

che

ckpo

int

ATM

/Chk

2/Cd

c25C

/Cdc

2-Cy

clin

B[4

3] [4

4] [4

5] [1

18] [

123]

Cdc2

(Cdk

1)[1

21] [

122]

[123

] [12

5] [1

26]

STAT

3[4

4] [9

5]N

F-κB

/Sur

vivi

n[6

5] [6

6]G

AD

D45

α (M

APK

3K-J

NK/

p39-

MA

PK)

[42]

[71]

[124

]Cd

c2 (C

dk1)

Flav

opir

idol

[45]

[102

] [12

6]Cd

c2-C

yclin

BW

EE1

inhi

bito

r[1

22]

Chk2

Caffe

ine

[118

]III

Redu

ced

drug

acc

umul

ation

P-gl

ycop

rote

in[6

5] [9

0] [1

02] [

129]

[130

] [13

1]P-

glyc

opro

tein

RNA

i: si

ABC

B1[1

35]

Ezri

nst

able

Ezr

in m

utan

t (sh

RNA

)[1

31]

Redu

ced

Fola

te C

arri

er[1

32] [

133]

[134

]IV

Incr

ease

d de

toxi

ficati

onG

luta

thio

ne S

-tra

nsfe

rase

P1

(GST

P1)

[39]

[124

]G

STP1

/JN

K- c

-Jun

/p38

-MA

PK/E

RK1/

2[1

36] [

137]

GST

P1-1

-TRA

F1N

BDH

EX[1

37]

GST

NBD

HEX

[136

] [13

7]V

Incr

ease

d D

NA

repa

ir(A

TM/D

NA-

PK) γ

H2A

X[1

38]

APE

-1[1

39]

ERCC

2 or

XPD

gen

e[1

40]

γH2A

Xsi

-γH

2AX

[138

]γH

2AX

miR

-138

ove

rexp

ress

ion

[138

]A

PE-1

si-A

PE1

[139

]

Mec

hani

sms

of R

esis

tanc

eM

olec

ular

invo

lvem

ent

Dru

g Ta

rget

Dru

gRe

fere

nces

VIA

popt

osis

resi

stan

cep5

3/BA

X/N

OXA

/PU

MA

/p53

AIP

1[6

5] [1

18]

p53/

MD

M2

Nut

lin(-

3a)

[145

] [14

6]Fa

s [3

8] [5

0] [6

9] [7

0] [7

6]

Fas

IL-1

2[5

0] [7

0] [7

6]

Fas

Gem

cita

bine

[7] [

69]

Fas

Lipo

som

al M

TP-P

E[5

4] [7

5] [7

8] [7

9] [1

47] [

148]

[149

] [1

50]

FasL

Ifosf

amid

e[6

9] [7

0]N

F-κB

/Sur

vivi

n[6

6]Cd

k7,C

dk9/

NF-

κBFl

avop

irid

ol[1

02] [

127]

JN

K/c-

Jun/

AP-

1[1

24] [

142]

JNK

JIP1

inhi

bito

r[1

42]

Wnt

/β-c

aten

in[3

2] [3

5] [6

6]W

nt- β

-cat

enin

/ Sy

ndec

an-2

[103

(Wnt

/β-c

aten

in) T

CF[1

03]

Synd

ecan

-2/e

ndot

helin

-1/

ERK1

/3/

-> P

IK3/

Akt

/Cal

pain

-6Ca

lpai

n-6

RNA

i si-c

alpa

in-6

[104

]Sy

ndec

an-2

/end

othe

lin-1

/NF-

κB[1

04]

mTO

R/PI

3K/A

kt[1

41]

mTO

RC1/

AM

PK[1

41]

CXCR

-3/-

4 /

MM

P-2/

-9[2

7] [3

3] [6

6] [8

0] [1

03] [

104]

[106

] [1

32] [

141]

CXCR

-4/C

XCL1

2/N

F-κB

[27]

[37]

[56]

[80]

Ever

olim

us[1

41]

EMT

[34]

[49]

[58]

TG

F-β

[26]

[61]

[64]

E-ca

dher

in/β

-cat

enin

(Wnt

/β-c

aten

in)

[35]

[60]

[61]

Aut

opha

gyH

MG

B1[1

44]

HM

GB1

[144

]

Tabl

e 3.

1 Co

ntinu

ed

38

Chapter 3

(1) Non-proliferative stateMost conventional cytotoxic treatments modalities induce DNA damage and target rapidly proliferating (cancer) cells. [25,89,105] Contrarily, non-proliferating cells are not (so much) susceptible to DNA damaging therapy and therefore the non-proliferative state can serve tumour cells a survival benefit after the administration of treatment. Non-proliferative states are found in quiescent cells, cells with stem-cell like phenotypes and dormant cells. [86,88,106]

(a) Stem-like cellsThere is growing evidence that tumours may develop from cancer stem cells that have characteristics similar to normal/healthy stem cells but give rise to cell populations that do not mature into fully differentiated cells that hold the ability to divide infinitely, thus producing cancer cells. Many solid tumours are known to contain small populations of stem-like cells, or cancer stem cells. [107,108] In OS, both in cell lines and in biopsies from human OS specimens, populations of stem-like cells have been detected. When cultured the stem-like cells can form so-called sarcopheres that express mesenchymal stem cell markers such as CD133, CD117, CD105 and CD44. [106,108,109] These cancer stem cells, alike normal stem cells, have self-renewal capacity, can be quiescent (that is, resting in G0 for prolonged periods of time) and can repair their DNA. [107] Zhang et al. [106] show that cancer stem cells in spheres exhibit insensitivity to cisplatinum and doxorubicin treatment. The quiescent state, by definition, renders the stem-like cells less vulnerable for DNA damaging treatments. In addition, the ability to repair DNA is proposed to allow for new mutations in cells exposed to cytotoxic agents, and/or for the selection of therapy resistant cells that, after reactivation, possess a high clonogenic capacity and develop into overt tumours in a later stage of the disease. Also, the expression of ABC transporters, particularly P-glycoprotein encoded by the multidrug-resistance (MDR) gene, is reported to provide a resistant phenotype in stem-like cancer cells. [107] The importance of P-glycoprotein is more elaborately discussed in the section on drug efflux.

In OS, little is known about the molecules that steer the survival and activation of OS cancer stem cells. It was reported that Transforming Growth Factor β1 (TGF-β1) is a controlling molecule that can promote self renewal capacity, proliferation and chemoresistance in OS cancer stem cells. Inhibition of TGF-β1 reduced tumorigenicity and increased chemosensitivity in cancer stem cell spheres. [106] Furthermore, it was reported that in stem-like cells, STAT3 is activated to maintain self renewal capacity and pluripotency. This transcription factor is also implicated in apoptosis resistance, as will be discussed further below in the section on cell cycle. [108] Pathways involving IGF, MAPK/ERK, Wnt, Notch and JNK signalling have also been implicated in the maintenance of cancer stem cells. In addition, the tumour microenvironment, and low oxygen tension in particular, is proposed to play a role in de

39


3

maintenance of the stem-like state of cancer stem cells. In rapidly growing tumours, areas of (relative) hypoxia arise in which tumour cells display reduced cell metabolism and cell division rates, leading to a stem-like phenotype with diminished sensitivity to DNA damaging treatments. In addition, hypoxia was found to be instrumental in the maintenance of the pluripotency encountered in cancer stem cells. [106] Other studies show that hypoxia can instigate inflammation-like conditions in the tumour microenvironment that are favourable to the survival of (stem-like) cancer cells. [27,30,80] The interactions between tumour cells and the microenvironment shall be discussed in the section on apoptotic signalling. (Pre)Clinical studies: Stem-like cellsWhile it is noted that most conventional cytotoxic therapies do not kill stem-like cancer cells, studies exploring these cells as therapeutic targets are extremely rare. Perhaps the fact that cancer stem cells only comprise a few percent of all cells within the tumour precludes the efficacy of targeting these cells on beforehand. Nonetheless, if indeed these cells are responsible for the development and growth of malignancies, it might be paramount to identify crucial survival molecules in cancer stem cells and gain new insights in how to combat this very small, though very essential cell population within OS.

(b) DormancyDormancy can also be considered a non-proliferative state and refers to a prolonged period of survival of single cancer cells or micro-metastases and is typically encountered during the process of metastasis. Dormant tumour cells can go undetected for years or even decades prior to overt outgrowth. This phenomenon can explain why patients can relapse or show progressive metastatic disease after a (sometimes considerable) disease-free interval. [84,85,89,110] The molecules regulating dormancy in OS remain largely unknown, however, it is likely that the molecules that ensure survival in single cells (i.e. in the case of anchorage independent survival) provide a survival benefit that also increases resistance to cytotoxic stimuli. The tumour microenvironment is also thought to play a role in both dormancy of tumour cells and the reactivation of dormant tumour cells via cell-cell interactions, cell-extracellular matrix interactions or through chemokines present in the microenvironment. [111] The influence of cell surface receptors and chemokines on cell survival shall be further discussed in the section on apoptosis resistance. In order to survive in the absence of cell-cell and cell-matrix signalling, the single tumour cells must rely on altered intracellular signalling to prevent apoptosis, or more particularly anoikis to occur. [112] It has been shown that ainoikis resistant OS cells can intrinsically activate PI3K and/or Akt survival pathways; both pathways that can possibly also contribute to therapy resistance. [25,31,65,113] The anti-apoptotic gene Bcl-XL, that is reported to provide prolonged survival of dormant tumour cells, can possibly also convey insensitivity to chemotherapy in these cells. [26,49] Another molecule considered important in the survival of dormant cells is Integrin-α5β1 that acts

40

Chapter 3

via activation of the NF-κB pathway and via modulation of the ratio between ERK and p38-MAPK kinases, tilting the intracellular balance towards survival. [25,86] (Pre)Clinical studies: DormancyThe insensitivity of dormant tumour cells to cytotoxic therapy might not be an issue per se as long as the cells remain dormant, however, once the cells switch into the proliferative state, we are confronted with highly proliferative therapy resistant cells. It could be argued that the cells should be kept dormant in order to address this issue optimally. In this light, some state that tumour dormancy could offer opportunities in anticancer treatment, however, the lack of appropriate models impairs the design of new strategies. Recently, Almog et al., [110] presented work on regulatory micro RNAs (miRs) that steer the switch between dormant cells and fast-growing cells of the ‘angiogenic phenotype’ in various cell types including OS. They show that, among others, overexpression of miR-190 could suppress the angiogenic phenotype and tumour growth in an in vivo model of OS and impose a dormant state of the tumour cells. Possible genes that play a role here are EphA5 and Angiomotin. Further insights in this switch from dormant to proliferative and efforts to interfere with this switch could offer promising opportunities for future treatments.

(2) Cell Cycle AlterationsAlterations in cell cycle regulation are encountered in many types of cancer, including OS. Cell cycle alterations are excessively implicated in tumorigenesis, genomic instability and drug resistance. [44,45,114] Arrest in the cell cycle grants the cells extra time to repair their damaged DNA. Normally, re-entry in the cell cycle will only occur after DNA repair has been realised, but in selected cases cells are allowed to progress through the cell cycle without repair being fully completed. This attributes to genetically more instable clones which often have a more aggressive and invasive phenotype. [7,44,46,63] The cell cycle is tightly controlled and has three important checkpoints to maintain the genome. In general, these checkpoints are meant to prevent cells from replicating with damaged DNA, and in the case of DNA damage, to obstruct the cells from entering mitosis. Figure 3.1 gives an overview of the cell cycle regulation, with key players involved in drug response.

Many malignant cells, including OS cells lack a functional p53 pathway. The rate of p53 alterations in patient tumour samples ranges between 22-45%. [14,115-117] In the absence of functional p53, the G1 cell cycle checkpoint is often surpassed and the other cell cycle checkpoints become of more importance to the tumour cell for DNA damage management and survival. [43-45,63,114,118] Both the S-phase and the G2/M-phase checkpoint have been described in drug response in OS and arrest of the cell cycle in G2/M-phase is commonly observed upon DNA damage in malignant cells of various types, including OS. [44,45,119-122] In the case of damage, DNA double strand breaks are detected by ATM, which relays a signal (mostly) to Chk2, which in turn phosphorylates Cdc25C. Cdc25C

41


3

can phosphorylate the universal controller regarding the onset of mitosis, namely Cdc2. Inhibitory phosphorylation of Cdc2 inactivates the Cdc2/Cyclin-B complex, prevents the G2/M transition and consequently prevents the cell from entering mitosis, thus allowing time for DNA repair. [43-45,118,123]

Many other molecules are involved in regulating the cell cycle at the various checkpoints. STAT3 is a transcription factor associated with resistance to chemotherapy-induced apoptosis in OS. While the exact mechanism of action remains unknown to date, there is an implication in cell cycle regulation at the G2/M checkpoint and possibly in the activation of survival stimulatory genes. [33,95]

Survivin is a protein known to have anti-apoptotic properties in OS, but it also has an effect on the cell cycle. It lies downstream in the NF-κB pathway and can regulate mitotic entry at the G2/M checkpoint. [65,66] Overexpression of Survivin allows for aberrant cell cycle progression of cells with minor DNA damage through the G2/M checkpoint to enter mitosis, resulting in cells with increased genomic instability and a more aggressive phenotype.

GADD45α is another, nuclear, protein known to interact with various cell cycle related proteins. Repression or inactivation of this protein by NF-κB can make tumour cells escape

p38

cFos

p38

G2/M progression

Cyclin B

Cdk1

Cdc25C

Chk2/Chk1

ATM/ATR

Wee1

G1progression

Sphaseprogression

Sensor protein

DNA damage

p53

Cyclin D pRb

JNK

cJun

Notch

AP1

STAT3

MAPK

Ras

Src

βcatenin

cmyc

Cyclin A Cyclin E

Cdk4/6

Cdk2 Cdk2

p21

Wnt

Survivin

ATM/ATR

Cyclin E

Cdk2

Cdc25A

Chk1

CaffeineMyt1

MDM2

Caffeine

Figure 3.1 Representation of the cell cycle with key molecules involved in OS therapy resistance.

42

Chapter 3

programmed cell death. Defects in the expression of GADD45α are associated with drug resistance. Furthermore, this molecule is also thought to interact with MAP3-kinase which can activate JNK and p38-MAPK signalling pathways, further stimulating cell survival. [42,71,124] (Pre)Clinical studies: Cell Cycle AlterationsSince cell cycle alterations can greatly benefit OS tumour cells in surviving cytotoxic treatments, molecules involved in cell cycle regulation represent promising therapeutic targets. As a result, a magnitude of research has been performed concerning this topic, ultimately to abolish these cell cycle alterations and either induce cell death directly or to sensitise OS cells to existing treatments.

Zoledronate has been shown to induce growth retardation and apoptosis by induction of arrest in the S- and G2/M checkpoint. Upon treatment with Zoledronate, double strand DNA breaks occur, which activate the ATR/Chk1/Cdc25A cascade, leading to S-phase arrest, growth inhibition and possibly better tumour control. [119,120]

CDKs are key regulators of the cell cycle, therefore they represent a potential target for therapy. Cdc2 (also known as Cdk1) is a Cyclin-Dependent Kinase that plays a pivotal role in the onset of mitosis or the induction of a G2/M arrest. Abrogation of its inhibitory phosphorylation can abrogate the G2 arrest, initiate an unwanted acceleration of the cell cycle and force DNA damaged cells into mitosis, ultimately leading to enhanced apoptotic cell death. [43,45,121-123,125,126] Another way to intervene in cell cycle regulation is to manipulate CDK-regulatory pathways, rather than targeting the CDKs directly. For example, inhibition of Chk1 with a small molecule inhibitor has been reported to sensitise cells to DNA damage by abrogation of the G2 checkpoint leading to pre-mature mitotic entry and consequently to apoptosis. [45,126] Similarly, inhibition of Wee1 with a small molecule inhibitor can sensitise OS cells to irradiation-induced DNA damage by abrogation of the G2/M arrest and consequent cell death via mitotic catastrophe. [122]

Flavopiridol, a pan-CDK-inhibitor, was shown to exert effects on cell cycle regulation and enhance chemotherapy-induced cell death in vitro, also in chemoresistant cell lines. As Flavopiridol inhibits multiple CDKs, it can influence both the G1 checkpoint (via Cdk2) and the G2 checkpoint (via Cdk1). It can induce sustained checkpoint activation, followed by apoptosis via the mitochondrial pathway. Additionally, it would reduce the transcription of NF-κB via inhibition Cdk7 and Cdk9 and thus initiate apoptosis. [102,127]

Caffeine is a naturally occurring substance which influences the cell cycle at both the G1- and G2/M checkpoint. Although contradictory reports exist about this compound, the general consensus is that additional treatment with caffeine can override a G2/M arrest after DNA damaging treatment, induce apoptosis and therefore enhance the cytotoxic effects of chemotherapy. The reported effect seemed to be most potent in cells lacking a functional p53. The mechanism behind this abrogation of the G2/M arrest is contributed to

43


3

an inhibitory effect of caffeine on Chk1, resulting in Cdc25C activation, subsequent Cdc2/Cyclin-B activation and ultimately G2/M progression. [118] Preclinical studies with caffeine have shown a sensitisation to cisplatin in OS cell lines. Moreover, in a retrospective study, addition of caffeine to chemotherapy regimens in patients resulted in an increased efficacy of doxorubicin and cisplatin treatment. Patients with metastatic OS receiving the so called ‘caffeine potentiated chemotherapy’ had prolonged survival compared to those treated with chemotherapy alone. This finding supports the notion that cell cycle manipulation could enhance conventional chemotherapy. In the case of caffeine however, concern exists regarding the dose that needs to be administered, which is high and adverse effects may eventually limit the feasibility of this compound as an additive to chemotherapy. [118,128]

(3) Reduced drug accumulationP-glycoprotein is a transmembrane surface protein encoded by the MultiDrug Resistance 1 (MDR1) gene, its name already highly suggestive of a role in drug response in OS. [65,90,102,129] P-glycoprotein is indeed involved in resistance to various drugs as it functions as an efflux pump of both hydrophilic and hydrophobic compounds. In the case of OS, P-glycoprotein is especially active as an efflux pump for doxorubicin, but also vincristine and vinblastine, granting lower levels of intracellular (cytotoxic) drug accumulation and thus resulting is less cellular damage. The level of P-glycoprotein expression is reported to be correlated to the degree of drug resistance in tumours and to poor survival outcomes in OS. [90,129-131]

Caronia et al. [130], conducted a so-called pharmaco-genetic study evaluating drug response and survival outcomes for 102 OS patients in relation to a set of 24 candidate genes known to be involved in drug efflux. They showed that two members of the MDR family, namely ABCC3 and ABCB1 (encoding P-glycoprotein) were associated with a higher risk of death and inferior survival outcomes, further indicating the importance of P-glycoprotein in drug resistance in OS.

It has been proposed that the resistant phenotype of OS cells is not so much defined by the level of P-glycoprotein overexpression per se, but by its functional localisation on the cell membrane. Recently is was shown that Ezrin, a membrane-cytoskeleton linker protein that was previously reported to be instrumental to OS metastasis and survival of OS cells, [36,83] plays an essential role in P-glycoprotein function by maintenance of P-glycoprotein-actin connections, thereby securing its cell membrane localisation and functionality. [131]

Contrarily to increased drug efflux, reduced drug accumulation can also be realised by reduced influx into the tumour cells. In OS, the intracellular uptake of methotrexate (MTX) is mediated by the Reduced Folate Carrier (RFC). It was shown that in resection samples from patients with a poor response to induction chemotherapy, RFC levels were significantly lower compared with patients with a good chemotherapy response, suggesting that low RFC

44

Chapter 3

expression may confer resistance to treatment with MTX, most probably by impaired drug influx into the OS cells. [132-134](Pre)Clinical studies: Reduced drug accumulationIt is highly likely that inhibition of P-glycoprotein would lead to sensitisation of OS cells to chemotherapy and therefore this could potentially represent a new strategy to overcome drug resistance in OS. It was shown that direct downregulation of P-glycoprotein expression using siRNA directed against ABCB1 leads to an increased sensitivity to doxorubicin and methotrexate in an OS cell line. [135] Apart from targeting of P-glycoprotein expression directly, interference with P-glycoprotein function also holds promise for reduction or reversion of the resistant phenotype. For example, Brambilla et al. [131] showed that stable expression of an Ezrin mutant in OS cells abrogated the co-localisation of Ezrin and P-glycoprotein at the plasma membrane, thereby hampering P-glycoprotein function and reinstating sensitivity to cytotoxic therapy. Thus, p-glycoprotein represents a putative target for sensitisation of OS to chemotherapy.

(4) Increased detoxificationSome cytotoxic treatments result in the formation of free oxygen radicals, the presence of which is potentially lethal to cells. Glutathione S-transferase P1 (GSTP1) is an intracellular enzyme that catalyses Glutathione (anti-oxidant) detoxification. Increased expression of the GSTP1 gene as well as an increase in enzymatic activity of this protein is beneficial for tumour cells to survive chemotherapy, as the detoxification following oxidative stress occurs more rapidly. Both mechanisms have been shown relevant in therapy response in OS. GSTP1 is also assumed to regulate apoptotic signalling, namely to have an anti-apoptotic effect through interaction with c-Jun/JNK and regulation of p38-MAPK and Erk1/2. Activation of each of these signalling pathways leads to enhanced survival, which will be further discussed in the section on apoptosis resistance. [39,124,136,137] (Pre)Clinical studies: Increased detoxificationTargeting GSTP1 (and other glutathione-S transferase (GST) enzymes) could be a promising strategy to obtain sensitisation of OS to cytotoxic therapies. It was shown that resistant cell lines exhibit an increase in the expression of one or more GST isoenzymes as the cell lines acquire resistance. [136] NBDHEX is a GST targeting compound that disrupts complex formation between GSTP1-1 and TRAF2, resulting in prolonged JNK activation and ultimately to the induction of apoptosis. [137] Administration of NBDHEX to OS cells primarily led to impaired proliferation via G2/M arrest and after 48h of treatment apoptosis could be observed. [136,137] In a murine in vivo model of OS lung metastasis, a daily dose of NBDHEX proved to reduce both the formation and size of lung nodules compared to control. [136] The effect of combination treatment of NBDHEX and conventional cytotoxic drugs such as doxorubicin, cisplatinum and methotrexate was also investigated. NBDHEX

45


3

combined with doxorubicin or cisplatinum was shown to have an additive to synergistic effect on cell death, whereas in the case of methotrexate the opposite was observed and thus NBDHEX must be considered to have an antagonistic interaction with methotrexate. One possible explanation for this phenomenon would be that the NBDHEX-induced G2/M arrest slows down the cellular proliferation rate and thus renders the OS cells less sensitive to methotrexate induced DNA damage. [136] It is also speculated that the different response to NBDHEX to different cytotoxic drugs and across different cell lines could be attributed to the release of TRAF2 in the cytoplasm, leading to the activation of various MAPKs which in turn could mediate opposite effects on cell survival in a cell- and context-dependent fashion. [137] Taken together, targeting GST enzymes in OS may be a potential strategy to achieve increase therapy efficacy, either through direct effects of GST inhibition on cell viability, or via alterations in apoptotic signalling.

(5) Increased DNA RepairMost conventional cancer treatments are designed to damage the DNA of cells. [43,45] Enhanced repair of damaged DNA is a very efficient manner in which tumour cells evade chemotherapy-induced apoptosis. To achieve successful DNA repair, the first step is to signal that DNA damage has occurred, upon which repair proteins and pathways are recruited. At the same time, it is necessary for the cell to create time for repair, which is mainly realised by an arrest of the cell cycle as discussed previously. [138] Histone γH2AX plays an important role in the detection and repair of DNA damage. After DNA damage is signalled, activated ATM and DNA-PK kinases phosphorylates histone γH2AX which in turn recruits DNA repair proteins. [138]

After DNA damage has been detected, repair can occur through four basic mechanisms, namely direct reversal, base-excision-repair, nucleotide-excision-repair and mismatch repair. All of these mechanisms are likely to be present in OS.

Base-excision-repair (BER) as a means to repair DNA has been studied by Wang et al. [139] In BER, damaged DNA fragments are excised from the DNA, followed by polymerase B repair DNA synthesis. Enhanced repair of DNA damage after cytotoxic treatment in OS has been reported via this mechanism by upregulation of the APE-1 enzyme, which is a key enzyme in the BER-pathway. The expression of the APE-1 enzyme in OS is speculated to be of predictive value in treatment response.

Nucleotide-excision-repair (NER) typically repairs bulky DNA damage and is believed to be of influence on the response to cisplatinum. Recently, the effect of single nucleotide polymorphisms (SNPs) in NER genes in patients with OS was studied. It was shown that excision SNPs in the repair cross-complementing group 2 (ERCC2/XPD) gene correlated significantly with improved event-free survival, and this effect was directly proportional to the number of variant alleles in the gene, i.e. patients with a homozygously mutated

46

Chapter 3

genotype for this particular gene had better event-free survival compared to patients with a heterozygous mutation. Contrarily, no correlations were observed between mutations in NER genes and overall survival or response to chemotherapy. Nonetheless it is speculated that patients with XPD mutations may have an increased therapeutic benefit from platinum-based chemotherapy. [140] (Pre)Clinical studies: Increased DNA repairStudies investigating DNA repair mechanism as therapeutic targets are scarce. However, a few reports have been published. For example, it was shown that microRNA-138 targets γH2AX leading to reduced γH2AX expression levels, γH2AX foci formation and increased sensitivity to DNA damaging agents and irradiation. Furthermore, γH2AX downregulation using siRNA leads to deficiencies in homologous recombination. Combination treatment studies showed that both siRNA-mediated downregulation of γH2AX and γH2AX downregulation via overexpression of miR-138 lead to enhanced sensitivity to cisplatinum treatment. It is therefore speculated that miR-138 could represent a therapeutic target in OS. [138]

Preclinical studies on an OS cell line showed that downregulation of APE-1 using siRNA lead to an increased sensitivity of OS cells to treatment with thiotepa, etoposide and ionising radiation, yielding higher levels of apoptosis compared to cells with functional APE-1 levels. [139]

(6) Apoptosis ResistanceApoptosis resistance is a major issue in anti-cancer treatment failure. It is believed that the efficacy of cytotoxic therapies is supported by the activation of apoptotic pathways. The failure to induce apoptosis upon treatment is thought to be the result of a misbalance between pro- and anti-apoptotic signalling, in which there is a preference for the expression of anti-apoptotic genes. Restoration of this balance, creating an environment favourable of pro-apoptotic signalling should enhance treatment with cytotoxic agents. [26,49,59,65,66,70,71] Apart from changes in gene-expression, apoptotic signalling can be influenced by several different mechanisms. For example, constitutively activated receptor tyrosine kinase pathways (i.e. IGF, Wnt, ERK, PI3K, mTOR) that play a role in the development of OS can induce enhanced survival after cytotoxic treatments through aberrant survival signalling. Also, there is increasing evidence that the interaction between tumour cells and their microenvironment, via cell surface receptors and chemokine signalling, can lead to increased downstream survival signalling and thus confer resistance to cytotoxic therapies. [66,103,104,106,132,134,141] A different phenomenon observed in tumour cells (although more often in cells from an epithelial rather than a mesenchymal lineage) is the Epithelial-Mesenchymal-Transition (EMT), which is a type of dedifferentiation of tumour cells associated with a more primitive phenotype, increased malignant behaviour, metastasis and apoptosis resistance. [26,34,48,49,58] Therefore, these pathways and interactions shall also be discussed in this section on apoptotic signalling.

47


3

p53 is probably the best known regulator in DNA-damage management. Apart from regulation of the G1 cell cycle checkpoint, it can induce the expression of the pro-apoptotic proteins BAX, NOXA, PUMA and p53AIP1 and thus create a situation in favour of apoptosis. [65,118] Unfortunately, as stated in the above section on cell cycle, many OS lack a functional p53, which could impair apoptosis after cytotoxic therapy and confer therapy resistance.

The Fas receptor is a member of the Tumor Necrosis Factor (TNF) receptor superfamily and can induce receptor-mediated apoptosis. This extrinsic apoptotic pathway is activated by binding of Fas to its Fas-ligand (FasL), either on neighbouring cells, or by cross-linkage on one and the same cell. Fas/FasL interaction leads to activation of Fas receptor signalling which consecutively leads to cell death. The Fas death receptor pathway and aberrations in the cell surface expression of the Fas receptor on OS has been extensively studied, mainly in the context of OS metastasis. It has been observed that metastatic cell populations show a down-regulation of the Fas receptor on the cell surface. Consequently, Fas activation is impaired and Fas-induced apoptosis is reduced. It was shown that absence of Fas expression in lung metastases from OS patients correlated with disease progression and poor survival outcome. [38,50,69,76] Furthermore, the Fas-pathway is also reported to be of influence on chemotherapy-induced apoptosis. [70] Given its role in apoptosis and the observed down-regulation in OS cells with increased malignant behaviour, the Fas receptor might represent a potential target for the sensitisation of OS to existing therapies.

The NF-κB pathway is also involved in regulation of apoptosis after chemotherapeutic treatment. Upon activation it activates multiple anti-apoptotic proteins such as Inhibitor of Apoptosis Proteins (IAPs) and Bcl-XL, thus putting more weight at the anti-apoptotic side of the balance and providing a survival benefit to the tumour cells. Survivin, which has been noted earlier, is a known anti-apoptotic protein that also belongs to the NF-κB pathway and influences drug response in OS. [66]

JNK pathway activation is observed following cellular stress such as DNA damage, cytotoxic stress and γ-irradiation. However, the exact influence of JNK and it downstream molecules c-Jun and AP-1 on cell death is ambiguous and reported to be cell and context dependent. It seems that JNK exerts anti-apoptotic properties mainly in p53 deficient cells lines. [124,142]

The Wnt/ β-catenin pathway is implicated in the development and metastatic progression of OS. [32,35,66] Recently, it was reported that Wnt signalling is likely to influence doxorubicin response in OS cells by repression of syndecan-2 expression after the administration of doxorubicin. Syndecan-2 is a proteoglycan reported to influence both caspase-dependent and independent apoptosis and is upregulated by doxorubicin treatment to allow cell death. However, β-catenin activation was reported to impair syndecan-2 upregulation and therefore reduce sensitivity to doxorubicin in OS cells. [103] Syndecan-2 is reported to modulate ERK, PI3K/Akt and NF-κB pathways and thus influence cell survival after treatment with cytotoxic drugs. The proposed mechanism through which this occurs is

48

Chapter 3

that syndecan-2 inhibits endothelin-1-induced activation of ERK1/2 and PI3K/Akt which leads to reduced calpain-6 gene expression. Calpain-6 is implicated in positive stimulation of cell survival and proliferation in OS. Furthermore, the syndecan-2 mediated decrease in endothelin-1 signalling leads to reduced NF-κB activation and impaired survival signalling. [104] Thus, downregulation of syndecan-2 provides OS cells with a possible survival benefit and restoration or overexpression on syndecan-2 might be a method to sensitise OS cells to therapy.

Mammalian target of rapamycin (mTOR) signalling has been implicated in survival and proliferation of OS cells via the PI3K/Akt/mTOR pathway. [54,59,90,141] Very recently, a study was conducted to analyse the influence of mTOR on Sorafenib (a multi-kinase inhibitor) response in OS. It is observed that Sorafenib response in OS is commonly short-lived and that resistance to this treatment develops rapidly. mTOR can form two active protein complexes, mTORC1 and mTORC2, both of which can act as a component of PI3K/Akt signalling. It was shown that Sorafenib inhibits mTORC1, but contrarily activates mTORC2. It is noted that mTORC1 can activate AMP-activated protein kinase (AMPK) via ERK1/2 signalling and thus initiate apoptosis. Possibly, mTORC2 and/or the crosstalk between both complexes conveys resistance to Sorafenib and interfering with complex formation might overcome the resistance. [141]

The tumour microenvironment is known to influence important features in OS development, metastasis and survival. It can stimulate tumour cell survival and therefore the response to cytotoxic treatments via cell-cell interactions, cell-extracellular matrix interactions and chemokine signalling. In OS it was shown that hypoxia can upregulate the expression of chemokine receptors CXCR-3 and CXCR-4. The upregulation of these CXCRs can start a cascade in which CXCR expression leads to increased matrix-metalloproteinase-2 (MMP-2) and MMP-6 expression on the cell surface, leading to and increased degradation of the extracellular matrix surrounding the OS cell, leading to the release of inflammatory cytokines such as interleukins, providing a survival benefit for the OS cells. [27,30,80] Binding of CXCR-3 and CXCR-4 to their corresponding ligands (i.e., CXCL9/-10/-11 and CXCL12 respectively) stimulates OS metastasis, providing not only adherence of OS cells to their microenvironment, but also survival and proliferation. Binding of CXCR-4 to CXCL12 can also activate the NF-κB pathway via ERK and further induce anti-apoptotic signalling in OS cells. Furthermore, it was found that CXCR-4 expression in patients samples inversely correlated with survival in these patients, suggesting that CXCR-4 is a relevant molecule for patients suffering OS. [27,37,56,80] Thus, CXCR/CXCL interactions in OS cells lead to apoptosis resistance and therefore the disruption of CXCR/CXCL interactions may be and opportunity to sensitise OS cell to cell death after cytotoxic treatments.

A different phenomenon that is speculated to confer and increased malignant phenotype and possibly resistance to therapy is the Epithelial-Mesenchymal-Transition (EMT). EMT is a

49


3

type of dedifferentiation that is accompanied by a change in expression of surface molecules, namely the cadherins, that define either an epithelial or mesenchymal phenotype. It has been broadly implicated in the process of metastasis. Loss of E-cadherin and/or switch to N-cadherin on the cell surface of cells characterises EMT, resulting in a more mesenchymal, primitive phenotype. [26,34,48,49,58] EMT can be induced by TGF-β signalling, but other pathways such as the canonical Wnt signalling pathway, Notch pathway and MAP signalling are also associated with this process. Via the abovementioned pathways does EMT not only provide motility to OS cells, but also apoptosis resistance. [26,61,64] At the intracellular level, cadherin switching can stimulate survival via β-catenin; a downstream molecule in the Wnt signalling pathway. β-catenin is bound to the intracellular domain of E-cadherin by α-catenin. Loss of E-cadherin at the cell surface releases β-catenin into the cytoplasm and thus increases its cytoplasmic availability, inducing survival. [35,60,61] Of note is that E-cadherin is typically expressed by epithelial cells rather than mesenchymal cells and EMT has been more extensively studied in epithelial than mesenchymal tumours. Its importance in mesenchymal tumours has not yet been fully elucidated. However, Wheelock et al. state that cadherin switching can also include situations in which E-cadherin expression levels do not alter so much while there is a significant increase in N-cadherin expression. [143] This situation may be applicable to changes in cadherin expression in OS. The therapeutic opportunities to target or counter EMT in an attempt to sensitise OS cell to cytotoxic treatments most likely lie at the level of β-catenin and/or the Wnt pathway.

A less commonly described mechanism through which OS cells can exhibit chemotherapy resistance is autophagy. Autophagy is a cellular protection mechanism through which a cell can dispose of dysfunctional or damaged cellular components, and in doing this can prevent apoptotic cell death after cytotoxic stress. High Mobility Group Box 1 (HMGB1) protein is a bone active cytokine and, when overexpressed, associated with all hallmarks of cancer, including evasion of apoptosis. It is also reported to be a critical regulator of autophagy through its influence on the formation of Beclin1/PI3KC3 complexes and consequent vesicle formation. In OS cells, HMGB1 is reported to be overexpressed following treatment with doxorubicin, cisplatinum and methotrexate, leading to autophagy of damaged structures and thus creating a survival benefit for the treated cells. [144] (Pre)Clinical studies: Apoptosis resistance(Re)activation of apoptotic signalling, or impairment of anti-apoptotic signalling could theoretically lead to an enhanced therapy efficacy in OS. This field is intensively studied and many publications describe research efforts to exploit apoptotic signalling pathways to gain sensitisation of OS to a variety of treatments.

As stated in previous sections, p53 is an important regulator of the G1 cell cycle checkpoint, but also an important modulator of pro-apoptotic signalling, for example via BAX, NOXA, PUMA and p53AIP1. In OS, inactivating p53 mutations are encountered, as

50

Chapter 3

well as increased expression of negative regulators of OS such as mouse double minute 2 (MDM2) protein. [65,118,145,146] MDM2 is a ligase that binds to p53 and subsequently promotes its degradation. Nutlin-3a (Nutlin in short) is a small molecule inhibitor of MDM2 by binding at the p53 binding-site thereby rendering MDM2 inactive through prevention of MDM2-p53 binding. [145,146] Previous work from our group demonstrated that treatment with Nutlin induced cell death in p53 wild-type OS cells. Combination treatment of Nutlin with an adenovirus introducing exogenous p53 further augmented tumour cell kill. [145] Thus, in principle, combining conventional therapies with Nutlin could potentially increase therapeutic efficacy. Recently, however, a report was published describing that treatment of cells with Nutlin could possibly lead to newly formed mutations in p53 which in turn may lead to selection for p53-mutated cells. [146] This unwanted side effect may eventually limit the applicability of Nutlin in clinical practice.

The Fas death receptor pathway is a promising pathway for therapeutic intervention. It was shown that OS cells with decreased Fas expression have a survival benefit. Restoration of the Fas death pathway has been tested in different preclinical models of OS and results are promising. Upregulation of Fas expression can be achieved by the cytokine Interleukin-12 (IL-12) and by Gemcitabine. Liposomal MTP-PE (muramyl tripeptide phosphatidyl ethanolamine) is a synthetic analogue of a component of bacterial cell walls that can induce endogenous IL-12 production and thus indirectly induce an upregulation of Fas. Additionally, MTP-PE was also reported to stimulate macrophages and monocytes to exert anti-tumour activity. Furthermore, IL-12 could stimulate cytotoxic T-cells and NK-cells to clear OS cells from the circulation of the host. Additionally, Ifosfamide, commonly used in the treatment of OS was reported to augment FasL expression on the surface of OS tumour cells. [7,50,69,70] In an in vivo model of OS, the intranasal delivery of IL-12 was shown to establish an overexpression of Fas death receptor in OS lung metastases, leading to a decrease in tumour burden as single agent treatment already. Combination therapy of intranasal IL-12 with Ifosfamide further increased the anti-tumour efficacy. [70] Despite these promising pre-clinical findings, it must be noted that a major drawback in the use of IL-12 is its potent immunostimulatory effect that could lead to serious side effects such as fever, chills, headache, myalgia, nephro- and hepatotoxicity when delivered systemically. [7,76]

The use of liposomal MTP-PE could provide a possible solution for this issue because it can stimulate endogenous IL-12 production without inducing the systemic toxicity encountered after direct administration of exogenous IL-12. [7] Combination therapy of liposome encapsuled MTP-PE with conventional treatment agents has been proven to enhance therapy efficacy in vitro, in vivo and in trial setting. [147-150] The addition of liposomal MTP-PE to conventional treatment schedules has been shown to give improved overall survival in a phase III clinical trial in patients with high-grade conventional OS. [54,75,78,79] These results suggest that stimulating extrinsic apoptotic pathways can enhance the cytotoxic

51


3

potential of chemotherapy and are beneficial for patients suffering from OS. The role of and interfering with JNK-signalling after DNA damaging treatment in OS has

also been studied. [124,142] Although JNK activation is reported to have both pro- and anti-apoptotic properties in a cell and context dependent manner, we recently showed that interference of JNK-signalling, either using RNAi techniques or by small-molecule inhibition renders OS cells more sensitive to doxorubicin therapy. [142] Literature on JNK modulation and chemotherapy response in OS remains scarce and further studies need to be conducted in order to appreciate the potential suitability of JNK and its pathway as a target for therapy in OS.

Dieudonné et al. [103] describe how the Wnt/β-catenin/T-Cell Factor (TCF) pathway impairs the apoptotic response to doxorubicin and that modulation of this pathway could increase doxorubicin sensitivity. They show that downregulation of TCF transcriptional activity using a small molecule inhibitor increased syndecan-2 levels and consequently increased doxorubicin sensitivity in both sensitive and resistant OS cell lines. Others confirm that re-establishment of syndecan-2 expression sensitises OS cells to drug induced apoptosis in OS cell lines and in a murine model for OS. RNAi mediated knockdown of calpain-6 is also shown to re-sensitise cells with a resistant phenotype. It is shown that syndecan-2 is crucial in the control of calpain-6 expression and that (syndecan-2 mediated) downregulation of calpain-6 leads to enhanced apoptosis after treatment with doxorubicin. Furthermore, cells that stably overexpressed syndecan-2 were less tumorigenic and showed less invasive growth compared to controls. [104] Thus, modulation of syndecan-2 expression, either via inhibition of TCF of calpain-6 may be a strategy for the sensitisation of OS to doxorubicin treatment.

mTOR is a molecule that can form two active complexes, mTORC1 and mTORC2 and it was shown that mTORC2 mediates resistance of OS to Sorafenib treatment. Combination therapy regimens of Sorafenib and the mTOR inhibitor Everolimus however, reversed the resistant phenotype in OS cells and resulted in increased apoptotic response to Sorafenib. In OS cell lines, combination therapy of Sorafenib and Everolimus showed a synergistic effect on apoptotic cell death. Furthermore, combination treatment resulted in decreased cell motility and reduced tumorigenicity and angiogenesis in an in vivo model for OS. This reversal is attributed to disassembly of the mTORC2 complex, possibly resulting in increased mTORC1 complex formation, activation of AMPK and apoptosis. [141] Apart from providing insights in mTOR signalling and its role in resistance, this study also provides proof that new combinations of existing and registered drugs can lead to improved treatment efficacy in OS.

As described above, HMBG1 protein is involved in autophagy in response to cytotoxic treatments in OS. Huang et al. described that RNAi mediated knockdown of HMGB1 resulted in increased sensitivity of OS cells and an increase in apoptosis after cytotoxic treatment. Moreover, in an in vivo model, it was shown that HMGB1 knockdown tumours had inferior

52

Chapter 3

growth potential compared to controls, and, more importantly, showed increased therapy response compared to control tumours. [144] These data would lead to believe that if cells are not permitted to execute autophagy after cytotoxic stress, apoptosis is increased and thus targeting this mechanism could potentially help overcome therapy resistance in OS.

C O N C L U S I O NIn conclusion, therapy resistance remains a major issue in the treatment of patients with OS. Conventional treatments have reached their limits to improve survival outcomes in OS due to the encountered therapy resistance. This review summarises mechanisms and molecules that may contribute to therapy resistance in OS. A multitude of research in this field has already been conducted to discover new therapeutic options to sensitise OS to existing treatments. We hypothesise that successful treatment of therapy resistant OS should target molecules or pathways that are crucial in survival after exposure to cytotoxic treatment, and be given as a complement to conventional treatment. Thus far, re-establishment or augmentation of apoptotic signalling upon cytotoxic treatment seems to be most successful in the sensitisation of OS. Additionally, the advent of small molecule inhibitors provides the possibility to target certain molecules or classes of molecules very specifically. Ultimately, the goal is to establish treatments with a higher specificity and lower toxicity for patients suffering from OS. Despite all the preclinical research efforts endeavoured very few new agents have reached clinical use for OS patients. It seems that on the one hand, gaining a profound understanding of mechanisms that underlie and induce therapy resistance in OS should give us the opportunity to design targeted treatments for OS on a rational basis. On the other hand, to pursue promising sensitising treatments into clinical use might be even more crucial for the improvement of survival of OS patients. Maybe, centralised discussions to identify the most promising sensitising modalities and to reach consensus on strategy development may push forward a new era of treating OS.

53


3

Chapter 4WEE1 inhibition sensitises

osteosarcoma to radiotherapy

J. PosthumaDeBoer, T. Würdinger, H.C.A. Graat, V.W. van Beusechem, M.N. Helder, B.J. van Royen, G.J.L. Kaspers

BMC Cancer. 2011 Apr 29, 11: 156

56

Chapter 4

A B S T R A C TBackground: The use of radiotherapy in osteosarcoma (OS) is controversial due to its radioresistance. OS patients currently treated with radiotherapy generally have unresectable tumours, have painful skeletal metastases, refuse surgery or have undergone an intralesional resection of the primary tumour. After irradiation-induced DNA damage, OS cells sustain a prolonged G2 cell cycle checkpoint arrest allowing DNA repair and evasion of cell death. Inhibition of WEE1 kinase leads to abrogation of the G2 arrest and could sensitise OS cells to irradiation induced cell death. Methods: WEE1 expression in OS was investigated by gene-expression data analysis and immunohistochemistry of tumor samples. WEE1 expression in OS cell lines and human osteoblasts was investigated by Western blot. The effect of WEE1 inhibition on the radiosensitivity of OS cells was assessed by cell viability and caspase activation analyses after combination treatment. The presence of DNA damage was visualised using immunofluorescence microscopy. Cell cycle effects were investigated by flow cytometry and WEE1 kinase regulation was analysed by Western blot. Results: WEE1 expression is found in the majority of tested OS tissue samples. Small molecule drug PD0166285 inhibits WEE1 kinase activity. In the presence of WEE1-inhibitor, irradiated cells fail to repair their damaged DNA and show higher levels of caspase activation. The inhibition of WEE1 effectively abrogates the irradiation-induced G2 arrest in OS cells, forcing the cells into premature, catastrophic mitosis thus enhancing cell death after irradiation treatment. Conclusion: We show that PD0166285, a small molecule WEE1 kinase inhibitor, can abrogate the G2 checkpoint in OS cells, pushing them into mitotic catastrophe and thus sensitising OS cells to irradiation-induced cell death. This suggests that WEE1 inhibition may be a promising strategy to enhance the efficacy of radiotherapy in patients suffering OS.

57

WEE1 inhibition sensitises osteosarcoma to radiotherapy

4

I N T R O D U C T I O NOsteosarcoma (OS) is the most common primary malignant bone tumour in children and adolescents. The gold standard for treatment of OS consists of multi-agent neoadjuvant chemotherapy, radical excision of the tumour and adjuvant chemotherapy. [99-101] With this treatment regimen, 5-year survival rates of approximately 65% are obtained in localised disease. In patients with axial, head-and-neck and/or unresectable OS, local control is difficult to achieve and there is a high risk of relapse and/or metastasis and the prognosis for these patients is worse with a 5-year survival of around 20%. [1,18,19,98] Clearly, alternative treatment options for OS are warranted, especially for patients in whom local control can be scarcely achieved and therefore have a high risk of recurrence. Radiotherapy as a treatment modality for cancer has evolved over the past decades, but its use in OS treatment is controversial because OS is considered to be a radioresistant tumour. [19,98-100] At present, radiotherapy is applied only in a select group of patients with OS, namely those who suffer from unresectable (advanced extremity, axial or head-and-neck) OS, patients with painful bone metastases and patients who refuse surgery. Radiotherapy can give local control in OS when applied as an adjuvant therapy in patients who have undergone an intralesional resection of the primary tumour with subsequent irradiation of the contaminated surgical margins. [98-101,151-154] Technical progression in the field of radiotherapy has facilitated a more precisely localised delivery of radiation and thus warranted dose-intensification at the site of the tumour. This is a valuable development since the high irradiation doses needed for tumour control are usually difficult to achieve in patients with tumours that lie in the proximity of delicate structures, as is often the case in axial OS. Regularly, adverse side effects limit the dose that can be applied. Although presently still considered an advanced technique, the use of proton radiotherapy can be localised with even more precision to deliver a higher irradiation dose in the tumour while sparing adjacent healthy tissues. The toxicity and efficacy of this method in bone sarcomas is currently under investigation in clinical trial setting. [154,155] Moreover, the use of radiosensitising drugs has further improved the anti-tumor efficacy of radiotherapy. [55,98,100,151,156] Conventional chemotherapy has been shown to enhance the effect of radiotherapy in OS. Gemcitabine (with or without Docetaxel) and Ifosfamide have been shown to be potent radiosensitisers. [54,100] Also, the use of 153-Samarium can enhance the anti-tumor effect of external beam radiotherapy in axial OS. [98,100,152,156] Thus, chemotherapeutic agents may be used as radiosensitisers in OS patients. Additionally, small molecule inhibitor drugs may also serve as radiosensitisers. [43,156]

Radiotherapy, like many other cancer treatments, induces damage to the DNA. Prolonged activation of cell cycle checkpoints (arrest) is one effective method exploited by cancer cells to repair DNA and thus evade apoptosis after DNA-damaging treatments. [43,45,121,125,157] When cells progress through the cell cycle despite the presence of

58

Chapter 4

DNA damage, as a result, they undergo a mitosis specific cell death programme called mitotic catastrophe. [43,45,125,157-160] Cancer cells often lack a functional G0/1 cell cycle checkpoint and as a result rely mainly on the G2 cell cycle arrest to gain time for DNA repair. [3,7,44,45,160,161] Therefore, one strategy to sensitise OS cells to DNA damaging treatments is to exploit their vulnerability in defective cell cycle regulation and prevent them from repairing the damaged DNA during G2 arrest. WEE1 kinase plays a dominant role

Figure 4.1 WEE1 kinase regulates the onset of mitosis and is expressed in human OS. (A) The Cyclin-B/CDC2 complex is considered the master switch for the G2/M transition and CDC2 is designated the general controller in the onset of mitosis. Inhibition of the Cyclin-B/CDC2 complex prevents mitotic entry. CDC2 activation is essential for G2/M transition and thus progression through the cell cycle and is regulated on multiple levels. Inhibitory phosphorylation of CDC2 is achieved by WEE1 and prevents proper association with Cyclin-B. Consequently, Cyclin-B and CDC2 cannot form a complex and this complex cannot be activated. This induces a G2 arrest and prevents the cell cycle to progress until DNA repair is completed. Dephosphorylation of CDC2 by the phosphatase CDC25C activates the Cyclin-B/CDC2 complex, allowing cell cycle progression. The activation status of the Cyclin-B/CDC2 complex is dependent on the balance between WEE1 kinase and CDC25C phosphatase activity, in which WEE1 kinase is the rate limiting, dominant molecule. (B) WEE1 expression is significantly higher in the OS samples when compared to the normal tissue samples (ANOVA: p < 0.0001). (C) Immunohistochemical staining of sections of primary OS (extremity: II, III and axial: I, V) and OS lung metastasis (IVa,b). The brown nuclear staining indicates WEE1 expression in OS tumour tissue. (D) Western blot analysis of the effect of PD0166285 on WEE1 function. Irradiated (IR) cells show increased expression levels of CDC2-pY15 (CDC2-p) compared to untreated cells. After subsequent treatment with WEE1-inhibitor PD0166285 (WEE1-i), CDC2-pY15 expression levels were diminished, indicative of inhibition of WEE1 kinase activity. (E) Baseline expression levels of phosphorylated CDC (CDC2-p), the most important molecular target of WEE1 kinase, are distinctly lower in human primary osteoblasts compared to human OS cells.

βActin

CDC2p

MG

63

U2O

SSa

OS

2O

RT

1

Hum

31

Hum

54

G2/M progression

Cyclin B

CDC2

CDC25C

CHK2 / CHK1

ATM / ATRSensor protein

WEE1

1433σ

DNA damage

Ctrl

OS

n = 27

VNT

n = 504

WEE

exp

ress

ion

Ctrl

βActin

CDC2p

WEE1

MG63 SaOS2U2OS

IR +

WEE

1i

IR

IR +

WEE

1i

IR

IR +

WEE

1i

IRCtrl

59


4

in the sensitivity of cancer cells to DNA damage by inhibitory phosphorylation of Cyclin-Dependent-Kinase 1 (CDC2), thereby preventing mitotic entry, which is illustrated in Figure 4.1A. [43,45,46,121,123,125,157,161-165] It has been previously shown that PD0166285, a small molecule WEE1 kinase inhibitor, can abrogate the G2 checkpoint in cancer cells, forcing DNA-damaged cells into premature mitotic entry, thus inducing mitotic catastrophe and sensitising the cells to apoptosis. The anti-tumor activity of WEE1 inhibition in combination with DNA damaging treatments has been demonstrated in vitro as well as in in vivo models for different malignancies. [43,158,162,163] These promising preclinical results have led to the testing of a small molecule WEE1-inhibitor in a phase I clinical trial. [161] The aim of our study is to investigate if irradiation in combination with WEE1 inhibition could be used as a new therapeutic strategy to improve local control in the treatment of OS.

M E T H O D SCell culture, irradiation and compoundsHuman osteosarcoma cell lines MG-63, U2OS and SaOS-2 were kindly provided by Dr. C. Löwik (Leiden University Medical Center, Leiden, the Netherlands), Dr. S. Lens (Dutch Cancer Institute, Amsterdam, the Netherlands) and Dr. F. van Valen (Westfalische Wilhelms-Universität, Münster, Germany) respectively. Human primary (short-term culture) osteoblasts (ORT-1, Hum31 and Hum54) were obtained from healthy patients undergoing total knee replacement after informed consent. Cells were cultured in D-MEM (Gibco, Invitrogen) supplemented with 10% fetal calf serum (FCS) and 1mg/mL Penicillin-Streptomycin (Gibco) at 37°C and 5% CO2 in a humidified incubator.

Cells were irradiated in a Gammacell® 220 Research Irradiator (MDS Nordion) at doses varying from 2 to 10 Gray (Gy). The WEE1-inhibitor PD0166285 (Pfizer, Ann Arbor, MI, USA) was diluted in phosphate-buffered saline (PBS) to the desired concentration of 0.5 μM.

ImmunohistochemistryParaffin embedded tissue samples of primary OS and OS lung metastases, obtained from excision specimens from our institute, were deparaffinised and rehydrated in xylene and a graded series of alcohol respectively. Endogenous peroxidase was inhibited by incubation of the sections in 0.3% H2O2, diluted in methanol, for 30 minutes at room temperature. Antigens were retrieved by boiling for 10 minutes in citrate buffer (pH 6), followed by successive rinses in PBS containing 0.5% Triton and then in PBS only. Slides were incubated for 10 minutes in 0.1 M glycine (diluted in PBS) and rinsed in PBS prior to incubation with mouse-anti-WEE1 (SantaCruz) O/N at 4°C. Visualisation was performed using the Power Vision+ Poly-HRP IHC Kit (Immunologic) and tissue staining was performed with DAB chromogen solution. Slides were counterstained with hematoxylin, dehydrated and mounted. Placenta tissue served

60

Chapter 4

as positive control, prostate tissue served as negative control (not shown). Images were acquired at 20x objective.

Western BlotBasic expression levels of WEE1 and phosphorylated CDC2 in human OS cell lines and human primary osteoblasts were assessed by Western blot. Cells were lysed in phospho-lysis buffer containing Protease and Phosphatase Inhibitor Cocktails (Sigma). Proteins were quantified with the BCA protein Assay Kit (Pierce). A total of 40 μg protein was separated on a SDS-PAGE gel and transferred to a PVDF membrane, followed by incubation with the primary antibodies: mouse anti-WEE1 (SantaCruz), mouse anti-β-actin (SantaCruz) and rabbit anti-CDC2-pY15 (Abcam) and subsequently incubated with secondary goat-anti-mouse and goat-anti-rabbit immunoglobulins (DAKO). Protein detection and visualization was performed using ECL+ Western Blotting Detection Reagents (Pierce).

Inhibition of WEE1 kinase activity and concomitant phosphorylation of CDC2 by the WEE1-inhibitor PD0166285 was also analysed by Western blot analysis. Cells were plated and irradiated at a dose of 4 Gy in the presence or absence of 0.5 μM PD0166285. After 4 h treatment with 0.5 μM PD0166285, cells were lysed in phospho-lysis buffer containing protease and phosphatase inhibitors, followed by Western blot analysis as described above.

Cell Viability and apoptosis assayFor cell viability analysis, OS cells and primary osteoblasts were plated in 96-well format and irradiated at doses of 2, 3, 4, 6, 8 and 10 Gy. Cells were incubated with 0.5 μM PD0166285 or PBS directly post-irradiation. At 4 days (OS) and 9 days (osteoblasts) after treatment cell viability was assessed using the CellTiter-Blue Cell Viability Assay (Promega) according to the manufacturer’s instructions.

To analyse apoptosis, OS cells were plated in white opaque 96-well plates and treated with 4 Gy irradiation or with combination treatment of 4 Gy and 0.5 μM PD0166285. At 6 h and 24 h post-irradiation, caspase activity was measured using the Caspase-Glo 3/7 assay (Promega) according to the manufacturer’s instructions.

Fluorescence and luminescence read-out was performed using a Tecan Infinite F200 Microplate Reader (Tecan Trading AG, Switzerland). Results were analysed using GraphPad Prism® Version 5.01 (GraphPad Software, Inc. San Diego, CA, USA).

Flow cytometryCell cycle distribution analysis and quantification of the percentage of mitotic cells were performed using flow cytometry. Cells were plated and treated with 4 Gy irradiation, 0.5 μM PD0166285 or combination treatment. At 20 h after treatment, cells were trypsinised, washed in PBS containing 1% FCS and fixed in 70% ice-cold ethanol for 24 h. After fixation,

61


4

cells were washed with PBS containing 1% FCS and incubated with rat-anti-phospho-histone H3 (PHH3) antibody (BD Pharmingen) in PBS containing 1% BSA for 2 h at room temperature, followed by secondary antibody incubation with rabbit-anti-rat/FITC immunoglobulins (DAKO) in PBS containing 1% BSA for 30 minutes at room temperature in the dark. Cells were washed once and DNA was stained with 50 μg/mL propidium iodide (PI) solution in the presence of 250 μg/mL RNAseA (Sigma). The DNA content and the percentage of PHH3 positive cells were measured using a FacsCalibur Flow Cytometer and Cell Quest Pro software (Becton Dickinson) and results were subsequently analysed using ModFitLT software (Verity Software House, Topsham, ME, USA).

ImmunofluorescenceOS cells were seeded on glass cover slips in 24-well plates and treated with 4 Gy irradiation or with combination treatment of 4 Gy and 0.5 μM PD0166285. At 1 h and 24 h post-irradiation, cells were fixed in 2% paraformaldehyde. Prior to staining, the cells were rinsed in PBS and permeabilised in PBS containing 0.1% Triton X-100 for 30 minutes at room temperature and blocked in PBS containing 5% FCS. Slips were incubated with mouse-anti-γ-histone-H2AX (Millipore) in PBS containing 5% FCS O/N at 4°C, followed by secondary antibody incubation rabbit-anti-mouse/FITC immunoglobulins (DAKO) in PBS containing 5% FCS for 30 minutes at room temperature in the dark. Slips were rinsed in PBS thrice and nuclei were stained with DAPI (1:10 000) in PBS at room temperature in the dark, followed by successive rinses in PBS and sterile water. The slips were then mounted on glass slides, fixed with Mowiol and analysed with a Carl Zeiss Axioskop 20 microscope at 100x objective.

R E S U LT STo investigate whether WEE1 could be a suitable drug target in human OS we first explored its expression levels. From publicly available gene expression data in the GEO Expression Omnibus ( GSE14827 (http://www.ncbi.nlm.nih.gov/geo) we analysed WEE1 expression in 27 OS samples and 504 various normal tissue samples using the software programme R2. [166] We determined that WEE1 kinase is overexpressed in OS compared to various normal tissues, as shown in Figure 4.1B. When comparing the mRNA expression level of WEE1 in OS samples to the normal various tissue samples, one-way analysis-of-variance (ANOVA) shows that WEE1 expression is significantly higher in the OS samples (p<0.0001). In addition, we determined WEE1 protein expression in human OS tissue sections by immunohistochemical staining. Five out of 6 tested tumors had positive nuclear WEE1 staining (Figure 4.1C). The nuclear localisation of the protein is in concordance with its role in cell cycle regulation. These data indicate that WEE1 is indeed expressed by OS and could thus serve as a potential drug target.

62

Chapter 4

Next, we assessed whether PD0166285 can inhibit WEE1 kinase function by determining phosphorylation levels of its target CDC2 (resulting in CDC2-pY15) using Western blot analysis. Irradiated cells showed a moderate increase in WEE1 expression and a more profound increase in expression of CDC2 -pY15 compared to untreated cells (Figure 4.1D). This supports the notion that WEE1 kinase plays a role in the response to DNA damage through phosphorylation of CDC2. Subsequent treatment with PD0166285 diminished the expression of CDC2 -pY15 after irradiation. Thus PD0166285 effectively inhibits WEE1 activity and consequently reduces the inhibitory phosphorylation of CDC2 in irradiated OS cells. To analyse how baseline WEE1 and CDC2- pY15 levels in OS cells compare to normal cells, we performed a western blot analysis on both cell types. Figure 4.1E shows that CDC2- pY15

levels in human primary osteoblasts are negligible in comparison to the OS cell lines. WEE1 expression in the osteoblasts could not be visualised.

To investigate the effects of WEE1 inhibition on OS cell survival after γ-irradiation-induced DNA damage, we compared cell viability of irradiated cells in the presence or absence of the WEE1-inhibitor PD0166285. Figure 4.2A shows that WEE1 inhibition using PD0166285 at a non-toxic dose (0.5 μM) increased cell death after 2 to 6 Gy γ-irradiation in the OS cell lines MG-63, U2OS and SaOS-2 (p<0.01), whereas treatment with 0.5 μM WEE1-inhibitor alone showed no effect on cell viability (data not shown). To ascertain that WEE1 inhibition does not radiosensitise normal cells, we compared cell viability of human primary osteoblasts to osteosarcoma cell lines after 4 Gy irradiation, in the presence or absence of 0.5 μM PD0166285. Figure 4.2B shows that in the osteosarcoma cell lines there is a clear sensitisation to irradiation treatment, with approximately a 2-fold reduction in cell viability after combination treatment versus irradiation alone. In contrast, in the human osteoblasts no such effect is seen. There is a minor decrease in cell viability due to the irradiation treatment, but WEE1 inhibition does not enhance cell death. The results were consistent for all three tested human primary osteoblasts. From this we conclude that OS cell are indeed sensitised to irradiation by WEE1 inhibition whereas normal cells are not.

To investigate whether the sensitising effect of WEE1 inhibition in OS could be explained by mitotic catastrophe, we investigated three aspects of this phenomenon. We performed FACS cell cycle analysis of cells treated with 4 Gy γ-irradiation, 0.5 μM PD0166285, and combination treatment. Cells were stained with PI to analyse DNA content and with PHH3 to distinguish the fraction of mitotic cells from the cells in G2/M phase. Treatment with the WEE1-inhibitor alone did not alter the cell cycle distribution (Figure 4.3A). Irradiation of the cells resulted in arrest in the G2/M-phase, indicated by an accumulation of cells with 4N DNA content, but a stable percentage of mitotic cells. However, after treatment of the irradiated cells with the WEE1-inhibitor, a clear abrogation of G2 arrest was observed. Additionally, there was a 2 to 4-fold increase in the percentage of mitotic cells. Thus, WEE1 inhibition in irradiated cells leads to a forced progression through the cell cycle into mitosis. To assess

63


4

the extent of γ-irradiation-induced double strand DNA breaks (DSBs), we visualized the number of ionising radiation-induced foci (IRIF) with DSB marker γ-H2AX at 1 h and 24 h after irradiation in cells irradiated at a dose of 4 Gy in the presence or absence of 0.5 μM PD0166285. Figure 4.3B shows that DNA damage is visible at 1 h after irradiation. In the irradiated cells, only a few residual foci are detectable after 24h compared to the 1 h time point, indicating that DNA repair has occurred or is still ongoing. The shape of the nuclei is regular and there are no clear signs of apoptosis. In contrast, the cells treated with irradiation in combination with WEE1-inhibitor show extensive remaining DNA damage after 24 h with irregularity and fragmentation of nuclei indicative of nuclear envelope disassembly and apoptosis. From this we derive that in WEE1 inhibited cells DNA repair is not effectively realised.

To verify that cell death occurs as a result of apoptosis we analysed caspase activation in irradiated cells in the presence or absence of WEE1-inhibitor (Figure 4.3C). At 6 h post-irradiation there is a mild caspase activation in cells treated with irradiation alone or with

Figure 4.2 WEE1-inhibition sensitises OS cells, but not healthy osteoblasts, to γ-irradiation.(A) Indicated are dose response curves of irradiated cells (squares) and cell treated with IR + WEE1-inhibitor (triangles). The curves represent three experiments, performed in triplicate; error bars indicate standard deviation (SD). The sensitising effect of WEE1 inhibition is significant in all three cell lines (student’s t-test: p < 0.01). (B) Analysis of cell viability of human OS cell lines and human primary osteoblasts treated with 4 Gray (Gy) IR in the presence or absence of WEE1-inhibitor. Bars represent experiments performed in triplicate; error bars indicate SD. The osteosarcoma cell lines MG-63, U2OS and SaOS-2 show a 2-fold decrease in cell viability when treated with combination therapy, whereas human primary osteoblasts ORT-1, Hum31 and Hum54 show no sensitisation to radiotherapy in the presence of the WEE1-inhibitor.

64

Chapter 4

combination treatment. However, at 24 h post-irradiation there is a distinct difference in caspase activation between irradiated cells (4 to 6-fold) and cells treated with the combination of irradiation and WEE1-inhibitor (11 to 22-fold) (p<0.01 for all three cell lines). Taken together, this implies that cells treated with the WEE1-inhibitor are forced to proceed through the G2 cell cycle checkpoint into mitotic entry despite the presence of DNA damage and are thus sensitised to γ-irradiation-induced apoptosis.

Figure 4.3 WEE1 inhibition abrogates the γ-irradiation induced G2/M cell cycle arrest in OS cells and leads to mitotic catastrophe. (A) Histograms of FACS cell cycle analysis of OS cells treated with 4 Gy IR, 0.5 μM WEE1-inhibitor, and combination treatment. Percentages of cells in G2/M-phase and the mitotic index (MI) in percentages are shown. After irradiation, the percentage of cells in G2/M-phase is increased, whereas the percentage of mitotic cells remains unaltered. After subsequent treatment with WEE1-inhibitor, the G2 arrest is completely abrogated; the percentages of cells in G2/M-phase return to control values, while the percentages of mitotic cells increase dramatically. This indicates a forced progression through the G2 cell cycle checkpoint. (B) Fluorescence microscopy images of nuclei (blue) with ionising radiation-induced foci (IRIF) (green) indicating DNA damage, visualised using immunofluorescent γ-histone-H2AX staining of DNA breaks at 1 h and 24 h post-irradiation. Cells treated with IR alone show fewer IRIF at 24 h after treatment than cells treated with IR + WEE1-inhibitor. (C) Caspase activation in OS cells treated with 4 Gy IR or 4 Gy IR + 0.5 μM WEE1-inhibitor after 6 h and 24 h. Bars represent experiments performed in triplicate; error bars indicate SD. After 6 h, there is a mild induction of caspase activity. Caspase activation levels are comparable between the two treatment groups. At the 24 h time point there is a significantly higher caspase activation in the cells treated with combination treatment for all cell lines (student’s t-test: p < 0.01).

D I S C U S S I O NIn this work, we explore the possibility to use WEE1 inhibition as a new therapeutic strategy in OS. The use of WEE1-inhibitor PD0166285 to obtain radiosensitisation in various malignancies has been reported previously by other groups. [125,158,162,163] The radiosensitising effect is described to be particularly effective in, if not limited to, p53

65


4

deficient malignancies. [125,162] Interestingly, we have found that our tested cell lines can all be sensitised to irradiation, regardless of their p53 status (wt in U2OS, mutated in MG-63 and null in SaOS-2). This, we ascribe to the idea that a defective G1 checkpoint is not necessarily caused by p53 mutations alone but rather a disruption in the p53 pathway, which can be caused by other aberrations within this pathway. After irradiation, OS cells accumulate in a predominant G2 arrest, the abrogation of which effectively leads to mitotic catastrophe.

As was reported previously, [158,161] our results confirm that normal cells remain unaffected by WEE1 inhibition after irradiation. We tested human primary osteoblasts for their response to irradiation in the presence or absence of WEE1-inhibitor. While there was a minor effect of irradiation alone on cell viability, no radiosensitisation by WEE1-inhibitor PD0166285 was observed. This is likely explained by a wild type p53 expression with a concurrent functional G1 checkpoint. This indicated that WEE1 inhibition is a safe strategy to apply in OS patients because the radiosensitisation would be cancer cell specific and healthy tissues would remain unaffected.

Apart from being an important regulator of mitotic entry, WEE1 has been described to also affect other important cellular processes, such as regulation of mitotic spindle formation, positioning and integrity, microtubule stabilisation and heat shock protein 90 (Hsp90) phosphorylation. [163,167,168] In this paper, we have not examined these phenomena, but it could be that the disruption of one of these processes contributes to the observed phenotype. It may be interesting to study these additional effects in the future.

The timing of combination therapy is important to obtain optimal treatment efficacy. It was reported that CDC2 is transiently phosphorylated to induce an arrest at the G2/M checkpoint for 12 h after irradiation treatment and that DNA damage could be repaired in 12-24 h after irradiation. [121] Our results support this; in irradiated cells, we observed only few remaining foci of DNA damage after 24h, whereas cells treated with irradiation and WEE1-inhibitor had many residual foci after 24h, indicating that these cells were unable to perform DNA repair. This suggests that DNA damaged cells are especially susceptible to WEE1-inhibitor in the first 12h after induction of DNA damage. In our experimental set-up, treatment of OS cells with WEE1-inhibitor directly after irradiation showed a good sensitisation. This suggests that cells do not have to be arrested in G2/M to be susceptible to WEE1 inhibition, but rather that the inability to activate (or maintain) the G2/M checkpoint in the presence of DNA damage leads to the sensitisation. In in vivo testing of WEE1 inhibitors, different approaches have been applied. Mir et al. [158] administered WEE1-inhibitor at 5 consecutive days around the irradiation dose, whereas Hirai et al. [125] first administered DNA damaging agents, followed by WEE1-inhibitor after a 24 hour interval. Both groups showed enhanced anti-tumour efficacy. What will be the most optimal schedule for radiotherapy combined with WEE1 inhibition in OS remains to be tested in vivo.

66

Chapter 4

C O N C L U S I O NRadiotherapy is a controversial topic in the treatment of OS. Its efficacy is limited in this cancer and therefore it is not widely applied. Novel small molecules, in particular WEE1-inhibitor drugs may serve as radiosensitisers in OS. WEE1 kinase is expressed in OS and plays a critical role in DNA repair by maintaining the G2 cell cycle arrest through inhibitory phosphorylation of CDC2. Our results show that WEE1-inhibitor PD0166285 can abrogate the DNA damage induced G2/M cell cycle arrest in OS cells, forcing the cells into mitotic catastrophe and thus causing radiosensitisation. WEE1 could therefore be a strategic, cancer cell specific drug target and its inhibition could be an effective strategy to enhance the efficacy of radiotherapy in OS.

67


4

Chapter 5 Targeting JNK-interacting protein 1 (JIP1)

sensitises osteosarcoma to doxorubicin

J. PosthumaDeBoer, P.W. van Egmond, M.N. Helder, R.X. de Menezes, A.M. Cleton-Jansen, J.A.M. Beliën, H. M. W. Verheul, B.J. van Royen,

G.J.L. Kaspers, V.W. van Beusechem

Oncotarget. 2012 Oct; 3(10): 1169-81

70

Chapter 5

A B S T R A C TOsteosarcoma (OS) is the most common primary malignant bone tumour in children and adolescents. Despite aggressive therapy, survival outcomes remain unsatisfactory, especially for patients with metastatic disease or patients with a poor chemotherapy response. Chemoresistance contributes to treatment failure. To increase the efficacy of conventional chemotherapy, essential survival pathways should be targeted concomitantly. Here, we performed a loss-of-function siRNA screen of the human kinome in SaOS-2 cells to identify critical survival kinase or kinase-associated genes after doxorubicin treatment. Gene silencing of JNK-interacting-protein-1 (JIP1) elicited the most potent sensitisation to doxorubicin. This candidate was further explored as potential target for chemosensitisation in OS. A panel of OS cell lines and human primary osteoblasts was examined for sensitisation to doxorubicin using small molecule JIP1-inhibitor BI-78D3. JIP1 expression and JIP1-inhibitor effects on JNK-signalling were investigated by Western blot analysis. JIP1 expression in human OS tumours was assessed by immunohistochemistry on tissue micro arrays. BI-78D3 blocked JNK-signalling and sensitised three out of four tested OS cell lines, but not healthy osteoblasts, to treatment with doxorubicin. Combination treatment increased the induction of apoptosis. JIP1 was found to be expressed in two-thirds of human primary OS tissue samples. Patients with JIP1 positive tumours showed a trend to inferior overall survival.

Collectively, JIP1 appears a clinically relevant novel target in OS to enhance the efficacy of doxorubicin treatment by means of RNA interference or pharmacological inhibition.

71

Targeting JNK-interacting protein 1 (JIP1) sensitises osteosarcoma to doxorubicin

5

I N T R O D U C T I O NOsteosarcoma (OS) is the most common primary malignant bone tumour in children and adolescents. The gold standard for therapy consists of a combination of multi-agent neoadjuvant chemotherapy, followed by radical surgery and adjuvant chemotherapy. With this aggressive regimen, 5-year survival rates of approximately 65% are obtained in patients with localised disease. However, in the case of metastatic or recurrent disease, 5-year survival rates are reduced to only 20%. [1,6,19,20] Chemoresistance, both intrinsic and acquired, is a key issue in the failure of current treatment to cure patients with OS. [90,124] A variety of combination therapy regimens and dose escalation of several therapeutics have not improved survival outcomes. Also, the current chemotherapy regimens are demanding for the patients and serious adverse effects, such as severe mucositis, bone-marrow depression and cardiotoxicity, are regularly encountered. [5,15-17] In order to improve treatment efficacy whilst limiting adverse effects, new treatment strategies for OS are warranted.

Targeting essential survival pathways in combination with conventional therapy could be a strategy to improve the efficacy of current drug regimens. The identification of key regulators of drug response is an essential step in the design of such new targeted treatment strategies. RNA interference (RNAi)-based screening is a powerful technology for the discovery of candidate drug targets in malignant cells. [105,169,170] Proteins of the human kinome are involved in many cellular processes, including inter- and intra-cellular signalling, gene transcription, metabolism, cell shape and motility, proliferation, differentiation, survival and apoptosis. Kinases are known to play essential roles in disease development [171,172] and the kinome is therefore likely to harbour potential drug targets. Furthermore, kinases have been the subject of the design and development of small molecule drugs that target specific pathways in malignant cells. [43,58,126,173,174] This has confirmed the feasibility of targeting kinases with specific small molecules and their utility as targets for therapy.

Here, we performed an siRNA screen of the human kinome to systematically identify genes that are involved in survival of OS cells treated with doxorubicin. One candidate, MAPK8IP1, which encodes mitogen-activated protein kinase 8 interacting protein 1, also known as JNK-interacting protein 1 (JIP1), was further analysed for its potential use as a target for sensitisation of OS to doxorubicin.

M AT E R I A L S A N D M E T H O D SCell culture and compoundsHuman osteosarcoma cell lines SaOS-2, MG-63, U2OS and LM7 [175] were kindly provided by Dr. F. van Valen (Westfalische Wilhelms-Universität, Münster, Germany), Dr. C. Löwik (Leiden University Medical Center, Leiden, the Netherlands), Dr. S. Lens (Netherlands Cancer Institute, Amsterdam, the Netherlands) and Prof. Dr. E.S. Kleinerman (MD Anderson Cancer

72

Chapter 5

Center, Houston, TX, USA), respectively. Human primary (short-term culture) osteoblasts Hum63, Hum65 and Hum71 were obtained from otherwise healthy patients undergoing total knee replacement after informed consent. All cells, with the exception of LM7, were cultured in D-MEM (PAA Laboratories) supplemented with 10% fetal calf serum (FCS) and 1 mg/mL Penicillin-Streptomycin (Gibco) at 37°C and 5% CO2 in a humidified incubator. LM7 was cultured in E-MEM (Lonza) supplemented with 10% FCS, 1 mg/mL Pen-Strep, 1% non-essential amino acids, 1% sodium pyruvate, 2 nM L-glutamine and 2% MEM-Vitamin solution (all: Gibco, Invitrogen) at 37°C and 5% CO2 in a humidified incubator. Doxorubicin (TEVA/Pharmachemie, Haarlem, the Netherlands) was diluted in D-MEM to the desired concentrations prior to use. The small molecule JIP1-inhibitor BI-78D3 (Sigma) [176] was dissolved in DMSO and diluted to the appropriate concentration in PBS directly prior to use.

siRNA library screeningThe siRNA screen was performed at the RNA Interference Functional Oncogenomics Laboratory (RIFOL) core facility of the VUmc Cancer Center Amsterdam, using an automated platform and the Human Protein Kinases ON-TARGETplus siRNA library from Thermo Fisher Scientific Dharmacon (Lafayette, CO). This library comprises siRNAs targeting 788 kinase and kinase-associated genes. Four different siRNA duplexes designed against each target gene were pooled and arrayed for screening in 96-well plate format. siRNA against PLK1 was used as a positive control for cell death and mock treated wells served as negative controls. For confirmation studies, individual siGENOME siRNAs (TFS Dharmacon) were used. Forward transfections were done according to the manufacturer’s recommendations, using 25 nM siRNA and 0.05% (v/v) Dharmafect 1 transfection reagent (TFS Dharmacon) in 100 μL culture medium. Figure 5.1A shows a schematic overview of the screen design. The screens were performed three times, each time including one set of plates with doxorubicin and one set without. SaOS-2 cells were plated at a density of 2,000 cells/well in 96-well plates (Greiner) and transfected with siRNAs the next day. Two days post-transfection, doxorubicin was added to one set of plates to a final concentration of 30 ng/mL; culture medium was added to the other set. Four days after the start of doxorubicin treatment, cell viability was determined using the CellTiter-Blue (CTB) Cell Viability Assay (Promega), measuring fluorescence at 540 nm excitation and 590 nm emission wavelengths using a Tecan Infinite F200 Microplate Reader (Tecan Trading AG, Switzerland). The CTB Cell Viability Assay measures metabolic capacity of viable cells by reduction of the indicator dye resazurin into fluorescent resofurin through the action of cellular enzymes, in which the measured fluorescent signal is proportional to the number of viable cells (www.promega.com). Plate data was read and configured in R (The R Project for Statistical Computing) [177] using the cellHTS2 software package. [178] We used the negative control-medians per

73


5

plate to centre plate-specific log2 intensities and then computed robust z-scores per screen. The z-score matrix containing 3 untreated and 3 treated screens was used to study the combination treatment effect (i.e., combined effect of doxorubicin and siRNA as discriminated from single agent effects) by means of an empirical-Bayes linear model, using the limma software package. [179] The obtained p-values for the combination treatment effect were then corrected for multiple testing using Benjamini & Hochberg’s step-up false discovery rate (FDR). [180] The magnitude of sensitisation to doxorubicin by siRNA transfection was estimated by calculating the ratio of mean cytotoxicity observed after combination treatment over mean cytotoxicity observed after doxorubicin treatment.

Doxorubicin dose response analysis and apoptosis assayHuman OS cells or primary osteoblasts were plated at a density of 2,000 cells/well in 96-well format and transfected with JIP1 siRNA as described above or incubated with BI-78D3 at a non-toxic dose of 10 nM. Two days after siRNA transfection or concurrently with BI-78D3 addition, cells were treated with a doxorubicin dose range and four days later cell viability was assessed using the CTB assay as described above. To analyse apoptosis, OS cells were plated in white opaque 96-wells plates and treated with doxorubicin (0.1 μg/mL) or combination treatment with doxorubicin plus BI-78D3 (10nM). At 24 h post-treatment, caspase activity was measured using the Caspase-Glo 3/7 assay (Promega) according to the manufacturer’s instructions.

Luminescence read-out was performed using a Tecan Infinite F200 Microplate Reader (Tecan Trading AG, Switzerland). Results were analysed using GraphPad Prism® Version 5.01 (GraphPad Software Inc.). Relative caspase activity was normalised to the signal measured in the control (PBS) condition.

Quantitative RT-PCRCells were plated, allowed to adhere and transfected with siRNA as described above. Two days post-transfection the cells were harvested by trypsinisation. Cellular RNA was isolated using the RNeasy Kit (Qiagen) and quantified by spectrophotometry using the NanoDrop ND-1000 and ND-1000 software version 3.3 (Isogen Lifescience). Per sample, 1 μg total RNA was reverse transcribed into cDNA using the SuperScript® III RT kit plus Random Primers (Invitrogen). Real time quantitative PCR for MAPK8IP1 / JIP1 was performed using the Quantitect primer assay designed for detection with SYBR-Green (Qiagen). Amplification was measured on a LightCycler® 480 and analysed using the corresponding software, Version 1.5 (Roche). Relative MAPK8IP1 / JIP1 gene expression was normalised to that of GAPDH using the ΔΔCt method. [181]

74

Chapter 5

Immunoprecipitation and Western blot analysisBaseline JIP1 protein expression levels across OS cell lines and primary osteoblasts and JIP1 downregulation at the protein level 1-5 days after transfection of SaOS-2 cells was assessed using Western blot. Cells were lysed in buffer containing Protease and Phosphatase Inhibitor Cocktails (Sigma). Proteins were quantified with the BCA protein Assay Kit (Pierce). A total of 30 μg protein per sample was separated on a SDS-PAGE gel and transferred to a PVDF membrane. Blots were incubated with primary antibodies rabbit-anti-MAPK8IP1 (181-193) at a dilution of 1:1,000 (Sigma-Aldrich) and mouse-anti-β-actin (Abcam) at a dilution of 1:10,000, followed by secondary antibody incubation with HRP-conjugated goat-anti-rabbit and goat-anti-mouse immunoglobulins (DAKO), respectively. Protein detection and visualisation was performed using ECL+ Western Blotting Detection Reagents (Pierce).

Regulation of JIP1 and JNK phosphorylation following doxorubicin and JIP1-inhibitor BI-78D3 treatment was analysed by immunoprecipitation followed by Western blot analysis. OS cells and human primary osteoblasts were plated and treated with doxorubicin (0.3 μg/mL) or combination treatment with doxorubicin plus BI-78D3 (10 nM) for 4 h and then lysed in phospho-lysis buffer (HEPES containing 0.5% β-glucose, 0.1% DTT, 0.1% Na3VO4) containing Protease and Phosphatase Inhibitor Cocktails (Sigma). For immunoprecipitation, 25 μL of IgG-sepharose beads (Pierce) were incubated with 2 μL rabbit-anti-MAPK8IP1 antibody for 1 h at 4°C under continuous motion. The antibody-bead conjugates were collected by centrifugation, mixed with 60 μg protein lysate for 1 h at 4°C under continuous motion. Immunoprecipitates were collected by centrifugation. Sample separation and transfer followed as described above. Primary antibody incubation was done using rabbit-anti-phospho-JNK (Cell Signalling) and secondary antibody incubation and protein detection were performed as described above. Protein levels were quantified using the Image J tool (National Institute of Health, USA). Intensities were normalised to β-actin levels in the Western blots and to the internal control sample in the immunoprecipitation.

Tissue micro arrays and immunohistochemistryTwo tissue micro arrays (TMAs) containing a total of 647 cores of human primary OS samples (corresponding to 130 OS patients) and 20 control tissue cores, were stained and analysed for JIP1 expression (Appendix A: Supplementary Figure 5.2A) and then correlated to clinical and survival data. The TMAs were crafted at the Leiden University Medical Center (LUMC, Leiden, the Netherlands) according to the protocol described in Mohseny et al. [182] All patients were treated for OS at the LUMC in the period between 1984-2009. Available clinical data includes: age, gender, location and side of the primary tumour, response to chemotherapy according to the Huvos grading system [183] (when available), metastasis, recurrence, date of recurrence, survival, date of death (when applicable) and time of follow-up. (See Supplementary Table 5.4) The tissue array slides were heated at 60°C for 20 minutes

https://www.dropbox.com/home/Supplementary%20tables%20to%20%27Towards%20targeted%20therapy%20for%20osteosarcoma%27

75


5

prior to deparaffinization in Xylene and rehydration in a graded series of alcohol. Endogenous peroxidase activity was inhibited by incubation with 0.3% H2O2 diluted in methanol for 30 minutes. The arrays were boiled in 10 mM citrate buffer (pH 6) for 10 minutes and subsequently rinsed in PBS. The slides were incubated with rabbit-anti-MAPK8IP1 (181-193) primary antibody at a dilution of 1:500 O/N at 4°C. Antigen visualisation was performed using the EnVision+ Poly-HRP IHC Kit (Immunologic) and DAB chromogen solution. Slides were counterstained with hematoxylin, dehydrated and mounted.

Tissue micro array scoringThe stained TMA slides were automatically scanned using a digital whole slide scanning system (Mirax slide Scanner system 3DHISTECH Ltd., Budapest, Hungary), equipped with a numerical aperture of 0.75 and a Sony DFW-X710 Fire Wire 1/3” type progressive SCAN IT CCD (pixel size 4.65 x 4.65 μm), with an actual scan resolution (effective pixel size in the sample plane) at 20x objective of 0.23 μm. All 647 samples were independently examined and scored by two of the authors (JP and PE). The scoring was performed using dedicated TMA scoring software (3DHISTECH Ltd., Budapest, Hungary) in a blinded fashion. To facilitate the scoring and improve the reproducibility of scoring, a consensus chart with exemplary staining patterns per category was created (Appendix A: Supplementary Figure 5.2B) and used by the observers during the scoring of the samples. The staining per tissue was assessed and valued as “negative” or “positive”. Due to loss of cores during the cutting and staining procedure, not all cores could be included for analysis. Samples were considered unsuitable for scoring when less than 30% of tissue was present on the digital copy of the tissue core. In case of insufficient tissue, the cores were given the value “no data”. Of each tumour, three cores are present on the TMA. The grading scale consistend of 3 values (0 = no data; 1 = negative; 2 = positive) To assure a robust staining score, i.e. reliable scoring per tumour sample, we applied a threshold of a minimum of 6 scores 1 or 2 (excluding “no data” observations) for tumours to be included in the statistical analysis. We used the mean of the scores to assign the final staining result (positive or negative) to a sample. The clinical data and the staining results were entered and statistically analysed in SPSS, version 17.0 (SPSS Software, Inc., Chicago, IL, USA). To assess inter- and intra-observer agreement in grading JIP1 staining, Kappa statistics were used. Because inter- and intra-observer reproducibility may be biased by an overemphasis on patients with grade 0 findings, kappa values were therefore also calculated with the exclusion of grade 0 findings (censored Kappa). Values between 0 and 1 were interpreted according to modified published guidelines. [184,185] (Supplementary Table 5.3) Kaplan-Meier analysis was used to assess survival and differential survival between groups was analysed using the Log Rank test. To determine significant differences between categorical groups, the Pearson chi-square test was used. In numerical groups, the independent t-test and one-way ANOVA were used. The threshold for statistical significance was set at p<0.05.


76

Chapter 5

Figure 5.1 siRNA library screen of the human kinome identifies enhancers of doxorubicin response in OS. (A) Schematic overview of the set-up of the screens performed with pools of 4 siRNAs against 788 human kinases and kinase-associated genes in SaOS-2 cells. (B) Screen results of the 10 selected candidate hits showing the effects of gene-silencing only (grey bars) versus gene-silencing + doxorubicin treatment (black bars) on cell viability. Bars represent the average cell viability measured in the 3 screens; error bars indicate standard deviations (SD). (C) Pie chart summarising the secondary screen results. The chart shows the number of separate siRNA duplexes of 4 individual siRNAs tested per gene, that reproduced the doxorubicin-sensitising phenotype of the pooled siRNAs. (Bar graphs for the separate siRNA duplexes are provided in panel D for JIP1 (see Appendix A: Supplementary Figure 5.1 for the other genes). (D) Confirmation of the doxorubicin-sensitising phenotype with siRNAs against JIP1. Cells were transfected with the indicated siRNA duplex and cultured in the presence (black bars) or absence (grey bars) of doxorubicin at IC20 concentration. Bars represent results from an experiment performed in triplicate; error bars indicate SD.

R E S U LT SsiRNA screening identifies regulators of doxorubicin response in OS cellsIn order to identify regulators of doxorubicin response in OS, we performed cell viability screens on SaOS-2 cells using an siRNA library targeting the human kinome (Figure 5.1A). Screens were performed three times, each time including pairs of plates with and without doxorubicin treatment at an approximate IC20 concentration. Assay metrics based on mock- versus siPLK1-treated wells revealed Z’-factors ranging from 0.69 to 0.82 in the three experiments, indicating strong assay resolution. [186] Supplementary Table 5.1 lists robust


77


5

z-scores of all tested genes per screen. The effects of doxorubicin plus siRNA treatment were analysed using an empirical-Bayes linear model. Supplementary Table 5.2 lists the computed treatment effects for all tested genes. Table 5.1 lists the genes that showed a most significant combination treatment effect (threshold p<0.025). As indicator of the sensitising potential of gene silencing to doxorubicin treatment we calculated relative cytotoxicities (i.e., doxorubicin plus siRNA effect/doxorubicin effect). We then selected 10 candidate genes that met the following criteria: p<0.025 and FDR<0.4 and/or p<0.025 and relative cytotoxicity >3-fold.

The mean relative cell viabilities of SaOS-2 cells treated with the 10 selected siRNAs in the presence or absence of doxorubicin are shown in Figure 5.1B. siRNA against JIP1 appeared to elicit the most potent and highly significant enhancement of doxorubicin-induced cell kill (relative cytotoxicity = 8.6; p = 1.0*E-04; FDR = 2%).To confirm the findings in the primary screen for these 10 candidates, they were reanalysed using 4 siRNAs directed against different sequences on their mRNA. For 8 candidate genes, the doxorubicin-sensitising phenotype

Accession number

Gene symbol Gene name p-value t-statistic FDR relativecytotoxicity

NM_000076 CDKN1C cyclin-dependent kinase inhibitor 1C (p57, Kip2)

6.3E-05 -4.00 0.02 5.4

NM_005456 JIP1 (MAPK8IP1) mitogen-activated protein kinase 8 interacting protein 1

1.0E-04 -3.89 0.02 8.6

NM_212469 CHKA choline kinase alpha 1.1E-03 -3.26 0.12 5.7NM_001319 CSNK1G2 casein kinase 1, gamma 2 3.7E-03 -2.90 0.22 2.7NM_001570 IRAK2 interleukin-1 receptor-associated kinase-like 2 3.8E-03 -2.89 0.22 8.0

NM_001381 DOK1 docking protein 1, 62kDa (downstream of tyrosine kinase 1)

5.3E-03 -2.79 0.28 4.4

NM_001291 CLK2 CDC-like kinase 2 7.1E-03 -2.69 0.35 1.9NM_020421 ADCK1 aarF domain containing kinase 1 9.1E-03 -2.61 0.41 1.3NM_031284 ADP-GK ADP-dependent glucokinase 0.011 -2.55 0.44 2.0

NM_004196 CDKL1 cyclin-dependent kinase-like 1 (CDC2-related kinase)

0.011 -2.54 0.44 3.6

NM_000586 IL2 interleukin 2 0.013 -2.48 0.49 5.2

NM_014826 CDC42BPA CDC42 binding protein kinase alpha (DMPK-like)

0.016 -2.42 0.49 11.1a

NM_052947 HAK (ALPK2) alpha-kinase 2 0.016 -2.42 0.49 1.1NM_198452 PNCK pregnancy up-regulated non-ubiquitously

expressed CaM kinase0.020 -2.33 0.54 3.0

NM_001025778 VRK3 vaccinia related kinase 3 0.021 -2.31 0.55 1.9NM_001001329 PRKCSH protein kinase C substrate 80K-H 0.024 -2.26 0.61 7.0

aexcluded from further analysis due to high direct cytotoxicity upon siRNA treatment alone

Table 5.1 Primary screen hit list The top 16 genes that, upon silencing elicit a statistically significant (p < 0.025) increase in sensitivity to doxorubicin treatment in osteosarcoma cells with corresponding p-values and t-statistics for treatment effect, and false discovery rates for each gene. Relative cytotoxicity is defined as doxorubicin plus siRNA effect/doxorubicin effect.


78

Chapter 5

could be reproduced with at least 3 individual siRNAs, suggesting that they represent genuine therapeutic targets (Figure 5.1C). Figure 5.1D shows the reanalysis results for JIP1. Three siRNAs (i.e., duplexes #2, #3, and #4) clearly enhanced cell kill after doxorubicin treatment, confirming the phenotype that was observed with the siRNA pool. In fact, these siRNAs exhibited a more selective effect, as they caused less direct cytotoxicity than the pool in the absence of doxorubicin. This could be explained by a profound cytotoxicity induced upon transfection of siRNA duplex #1. For this reason, siRNA duplex #1 was considered to elicit an off-target effect and was excluded from further analyses. Supplementary Figure 5.1 (Appendix A) shows the reanalysis results of the remaining 9 candidate genes. For 7 candidate genes, i.e. CDKN1C, JIP1, CHKA, CSNK1G2, IRAK2, DOK1, CLK2 and IL2, the doxorubicin-sensitising phenotype could be reproduced with at least 3 individual siRNAs. Two genes could not be confirmed; CDKL1 had only 2 effective duplexes and PRKCSH was excluded from further analysis because three of the tested duplexes induced an increase in cell viability, yielding only 1 effective duplex for this gene.

dupl

ex 2

dupl

ex 3

dupl

ex 4

0

20

40

60

80

100

120

Rela

tive

JIP1

mRN

A ex

pres

sion

(%)

0.1 1 100

20

40

60

80

100

120

CtrlsiJIP1 seq 2

Log [doxorubicin]

Rel

ative

Via

bility

(%)

βActin

JIP1

Ctrl 48 h 72 h 120 h

Figure 5.2 siRNA targeting JIP1 reduces JIP1 mRNA and protein expression in SaOS2 cells and causes increased sensitivity to doxorubicin. (A) JIP1 mRNA silencing after transfection of JIP1 siRNA duplexes, analysed by RT-qPCR. JIP1 mRNA expression levels were normalised to GAPDH mRNA levels for each sample. Bars represent experiments performed in triplicate, error bars indicate SD. Results could not be obtained for JIP1 siRNA duplex #1 due to major cytotoxicity upon transfection of this siRNA. (B) Western blot analysis of JIP1 protein depletion at the indicated time-points after transfection of JIP1 siRNA duplex #2. (C) Dose-response curve of cells treated with doxorubicin (closed squares) and cells treated with doxorubicin after JIP1 gene silencing using JIP1 siRNA duplex #2 (open triangles). Results were obtained in an experiment performed in triplicate; error bars indicate SD. The JIP1 silenced cells show a significantly increased sensitivity to doxorubicin treatment (student’s t-test at IC50; p<0.0001).

79


5

Depleting OS cells from JIP1 protein increases doxorubicin-induced cell death. Based on the siRNA screens, we selected JIP1 for further investigations. First, we assessed the gene-silencing efficiency obtained with the three functional JIP1 siRNA duplexes using qRT-PCR. Figure 5.2A shows that JIP1 siRNA duplex #2 was the most effective in suppressing JIP1 mRNA. Silencing JIP1 also depleted SaOS-2 cells of the protein product, as shown by Western blot analysis (Figure 5.2B). Decreased JIP1 protein expression became evident 2

Figure 5.3 JIP1 inhibition sensitises OS cells but not normal osteoblasts to doxorubicin-induced apoptosis. (A) Cells were treated in triplicate with doxorubicin (closed squares) or with doxorubicin and JIP1-inhibitor BI-78D3 (open triangles) and cell viability was determined after four days. Sigmoidal dose-response curves were created and IC50 values were calculated. The sensitising effect of JIP1 inhibition to doxorubicin treatment is significant in three out of four OS cell lines, (student’s t-test at IC50; U2OS, p=0.19, n.s; SaOS-2, p<0.05; LM7 and MG-63, p<0.0001). Human primary osteoblasts Hum63 and Hum65 do not show sensitisation to doxorubicin treatment in the presence of the JIP1-inhibitor (student’s t-test at IC50; p=0.459 and p=0.417 respectively). (B) Caspase activation in OS cells treated with doxorubicin (grey bars) or with doxorubicin and JIP1-inhibitor BI-78D3 (black bars). White bars represent the control condition, which was set to 1. Bars represent an experiment performed in triplicate, error bars indicate SD. After combination treatment, caspase activity is distinctly higher in SaOS2, LM7 and MG-63 compared to doxorubicin treatment alone (student’s t-test; LM7: **, p<0.01; MG-63: *, p<0.05; SaOS2, p=0.08). In U2OS cells, caspase activity is comparable either with or without JIP1 inhibition (p=0.48).

80

Chapter 5

days after transfection and was most pronounced at day 3. Doxorubicin dose-response curves for SaOS-2 cells and SaOS-2 cells transfected with JIP1 siRNA #2 (Figure 5.2C) demonstrated a significant shift in IC50, from 0.7 μg/mL for the control treated cells to 0.3 μg/mL for the JIP1 silenced cells (student’s t-test, p<0.0001). Hence, efficient siRNA-mediated silencing of JIP1 in SaOS-2 cells reduced JIP1 protein expression, sensitising the cells to doxorubicin treatment.

Selective sensitisation of OS cells to doxorubicin-induced apoptosis using a small molecule JIP1 inhibitor drug.Next, we investigated the effect of JIP1 inhibition using a small molecule drug on sensitivity to doxorubicin in OS cells and primary (non-malignant) human osteoblasts. To this end, cells were subjected to a dose range of doxorubicin concentrations in the presence or absence of a non-toxic dose (10 nM) of the small molecule JIP1-inhibitor BI-78D3 [176] that binds competitively at the JNK-binding site of JIP1. BI-78D3 increased sensitivity to doxorubicin treatment in 3 out of 4 tested OS cell lines, but not in primary osteoblasts (Figure 5.3A). IC50s were significantly decreased in SaOS-2 (p<0.05), LM7 (p<0.0001) and MG-63 (p<0.0001) cells. To investigate if the combined effect of doxorubicin treatment and JIP1 inhibition was associated with stimulation of apoptotic cell death, we measured caspase-3 and caspase-7 activity in OS cells treated with doxorubicin in the presence or absence of 10nM BI-78D3. In all four OS cell lines, doxorubicin increased caspase activity (by 1.3 to 2.8-fold). In SaOS2, LM7 and MG-63 cells, caspase activity was further increased (to 2.5 to 3.8-fold) by addition

1.0 0.11.2 1.0 0.10.8

βActin

JIP1

Hum

71

U2O

S

MG

63

LM7

SaO

S2 Doxorubicin

BI78D3+

+ ++ + +

U2OSLM7

βActin

JIP1

pJNKJIP1 IP1.2 0.5 0.20.6 0.8

0.4 0.4 0.4 0.40.4 0.5

Figure 5.4 JIP1-inhibitor BI-78D3 reduces JIP1/p-JNK complexes in doxorubicin treated OS cells. (A) Western blot analysis of baseline JIP1 levels in human OS cell lines and human primary osteoblast culture Hum71. All OS cell lines exhibit higher JIP1 levels than the human primary osteoblasts. (B) Immunoprecipitation and Western blot analysis of the JIP1/JNK activation module. LM7 and U2OS cells were treated with doxorubicin either in the presence or absence of BI-78D3. Whole cell lysates were analysed for JIP1 expression with β-actin serving as control for equal loading. In addition, cell lysates were immunoprecipitated with an anti-JIP1 antibody and Western blot analysis was done with an anti-p-JNK antibody. Doxorubicin-treated LM7 cells show an increased expression level of p-JNK compared to untreated cells, whereas in U2OS there is a slight decrease in p-JNK after doxorubicin treatment. After concurrent treatment with BI-78D3, p-JNK expression levels are diminished, indicative of inhibition of JIP1 scaffold function.

81


5

of the JIP1-inhibitor (Figure 5.3B). This was significant in LM7 (p<0.01) and MG-63 (p=0.01) cells and approached significance in SaOS2 cells (p=0.08). Contrarily, BI-78D3 did not augment caspase activation in U2OS cells (p=0.48). Thus, the results of the caspase activity assay were in agreement with the observations made in the dose-response experiments. We therefore conclude that sensitisation of OS cells to doxorubicin treatment by JIP1 inhibition is probably caused by promotion of doxorubicin-induced apoptosis. Collectively, these data independently validate JIP1 as a regulator of doxorubicin response in OS and suggest that pharmacological inhibition of JIP1 could provide increased efficacy of doxorubicin treatment in OS, while sparing healthy bone cells.

BI-78D3 inhibits JIP1-JNK interaction and JNK-phosphorylation in OS cells. JIP1 is a scaffold protein that selectively mediates JNK signalling by assembling specific components of the MAPK cascade, including MLK, MKK4 and MKK7, in a signalling complex and JIP1 appears essential for JNK activation or maintenance of JNK phosphorylation. [176,187-192] In response to environmental stress, JNK is activated by phosphorylation of residues in its activation loop by MKK4 and MKK7, which in turn are activated by MLK. Apart from this, JIP1 has also been implicated in Akt1 activation and suppression of Notch1 activity. [193,194] Western blot analysis demonstrated that JIP1 is expressed in all tested OS cell lines and at a lower level in primary osteoblasts (Figure 5.4A). We then investigated JIP1 expression and JNK phosphorylation in response to doxorubicin and BI-78D3 treatment in LM7 and U2OS cells. These cell lines respectively showed a strong and absent sensitisation to doxorubicin efficacy by treatment with JIP1-inhibitor BI-78D3. Treatment with doxorubicin did not affect JIP1 expression (Figure 5.4B). p-JNK could not be detected in whole cell lysates (not shown). However, upon immunoprecipitation with an anti-JIP1 antibody, p-JNK could be detected in both cell lines (Figure 5.4B). Doxorubicin treatment increased p-JNK in complex with JIP1 in LM7 cells, but not in U2OS cells. Concurrent treatment with BI-78D3 diminished p-JNK in both cell lines to undetectable levels, indicating successful inhibition of the JIP1-JNK interaction and JNK phosphorylation (Figure 5.4B). Together, these results suggest that inhibition of JIP1 increases the cytotoxicity of doxorubicin in cells that respond to doxorubicin treatment by increased JNK-signalling via assembly of the JIP1-JNK signalling complex.

JIP1 is expressed in a majority of OS tumour specimens and is associated with poor survivalTo assess the clinical relevance of JIP1 and to investigate if JIP1 could be considered a biomarker for OS, tissue micro arrays (TMAs) containing 647 cores of human primary OS samples (corresponding to 130 OS patients) were stained and analysed for JIP1 expression. Every tumour was represented by three cores on the TMAs and certain patients were represented more than once on the TMA i.e., with cores belonging to primary biopsies,

82

Chapter 5

first resections and metastases. Staining results were then correlated to clinical data. All samples were independently scored by two of the authors, with high observer agreement (Supplementary Table 5.3).

Due to loss of tissue from the slide as a result of the cutting and staining procedures, 200 cores were unsuited for scoring, leaving 447 scored tissue cores. Of these, 67% were valued “positive” and 33% were valued “negative”. To study the predictive value of JIP1 staining on overall survival, we selected first biopsy or resection samples only; first biopsies had not been exposed to pre-operative chemotherapy and resection samples had been exposed to pre-operative chemotherapy. Samples of recurrences (both metastatic and local) were excluded to avoid confounding for inferior survival outcome as a result of metastatic and/or recurrent disease. After applying these criteria, data of 71 patients remained suitable for analysis. Figure 5.5 shows a Kaplan-Meier univariate analysis of JIP1 staining in tumour tissue as predictor of overall survival.

Patients with JIP1 negative tumours showed a better survival outcome (mean 12.0 years) compared to patients with JIP1 positive tumours (mean 9.1 years); this difference in overall survival is borderline significant (LogRank, p=0.056). JIP1 staining did not significantly correlate with event-free survival (p=0.3) or with relapse (p=0.2). Moreover, we did not find a significant association between JIP1 staining and response to multi-agent chemotherapy (p=0.9) (Supplementary Table 5.4). Thus, while JIP1 staining did not directly correlate to the response to chemotherapy, JIP1 positivity did show a strong trend towards inferior overall

Figure 5.5 Kaplan-Meier overall survival analysis of OS patients with JIP1 positive or negative tumours. Kaplan-Meier survival plot showing the cumulative survival of patients suffering from localised OS. Patients were divided into groups of patients with JIP1 positive tumours (55 samples) and those with JIP1 negative tumours (16 samples). The observed difference in cumulative survival between the two groups shows a trend towards inferior survival outcomes for patients with positive JIP1 staining (p=0.056).



83


5

survival outcome, suggesting a possible role for outcome prediction in patients with OS. Importantly, JIP1 was found to be expressed in a majority of primary OS tumour samples, suggesting that JIP1 could be considered a clinically relevant target for treatment.

D I S C U S S I O NIn this work, we applied a functional genomics approach to systematically identify kinases and kinase-associated proteins that could be targets to enhance response to doxorubicin treatment in osteosarcoma. Doxorubicin is one of the major components of OS chemotherapy treatment, but its efficacy knows limitations in view of the 65% 5-year survival rate for patients with localised OS. Improving its efficacy could lead to enhanced tumour control or chemotherapy dose reduction in patients. The latter is relevant as doxorubicin is renowned for the serious adverse effects that regularly arise upon administration of this drug, especially cardiotoxicity. [15] High throughput RNAi screens have previously been used successfully to identify essential oncogenes, modulators of response to anti-cancer drugs and cancer-specific targets for therapy. [105,169,170,195-197] Previous work in OS has led to the identification of potential therapeutic targets in OS by RNAi kinome library screening. [198,199]

In our siRNA library screen on SaOS-2 human OS cells, silencing of JIP1 led to the most potent and highly significant sensitisation to doxorubicin. This finding could be validated on SaOS-2 cells and two other p53-deficient OS cell lines using small molecule JIP1-inhibitor BI-78D3. OS cells expressed JIP1 at higher levels than normal osteoblasts and the sensitising effect of JIP1 inhibition to doxorubicin treatment was not observed in normal osteoblasts. This suggests that JIP1 is important for the survival of malignant, but not healthy cells subjected to chemotherapy, which makes JIP1 a potential target for selective anticancer therapy.

JIP1 is a scaffold protein that, in response to cellular stress, assembles a JNK activation module containing various kinases upstream of JNK, such as MKK4, MKK7 and MLK family members. The physical proximity between the JNK-signalling components in the JIP1/JNK signalling complex facilitates the phosphorylation of JNK and subsequent activation of the JNK signalling pathways. JIP1 inhibition is expected to alter or impair JNK-signalling, possibly leading to reduced c-Jun phosphorylation and downregulation of AP-1 transcriptional activation. [176,187-191] JNK pathway activation is observed following cellular stress such as DNA damage, cytotoxic stress and γ-irradiation. However, the exact influence of JNK and AP-1 on cell death is ambiguous and reported to be cell and context dependent. [41,124,188,191,192,200] JNK has been found to promote as well as inhibit apoptosis, possibly depending on different upstream stimuli. In addition, JNK pathway activation has been suggested to promote p53-induced apoptosis. Hence, JNK might only exert its anti-

84

Chapter 5

apoptotic effects in p53-deficient tumour cells. [192] Our observations are in line with the described regulation of the JIP1/JNK signalling module. Treatment with doxorubicin induced JNK phosphorylation in p53 deficient LM7 cells, but not in p53 wild type U2OS cells, and inhibition of the JIP1-JNK interaction completely abrogated JNK phosphorylation. In addition, p53-null SaOS-2 and LM7 cells and p53-mutated MG-63 cells, but not p53 wild type U2OS cells [120,145,175] and normal osteoblasts, were sensitised to doxorubicin cytotoxicity upon JIP1 inhibition. Furthermore, increased cytotoxicity upon combination treatment was associated with promotion of apoptosis induction. This leads us to hypothesise that p53 deficient OS cells are sensitised to doxorubicin by inhibition of anti-apoptotic JNK signalling. However, given the relatively small number of cell lines analysed and the redundancy in survival pathways in tumour cells, this theory remains speculative.

The pharmacological verification of the chemosensitising effect of JIP1 inhibition in OS cells with BI-78D3 is promising in view of possible clinical translation of this treatment strategy, because whereas RNAi is a powerful, efficient method for the systematic discovery of drug targets, its applicability in a clinical setting is limited, mainly due to delivery issues. Although there is progression in the field of siRNA delivery techniques and some siRNAs are currently in clinical development, small molecule inhibitors generally have better pharmacokinetic properties and the road to clinical application seems less cumbersome than for siRNA. [201] Additional treatment with a JIP1-inhibitor in OS patients receiving doxorubicin seems a realistic scenario that may elicit improved treatment efficacy and thus requires clinical studies.

To assess the potential clinical relevance of JIP1 expression in OS patients, we analysed human OS tissue samples by immunohistochemical staining of TMAs and correlated JIP1 staining to clinical outcome. At present, there is no specific predictive or prognostic marker in OS. [7,202] Prediction of treatment and/or survival outcomes might be of importance for the stratification of patients to treatment regimens and enable a more individualised treatment of OS patients by offering those patients additional targeted therapy when it could be anticipated that they will benefit from this specific treatment strategy. The importance of the discovery of reliable biomarkers in OS and their possible role in targeted treatment design has recently been highlighted by other research groups. [7,202-204] In this work, we were particularly interested in the correlations between JIP1 staining and survival and between JIP1 staining and chemotherapy response. Since, as demonstrated herein, silencing of JIP1 in OS cells enhances response to doxorubicin, high JIP1 expression levels in OS could perhaps correlate with inferior response to chemotherapy with doxorubicin and consequently inferior survival outcomes. In our tested dataset, we could not correlate JIP1 expression to inferior response to chemotherapy, possibly because all tested specimens were subdued to multi-agent chemotherapy and not doxorubicin mono-therapy. Also, given the presumed effect of JIP1 inhibition being dependent on defective p53, it might

85


5

be worthwhile to analyse the tested specimens for p53 status. We did observe a distinct difference in overall survival between patients with positive and negative JIP1 staining, where positive JIP1 staining appeared associated with inferior survival outcome. While this finding needs confirmation in an independent dataset, our observation suggests that JIP1 could be a predictor for overall survival in patients with localised OS. Furthermore, JIP1 was found to be expressed in the majority of patient tissue samples, indicating that JIP1 is a relevant molecule in OS and that exploitation of JIP1 inhibition as additional treatment to current standard chemotherapy regimens could be beneficial to a majority of OS patients. In conclusion, based on our results, we propose JIP1 as a potential new drug target for OS to enhance the efficacy of chemotherapies including doxorubicin.

Chapter 6Surface proteomic analysis of

osteosarcoma identifies EPHA2 as

receptor for targeted drug delivery

J. PosthumaDeBoer, S.R. Piersma, T.V. Pham, P.W. van Egmond, J.C. Knol, A.M. Cleton-Jansen, M.A. van Geer, V.W. van Beusechem, G.J.L. Kaspers,

B.J. van Royen, C.R. Jiménez, M.N. Helder

British Journal of Cancer. 2013 Oct 15; 109(8): 2142-54

88

Chapter 6

A B S T R A C TBackground: Osteosarcoma (OS) is the most common bone tumour in children and adolescents. Despite aggressive therapy regimens, treatment outcomes are unsatisfactory. Targeted delivery of drugs can provide higher effective doses at the site of the tumour, ultimately improving the efficacy of existing therapy. Identification of suitable receptors for drug targeting is an essential step in the design of targeted therapy for osteosarcoma. Methods: We conducted a comparative analysis of the surface proteome of human OS cells and osteoblasts using cell surface biotinylation combined with nano-liquid chromatography - tandem mass spectrometry-based proteomics to identify surface proteins specifically upregulated on OS cells. This approach generated an extensive dataset from which we selected a candidate to study for its suitability as receptor for targeted treatment delivery to OS. First, surface expression of the Ephrin type-A receptor (EPHA2) receptor was confirmed using FACS analysis. Receptor targeting and internalisation studies were conducted to assess intracellular uptake of targeted modalities via EPHA2. EPHA2 expression in human tumour tissue was tested using immunohistochemistry. Finally, tissue micro arrays containing cores of human osteosarcoma tissue were stained using immunohistochemistry and EPHA2 staining was correlated to clinical outcome measures. Results: Using mass spectrometry, a total of 2841 proteins were identified of which 156 were surface proteins significantly upregulated on OS cells compared to human primary osteoblasts. EPHA2 was highly upregulated and the most abundant surface protein on OS cells. EPHA2 effectively mediates internalisation of targeted adenoviral vectors into OS cells. Additionally, EPHA2 was expressed in a vast majority of human osteosarcoma samples. Patients with EPHA2 positive tumours showed a trend toward inferior overall survival. Conclusion: The results presented here suggest that the EPHA2 receptor can be considered an attractive candidate receptor for targeted delivery of therapeutics to OS.

89

Surface proteomic analysis of osteosarcoma identifies EPHA2 as receptor for targeted drug delivery

6

I N T R O D U C T I O NOsteosarcoma (OS) is the most common primary malignant bone tumour in children and adolescents. The treatment for OS currently consists of a combination of multi-agent induction chemotherapy, radical excision of the tumour and metastases (when feasible), followed by adjuvant chemotherapy. Despite this aggressive regimen, the survival for patients affected by this tumour remains unsatisfactory. In patients with localised disease, 5-year survival rates of approximately 65% are obtained, however, in the case of metastatic or recurrent disease, 5-year survival rates are reduced to only 20%. [1,5] Clearly, new, more effective treatment regimens are desirable for OS, preferentially enhancing the existing regimens. In the recent past, so-called targeted therapy has gained tremendous interest in anti-cancer treatment, exploiting tumour-specific molecules for therapeutic goals. In the case of cancer-specific treatment by targeted delivery of drugs, therapy should ideally be targeted to a cell surface molecule that is specific for the tumour and is highly expressed on the surface of tumour cells, but not on healthy tissues. The identification of suitable receptors for targeted delivery of therapeutics to OS is essential in the design and development of novel targeted treatment strategies. In this work, we aimed to identify surface markers that are specifically upregulated on OS by performing a proteomic analysis of the surface proteomes of OS cells compared to human primary osteoblasts (hp-OBs) using a mass spectrometry approach. Mass spectrometry-based proteomics has made enormous and rapid technical advancements over the past decade and currently allows for comprehensive comparative analyses to systematically identify and quantify thousands of proteins across multiple biological samples, for example between tumour cells versus healthy controls. [205-207]

Cell surface molecules are known to be involved in important biological processes, including proliferation, differentiation, migration and survival. [206,208-211] Also, surface molecules such as growth factor receptors, cytokines, metalloproteases and integrins are implicated in the development and progression of malignancies, including OS. [25,32,34,36,37,60,83] Given their prominent role in cancer, surface proteins that are highly differentially expressed on OS cells compared with their healthy counterparts and can serve as potential delivery targets are likely to be found. Here, we present a comprehensive, mass spectrometry-based study of the surface proteome of OS, in which we identify multiple highly upregulated surface molecules that could potentially serve as receptors for the targeted delivery of drugs to OS. We select one candidate, the Ephrin type-A receptor 2 (EPHA2), and show its potential as receptor for the intracellular delivery of targeted vectors to OS cells.

90

Chapter 6

M AT E R I A L S A N D M E T H O D SCell cultureHuman osteosarcoma cell lines MG-63, U2OS, Cal-72, [212] SaOS-2 and SaOS-LM7 [175] (LM7) were kindly provided by Dr. C. Löwik (Leiden University Medical Center, Leiden, NL), Dr. S. Lens (Dutch Cancer Institute, Amsterdam, NL), Dr. J. Gioanni (Faculté de Medicine, Nice, FR), Dr. F. van Valen (Westfalische Wilhelms-Universität, Münster, GER) and Prof. Dr. E.S. Kleinerman (MD Anderson Cancer Center, Houston, TX, USA), respectively. Human primary (short-term culture; i.e. passage <10) osteoblasts (ORT-1, Hum31, Hum54, Hum63 and Hum65) were obtained from healthy patients undergoing total knee replacement after informed consent. All cells, with the exception of LM7, were cultured in D-MEM (Gibco, Invitrogen) supplemented with 10% fetal calf serum (FCS) and 1mg/mL Penicillin-Streptomycin (Gibco) at 37°C and 5% CO2 in a humidified incubator. LM7 was cultured in Eagle’s-MEM supplemented with 10% FCS, 1 mg/mL Pen-Strep, 1% non-essential amino acids, 1% sodium pyruvate, 2nM L-glutamine and 2% MEM-Vitamin solution (Gibco, Invitrogen) at 37°C and 5% CO2 in a humidified incubator.

Cell surface protein isolation and gel-electrophoresisFor the isolation and collection of surface proteins, we used the Pierce® Cell Surface Protein Isolation Kit (ThermoScientific), following the manufacturer’s instructions adjusted according to the protocol described by de Wit et al. [213] (Figure 6.1A) Supplementary File 6.1 (Appendix B) provides an elaborate description of the cell surface protein isolation and mass spectrometry protocol. Per biological replicate, 3x107 cells were cultured in five 75 cm2

flasks. In brief, the cells were incubated with Sulfo-NHS-SS-Biotin for 30 minutes at 4°C after which the biotinylation reaction was quenched. The cells were washed, harvested by gentle scraping and lysed using the provided lysis buffer in the presence of a protease inhibitor cocktail (Sigma-Aldrich). To capture biotinylated (surface) proteins, protein lysates were incubated with Neutravidin Agarose gel for 2h in a column. The unbound (unbiotinylated) proteins, representing the intracellular fraction, were separated from the captured surface proteins by centrifugation of the column. The intracellular fraction was stored at -20°C to serve as an internal control for the surface protein isolation process. (see Appendix B: Supplementary File 6.1) Finally, the captured surface proteins were eluted from the biotin-NeutrAvidin Agarose by incubation with dithiothreitol (DTT) in PBS. The eluted proteins, that is, the cell surface proteins, were collected by column centrifugation.

For all cell lines, three biological replicates were obtained; per cell line, the cell surface proteins were pooled and concentrated ten times using a Microcon YM-10 filter (Millipore) to obtain adequate protein concentrations for gel-electrophoresis. Protein concentrations were quantified using the BCA protein Assay Kit (Pierce) and the lysates were stored at -20°C until use.

91


6

Cell surface protein lysates were separated by 1D gel-electrophoresis. The obtained gel was fixed in 50% ethanol containing 3% phosphoric acid for 1h, rinsed in Milli-Q water (MQ) and stained with Coomassie-R250 overnight (O/N) to visualise the protein bands. After staining, the gel was washed vigorously with MQ to rid the Coomassie and stored in MQ at 4°C until further processing.

In-gel digestion and mass spectrometryThe proteins were further processed into tryptic peptides by in-gel digestion according to the protocol described by Piersma et al., [214] which was modified so that the pre-treatment phase of this protocol was applied to the whole gel instead of to protein fractions. This allows for as good retrieval of peptides while reducing the laboriousness of this procedure. [207]

Peptide separation was performed by nano-liquid chromatography using an Ultimate 3000 nanoLC system (Dionex LC-Packings, Amsterdam, The Netherlands). Intact peptide MS spectra and MS/MS spectra were acquired on a LTQ-FT hybrid mass spectrometer (ThermoFisher, Bremen, Germany) as described in detail in [214,215] (see Appendix B: Supplementary File 6.1).

Protein identification and quantificationTo identify proteins from the acquired data, MS/MS spectra were searched against the human IPI database 3.62 (83.947 entries) using Sequest (version 27, rev 12). Scaffold 3.00.04 (Proteomesoftware, Portland, OR) was used to organize the gel-slice data and to validate peptide and protein identifications. For quantitative protein analysis spectral counting (the number of assigned MS/MS spectra for each identified protein) was used. For quantification across samples, the spectral counts were normalised to the sum of the spectral counts per biological sample. Differential analysis of samples was performed using the beta-binominal test as described previously. [216] Protein identification and quantification details can be found in [214,216]. The obtained dataset was exported to Excel for further use.

Data miningSubcellular protein localisations were verified using the Uniprot Knowledgebase (www.uniprot.org), searching under the header “GO annotation” for evidence of expression at the cell and/or plasma membrane. Protein-protein interactions (PPIs) were investigated using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) version 9.0 (www.string-db.org). For cluster and gene ontology (GO) analyses we used the Cytoscape platform for network analysis (www.cytoscape.org), using the plug-ins CluterONE version 0.91 (http://chianti.ucsd.edu/cyto_web/plugins/) for the clustering of proteins and BINGO version 2.44 (http://www.psb.ugent.be/cbd/papers/BiNGO/) for the analysis of GO annotations of

92

Chapter 6

biological processes associated with our obtained protein networks. In addition, verification of subcellular protein localisation was performed using Ingenuity pathway analysis software (IPA, Ingenuity Systems, Inc. CA, USA) and Biomart portal, version 7.0 (www.biomart.org).

FACSSurface expression of EPHA2 on OS cells and hp-OBs was verified by flow cytometric analysis. Per sample, cells from three T25 culture flasks were harvested and pooled. Cells were trypsinised, washed in PBS containing 1% Bovine Serum Albumin (BSA) and fixed in 70% ice-cold ethanol for 24h. After fixation, cells were washed in PBS with 1% BSA, incubated with 0.25 µg mouse-anti-EPHA2 (Sigma-Aldrich) diluted in PBS containing 1% BSA for 1 h at 4°C, rinsed and then incubated with the secondary FITC-conjugated rabbit-anti-mouse antibody (DAKO) in PBS containing 1% BSA for 30 minutes at RT in the dark. Finally, cells were rinsed and diluted in 200 µL PBS containing 1% BSA. Measurements were performed using a FacsCalibur II Flow Cytometer and the Cell Quest Pro programme (Becton Dickinson).

Receptor targeting and internalisation To investigate EPHA2-mediated intracellular uptake of targeted moieties, we performed internalisation and competition studies using AdYSA. AdYSA is a GFP-expressing, non-replicative adenoviral vector specifically targeting EPHA2 due to the insertion of a small peptide (YSA) with high EPHA2 binding-affinity into the HI loop of the adenovirus serotype 5 fibre knob. This insertion was combined with ablation of binding sites for the native adenoviral receptors CAR and αv-integrins to further ensure specificity towards EPHA2. [217] Osteosarcoma cells and hp-OBs were plated in 96-well format and incubated with 100 µL complete medium at multiplicities of infection (MOIs) ranging from 0.2 to 100 infectious units (IU) per cell. Transduction efficiency was analysed after 48h by fluorescence read-out using an Acumen eX3® microplate cytometer (TTP LabTech Ltd., UK). Cells were incubated with 0.06 nM Hoechst for 30 minutes at 37°C in the dark to stain nuclei. Cells were excited with a 405 nm laser for Hoechst (blue) and a 448 nm laser for GFP (green) and the emissions of the positive cells for Hoechst and GFP were counted in their respective channels. The ratio of GFP-positive to Hoechst-positive objects was used as the measure for transduction efficiency and recorded as percentages. AdGFP vector with native adenovirus serotype 5 tropism (MOI-10 and MOI-100) was used as a control for transduction for all cell lines.

Competition experiments were performed by the addition of synthetic YSA peptide (YSAPDSVPMMS) to the cells, thereby blocking the EPHA2 receptor-mediated uptake of AdYSA. The irrelevant peptide Cys.S (SSSKEENRIIPGG) was used as negative control. Cells were plated and pre-incubated with 250 µM of either peptide in 50 µL PBS for 20 minutes at RT. Then, 50 µL of complete medium with AdYSA was added to the cells to a final MOI-100 and left to incubate for 30 minutes at 37°C. The cells were washed with PBS and

93


6

incubated with complete medium for 48h after which transduction efficiency was analysed as described above.

FACS and Acumen data analysis was performed using GraphPad Prism® Version 5.01 (GraphPad Software, Inc. San Diego, CA, USA).

ImmunohistochemistrySlides containing paraffin embedded tissue samples of primary OS (n=10), OS lung metastases (n=8) and normal bone (n=11) tissue were obtained from excision specimens from our institute. The primary tumours and metastases were not obtained from the same patients. The primary OS lesions comprise five biopsies (chemonaïve) and five resections (after neo-adjuvant chemotherapy). All patients with metastatic lesions had received chemotherapy in an earlier phase of their disease, but had not yet received treatment for their metastases. The slides were deparaffinised in Xylene and rehydrated in a graded series of alcohol. Endogenous peroxidase activity was inhibited by incubation with 0.3% H2O2 diluted in methanol for 30 minutes. Tissue micro arrays (TMAs) were boiled in 10 mM citrate buffer (pH 6) for 10 minutes and subsequently rinsed in PBS. The slides were incubated with mouse-anti-EPHA2 (Sigma-Aldrich) O/N at 4°C. Antigen visualisation was performed using the EnVision+ Poly-HRP IHC Kit (Immunologic) and DAB chromogen solution. Slides were counterstained with hematoxylin, dehydrated and mounted.

Tissue micro arrays Two tissue micro arrays (TMAs) containing a total of 647 cores of human primary OS samples (corresponding to 130 OS patients) and 20 control tissue cores, were stained and analysed for EPHA2 expression (Appendix B: Supplementary Figure 6.1A) and then correlated to clinical and survival data. The TMAs were crafted at the Leiden University Medical Center (LUMC, Leiden, the Netherlands) according to the protocol described in Mohseny et al. [182] In brief, tissue cores were obtained from tumour areas selected by a pathologist based on a haematoxylin and eosin staining of each specimen, to assure that tumour tissue was sampled. 144 fresh frozen paraffin embedded OS samples were used. At least three cores per tumour were sampled in order to intercept intra-tumoural heterogeneity. All patients were treated for OS at the LUMC in the period between 1984 and 2009. Available clinical data includes: age, gender, location and side of the primary tumour, response to chemotherapy according to the Huvos grading system [183] (when available), metastasis, recurrence, date of recurrence, survival, date of death (when applicable) and time of follow-up. The TMA slides were heated at 60°C for 20 minutes prior to deparaffinisation and rehydration. Immunohistochemical staining followed as described above.

94

Chapter 6

Tissue micro array scoring and analysisThe stained tissue micro array slides were automatically scanned as described previously. [142] All 647 samples were independently examined and scored twice by two of the authors (JP and PE), in two separate sessions. The scoring was performed using dedicated TMA scoring software (3DHISTECH Ltd., Budapest, Hungary) in a blinded manner. To facilitate the scoring and improve the reproducibility of scoring, a consensus chart with exemplary staining patterns per category was created (Appendix B: Supplementary Figure 6.1B) and used by the observers during scoring. The staining per tissue was assessed and valued as “negative”, “weak”, “moderate” or “positive”. Owing to loss of cores during the cutting and staining procedure, not all cores could be included for analysis. Samples were considered unsuitable for scoring when less than 30% of tissue was present at the site of the core. In case of insufficient tissue, the cores were given the value “no data”. The grading scale eventually consisted of 5 values (0 = no data; 1 = negative; 2 = weak; 3 = moderate; 4 = positive). Each tumour was represented by three cores on the TMAs allowing for a maximum of 12 observations per tumour (i.e., 3 cores x 2 observers x 2 scoring sessions). To assure reliable scoring per tumour sample, a tumour had to be assigned minimally 8 scores of 1 or higher (thus excluding “no data” observations), to be included in the statistical analysis. We used the mean of the observations to assign the final staining score to each sample. The clinical data and the staining results were entered and statistically analysed in SPSS, version 17.0 (SPSS Software, Inc., Chicago, IL, USA). To assess inter- and intra-observer agreement in grading EPHA2 staining, Kappa statistics were used. As inter- and intra-observer reproducibility may be biased by an overemphasis of patients with grade 0 findings, kappa values were therefore also calculated with the exclusion of grade 0 findings (censored Kappa). Values between 0 and 1 were interpreted according to modified published guidelines (Supplementary Table 6.1). [184,185] Kaplan-Meier analysis was used to assess survival and differential survival between groups was analysed using the Log Rank test. To determine significant differences between categorical groups, the Pearson chi-square test was used. In numerical groups, the independent t-test and one-way ANOVA were used. The threshold for statistical significance was set at p<0.05.

R E S U LT SSurface proteomics identifies proteins with differential abundance on osteosarcoma cells and human primary osteoblastsIn our search for surface proteins that could serve as receptors for targeted drug delivery to OS, we performed surface biotinylation of five OS cell lines and three human primary OB cell cultures for cell surface protein isolation and combined this with a comprehensive proteomics analysis using gel fractionation and mass spectrometry (Figure 6.1A).



95


6

Cal-72

SaOS-2

Hum54Hum31ORT-1

U2OS

LM7MG-63

U2OS

MG-63

LM7

SaOS-2

Hum63 Hum65

EPHA2 staining intensity

Cell

coun

ts

Figure 6.1 cell surface protein isolation, visualisation of the proteomic data and EPHA2 expression levels. (A) General workflow of cell surface protein isolation. Cells were cultured in 75 cm2 culture flasks and incubated with Sulfo-NHS-SS-Biotin that covalently binds to primary amino groups of extracellular proteins. The cells were lysed and lysates were incubated with Neutravidin Agarose beads to capture the biotinylated proteins. The biotinylated (cell surface) proteins were separated from the unbound (intracellular) proteins by centrifugation on a column and then eluted from the biotin-Neutravidin Agarose beads using DDT. (B) A heat map of a supervised cluster analysis of our obtained dataset, visualising the differentially expressed proteins (cut-off: p<0.05) between the three human primary OBs (Hum31, Hum54 and ORT-1) and the five OS cell lines (SaOS-2, MG-63, U2OS, Cal-72 and LM7). (C) EPHA2 expression data acquired by nanoLC-MS/MS. Differential expression of EPHA2 between the OS cell lines (black bars) and hp-OBs (grey bars) is evident, approximately 12-fold across the two types of cell, a difference that is highly significant (beta-binominal test; ***, p=0.0005). (D) EPHA2 expression data acquired by flow cytometry in OS cell lines SaOS-2, LM7, MG-63 and U2OS and hp-OBs Hum63 and Hum65, expressed as median (signal-to-background-ratio) fluorescence intensities per cell line. Bars represent experiments performed in triplicate; error bars indicate standard deviation (SD). The OS cell lines (black bars) show a convincingly higher EPHA2 expression than the hp-OBs (grey bars). On average, the OS cells were found to express 4-fold higher levels of EPHA2 than the healthy bone cells and this difference is significant (student’s t-test; **, p<0.01). (E) Histograms of EPHA2 expression in OS cell lines; visualisation of flow cytometry results. Staining positivity for EPHA2 is indicated by an increase in fluorescent signal and a concomitant right-shift of the histograms.

96

Chapter 6

Acc

essi

onnu

mbe

rG

ene

sym

bol

Nam

eSp

ectr

al C

ount

s pe

r ce

ll lin

ep-

valu

eFr

eque

ncy

Fold

ch

ange

Abu

ndan

ce

SaOS-2

MG-63

U2OS

Cal-72

LM7

Hum54

Hum31

ORT-1

IPI0

0021

267

EPHA

2Ep

hrin

type

-A re

cept

or 2

156

6943

119

112

714

60,

0005

3340

512

100

IPI0

0435

020

NCA

M1

Neu

ral c

ell a

dhes

ion

mol

ecul

e 1

160

5432

1756

01

00,

0008

6432

513

464

IPI0

0009

803

ITGA

4In

tegr

in a

lpha

-416

129

1335

552

40

0,00

8530

975

2959

IPI0

0216

221

ITGA

6In

tegr

in a

lpha

-640

6756

9826

70

00,

0001

4465

526

57

IPI0

0398

020

ODZ

3Te

neur

in-3

6698

5723

190

00

0,00

0037

505

∞53

IPI0

0219

421

EPHB

2Ep

hrin

type

-B re

cept

or 2

4578

5216

440

12

0,00

0232

775

4347

IPI0

0002

406

BCAM

Basa

l cel

l adh

esio

n m

olec

ule

8346

3830

82

10

0,00

2546

365

3741

IPI0

0027

232

IGF1

RIn

sulin

-like

gro

wth

fact

or 1

rece

ptor

7641

3722

225

40

0,00

0745

435

1340

IPI0

0220

845

ITGB

4In

tegr

in b

eta-

474

2327

1840

00

00,

0000

2229

5∞

37

IPI0

0152

871

LRRC

15Le

ucin

e-ric

h re

peat

-con

tain

ing

prot

ein

1595

2419

622

00

00,

0003

4199

5∞

34

IPI0

0170

706

TMEM

2Tr

ansm

embr

ane

prot

ein

257

2228

4113

00

00,

0000

1970

5∞

32

IPI0

0296

992

AXL

Tyro

sine-

prot

ein

kina

se re

cept

or U

FO13

4435

238

20

00,

0008

8052

539

25

IPI0

0289

342

EPHB

4Ep

hrin

type

-B re

cept

or 4

3031

2520

62

10

0,00

0930

085

2022

IPI0

0031

713

CD70

CD70

anti

gen

5414

95

262

00

0,00

5377

825

3421

IPI0

0874

147

CXAD

RCo

xsac

kiev

irus a

nd a

deno

viru

s rec

epto

r40

2320

315

01

00,

0021

8668

543

20

IPI0

0942

398

ODZ

2Te

neur

in-2

4124

151

160

00

0,00

1158

665

∞19

IPI0

0290

328

PTPR

JRe

cept

or-t

ype

tyro

sine-

prot

ein

phos

phat

ase

eta

1931

1318

120

10

0,00

0050

885

4019

IPI0

0412

492

PLXN

D1Pl

exin

-D1

3327

726

15

00

0,02

0867

845

1219

IPI0

0940

698

FAT1

Prot

ocad

herin

Fat

147

61

1622

00

00,

0013

0381

5∞

19

Acc

essi

onnu

mbe

rG

ene

sym

bol

Nam

eSp

ectr

al C

ount

s pe

r ce

ll lin

ep-

valu

eFr

eque

ncy

Fold

ch

ange

Abu

ndan

ce

SaOS-2

MG-63

U2OS

Cal-72

LM7

Hum54

Hum31

ORT-1

IPI0

0013

897

ADAM

10Di

sinte

grin

and

met

allo

prot

eina

se d

omai

n-co

ntai

ning

pro

tein

10

4212

818

121

00

0,00

0738

455

5919

IPI0

0297

224

SUSD

5Su

shi d

omai

n-co

ntai

ning

534

2018

119

30

00,

0005

0536

520

19

IPI0

0337

612

DCBL

D1Di

scoi

din,

CU

B an

d LC

CL d

omai

n co

ntai

ning

110

2722

217

00

00,

0000

4458

5∞

17

IPI0

0021

058

SLC4

A7So

dium

bic

arbo

nate

cot

rans

port

er 3

2914

1316

105

00

0,00

0434

085

1117

IPI0

0418

426

CNN

M4

Met

al tr

ansp

orte

r CN

NM

425

2320

82

10

00,

0032

0870

549

16

IPI0

0177

968

SCAR

B1Sc

aven

ger r

ecep

tor c

lass

B m

embe

r 117

2322

111

00

00,

0006

0759

5∞

15

IPI0

0011

564

SDC4

Synd

ecan

-46

2828

101

11

00,

0335

9498

519

15

IPI0

0010

676

PLAU

RU

roki

nase

pla

smin

ogen

acti

vato

r sur

face

rece

ptor

1713

116

211

00

0,00

0167

795

4414

IPI0

0168

565

CNN

M3

Met

al tr

ansp

orte

r CN

NM

324

1214

87

00

00,

0000

4022

5∞

13

IPI0

0220

325

INSR

Insu

lin re

cept

or24

149

104

00

00,

0000

9844

5∞

12

IPI0

0031

822

SLC5

A6So

dium

-dep

ende

nt m

ultiv

itam

in tr

ansp

orte

r14

1514

61

00

00,

0007

1568

5∞

10

IPI0

0852

735

FAT4

Prot

ocad

herin

Fat

434

41

12

00

00,

0126

7229

5∞

8

IPI0

0008

085

SLC3

9A10

Zinc

tran

spor

ter Z

IP10

1016

103

10

00

0,00

1318

045

∞8

IPI0

0219

187

APP

Amyl

oid

beta

A4

prot

ein

116

714

10

00

0,00

0599

075

∞8

IPI0

0000

070

LDLR

Low

-den

sity

lipop

rote

in re

cept

or5

1110

131

00

00,

0005

6815

5∞

8

IPI0

0002

441

SDC1

Synd

ecan

-12

1815

21

00

20,

0417

9604

513

8

IPI0

0001

893

PCDH

7Pr

otoc

adhe

rin-7

253

15

30

00

0,00

3665

965

∞7

IPI0

0060

569

ABHD

12M

onoa

cylg

lyce

rol l

ipas

e AB

HD12

132

46

121

00

0,00

7348

715

237

IPI0

0027

728

SLC7

A1Hi

gh a

ffini

ty c

ation

ic a

min

o ac

id tr

ansp

orte

r 16

99

23

00

00,

0002

8837

5∞

6

IPI0

0010

474

SLC7

A11

Cysti

ne/g

luta

mat

e tr

ansp

orte

r2

62

162

00

00,

0027

9280

5∞

6

IPI0

0550

382

SLC2

9A1

Equi

libra

tive

nucl

eosid

e tr

ansp

orte

r 15

117

31

10

00,

0080

4474

518

6

IPI0

0012

442

G3BP

1Ra

s GTP

ase-

activ

ating

pro

tein

-bin

ding

pro

tein

11

68

83

00

00,

0006

3535

5∞

5

IPI0

0024

012

FZD7

Frizz

led-

76

31

103

00

00,

0024

4694

5∞

5

IPI0

0215

980

PVRL

2Po

liovi

rus r

ecep

tor-r

elat

ed p

rote

in 2

, iso

form

alp

ha5

65

42

00

00,

0000

8064

5∞

5

Tabl

e 6.

1 To

p lis

t of u

preg

ulat

ed c

ell s

urfa

ce p

rote

ins

on o

steo

sarc

oma.

The

top

43 p

rote

ins

iden

tified

by

nano

LC-M

S/M

S th

at m

eet

the

follo

win

g cr

iteri

a: s

igni

fican

t up

regu

latio

n on

OS

cells

com

pare

d to

hum

an p

rim

ary

oste

obla

sts

(p <

0.

05);

expr

essi

on in

five

out

of fi

ve t

umou

r ce

ll lin

es; 1

0-fo

ld u

preg

ulati

on o

n O

S ce

lls c

ompa

red

to h

uman

pri

mar

y os

teob

last

s an

d hi

gh a

bund

ance

(spe

ctra

l cou

nts

> 5

per

tum

our

cell

line)

.

97


6

Acc

essi

onnu

mbe

rG

ene

sym

bol

Nam

eSp

ectr

al C

ount

s pe

r ce

ll lin

ep-

valu

eFr

eque

ncy

Fold

ch

ange

Abu

ndan

ce

SaOS-2

MG-63

U2OS

Cal-72

LM7

Hum54

Hum31

ORT-1

IPI0

0021

267

EPHA

2Ep

hrin

type

-A re

cept

or 2

156

6943

119

112

714

60,

0005

3340

512

100

IPI0

0435

020

NCA

M1

Neu

ral c

ell a

dhes

ion

mol

ecul

e 1

160

5432

1756

01

00,

0008

6432

513

464

IPI0

0009

803

ITGA

4In

tegr

in a

lpha

-416

129

1335

552

40

0,00

8530

975

2959

IPI0

0216

221

ITGA

6In

tegr

in a

lpha

-640

6756

9826

70

00,

0001

4465

526

57

IPI0

0398

020

ODZ

3Te

neur

in-3

6698

5723

190

00

0,00

0037

505

∞53

IPI0

0219

421

EPHB

2Ep

hrin

type

-B re

cept

or 2

4578

5216

440

12

0,00

0232

775

4347

IPI0

0002

406

BCAM

Basa

l cel

l adh

esio

n m

olec

ule

8346

3830

82

10

0,00

2546

365

3741

IPI0

0027

232

IGF1

RIn

sulin

-like

gro

wth

fact

or 1

rece

ptor

7641

3722

225

40

0,00

0745

435

1340

IPI0

0220

845

ITGB

4In

tegr

in b

eta-

474

2327

1840

00

00,

0000

2229

5∞

37

IPI0

0152

871

LRRC

15Le

ucin

e-ric

h re

peat

-con

tain

ing

prot

ein

1595

2419

622

00

00,

0003

4199

5∞

34

IPI0

0170

706

TMEM

2Tr

ansm

embr

ane

prot

ein

257

2228

4113

00

00,

0000

1970

5∞

32

IPI0

0296

992

AXL

Tyro

sine-

prot

ein

kina

se re

cept

or U

FO13

4435

238

20

00,

0008

8052

539

25

IPI0

0289

342

EPHB

4Ep

hrin

type

-B re

cept

or 4

3031

2520

62

10

0,00

0930

085

2022

IPI0

0031

713

CD70

CD70

anti

gen

5414

95

262

00

0,00

5377

825

3421

IPI0

0874

147

CXAD

RCo

xsac

kiev

irus a

nd a

deno

viru

s rec

epto

r40

2320

315

01

00,

0021

8668

543

20

IPI0

0942

398

ODZ

2Te

neur

in-2

4124

151

160

00

0,00

1158

665

∞19

IPI0

0290

328

PTPR

JRe

cept

or-t

ype

tyro

sine-

prot

ein

phos

phat

ase

eta

1931

1318

120

10

0,00

0050

885

4019

IPI0

0412

492

PLXN

D1Pl

exin

-D1

3327

726

15

00

0,02

0867

845

1219

IPI0

0940

698

FAT1

Prot

ocad

herin

Fat

147

61

1622

00

00,

0013

0381

5∞

19

Acc

essi

onnu

mbe

rG

ene

sym

bol

Nam

eSp

ectr

al C

ount

s pe

r ce

ll lin

ep-

valu

eFr

eque

ncy

Fold

ch

ange

Abu

ndan

ce

SaOS-2

MG-63

U2OS

Cal-72

LM7

Hum54

Hum31

ORT-1

IPI0

0013

897

ADAM

10Di

sinte

grin

and

met

allo

prot

eina

se d

omai

n-co

ntai

ning

pro

tein

10

4212

818

121

00

0,00

0738

455

5919

IPI0

0297

224

SUSD

5Su

shi d

omai

n-co

ntai

ning

534

2018

119

30

00,

0005

0536

520

19

IPI0

0337

612

DCBL

D1Di

scoi

din,

CU

B an

d LC

CL d

omai

n co

ntai

ning

110

2722

217

00

00,

0000

4458

5∞

17

IPI0

0021

058

SLC4

A7So

dium

bic

arbo

nate

cot

rans

port

er 3

2914

1316

105

00

0,00

0434

085

1117

IPI0

0418

426

CNN

M4

Met

al tr

ansp

orte

r CN

NM

425

2320

82

10

00,

0032

0870

549

16

IPI0

0177

968

SCAR

B1Sc

aven

ger r

ecep

tor c

lass

B m

embe

r 117

2322

111

00

00,

0006

0759

5∞

15

IPI0

0011

564

SDC4

Synd

ecan

-46

2828

101

11

00,

0335

9498

519

15

IPI0

0010

676

PLAU

RU

roki

nase

pla

smin

ogen

acti

vato

r sur

face

rece

ptor

1713

116

211

00

0,00

0167

795

4414

IPI0

0168

565

CNN

M3

Met

al tr

ansp

orte

r CN

NM

324

1214

87

00

00,

0000

4022

5∞

13

IPI0

0220

325

INSR

Insu

lin re

cept

or24

149

104

00

00,

0000

9844

5∞

12

IPI0

0031

822

SLC5

A6So

dium

-dep

ende

nt m

ultiv

itam

in tr

ansp

orte

r14

1514

61

00

00,

0007

1568

5∞

10

IPI0

0852

735

FAT4

Prot

ocad

herin

Fat

434

41

12

00

00,

0126

7229

5∞

8

IPI0

0008

085

SLC3

9A10

Zinc

tran

spor

ter Z

IP10

1016

103

10

00

0,00

1318

045

∞8

IPI0

0219

187

APP

Amyl

oid

beta

A4

prot

ein

116

714

10

00

0,00

0599

075

∞8

IPI0

0000

070

LDLR

Low

-den

sity

lipop

rote

in re

cept

or5

1110

131

00

00,

0005

6815

5∞

8

IPI0

0002

441

SDC1

Synd

ecan

-12

1815

21

00

20,

0417

9604

513

8

IPI0

0001

893

PCDH

7Pr

otoc

adhe

rin-7

253

15

30

00

0,00

3665

965

∞7

IPI0

0060

569

ABHD

12M

onoa

cylg

lyce

rol l

ipas

e AB

HD12

132

46

121

00

0,00

7348

715

237

IPI0

0027

728

SLC7

A1Hi

gh a

ffini

ty c

ation

ic a

min

o ac

id tr

ansp

orte

r 16

99

23

00

00,

0002

8837

5∞

6

IPI0

0010

474

SLC7

A11

Cysti

ne/g

luta

mat

e tr

ansp

orte

r2

62

162

00

00,

0027

9280

5∞

6

IPI0

0550

382

SLC2

9A1

Equi

libra

tive

nucl

eosid

e tr

ansp

orte

r 15

117

31

10

00,

0080

4474

518

6

IPI0

0012

442

G3BP

1Ra

s GTP

ase-

activ

ating

pro

tein

-bin

ding

pro

tein

11

68

83

00

00,

0006

3535

5∞

5

IPI0

0024

012

FZD7

Frizz

led-

76

31

103

00

00,

0024

4694

5∞

5

IPI0

0215

980

PVRL

2Po

liovi

rus r

ecep

tor-r

elat

ed p

rote

in 2

, iso

form

alp

ha5

65

42

00

00,

0000

8064

5∞

5

Tabl

e 1

Conti

nued

98

Chapter 6

In total, 2841 proteins were identified in the eight different surface lysates (Supplementary Table 6.2). In all, 684 proteins were significantly upregulated in OS cells compared with hp-OBs (p<0.05) (Figure 6.1B). A total of 151 proteins were confirmed as being cell surface proteins, based on their cellular component ontology term in the Uniprot Knowledgebase (www.uniprot.org). We further investigated PPIs, protein clusters, networks, subcellular localisations and the biological processes of a selected set of identified proteins, using the STRING 9.0 tool, Cytoscape software with packages ClusterONE and BINGO, and Ingenuity Pathway Analysis. The results of these elaborate analyses are provided in Supplementary File 6.2 (Appendix B) and Supplementary Tables 6.3 and 6.4.

Proteins of interest were further selected as putative receptors for drug targeting to OS according to the following considerations: (1) OS is a heterogeneous tumour, therefore we applied a frequency filter stating that candidate surface proteins should be expressed consistently, that is, on all five analysed OS cell lines; (2) to ascertain a therapeutic window that is wide enough to realise a dose increase at the site of the tumour while sparing other tissues, candidate surface proteins should be >10-fold upregulated on OS cells compared with hp-OBs; and (3) to allow effective drug delivery, candidate surface receptors should be abundantly expressed. Abundance was defined as an average of >5 spectral counts (sc) per sample. When applying these criteria to our dataset, we retrieved 97 cell surface proteins that were found to be expressed in 5 out of 5 OS cell lines; 55 of which were >10-fold upregulated in OS and, finally, 43 proteins were abundant (Table 6.1). These 43 proteins are considered promising candidate receptors for targeted drug delivery to OS. The hit list contains surface proteins that are well known in tumour biology, such as Integrins, Ephrins and Ephrin receptors, growth factor receptors (i.e., the Insulin Receptor and the Insulin-like Growth Factor 1 Receptor), (proto)cadherins and various transmembrane transporters.

Among the selected hits, the Ephrin type-A receptor 2 (EPHA2) was found to be the most abundant surface protein on OS cells, with an average of 100 sc per OS cell line, significantly differentially expressed (p=0.0005) and 12-fold upregulated on OS cells compared with hp-OBs (Table 6.1, Figure 6.1C). Based on its abundance and degree of differential expression, we chose to further investigate the EPHA2 receptor for its suitability as a receptor for the targeted delivery of therapeutics to OS.

Validation of cell surface location of EPHA2 by FACS analysisTo validate our mass spectrometry data and ascertain high levels of cell surface localisation of EPHA2 on OS cells and low levels on hp-OBs, cells were subjected to FACS analysis. The measure of EPHA2 surface expression was defined as the ratio of median fluorescence-signal-intensity between EPHA2-labelled cells vs. non-labelled cells. Figure 6.1D shows that EPHA2 expression is considerably higher in OS cells compared to hp-OBs (student’s t-test;




99


6

p<0.01), which is in concordance with our mass spectrometry data (Figure 6.1C). Figure 6.1E shows that all the tested OS cell lines convincingly express EPHA2, as we observe a substantial right-shift of the histograms of the EPHA2-labelled cells compared with the control samples. Thus, the flow cytometry results confirm the primary mass spectrometry findings.

EPHA2 receptor targeting and uptake studiesTo confirm that the EPHA2 receptor can be used for the specific delivery and uptake of targeted vectors to OS, we performed internalisation studies with AdYSA, a GFP-expressing, replication defective, adenoviral vector that was designed to specifically target EPHA2. [217] First, transduction efficiency on OS cells and hp-OBs was assessed. Cells were plated and subjected to increasing MOIs of AdYSA up to 100 IU/cell. At 48 h post-infection, GFP expression was analysed. All OS cell lines showed a dose-dependent transduction by AdYSA. Contrarily, the hp-OBs were poorly transduced with AdYSA. Figure 6.2A shows typical fluorescence image overlays of two OS cell lines (SaOS-2 and MG-63) and one human primary OB (Hum63) subjected to different MOIs of AdYSA. As can be appreciated, SaOS-2 showed clear GFP expression already at a low MOI (6.25) and MG-63 is effectively transduced at higher MOI. In contrast, Hum63 showed a very low response to incubation with AdYSA. GFP expression was essentially absent, up to the highest virus dose used. The difference in transduction efficiency between the cell types was tested at MOI-100 (Figure 6.2B). Osteosarcoma cells showed a three- to six-fold higher GFP expression than hp-OBs, which was significant (student’s t-test; p<0.05). From this we conclude that vectors targeted to the EPHA2 receptor can be successfully internalised into EPHA2-expressing OS cells, whereas hp-OB cells scarcely expressing EPHA2 remain essentially unaffected.

To prove that AdYSA uptake is specifically mediated by the EPHA2 receptor, we performed competition experiments with synthetic YSA peptide. Cells were plated and pre-incubated with the synthetic YSA peptide to block EPHA2 receptor-mediated internalisation. As a control, cells were pre-incubated with an irrelevant peptide Cys.S and with AdYSA alone. Cells that were pre-incubated with YSA peptide showed a strong decrease in viral uptake compared to cells incubated with virus alone or cells treated with Cys.S peptide (Figure 6.2C). Blocking the receptor with YSA peptide reduced AdYSA transduction efficiency four- to seven-fold in different OS cell lines (significance p<0.001 to p<0.01), whereas Cys.S peptide did not hamper the uptake of viral particles. This indicates that YSA peptide specifically blocks the EPHA2 receptor, thereby decreasing transduction by AdYSA. In hp-OBs that were essentially poorly transduced with AdYSA, no decrease in transduction efficiency by receptor blockage could be detected. Thus, internalisation of AdYSA into OS cells was indeed specifically mediated by the EPHA2 receptor.

100

Chapter 6

Taken together, the data from these experiments demonstrate that targeted vectors can be internalised into OS cells via the EPHA2 receptor and that this uptake is convincingly higher in OS cells than in hp-OBs. Furthermore, the uptake of the targeted vector relies specifically on the EPHA2 receptor. Given the differential uptake and the specificity thereof, EPHA2 can be considered a putative receptor for the targeted delivery of treatment to OS.

EPHA2 is differentially expressed in human osteosarcoma tissue compared with healthy bone tissue To investigate whether EPHA2 is, in addition to on OS cell lines, also expressed on human OS tissue, we performed immunohistochemical staining on archival OS tumour sections (n=18), including OS lung metastases (n=8) and compared EPHA2 staining with healthy bone tissue (n=11). Clinical details of the tested samples are provided in Supplementary Table 6.5. Figure 6.3A shows exemplary staining results of OS sections with positive, moderate and weak staining. The negative samples represent the normal bone sections. Among the OS specimens, 67% showed strong positive EPHA2 staining, 22% showed a moderately positive

SaOS-2 MOI-1.6 MG-63 MOI-1.6 Hum63 MOI-1.6

SaOS-2 MOI-100 MG-63 MOI-100 Hum63 MOI-100







Figure 6.2 EPHA2 receptor specifically mediates targeted adenoviral vector internalisation into OS cells. (A) Representative fluorescence images of SaOS-2 and MG-63 OS cells and Hum63 hp-OBs subjected to transduction with AdYSA at the indicated MOIs, acquired using the Acumen eX3® microplate cytometer. Composite images show Hoechst stained cell nuclei (blue) and GFP-expression (green). (B) Transduction of OS cell lines (green bars) and hp-OBs (blue bars) with EPHA2-targeted adenoviral vector AdYSA, as measured by GFP expression 48 h after subjecting the cells to virus at MOI-100. Bars represent mean results of an experiment performed in triplicate; error bars indicate SD. The observed difference between OS cells and hp-OBs is 3- to 6-fold (student’s t-test; *, p<0.05). (C) Competition by pre-incubation of the cells with synthetic peptides prior to incubation with AdYSA. Bars represent mean results of an experiment performed in triplicate; error bars indicate SD. Green bars represent the control condition in which the cells were incubated with AdYSA alone, grey bars represent the control condition in which cells were pre-incubated with Cys.S peptide, followed by incubation with AdYSA and blue bars represent the competition condition in which cells were pre-incubated with YSA peptide followed by incubation with AdYSA. Receptor blocking with YSA peptide results in significant reduction of transduction efficiency in the OS cell lines (****, p<0.001 for SaOS-2; ***, p<0.005 for LM7 and U2OS; **, p<0.01 for MG-63), but not in the already poorly transduced healthy bone cells (NS).


101


6

staining and 11% stained negative (Figure 6.3B). All but one metastatic lesions (87%) stained positive for EPHA2. We predominantly observed mixed cytoplasmic and plasma membrane staining patterns. In contrast, none of the healthy bone specimens showed clear positivity for EPHA2. Two bone specimens (18%) had a weak staining intensity and 1 (9%) stained moderately positive. In summary, the staining results confirm that there is a significant differential EPHA2 expression on human OS tissue compared with normal bone (chi-square test; p<0.005). When analysed separately both primary and metastatic lesions show significantly increased levels of EPHA2 compared with healthy bone (p<0.005). As a majority of OS expressed EPHA2, this molecule can be considered a relevant receptor for targeted drug delivery to this tumour and targeting this receptor might be beneficial for a majority of patients.

(n = 10) (n = 8)

5 7 4 1 1

χsquare test p = 0.0046 p = 0.004 p = 0.0023

Positive Moderate Weak Negative

Figure 6.3 EPHA2 is expressed on human OS tissue. (A) Immunohistochemical staining results for EPHA2 on human OS and normal bone tissue sections. Per category, two examples are shown. (B) Tissue staining per group. Staining was scored based on the percentage of positive cells and the intensity of the staining of the cells and allotted to the categories: negative, weak, moderate, or positive. One bone sample could not be scored because of limited sample quality (N/A). The staining intensity was significantly higher in the OS samples compared to healthy bone tissue, both in the primary and metastatic lesions (chi-square; ***; p<0.005). Clinical details of all samples are provided in Supplementary Table 6.5.


102

Chapter 6

EPHA2 is expressed in a majority of OS tumour samples and is associated with poor survivalTo assess the clinical relevance of EPHA2 expression in OS and to investigate if EPHA2 could be considered a biomarker for OS, tissue micro arrays (TMAs) containing 647 cores of human primary OS samples were stained and analysed for EPHA2 expression. Every tumour was represented by three cores on the TMAs and certain patients were represented more than once on the TMA, i.e., with cores corresponding to primary biopsies, first resections and resections of metastases and/or relapses. Staining results were correlated to clinical data. All samples were independently scored twice by two of the authors (JP and PE) with high observer agreement (Supplementary Table 6.1). Owing to loss of tissue from the array slides as a result of the cutting and staining procedures, 200 cores were considered unsuitable for scoring, leaving 447 scored tissue cores for analysis. Eventually, 127 primary tumours were assigned a staining score. Of these, 85 tumours (67%) showed EPHA2 positivity and 42 tumours (37%) were EPHA2 negative. In addition, 33 metastatic lesions and 10 local relapses were assigned a staining score. These lesions showed an EPHA2 positivity of 58% and 40%, respectively. To study the predictive value of EPHA2 staining on overall survival, we selected first biopsy and/or resection samples only; diagnostic biopsies had not been exposed to pre-operative chemotherapy while resection samples had been exposed to pre-operative chemotherapy. Samples of recurrences (both metastatic and local) were excluded to avoid confounding for inferior survival as a result of metastatic and/or recurrent disease.

Figure 6.4 Kaplan-Meier curves showing overall survival of OS patients with EPHA2 positive or negative tumours. Kaplan-Meier survival plot showing the cumulative survival of patients suffering from localised OS. Patients were divided into two groups of patients with EPHA2 positive tumours (57 samples) and those with EPHA2 negative tumours (11 samples). The observed difference in the probability of survival between the two groups shows a trend towards inferior survival for patients with positive EPHA2 staining (Log Rank; p=0.065).


103


6

After applying these criteria, 68 patients remained eligible for analysis. Of the analysed patients, 57 (84%) had EPHA2-positive and 11 (16%) had EPHA2-negative tumours. Baseline characteristics did not differ significantly between the EPHA2-positive and EPHA2-negative tumours (Table 6.2). Figure 6.4 shows Kaplan-Meier curves of EPHA2-positive and -negative staining as predictor for overall survival. The difference in cumulative overall survival between patients with EPHA2-positive and -negative tumours shows a trend towards inferior survival

Categorical N total EPHA2 negative EPHA2 positive p-value

Gender Female 35 5 30Male 33 6 27

0.663Location Femur 36 7 29

Humerus 7 2 5Tibia/Fibula 23 2 21Other 2 0 2

0.491Side Left 35 6 29

Right 33 5 280.824

Relapse No 48 7 41Yes 20 4 16

0.580Chemotherapy Poor 34 6 28response Good 21 3 18

Huvos 1 8 0 80.743

Huvos 2 26 6 20Huvos 3 18 2 16Huvos 4 3 1 2

0.332Survival Alive 45 10 35

Deceased 23 1 220.058

Numerical N total EPHA2 negative EPHA2 positive p-valuemean SD range mean SD range

Age (years) Diagnosis 63 14.6 3.5 (7.9-19.2) 15.0 5.0 (5.6-36.8) 0.842Death 23 17.6 - - 16.4 4.4 (7.0 - 25.9) 0.803

Follow-up (months) 63 115.5 98.4 (11-267) 70.1 3.0 (3-258) 0.084Event-free survival (months) 32 2.5 4.4 (0.0-7.6) 9.7 11.0 (0.0-40.2) 0.273Overall survival (years) Alive 40 10.5 8.1 (1.2-22.3) 8.7 6.6 (1.2-21.5) 0.491

Deceased 23 0.9 - - 1.9 1.2 (0.3-4.6) 0.428

Table 6.2 Clinical characteristics and outcome data classified according to EPHA2 stainingBaseline characteristics and clinical outcome corresponding to all analysed first biopsy or resection samples present on the tissue micro arrays. p-values were calculated using the Pearson chi-square test for categorical groups and the independent T-test and One-way ANOVA in numerical groups.

104

Chapter 6

for patients with EPHA2 positive tumours (Log Rank, p=0.065). Among 23 deceased patients, only one patient had an EPHA2-negative tumour (chi-square, p=0.058). Patients with EPHA2-negative tumours showed a better survival outcome (mean 10.5 years) compared to patients with EPHA2-positive tumours (mean 8.7 years), however, this difference was not statistically significant. Also, EPHA2 staining did not significantly influence relapse (p=0.580), response to chemotherapy (p=0.743) or event-free survival (p=0.273) in our dataset. Thus, although EPHA2 staining only shows modest influence on several early clinical outcome parameters, EPHA2 positivity does seem to predict an unfavourable survival outcome. More importantly, the majority of tested primary OS specimens express EPHA2, indicating that when applying targeted drug delivery towards this receptor, a majority of patients may benefit.

D I S C U S S I O NIn this work, we conducted a comparative proteomic analysis of the cell surface proteomes of five OS cell lines and three osteoblast cultures in order to identify a receptor for the targeted delivery of treatment to OS. For this purpose, we combined the biotinylation of the cell surface for the isolation and retrieval of surface proteins with a high-resolution mass spectrometric analysis. Doing this, we obtained an extensive and robust dataset that potentially harbours various interesting novel biomarkers or treatment targets for OS.

The number of published proteomic studies investigating OS is limited. Previous research includes efforts to unravel OS development, to study mechanisms of drug response, to identify new biomarkers for prognosis and therapy response, and, to identify putative therapeutic candidates for OS by studying differences within the entire proteome of OS cells or cell lines compared with human osteoblasts. [211,218-220]

To study the surface proteome of OS cells, we used an in-depth, high resolution approach combining cell surface biotinylation for the isolation and retrieval of plasma membrane proteins, gel fractionation and mass spectrometric analysis using nano-liquid chromatography - tandem mass spectrometry (nanoLC-MS/MS) based proteomics. Typically, surface proteomics analysis consists of surface biotinylation and surface protein purification, protein solubilisation, protein and/or peptide fractionation, peptide digestion and extraction, tandem mass spectrometry and database searching, protein identification and quantification. [206,216] Although plasma membrane proteins have fundamental roles in tumour biology, they tend to be underrepresented in datasets, especially in the studies that employed 2D gel-electrophoresis for protein fractionation. In part, this is attributable to their composition with both hydrophobic and hydrophilic regions, which poses solubility issues in the workup of surface proteins for experimental analyses. In addition, membrane proteins are mostly low-abundant and therefore high-abundant intracellular proteins could potentially overshadow the presence of low-abundant surface proteins, thereby hampering

105


6

their identification and quantification. To subvert these issues, we performed an enrichment of surface proteins by surface biotinylation [206,209,213] and a protein fractionation method highly compatible with membrane proteins (SDS-PAGE) prior to our nanoLC-MS/MS analysis. To our knowledge, this is the first report of cell surface proteomics in OS in a panel of cell lines. Two recent studies reported a preliminary proteomics analysis of the plasma membrane proteome of two OS cell types as isolated by a crude two-phase partitioning method in conjunction with a 2D gel-based analysis or an iTRAQ-LC-MS/MS-based approach. [221,222] Importantly, the key molecules that were selected in these studies as putative biomarkers (i.e., NDRG1 and CD151) out of relatively small datasets (343 proteins in the study of Zhang et al. and 7 proteins in the study of Hua et al.) are also regulated in our extended cell surface proteome dataset, retrieved in a panel of 5 OS cell lines. Such independent, bilateral verification adds to the robustness of our dataset. Furthermore, our hitlist contains several well known cell surface proteins that were previously reported to be expressed and of biological importance in OS, supporting the relevance of our obtained dataset. These proteins include the Insulin Growth Factor Receptor 1 (IGF-1R) [91,97] and Insulin receptor (INSR) [223] that have been reported to stimulate tumour growth in OS; the Wnt-receptor family member Frizzled-7 receptor that is involved in survival, migration and disease progression in OS [32,61,67,224,225]; Integrin-α4, described to be of influence on cell survival in OS [113] and Integrin-β4 that has previously been reported to co-localise with Ezrin and enhance OS metastasis [83]; and, finally, Ephrin receptors (EPHA2, EPHB2, EPHB4) that have been previously implicated in tumour growth and prognosis in OS. [226-228]

EPHA2 was the most abundant upregulated cell surface receptor in our dataset and thus we chose this molecule for follow-up investigations. EPHA2 is a cell surface receptor of the erythropoietin-producing hepatocellular (Eph) tyrosine kinases receptor family and is reported to be overexpressed in various types of cancer, while being comparatively lowly expressed in normal tissues. [229-231] Binding of EPHA2 to its corresponding ligand (EphrinA1), causes receptor phosphorylation and subsequent internalisation and degradation. Phosphorylation of the receptor results in downstream signalling via several well known kinases (PI3K, FAK, Akt, RHOA, MEK, Src family- and MAP-kinases), influencing multiple fundamental processes such as cell morphology, adhesion, proliferation, differentiation, survival, migration, invasion and metastasis. Additionally, EPHA2 induces angiogenesis. [227,229-231] The substantially higher expression levels of EPHA2 on OS compared with healthy bone, in addition to its role in cancer progression, implies that EPHA2 could be a suitable therapeutic target itself and this has been the topic of studies in several malignancies. Preclinical studies in various tumour models showed that targeting of EPHA2 using antibodies or RNAi techniques led to reduced tumorigenicity and restored treatment sensitivity in vitro and in vivo. [229-233] Although some studies report that EPHA2 receptor binding, internalisation and degradation influences cell survival, we did not observe a cytotoxic effect of incubation with AdYSA. It

106

Chapter 6

might be that binding and internalisation of EPHA2 alone is not sufficient to affect OS cell survival and that additional stimuli are needed. Furthermore, in previous work performed in our group, we conducted an siRNA screen in which we analysed the effect of gene-silencing of kinases and kinase-associated genes on the viability of OS cells. [142] The tested library also contained siRNA targeting EPHA2. The results obtained in this screen showed no significant decrease in cell viability upon silencing of EPHA2. Thus, we have evidence suggesting that EPHA2 silencing in OS cells does not influence proliferation rates or cell viability. Nonetheless, it will be interesting to investigate whether EPHA2 could be used as a therapeutic target itself via targeting with other treatment modalities in OS.

Literature on EPHA2 expression in OS is extremely limited, however, recently Fritsche-Guenther et al. reported EPHA2 mRNA overexpression in OS and speculated that this receptor could be a therapeutic target for the treatment of OS. [227] In our study, we used an unbiased, comprehensive mass spectrometry approach to study the OS surface proteome and independently observed EPHA2 receptor overexpression in OS. We could thus conclude that increased EPAH2 gene expression translates into elevated EPHA2 surface protein levels in OS. Furthermore, we found EPHA2 receptor expression in human osteosarcoma patient material. In our studies on 18 archival OS tumour sections and 146 tumour sections on TMAs, we found that the majority of the analysed tumours expressed EPHA2. We then proceeded to assess the potential clinical relevance of EPHA2 expression in patients. For this purpose, we correlated EPHA2 staining to clinical parameters. At present, there is no specific predictive or prognostic marker for OS. [7] EPHA2 expression in tumours has been linked to increased malignancy and reduced survival rates in previous studies. [229,230,232] The Kaplan-Meier analysis of patients with EPHA2-positive versus EPHA2-negative tumours showed a trend toward inferior survival in patients with EPHA2-expressing tumours (p=0.065). We could not find significant differences nor trends between EPHA2-positive and EPHA2-negative tumours in terms of relapse rate, response to chemotherapy or event-free survival. A possible explanation would be the limited number of patients analysed and the relative disproportion between the number of EPHA2-positive (57) and-negative (11) tumours. Our findings suggest that possibly EPHA2 could be a predictor for inferior survival of OS patients deriving from the trend that we observed in our dataset. However, given the fact that this did not reach statistical significance in our relatively small number of samples, this observation needs further confirmation in a larger, independent dataset. Our focus was centered on researching the possibility to use EPHA2 as a receptor for the targeted delivery of therapeutic moieties to OS. The use of EPHA2 for this purpose was studied previously in other cancer types. In preclinical models of gynaecological malignancies, immunoconjugates were reported to specifically bind and reduce the viability of EPHA2-expressing cells. [234] Apart from antibody-based targeting, the synthetic YSA peptide can also efficiently direct various moieties to the EPHA2 receptor. Recently,

107


6

Mitra et al. [235] elucidated functional properties of the YSA peptide as a specific binding molecule to EPHA2 with subsequent downstream signalling events in a prostate cancer cell line. Dickerson et al. [236] described the use of hydrogel nanoparticles coated with YSA peptide for the specific delivery of siRNA to EPHA2-expressing tumour cells, with the aim of enhancing chemosensitivity. YSA has also been coupled to magnetic nanoparticles that were used to specifically bind EPHA2-expressing tumour cells and clear these cells from the circulation and peritoneal fluid of tumour-bearing mice using a strong magnetic field. [237] In this work we showed, in line with previous work in pancreatic cancer, [217,238] that YSA peptide-carrying adenoviral vectors can be internalised efficiently and specifically by EPHA2-expressing OS cells. This could provide a promising strategy to deliver therapy to the tumour rather than to healthy tissues and thus realise a selective anticancer therapy. Importantly, EPHA2 was found expressed in a majority of OS tumour samples, implying that EPHA2 can be considered a clinically relevant molecule for the targeted delivery of drugs to OS, and that exploitation of EPHA2 as a receptor for drug delivery could be beneficial to a majority of OS patients. Future development of EPHA2 as a receptor for targeted drug delivery would need confirmation of our findings in an animal model of osteosarcoma, to investigate whether the specific delivery and uptake of EPHA2-targeted moieties can also be realised in vivo.

In summary, the combined knowledge discussed above plus the new findings that we have presented in this work lead us to conclude that targeted delivery of therapeutic agents via the EPHA2 receptor could be a useful strategy to increase the efficacy of OS treatment.

Chapter 7Summary and discussion

110

Chapter 7

S U M M A R YDespite aggressive treatment regimens, survival outcomes for patients suffering from osteosarcoma, especially for those with metastatic disease, remain dismal. The survival outcome for osteosarcoma has currently stagnated at approximately 65% 5-year survival for patients with localised disease, whereas patients with metastatic or relapsed disease have even lower survival rates of approximately 20% 5-year survival. [1,2] As posed in the introductory section, it seems that the current treatments for osteosarcoma lack efficacy, particularly in treating metastatic disease, and this inefficacy could, at least in part, be attributed to therapy resistance encountered in osteosarcoma cells. The observation that osteosarcoma metastases generally show a poorer response to chemotherapy than do primary tumours suggests that the metastatic cell has characteristics that not only allows it to metastasise, but also to (better) evade cell death after cytotoxic treatment. This is thoroughly discussed in chapter 2. Targeting these features in the metastatic cells might enhance the efficacy of current osteosarcoma treatments, however, OS metastasis is not yet well understood. Further unravelling the biology of OS metastasis should provide new molecular leads that could be used as a rational basis for innovative OS treatments. In chapter 3 we further evaluate the current knowledge that exists on the biological processes that contribute to therapy resistance in osteosarcoma cells and summarise what research efforts have thus far been undertaken to design novel treatment strategies to reverse or circumvent therapy resistance in osteosarcoma. From these first two chapters we conclude that, despite the multitude of pre-clinical and clinical research both in the treatment of osteosarcoma lung metastases and in overcoming treatment resistance in osteosarcoma, very few new agents or strategies have reached clinical application. A profound understanding of mechanisms that underlie therapy resistance in osteosarcoma should provide opportunities to design targeted treatments for osteosarcoma on a rational basis. In the research described in this thesis, we sought for new targeted treatment opportunities to subvert or circumvent the relative resistance to therapy observed in osteosarcoma. This targeted therapy could be achieved either by selectively targeting intracellular proteins essential for tumours to survive, or by a targeted delivery of therapeutics to the tumour by directing therapy selectively to extracellular surface proteins or receptors on tumour cells. Ideally, such new therapeutic strategies should target osteosarcoma-specific molecules, thereby selectively enhancing tumour cell kill while sparing healthy cells.

In search of intracellular treatment targets, we studied the sensitisation of osteosarcoma cells to radiotherapy and chemotherapy. In chapter 4 we describe the investigation of a candidate therapeutic target that was previously reported to be a radiosensitiser in other malignancies, [125,158,162,163] and studied its potential use for the radiosensitisation of osteosarcoma. We show that osteosarcoma can be sensitised to radiation therapy using a small molecule WEE1-inhibitor. Radiotherapy is not commonly applied in the treatment

111

Summary and discussion

7

of osteosarcoma because it is considered a radio-resistant tumour. However, a subgroup of patients, namely those with unresectable tumours (mainly located in the axial skeleton or head-and-neck region), painful skeletal metastases or patients who have undergone an intralesional resection of the primary tumour, could benefit from radiotherapy. Nonetheless, radiotherapy is only moderately effective and therefore we studied a possibility to improve the efficacy of this treatment modality. We observed that after irradiation-induced DNA damage, osteosarcoma cells sustain a prolonged G2 cell cycle checkpoint arrest, granting the cells time to perform DNA repair prior to entry into mitosis and thus evade cell death. WEE1 kinase can indirectly prevent cells from entering mitosis through inhibitory phosphorylation of CDC2, which is then hampered to bind to Cyclin-B, resulting in a G2 phase arrest. After WEE1 inhibition, osteosarcoma cells with damaged DNA are unable to sustain a prolonged G2 phase and are consequently forced into mitosis, leading to a type of apoptosis referred to as mitotic catastrophe. WEE1 expression analysis revealed that WEE1 gene- and protein expression is increased in osteosarcoma compared to healthy bone tissue, indicating that osteosarcoma patients treated with radiotherapy could benefit from concomitant WEE1 inhibition. Further in vivo work needs to be performed before translating these pre-clinical results to clinical testing of WEE1 inhibition combined with radiotherapy in osteosarcoma patients. The further development of WEE1 inhibition as a novel treatment strategy for the radiosensitisation of osteosarcoma is helped by the fact that WEE1 inhibitors have already been tested in clinical trials for other indications in oncology and therefore safety and dosage data are available. [161] Nonetheless, we expect it to be very challenging to gather sufficient patients in which radiotherapy is indicated, to study the effectiveness of this new combination therapy.

In the following chapter, (chapter 5) we focussed on chemosensitisation. We hypothesised that targeting essential survival pathways in combination with administering conventional therapy may lead to enhanced cell death. The identification of essential regulators of drug response in osteosarcoma is crucial in the design of such a new treatment strategy. To analyse the possibilities to sensitise osteosarcoma cells to chemotherapy, we chose an unbiased functional genomics approach in which we performed siRNA library screening of kinase and kinase-associated genes on SaOS-2 osteosarcoma cells to detect critical survival kinases after cytotoxic treatment. Gene silencing of JNK-interacting protein 1 (JIP1) elicited the most potent sensitisation to doxorubicin. Using a small molecule JIP1-inhibitor we could confirm our findings in a panel of osteosarcoma cell lines. The observed sensitisation to doxorubicin treatment seemed to be partly dependent on p53 status, where osteosarcoma cells harbouring functional p53 showed a less pronounced sensitisation than osteosarcoma cells with mutated or absent p53. JIP1 is known to be a scaffold protein in the JNK-signalling pathway. It plays an important role in the activation of JNK signalling by assembling specific MAP kinases indispensable for JNK activation or for maintaining JNK phosphorylation.

112

Chapter 7

Therefore, downregulation of JIP1 leads to defective JNK signalling in response to cytotoxic damage. Protein expression analysis of JIP1 revealed that approximately two-thirds of human osteosarcoma samples express JIP1 and that patients with JIP1 positive tumours show a trend toward inferior overall survival. Thus, in this chapter we have successfully applied a functional genomics approach to identify a candidate drug target to exploit as a chemosensitiser to doxorubicin in osteosarcoma. Through downregulation of JIP1, the cellular balance is tilted further toward cell death leading to increased cell kill after doxorubicin treatment. Also here, the results provide a promising lead to pursue in the development of an innovative sensitising treatment for osteosarcoma. Many steps remain to be taken prior to realising clinical trials or clinical implementation of a JIP1-inhibitor for osteosarcoma patients. For example, it might be interesting to test whether JIP1 inhibition can induce sensitisation to other compounds commonly used in osteosarcoma treatment, such as cisplatinum or methotrexate, as well. At present, no such data is available. Also, in vivo validation of the observed increased anti-tumour effect is needed, as well as information on dose and timing of administration of the JIP1-inhibitor in relation to doxorubicin. Previous in vivo work on diabetes mellitus showed that the JIP1-inhibitor we used in our investigations can be safely administered to mice. [176] The in vivo administration of JIP1-inhibitor is currently also under investigation by a group that focuses on Parkinson’s disease. (www.michaeljfox.org/foundation) If the observed chemosensitisation can be verified in an in vivo model of osteosarcoma, phase I clinical trails might be designed to progress to toxicity and dose testing in osteosarcoma patients.

Apart from shifting molecular balances within osteosarcoma cells to favour cell death after radiation and chemotherapy, another strategy to improve treatment efficacy could be to enhance the intracellular delivery of currently applied drugs to the tumour cells. In the case of cancer-specific treatment by targeted delivery of drugs, therapy should ideally be targeted to a cell surface molecule that is specific for the tumour, i.e. that is highly expressed on the surface of tumour cells, but not on healthy tissues. The identification of suitable receptors for targeted delivery of therapeutics to osteosarcoma is essential in the design and development of novel targeted treatment strategies and is described in chapter 6. In order to identify such a suitable receptor we performed a comparative surface proteomic analysis of five osteosarcoma cell lines compared to three healthy human bone cell cultures. As surface proteins are difficult to solubilise and are of relatively low abundance compared to intracellular proteins, we first enriched our study samples for surface proteins by biotinylation and selective retrieval of the surface proteins prior to high resolution mass spectrometric analysis. From our proteomic analysis we obtained a comprehensive dataset containing many surface proteins of interest. We found the Ephrin type-A receptor 2 (EPHA2) to be the most abundant surface protein on osteosarcoma cells, with a 12-fold upregulation compared to the healthy human primary osteoblasts. We validated EPHA2 surface expression

113


7

and significant EPHA2 upregulation on osteosarcoma cell lines compared to human primary osteoblasts using FACS analysis and EPHA2 expression in human tumour tissue samples. Analysis of human osteosarcoma tissue micro arrays revealed that around 80% of the tumour samples present on the array showed EPHA2 expression. Patients with EPHA2-expressing tumours showed a trend toward inferior overall survival. As a proof-of-concept we performed binding and internalisation studies of EPHA2 using a specifically designed GFP-expressing adenovirus (AdYSA) that is targeted to EPHA2 via the YSA peptide in its fibre knob. In these studies we showed that EPHA2 can specifically mediate the uptake of AdYSA into osteosarcoma cells, whereas the human primary osteoblasts remained unaffected due to their low EPHA2 expression. The results presented in this chapter show that EPHA2 is a highly upregulated surface receptor on osteosarcoma cells that can be used for the specific intracellular uptake of vectors directed toward EPHA2. The use of EPHA2 as receptor for targeted delivery does not need to be restricted to adenoviral vectors alone. Other groups have shown the application of nanoparticles coated with YSA peptide to direct them to EPHA2-expressing cells. One group showed the application of magnetic nanoparticles that can capture EPHA2-expressing cells and then clear them from the circulation of tumour bearing mice. [237] Another group demonstrated the use of YSA coated nanoparticles to specifically deliver siRNA to EPHA2-expressing tumour cells to increase chemosensitivity in vivo. [236] In the case of osteosarcoma, the targeted delivery of treatment modalities using the EPHA2 needs to be tested in a pre-clinical animal model as well. Eventually, the treatment and the vehicle used to deliver therapy to osteosarcoma cells might be various. For example, perhaps in future research we can combine the knowledge obtained in this and the previous chapter and deliver a JIP1 inhibitor to osteosarcoma cells specifically through internalisation via EPHA2.

D I S C U S S I O NIn the recent past, so-called targeted therapy has gained tremendous interest in anti-cancer treatment, exploiting tumour-specific molecules for therapeutic goals. In this era of targeted treatments, the use of so-called ‘-omics’ approaches such as genomics, proteomics, and sequencing techniques allows the acquisition of vast amounts of information on tumour biology. It has allowed for the identification of crucial growth and survival genes and proteins in cancer cells; i.e. molecules that can be considered potential targets for targeted therapy. Thus, -omics approaches could be considered the cornerstone of the design of targeted anticancer therapy. Additionally, the advent of small molecule drugs that specifically target proteins or classes of proteins have further allowed the development of a very specific, or tailored, treatment of certain tumours based on the information provided by -omics techniques. Ultimately, tailored treatments should provide high anti-tumour efficacy

114

Chapter 7

combined with minimal toxicity for cancer patients. The use of -omics approaches can also be applied to identify cancer specific genetic profiles to predict therapy response or predictive or prognostic markers for clinical outcome. This can aid in treatment and dosing decisions and possible future stratification of patients to specific treatments that target a certain characteristic of their tumour and thus hopefully achieve improved treatment and survival outcomes.

However, both the investigational techniques for the identification of candidate drug targets and the development and implementation of novel drugs in the clinical setting is costly and cumbersome. The affordability of health care in general and cancer care in particular is a contemporary topic of political debate and even though it is clear that the development of novel treatment strategies for osteosarcoma is warranted, there are political and economical considerations to be made. Currently, it is estimated that the costs for anti-cancer medicine are approximately 1% of all health care spending. [239] As research techniques have become more sophisticated and high-tech, and the research questions surrounding cancer have become increasingly complex, the costs associated with the development of novel treatments are likely to increase as well. Furthermore, the value of the research and development of innovative treatments for small numbers of individuals is being questioned. For example, even though osteosarcoma is the most common primary bone tumour in children and adolescence, its incidence is extremely low (approximately 4/million/year). [1,2] In the Netherlands this would amount to only approximately 50-60 new patients each year. The combination of high expenditure in the development of novel treatments and only very few patients to be treated tends to lead to discussions about funding, pricing and costs of such anticancer treatments. [239] Multiple challenges exist surrounding the research and development of new treatment strategies for extremely rare diseases. Owing to its rarity, osteosarcoma is considered an orphan drug indication in oncology. One might state that there is no financial incentive for pharmaceutical companies to invest in the discovery and development of drugs for orphan indications in oncology because the expected returns are negligible. Therefore, legislation exists to encourage companies, universities and other research institutions to develop and manufacture treatments for orphan drug indications by for example granting the manufacturer market exclusivity for a specified period of time, granting a prolonged duration of patent, providing fiscal incentives for investors, as well as providing other forms of funding and support. Also, regulatory requirements for clinical trials might be adjusted for drug testing in orphan indications to reduce the complexity of executing a clinical trial, consequently reducing procedural costs as well, thus making drug development for these indications more attainable. [239,240]

Another issue could be that, after innovative (perhaps unregistered) treatments for orphan indications in oncology have reached clinical implementation, it is not unlikely that they will not be reimbursed, or only partially reimbursed by health insurance companies.

115


7

High pricing of the novel drugs could be the result of high development costs and the need to produce and supply only low volumes of the drug. Furthermore, if the efficacy of a novel treatment is not fully proven yet, insurers are likely to be reluctant to reimburse the costs for the treatment. Consequently these novel treatments may become (or remain) unavailable to those patients that cannot afford to pay for their own treatment. Policymakers, regulatory organs or governments in the EU can hopefully achieve a way to balance development costs, pricing, drug regulation and availability for novel treatments for orphan indications in oncology such as osteosarcoma, to ensure fair access for patients to the best possible therapy for their disease. [239,240]

A different difficulty in the development of new treatments for osteosarcoma is the accrual of sufficient patients to perform adequately powered clinical trials. It is appreciated worldwide that appropriately performed clinical trials provide the best evidence on safety and efficacy of new treatments. To collect data and material from large enough numbers of patients is both cumbersome, extremely time consuming and also potentially expensive. In the case of osteosarcoma, international collaborations have been formed through which a more extensive patient population becomes available for centralised research. The European and American Osteosarcoma Study Group (EURAMOS) is an important example. EURAMOS was founded in 2001 and is a collaboration of four other international study groups, namely the COSS (Collaborative German-Austrian-Swiss Osteosarcoma Study Group), the COG (Children’s Oncology Group, USA), the EOI (European Osteosarcoma Group) and the SSG (Scandinavian Sarcoma Group). Their ultimate aim is to improve the survival from osteosarcoma by conducting large randomised clinical trials, to undertake extensive parallel biological studies on tumour material obtained from patients enrolled in their trials and to eventually develop a common understanding and language about osteosarcoma worldwide. (www.ctu.mrc.ac.uk/euramos) Their first clinical trial (EURAMOS-1) has now been completed and first results have been presented. These results are beyond the scope of the material discussed here. Part of the secondary objectives of EURAMOS-1 was to assess the feasibility of such an international operation in performing clinical trials for osteosarcoma. The presentation of first results suggests that this in fact, is the case.

Apart from testing the safety and efficacy of novel treatments or drugs, the collection of adequate amounts of human tumour tissue samples is also a very important achievement that can be realised by international collaborations, especially when follow-up data on disease progression and survival are present and complete. Commonly, in osteosarcoma research, the absence of sufficient primary tumour material or primary cell cultures leads to the use of OS cell lines to perform in vitro experiments to study osteosarcoma biology. We ourselves have also done so. Recently, Mohseny et al. [241] published a report in which they studied the appropriateness of OS cell lines to study the human disease. They identified at least eight cell lines that are representative for the human disease that could be used as

116

Chapter 7

a model for osteosarcoma. Apart from this, various murine models exist (of both primary and metastatic osteosarcoma) to extrapolate in vitro data to an in vivo situation. [175,242] Although such pre-clinical models for osteosarcoma are suitable to study various aspects of osteosarcoma biology and behaviour, the study of human primary tumour tissue is absolutely indispensable. For example, protein expression levels of newly identified candidate drug targets need to be investigated in human tumour samples to validate their possible clinical relevance for patients suffering from osteosarcoma; this comprises an essential step from laboratory investigations to implementation of new strategies in the clinic. Apart from validation purposes, the collected human tissue samples and follow-up data could also be investigated to identify osteosarcoma specific predictive markers for therapy response or prognostic markers for survival outcome, neither of which currently exist for osteosarcoma. [7] Ultimately, prediction of treatment and/or survival outcomes may in the future allow for the stratification of patients to certain treatment schedules, thereby offering these patients a tailored therapy when it could be anticipated that they will benefit from specific targeted treatments.

As noted above, molecular profiling of tumours using -omics techniques allows the generation of extensive amounts of data on tumour biology. Apart from the identification of oncogenes, crucial survival genes, predictive or prognostic biomarkers, etc. these techniques have also unveiled the extent of tumour subtypes and/or tumour heterogeneity within tumour types. [243] Osteosarcoma is known to be a highly heterogeneous tumour that harbours complex chromosomal aberrancies and regularly displays chromosomal instability. [5,7,243] This heterogeneity, both between tumours in different individuals but particularly within one tumour and between a primary tumour and its metastases, has implications for biomarker discovery, response to (chemo)therapy and for the design of targeted therapy. In this sense, the tumour heterogeneity may have negative implications for the newly identified drug targets described in this thesis. Although we have found relatively high expression levels of JIP1 (67%) and EPHA2 (84%) in our tested osteosarcoma samples, verification of this expression in independent, preferably large datasets is very much needed to ascertain that indeed, these therapeutic targets are broadly expressed in osteosarcoma patients. It has been observed, in osteosarcoma as well as other types of cancer, that genomic or chromosomal instability can give rise to a genetic diversity of cells within one tumour that exhibit a different behaviour than the general cell population within that tumour. Apart from chromosomal instability, (miR-induced) post-translational modifications and epigenetic changes can lead to the development of different clonal populations within one tumour. After a cycle of treatment, resistant clones remain to give rise to tumour recurrence in a later stage of the disease. Thus, tumour heterogeneity can allow a tumour to adapt to extra-tumoural circumstances and give rise to, for example, therapy resistance but also other phenomena such as metastasis. [47,51,243] It is essential

117


7

to understand this heterogeneity and try to ascertain that newly defined drug targets are very broadly and preferably stably expressed among osteosarcomas and osteosarcoma metastases to be sure that most patients could potentially benefit from the newly designed therapy. If the tumour might ‘adapt’ to treatment, than perhaps so should the treatment schedule be adapted to the tumour behaviour. Among others, this is one reason why single agent therapy is very unlikely to be successful in heterogeneous tumours. It may eventually be necessary to switch from one compound to the other along the course of osteosarcoma treatment.

Contrary to this perhaps disturbing notion that -omics approaches revealed a heterogeneity within tumours which could potentially impede the design of one single targeted therapy for a certain type of cancer, whole genome sequencing, for example, can also provide valuable information on common gene mutations across different types of cancer. Recently, in the Netherlands, a Center for Personalised Cancer Treatment (CPCT) has been founded where patients suffering metastatic disease are included, biopsies from metastases are obtained and then subjected to ‘Next Generation Sequencing’ (NGS). Two thousand genes per tumour biopsy are analysed and thus the biopsies are screened for known predictive or prognostic markers and for novel biomarkers for therapy response and survival. The ultimate goal of this project is to provide each patient suffering from metastatic disease with a personalised treatment schedule, based on the molecular profile of their metastases and not the primary tumour. Also, patients can be enrolled in phase 1 clinical trials based on the genetic profile of their tumour rather than on the type of cancer that they suffer. Patients suffering from metastatic osteosarcoma are also enrolled in phase 1 clinical testing of novel compounds targeting their specific molecular profile. (www.cpct.nl) This formulation of pharmacogenetic profiles of tumours may eventually allow a better use of already existing anti-cancer drugs, where drugs initially developed and registered for other indications might be rationally administered to (and tested for efficacy in) patients suffering from osteosarcoma with a specific molecular profile. Although at present, the sequencing techniques are still costly and relatively time-consuming, in the future it may prove to be cost-effective to selectively stratify osteosarcoma patients to treatment with readily available anti-cancer compounds for other indications.

F U T U R E P E R S P E C T I V EThe research presented in this dissertation was mainly aimed at the identification of novel treatment targets in osteosarcoma to formulate strategies to improve the current treatment regimens. We speculate that therapy resistance in osteosarcoma cells is accountable for therapeutic failure and poor outcome and that research efforts for the design of novel treatment strategies should be aimed at overcoming or subverting this resistance. This

118

Chapter 7

can be achieved either by the selective targeting of intracellular proteins that grant the osteosarcoma cells a survival benefit after cytotoxic treatment, or by selectively targeting treatment modalities (be it conventional, small molecules or gene therapy) to osteosarcoma cells in order to obtain a higher effective dose at the site of the tumour whilst sparing normal cells and tissues.

By performing an siRNA library screening and mass spectrometry based proteomics we obtained two extensive datasets on intracellular regulators of doxorubicin response and osteosarcoma specific surface molecules, respectively. We have selected one candidate from each dataset for follow-up studies and present these as molecules to be exploited for the design of a new, targeted, treatment for osteosarcoma. In the case of EPHA2, the specific delivery of treatment modalities via this receptor can be realised by conjugating the particular modality with the YSA peptide which covalently binds to EPHA2, leading to receptor internalisation and thus the intracellular delivery of the YSA-coupled modality. The moieties delivered to osteosarcoma cells could encompass various entities. YSA could be conjugated to multiple modalities, for example liposome encapsuled chemotherapy or small molecules, but also gene therapy such as siRNAs, short hairpin RNAs or microRNAs, also encapsuled in a delivery vehicle of some sort.

siRNA library screening identified JIP1 (among others) as chemosensitiser to doxorubicin treatment in osteosarcoma. Validation studies with separate siRNA clones targeting the JIP1 gene verified that siRNA-mediated gene silencing led to a depletion of intracellular JIP1 with consequent increased cell death after cytotoxic treatment with doxorubicin. We then proceeded to use a small molecule JIP1-inhibitor for further combination treatment studies. The systemic administration of siRNA to treat tumours remains a challenge today and therefore the use of small molecule drugs that can selectively inhibit proteins or protein classes is an attractive method to use for in vivo and clinical studies following target identification. Whereas RNA interference is a powerful and efficient method for the systematic discovery of potential drug targets, the applicability of siRNAs in the clinical setting is still limited, mainly due to delivery issues. For example, the intracellular uptake of naked siRNA is hampered by its low membrane permeability. Also, proteases in the blood stream lead to enzymatic degradation, combined with rapid renal clearance results in a very short circulation half-life of siRNAs and therefore a limited biodistribution. Furthermore, siRNAs can elicit an interferon response and thus an undesirable induction of the immune system when systemically administered. [244,245] Notwithstanding delivery issues, the use of siRNA for targeted treatment has advantages over the use of small molecules. For example, siRNAs have extremely high target selectivity and therefore the chance of off-target effects are smaller compared to small molecules. Additionally, the design of siRNA sequences is relatively uncomplicated, rapid and potentially less expensive. [244] Coupling of siRNA to a targeting ligand can improve cell specific delivery of siRNAs, however, due to

119


7

their charge siRNAs cannot cross the cell membrane by diffusion and so a packaging modality remains necessary for the siRNA to reach the intracellular compartment. [244] Thus, prior to the successful application of siRNAs in targeted treatment in humans, the development of a delivery system that protects the siRNA against degradation and elimination from the bloodstream is needed.

Viral delivery of gene therapy to osteosarcoma cells has been successfully shown by our group in the past. [145,246,247] Viral vectors can also be adjusted to specifically infect tumour cells, an approach that we used for proof-of-principle studies in chapter 5. Viral vectors are reported to have a high transduction efficiency and have a profound capacity to deliver the gene therapy intracellularly, potentially granting high exogenous siRNA and thus ‘gene-silencing’ levels. Furthermore, they can be quickly produced in high titres. The use of (replication deficient) viruses in patients has been subject of discussion due to concerns surrounding safety, such as the possibility of virus induced insertional mutagenesis and immune and/or toxic reactions in the patient that have been observed in the past. Currently, the use of oncolytic viruses in patients has been proven safe, however, therapeutic efficacy remains relatively low due to delivery and tumoural uptake issues.

Nanotechnology is a newly developing technique that has evolved dramatically over the past decade and that may provide an important method to deliver therapeutics to tumour cells in the future. Nanoparticles can be lipid-, polymer- or peptide/protein-based and can be designed to form various different special configurations, depending on the requested properties of the particle, the route of administration and the moiety that it should encapsulate. It is reported that nanoparticles can also carry combinations of, for example, doxorubicin and siRNA, as well as small molecule inhibitors, leading to a ‘dual action’ particle where both cytotoxic agent and sensitiser can be delivered concomitantly. Furthermore, they can be coated with targeting ligands to stimulate a more targeted delivery of the nanoparticles to the tumour cells. Importantly, nanoparticles can be designed to be both biocompatible and -degradable and thus induce relatively little toxicity. [244,248,249] As mentioned above in the summary section, the use of the EPHA2 receptor for the specific targeting of EPHA2-expressing tumour cells with YSA-coated magnetic nanoparticles has been described to clear tumour cells from the circulation and peritoneal fluid of tumour-bearing mice, thus indicating that EPHA2 could be suitable as a targeting molecule for nanoparticles. [237]

Very recently, an in vitro study presented the successful delivery of doxorubicin and siRNA using biocompatible and biodegradable polymeric nanoparticles to osteosarcoma cells. The particles were loaded either with doxorubicin or siRNA and treatment of cells with doxorubicin-loaded nanoparticles led to increased cytotoxicity compared to doxorubicin treatment alone; treatment of cells with siRNA-loaded nanoparticles led to gene-silencing of the target gene with accompanying depletion of the protein product. [245]

120

Chapter 7

Thus, nanotechnology can be applied to osteosarcoma cells in laboratory investigations. Theoretically, the combined findings of our functional genomics and proteomics analyses could provide a recipe for the design of an EPHA2-directed, doxorubicin/siJIP1-carrying nanoparticle that could be used for the sensitisation of osteosarcoma cells to doxorubicin treatment. Alike, an EPHA2-directed nanoparticle carrying siWEE1 or small molecule WEE1-inhibitor might be used to achieve radiosensitisation in patients that need to undergo radiation therapy for their tumour.

At present, nanotechnology is still a developing field and therefore not (broadly) applied in clinical practice. The novelty and complexity is still high and with that the development costs also. It was noted that the entry of nanomedicines onto the market was delayed because there was little collaboration between research institutes and pharmaceutical companies to actually develop and produce the medicine. [248] As with any evolving technology it could be anticipated that as this discipline progresses further, the development and production costs of nanosystems for drug delivery will decrease and the use of this new technique will become a more realistic therapeutic entity in anticancer treatment somewhere in the future.

C O N C L U S I O NThe research presented in this dissertation was mainly focussed at the identification of novel therapeutic targets in osteosarcoma in order to formulate new, targeted treatment strategies to improve the current treatment regimens for this disease. Resistance to therapy remains an issue in the treatment of cancer, particularly in disseminated disease today. The results from our investigations provide leads (molecules) through which the cytotoxicity of existing therapies may be enhanced in osteosarcoma, ultimately leading to higher treatment efficacy, reduced toxicity and improved clinical outcomes for osteosarcoma patients. The exact development of targeted treatment based on our candidate molecules has yet to take form. We live in an era where growing understanding of tumour biology and rapid technological advancements in research platforms seem to provide endless opportunities in designing novel, targeted treatments for cancer. At the same time, the economical climate, policymakers, financiers, etc., ask for a certain cost-effectiveness in the development and clinical use of newly designed treatments. Therefore, we believe that in future osteosarcoma treatment smart use of available techniques and data is required. For example, -omics approaches may reveal common mutations in osteosarcoma and other tumours, such that compounds readily available for other indications can be applied in the treatment of osteosarcoma. Furthermore, in the case of osteosarcoma, its rarity poses difficulties in (pre-) clinical research and the design, development and production of novel treatments. Therefore, orchestrated international collaborations are of absolute importance in order to enable osteosarcoma research to develop innovative treatment strategies.

121


7

Eventually, application of -omics approaches on tumour material, international collaboration between research institutes and, perhaps most importantly, the determination of researchers and clinicians employed in this field, will be conducive to a new era of targeted treatments for osteosarcoma.

Chapter 8 Nederlandse samenvatting

124

Chapter 8

S A M E N VAT T I N GHet osteosarcoom is de meest voorkomende kwaadaardige bottumor op kinderleeftijd. De geschatte incidentie wereldwijd is 4 per miljoen per jaar met een piekincidentie rond het 15e tot 19e levensjaar en een tweede, echter aanzienlijke kleinere, na het 60e levensjaar. De tumor komt 1,4x vaker voor bij mannen dan bij vrouwen. In Nederland wordt jaarlijks de diagnose osteosarcoom bij ongeveer 40 tot 50 mensen gesteld. De etiologie is tot op heden onbekend. Het conventionele osteosarcoom ontstaat gewoonlijk uit de groeischijf (metafyse) van de lange pijpbeenderen; in regio’s van het skelet waar een grote lengtegroei wordt gerealiseerd. In het absolute merendeel van de gevallen (70%) ontstaan osteosarcomen rond de knie; in het distale femur of de proximale tibia, gevolgd door de proximale humerus (10%) en het bekken (7%). Op oudere leeftijd (>60) volgt het osteosarcoom een ander patroon en wordt dan vaker aangetroffen in het axiale skelet. Op deze leeftijd kan het osteosarcoom ontstaan zijn als een bijwerking na bestraling voor een eerdere maligniteit of in een deel van het skelet waarin er sprake is van een metabole botzieke (bijvoorbeeld Paget’s disease).

De symptomen waarmee het osteosarcoom zich presenteert zijn zeer aspecifiek, veelal relatief mild en kennen een sluipend ontstaan. Pijn is het meest gebruikelijke symptoom waarbij opmerkelijk is dat er ook pijn in rust en nachtelijke pijn is. Soms is een zwelling zichtbaar en een enkele keer kan er spontaan een (pathololgische) fractuur ontstaan. Hierdoor is er vaak sprake van zowel een ‘patient-delay’ als een ‘doctor-delay’ in het stellen van de diagnose. Op een conventionele röntgenopname kenmerkt het osteosarcoom zich als een expansieve laesie danwel botmassa ter plaatse van de tumor met een begeleidende periost-reactie en, in het geval van infiltratieve groei, calcificaties in de omliggende weke delen. Een aanvullende MRI-scan kan waardevolle informatie geven ten aanzien van de dimensies van de tumor, matrixvorming en de mate van weke-delen uitbreiding. De uiteindelijke diagnose wordt gesteld door de patholoog na weefselonderzoek van een biopt van de afwijking. De vorming van onrijp bot (osteoid), aanwezig zowel op röntgenopnames als bij histopathologisch onderzoek is pathognomonisch voor osteosarcoom.

Het osteosarcoom heeft een sterke neiging tot uitzaaien (metastasering). Circa 30% van de patiënten presenteert zich met metastasen ten tijde van de initiële diagnose en in nog eens 40% van de patiënten ontstaan er uitzaaiingen in een latere fase van hun ziekte. Het overgrote merendeel van de metastasen ontstaat in de longen (80%); de overige metastasen ontstaan vaak in andere skeletdelen. Metastasering naar andere organen is zeer zeldzaam. Patiënten met longmetastasen overlijden vaak aan respiratoire insufficiëntie. Botmetastasen zijn niet dodelijk, maar veroorzaken wel pijn en zijn vaak ernstig invaliderend voor de patiënt. Ook correleren botmetastasen met verminderde overlevingskansen. De standaardbehandeling voor patiënten met osteosarcoom bestaat uit neo-adjuvante chemotherapie (bestaand uit Doxorubicine, Cisplatinum, Methotrexaat en Ifosfamide), radicale chirurgische excisie van de primaire tumor en indien mogelijk alle metastasen,

125

Nederlandse samenvatting

8

gevolgd door adjuvante chemotherapie. Om curatief te kunnen behandelen is het essentieel om de primaire tumor radicaal te excideren. Radiotherapie behoort in principe niet tot de standaardbehandeling van het osteosarcoom, echter in sommige gevallen, zoals patiënten met irresectabele (axiale of hoofd-hals) tumoren, pijnlijke botmetastasen of bij patiënten die operatief ingrijpen weigeren, kan radiotherapie wel verbeterde overlevingskansen geven.

De 5-jaars overleving van patiënten met locaal osteosarcoom is ongeveer 65%. Echter, in het geval van uitzaaiing of terugval (recurrence) is de prognose veel slechter en ligt de 5-jaars overleving rond de 20%. Verreweg de meeste patiënten die overlijden doen dat als gevolg van uitgezaaide ziekte. Ondanks de agressieve therapie die nu gegeven wordt voor deze aandoening, zijn de overlevingskansen matig. Verhoging van intensiteit en doseringen van chemotherapeutische behandeling hebben verbeterde overleving opgeleverd. Een groot probleem, en mogelijk ook de verklaring voor het falen van huidige therapieën in de behandeling van het osteosarcoom, is de relatieve ongevoeligheid (ook wel resistentie) voor de therapieën die nu gegeven worden. Aan resistentie tegen therapie kunnen velerlei mechanismen ten grondslag liggen. De meeste chemotherapie en ook radiotherapie zijn erop gericht om het DNA in de cel te beschadigen. Deze therapieën worden ook wel cytotoxisch genoemd. De aanwezigheid van beschadigd DNA, en met name celdeling in de aanwezigheid van beschadigd DNA, leidt tot geprogrammeerde celdood (apoptose). Sneldelende cellen (dus bij uitstek tumorcellen) zijn over het algemeen meer gevoelig voor DNA schade dan gezonde cellen. De tumorcel met beschadigd DNA kan echter verschillende mechanismen aanwenden om aan apoptose te ontsnappen. Voorbeelden hiervan zijn stoppen met celdelen, toegenomen DNA herstel of veranderingen in de balans van apoptose/anti-apoptose eiwitten. De eiwitten die hiervoor verantwoordelijk zijn kunnen mogelijk als nieuw therapeutisch doelwit gelden. Door deze moleculen te behandelen zou de gevoeligheid van het osteosarcoom voor de conventionele therapieën verbeterd kunnen worden. Deze moleculen benoemen we hier als zogenaamde ‘sensitisers’. Het uiteindelijke doel van het onderzoek beschreven in dit proefschrift is dan ook om nieuwe therapeutische doelwitten (ook wel ‘targets’) voor het osteosarcoom te identificeren en te testen op hun geschiktheid als zogenaamde ‘sensitiser’.

Het is opmerkelijk dat de metastasen van het osteosarcoom minder goed reageren op chemotherapie dan de primaire tumor, waardoor de uitzaaiingen moeilijker te behandelen zijn. Cellen die kunnen metastaseren moeten beschikken over verschillende eigenschappen die het ze mogelijk maakt om te overleven in omstandigheden waaraan tumorcellen in een primaire tumor niet worden blootgesteld. Zo kan een metastaserende cel bijvoorbeeld overleven (in de bloedbaan) zonder cel-cel contacten en zonder overlevingssignalen uit de extracellulaire matrix, waaraan cellen in weefsels normaliter wel zijn blootgesteld. Deze zelfde eigenschappen kunnen de gemetastaseerde tumorcel ook een overlevingsvoordeel opleveren na behandeling met cytotoxische therapie. In hoofdstuk 2 geven we een

126

Chapter 8

uiteenzetting van de afzonderlijke stappen die een metastaserende cel moet afleggen om tot een uitzaaiing te komen, we bespreken de cellulaire eigenschappen en moleculaire veranderingen die aan elk proces ten grondslag liggen en bediscussiëren hoe deze moleculaire veranderingen mogelijk als therapeutisch target kunnen worden uitgebuit.

In hoofdstuk 3 volgt een tweede review artikel waarin de nadruk meer ligt op de cellulaire mechanismen van resistentie voor therapie en de moleculen en intracellulaire signalen (ook wel ‘signalling pathways’) die hieraan bijdragen. Uit deze twee eerste hoofdstukken blijkt dat ondanks een veelheid aan (pre)klinisch onderzoek naar het osteosarcoom, metastasering en therapieresistentie, er weinig nieuwe middelen de kliniek hebben bereikt. Het uitdiepen en begrijpen van de biologische processen waardoor resistentie ontstaat is van belang om deze resistentie te bestrijden en ligt aan de basis van het ontwerpen van nieuwe therapeutische strategieën waardoor de effectiviteit van huidige behandelingen verbeterd kan worden.

Het onderzoek dat is uitgevoerd ter totstandkoming van dit proefschrift was gericht op de identificatie van nieuwe therapeutische ‘targets’, die kunnen dienen als ‘sensitiser’ van het osteosarcoom voor bestaande therapieën. De laatste decennia is er een groeiende trend naar zogenaamde ‘targeted therapy’, of ‘therapie op maat’ voor patiënten met tumoren. Behalve dat het mogelijk is om de biologische processen te ontrafelen die zorgen voor de ontwikkeling, groei en overleving van tumoren, zijn er ook nieuwe soorten medicijnen ontwikkeld zoals (monoclonale) antilichamen en ‘small molecule drugs’ die aangrijpen op één bepaald eiwit, één klasse eiwitten of op ‘signalling pathway(s)’ en dus een zeer specifieke therapeutische werking hebben. Deze medicijnen kunnen gebruikt worden om specifieke moleculen in een tumor te behandelen. Door het combineren van deze ‘small molecule drugs’ en bestaande DNA-beschadigende (cytotoxische) therapieën kan de gevoeligheid van de tumor worden vergroot en de effectiviteit van de therapie worden verbeterd.

In hoofdstuk 4 beschrijven wij onderzoek naar de mogelijkheid om het osteosarcoom gevoelig te maken voor radiotherapie door een bekende ‘radiosensitiser’ voor andere tumoren, het WEE1 eiwit, op werkzaamheid te testen in osteosarcoomcellen. Radiotherapie wordt gewoonlijk niet toegepast in de behandeling van patiënten met osteosarcoom, echter, voor patiënten met irresectabele tumoren (veelal tumoren in het axiale skelet of in het hoofd-halsgebied), patiënten met pijnlijke botmetastasen en patiënten die chirurgische behandeling weigeren, kan bestraling toegevoegde waarde hebben in hun behandeling, zij het met beperkte effectiviteit. Derhalve bestudeerden wij de mogelijkheid om de effectiviteit van radiotherapie te verbeteren. Uit studies op andere tumoren was reeds gebleken dat remming van het WEE1 eiwit de gevoeligheid van deze tumoren voor bestraling doet toenemen. Door met een ‘small molecule drug’ het WEE1 eiwit te remmen in osteosarcoomcellen, werden deze cellen meer gevoelig voor radiotherapie, hetgeen zich vertaalde in toegenomen celdood in cellen die tegelijk werden behandeld met WEE1-inhibitor en bestraling, vergeleken met cellen die alleen werden bestraald. Na bestraling

127


8

ontstaat DNA schade in de osteosarcoomcellen. Als reactie hierop stoppen de cellen met het doorlopen van de celcyclus in het G2-checkpoint (een zg. G2-arrest), ten einde extra tijd te creëren voor DNA herstel voordat celdeling plaats kan vinden. Op deze manier kunnen de cellen aan celdood ontsnappen. WEE1 speelt een belangrijke rol in het G2-arrest doordat het CDC2 fosforyleert, dat vervolgens niet kan binden aan het celcyclus eiwit Cyclin-B en uiteindelijk een cel cyclus arrest in het G2-checkpoint tot gevolg heeft. Als het WEE1 eiwit geremd wordt, kan de osteosarcoomcel geen G2-arrest bewerkstelligen waardoor de cel gedwongen wordt om te delen in aanwezigheid van DNA schade, met als gevolg apoptose. Apoptose die specifiek plaatsvindt tijdens de celdeling wordt ook wel mitotic catastrophe genoemd. WEE1 genen eiwitexpressie werd respectievelijk in silico en immunohistochemisch onderzocht. Hieruit bleek dat WEE1 eiwit verhoogd tot expressie komt in het osteosarcoom vergeleken met gezonde weefsels en dat WEE1 zowel in primaire tumoren als in metastasen tot expressie komt. Dit betekent dat patiënten met osteosarcoom baat zouden kunnen hebben bij additionele behandeling met een WEE1 remmer tijdens radiotherapie. Voordat het gebruik van WEE1 inhibitie een plaats kan hebben in de klinische behandeling van osteosarcoom dienen eerst in vivo experimenten uitgevoerd te worden om te verifiëren dat WEE1 remming inderdaad kan leiden tot sensitisatie van het osteosarcoom voor radiotherapie. Een bijkomend gegeven dat de weg van ‘bench to bedside’ eventueel kan vergemakkelijken voor WEE1 inhibitie bij patiënten met osteosarcoom, is dat ‘small molecule’ WEE1 remmers (inhibitors) reeds getest zijn in phase 1 klinische trials voor andere oncologische indicaties en er dus al gegevens beschikbaar zijn omtrent veiligheid en dosering. Ondanks deze beschikbare data zal het in de kliniek testen van WEE1 inhibitie in combinatie met radiotherapie bij patiënten met irresectabel osteosarcoom zeer moeizaam zijn, met name gezien het zeer beperkte aanbod patiënten die geschikt zijn voor deelname aan een dergelijke klinische trial.

In het volgende hoofdstuk (hoofdstuk 5) vestigen we de aandacht op zogenaamde chemosenstitisatie van het osteosarcoom voor doxorubicine behandeling. We stellen dat het behandelen van essentiële ‘survival’ eiwitten (d.w.z. eiwitten die overlevingssignalen doorgeven in de cel en dus essentieel zijn voor het overleven van de tumorcel na het optreden van DNA schade) tegelijkertijd met het toedienen van chemotherapie, kan leiden tot toegenomen celdood na chemotherapeutische behandeling. De identificatie van essentiële ‘survival’ eiwitten is van cruciaal belang in het ontwerp van een dergelijke behandelstrategie. Ter identificatie van een essentieel overlevingseiwit en dus ‘sensitiser’ voor doxorubicine voerden wij een (unbiased) siRNA screen uit op SaOS-2 osteosarcoomcellen waarin wij systematisch alle bekende kinase- en ‘kinase-associated’-genen uitschakelden om vervolgens het therapeutisch effect van doxorubicine te analyseren. Hieruit bleek dat het uitschakelen van het JIP1 gen, en daarmee uitschakelen van het JIP1 eiwit, de meest sterke toename in celdood na doxorubicine behandeling tot gevolg heeft. Remming van het JIP1 eiwit met

128

Chapter 8

een ‘small molecule drug’ valideerden de resultaten die verkregen waren op genetisch (RNA) niveau. Deze resultaten konden ook bevestigd worden in een panel osteosarcoom cellijnen, terwijl gezonde botcellen onaangedaan bleven. Dit betekent dat JIP1 remming een tumorspecifieke behandeling kan zijn die in tumorcellen toegenomen celdood geeft terwijl gezonde cellen relatief gespaard blijven. De gevonden toename in celdood na doxorubicine behandeling bleek afhankelijk te zijn van de p53 status in de onderzochte cellen, waarbij in cellen met een functioneel p53 eiwit een minder sterke sensitisatie voor doxorubicine verkregen werd.

De functie van het JIP1 eiwit is, om na het optreden van DNA schade, een complex te vormen met verschillende andere eiwitten die zorgen voor de fosforylering en daarmee activering van het JNK eiwit. JNK is een eiwit dat anti-apoptose signalen doorgeeft in de cel en dus resistentie tegen celdood na DNA schade kan bewerkstelligen. Door het uitschakelen van JIP1 kan JNK niet geactiveerd worden en zal er dus meer apoptose optreden na behandeling met het cytotoxische doxorubicine. Uit eiwitexpressie analyses op ‘tissue micro arrays’ uit het LUMC bleek dat het JIP1 eiwit tot expressie komt in ongeveer tweederde van de onderzochte tumoren en dus een klinisch relevant drug target kan zijn voor patiënten met osteosarcoom. De aanwezigheid van JIP1 expressie liet sterke trend naar slechtere overleving van de patiënt zien. Samengevat werd met behulp van een functional genomics (siRNA) screen, JIP1 geïdentificeerd als essentieel ‘survival’ eiwit in het osteosarcoom. Remming van JIP1 met een ‘small molecule drug’ leidt tot toegenomen celdood na doxorubicine behandeling doordat er geen JNK-signalling kan optreden en er dus een intracellulaire balans ten faveure van apoptose ontstaat. Aangezien dit effect beperkt blijft tot tumorcellen en niet optreedt in gezonde botcellen, én het merendeel van de onderzochte osteosarcomen JIP1 tot expressie brengt, kan JIP1 worden aangemerkt als een tumorspecifiek therapeutisch target dat, wanneer gelijktijdig JIP1 inhibitie en doxorubicine wordt gegeven, mogelijk in een merendeel van de patiënten een verbeterde effectiviteit van de huidige behandeling zou kunnen geven. Ook hier geldt dat, voordat JIP1 als innovatief therapeutisch target verder ontwikkeld kan worden voor klinisch gebruik in osteosarcoompatiënten, er nog verscheidene experimentele stappen resten. De eerste stap is dierexperimenteel onderzoek ter validatie van de in vitro verkregen resultaten. In het geval dat de sensitisatie voor doxorubicine ook in een in vivo setting kan worden geobjectiveerd zullen aanvullende studies omtrent dosering en timing van toediening van de JIP1 remmer moeten plaatsvinden. Uit voorgaand dierexperimenteel onderzoek naar diabetes mellitus is gebleken dat de door ons gebruikte ‘small molecule JIP1 inhibitor’ veilig kan worden toegediend aan muizen. Een andere groep die zich toelegt op onderzoek naar de ziekte van Parkinson onderzoekt op dit moment eveneens de veiligheid van toediening van JIP inhibitor aan muizen. Wanneer de chemosensitisatie van het osteosarcoom voor doxorubicine door JIP1 remming inderdaad aangetoond kan worden in in vivo experimenten, dan kan dit onderzoek mogelijk voortgezet

129


8

worden naar een phase 1 klinische trial waarin toxiciteit en dosis getest kunnen worden bij patiënten met osteosarcoom.

In hoofdstuk 4 en 5 hebben wij intracellulaire therapeutische targets onderzocht die de gevoeligheid van het osteosarcoom voor radiotherapie en chemotherapie kunnen doen toenemen door middel van geforceerde progressie door de celcyclus en verschuivingen in apoptotische signalen. Echter, sensitisatie kan ook bewerkstelligd worden door extracellulaire ‘targeting’ van het osteosarcoom, hetgeen wordt beschreven in hoofdstuk 6. Deze strategie richt zich op het specifiek afleveren van chemotherapie ter plaatse van de tumor (‘drug targeting’ of ‘targeted delivery’) door gebruik te maken van oppervlakte eiwitten (receptoren) die hoog tot expressie komen op de tumor cellen en maar weinig voorkomen op de oppervlakte van gezonde cellen. Op deze manier kan er ter plaatse van de tumor een hoge concentratie van de chemotherapie bereikt worden terwijl gezonde weefsels relatief gespaard blijven. Ook hier is voor het ontwikkelen van een nieuwe behandelstrategie de identificatie van geschikte doelwitten van essentieel belang. Ter identificatie van een receptor op het osteosarcoom die kan dienen als doelwit voor ‘targeted drug delivery’ hebben wij een proteomics analyse uitgevoerd waarin de oppervlakte eiwitten van vijf osteosarcoom cellijnen en drie osteoblast (gezonde bot) cellijnen met elkaar zijn vergeleken. Allereerst werden de oppervlakte eiwitten gescheiden van overige intracellulaire eiwitten en vervolgens werd een massa spectrometrie analyse uitgevoerd. Middels deze techniek werd een uitgebreide dataset verkregen waaruit bleek dat EPHA2 de meest voorkomende receptor is op osteosarcoomcellen, en 12x hoger tot expressie komt op osteosarcoomcellen dan op gezonde botcellen. De oppervlakte expressie van EPHA2 op osteosarcoomcellen werd bevestigd met behulp van FACS analyse. EPHA2 expressie op osteosarcoom weefsel van patiënten werd bevestigd middels immunohistochemie op osteosarcoom samples uit ons eigen archief en op ‘tissue micro arrays’ uit het Leids Universitair Medisch Centrum (LUMC). Nadat oppervlakte expressie van de EPHA2 receptor op osteosarcoomcellen en osteosarcoom weefsel was bevestigd, onderzochten wij of EPHA2 geschikt is voor de specifieke intracellulaire opname van middelen via deze receptor. Hiertoe werd gebruik gemaakt van een gemodificeerd adenovirus, waarbij in de fibre knop (het uitsteeksel waarmee een virus aan een cel kan binden) het YSA peptide is geïncorporeerd. Het YSA peptide kan covalent binden aan de EPHA2 receptor, waarna de receptor tezamen met het gebonden virus geïnternaliseerd wordt door de cel. Uit deze zogenaamde internalisatie studies bleek dat osteosarcoomcellen met veel EPHA2 expressie gemakkelijk geïnfecteerd werden door het specifieke virus terwijl in de gezonde botcellen geen opname van het virus werd gezien. Hieruit concluderen wij dat de EPHA2 receptor gebruikt kan worden voor targeted delivery van specifiek EPHA2-gerichte entiteiten, en dus uiteindelijk voor het gericht afleveren van chemotherapie naar de tumorcellen. Uit de eiwitexpressie analyse bleek dat het overgrote merendeel van de door ons onderzochte

130

Chapter 8

osteosarcoom samples EPHA2 tot expressie brengt (80%). EPHA2 expressie werd gevonden in zowel primaire tumoren, recidieven en metastasen, en EPHA2 expressie was significant hoger in osteosarcoom vergeleken met gezond botweefsel. Patiënten met osteosarcomen die EPHA2 tot expressie brengen laten een trend van slechtere overleving zien. Samenvattend wijst dit onderzoek uit dat EPHA2 een relevante oppervlakte receptor is voor osteosarcoompatiënten die, gegeven zijn significant verhoogde expressie op het celoppervlak en hoge opnamecapaciteit, gebruikt kan worden voor de gerichte aflevering en opname van therapeutische entiteiten naar osteosarcoomcellen waardoor een hogere effectieve dosis ter plaatse van de tumor verkregen kan worden terwijl gezonde cellen relatief gespaard blijven. De werkzaamheid van ‘targeted delivery’ naar osteosarcoomcellen zal evengoed in een dierexperimenteel model gevalideerd moeten worden. Uiteindelijk zal de EPHA2 receptor verscheidene mogelijkheden kunnen bieden in een nieuwe behandelstrategie voor het osteosarcoom. Het type therapeutische entiteit welke via EPHA2 in osteosarcoomcellen kan worden afgeleverd hoeft zich niet te beperken tot virale vectoren, maar zou ook kunnen bestaan uit bijvoorbeeld liposomale chemotherapeutica, nanoparticles geladen met siRNA, shRNA of miRs. Andere onderzoeksgroepen hebben zowel in laboratoriumonderzoek als in dierexperimenteel onderzoek het YSA peptide gebruikt om bijvoorbeeld nanoparticles af te leveren aan tumorcellen om zo tumorcellen gevoeliger te maken voor chemotherapie. Naast de gerichte aflevering van bijvoorbeeld conventionele chemotherapie, zou de EPHA2 receptor in het osteosarcoom ook voor een soortgelijk doeleinde gebruikt kunnen worden.

Samengevat stellen wij dat de teleurstellende behandeluitkomsten voor patiënten met osteosarcoom kunnen worden toegeschreven aan de relatieve ongevoeligheid, ofwel resistentie, van de osteosarcoomcellen voor de huidige conventionele behandelingen. Initiatieven ter ontwerp van innovatieve behandelstrategieën zouden gericht moeten zijn op de mechanismen van resistentie ten einde deze te kunnen omverwerpen danwel omzeilen. Dit zou bewerkstelligd kunnen worden door intracellulaire targeting van essentiële overlevingsmechanismen van de osteosarcoomcel gelijktijdig met de toediening van conventionele behandeling, óf door ‘targeted delivery’ van chemotherapie of sensitisers naar de osteosarcoomcellen zodat er een hoge therapeutische dosis ter plaatse van de tumor komt terwijl gezonde weefsels relatief gespaard blijven. Wij bevestigden de werkzaamheid van WEE1 remming voor radiosensitisatie in osteosarcoomcellen. Daarnaast identificeerden wij door middel van een siRNA screen en een proteomics analyse zowel een intracellulair (JIP1) als een extracellulair (EPHA2) therapeutisch doelwit. Uiteindelijk zou, op basis van de gegevens opgedaan in dit proefschrift, toekomstige osteosarcoom therapie kunnen bestaan uit het combineren van ‘small molecule inhibitors’ van JIP1 en WEE1, gecombineerd met respectievelijk chemo- en radiotherapie. Toekomstig onderzoek zou ook gericht kunnen zijn op het ontwikkelen van een nanoparticle dat specifiek wordt gestuurd en afgeleverd aan de

131


8

EPHA2 receptor en dat bijvoorbeeld een WEE1 remmer, doxorubicine en/of doxorubicine + een JIP1 remmer bij zich draagt. Ondanks dat nanotechnologie nog een jong veld van onderzoek is, is het de laatste jaren volop in ontwikkeling en kan het in de toekomst mogelijk een belangrijke rol in de behandeling van tumoren gaan spelen. ‘Nanomedicine’ geeft, net zoals de ‘small molecule drugs’ die specifieke eiwitten en klassen eiwitten behandelen, de mogelijkheid om therapie op maat te verschaffen aan patiënten met een osteosarcoom, afhankelijk van het type therapie dat het nanoparticle bij zich draagt en aflevert.

In het huidige tijdperk van onderzoek naar maligne tumoren en het ontwerpen van zogenaamde therapie op maat is het ontrafelen van tumorbiologie van toenemend belang. De therapeutische ‘targets’ die in dit werk zijn geïdentificeerd en gevalideerd kunnen mogelijk een bijdrage leveren aan het ontwerp van een innovatieve behandelstrategie voor het osteosarcoom gericht op het doorbreken van therapieresistentie.

Gegeven de complexiteit van de tumorbiologie en de extreme zeldzaamheid van het osteosarcoom is het voor wetenschappelijk onderzoek naar (en uiteindelijk de behandeling van) deze tumor van belang om (internationale) samenwerkingsverbanden te realiseren. Deze geven de mogelijkheid om grotere hoeveelheid tumormateriaal te verzamelen voor bijvoorbeeld genen eiwitexpressie analyse of voor onderzoek naar klinische uitkomstmaten, maar ook bijvoorbeeld voor het optimaal gebruik van onderzoeks- en ontwikkeltechnieken door instituten die zich met onderzoek naar het osteosarcoom bezig houden. In het geval van klinische trials, zouden internationale studieverbanden ook kunnen voorzien in de werving van grote aantallen patiënten.

De gezamenlijke inzet van wetenschappers en artsen wereldwijd zal uiteindelijk meer en meer inzicht geven in de biologische processen die essentieel zijn voor therapieresistentie van het osteosarcoom en door deze te doorbreken zullen we uiteindelijk betere behandelresultaten en overleving bereiken voor patiënten die lijden aan een osteosarcoom.

8

Reference list

134

R E F E R E N C E S

1. Bielack SS, Kempf-Bielack B, Delling G et al (2002) Prognostic factors in high-grade osteosarcoma of the extremities or trunk: an analysis of 1,702 patients treated on neoadjuvant cooperative osteosarcoma study group protocols. J Clin Oncol 20(3):776-790

2. Mirabello L, Troisi RJ, Savage SA (2009) Osteosarcoma incidence and survival rates from 1973 to 2004: data from the Surveillance, Epidemiology, and End Results Program. Cancer 115(7):1531-1543

3. Hayden JB, Hoang BH (2006) Osteosarcoma: basic science and clinical implications. Orthop Clin North Am 37(1):1-7

4. Ta HT, Dass CR, Choong PF et al (2009) Osteosarcoma treatment: state of the art. Cancer Metastasis Rev 28(1-2):247-263

5. Hughes DP (2009) Strategies for the targeted delivery of therapeutics for osteosarcoma. Expert Opin Drug Deliv 6(12):1311-1321

6. Bacci G, Ferrari S, Bertoni F et al (2001) Histologic response of high-grade nonmetastatic osteosarcoma of the extremity to chemotherapy. Clin Orthop Relat Res (386):186-196

7. Gorlick R, Anderson P, Andrulis I et al (2003) Biology of childhood osteogenic sarcoma and potential targets for therapeutic development: meeting summary. Clin Cancer Res 9(15):5442-5453

8. Schajowicz F, Sissons HA, Sobin LH (1995) The World Health Organization’s histologic classification of bone tumors. A commentary on the second edition. Cancer 75(5):1208-1214

9. Both J, Wu T, Bras J et al (2012) Identification of novel candidate oncogenes in chromosome region 17p11.2-p12 in human osteosarcoma. PLoS One 7(1):e30907

10. Atiye J, Wolf M, Kaur S et al (2005) Gene amplifications in osteosarcoma-CGH microarray analysis. Genes Chromosomes Cancer 42(2):158-163

11. Berman SD, Calo E, Landman AS et al (2008) Metastatic osteosarcoma induced by inactivation of Rb and p53 in the osteoblast lineage. Proc Natl Acad Sci U S A 105(33):11851-11856

12. Fuchs B, Pritchard DJ (2002) Etiology of osteosarcoma. Clin Orthop Relat Res (397):40-52

13. Freeman SS, Allen SW, Ganti R et al (2008) Copy number gains in EGFR and copy number losses in PTEN are common events in osteosarcoma tumors. Cancer 113(6):1453-1461

14. Wunder JS, Gokgoz N, Parkes R et al (2005) TP53 mutations and outcome in osteosarcoma: a prospective, multicenter study. J Clin Oncol 23(7):1483-1490

15. Blay JY (2007) Chemotherapy for osteosarcoma without high-dose methotrexate: another piece in the puzzle. Onkologie 30(5):226-227

16. Lewis IJ, Nooij MA, Whelan J et al (2007) Improvement in histologic response but not survival in osteosarcoma patients treated with intensified chemotherapy: a randomized phase III trial of the European Osteosarcoma Intergroup. J Natl Cancer Inst 99(2):112-128

17. Smeland S, Wiebe T, Bohling T et al (2004) Chemotherapy in osteosarcoma. The Scandinavian Sarcoma Group experience. Acta Orthop Scand Suppl 75(311):92-98

18. Anderson PM (2003) Effectiveness of radiotherapy for osteosarcoma that responds to chemotherapy. Mayo Clin Proc 78(2):145-146

19. Bielack SS, Carrle D, Hardes J et al (2008) Bone tumors in adolescents and young adults. Curr Treat Options Oncol 9(1):67-80

20. Bacci G, Rocca M, Salone M et al (2008) High grade osteosarcoma of the extremities with lung metastases at presentation: treatment with neoadjuvant chemotherapy and simultaneous resection of primary and metastatic lesions. J Surg Oncol 98(6):415-420

21. Harting MT, Blakely ML (2006) Management of osteosarcoma pulmonary metastases. Semin Pediatr Surg 15(1):25-29

22. Kager L, Zoubek A, Potschger U et al (2003) Primary metastatic osteosarcoma: presentation and outcome of patients treated on neoadjuvant Cooperative Osteosarcoma Study Group protocols. J Clin Oncol 21(10):2011-2018

23. Messerschmitt PJ, Garcia RM, bdul-Karim FW et al (2009) Osteosarcoma. J Am Acad Orthop Surg 17(8):515-527

135

8

24. Fuchs J, Seitz G, Ellerkamp V et al (2008) Analysis of sternotomy as treatment option for the resection of bilateral pulmonary metastases in pediatric solid tumors. Surg Oncol 17(4):323-330

25. Krishnan K, Khanna C, Helman LJ (2005) The biology of metastases in pediatric sarcomas. Cancer J 11(4):306-313

26. Eccles SA, Welch DR (2007) Metastasis: recent discoveries and novel treatment strategies. Lancet 369(9574):1742-1757

27. Pradelli E, Karimdjee-Soilihi B, Michiels JF et al (2009) Antagonism of chemokine receptor CXCR3 inhibits osteosarcoma metastasis to lungs. Int J Cancer 125(11):2586-2594

28. Rodriguez NI, Hoots WK, Koshkina NV et al (2008) COX-2 expression correlates with survival in patients with osteosarcoma lung metastases. J Pediatr Hematol Oncol 30(7):507-512

29. Briccoli A, Rocca M, Salone M et al (2010) High grade osteosarcoma of the extremities metastatic to the lung: long-term results in 323 patients treated combining surgery and chemotherapy, 1985-2005. Surg Oncol 19(4):193-199

30. deNigris F., Rossiello R, Schiano C et al (2008) Deletion of Yin Yang 1 protein in osteosarcoma cells on cell invasion and CXCR4/angiogenesis and metastasis. Cancer Res 68(6):1797-1808

31. Diaz-Montero CM, Wygant JN, McIntyre BW (2006) PI3-K/Akt-mediated anoikis resistance of human osteosarcoma cells requires Src activation. Eur J Cancer 42(10):1491-1500

32. Guo Y, Rubin EM, Xie J et al (2008) Dominant negative LRP5 decreases tumorigenicity and metastasis of osteosarcoma in an animal model. Clin Orthop Relat Res 466(9):2039-2045

33. Hingorani P, Zhang W, Gorlick R et al (2009) Inhibition of Src phosphorylation alters metastatic potential of osteosarcoma in vitro but not in vivo. Clin Cancer Res 15(10):3416-3422

34. Kansara M, Thomas DM (2007) Molecular pathogenesis of osteosarcoma. DNA Cell Biol 26(1):1-18

35. Kansara M, Tsang M, Kodjabachian L et al (2009) Wnt inhibitory factor 1 is epigenetically silenced in human osteosarcoma, and targeted disruption accelerates osteosarcomagenesis in mice. J Clin Invest 119(4):837-851

36. Khanna C, Wan X, Bose S et al (2004) The membrane-cytoskeleton linker ezrin is necessary for osteosarcoma metastasis. Nat Med 10(2):182-186

37. Kim SY, Lee CH, Midura BV et al (2008) Inhibition of the CXCR4/CXCL12 chemokine pathway reduces the development of murine pulmonary metastases. Clin Exp Metastasis 25(3):201-211

38. Lafleur EA, Koshkina NV, Stewart J et al (2004) Increased Fas expression reduces the metastatic potential of human osteosarcoma cells. Clin Cancer Res 10(23):8114-8119

39. Pasello M, Michelacci F, Scionti I et al (2008) Overcoming glutathione S-transferase P1-related cisplatin resistance in osteosarcoma. Cancer Res 68(16):6661-6668

40. Ren L, Hong SH, Cassavaugh J et al (2009) The actin-cytoskeleton linker protein ezrin is regulated during osteosarcoma metastasis by PKC. Oncogene 28(6):792-802

41. Wang IC, Chen YJ, Hughes DE et al (2008) FoxM1 regulates transcription of JNK1 to promote the G1/S transition and tumor cell invasiveness. J Biol Chem 283(30):20770-20778

42. Zucchini C, Rocchi A, Manara MC et al (2008) Apoptotic genes as potential markers of metastatic phenotype in human osteosarcoma cell lines. Int J Oncol 32(1):17-31

43. Kawabe T (2004) G2 checkpoint abrogators as anticancer drugs. Mol Cancer Ther 3(4):513-519

44. Nyberg KA, Michelson RJ, Putnam CW et al (2002) Toward maintaining the genome: DNA damage and replication checkpoints. Annu Rev Genet 36:617-656

45. Reinhardt HC, Aslanian AS, Lees JA et al (2007) p53-deficient cells rely on ATM- and ATR-mediated checkpoint signaling through the p38MAPK/MK2 pathway for survival after DNA damage. Cancer Cell 11(2):175-189

46. Syljuasen RG, Jensen S, Bartek J et al (2006) Adaptation to the ionizing radiation-induced G2 checkpoint occurs in human cells and depends on checkpoint kinase 1 and Polo-like kinase 1 kinases. Cancer Res 66(21):10253-10257

47. Hanahan D, Weinberg RA (2000) The hallmarks of cancer. Cell 100(1):57-70

48. Kashima T, Nakamura K, Kawaguchi J et al (2003) Overexpression of cadherins suppresses pulmonary metastasis of osteosarcoma in vivo. Int J Cancer 104(2):147-154

49. Steeg PS (2006) Tumor metastasis: mechanistic insights and clinical challenges. Nat Med 12(8):895-904

50. Worth LL, Lafleur EA, Jia SF et al (2002) Fas expression inversely correlates with metastatic potential in

136

osteosarcoma cells. Oncol Rep 9(4):823-827

51. Hynes RO (2003) Metastatic potential: generic predisposition of the primary tumor or rare, metastatic variants-or both? Cell 113(7):821-823

52. Khanna C, Khan J, Nguyen P et al (2001) Metastasis-associated differences in gene expression in a murine model of osteosarcoma. Cancer Res 61(9):3750-3759

53. Minn AJ, Gupta GP, Siegel PM et al (2005) Genes that mediate breast cancer metastasis to lung. Nature 436(7050):518-524

54. Anderson P, Kopp L, Anderson N et al (2008) Novel bone cancer drugs: investigational agents and control paradigms for primary bone sarcomas (Ewing’s sarcoma and osteosarcoma). Expert Opin Investig Drugs 17(11):1703-1715

55. Hughes DP (2010) How the NOTCH Pathway Contributes to the Ability of Osteosarcoma Cells to Metastasize. Cancer Treat Res 152:479-496

56. Laverdiere C, Hoang BH, Yang R et al (2005) Messenger RNA expression levels of CXCR4 correlate with metastatic behavior and outcome in patients with osteosarcoma. Clin Cancer Res 11(7):2561-2567

57. Fan DG, Dai JY, Tang J et al (2009) Silencing of calpain expression reduces the metastatic potential of human osteosarcoma cells. Cell Biol Int 33(12):1263-1267

58. Kim LC, Song L, Haura EB (2009) Src kinases as therapeutic targets for cancer. Nat Rev Clin Oncol 6(10):587-595

59. Pignochino Y, Grignani G, Cavalloni G et al (2009) Sorafenib blocks tumour growth, angiogenesis and metastatic potential in preclinical models of osteosarcoma through a mechanism potentially involving the inhibition of ERK1/2, MCL-1 and ezrin pathways. Mol Cancer 8:118

60. Rubin EM, Guo Y, Tu K et al (2010) Wnt inhibitory factor 1 decreases tumorigenesis and metastasis in osteosarcoma. Mol Cancer Ther 9(3):731-741

61. Guo Y, Zi X, Koontz Z et al (2007) Blocking Wnt/LRP5 signaling by a soluble receptor modulates the epithelial to mesenchymal transition and suppresses met and metalloproteinases in osteosarcoma Saos-2 cells. J Orthop Res 25(7):964-971

62. Engin F, Bertin T, Ma O et al (2009) Notch signaling contributes to the pathogenesis of human osteosarcomas. Hum Mol Genet 18(8):1464-1470

63. Tanaka M, Setoguchi T, Hirotsu M et al (2009) Inhibition of Notch pathway prevents osteosarcoma growth by cell cycle regulation. Br J Cancer 100(12):1957-1965

64. Zhang P, Yang Y, Zweidler-McKay PA et al (2008) Critical role of notch signaling in osteosarcoma invasion and metastasis. Clin Cancer Res 14(10):2962-2969

65. Igney FH, Krammer PH (2002) Death and anti-death: tumour resistance to apoptosis. Nat Rev Cancer 2(4):277-288

66. Leow PC, Tian Q, Ong ZY et al (2009) Antitumor activity of natural compounds, curcumin and PKF118-310, as Wnt/beta-catenin antagonists against human osteosarcoma cells. Invest New Drugs

67. Cai Y, Mohseny AB, Karperien M et al (2010) Inactive Wnt/beta-catenin pathway in conventional high-grade osteosarcoma. J Pathol 220(1):24-33

68. Jeanes A, Gottardi CJ, Yap AS (2008) Cadherins and cancer: how does cadherin dysfunction promote tumor progression? Oncogene 27(55):6920-6929

69. Gordon N, Koshkina NV, Jia SF et al (2007) Corruption of the Fas pathway delays the pulmonary clearance of murine osteosarcoma cells, enhances their metastatic potential, and reduces the effect of aerosol gemcitabine. Clin Cancer Res 13(15 Pt 1):4503-4510

70. Duan X, Jia SF, Koshkina N et al (2006) Intranasal interleukin-12 gene therapy enhanced the activity of ifosfamide against osteosarcoma lung metastases. Cancer 106(6):1382-1388

71. Yang C, Yang S, Wood KB et al (2009) Multidrug resistant osteosarcoma cell lines exhibit deficiency of GADD45alpha expression. Apoptosis 14(1):124-133

72. Yuan XW, Zhu XF, Huang XF et al (2007) Interferon-alpha enhances sensitivity of human osteosarcoma U2OS cells to doxorubicin by p53-dependent apoptosis. Acta Pharmacol Sin 28(11):1835-1841

73. Hoang BH, Kubo T, Healey JH et al (2004) Expression of LDL receptor-related protein 5 (LRP5) as a novel marker for disease progression in high-grade osteosarcoma. Int J Cancer 109(1):106-111

Reference list

137

8

74. Schuetze S, Wathen K, Choy E et al (2010) Results of a Sarcoma Alliance for Research through Collaboration (SARC) phase II trial of dasatinib in previously treated, high-grade, advanced sarcoma. J Clin Oncol 28(15):10009

75. Whelan J, Patterson D, Perisoglou M et al (2010) The role of interferons in the treatment of osteosarcoma. Pediatr Blood Cancer 54(3):350-354

76. Lafleur EA, Jia SF, Worth LL et al (2001) Interleukin (IL)-12 and IL-12 gene transfer up-regulate Fas expression in human osteosarcoma and breast cancer cells. Cancer Res 61(10):4066-4071

77. Nakamura Y, Yamada N, Ohyama H et al (2006) Effect of interleukin-18 on metastasis of mouse osteosarcoma cells. Cancer Immunol Immunother 55(9):1151-1158

78. Chou AJ, Kleinerman ES, Krailo MD et al (2009) Addition of muramyl tripeptide to chemotherapy for patients with newly diagnosed metastatic osteosarcoma: a report from the Children’s Oncology Group. Cancer 115(22):5339-5348

79. Lewis VO (2009) What’s new in musculoskeletal oncology. J Bone Joint Surg Am 91(6):1546-1556

80. Huang CY, Lee CY, Chen MY et al (2009) Stromal cell-derived factor-1/CXCR4 enhanced motility of human osteosarcoma cells involves MEK1/2, ERK and NF-kappaB-dependent pathways. J Cell Physiol 221(1):204-212

81. Kaplan RN, Psaila B, Lyden D (2006) Bone marrow cells in the ‘pre-metastatic niche’: within bone and beyond. Cancer Metastasis Rev 25(4):521-529

82. Murphy PM (2001) Chemokines and the molecular basis of cancer metastasis. N Engl J Med 345(11):833-835

83. Wan X, Kim SY, Guenther LM et al (2009) Beta4 integrin promotes osteosarcoma metastasis and interacts with ezrin. Oncogene 28(38):3401-3411

84. Hauben EI, Bielack S, Grimer R et al (2006) Clinico-histologic parameters of osteosarcoma patients with late relapse. Eur J Cancer 42(4):460-466

85. Strauss SJ, McTiernan A, Whelan JS (2004) Late relapse of osteosarcoma: implications for follow-up and screening. Pediatr Blood Cancer 43(6):692-697

86. Barkan D, Green JE, Chambers AF (2010) Extracellular matrix: a gatekeeper in the transition from dormancy to metastatic growth. Eur J Cancer 46(7):1181-1188

87. Favaro E, Amadori A, Indraccolo S (2008) Cellular interactions in the vascular niche: implications in the regulation of tumor dormancy. APMIS 116(7-8):648-659

88. Almog N, Ma L, Raychowdhury R et al (2009) Transcriptional switch of dormant tumors to fast-growing angiogenic phenotype. Cancer Res 69(3):836-844

89. Liu X, Lei M, Erikson RL (2006) Normal cells, but not cancer cells, survive severe Plk1 depletion. Mol Cell Biol 26(6):2093-2108

90. Scotlandi K, Picci P, Kovar H (2009) Targeted therapies in bone sarcomas. Curr Cancer Drug Targets 9(7):843-853

91. Mansky PJ, Liewehr DJ, Steinberg SM et al (2002) Treatment of metastatic osteosarcoma with the somatostatin analog OncoLar: significant reduction of insulin-like growth factor-1 serum levels. J Pediatr Hematol Oncol 24(6):440-446

92. Kolb EA, Gorlick R, Houghton PJ et al (2008) Initial testing (stage 1) of a monoclonal antibody (SCH 717454) against the IGF-1 receptor by the pediatric preclinical testing program. Pediatr Blood Cancer 50(6):1190-1197

93. Abdeen A, Chou AJ, Healey JH et al (2009) Correlation between clinical outcome and growth factor pathway expression in osteogenic sarcoma. Cancer 115(22):5243-5250

94. Kim HS, Lim SJ, Park YK (2009) Anti-angiogenic factor endostatin in osteosarcoma. APMIS 117(10):716-723

95. Ryu K, Choy E, Yang C et al (2010) Activation of signal transducer and activator of transcription 3 (Stat3) pathway in osteosarcoma cells and overexpression of phosphorylated-Stat3 correlates with poor prognosis. J Orthop Res

96. Kaya M, Wada T, Nagoya S et al (2007) Prevention of postoperative progression of pulmonary metastases in osteosarcoma by antiangiogenic therapy using endostatin. J Orthop Sci 12(6):562-567

97. Kolb EA, Kamara D, Zhang W et al (2010) R1507, a fully human monoclonal antibody targeting IGF-1R, is effective alone and in combination with rapamycin in inhibiting growth of osteosarcoma xenografts. Pediatr Blood Cancer 55(1):67-75

98. Anderson PM, Wiseman GA, Erlandson L et al (2005) Gemcitabine radiosensitization after high-dose samarium for osteoblastic osteosarcoma. Clin Cancer Res 11(19 Pt 1):6895-6900

138

99. Machak GN, Tkachev SI, Solovyev YN et al (2003) Neoadjuvant chemotherapy and local radiotherapy for high-grade osteosarcoma of the extremities. Mayo Clin Proc 78(2):147-155

100. Schwarz R, Bruland O, Cassoni A et al (2010) The role of radiotherapy in oseosarcoma. Cancer Treat Res 152:147-164

101. Delaney TF, Park L, Goldberg SI et al (2005) Radiotherapy for local control of osteosarcoma. Int J Radiat Oncol Biol Phys 61(2):492-498

102. Li Y, Tanaka K, Li X et al (2007) Cyclin-dependent kinase inhibitor, flavopiridol, induces apoptosis and inhibits tumor growth in drug-resistant osteosarcoma and Ewing’s family tumor cells. Int J Cancer 121(6):1212-1218

103. Dieudonne FX, Marion A, Marie PJ et al (2012) Targeted inhibition of T-cell factor activity promotes syndecan-2 expression and sensitization to doxorubicin in osteosarcoma cells and bone tumors in mice. J Bone Miner Res 27(10):2118-2129

104. Marion A, Dieudonne FX, Patino-Garcia A et al (2012) Calpain-6 is an endothelin-1 signaling dependent protective factor in chemoresistant osteosarcoma. Int J Cancer 130(11):2514-2525

105. MacKeigan JP, Murphy LO, Blenis J (2005) Sensitized RNAi screen of human kinases and phosphatases identifies new regulators of apoptosis and chemoresistance. Nat Cell Biol 7(6):591-600

106. Zhang H, Wu H, Zheng J et al (2013) Transforming Growth Factor beta1 Signal is Crucial for Dedifferentiation of Cancer Cells to Cancer Stem Cells in Osteosarcoma. Stem Cells 31(3):433-446

107. Dean M, Fojo T, Bates S (2005) Tumour stem cells and drug resistance. Nat Rev Cancer 5(4):275-284

108. Gibbs CP, Kukekov VG, Reith JD et al (2005) Stem-like cells in bone sarcomas: implications for tumorigenesis. Neoplasia 7(11):967-976

109. Tirino V, Desiderio V, d’Aquino R et al (2008) Detection and characterization of CD133+ cancer stem cells in human solid tumours. PLoS One 3(10):e3469

110. Almog N, Ma L, Schwager C et al (2012) Consensus micro RNAs governing the switch of dormant tumors to the fast-growing angiogenic phenotype. PLoS One 7(8):e44001

111. Chou AJ, Geller DS, Gorlick R (2008) Therapy for osteosarcoma: where do we go from here? Paediatr Drugs 10(5):315-327

112. Frisch SM, Screaton RA (2001) Anoikis mechanisms. Curr Opin Cell Biol 13(5):555-562

113. Marco RA, Diaz-Montero CM, Wygant JN et al (2003) Alpha 4 integrin increases anoikis of human osteosarcoma cells. J Cell Biochem 88(5):1038-1047

114. Lopez-Guerrero JA, Lopez-Gines C, Pellin A et al (2004) Deregulation of the G1 to S-phase cell cycle checkpoint is involved in the pathogenesis of human osteosarcoma. Diagn Mol Pathol 13(2):81-91

115. Lonardo F, Ueda T, Huvos AG et al (1997) p53 and MDM2 alterations in osteosarcomas: correlation with clinicopathologic features and proliferative rate. Cancer 79(8):1541-1547

116. Miller CW, Aslo A, Won A et al (1996) Alterations of the p53, Rb and MDM2 genes in osteosarcoma. J Cancer Res Clin Oncol 122(9):559-565

117. Patino-Garcia A, Pineiro ES, Diez MZ et al (2003) Genetic and epigenetic alterations of the cell cycle regulators and tumor suppressor genes in pediatric osteosarcomas. J Pediatr Hematol Oncol 25(5):362-367

118. Bode AM, Dong Z (2007) The enigmatic effects of caffeine in cell cycle and cancer. Cancer Lett 247(1):26-39

119. Iguchi T, Miyakawa Y, Saito K et al (2007) Zoledronate-induced S phase arrest and apoptosis accompanied by DNA damage and activation of the ATM/Chk1/cdc25 pathway in human osteosarcoma cells. Int J Oncol 31(2):285-291

120. Ory B, Blanchard F, Battaglia S et al (2007) Zoledronic acid activates the DNA S-phase checkpoint and induces osteosarcoma cell death characterized by apoptosis-inducing factor and endonuclease-G translocation independently of p53 and retinoblastoma status. Mol Pharmacol 71(1):333-343

121. Kim MJ, Lee JY, Lee SJ (2008) Transient suppression of nuclear Cdc2 activity in response to ionizing radiation. Oncol Rep 19(5):1323-1329

122. Posthumadeboer J, Wurdinger T, Graat HC et al (2011) WEE1 inhibition sensitizes osteosarcoma to radiotherapy. BMC Cancer 11:156

123. Wells NJ, Watanabe N, Tokusumi T et al (1999) The C-terminal domain of the Cdc2 inhibitory kinase Myt1 interacts with Cdc2 complexes and is required for inhibition of G(2)/M progression. J Cell Sci 112 ( Pt 19):3361-3371

Reference list

139

8

124. Huang G, Mills L, Worth LL (2007) Expression of human glutathione S-transferase P1 mediates the chemosensitivity of osteosarcoma cells. Mol Cancer Ther 6(5):1610-1619

125. Hirai H, Iwasawa Y, Okada M et al (2009) Small-molecule inhibition of Wee1 kinase by MK-1775 selectively sensitizes p53-deficient tumor cells to DNA-damaging agents. Mol Cancer Ther 8(11):2992-3000

126. Senderowicz AM, Sausville EA (2000) Preclinical and clinical development of cyclin-dependent kinase modulators. J Natl Cancer Inst 92(5):376-387

127. Senderowicz AM (1999) Flavopiridol: the first cyclin-dependent kinase inhibitor in human clinical trials. Invest New Drugs 17(3):313-320

128. Kimura H, Tsuchiya H, Shirai T et al (2009) Caffeine-potentiated chemotherapy for metastatic osteosarcoma. J Orthop Sci 14(5):556-565

129. Scotlandi K, Serra M, Manara MC et al (1994) Pre-treatment of human osteosarcoma cells with N-methylformamide enhances P-glycoprotein expression and resistance to doxorubicin. Int J Cancer 58(1):95-101

130. Caronia D, Patino-Garcia A, Perez-Martinez A et al (2011) Effect of ABCB1 and ABCC3 polymorphisms on osteosarcoma survival after chemotherapy: a pharmacogenetic study. PLoS One 6(10):e26091

131. Brambilla D, Zamboni S, Federici C et al (2012) P-glycoprotein binds to ezrin at amino acid residues 149-242 in the FERM domain and plays a key role in the multidrug resistance of human osteosarcoma. Int J Cancer 130(12):2824-2834

132. Chou AJ, Gorlick R (2006) Chemotherapy resistance in osteosarcoma: current challenges and future directions. Expert Rev Anticancer Ther 6(7):1075-1085

133. Fayette J, Blay JY (2005) Genetic predictors for drug resistance in soft tissue sarcoma: a review of publications in 2004. Curr Opin Oncol 17(4):370-375

134. Ifergan I, Meller I, Issakov J et al (2003) Reduced folate carrier protein expression in osteosarcoma: implications for the prediction of tumor chemosensitivity. Cancer 98(9):1958-1966

135. Perez J, Bardin C, Rigal C et al (2011) Anti-MDR1 siRNA restores chemosensitivity in chemoresistant breast carcinoma and osteosarcoma cell lines. Anticancer Res 31(9):2813-2820

136. Pasello M, Manara MC, Michelacci F et al (2011) Targeting glutathione-S transferase enzymes in musculoskeletal sarcomas: a promising therapeutic strategy. Anal Cell Pathol (Amst) 34(3):131-145

137. Sau A, Filomeni G, Pezzola S et al (2012) Targeting GSTP1-1 induces JNK activation and leads to apoptosis in cisplatin-sensitive and -resistant human osteosarcoma cell lines. Mol Biosyst 8(4):994-1006

138. Wang Y, Huang JW, Li M et al (2011) MicroRNA-138 modulates DNA damage response by repressing histone H2AX expression. Mol Cancer Res 9(8):1100-1111

139. Wang D, Luo M, Kelley MR (2004) Human apurinic endonuclease 1 (APE1) expression and prognostic significance in osteosarcoma: enhanced sensitivity of osteosarcoma to DNA damaging agents using silencing RNA APE1 expression inhibition. Mol Cancer Ther 3(6):679-686

140. Biason P, Hattinger CM, Innocenti F et al (2012) Nucleotide excision repair gene variants and association with survival in osteosarcoma patients treated with neoadjuvant chemotherapy. Pharmacogenomics J 12(6):476-483

141. Pignochino Y, Dell’aglio C, Basirico M et al (2013) The Combination of Sorafenib and Everolimus abrogates mTORC-1 and mTORC2 up-regulation in preclinical models of Osteosarcoma. Clin Cancer Res

142. Posthumadeboer J, van Egmond PW, Helder MN et al (2012) Targeting JNK-interacting-protein-1 (JIP1) sensitises osteosarcoma to doxorubicin. Oncotarget 3(10):1169-1181

143. Wheelock MJ, Shintani Y, Maeda M et al (2008) Cadherin switching. J Cell Sci 121(Pt 6):727-735

144. Huang J, Ni J, Liu K et al (2012) HMGB1 promotes drug resistance in osteosarcoma. Cancer Res 72(1):230-238

145. Graat HC, Carette JE, Schagen FH et al (2007) Enhanced tumor cell kill by combined treatment with a small-molecule antagonist of mouse double minute 2 and adenoviruses encoding p53. Mol Cancer Ther 6(5):1552-1561

146. Aziz MH, Shen H, Maki CG (2011) Acquisition of p53 mutations in response to the non-genotoxic p53 activator Nutlin-3. Oncogene 30(46):4678-4686

147. Nardin A, Lefebvre ML, Labroquere K et al (2006) Liposomal muramyl tripeptide phosphatidylethanolamine: Targeting and activating macrophages for adjuvant treatment of osteosarcoma. Curr Cancer Drug Targets 6(2):123-133

140

148. Kleinerman ES, Gano JB, Johnston DA et al (1995) Efficacy of liposomal muramyl tripeptide (CGP 19835A) in the treatment of relapsed osteosarcoma. Am J Clin Oncol 18(2):93-99

149. Kager L, Potschger U, Bielack S (2010) Review of mifamurtide in the treatment of patients with osteosarcoma. Ther Clin Risk Manag 6:279-286

150. Kleinerman ES, Maeda M, Jaffe N (1993) Liposome-encapsulated muramyl tripeptide: a new biologic response modifier for the treatment of osteosarcoma. Cancer Treat Res 62:101-107

151. Hundsdoerfer P, Albrecht M, Ruhl U et al (2009) Long-term outcome after polychemotherapy and intensive local radiation therapy of high-grade osteosarcoma. Eur J Cancer 45(14):2447-2451

152. Mahajan A, Woo SY, Kornguth DG et al (2008) Multimodality treatment of osteosarcoma: radiation in a high-risk cohort. Pediatr Blood Cancer 50(5):976-982

153. Wagner TD, Kobayashi W, Dean S et al (2009) Combination short-course preoperative irradiation, surgical resection, and reduced-field high-dose postoperative irradiation in the treatment of tumors involving the bone. Int J Radiat Oncol Biol Phys 73(1):259-266

154. Blattmann C, Oertel S, Schulz-Ertner D et al (2010) Non-randomized therapy trial to determine the safety and efficacy of heavy ion radiotherapy in patients with non-resectable osteosarcoma. BMC Cancer 10:96

155. Kamada T, Tsujii H, Tsuji H et al (2002) Efficacy and safety of carbon ion radiotherapy in bone and soft tissue sarcomas. J Clin Oncol 20(22):4466-4471

156. Anderson P, Aguilera D, Pearson M et al (2008) Outpatient chemotherapy plus radiotherapy in sarcomas: improving cancer control with radiosensitizing agents. Cancer Control 15(1):38-46

157. Castedo M, Perfettini JL, Roumier T et al (2004) Cell death by mitotic catastrophe: a molecular definition. Oncogene 23(16):2825-2837

158. Mir SE, De Witt Hamer PC, Krawczyk PM et al (2010) In Silico Analysis of Kinase Expression Identifies WEE1 as a Gatekeeper against Mitotic Catastrophe in Glioblastoma. Cancer Cell 18(3):244-257

159. Jiang H, Reinhardt HC, Bartkova J et al (2009) The combined status of ATM and p53 link tumor development with therapeutic response. Genes Dev 23(16):1895-1909

160. Raleigh JM, O’Connell MJ (2000) The G(2) DNA damage checkpoint targets both Wee1 and Cdc25. J Cell Sci 113 ( Pt 10):1727-1736

161. Stathis A, Oza A (2010) Targeting Wee1-like protein kinase to treat cancer. Drug News Perspect 23(7):425-429

162. Wang Y, Li J, Booher RN et al (2001) Radiosensitization of p53 mutant cells by PD0166285, a novel G(2) checkpoint abrogator. Cancer Res 61(22):8211-8217

163. Hashimoto O, Shinkawa M, Torimura T et al (2006) Cell cycle regulation by the Wee1 inhibitor PD0166285, pyrido [2,3-d] pyimidine, in the B16 mouse melanoma cell line. BMC Cancer 6:292

164. Iurisci I, Filipski E, Reinhardt J et al (2006) Improved tumor control through circadian clock induction by Seliciclib, a cyclin-dependent kinase inhibitor. Cancer Res 66(22):10720-10728

165. Enders GH (2010) Gauchos and ochos: a Wee1-Cdk tango regulating mitotic entry. Cell Div 5:12

166. R2. Microarray Analysis and Visualisation Platform. Jan Koster, editor. http://r2.amc.nl . 2010. Human Genetics Department, Academic Medical Center, Amsterdam, the Netherlands. 3-9-2010. Ref Type: Electronic Citation

167. Mollapour M, Tsutsumi S, Donnelly AC et al (2010) Swe1Wee1-dependent tyrosine phosphorylation of Hsp90 regulates distinct facets of chaperone function. Mol Cell 37(3):333-343

168. Garcia K, Stumpff J, Duncan T et al (2009) Tyrosines in the kinesin-5 head domain are necessary for phosphorylation by Wee1 and for mitotic spindle integrity. Curr Biol 19(19):1670-1676

169. Mullenders J, Bernards R (2009) Loss-of-function genetic screens as a tool to improve the diagnosis and treatment of cancer. Oncogene 28(50):4409-4420

170. Swanton C, Nicke B, Marani M et al (2007) Initiation of high frequency multi-drug resistance following kinase targeting by siRNAs. Cell Cycle 6(16):2001-2004

171. Johnson SA, Hunter T (2005) Kinomics: methods for deciphering the kinome. Nat Methods 2(1):17-25

172. Manning G, Whyte DB, Martinez R et al (2002) The protein kinase complement of the human genome. Science 298(5600):1912-1934

173. Ivy SP, Wick JY, Kaufman BM (2009) An overview of small-molecule inhibitors of VEGFR signaling. Nat Rev Clin Oncol 6(10):569-579

Reference list

141

8

174. Gotink KJ, Verheul HM (2010) Anti-angiogenic tyrosine kinase inhibitors: what is their mechanism of action? Angiogenesis 13(1):1-14

175. Jia SF, Worth LL, Kleinerman ES (1999) A nude mouse model of human osteosarcoma lung metastases for evaluating new therapeutic strategies. Clin Exp Metastasis 17(6):501-506

176. Stebbins JL, De SK, Machleidt T et al (2008) Identification of a new JNK inhibitor targeting the JNK-JIP interaction site. Proc Natl Acad Sci U S A 105(43):16809-16813

177. R Development Core Team (2008) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria

178. Boutros M, Bras LP, Huber W (2006) Analysis of cell-based RNAi screens. Genome Biol 7(7):R66

179. Smyth GK (2005) Limma: linear models for microarray data. In: Gentleman R, Carey V, Dudoit S, Irizarry R, Huber W (eds) Bioinformatics and Computational Biology Solutions using R and BioconductorSpringer, New York, pp397-420

180. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Roy Statist Soc Ser (B 57):289-300

181. Schmittgen TD, Livak KJ (2008) Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc 3(6):1101-1108

182. Mohseny AB, Szuhai K, Romeo S et al (2009) Osteosarcoma originates from mesenchymal stem cells in consequence of aneuploidization and genomic loss of Cdkn2. J Pathol 219(3):294-305

183. Rosen G, Caparros B, Huvos AG et al (1982) Preoperative chemotherapy for osteogenic sarcoma: selection of postoperative adjuvant chemotherapy based on the response of the primary tumor to preoperative chemotherapy. Cancer 49(6):1221-1230

184. Altman DG (1991) Practical statistics for medical research.Chapmann&Hall, London

185. Landis JR, Koch GG (1977) The measurement of observer agreement for categorical data. Biometrics 33(1):159-174

186. Zhang JH, Chung TD, Oldenburg KR (1999) A Simple Statistical Parameter for Use in Evaluation and Validation of High Throughput Screening Assays. J Biomol Screen 4(2):67-73

187. Ikezoe T, Yang Y, Taguchi H et al (2004) JNK interacting protein 1 (JIP-1) protects LNCaP prostate cancer cells from growth arrest and apoptosis mediated by 12-0-tetradecanoylphorbol-13-acetate (TPA). Br J Cancer 90(10):2017-2024

188. Vaishnav M, MacFarlane M, Dickens M (2011) Disassembly of the JIP1/JNK molecular scaffold by caspase-3-mediated cleavage of JIP1 during apoptosis. Exp Cell Res 317(7):1028-1039

189. Whitmarsh AJ, Kuan CY, Kennedy NJ et al (2001) Requirement of the JIP1 scaffold protein for stress-induced JNK activation. Genes Dev 15(18):2421-2432

190. Blanco S, Sanz-Garcia M, Santos CR et al (2008) Modulation of interleukin-1 transcriptional response by the interaction between VRK2 and the JIP1 scaffold protein. PLoS One 3(2):e1660

191. Dunn C, Wiltshire C, MacLaren A et al (2002) Molecular mechanism and biological functions of c-Jun N-terminal kinase signalling via the c-Jun transcription factor. Cell Signal 14(7):585-593

192. Liu J, Lin A (2005) Role of JNK activation in apoptosis: a double-edged sword. Cell Res 15(1):36-42

193. Kim JW, Kim MJ, Kim KJ et al (2005) Notch interferes with the scaffold function of JNK-interacting protein 1 to inhibit the JNK signaling pathway. Proc Natl Acad Sci U S A 102(40):14308-14313

194. Engstrom W, Ward A, Moorwood K (2010) The role of scaffold proteins in JNK signalling. Cell Prolif 43(1):56-66

195. Schlabach MR, Luo J, Solimini NL et al (2008) Cancer proliferation gene discovery through functional genomics. Science 319(5863):620-624

196. Peterson D, Lee J, Lei XC et al (2010) A chemosensitization screen identifies TP53RK, a kinase that restrains apoptosis after mitotic stress. Cancer Res 70(15):6325-6335

197. Whitehurst AW, Bodemann BO, Cardenas J et al (2007) Synthetic lethal screen identification of chemosensitizer loci in cancer cells. Nature 446(7137):815-819

198. Yamaguchi U, Honda K, Satow R et al (2009) Functional genome screen for therapeutic targets of osteosarcoma. Cancer Sci 100(12):2268-2274

199. Yang C, Ji D, Weinstein EJ et al (2010) The kinase Mirk is a potential therapeutic target in osteosarcoma. Carcinogenesis 31(4):552-558

142

Reference list

200. Shaulian E, Karin M (2002) AP-1 as a regulator of cell life and death. Nat Cell Biol 4(5):E131-E136

201. Grant SK (2009) Therapeutic protein kinase inhibitors. Cell Mol Life Sci 66(7):1163-1177

202. Baumhoer D, Elsner M, Smida J et al (2011) CRIP1 expression is correlated with a favorable outcome and less metastases in osteosarcoma patients. Oncotarget 2(12):970-975

203. Chen Y, Yang Y, Yuan Z et al (2012) Predicting chemosensitivity in osteosarcoma prior to chemotherapy: An investigational study of biomarkers with immunohistochemistry. Oncol Lett 3(5):1011-1016

204. Saini V, Hose CD, Monks A et al (2012) Identification of CBX3 and ABCA5 as Putative Biomarkers for Tumor Stem Cells in Osteosarcoma. PLoS One 7(8):e41401

205. Cox J, Mann M (2007) Is proteomics the new genomics? Cell 130(3):395-398

206. Wu CC, Yates JR, III (2003) The application of mass spectrometry to membrane proteomics. Nat Biotechnol 21(3):262-267

207. Pham TV, Piersma SR, Oudgenoeg G et al (2012) Label-free mass spectrometry-based proteomics for biomarker discovery and validation. Expert Rev Mol Diagn 12(4):343-359

208. Lund R, Leth-Larsen R, Jensen ON et al (2009) Efficient isolation and quantitative proteomic analysis of cancer cell plasma membrane proteins for identification of metastasis-associated cell surface markers. J Proteome Res 8(6):3078-3090

209. Scheurer SB, Rybak JN, Roesli C et al (2005) Identification and relative quantification of membrane proteins by surface biotinylation and two-dimensional peptide mapping. Proteomics 5(11):2718-2728

210. Stevens TJ, Arkin IT (2000) Do more complex organisms have a greater proportion of membrane proteins in their genomes? Proteins 39(4):417-420

211. Zhang YK, Zhang XH, Li JM et al (2009) A proteomic study on a human osteosarcoma cell line Saos-2 treated with diallyl trisulfide. Anticancer Drugs 20(8):702-712

212. Rochet N, Dubousset J, Mazeau C et al (1999) Establishment, characterisation and partial cytokine expression profile of a new human osteosarcoma cell line (CAL 72). Int J Cancer 82(2):282-285

213. de Wit M., Jimenez CR, Carvalho B et al (2012) Cell surface proteomics identifies glucose transporter type 1 and prion protein as candidate biomarkers for colorectal adenoma-to-carcinoma progression. Gut 61(6):855-864

214. Piersma SR, Fiedler U, Span S et al (2010) Workflow comparison for label-free, quantitative secretome proteomics for cancer biomarker discovery: method evaluation, differential analysis, and verification in serum. J Proteome Res 9(4):1913-1922

215. Albrethsen J, Knol JC, Piersma SR et al (2010) Subnuclear proteomics in colorectal cancer: identification of proteins enriched in the nuclear matrix fraction and regulation in adenoma to carcinoma progression. Mol Cell Proteomics 9(5):988-1005

216. Pham TV, Piersma SR, Warmoes M et al (2010) On the beta-binomial model for analysis of spectral count data in label-free tandem mass spectrometry-based proteomics. Bioinformatics 26(3):363-369

217. Van Geer MA, Brevoord D, Kuhlmann KF et al (2010) A fiber modified adenovirus vector that targets to the EphrinA2 receptor reveals enhanced gene transfer to ex vivo pancreatic cancer. Int J Oncol 36(1):233-244

218. Liu X, Zeng B, Ma J et al (2009) Comparative proteomic analysis of osteosarcoma cell and human primary cultured osteoblastic cell. Cancer Invest 27(3):345-352

219. Guo QC, Shen JN, Jin S et al (2007) Comparative proteomic analysis of human osteosarcoma and SV40-immortalized normal osteoblastic cell lines. Acta Pharmacol Sin 28(6):850-858

220. Folio C, Mora MI, Zalacain M et al (2009) Proteomic analysis of chemonaive pediatric osteosarcomas and corresponding normal bone reveals multiple altered molecular targets. J Proteome Res 8(8):3882-3888

221. Hua Y, Jia X, Sun M et al (2011) Plasma membrane proteomic analysis of human osteosarcoma and osteoblastic cells: revealing NDRG1 as a marker for osteosarcoma. Tumour Biol 32(5):1013-1021

222. Zhang Z, Zhang L, Hua Y et al (2010) Comparative proteomic analysis of plasma membrane proteins between human osteosarcoma and normal osteoblastic cell lines. BMC Cancer 10:206

223. Avnet S, Sciacca L, Salerno M et al (2009) Insulin receptor isoform A and insulin-like growth factor II as additional treatment targets in human osteosarcoma. Cancer Res 69(6):2443-2452

224. Chen K, Fallen S, Abaan HO et al (2008) Wnt10b induces chemotaxis of osteosarcoma and correlates with reduced survival. Pediatr Blood Cancer 51(3):349-355

143

8

225. McQueen P, Ghaffar S, Guo Y et al (2011) The Wnt signaling pathway: implications for therapy in osteosarcoma. Expert Rev Anticancer Ther 11(8):1223-1232

226. Abdou AG, Abd el-Wahed MM, Asaad NY et al (2010) Ephrin A4 expression in osteosarcoma, impact on prognosis, and patient outcome. Indian J Cancer 47(1):46-52

227. Fritsche-Guenther R, Noske A, Ungethum U et al (2010) De novo expression of EphA2 in osteosarcoma modulates activation of the mitogenic signalling pathway. Histopathology 57(6):836-850

228. Varelias A, Koblar SA, Cowled PA et al (2002) Human osteosarcoma expresses specific ephrin profiles: implications for tumorigenicity and prognosis. Cancer 95(4):862-869

229. Pasquale EB (2010) Eph receptors and ephrins in cancer: bidirectional signalling and beyond. Nat Rev Cancer 10(3):165-180

230. Tandon M, Vemula SV, Mittal SK (2011) Emerging strategies for EphA2 receptor targeting for cancer therapeutics. Expert Opin Ther Targets 15(1):31-51

231. Udayakumar D, Zhang G, Ji Z et al (2011) Epha2 is a critical oncogene in melanoma. Oncogene 30(50):4921-4929

232. Landen CN, Jr., Chavez-Reyes A, Bucana C et al (2005) Therapeutic EphA2 gene targeting in vivo using neutral liposomal small interfering RNA delivery. Cancer Res 65(15):6910-6918

233. Hammond SA, Lutterbuese R, Roff S et al (2007) Selective targeting and potent control of tumor growth using an EphA2/CD3-Bispecific single-chain antibody construct. Cancer Res 67(8):3927-3935

234. Lee JW, Stone RL, Lee SJ et al (2010) EphA2 targeted chemotherapy using an antibody drug conjugate in endometrial carcinoma. Clin Cancer Res 16(9):2562-2570

235. Mitra S, Duggineni S, Koolpe M et al (2010) Structure-activity relationship analysis of peptides targeting the EphA2 receptor. Biochemistry 49(31):6687-6695

236. Dickerson EB, Blackburn WH, Smith MH et al (2010) Chemosensitization of cancer cells by siRNA using targeted nanogel delivery. BMC Cancer 10:10

237. Scarberry KE, Dickerson EB, McDonald JF et al (2008) Magnetic nanoparticle-peptide conjugates for in vitro and in vivo targeting and extraction of cancer cells. J Am Chem Soc 130(31):10258-10262

238. Van Geer MA, Bakker CT, Koizumi N et al (2009) Ephrin A2 receptor targeting does not increase adenoviral pancreatic cancer transduction in vivo. World J Gastroenterol 15(22):2754-2762

239. Sullivan R, Peppercorn J, Sikora K et al (2011) Delivering affordable cancer care in high-income countries. Lancet Oncol 12(10):933-980

240. Davies JE, Neidle S, Taylor DG (2012) Developing and paying for medicines for orphan indications in oncology: utilitarian regulation vs equitable care? Br J Cancer 106(1):14-17

241. Mohseny AB, Machado I, Cai Y et al (2011) Functional characterization of osteosarcoma cell lines provides representative models to study the human disease. Lab Invest 91(8):1195-1205

242. Janeway KA, Walkley CR (2010) Modeling human osteosarcoma in the mouse: From bedside to bench. Bone 47(5):859-865

243. Fisher R, Pusztai L, Swanton C (2013) Cancer heterogeneity: implications for targeted therapeutics. Br J Cancer 108(3):479-485

244. Reischl D, Zimmer A (2008) Drug delivery of siRNA therapeutics: potentials and limits of nanosystems. Nanomedicine

245. Kobayashi E, Iyer AK, Hornicek FJ et al (2013) Lipid-functionalized dextran nanosystems to overcome multidrug resistance in cancer: a pilot study. Clin Orthop Relat Res 471(3):915-925

246. Oosterhoff D, Witlox MA, Van B, V et al (2003) Gene-directed enzyme prodrug therapy for osteosarcoma: sensitization to CPT-11 in vitro and in vivo by adenoviral delivery of a gene encoding secreted carboxylesterase-2. Mol Cancer Ther 2(8):765-771

247. Witlox MA, Van B, V, Grill J et al (2002) Epidermal growth factor receptor targeting enhances adenoviral vector based suicide gene therapy of osteosarcoma. J Gene Med 4(5):510-516

248. Lammers T, Hennink WE, Storm G (2008) Tumour-targeted nanomedicines: principles and practice. Br J Cancer 99(3):392-397

249. Chen ZG (2010) Small-molecule delivery by nanoparticles for anticancer therapy. Trends Mol Med 16(12):594-602

List of abbreviations

146

List of abbreviations

L I S T O F A B B R E V I AT I O N S

ABC ATP binding cassette (transporter)Akt v-akt murine thymoma viral oncogene homolog 1AMPK 5’ AMP-activated protein kinaseAP-1 Transcription factor AP-1APE-1 Apurinic-apyrimidinic endonuclease 1ATM Ataxia telangiectasia mutatedATR ATM and Rad3-relatedBAX Bcl2-like protein 4Bcl-2 B cell lymphoma 2 associated oncogeneBcl-XL Bcl2-like 1BER Base excision repairBSA Bovine serum albuminCD Cluster of differentiation / designationCdc Cyclin dependent kinaseCDC2 Cyclin Dependent Kinase 1CDK Cyclin dependent kinaseChk Checkpoint kinaseCXCR Chemokine (C-X-C-motif) receptorCXCL Chemokine (C-X-C-motif) ligandDSB Double strand breakDTT DithiothreitolECM Extracellular matrix EGFR Epidermal growth factor receptorEMT Epithelial-mesenchymal-transitionEPHA Ephrin type-A receptorEPHB Ephrin type-B receptorERK Extracellular signal regulated kinaseFAK Focal adhesion kinaseFasL Fas ligandγ-H2AX {Gamma}-Histone H2AXGADD45α Growth arrest and DNA damage-inducible protein GADD45 alphaGFP Green fluorescent proteinGH Growth hormoneGO Gene ontologyGST(P1) Glutathione S-transferase (pi 1)Gy GrayHLA Human leukocyte antigenHMGB1 High mobility group protein B1Hp-OB Human primary osteoblastHsp90 Heat shock protein 90IAP Inhibitor of apoptosisIFN-α Interferon-alpha IGF Insulin-like growth factorIGF-1R Insulin-like growth factor 1 receptorIL InterleukinInh InhibitorINS Insulin receptorIR IrradiationiTRAQ Isobaric tags for relative and absolute quantitationIU Infectious unit

147

8

JAK Janus kinaseJNK c-Jun N-terminal kinasemAB Monoclonal antibodyMAP(K) Mitogen-activated protein (kinase)MDM2 Mouse Double minute 2 proteinMDR Multidrug resistanceMEK Mitogen-activated protein kinase kinase 1 MI Mitotic indexmiR Micro RNAMMP Matrix metalloproteinaseMOI Multiplicity of infectionmTOR(C) Mammalian target of Rapamycin (complex)MS/MS Tandem mass spectrometryMTP-PE Muramyl tripeptide phophatidyl ethanolaminenanoLC Nano liquid chromatographyNBDHEX 6-(7-nitro-2,1,3-benzoxadiazol-4-ylthio)hexanolNDRG1 N-myc downstream regulated 1NER Nucleotide excision repairNF-κB Nuclear factor-kappa BNK cells Natural killer cellsNOXA Phorbol-12-myristate-13-acetate-induced protein 1OS Osteosarcomap53AIP1 p53-regulated apoptosis-inducing protein 1PBS Phosphate-buffered salinePD Pro-drugPDGF-R Platelet-derived growht factor receptorPHH3 Phospho-Histone H3 PI3K Phosphatidylinositol 3-kinasePPI Protein-protein interactionPUMA Bcl2-binding component 3RFC Reduced folate carrierRHOA Ras homolog family member ARNAi RNA interferencesc Spectral countsSD Standard deviationshRNA Short hairpin RNAsiRNA Small interfering RNASNP Single nucleotide polymorphismSTAT(3) Signal transducer and activator of transcription (3)TCF T-cell factorTGF- β Transforming growth factor - betaTMA Tissue micro arrayTNF Tumor necrosis factorTRAF2 TNF receptor-associated factor 2VEGF Vascular endothelial growth factorWIF-1 Wnt inhibitory factor 1

Appendix A: Supplementary files

and figures corresponding to Chapter 5

150

Appendix A

Supplementary figure 5.1 siRNA deconvolution analysis of selected screen hits. Results for JIP1 are shown in Figure 5.2 of the main manuscript; the other nine selected hits are shown here. Confirmation scores of all 10 analyses are summarised in Figure 5.1B of the main manuscript. Two genes could not be confirmed: CDKL1 has 2 effective duplexes and PRKCSH was excluded from further analysis since gene-silencing alone induced an increase in cell viability in 3 duplexes, which was considered to be undesirable and leaving one effective duplex for this gene.

151

8

Supplementary Figure 5.2 Tissue micro array (TMA) layout and staining consensus chart. (A) Overview of the layouts of the two tissue micro arrays analysed for JIP1 staining. Corresponding clinical data is summarised in Supplementary Table 5.3. (B) Consensus chart showing six exemplary tissue cores per category (negative or positive) used as guide while scoring JIP1 staining.

Appendix B: Supplementary files

and figures corresponding to Chapter 6

154

Appendix B

Supplementary file 6.1: cell surface protein isolation and mass spectrometry protocol.

Cell surface protein isolationFor the isolation and collection of surface proteins, we used the Pierce® Cell Surface Protein Isolation Kit (Thermo Scientific), generally following the protocol described by de Wit et al. [de Wit M. 2012] (see Figure 6.1 of the main manuscript). Per biological replicate, 3x107 cells were cultured in five 75 cm2 flasks. Prior to surface protein biotinylation, all reagents were cooled to 4°C. The cells were washed four times with ice-cold phosphate buffered saline (PBS) followed by incubation with 0.25 mg/mL Sulfo-NHS-SS-Biotin in 48 mL ice-cold PBS per flask on a rocking platform for 30 minutes at 4°C. The biotinylation reaction was quenched by adding 500 μL of the provided Quenching Solution (Pierce). Cells were washed with ice-cold PBS, harvested by gentle scraping and pelleted by centrifugation. The cells were lysed using the provided Lysis Buffer (Pierce) containing a protease inhibitor cocktail (Sigma) for 30 minutes on ice with intermittent vortexing. To rid cell remnants, the lysates were centrifuged at 16,000 x g for 2 minutes at 4°C. The clarified supernatant was used for purification of biotinylated proteins on NeutrAvidin Agarose. Before use, 500 μL of NeutrAvidin Agarose slurry was washed thrice with Pierce Wash Buffer in a provided column (Pierce). The clarified supernatant was added to the slurry and incubated for 2 h at 4°C in the closed column using an end-over-end tumbler to mix vigorously and allow the biotinylated proteins to bind to the NeutrAvidin Agarose slurry. The unbound proteins, representing the intracellular fraction, were collected by centrifugation of the column at 1,000 x g for 2 minutes and stored at -20°C to serve as an internal control for the surface protein isolation process. Any remaining unbound proteins were removed by repetitive washing; thrice with 1% Nonidet-P40 and 0.1% SDS in 500 μL PBS and thrice with 0.1% Nonidet-P40 and 0.5 M NaCl in 500 μL PBS. Finally, the captured surface proteins were eluted from the biotin-NeutrAvidin Agarose by incubation with 330 μL of 50 nM dithiothreitol (DTT) in PBS containing 62.5 mM Tris-HCl for 1 h at room temperature (RT) in the end-over-end tumbler. The eluted proteins, representing the cell surface proteins, were collected by column centrifugation at 1,000 x g for 2 minutes. For all cell lines, three biological replicates were obtained; the three cell surface protein lysates per cell line were pooled and concentrated ten times using a Microcon YM-10 filter (Millipore) with a Mw cut-off of 10 kDa that was operated at 14,000 x g for 30 minutes at 25°C to obtain adequate protein concentrations for gel-electrophoresis. Protein concentrations were quantified using the BCA protein Assay Kit (Pierce) and the lysates were stored at -20°C until use.

Gel-electrophoresisOf each sample, approximately 30 μg of cell surface protein lysate was separated by 1D

155

8

gel-electrophoresis on a NuPAGE 4-12% Bis-Tris Gel (Invitrogen). The obtained gel was fixed in 50% ethanol containing 3% phosphoric acid for 1 h, rinsed for three consecutive times in Milli-Q water (MQ) and subsequently stained with Coomassie-R250 O/N on a rocking platform to visualise the protein bands (Figure 1). After staining, the gel was washed vigorously with MQ to rid the Coomassie and stored in MQ at 4°C until further processing.

Figure 1 The Coomassie stained gel with the eight separated cell surface protein samples. This figure illustrates that the protein levels loaded onto the gel are equal across samples, which is of paramount importance when performing comparative proteomic studies based on spectral counting. Also, a different band pattern between the different cell lines, and, more importantly, different cell types can be observed.

In-gel digestionThe proteins were further processed into tryptic peptides according to the protocol described by Piersma et al., [Piersma SR. 2010] which was modified so that the pre-treatment phase of this protocol was applied to the whole gel in stead of to protein fractions. This allows for as good retrieval of peptides whilst reducing laboriousness of this procedure. [Pham TV. 2012] In brief, the entire gel was washed and dehydrated in 50 nM ammonium bicarbonate pH 7.9 (ABC) and 50 nM ammonium bicarbonate (ABC) containing 50% acetonitrile (ACN), respectively. Cysteine bonds were reduced by incubation with 10 mM DTT for 1 h at 56°C and alkylated with 50 mM iodoacetamide for 45 minutes at RT in the dark. The gel was washed, dehydrated, and washed again. Then, the gel was sliced into 10 bands per lane using a clean razor blade under keratin-free conditions. Thus, 10 protein fractions per cell line were obtained (Figure 1). Each fraction was further cut into 1 mm3 dimensioned gel pieces and washed in ABC/ACN. The slices were dried completely in a vacuum centrifuge (Thermo Fisher) and then incubated O/N with 100 µL of 6.25 ng/µL trypsin in 50 nM ABC at 25°C. The peptides were extracted in 100 µL of 1% formic acid once, followed by extraction

156

Appendix B

in 100 µL of 50% ACN in 5% formic acid twice. The combined volume of peptide fractions to be subjected to mass spectrometry by nano-liquid-chromatography tandem mass-spectrometry (nanoLC-MS/MS) was reduced to 50 µL in a vacuum centrifuge.

Nano-LC separationPeptide separation was performed using an Ultimate 3000 nanoLC system (Dionex LC-Packings, Amsterdam, The Netherlands) equipped with a 20 cm × 75 µm ID fused silica column custom packed with 3 µm 120 Å ReproSil Pur C18 aqua (Dr Maisch GMBH, Ammerbuch-Entringen, Germany). After injection, peptides were trapped at 6 µL/min (1.6% ACN in 0.05% formic acid (FA)) on a 1 cm × 100 µm ID precolumn packed with 5 µm ReproSil Pur C18 aqua. Peptides were separated in a 60 minutes gradient (8-32% ACN in 0.05% FA) at 300 nL/min followed by washing (72% ACN in 0.05% FA) and equilibration (4% ACN in 0.05% FA). The inject-to-inject time was 90 minutes.

Mass spectrometryIntact peptide MS spectra and MS/MS spectra were acquired on a LTQ-FT hybrid mass spectrometer (Thermo Fisher, Bremen, Germany) as described in detail in [Albrethsen J. 2010;Piersma SR. 2010]. Intact masses were measured at 50,000 resolution in the ICR cell. In parallel, following an FT pre-scan, the top 5 peptide signals (charge-states 2+ and higher) were submitted to MS/MS in the linear ion trap (3 amu isolation width, 30 ms activation, 35% normalized activation energy, Q value of 0.25 and a threshold of 5,000 counts). Dynamic exclusion was applied with a repeat count of 1 and an exclusion time of 30 seconds.

Protein identification and quantificationTo identify proteins from the acquired data, MS/MS spectra were searched against the human IPI database 3.62 (83,947 entries) using Sequest (version 27, rev 12) with a maximum allowed deviation of 10 ppm for the precursor mass and 1 amu for fragment masses. Methionine oxidation and cysteine carboxamidomethylation were allowed as variable modifications; two missed cleavages were also allowed. Scaffold 3.00.04 (Proteomesoftware, Portland, OR, USA) was used to organize the gel-slice data and to validate peptide and protein identifications. First, peptide identifications with a Peptide Prophet probability >95% were retained. Subsequently, protein identifications with 2 peptides or more in at least one of the samples and a Protein Prophet probability of >99% were retained. The proteins that contained similar peptides and could not be differentiated based on MS/MS analysis alone were grouped. For quantitative protein analysis spectral counting (the number of assigned MS/MS spectra for each identified protein) was used. For quantification across samples, the spectral counts were normalised to the sum of the spectral counts per biological sample. Differential analysis of samples was performed using the Beta-Binominal test as described previously. [Pham TV. 2010] Protein

157

8

identification and quantification details can be found in [Pham TV. 2010;Piersma SR. 2010]. The obtained dataset was exported to Excel for further use.

Data miningSubcellular protein localisations were verified using the Uniprot Knowledgebase (www.uniprot.org), searching under the header “GO annotation” for evidence of expression at the cell and/or plasma membrane. Protein-protein interactions (PPIs) were investigated using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) version 9.0 (www.string-db.org). [Szlarczyk D. 2011] For cluster and gene ontology (GO) analyses we used the Cytoscape platform for network analysis (www.cytoscape.org), using the plug-ins CluterONE version 0.91 (http://chianti.ucsd.edu/cyto_web/plugins/) for clustering and BINGO version 2.44 (http://www.psb.ugent.be/cbd/papers/BiNGO/) for the analysis of biological processes within our obtained networks, based on GO annotations. Additionally, verification of subcellular protein localisation was performed using Ingenuity pathway analysis software (IPA, Ingenuity Systems, Inc. CA, USA) and Biomart portal, version 7.0 (www.biomart.org).

Analysis of surface protein enrichmentTo verify enrichment of cell surface proteome in our biotinylated fraction, we compared our surface-enriched (biotinylated) protein lysates to intracellular (unbound) protein lysates and analysed the percentage of surface proteins in both fractions. The intracellular fractions of the osteosarcoma cell lines and human primary osteoblasts were pooled prior to work-up and mass spectrometric analysis. In this quality control, we only considered proteins with > 3 spectral counts per OS cell line. Proteins with a transmembrane and/or a signal peptide annotation using the Biomart portal version 7.0 (www.biomart.org) were considered predicted surface proteins. Within our surface protein enriched fraction, we retrieved 37% predicted surface proteins and compared to the whole lysate (10%). Thus, we indeed enriched our protein samples with surface molecules.

Figure 2 Enrichment of surface proteins in whole lysate (intracellular fraction) and surface enriched (biotinylated) fraction. Analysis of the intracellular proteome also allowed for a comparison of intracellular EPHA2 expression in OS cell lines and primary human osteoblasts. The average spectral counts for intracellular EPHA2 in the OS cell lines was 11 compared to 0 in the human primary osteoblasts. Thus, apart from differential surface expression of EPHA2, the total protein expression of EPHA2 is also differentially regulated between OS cells and healthy bone cells.

158

Appendix B

ReferencesAlbrethsen J, Knol JC, Piersma SR, Pham TV, de Wit M, Mongera S, Carvalho B, Verheul HM, Fijneman RJ, Meijer

GA, Jimenez CR (2010) Subnuclear proteomics in colorectal cancer: identification of proteins enriched in the nuclear matrix fraction and regulation in adenoma to carcinoma progression. Mol Cell Proteomics 9: 988-1005

de Wit M., Jimenez CR, Carvalho B, Belien JA, Delis-van Diemen PM, Mongera S, Piersma SR, Vikas M, Navani S, Ponten F, Meijer GA, Fijneman RJ (2012) Cell surface proteomics identifies glucose transporter type 1 and prion protein as candidate biomarkers for colorectal adenoma-to-carcinoma progression. Gut 61: 855-864

Pham TV, Piersma SR, Oudgenoeg G, Jimenez CR (2012) Label-free mass spectrometry-based proteomics for biomarker discovery and validation. Expert Rev Mol Diagn 12: 343-359

Pham TV, Piersma SR, Warmoes M, Jimenez CR (2010) On the beta-binomial model for analysis of spectral count data in label-free tandem mass spectrometry-based proteomics. Bioinformatics 26: 363-369

Piersma SR, Fiedler U, Span S, Lingnau A, Pham TV, Hoffmann S, Kubbutat MH, Jimenez CR (2010) Workflow comparison for label-free, quantitative secretome proteomics for cancer biomarker discovery: method evaluation, differential analysis, and verification in serum. J Proteome Res 9: 1913-1922

Szklarczyk D, Franceschini A, Kuhn M, simonovic M, Roth A, Minguez P, Doerks T, Stark M, Muller J, Bork P, Jensen LJ, von Mering C (2011) The STRING database in 2001: functional interaction networks of proteins, globally integrated and scored. Nucleid Acids Res 39: D561-D568

Supplementary File 6.2: in silico analysis of highly upregulated proteins on osteosarcoma. To investigate protein-protein-interactions (PPIs), protein clustering, networks, subcellular localisation of the proteins in our hit list and biological processes wherein the proteins are involved, we used slightly less stringent criteria for protein selection than the criteria that were applied to obtain the proteins included in Table 6.1 of the main manuscript. We filtered proteins based on their (1) expression in 4/5 OS cell lines, (2) >3-fold upregulation on OS cells vs. hp-OBs and (3) with p<0.05 significance. This retrieved 123 highly upregulated proteins from our dataset which were entered into STRING 9.0 (www.string-db.org) using default settings. Figure 1 shows a PPI map of physical interactions between upregulated surface proteins on OS. In this map, we can identify 6 high density clusters, which are enlarged and depicted schematically in the inlays in Figure 1. These clusters contain (1) an integrin network connected to several growth factor receptors (MET, FGFR1 and -2); (2) a cluster containing several Ephrins and Ephrin Receptors (EFNB2 and -3, EFNA5, EPHA2 and EPHB2 and -4); (3) a cluster of transporter molecules; (4) a cluster containing both the insulin (INSR) and a insulin growth factor receptor (IGF-1R) and a tumour necrosis factor receptor (TNFRSF10A); (5) a slightly comparable cluster containing another member of the tumour necrosis factor receptors (LTBR) plus a TNF ligand (CD70) alongside the insulin receptor; and, (6) a small cluster containing Wnt-Receptors (FZD7, LRP6 and ROR2). The PPI map was exported into CytoScape software (www.cytoscape.org) and the above-mentioned clusters were isolated (ClusterONE) for functional analysis (BINGO). For each cluster, GO-annotations for biological processes were assigned and the number of involved proteins and corrected p-values were calculated to assess enrichment of biological functions in the clusters compared to

159

8

the entire human proteome (Supplementary Table 6.4 provides GO annotations of all 6 identified clusters). Within cluster 1 the proteins are associated with cell surface receptor signalling (63%), integrin signalling (19%), cell adhesion (50%), cell differentiation (44%) and cell migration (19%). Proteins in cluster 2 are involved in transmembrane receptor protein kinase signalling (38%) and Ephrin signalling in particular (23%), as well as cell differentiation (46%), angiogenesis (15%) and, interestingly, neurogenesis (38%). Cluster 3 consists of various transporters responsible for organic acid transport (100%), amino acid transport (67%) and ion transport (50%) across the cell membrane. The proteins in cluster 4 are involved in signal transduction again (67%), particularly in the insulin receptor pathway (33%). The HLA molecules in combination with the IGF-1R play a role in the immune response (50%). Cluster 5 is involved in signal transduction (67%), cell migration (33%) and proliferation (33%). Finally, cluster 6 is a small network of three receptors of the Wnt-signalling pathway and are thus involved in Wnt-signalling (100%), in tissue development, particularly neurogenesis (67%) and cell proliferation (67%).

Additionally, to visualise cell surface localisation of these proteins, we performed an Ingenuity Pathway Analysis (Figure 2), which confirmed that the vast majority of identified proteins are indeed cell surface molecules. To verify the subcellular localisation and investigate functional networks in the set of highly upregulated surface proteins we built an

Figure 1 Protein-protein Interaction map of 123 significantly upregulated surface proteins on OS compared to healthy bone cells. This PPI map was generated by STRING 9.0. Thicker lines represent stronger associations. The squares highlight functional clusters identified within the STRING network, based on molecule density. The inlays depict these functional clusters schematically (analysed with ClusterONE in Cytoscape).

160

Appendix B

IPA network of our entries. IPA generates a network of 53 proteins and consecutively shows the subcellular localisation of the connected proteins.

Figure 2 Ingenuity Pathway Analysis (IPA) of 123 significantly upregulated surface proteins on OS compared to healthy bone cells.The measure of fold-change upregulation on OS cells compared to hp-OBs is indicated in a colour key; red for > 100-fold upregulation, dark pink for 41 to 99-fold and light pink for 21 to 40-fold. Solid arrows represent direct interactions and dashed arrows represent indirect interactions between the proteins.

161

8

The vast majority (51) of these 53 proteins are indeed annotated to be localised in the plasma membrane, supporting our proteomic data. In the network, one protein is reported to be in the extracellular space, and one, interestingly, is considered a nuclear protein. The remaining 70 proteins could not be connected to the generated network. Of the remaining unconnected proteins, 5 were reported to be in the extracellular space, 4 in the cytoplasm and for 12 proteins the subcellular localisation was unknown. The full list is provided in Supplementary Table 6.5, including subcellular localisation and molecular functions as annotated by IPA.

Supplementary figure 6.1 Tissue micro array (TMA) layouts and staining consensus chart. (A) Overview of the layouts of the two tissue micro arrays analysed for EPHA2 staining. Corresponding clinical data is found in Table 6.2 of the main manuscript. (B) Consensus chart showing three exemplary tissue cores per category (positive, moderate, weak or negative) used by the observers as guide while scoring EPHA2 staining.

Acknowledgement

an expression of thanks or gratitude;

a token of appreciation

Acknowledge - to express obligation, thanks, gratitude or appreciation for

Support - to give aid or courage to; to give strength to

Curriculum Vitae

List of publications

168

C U R R I C U L U M V I TA E

Jantine Posthuma de Boer was born in on the 9th of September 1982 in Blackburn (UK) as a twin to her brother Pier. After attending the bilingual department at the OSG Wolfert van Borselen in Rotterdam, she started her study Medicine in the year 2000 at the Erasmus University in Rotterdam. For her Masters degree she performed a comparative retrospective study on the treatment outcomes of solitary posterior cruciate ligament tears, supervised by dr. M. Heijboer, department of orthopaedic surgery, Erasmus MC, Rotterdam. During her final rotation at the department of orthopaedic surgery of the Erasmus MC in Rotterdam she participated in research studying radiological parameters in Perthes’ Disease, supervised by dr. G. Bessems. This led to a poster presentation at an international conference.

After completing medical school she started to work as a junior resident orthopaedic surgery at the VU University Medical Center in Amsterdam in 2007. In 2008 she started a PhD project at the department of orthopaedic surgery (supervised by prof. dr. B.J. van Royen and dr. M.N. Helder) in collaboration with the department of paediatric oncology (prof. dr. Gert-Jan Kaspers) and Medical Oncology (dr. V.W. van Beusechem). The scientific research conducted during this period led to the production and publication of this thesis as well as to two scientific prices for best oral presentations at national and international conferences. In 2012 she started her residency orthopaedic surgery at the department of general surgery of the Spaarne Ziekenhuis in Hoofddorp. Presently she is a registrar at the department of orthopaedic surgery of the MCA Alkmaar (supervisor: dr. B.J.Burger) and anticipates to complete her training as orthopaedic surgeon in 2018.

169

8

L I S T O F P U B L I C AT I O N S

Publications

J. PosthumaDeBoer, B.J. van Royen, M.N. HelderMechanisms of therapy resistance in osteosarcoma: a reviewOncology Discovery 2013 Dec 31; 1:8

J. PosthumaDeBoer, S.R. Piersma, T.V. Pham, P.W. van Egmond, J.C. Knol, A.M. Cleton-Jansen, M.A. van Geer, V.W. van Beusechem, G.J.L. Kaspers, B.J. van Royen, C.R. Jiménez, M.N. HelderSurface proteomic analysis of osteosarcoma identifies EPHA2 as receptor for targeted drug deliveryBr J Cancer 2013 Oct 15; 109(8): 2142-54

J. PosthumaDeBoer, P.W. van Egmond, M.N. Helder, R.X. de Menezes, A.M. Cleton-Jansen, J.A.M. Beliën, H. Verheul, B.J. van Royen, G.J.L. Kaspers, V.W. van BeusechemLoss-of-function RNA interference screen of the human kinome identifies JIP1 as regulator of doxorubicin response in osteosarcomaOncotarget 2012 Oct;3(10):1169-81

J. PosthumaDeBoer, T. Würdinger, H.C.A. Graat, V.W. van Beusechem, M.N. Helder, B.J. van Royen, G.J.L. KaspersWEE1 inhibition sensitizes Osteosarcoma to RadiotherapyBMC Cancer 2011, 11:156

J. PosthumaDeBoer, M.A. Witlox, G.J.L. Kaspers, B.J. van RoyenMolecular Alterations as Target for Therapy in Metastatic Osteosarcoma: A Review of Literature Clin Exp Metastasis (2011) 28:493–503

M.P. van de Kerkhove, J. PosthumaDeBoer, T.U. Jiya, H.A.H. Winters, R. SaoutiPelvic Reconstruction with a Free Vascularised Distal Femur for Revision Total Hip Arthroplasty. A Case ReportJournal of Arthroplasty 2012 Mar;27(3):493.e19-22. [Epub (2011) Jul 1]

Y. Könst, J. PosthumaDeBoer, R. SaoutiFractuur van een Ceramic Coated Implant knie protheseNTvO Vol 17, Nr 3, september 2010

J. PosthumaDeBoer, H.C.A. Graat, J. Bras, R. SaoutiSmall Cell Osteosarcoma of a Toe Phalanx: a Case Report and Review of LiteratureJournal of Orthopaedic Surgery and Research 2010, 5:36

170

Presentations

J. PosthumaDeBoer, P.W. van Egmond, M.N. Helder, R.X. de Menezes, A.M. Cleton-Jansen, J.A.M. Beliën, H.M.W. Verheul, B.J. van Royen, G.J.L. Kaspers, V.W. van BeusechemJIP1-inhibition sensitises osteosarcoma to doxorubicinNOV, jaarcongres 2013, Amsterdam

J. PosthumaDeBoer, P.W. van Egmond, M.N. Helder, R.X. de Menezes, A.M. Cleton-Jansen, J.A.M. Beliën, H. Verheul, B.J. van Royen, G.J.L. Kaspers, V.W. van BeusechemLoss-of-function RNA interference screen of the human kinome identifies JIP1 as regulator of doxorubicin response in osteosarcomaEuropean Orthopaedic Research Society (EORS) 2012, 20th annual meeting Amsterdam

J. PosthumaDeBoer, S.R. Piersma, T.V. Pham, V.W. van Beusechem, C.R. Jiménez, G.J.L. Kaspers, M.N. Helder, B.J. van RoyenSurface proteomic analysis of osteosarcoma identifies EPHA2 as receptor for targeted drug delivery*European Orthopaedic Reseach Society (EORS), 20th annual meeting Amsterdam* Winner Best Oral presentation - basic science Award

J. PosthumaDeBoer, S.R. Piersma, T.V. Pham, V.W. van Beusechem, C.R. Jiménez, G.J.L. Kaspers, M.N. Helder, B.J. van RoyenSurface proteomic analysis of osteosarcoma identifies EPHA2 as receptor for targeted drug delivery*NOV, jaarcongres 2012, Den Haag* Winner Rik Huiskes Prijs

J. PosthumaDeBoer, T. Würdinger, H.C.A. Graat, V.W. van Beusechem, M.N. Helder, B.J. van Royen, G.J.L. KaspersWEE1 inhibition by PD0166285 sensitizes osteosarcoma cells to irradiation-induced cell deathEuropean Musculo-Skeletal Oncology Society (EMSOS) 2011, 24th annual meeting Gent, België

J. PosthumaDeBoer, T. Würdinger, H.C.A. Graat, V.W. van Beusechem, M.N. Helder, B.J. van Royen, G.J.L. KaspersWEE1 inhibition sensitizes Osteosarcoma to RadiotherapyNOV, jaarcongres 2011, Groningen

Poster presentations

J. PosthumaDeBoer, S.R. Piersma, T.V. Pham, V.W. van Beusechem, C.R. Jiménez, G.J.L. Kaspers, M.N. Helder, B.J. van RoyenSurface proteomic analysis of osteosarcoma identifies EPHA2 as receptor for targeted drug deliveryORS, 2012 Annual Meeting of the Orthopaedic Research Society, San Francisco, CA, USA

J. PosthumadeBoer, J.M.J.H. Bessems, M. Reijman, J.A.N. VerhaarAlternative reference lines for Hilgenreiner’s lineNOF, 54th Nordic Orthopaedic Federation Congress 2008, Amsterdam

171

8

172

Towards targeted treatment for osteosarcoma · 2015. 1. 12. · provides an overview of (pre)clinical research efforts that were endeavoured upon in the past decade to exploit specific

Documents