Tumour Survival Signals and Epigenetic Gene Silencing in Multiple Myeloma: Implications for Biology and Therapy

(1)

UNIVERSITATISACTA UPSALIENSIS

Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1071

Tumour Survival Signals and Epigenetic Gene Silencing in Multiple Myeloma

Implications for Biology and Therapy

CHARLOTTE FRISTEDT DUVEFELT

ISSN 1651-6206 ISBN 978-91-554-9159-8

(2)

Dissertation presented at Uppsala University to be publicly examined in Fåhraeussalen, Rudbeck Laboratory, Dag Hammarskjölds väg 20, Uppsala, Wednesday, 25 March 2015 at 09:15 for the degree of Doctor of Philosophy (Faculty of Medicine). The examination will be conducted in English.

Abstract

Fristedt Duvefelt, C. 2015. Tumour Survival Signals and Epigenetic Gene Silencing in Multiple Myeloma. Implications for Biology and Therapy. Digital Comprehensive

Summaries of Uppsala Dissertations from the Faculty of Medicine 1071. 45 pp. Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-554-9159-8.

This thesis is focused on multiple myeloma (MM), a haematological malignancy that still remains incurable. The pathogenesis of MM is not fully understood and there is a large intra- tumour and interclonal genetic variation in MM patients. One of the most challenging areas in MM research is to find mechanisms for initiation and progression of MM, but also to overcome the arising resistance to therapy.

In paper I, a signature of under-expressed genes in MM was found to significantly correlate with already defined Polycomb target genes. In selected genes from the profile we found an enrichment of H3K27me3, a repressive mark catalysed by Polycomb repressive complex 2 (PRC2), in MM patients and MM cell lines. Treatment with LBH589 (HDAC inhibitor) and DZNep (methyltransferase inhibitor) reactivated the H3K27me3 target genes and induced apoptosis in MM cell lines. LBH589 reduced tumour load and increased overall survival in the 5T33MM mice. These results suggest an important role for Polycomb complex in MM development and highlight PRC2 as a drug target in MM.

In paper II, the insulin-like growth factor type 1 receptor tyrosine kinase (IGF-1RTK) inhibitor picropodophyllin (PPP) in combination with LBH589 synergistically inhibited cell proliferation and enhanced the apoptotic effect in MM. Since the bone marrow microenvironment has an important role in MM disease and also contributes to drug-resistance, we therefore evaluated the drug combination in the immunocompetent 5T33MM murine model.

The drug combination significantly prolonged the survival of the 5T33MM mice compared to single drug treatment. We conclude that the combination of PPP and LBH589 has a therapeutic potential in MM.

In paper III, the role of the cellular inhibitor of apoptosis protein 2 (cIAP2) was evaluated in MM cells harbouring TRAF3 deletion/mutation. By overexpressing cIAP2 in these cells we found an increased resistance to proteasome inhibitors. cIAP2 over-expression by lentiviral constructs led to decreased caspase activation, activation of the canonical NF-κB pathway, and down-regulation of tumour suppressor genes and genes that contribute to apoptosis. Supporting the role of cIAP2 mediated drug-resistance, we here demonstrate that inhibiting cIAP2 using an IAP antagonist, increased the sensitivity to the proteasome inhibitor, bortezomib.

Keywords: Multiple myeloma, Polycomb, apoptosis, cIAP2, IAP-inbibitors, proteasome inhibitors, LBH589, PPP, DZNep

Charlotte Fristedt Duvefelt, Department of Immunology, Genetics and Pathology, Rudbecklaboratoriet, Uppsala University, SE-751 85 Uppsala, Sweden.

ISBN 978-91-554-9159-8

urn:nbn:se:uu:diva-242571 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-242571)

(3)

To my lovely family

(4)

(5)

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Kalushkova, A.*, Fryknäs, M.*, Lemaire, M., Fristedt, C., Agarwal, P., Eriksson, M., Deleu, S., Atadja, P., Österborg, A., Nilsson, K., Vanderkerken, K., Öberg, F., and Jernberg- Wiklund, H. (2010) Polycomb Target Genes Are Silenced in Multiple Myeloma. PLoS ONE 5(7): e11483 doi:10.1371/

journal.pone.0011483

II Lemaire, M.*, Fristedt, C.*, Agarwal, P., Van Valckenborgh, E., De Bruyne, E., Österborg, A., Atadja, P., Larsson, O., Axelson, M., Van Camp, B., Jernberg-Wiklund, H.°, and Vanderkerken, K.° (2012) The HDAC Inhibitor LBH589 Enhances the Antimyeloma Effects of the IGF-1RTK Inhibitor Picropodophyllin. Clinical Cancer Research 18(8):2230-2239.

doi:10.1158/1078-0432.CCR-11-1764

III Fristedt Duvefelt, C.*, Lub, S.*, Agarwal, P., Arngården, L., Van Valckenborgh, E., Vanderkerken, K., and Jernberg- Wiklund, H. (2014) Increased resistance to proteasome inhibitors in multiple myeloma mediated by cIAP2 – implications for a combinatorial treatment. Manuscript.

* Equally contributing authors

° Co-last authors

Reprints were made with permission from the respective publishers.

(6)

Publications not included in this thesis

Kharaziha, P., De Raeve, H., Fristedt, C., Li, Q., Gruber, A., Johnsson, P., Kokaraki, G., Panzar, M., Laane, E., Österborg, A., Zhivotoysky, B., Jernberg-Wiklund, H., Grandér, D., Celsing, F., Björkholm, M., Vanderkerken, K., Panaretakis, T. (2012) Sorafenib Has Potent Antitumor Activity against Multiple Myeloma In Vitro, Ex Vivo, and In Vivo in the 5T33MM Mouse Model Cancer Research 72(20):5348-5362. Doi:

10.1158/0008-5472.CAN-12-0658

(7)

Abbreviations

AID Apaf-1 BH BIR BM BMSC CARD cIAP1 cIAP2 Cyt C DAP DISC DR ECM EZH2 FADD FGFR3 IAP Ig IGF-1 IGF-1R IgH IKK IκB IL-6 IMiD JAK mAb MAPK MGUS MM MMSET NF-κB NIK PcG

Activation-Induced Deaminase Adaptor protease activating factor-1 Bcl-2 homology domains

Baculoviral IAP repeat Bone marrow

Bone marrow stromal cells

Caspase activation and recruitment domain Cellular inhibitor of apoptosis 1

Cellular inhibitor of apoptosis 2 Cytochrome C

Death-associated protein

Death-inducing signalling complex Death receptor

Extracellular matrix

Enhancer of zeste homolog 2 Fas associated death domain Fibroblast growth factor receptor Inhibitors of apoptosis

Immunoglobulin

Insulin-like growth factor

Insulin-like growth factor receptor Immunoglobulin heavy-chain Inhibitor of IκB kinase Inhibitors of NF-κB Interleukin-6

Immune modulatory drug Janus kinase

Monoclonal antibody

Mitogen-activated protein kinases

Monoclonal gammopathy of undetermined significance Multiple myeloma

Multiple myeloma SET domain

Nuclear factor κ-light-chain-enhancer of activated B-cells NF-κB inducing kinase

Polycomb group

(9)

PCL PI-3K PPP PRC RING SCID SMM STAT TNF TRADD

Plasma cell leukaemia

Phosphatidylinositol-4,5-bisphosphate 3-kinase Picropodophyllin

Polycomb repressive complex Really interesting new gene

Severe combined immunodeficiency Smoldering myeloma

Signal transducer and activator of transcription Tumour necrosis factor

TNFR1-associated death domain

(10)

(11)

Multiple myeloma

Multiple myeloma (MM) is a genetically heterogeneous B-cell malignancy that is characterized by a clonal accumulation of malignant plasma blast / plasma cells which produce a large amount of monoclonal proteins in the bone marrow (BM).^1,2 Although the origin of the malignant MM cells still remains unclear, the immunoglobulin (Ig) isotype is mostly IgG or IgA demonstrating that the MM cells are predominantly derived from a post- switch B cell.³

MM is thought to be preceded by a pre-malignant condition without clinical symptoms called monoclonal gammopathy of undetermined significance (MGUS), that in turn can progress to smoldering MM (SMM), an asymptomatic precursor of active MM and finally, to symptomatic MM,⁴ Figure 1. MGUS is reported to evolve to MM at a rate of 1% per year,⁵ however, not all cases of MGUS develop into MM.

MM is a malignancy restricted to the BM and the BM is thought to be essential for the development of MM until the late stages of the disease. In later stages of the disease the MM cells may obtain additional genetic abnormalities enabling the tumour cells to be independent of the interactions with the BM now gaining the capacity to survive outside the BM and evolve into plasma cell leukaemia (PCL).^1,6 In the BM, the malignant MM cells displace the normal blood-forming cells causing anaemia, one of the symptoms of MM. Other clinical symptoms and signs of MM include bone pain and fractures, renal failure and hypercalcaemia. Worldwide the incidence of MM is 1-4 per 100 000 people per year⁷ and in Sweden approximately 500-600 people are diagnosed with MM annually. The introduction of the current therapies such as the proteasome inhibitor bortezomib and the immune modulatory drugs (IMiDs) thalidomide and lenalidomide has improved the clinical outcome of MM patients and increased the median survival to more than seven years.^8-10 Unfortunately, the current therapies have so far not proven to be curative for the MM patients^8,9 and the development of drug-resistant clones leads to recurrence of the tumour. Obviously, new treatment approaches are necessary to improve the outcome for MM patients.

(12)

Genetic aberrations in MM

Chromosomal abnormalities resulting in altered expression and abnormal functions of genes that has an important role in cell growth and differentiation in normal cells are important hallmarks of cancer. MM is a heterogeneous disease with a large variation of genetic alterations but heterogeneity is also described in the different response and survival of MM patients although receiving the same treatment.¹ It is known that genetic alterations, epigenetic alterations and/or abnormal microRNA expression contribute to MM pathogenesis.^1,11-13 However, a common pattern of alterations that occurs in all MM tumours has so far not been found.

MM can be divided in two different groups; the hyperdiploid and non- hyperdiploid MM, based on the type of chromosomal alterations observed.

The hyperdiploid group is characterized by trisomies of odd-numbered chromosomes, while the non-hyperdiploid group is characterized by translocations to the 14q32 loci as well as monosomy of 13 and gains of 1p.¹⁴ Hyperdiploidy and t(11;14) have been associated with a favourable prognosis for MM patients, while t(4;14) or del(17p) have been associated to a poor prognosis.¹⁴

A diversity of chromosomal loci and partner genes involved in Ig heavy- chain (IgH) translocations may provide indications of an early immortalizing event. Four different loci are frequently involved in MM tumours: 11q13 (cyclin D1), 4p16 (fibroblast growth factor receptor [FGFR3] and multiple myeloma SET domain [MMSET], 16q23 (c-maf) and 6p21 (cyclin D3).^15,16 Importantly, the prevalence of IgH translocation is nearly as high in MGUS as in MM. These translocations are therefore referred to as early genetic events in MM transformation, but still not sufficient to yield the full malignant potential.¹⁵ Several secondary genetic events occur in MM e.g.

activating mutations of KRAS and NRAS, inactivating mutations of TP53, RB1 and PTEN.^17,18 For long, c-myc has been considered a late event in the MM pathogenesis.¹⁹ Recently however, c-myc was found to be activated during the transition from MGUS to MM, and thus suggested to constitute a universal and important step in the transformation process.²⁰ In addition, genetic aberrations that involve activating mutations in the NF-κB signalling pathway have been described and been suggested to be important during pathogenesis of MM. In a study by Keats et al²¹ a majority of the MM patient samples and MM cell lines were shown to display elevated gene expression in NF-κB target genes, and this expression is often associated with alterations of NIK, TRAF3, CYLD, cIAP1/cIAP2, NFKB1 or NFKB2.

(13)

MGUS Smouldering MM IntramedullaryMM Extramedullary MM Germinal-centre

B cell

Karyotypic instability

Primary Ig translocation

Secondary Ig translocation

13q14 deletion / monosomy

Activating mutations (NRAS, KRAS, FGFR3)

TP53 mutations

Figure 1. Schematic model of the MM stages and the molecular pathogenesis. MM is thought to develop from a normal germinal-centre B cell. In most cases, MM is preceded by the pre-malignant condition MGUS. The progression may in some cases include the stage of smouldering MM. Initially, intramedullary MM is restricted to the BM. After some time, the MM tumour cells obtain the capacity to grow in extramedullary places e.g. in the blood. Several oncogenic events are now identified in MM i.e. karyotypic instability, primary as well as secondary Ig translocations, deletion of chromosome 13q, activating mutations of FGFR3, NRAS or KRAS2 and TP53 mutations. Figure is adapted from Kuehl et al 2002.²²

MM models

A better understanding in the MM biology and advances in bone marrow transplantation have led to significant advances in the overall survival of MM patients.²³ By the use of improved laboratory based models, new agents may be pre-clinically evaluated and introduced into clinical practice.^23,24

In vitro models

Purified human MM cells and authentic MM cell lines have for long been used to understand the biology of MM. Historically it has also been proven very difficult to establish cell lines from plasma cell neoplasms. Therefore in several studies EBV+ non-malignant lymphoblastoid cell lines, highly irrelevant to the biology of MM, have been used in studies of MM biology.

The establishment of MM cell lines proved to be more successful when the growth requirement of the MM cells and dependence of the bone marrow

(14)

interaction could be defined.^25-27 During in vitro growth some of these cell lines grow independent of these supportive factors.²⁸ Today, approximately 80 authentic MM cell lines and corresponding to the primary patient sample now constitute a relevant in vitro model for MM.^29-33 Primary MM cells from patients are, still to date, difficult to grow both in vitro as well as in vivo.³⁴

In vivo models

A functional in vivo model of MM must fulfil a number of characteristic features of the human MM disease i.e. similarity in tumour cell biology and genetics. Preferably the tumour is localized to relevant organs of the mouse i.e. the bone marrow. This is important for studying tumour initiating capacity or new treatment options targeting the tumour and the microenvironment. Several in vivo models fulfilling these criteria have been described and are used to study the biology of MM.³⁵ There are two main models that are used in MM, the 5TMM models and the NOD/SCID models.³⁶ The use of transgenic mice driven by myc activation at late stages of B cell differentiation have also been described to be a highly relevant in vivo model for MM.^37,38

5TMM murine myeloma model

The 5T mouse model of MM is a syngeneic model originally established from spontaneously occurring MM in C57BL/KaLwRij mice.³⁹ Although the 5TMM is a model of murine MM, the model is well representative to the human disease both on cellular and molecular levels e.g. localization of the MM cells in the bone marrow, induction of angiogenesis and development of an M component.^35,40,41 Some advantage with these models include that they are dependent on the microenvironment and are immune-competent and thus can be used to study tumour-bone marrow interactions as well as immune- therapeutic strategies.³⁶ Of the 5TMM models, 5T2MM and 5T33MM models are the most well characterized and widely used.⁴² The 5T2MM is a model characterized slow/moderate growth and osteolytic lesions, while the 5T33MM is an aggressive form with rapid tumour growth.³⁵ For investigations of the pathobiology and for preclinical studies of human MM the 5TMM models are considered a highly relevant in vivo model because of the simple and reproducible continuation of the model and characteristics of the 5TMM models similar to that in human MM.³⁵ The limitation with these models is that the two models, 5T33MM and 5T2MM, currently in use might not fully reflect the heterogeneity of the human MM disease.³⁶

SCID-models

Point mutations in chromosome 16 in the CB-17 inbred mouse strain give rise to the SCID (severe combined immunodeficiency) mouse. In the

(15)

NOD/SCID models, either human MM cells or MM cell lines are injected intravenously or locally in the mice. The injected MM cells home to the BM and cause osteolysis but the injected cells can also grow extramedullary.

Since the mice in these models are immune-compromised the interaction between the MM cells and the BM cannot be studied therefore they can never reflect the complexity of the MM disease.³⁶ Development of this model has resulted in the SCID-hu model. SCID-hu model is a human MM model in SCID mice where the SCID mice serve as host for human embryonal bone grafts to create a human BM microenvironment which makes it possible for both human primary cells and human cell lines to grow.^43,44 The SCID-hu model has enabled studies of the interaction between the primary MM cells and the human bone marrow microenvironment⁴⁵ but major disadvantages are the ethical concerns by use of foetal bones.

The Vk*MYC mouse

In the Vk*MYC mouse model the misdirection of the Activation-Induced Deaminase (AID) activity to a conditional MYC transgene leads to indolent MM in mice.³⁷ The AID activity is required for Ig genes of B-cells in the germinal centre to undergo class switch recombination and somatic hypermutation (SHM).³⁷ AID has also been shown to be linked to oncogene mutations and IgH/MYC translocations.^46,47 In this MM model, an exon in the Vk*MYC vector is mutated to a stop codon resulting in no expression of MYC protein. During SHM, different mechanisms introduce mutations that sporadically revert the stop codon and MYC can be translated and expressed.³⁷ Thus, in the transgenic Vk*MYC mouse conditional activation of MYC occurs sporadically through the mechanisms of somatic hypermutation in post germinal centre B-cells.

These mice develop plasma cell tumours representative of the human MM both the biological and the clinical features. This model may be used for studying the biology in MM in a native environment.³⁷

(16)

Regulation of survival and apoptotic pathways in MM

In a normal cell the processes of cell division and growth are strictly controlled and there are some specific characteristics that are crucial for the organism to function properly. A balance between proliferation and apoptosis is required for homeostasis. In cancer cells this balance is disturbed and cells can therefore grow unrestricted and respond poorly to essential signals that control cellular growth and apoptosis. In fact, both uncontrolled growth and evasion of apoptosis are two true hallmarks of cancer cells.⁴⁸

The microenvironment and selected growth and survival factors in MM

The microenvironment of the malignant MM cells consists of endothelial cells, bone marrow stromal cells (BMSC), osteoclast, osteoblasts and various hematopoietic cells. The non-cellular compartment consists of extracellular matrix (ECM) proteins and a liquid milieu containing growth factors, cytokines and chemokines.^22,49 MM progresses almost exclusively in the bone marrow and signals from the microenvironment plays a critical role in maintaining the growth, migration and survival of the tumour cells.^16,50 The interaction between the MM cells and the cells within the BM microenvironment results in the production of several cytokines and growth factors that are important factors regulating growth and survival of MM cells.^27,51-53

One of the first growth factors described for MM was interleukin-6 (IL-6) that is mainly produced by the BM stromal cells. IL-6 induces proliferation, inhibits apoptosis and may also override the apoptotic signals mediated by cytotoxic drugs.⁵⁴ IL-6 mediates its signals, when binding to its receptor, through activation of the JAK/STAT,⁵⁵ Ras/MAPK⁵⁶ and PI-3K/AKT⁵⁷ signalling pathways.

Insulin-like growth factor-1 (IGF-1) is an important growth factor in MM that has a key role in tumour growth. IGF-1 stimulates proliferation of both IL-6-dependent⁵⁸ and -independent⁵⁹ MM cell lines. Via binding and activation to its receptor (IGF-1R), IGF-1 activates both PI-3K/AKT and

(17)

MAPK signalling pathways, the PI-3K/AKT pathway functioning as the major regulator of both proliferation and apoptosis with only a small contribution from the MAPK cascade.^59,60 Overexpression of IGF-1 and its receptor IGF-1R is a frequent feature of MM^51,61-63 and higher expression levels of IGF-1R in MM cells has in some reports been correlated to poor patient prognosis.^64,65 IGF-1 has also been implicated in drug resistance. The proteasome inhibitor, bortezomib, included in the current therapies for MM has, although not being a curative treatment, substantially prolonged the survival for MM patients.^8,9 Resistance to bortezomib in patients eventually occurs. Numerous mechanisms have been suggested to underlie bortezomib resistance or be associated with the bortezomib resistant phenotype including an enhanced IGF-1 signalling both caused by increased IGF-1 secretion and IGF-1R activation.⁶⁶ By combining an IGF-1R tyrosine kinase inhibitor with bortezomib the anti-MM effect was enhanced and bortezomib resistance in vivo could be overcome indicating that dysregulation of IGF-1/IGF-1R pathway may be involved in bortezomib resistance in MM.⁶⁶

IGF-1R signalling does not seem crucial for adult normal tissue, even though the IGF-1R is ubiquitously expressed. Selective inhibition of IGF-1R triggers both anti-proliferative and pro-apoptotic events in human MM cells.

This has encouraged the development of ways to inhibit IGF-1R signalling for clinical use in MM.^63,67,68 Unfortunately, toxicity issues regarding the non-specific inhibition of the insulin receptor has for long hindered the development of IGF-1R inhibitors.³⁶ However, new more selective drugs have been developed. As of today, several inhibitors of IGF-1R signalling in cancer are currently in clinical development and can be divided in three groups, e.g. IGF-1R tyrosine kinase inhibitors, monoclonal antibodies (mAb) against IGF-1R ligand and mAbs against IGF-1R⁶⁹ of which the mAb against IGF-1R are the most advanced in the clinical development.⁷⁰ Although, only targeting the IGF-1R may not be curative for MM or other cancer diseases but the results so far strongly suggest that further studies on specific IGF-1R inhibitors in combination with other drugs will yield encouraging results.⁶⁹

Inhibitors of apoptosis proteins

The inhibitors of apoptosis (IAP) proteins were first discovered in baculovirus.⁷¹ The family of IAP proteins have been shown to be involved in several processes such as cell death, cell cycle, inflammation, immunity and migration.⁷² The IAP family are now composed of eight human homologues, XIAP (X-chromosome linked IAP), cIAP1 (cellular IAP 1), cIAP2, (cellular IAP 2), NAIP (baculoviral IAP repeat-containing protein 1), ILP2 (IAP like protein), survivin, BRUCE and livin.⁷³ All IAPs contain one to three baculoviral IAP repeat (BIR) domains⁷⁴ which is essential for the interaction

(18)

with a number of pro-apoptotic factors. Some of the members of the IAP family also contain the Really Interesting New Gene (RING) domain which possesses E3 ligase activity resulting in ubiquitination and proteasomal degradation of substrates.⁷⁵ The cIAP1 and cIAP2 also contains a caspase activation and recruitment domain (CARD) which mediate protein-protein interactions⁷⁶ but still the function of the CARD domain is not completely understood in cIAP1 and cIAP2.⁷⁷

Currently, the most well characterized IAP is the XIAP. XIAP is also the one that retains the strongest anti-apoptotic properties.⁷⁸ XIAP have been described to inhibit apoptosis through inhibition of caspases-3, -7 and -9,⁷⁹ and also via the RING domain by targeting pro-apoptotic proteins for proteasomal degradation.⁷⁵ While cIAP1 and cIAP2 have been reported to bind to caspases, the anti-apoptotic effect of cIAP1 and cIAP2 seems not associated with the direct binding and sequestering of caspase activity.^78,79 Instead they have an effect on caspase activation, through their RING domain, by targeting caspases for proteasomal degradation⁸⁰ and through supporting survival by promoting activation of the canonical (classical) NF- κB pathway.⁸¹

IAP proteins are often found to be overexpressed in cancer.⁸² In MM patients that developed drug-resistance, over-expression of IAPs e.g.

surviving, cIAP1, cIAP2 and XIAP was associated with poor prognosis.⁸³ The expression of IAP in cancer, their ability to inhibit apoptosis and mediate pro-survival signals through activation of NF-κB and MAPK signalling makes the IAP proteins attractive targets for anti-tumour therapy.⁷⁷

NF-κB signalling pathway

The nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) transcription factors are involved in regulating different cellular responses.

The mammalian family of NF-κB transcription factors is composed of five subunits NF-κB1 (p50/p105), NF-κB2 (p52/p100), RelA (p65), RelB and c- Rel that dimerize into hetero- and homo-dimers. The NF-κB transcription factors are held in the cytoplasm by the inhibitors of NF-κB (IκBs).⁸⁴ There are two known NF-κB activation pathways, the canonical (classical) and the non-canonical (alternative) pathway,^85,86 Figure 2. In the canonical pathway, activation of the inhibitor of IκB kinase (IKK) β phosphorylates the IκBs causing their degradation in the 26S proteasome⁸⁷ resulting in translocation of p50/p65 and c-Rel/p65 to the nucleus.⁸⁸ In the non-canonical pathway the NF-κB inducing kinase (NIK) activates IKKα which phosphorylates NF-κB2 (p100) resulting in a proteasomal removal of the inhibitory C-terminal IκB-δ domain, generating the p52 mature form and accumulation of p52/RelB in the nucleus.⁸⁹

(19)

NF-κB transcription factors are not only regulators of immune inflammatory response but also involved in the control of apoptosis and cell proliferation in many types of cancer including MM.⁹⁰ Both the canonical and the non-canonical NF-κB pathway have been associated with carcinogenesis.^21,91

In MM, the dysregulation of NF-κB signalling has mainly been related to the ligand-dependent interactions taking place in the bone marrow and seldom being associated to an underlying genetic mutation.²¹ However, several studies now show genetic abnormalities resulting in uncontrolled NF-κB activation/signalling.^21,91,92 In MM patients, approximately 20 % were found to have genetic lesions in the NF-κB pathway. Overexpression and/or gain of function mutations in the positive regulators of NF-κB signalling were found, e.g. NIK, NFκB1, NFκB2, LTBR, CD40 and TACI and inactivating mutations of the negative regulators e.g. TRAF3, TRAF2, cIAP1/cIAP2 and CYLD. Among these, the loss of function of the tumour suppressor TRAF3 is the most common gene deletion/mutation.²¹ TRAF3 together with TRAF2 and cIAP1/cIAP2 has repressive roles in the non- canonical NF-κB pathway by promoting the ubiquitination and degradation of NIK.⁹² However, a more complex picture exists regarding the function of cIAP1 and cIAP2. Although, cIAP1 and cIAP2 have been reported to have a repressive role in the non-canonical pathway they are reported to be positive regulators of TNFα mediated activation in the canonical pathway.⁹² Also by inducing expression of several genes, including cIAP1 and cIAP2, the NF- κB can suppress apoptosis.⁹³

(20)

TRAF2/3 cIAP1/2

Non-canonical NF-κB pathway Canonical NF-κB pathway

NIK

IKKα P IKKβ

P RIP

RelB P100

p50 p65

IκBα ^P

RelB Ub P100

UbUb Ub

P

p50 p65

IκBα _Ub^Ub^Ub Ub

Proteasome

Nucleus

TNFR family receptors

p65

p50 p52 RelB

Figure 2. A schematic view of the NF-κB pathway. In the canonical NF-κB pathway, p50 and p65 (the NF-κB dimers) are held in the cytoplasm because of the interaction with IκB. Upon ligand binding to the TNF receptor, cIAP1/2 activates RIP through ubiquitination resulting in activation of the IKK complex. The IKK complex phosphorylates the IκB which is degraded and p50/p65 can now translocate to the nucleus resulting in activation of targeted genes. In the non-canonical NF-κB pathway, in non-stimulated cells, cIAP1/2 together with TRAF proteins, ubiquitinate NIK leading to its degradation. NIK is required for the phosphorylation of the p100 which is then processed to p52 and can together with RelB translocate to the nucleus resulting in activation of targeted genes. Adapted from LaCasse et al 2008.⁹⁴

(21)

Apoptosis

A physiological process, that is evolutionary conserved, by which both damaged and excessive cells are removed is called apoptosis. Nuclear fragmentation, cell shrinking and membrane blebbing are typical features of apoptosis.⁹⁵ Important effector molecules in apoptosis are the proteolytic enzymes e.g the caspases.⁹⁶ Caspases can be initiated through activation of the death receptor (extrinsic) pathway or through the mitochondrial (intrinsic) pathway.⁹⁷ It is important that apoptotic pathways are strictly controlled to maintain homeostasis. Cancer cells are capable of avoiding apoptosis and thus anti-apoptotic mechanisms have been associated with drug-resistance in tumour cells.⁹⁷

Mitochondrial pathway

The mitochondrial pathway, Figure 3, is initiated from within the cell and is triggered by DNA damage, a defective cell cycle, hypoxia, loss of cell survival factors and severe cell stress. This pathway is primarily regulated by members of the Bcl-2 family. The Bcl-2 family consists of both pro- apoptotic and anti-apoptotic proteins and is divided into three different groups depending on presence of up to four Bcl-2 homology domains (BH1- 4). The anti-apoptotic Bcl-2 members Bcl-2, Bcl-XL and Mcl-1 contain BH- domains 1-4.^98,99 These proteins bind and sequester pro-apoptotic proteins and thereby prevent apoptosis. The pro-apoptotic members are divided into two groups, the first group is represented by members containing BH1-3 such as BAK and BAX and the other group is represented by members containing BH3 only. The BH3-only proteins include BAD, BID, BIK, puma, noxa and BIM that by protein-protein interactions with anti-apoptotic Bcl-2 proteins neutralize the pro-survival function of these anti-apoptotic proteins.⁹⁹ Bax and Bak undergoes conformational changes after activation that contributes to increased permeability of the mitochondrial membrane causing the release of cytochrome c (Cyt c) leading to activation of the caspase cascade. Cytosolic Cyt c forms a large multiprotein structure with the adaptor protease activating factor-1 (Apaf-1) and pro-caspase-9 called the apoptosome.¹⁰⁰ Pro-caspase-9 is cleaved and activated which in turn activates the downstream effector caspases-3, -6 and/or -7. SMAC/DIABLO and OMI/HtrA2 are molecules released from the mitochondria and promote caspase activation by antagonizing the members of the IAP family.^101,102

(22)

DR ligand DR

Death receptor (DR) pathway

FADD

Pro-caspase-8 DISC

Caspase-8

Mitochondrial pathway Apoptotic stimuli

Cyt c

SMAC/DIABLO Omi/HtrA2 IAP

APOPTOSIS Caspase-3

Caspase-9 Apoptosome

Apaf-1

Pro-caspase-9

Bid

tBid

Ψ

Bax Bak

Bcl-2 Bcl-xL

Figure 3. Apoptosis signalling pathways. In the DR pathway, activation upon ligand binding leads to assembly of the DISC complex resulting in caspase-8 activation which in turn activates caspase-3. Upon activation of the mitochondrial pathway the pro-apoptotic molecules promote permeabilisation of the mitochondrial membrane.

This results in Cyt c release and the formation of the apoptosome resulting in caspase-9 activation, which in turn activates caspase-3.

Death receptor mediated pathway

The death receptor (DR) mediated pathway, Figure 3, begins outside the cell and is activated by ligation of members belonging to the tumour necrosis factor (TNF) family such as Fas ligand, TNF-α and Apo2 ligand (also called TRAIL) to the death receptor e.g. Fas, TNFR1 and DR4/DR5 – members of the TNF receptor super family. Ligand binding induces recruitment of adaptor proteins Fas associated death domain (FADD) and/or TNFR1- associated death domain (TRADD) and the initiator caspase-8 as pro-caspase forming the death-inducing signalling complex (DISC).¹⁰³ Pro-caspase-8 is activated by proteolytic cleavage within the DISC which then activates effector caspases-3, -6 and/or -7. Caspase-8 and caspase-9 mediates major cellular death signalling and the lack of activation of either or both caspases may contribute to drug resistance. The combination of drugs that trigger both the mitochondrial and the death receptor mediated pathway or a downstream regulator of both these pathways may be a way to overcome drug resistance.

(23)

The role of epigenetics in MM

Epigenetics is the combined study of heritable changes in gene expression or cellular phenotype caused by mechanisms other than alterations of DNA nucleotide sequence.¹⁰⁴ The development of cancer has usually been considered to occur as a result of genetic abnormalities resulting in overexpression of oncogenes or loss of tumour suppressor genes. Studies on epigenetics have now shown that alterations in chromatin structure may also contribute to the oncogenic transformation.¹⁰⁵

It is now well-known that methylation of DNA and histone modifications can modify and control gene expression Figure 4. In MM, aberrant histone modifications, DNA methylation, and noncoding RNA expression are three common epigenetic mechanisms that cause abnormal signalling of important pathways.¹ DNA methylation is the most well-characterized epigenetic mechanism. In MM, promotor regions of some defined tumour suppressor genes have been reported to be hypermethylated although the mechanisms that causes this aberrant DNA methylation is not known.¹ An important role of the non-coding RNAs is also emerging, however in this thesis the epigenetic focus has been the histone modifications.

Histone modifications

Apart from gene silencing by DNA methylation, the epigenetic modification state of histones, a temporary silencing of genes during linage differentiation of pluripotent stem cells, is an important mechanism and is also implicated in oncogenesis.^106-109 Histones are proteins forming an octamer complex around which DNA is wrapped to form the nucleosome, and further organized in the higher order structure known as chromatin.¹ The interaction of histones with DNA and other DNA-binding proteins are affected by posttranslational changes of the histone tail, for example acetylation, methylation and phosphorylation and these histone modifications plays an essential role in gene regulation.¹¹⁰ Perturbations of the chromatin structure can result in incorrect gene expression and genomic instability that may lead to tumour growth. Proteins that control chromatin organization are therefore referred to as key players in cancer pathogenesis. The Polycomb group (PcG) proteins are such proteins and function as epigenetic regulators. They were first described in Drosophila where they function as key repressors of Hox genes

(24)

during embryonic development.¹¹¹ PcG are highly conserved from Drosophila to human and are now seen as a prominent player in carcinogenesis. The PcG proteins form multimeric complexes and are known to exert their functions through the formation of Polycomb Repressive Complexes (PRCs). The most well characterized PRCs are PRC1 and PRC2, responsible for the posttranslational modification of histones. PRC2 consists of the core proteins SUZ12, EED, RbAp46/48 and EZH1/2. EZH1/2 are the catalytic subunits and through their methyltransferase activity mediate trimethylation on histone H3 at lysine 27 (H3K27me3).¹¹² The PRC1 complex possesses histone ubiquitination E3 ligase activity and is responsible for ubiquitination of histone H2A at lysine 119 (H2AK119ub).

Both H3K27me3 and H2AK119ub are strongly associated with transcriptional repression.¹¹³ Overexpression of EZH2 has been linked to several types of cancer e.g. melanoma, prostate and breast cancer.^114,115 In MM, EZH2 overexpression has also been reported.¹¹⁶

Another member of the family of the PcG proteins that has been shown to be important in MM cells both in vitro as well as in vivo is the core component of the PRC1, Bmi-1.¹¹⁷ Moreover, in the human MM cell line RPMI 8226 it was shown that the pro-apoptotic gene Bim was negatively regulated by Bmi-1.¹¹⁷ In this paper, Bim was found to be up-regulated upon Bmi-1 knockdown.¹¹⁷ Furthermore, inactivating mutations of histone demethylase UTX have also been implicated in oncogenesis. UTX is a demethylase that removes the H3K27me3 mark and mutations affecting UTX have been found in a variety of tumours also including MM.¹¹⁸ Up- regulation of the histone-modifying enzyme, MMSET, was found in all MM patients with a t(4;14) translocation.¹¹⁹ MMSET catalyses the tri-methylation of H4K20¹²⁰ and di-methylation of H3K36 (H3K36me2). These modifications result in accumulation of H3K36me2 and a reduction of total H3K27me3 thus possibly leading to the transcriptional activation of oncogenes.^121,122

A whole genome and exome sequencing was performed by Chapman et al 2011 on 38 MM patients.¹² Both previously identified as well as novel genetic alterations and mutations were found. Interestingly, the transcription factor, HOXA9, was ubiquitously expressed and a significant enrichment of mutations in genes known to directly regulate HOXA9 e.g. UTX, WHSC1 (MMSET), WHSC1L1, MLL, MLL2 and MLL3 was found. Thus highlighting the cooperation between genetic and epigenetic mechanisms that underlie cancer etiology.¹²

(25)

Pol II

H3K4me3

H3K27me3

H3K36me3 Histone acetyltransferase Histone deacetylase

Histone acetylation Unmethylated cytosine Methylated cytosine Active gene

Silencing through DNA methylation

Silencing through PRC2 and H3k27me3

Figure 4. Schematic model of epigenetic modifications. Transcriptionally active genes are characterized by H3k4me3, H3K79me3 and acetylation in the promotor region and H4K20me1 and H3K36me3 in the gene body. At the same time in the promotor region, 5'cytosine-phosphodiester bond-guanine-3´ (CpG) islands are unmethylated while there is DNA methylation in the gene body. Silencing of gene expression could occur via DNA methylation or via the PRC2 complex that causes H3K27me3 gene silencing. Figure is adapted from Dimopoulos et al 2014.¹

(26)

Drug resistance

One of the most challenging areas for current cancer research and clinical treatment is the development of resistance to anticancer drugs. As of today, one of the primary approaches to treat cancer is chemotherapy.¹²³ The resistance that develops to chemotherapeutics could either be that resistance- mediating features already exist in the tumour cells and thereby the therapy will be unsuccessful (so called intrinsic resistance). Resistance can also develop during treatment and can be initiated by mutations/epimutations that arise during chemotherapy or activation of different compensatory pathways of survival (so called acquired resistance).^123,124 Some of the molecular mechanisms that have been associated with drug resistance are the alterations in drug metabolism, increased efflux and mutation of drug targets.^125-128 Other mechanisms that could lead to drug resistance are inactivation of death signalling pathways, activation of survival signalling pathway^129,130 influence of the tumour microenvironment and epigenetic changes.^131,132 One way to overcome drug resistance might be the use of combinatorial treatments. The drug combinations that are used are based on agents that have shown a synergistic effect in in vitro and in vivo studies.

A successful drug combination in MM is the use of IMiDs e.g.

thalidomide and lenalidomide together with dexamethasone which yield responses in up to 80% of MM patients.¹⁰ The anti-tumour mechanisms underlying the effect of the IMiDs were until recently largely unknown.

Thalidomide was found to bind to cereblon and inhibited autoubiquitylation of cereblon,¹³³ this was also found in MM.¹³⁴ Recently it was also shown that the IMiDs induce proteasomal degradation of both Ikaros and Aiolos, in a cereblon-dependent manner. Both Ikaros and Aiolos expression play essential roles in cell viability in many MM cell lines.^135-137

Usage of drugs that target the same survival pathway, may lead to the emergence of compensatory pathways of survival.¹²⁴ Therefore, by hitting the tumour cell with drugs that target completely different and independent pathways might be a way to limit the development of drug resistance.¹²³

In MM the clonal heterogeneity contributes to resistance to the current treatments. MM disease is now thought to be composed of clonally different subgroups of MM tumour cells harbouring a vast genetic diversity.^138,139 Following one t(4;14) MM patient made it possible to understand the relative frequency of the different MM subclones and how they change over time and with treatments.¹³⁹ The results from the study by Keats et al 2012¹³⁹

(27)

indicated several important clinical implications. First, partial response may not reflect a partial suppression of the whole tumour but rather be a result of suppression of sensitive (dominant) subclone while other subclones remain stable and could eventually grow to a dominant mass. Second, especially for MM patients with high-risk tumours that are suggested to be less stable and more likely to change over time, it is important to use therapies that target all of the existing subclones. Third, re-treatment might be effective since a sensitive subclone could have re-emerged. Fourth and finally, a more indolent subclone can be eliminated with early suboptimal treatment creating room for a more aggressive subclone.

The heterogeneity of the MM disease, how the different MM subclones progress during the disease and the changes of dominant subclones during treatment and relapse is a challenging research area.¹³⁸ Although we have a better understanding of MM biology, resistance to the current treatments will eventually occur in MM patients.¹⁴⁰ Development of new treatment approaches that either are active against all coexisting MM clones or agents that enhances the efficacy of the already existing treatments is necessary to overcome MM disease.

(28)

Present investigation

Polycomb target genes are silenced in multiple myeloma (Paper I)

Aims:

i. Identification of under-expressed genes in MM and determining if they have a common denominator.

ii. Investigate the presence of tri-methylation at H3 lysine 27 on several genes from the identified gene list.

iii. Investigate if H3K27me3 marks at these set of genes could be reverted by the use of chemical inhibitors

Results:

MM remains an incurable disease, and the mechanisms underlying its pathogenesis is only fragmentarily understood. To dissect the nature of the tumour initiating capacity and if a common profile for MM development and tumour maintenance exists we first analysed gene expression using data that are publicly available,^61,141,142 by integrative genomic analysis. We found a strong correlation between the under-expressed genes in MM patients and genes that were previously described as silenced by/associated with H3K27me3 in human embryonic fibroblast.¹⁴³ More importantly, in the advanced stages of MM progression these genes were more suppressed.

We then analysed the H3K27me3, mediated by the catalytic subunit EZH2 in the PRC2 complex, in four MM patients and found enrichment for H3K27me3. In two MM cell lines, RPMI 8226 and U-266-1984, we determine the chromatin modifications, i.e. H3K27me3 and H3K9ac at five PRC2 targets commonly under-expressed in MM (CIITA, CXCL12, GATA2, CDH6 and ICSBP/IRF8). In the patient samples and both cell lines an enrichment of H3K27me3 was found in the analysed genes. In the cell lines these genes also lacked the positive mark H3K9ac.

Next we analysed if the genes enriched for H3K27me3 could be reactivated using two chemical inhibitors, the HDAC inhibitor LBH589 and the S-adenosylhomocystein hydrolase inhibitor 3-Deazaneplanocin (DZNep): a histone deacetylase inhibitor and a global histone methylation inhibitor respectively. In RPMI 8226, both LBH589 and DZNep increased

(29)

mRNA levels of CIITA, GATA2 and CDH6. Up-regulation of CXCL12 could only been seen with DZNep treatment. In U-266-1984 an increased mRNA level of CIITA with DZNep treatment and ICSBP/IRF8 with LBH589 treatment was seen. Treatment with LBH589 and DZNep has been reported to deplete EZH2 protein from both breast cancer and AML.^144,145 Likewise, treatment with LBH589 or DZNep also depleted EZH2 protein, decreased the relative number of viable MM cells and increased the fraction of apoptotic cells. Furthermore, treatment with LBH589 reactivated CIITA and CXCL12 in vivo using 5T33MM murine model. It also resulted in a tumour load decrease and increased survival of the 5T33MM mice.

Discussion and conclusions:

Silencing mediated by the Polycomb complex is now a known factor in several types of cancer. Here we for the first time found a set of under- expressed genes in MM that correlated, significantly, to already defined Polycomb target genes.¹⁴³ The silencing of the genes found in this signature was more pronounced in the late stages of MM progression. Importantly, our data suggest that this profile of under-expressed genes is a common feature among the MM cells rather than representing a small specific subclone or subpopulation. If targeting the common denominator for the silenced genes could emerge as an attractive therapeutic strategy still remains to be answered. However, gene silencing via epigenetic mechanisms can be reverted and therefore constitutes an attractive drug target. Our data strongly supports that Polycomb mediated gene silencing plays an important role in MM and supports exploring more specific EZH2 inhibitors as new therapeutic options in MM.

(30)

The HDAC inhibitor LBH589 enhances the anti- myeloma effect of the IGF-1RTK inhibitor picropodophyllin (Paper II)

Aims:

i. Overcoming drug resistance by combining the IGF-1R inhibitor PPP and the HDAC inhibitor LBH589.

ii. Reveal the molecular mechanism underlying this combinatorial effect.

Results:

In paper II, by combining the IGF-1RTK inhibitor, PPP, and the HDACi, LBH589, we were able to enhance the anti-MM effect compared to single drug treatment. First we studied the cell viability after treatment with PPP and/or LBH589 using four human MM cell lines. All cell lines responded dose-dependently to LBH589 and PPP as single drugs. Simultaneously treatment with PPP and LBH589 resulted in a synergistic effect with a combination index <0.90 and a significant decrease in cell survival compared to single drug treatment. This was also true for the 5T33MM cells.

In the human MM cell line, RPMI 8226, the combinatorial treatment resulted in a significant increase of apoptotic cells, an additive effect on cleavage of caspase-8 and a more pronounced down-regulation of the anti-apoptotic proteins Mcl-1 and Bcl-2 compared to single drug treatment. In the 5T33MM cells a down-regulation of Bcl-2 and BCL-xL was observed in the combinatorial treatment.

We also analysed the effect of the drugs on cell cycle proteins and could identify a more prominent down-regulation in the combinatorial design of cell cycle regulating proteins such as cyclin E, cyclin B1 and cyclin D2 in RPMI 8226.

Moreover, we studied the anti-tumour effects of the drug combination in vivo using the 5T33MM murine model. A significant increase in overall survival as compared to the single-drug treatment was shown.

(31)

The expression of IGF-1R has been associated with progressive disease and drug resistance.¹⁴⁶ Several factors support the notion that IGF-1R is a candidate for novel MM therapy. Unfortunately compensatory survival pathways will emerge as a result of drug exposure¹²⁴ and targeting the IGF- 1R are therefore unlikely to be curative as single therapy in MM.

HDAC inhibitors (HDACi) are agents that are in clinical trials for treatment of MM patients¹⁴⁷ and the HDACi LBH589 has been shown to have synergistical anti-myeloma activity in combination with bortezomib, dexamethasone or melphalan.^148,149

We found that by combining PPP with LBH589, the anti-MM activity was enhanced. A factor that plays an important role in the expansion and survival of MM cells is the BM microenvironment.¹⁵⁰ By using the 5T33MM in vivo model we demonstrated that the combined treatment resulted in a significantly prolonged survival of the 5T33MM inoculated mice as compared to single drug treatment. This suggests that this therapy may indeed evade the favourable impact of the tumour microenvironment on tumour progression and expansion. Our data suggest that the combinatorial treatment of PPP and LBH589 is a candidate for new treatment strategies of MM patients.

(32)

Increased resistance to proteasome inhibitors in multiple myeloma mediated by cIAP2 expression – implication for a combinatorial treatment (Paper III)

Aims:

i. Investigate the role of cIAP2 in MM cell lines harbouring a TRAF3 deletion/mutation.

ii. cIAP2 role in resistance to proteasome inhibition.

Results:

In this study, we focused on exploring the role of cIAP2 in emerging drug resistance in MM cells. By lentiviral transduction of the human MM cell lines LP-1 and ANBL-6, we generated cIAP2 overexpressing cells, LP-1- cIAP2-eGFP and ANBL-6-cIAP2-eGFP, and control cells, LP-1-eGFP and ANBL-6-eGFP. Both MM cell lines used in this study harbour a TRAF3 deletion/mutation. A functional TRAF3 is needed for the cIAPs to function as a negative regulator in the non-canonical NF-κB pathway. The cIAP2 overexpressing cells were then exposed to a panel of different drugs. In both cell lines an increased tolerance to proteasome inhibitors was found as compared to each control. cIAP2 was also found to have an effect on apoptosis. In the LP-1-cIAP2-eGFP cells upon bortezomib treatment, a significant lower amount of apoptotic cells and also a decrease of caspase activation were found compared to the control.

We could also show that cIAP2 binds to caspase-3, caspase-8 and caspase-9 and that the interaction with caspase-3 and caspase-8 was significantly increased in the LP-1-cIAP2-eGFP cells upon bortezomib treatment. Furthermore, the NF-κB transcription factors, p105/p50 (canonical pathway) and p100/p52 (non-canonical pathway) were analysed when treated with bortezomib. Here we found that cIAP2 had an effect on the canonical pathway proteins and increased levels of p105 and p50 were found in the LP-1-cIAP2-eGFP cells compared to the control. No alteration of the non-canonical proteins p100 or p52 was seen in the LP-1-cIAP2-eGFP cells or in the control upon bortezomib treatment indicating that cIAP2 expression had no effect on the non-canonical pathway.

A gene expression microarray was performed in an effort to identify the molecular mechanisms underlying the effect of cIAP2. Increased expressions of genes which may function as nodes for a large number of direct interacting partners were found in the LP-1-cIAP2-eGFP cells. Gene

(33)

expression array also revealed 12 target genes of the NF-κB pathway to be down-regulated. Several of these 12 genes have been reported as tumour suppressors, ASS1 ¹⁵¹ and IRF1 ¹⁵² or to contribute to apoptosis, IL1RN, FTH1 and FAS.

In the LP-1-cIAP2-eGFP cells an up-regulation of IGF-1, an important growth factor in MM, was found both in the untreated cells and with bortezomib treatment whereas IGF-1 expression decreased in the control cells upon bortezomib treatment.

By combining bortezomib with the IAP inhibitor, AT-406, a more pronounced decrease in cell survival was found compared to single drug treatment in the MM cells.

Although current therapies prolong the survival for MM patients, drug resistant clones will eventually emerge. The mechanisms underlying the drug resistance are not fully understood but one mechanism underlying the insensitivity to therapy is the emergence of genetic alterations in survival signalling pathways. The members of the inhibitor of apoptosis (IAP) family are often found to be altered in human malignancies. Alterations of the IAP family are associated with chemo-resistance, disease progression and poor patient prognosis.^92,94,153 IAP proteins are also capable of binding to caspases and thereby protecting the cell from apoptosis. In addition they also have an important function as regulators of the NF-κB pathway. The cellular inhibitor of apoptosis 2 (cIAP2) together with cIAP1, has an important role as a positive regulator in the canonical, while having a negative role in the non-canonical NF-κB pathway.

We suggest based on our findings, that in TRAF3 deleted/mutated MM cell lines cIAP2 expression contributes to bortezomib resistance. One way to overcome this resistance might be by combining bortezomib treatment with an IAP antagonist. We found that the IAP-antagonist AT-406 could down- regulate both cIAPs and decrease the cell survival when used in a combinatorial design together with bortezomib. These results suggest that this combination might be favourable for MM patients harbouring TRAF3 deletions/mutations and thus being more dependent on the canonical NF-κB pathway.

(34)

Acknowledgements

First I would like to thank all the colleagues, co-authors and everyone that has supported, helped or contributed to the work presented in this thesis that was carried out at the Department of Immunology, Genetics and Pathology (IGP), Rudbeck laboratory, Uppsala University. The work in this thesis was financially supported by the Swedish Research Council, the Multiple Myeloma Research Foundation (MMRF), the Swedish Cancer Society and von Kantzows Stiftelse.

My supervisor Helena Jernberg Wiklund, thank you for giving me this huge opportunity to grow as a researcher and as a person. For always, maybe not physically, been there when I was lost or just needed your help and guidance.

My first co-supervisor Mårten Fryknäs, for being there and guided me through the bioinformatic mesh. For helping me correcting this thesis and for your helping and supporting words when I needed it the most, that help me a lot!

My second co-supervisor Fredrik Öberg, for your knowledge and valuable comments. For the well-cooked spring lunches, don’t forget that we are considering it as a tradition.

To all the present and former member of the Sten Sture lab, (someone has to explain where the name comes from). Kenneth, for sharing your scientific knowledge and for being a source of positive energy. Inger, the nice mushroom picking expeditions and always sharing your home-grown vegetables. Antonia, although we don’t work after the same watch, you were always there to help. Thanks for including me in your life, both in the work related as well in the non-work related life. Prasoon, for keeping me company in the office during all my years as a PhD-student. For the nice time we spent in Paris and for introducing us to the India kitchen.

Mohammad, for all the nice talks and scientific discussions. For making my pile of article to read never decrease. Alba, a friendly soul that always helps when its needed. Lotta, for welcoming me at Rudbeck and for all the help and nice talks. Pernilla, where to start… I would never have made it without you! Thanks for all your help with everything from growing cells to western

(35)

blot and cell sorting. Also for knowing that I always can count on you and talk to you about everything.

Umash, for the nice discussion breaks in the office and for helping me correcting this thesis.

To all my collaborators and co-authors, especially to Karin Vanderkerken and her research group (Susanne, Els and Miguel thanks for the nice papers we produced). Theoharis Panaretakis (Aris) group for welcoming me to KI and helped me with my project. Linda Arngården, for teaching me how PLA works.

I would like to thank all people at Rudbeck for making this a wonderful place to come to every morning, no one mention no one forgotten. To all the people at the administration, especially Christina Magnusson, you are the right person at the right place! It always feels so much better when I leave your office.

My most amazing family, my mother, Marie, and my father, Stig, you are the best parents anyone could wish for! Always supporting, encouraging me in whatever I decided to do. For all the help with Valter and Molly when I had to stay late in Uppsala or work at home. Even though you really don’t understand all I’m doing I know you are very proud of me. My sister and brother, Caroline and Fredrik, thanks for being the best siblings in the world! You and your families are the greatest!

My parents in laws, Lis and Bosse for your interest in my work and keeping my sugar levels constant high. I might need a detox after this 

And last I would like to thank my own little family that is increasing. Niklas, what would I have done without you?! You have pushed me when I need a push, supported me through years and years of commuting and been there during all the good and bad times. Valter, my little sun shine! Always a happy face when entering the door, you make me forget even the toughest day. Our coming baby that kept me company during the stressful time of writing this thesis.

(36)

References

1 Dimopoulos, K., Gimsing, P. & Gronbaek, K. The role of epigenetics in the biology of multiple myeloma. Blood cancer journal 4, e207, doi:10.1038/bcj.2014.29 (2014).

2 Morgan, G. J., Walker, B. A. & Davies, F. E. The genetic architecture of multiple myeloma. Nature reviews. Cancer 12, 335-348, doi:10.1038/

nrc3257 (2012).

3 Gahrton, G., Durie, B. G. M. & Samson, D. M. Multiple myeloma and related disorders. (Arnold, 2004).

4 Lopez-Corral, L. et al. The progression from MGUS to smoldering myeloma and eventually to multiple myeloma involves a clonal expansion of genetically abnormal plasma cells. Clinical cancer research : an official journal of the American Association for Cancer Research 17, 1692-1700, doi:10.1158/1078-0432.CCR-10-1066 (2011).

5 Dalton, W. S., Bergsagel, P. L., Kuehl, W. M., Anderson, K. C. &

Harousseau, J. L. Multiple myeloma. Hematology / the Education Program of the American Society of Hematology. American Society of Hematology.

Education Program, 157-177 (2001).

6 Palumbo, A. & Anderson, K. Multiple myeloma. The New England journal of medicine 364, 1046-1060, doi:10.1056/NEJMra1011442 (2011).

7 Raab, M. S., Podar, K., Breitkreutz, I., Richardson, P. G. & Anderson, K. C.

Multiple myeloma. Lancet 374, 324-339, doi:10.1016/S0140-6736(09) 60221-X (2009).

8 Staudt, L. M. Oncogenic activation of NF-kappaB. Cold Spring Harbor perspectives in biology 2, a000109, doi:10.1101/cshperspect.a000109 (2010).

9 Gozzetti, A., Candi, V., Papini, G. & Bocchia, M. Therapeutic advancements in multiple myeloma. Frontiers in oncology 4, 241, doi:10.3389/

fonc.2014.00241 (2014).

10 Licht, J. D., Shortt, J. & Johnstone, R. From anecdote to targeted therapy: the curious case of thalidomide in multiple myeloma. Cancer cell 25, 9-11, doi:10.1016/j.ccr.2013.12.019 (2014).

11 Sharma, A. et al. DNA methylation alterations in multiple myeloma as a model for epigenetic changes in cancer. Wiley interdisciplinary reviews.

Systems biology and medicine 2, 654-669, doi:10.1002/wsbm.89 (2010).

12 Chapman, M. A. et al. Initial genome sequencing and analysis of multiple myeloma. Nature 471, 467-472, doi:10.1038/nature09837 (2011).

13 Dimopoulos, K., Gimsing, P. & Gronbaek, K. Aberrant microRNA expression in multiple myeloma. European journal of haematology 91, 95- 105, doi:10.1111/ejh.12124 (2013).

14 Fonseca, R. et al. International Myeloma Working Group molecular classification of multiple myeloma: spotlight review. Leukemia 23, 2210- 2221, doi:10.1038/leu.2009.174 (2009).

(37)

15 Fonseca, R. et al. Genetics and cytogenetics of multiple myeloma: a workshop report. Cancer Res 64, 1546-1558 (2004).

16 Bergsagel, P. L. & Kuehl, W. M. Chromosome translocations in multiple myeloma. Oncogene 20, 5611-5622, doi:10.1038/sj.onc.1204641 (2001).

17 Hideshima, T., Mitsiades, C., Tonon, G., Richardson, P. G. & Anderson, K.

C. Understanding multiple myeloma pathogenesis in the bone marrow to identify new therapeutic targets. Nature reviews. Cancer 7, 585-598, doi:10.1038/nrc2189 (2007).

18 Bergsagel, P. L. & Kuehl, W. M. Molecular pathogenesis and a consequent classification of multiple myeloma. J Clin Oncol 23, 6333-6338, doi:23/26/6333 [pii]10.1200/JCO.2005.05.021 (2005).

19 Shou, Y. et al. Diverse karyotypic abnormalities of the c-myc locus associated with c-myc dysregulation and tumor progression in multiple myeloma. Proc Natl Acad Sci U S A 97, 228-233 (2000).

20 Chng, W. J. & Fonseca, R. Genomics in multiple myeloma: biology and clinical implications. Pharmacogenomics 6, 563-573, doi:10.2217/

14622416.6.6.563 (2005).

21 Keats, J. J. et al. Promiscuous mutations activate the noncanonical NF- kappaB pathway in multiple myeloma. Cancer cell 12, 131-144, doi:10.1016/j.ccr.2007.07.003 (2007).

22 Kuehl, W. M. & Bergsagel, P. L. Multiple myeloma: evolving genetic events and host interactions. Nat Rev Cancer 2, 175-187, doi:10.1038/nrc746 (2002).

23 Caers, J., Asosingh, K., Van Riet, I., Van Camp, B. & Vanderkerken, K., . Of mice and men: disese models of multiple myeloma. Drug Discovery Today:

Disease Models 1, 373-380, doi:10.1016/j.ddmod.2004.11.010 (2004).

24 Kaufman, J. & Lonial, S. Multiple myeloma: the role of transplant and novel treatment strategies. Semin Oncol 31, 99-105, doi:S0093775404000740 [pii]

(2004).

25 Kawano, M. et al. Autocrine generation and requirement of BSF-2/IL-6 for human multiple myelomas. Nature 332, 83-85, doi:10.1038/332083a0 (1988).

26 Klein, B. et al. Paracrine rather than autocrine regulation of myeloma-cell growth and differentiation by interleukin-6. Blood 73, 517-526 (1989).

27 Jernberg, H., Pettersson, M., Kishimoto, T. & Nilsson, K. Heterogeneity in response to interleukin 6 (IL-6), expression of IL-6 and IL-6 receptor mRNA in a panel of established human multiple myeloma cell lines. Leukemia 5, 255-265 (1991).

28 Barut, B. A. et al. Role of interleukin 6 in the growth of myeloma-derived cell lines. Leukemia research 16, 951-959 (1992).

29 Nilsson, K. Established cell lines as tools in the study of human lymphoma and myeloma cell characteristics. Haematol Blood Transfus 20, 253-264 (1977).

30 Nilsson, K. & Ponten, J. Classification and biological nature of established human hematopoietic cell lines. Int J Cancer 15, 321-341 (1975).

31 Ralph, P. The human B-cell lineage cell line ARH-77. Cancer 56, 2544-2545 (1985).

32 Pellat-Deceunynck, C. et al. Adhesion molecules on human myeloma cells:

significant changes in expression related to malignancy, tumor spreading, and immortalization. Cancer Res 55, 3647-3653 (1995).

Tumour Survival Signals and Epigenetic Gene Silencing in Multiple Myeloma: Implications for Biology and Therapy