Potential Tumor-Promoting Effects of Ectopic Cd137 Expression on Hodgkin Lymphoma

POTENTIAL TUMOR-PROMOTING EFFECTS OF ECTOPIC CD137 EXPRESSION ON HODGKIN LYMPHOMA

HO WENG TONG (B. Sci (Biomedical Sci.), UPM, Malaysia)

A THESIS SUBMITTED FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

DEPARTMENT OF PHYSIOLOGY

NATIONAL UNIVERSITY OF SINGAPORE

2013

Acknowledgement

I would like to take this opportunity to address my appreciation to my thesis supervisor, Associate Professor Herbert Schwarz, who had provided me with exceptional guidance and advice throughout my candidature. Besides that, he also gave me a lot of supports and this study would not be possible without his truly contributions.

I would also like to thank Dr. Shao Zhe, Dr Jiang Dongsheng and Dr

Shaqireen for guiding me through basic laboratory technique when I first joined the group, Dr. Gan Shu Uin, Dr. Angela Moh and Ms Tan Teng Ee for assisting me in the generation of transfected and knock down cell lines, and

Ms Lee Shu Ying from the confocal unit, NUHS, for providing the necessary facility for my confocal imaging. In addition, I would also like to show my appreciation to Ms Pang Wan Lu who worked closely with me in this Hodgkin lymphoma project.

Last but not least, I would also like to express my pleasure to work with all the members from Herbert Schwarz' laboratory, especially Zulkarnain, Qianqiao,

Liang Kai and Andy who gave me a lot of support both technically and morally.

ii TABLE OF CONTENTS

DECLARATION i

ACKNOWLEDGEMENT ii

TABLE OF CONTENTS iii

ABSTRACT vii

LIST OF TABLES ix

LIST OF FIGURES x

LIST OF ABBREVIATION xii

1. INTRODUCTION 1

1.1 Hodgkin lymphoma 2

1.1.1 Etiology and pathophysiology 3

1.1.2 Immunosuppressive microenvironment 6

1.1.3 Association of HL with members of tumor necrosis 8 factor receptor family

1.2 CD137 and CD137L

1.2.1 Expression and characteristic of CD137 and 12 CD137L

1.2.2 Targeting CD137 for immunotherapy 16

1.2.3 CD137 and CD137L in malignant diseases 18

1.3 Trogocytosis 19

1.4 Research objectives 22

iii 2. MATERIALS AND METHODS 24

2.1 Cells and cell culture 24

2.1.1 Cell lines 24

2.1.2 CD137 overexpressing cell lines 24

2.1.3 PBMC isolation 25

2.1.4 T cells, B cells and Monocytes purification 26

2.1.5 Storage of cell lines and primary cells 27

2.2 Antibodies and reagents 27

2.3 Flow cytometry 28

2.4 Enzyme linked immunosorbent assay (ELISA) 29

2.5 Western Blot 29

2.5.1 Protein purification 29

2.5.2 Electrophoresis and electroblotting 30

2.5.3 Antibody probing and visualization 30

2.6 Co-immunoprecipitation 31

2.7 Confocal microscopy 31

2.8 Endocytosis inhibition assay 32

2.9 Cell viability assay 32

2.10 Trogocytosis assay 33

2.11 Reserve-transcriptase polymerase chain reaction 33

2.11.1 RNA extraction 33

2.11.2 Reverse transcription 34

2.11.3 Polymerase chain reaction (PCR) 35

2.12 Light microscopic examination 36

2.13 Antibody treatment 36

iv 2.13.1 CD137 and CD137L neutralization 36

2.13.2 Agonistic CD137 stimulation 36

2.14 Statistics 37

3. RESULT 38

3.1 Screening and generation of RS cell lines 38

3.2 Ectopically expressed CD137 reduces the T cell 42 stimulatory capacity of HRS cells

3.2.1 Co-culturing HRS cells with PBMC and T cells 42

3.2.2 Ectopically expressed CD137 down-regulates 46 CD137L expression on RS cells

3.2.3 Induction of IFN-γ release is due to CD137L 48 upregulation after CD137 silencing

3.2.4 Summary 51

3.3 Mechanism of CD137L disappearance 52

3.3.1 CD137 neutralization does not affect CD137L 53 mRNA expression

3.3.2 CD137L protein expression increases after CD137 54 neutralization

3.3.3 CD137-CD137L co-localization in RS cells 56

3.3.4 CD137 and CD137L are co-internalized via 61

endocytosis

3.3.5 Trogocytosis mediates CD137 transfer which 65 causes CD137L downregulation

3.3.6 Summary 74

3.4 Transfer of CD137 to surrounding monocytes and B cells 75

3.4.1 Ectopically expressed CD137 induces CD137L 76 downregulation in monocytes and B cells

v 3.4.2 CD137 overexpressing cell lines abrogate IFN-γ 83 release from PBMC

3.4.3 Summary 84

3.5 Potential of CD137 and CD137L signaling on HRS cells 85

3.5.1 Agonistic anti-CD137 antibody 85

3.5.2 CD137 stimulation causes morphological changes in 87 HRS cells

3.5.3 Involvement of the cytoplasmic domain of CD137L in 89 HRS cell signaling

3.5.4 Summary 91

4. DISCUSSION 92

4.1 Ectopically expressed CD137 abrogates T cell activity 92

4.2 CD137 trogocytosis 99

4.2.1 CD137 and CD137L transfer 101

4.2.2 Aggregation of CD137 and CD137L in the cytoplasm 103

4.3 CD137 and CD137L induce signaling into HRS cells 104

4.4 Limitations 106

4.5 Future works 107

4.6 Conclusion 110

REFERENCES 111

APPENDIX I ACRYLAMIDE GEL CASTING 125

APPENDIX II MEDIA AND BUFFERS 126

APPENDIX III ANTIBODIES LIST 134

APPENDIX IV CD137L DOWNREGULATION ON 136 PRIMARY MONOCYTES AND B CELLS

vi ABSTRACT

Hodgkin lymphoma (HL) is a hematological malignancy. The malignant cells in HL, the Hodgkin and Reed Sternberg (HRS) cells, comprise only a minority of the entire tumor mass while infiltrating inflammatory cells constitute the vast majority of the tumor mass. HRS cells were reported to express CD137 ectopically but the function(s) of CD137 in the pathogenesis of HL remained unidentified. CD137 is a potent co-stimulatory molecule expressed by activated T cells. Upon ligation by CD137 ligand (CD137L) which is mainly expressed by antigen presenting cells (APC), CD137 signalling enhances T cell activity. This study shows that ectopic CD137 expression on HRS cells reduces IFN-γ release from T cells by downregulating CD137L expression on

HRS cells. The IFN-γ suppression due to ectopic CD137 expression can be reversed by neutralizing CD137 with antibodies or by knocking down CD137 with siRNA. The downregulation of CD137L is not due to a reduction of de

novo synthesis of CD137L but due to an increased CD137L turnover. When

CD137 binds to CD137L, the CD137-CD137L complex will be internalized, and hence the surface CD137L levels available for T cell co-stimulation are reduced. Ectopic CD137 also gets transferred from HRS cells to surrounding cells including B cells and monocytes and leads to the downregulation of

CD137L on monocytes. The CD137L downregulation on monocytes results in a lower T cell co-stimulation and a reduction of IFN-γ release from T cells. In

addition, this study finds that ectopic CD137 on HRS cells might induce

signalling into HRS cells but the benefits that HRS cells may gain from this

signalling are yet to be identified. In conclusion, this study provides new

vii insights into the immune escape mechanism of HL, and opens new research areas for the development of novel therapeutic approaches.

viii LIST OF TABLES

Table 1. Characteristic of Hodgkin lymphoma subtype 4

Table 2. RT-PCR reaction mix 34

Table 3. Standard PCR reaction mix 35

Table 4. Primers set for CD137 and cyclophilin PCR 35

Table 5. PCR thermal cycling for CD137L amplification 36

Table 6. Changes in CD137L expression on monocytes and B 82 cells after co-culturing with HRS cell lines

LIST OF FIGURES

Figure 1. CD137 expression on Hodgkin Lymphoma 11

Figure 2. CD137 and CD137L bidirectional signaling 15

Figure 3. PBMC isolation with density gradient centrifugation. 26

Figure 4. Expression of CD137 on Reed Sternberg cell lines 40

Figure 5. CD137 silencing of KM-H2 cells 41

Figure 6. CD137 expression on L428, L540 and L1236 cells after 41 CD137 transfection

Figure 7. CD137 neutralization on KM-H2 cells induces higher 44 IFN-γ release from PBMC

Figure 8. Impaired CD137 expression on KM-H2 cells leads to 45 higher IFN-g realease from PBMC and T cells

Figure 9. CD137 and CD137L expression in KM-H2 cells after 47 CD137 neutralization

Figure 10. CD137L is upregulated in KM-H2 cells after CD137 47 silencing

Figure 11. CD137L neutralizing antibodies abrogate the 49 induction of IFN-γ release from PBMC during co- culture with KM-H2-CD137 - cells.

Figure 12. CD137L neutralizing antibodies abrogate the 50 induction of IFN-γ release from T cells during co- culture with KM-H2-CD137 - cells.

Figure 13. CD137 neutralization does not affect mRNA expression 53 level of CD137L in KM-H2 cells

Figure 14. Total CD137L protein in KM-H2 cells was upregulated 55 after CD137 neutralization

Figure 15. Upregulation of CD137L protein in KM-H2 cells after 55 CD137 neutralization

Figure 16. CD137L co-immunoprecipitates with CD137 in KM-H2 58 cells.

x Figure 17. CD137 neutralization reduces CD137 and CD137L 58 interaction in KM-H2 cells. Figure 18. Colocalization of CD137 and CD137L in KMH2 cells 60

Figure 19. Toxicity titration of MDC on KM-H2 cells 63

Figure 20. Inhibition of endocytosis increases CD137 and CD137L 64 expression on KM-H2 cells

Figure 21. CD137 and CD137L induce trogocytosis between donor 67 and recipient cells

Figure 22. The presence of CD137 and CD137L induce L-428 and L- 68 1236 cell aggregation with KMS-11 cells 70 Figure 23. Downregulation of CD137L on KMS-11 cells following CD137L dependent transfer of CD137 from L-428 cells to KMS-11 cells

Figure 24. Downregulation of CD137L on KMS-11 cells following 71 CD137L dependent transfer of CD137 from L-1236 cells to KMS-11 cells

Figure 25. CD137L transfer from KMS-11-CD137L cells to L-428- 72 CD137 cells

Figure 26. CD137L transfer from KMS-11-CD137L cells to L-1236- 73 CD137 cells

Figure 27. Transfer of CD137 from HRS cell lines to monocytes 78

Figure 28. Transfer of CD137 from HRS cell lines to B cells 79

Figure 29. HRS cells induce CD137L downregulation on monocytes 82

Figure 30. CD137 expressing L-428 and L-1236 cells inhibit IFN-γ 84 releases from PBMC

Figure 31. Anti-CD137 antibody clone JG1.6a induces more IFN- γ 86 release from PBMC than the anti-CD137 antibody clone BBK-2

Figure 32. The CD137 agonistic antibody causes morphological 88 changes on CD137 expressing L-428 cells

Figure 33. The CD137L cytoplasmic domain is present in the 90 nucleus of non-proliferating KM-H2 cells

Figure 34. Ectopic CD137 promotes HRS cells survival via 98 downregulation of CD137L expression

xi LIST OF ABBREVIATIONS

4-1BBL 4-1BB ligand

7AAD 7-Aminoactinomycin

AICD Activation induced cell death

AID Activation-induced deaminase

APC Antigen presenting cells

BCMA B cell Maturation antigen

BCR B cell receptor cCD137L cytoplasmic domain of CD137L

CD Cluster of differentiation

CD137L CD137 ligand

CTL Cytotoxic T lymphocytes

CTLA-4 Cytotoxic T-lymphocyte antigen 4

D Dalton

DC Dendritic cells

DR Death Receptor

EBV Epstein–Barr virus

ELISA Enzyme linked immunosorbent assay

FACS Fluorescence-activated cell sorting

FBS Fetal bovine serum g gram

GC Germinal center

GM-CSF Granulocyte macrophage colony-stimulating factor h Hour

xii HL Hodgkin lymphoma

HRS Hodgkin and Reed-Sternberg

IFN-γ Interferon-gamma

IL Interleukin

ILA Inducible by lymphocyte activation

IP Immunoprecipitation

IRF4 Interferon regulating factor 4 k Kilo (1 x 10 3)

KIR Killer-cell Immunoglobulin-like Receptor l liter

L&H Lymphocytic and histiocytic

LMP Latent membrane protein m Milli (1 x 10 -3)

M Molar (mmol/L)

MACS Magnetic activated cell sorting

MDC Monodansylcadaverine

MFI Mean fluorescent intensity

MHC Major histocompatibility complex min Minute

MM Multiple myeloma mRNA Messenger RNA n Nano (1 x 10 -9)

NF-κB Nuclear Factor Kappa B

NHL Non-Hodgkin lymphoma

NICD Notch intracellular domain

xiii NK Natural killer

NKG2D Natural-killer group 2, member D

NLPHL Nodular lymphocyte predominant Hodgkin lymphoma

NLS nuclear localization sequence

NOG NOG (NOD/Shi-scid/IL-2R γnull) oC Degree Celsius

PBMC Peripheral blood mononuclear cell

PBS Phosphate buffer saline

PCR Polymerase chain reaction

PD-1 Programmed cell death 1

PD-L1 Programmed cell death 1 ligand 1

PFA Paraformaldehyde

RA Rheumatoid arthritis

RANK Receptor activator of NF-κB

RBC Red blood cell

rpm Revolutions per minute

RS Reed-Sternberg

RT-PCR Reserve-transcriptase polymerase chain reaction

sCD137 Soluble CD137

SD Standard deviation

SDS-PAGE SDS-polyacrylamide gel electrophoresis

SHM Somatic hypermutation

SLE Systemic lupus erythematosus

TACI Transmembrane activator and calcium modulator and

cyclophilin ligand interactor

xiv TAP Transporter-associated proteins

TGF-β Transforming growth factor beta

Th1 T helper 1 cells

Th2 T helper 2 cells

TIL Tumor infiltrating lymphocytes

TNF Tumor necrosis factor

TNFR Tumor necrosis factor receptor

TNF-α Tumor necrosis factor alpha

TRAIL TNF-related apoptosis-inducing ligand

Treg Regulatory T cells

U Unit

µ Micro (1 x 10 -6)

xv CHAPTER 1 INTRODUCTION

Malignant disease is one of the most common causes of human mortality.

Despite all the efforts in improving patients’ prognosis, cancer remains one of the deadliest diseases. The bottleneck which we face in cancer treatment is probably due to restricted knowledge on cancer biology and its pathogenesis.

Using Hodgkin lymphoma as an example, there has been little breakthrough in the understanding of its pathogenesis since Hodgkin and Reed Sternberg cells were reported to be originated from B cells in the 90's (Kuppers et al., 1994).

Lacking of knowledge has restricted the effort to improve treatments and reduce side effects that accompany the usage of conventional treatment. The primary strategy of Hodgkin lymphoma treatment is mainly comprised of chemotherapy and radiotherapy. These kind of therapeutic methods have significantly prolonged the survival rate of Hodgkin lymphoma patients, nonetheless these therapies produce substantial toxicities as a side effect and still result in a high frequency of relapse in the long term (Evens et al., 2008;

Re et al., 2005).

Thus the motivation of this study is to find out a novel mechanism which could further explain the pathogenesis of Hodgkin lymphoma. The new knowledge is expected to lead to a new immunotherapy approach and ultimately improve the management for Hodgkin lymphoma in patients.

In this section, the current knowledge of Hodgkin lymphoma, its pathogenesis and roles of microenvironment in sustaining its growth will be discussed.

1 Apart from that, the functionality of CD137 and CD137L and their involvement in the development of various diseases will also be discussed.

1.1 Hodgkin Lymphoma

Hodgkin lymphoma (HL) is a type of hematological malignancy which usually affects lymph nodes of the patient. It was first characterized by Thomas

Hodgkin in 1832 (Hodgkin, 1832). Malignant growth of HL is commonly found at cervical of supraclavicular region and often coupled with other clinical symptoms like cyclical fever, night sweat and loss of body weight.

Due to the similarity of these symptoms with viral infection and other malignancies, histology is necessary to confirm the diagnosis of HL

(Townsend and Linch, 2012).

Unlike other types of solid tumors, HL is mainly comprised of infiltrating inflammatory cells, e.g. histiocytes (old name for tissue macrophages), lymphocytes, eosinophils, etc. The malignant cells in HL, multinucleated

Reed-Sternberg (RS) cells and their mononucleated variant Hodgkin cells usually comprise as little as 1% of the tumor mass (Re et al., 2005). HRS cells visually appear as a symmetrical bi-nucleated cell resembles an “Owl’s eye”

(Mani and Jaffe, 2009) which is very characteristic and important for the diagnosis of HL. Hodgkin cells and Reed-Sternberg cells have a similar immunophenotype (Abe et al., 1988) and are often combined and called

Hodgkin and Reed Sternberg (HRS) cells. Studies found that the majority of the HRS cells are derived from germinal center (GC) B cells which produce no functional BCR after undergoing the somatic hypermutation process. Under

2 physiological condition, these GC B cells which do not produce highly specific antibodies will be removed quickly in the GC via apoptosis but precursor of HRS cells are resistant to apoptosis and transform into HRS cells

(Kuppers and Rajewsky, 1998).

1.1.1 Etiology and pathophysiology

HL is classified into 2 major subtypes, classical HL and nodular lymphocyte predominant Hodgkin lymphoma (NLPHL). NLPHL is not recognized as classical HL because the malignant cells in NLPHL, lymphocytic and histiocytic (L&H) cells, have low or null expression of classical HRS cell markers (CD30, CD15, bcl-2) but are positive for CD20 which is a B cell marker (Uherova et al., 2003). Molecular studies have shown that both L&H cells and HRS cells might derive from GC cells but L&H cells are derived from more mature and selected GC B cells (Kuppers et al., 1998). Classical

HL is further grouped into 4 subtypes namely: lymphocyte predominance, mixed cellularity, lymphocyte depleted and nodular sclerosing. Each of these subtypes has a distinct histological morphology and phenotype as summarized in Table 1. The rate of occurrence and prognosis vary between each subtype

(Mani and Jaffe, 2009; Townsend and Linch, 2012).

As compared to reactive lymph nodes, HL have increased number of T cells

(~70%) in their microenvironment. Notably most of the T cells are CD4 + T cells, but not CD8 + T cells which posses cytotoxic activity (Gorczyca et al.,

2002). This might be one of the factors which lead to the immune evasion of

3 HRS cells. Besides T cells, the HL microenvironment mainly consists of macrophages, B cells and eosinophils.

Table 1. Characteristic of Hodgkin lymphoma subtype (Schmitz et al., 2009)

Lymphocyte predominance Nodular growth with large number of non-malignant infiltrating B cells

Lymphocyte depleted High amount of HRS cells with little lymphocytes and histiocytes

Mixed cellularity Infiltrating of lymphocytes, histiocytes, eosinophils and neutrophils without sclerosis

Nodular sclerosis The most common subtype. Similar with mixed cellularity subtype but with collagen bands surrounding tumor nodular cluster

In 1994, Kuppers et al. showed that HRS cells are originated from GC B cells

after discovering the rearrangement of variable region of immunoglobulin

genes in HRS cells. This finding suggests that the precursors of HRS cells are

derived from GC B cells which have undergone somatic hypermutation

(SHM). During SHM, activation-induced deaminase (AID) expression in GC

B cell is upregulated and this initiates a very active single nucleotide

substitution on variable region of immunoglobulin sequence in order to

generate antibodies which will be more effective and specific against the

foreign antigen (Teng and Papavasiliou, 2007). However, this process is

highly randomized and error-prone. Thus as a protective mechanism, any GC

B cell which has low antigenic affinity or with crippled B cell receptor (BCR)

will be eliminated by apoptosis.

4 Further characterization of SHM on HRS cells revealed that at least part of the

HRS cells have crippled BCR (Kuppers and Rajewsky, 1998). This means that the precursor of HRS cells that are supposed to be apoptotic in GC after SHM but somehow these precursor cells escaped the elimination pathway and transformed into HRS cells. The detailed etiology of HL and transformation of

HRS cells is still unknown but the disease has been associated with Epstein–

Barr virus (EBV) infection. EBV infection was found in 30-50% of HL cases in developed countries, and escalated to nearly 50-100% in some other countries in Asia and Africa. The high association of EBV with HL suggests its potential involvement in the pathogenesis of HL. In addition to that, EBV infection might be negatively associated with a favorable prognosis of HL, although this finding is still controversial (Ambinder, 2007).

EBV only expresses a handful of viral proteins but has a strong association with Burkitt lymphoma and HL pathogenesis. It was shown that EBV infection can immortalize GC B cells which do not have functional BCR molecules (Bechtel et al., 2005). This observation may explain how HRS cells manage to evade apoptosis in GC after they acquire crippled BCR sequences during SHM. Studies have also shown that EBV viral proteins can substitute some of the missing but essential survival signalling in HRS cells after SHM.

Latent Membrane Protein 1 (LMP1) and LMP2 are among the most well characterized viral protein. LMP1 has been shown to induce a similar signalling pathway and function as compared to CD40, an important B cell co- stimulatory molecule. LMP1 is also constitutively activated in infected cells

(Lam and Sugden, 2003; Rastelli et al., 2008). On the other hand, LMP2 has

5 been shown to mimic BCR function and signalling while at the same time inhibiting normal antigen presentation function of BCR (Dykstra et al., 2001).

Since BCR and CD40 signalling are the most important signals required for B cell activation, one would expect that EBV infected B cells become self sustaining in survival signalling and increase proliferation which are two of the hallmarks for cancer. Surprisingly, LMP1 or LMP2 overexpression in B cells has been shown to downregulate B cell related markers like BCR and

CD20 on B cells and at the same time upregulate transcription of the genes which are related to the HRS cells phenotype (Portis et al., 2003; Vockerodt et al., 2008). These results have undoubtedly supported the involvement of EBV infected GC B cells in HRS cell emergence.

1.1.2 Immunosuppressive microenvironment

Like many other solid malignancies, HL is infiltrated with tumor infiltrating lymphocytes (TIL) which are unable to eradicate established tumors in vivo . It is suggested that this immunosuppressive nature of tumor microenvironment is due to predominance of a T helper 2 (Th2) and Regulatory T (Treg) cell response, which will be discussed below.

TIL which are found in HL lesions are predominantly CD4 + CD25 + Treg cells with high expression of IL-10 and Cytotoxic T-Lymphocyte Antigen 4

(CTLA-4). The TIL is also found to have a lower proliferation capacity and a reduced Interferon-gamma (IFN-γ) release (Koenecke et al., 2008; Marshall et

al., 2004). Both of these are very important measurements for the anti-tumor

activity of T cells. This is because the release of IFN-γ modulates T helper 1

6 (Th1) or cell-mediated responses and that the proliferation is required for the expansion of antigen specific T cells. Downregulation of these T cell responses might reduce T cell cytotoxicity, and might be the reason that leads to the immune evasion of HRS cells. Infiltration of Treg cells into HL might be partly due to the upregulation CCL17 and CCL22, important chemoattractants of CD4+ CD25 + Treg in HL patients (Niens et al., 2008).

Interestingly, co-culturing of CD4 + T cells with a HRS cell line alone was shown to be able to induce Foxp3 + Treg cell differentiation (Tanijiri et al.,

2007). This suggests that at least some of Treg cells might be induced in the

tumor microenvironment due to the influences of HRS cells.

Besides that, HRS cells were also shown to preferentially secrete Th2

cytokines and chemokines, especially Interleukin-6 (IL-6), IL-13 and IL-21

(Lamprecht et al., 2008; Skinnider and Mak, 2002). The bias toward a Th2

response in tumor microenvironment is commonly regarded as a survival

advantage for cancer cells, as it suppresses Th1 responses and proper cell-

mediated immunity which are required for tumor rejection. Besides that, IL-13

and IL-21 were also shown to be important autocrine growth factors for HRS

cells (Kapp et al., 1999). However, c-Maf + Th2 phenotypic cells in TIL have

been shown to be associated with improved prognosis of HL patients.

(Schreck et al., 2009). The reason of this contradictory observation has yet to

be elucidated but suggests that it may be the ratio and functional balance

between Th1 and Th2 cells in the HL microenvironment that is more

important than their abundance in the pathogenesis of the tumor.

7 It is interesting to learn that EBV infection plays a role in the immunosuppressive environment in HL. Apart from inducing a HRS phenotype, LMP1 can also induce IL-10 secretion from CD4 + T cells, and

subsequently suppress T cell proliferation and IFN-γ release (Marshall et al.,

2003). Besides that, LMP1 was shown to increase the promoter activity of

Programmed cell Death 1 Ligand 1 (PD-L1) and lead to the upregulation of

PD-L1 in HL (Green et al., 2012). PD-L1 will then interact with Programmed

cell Death 1 (PD-1) which is expressed by T cells in HL and reduce IFN-γ release from T cells (Yamamoto et al., 2008). These are all potential mechanisms by which EBV can manipulate the microenvironment in HL, reduce Th1 responses and lead to a less cytotoxic environment. This immunosuppressive mechanism has been developed evolutionarily by virus to prevent cell mediated immunity against EBV-infected cells. However, EBV infection also protects HRS cells from an anti-tumor response and potentially assists in their survival. Moreover, EBV infected tumor cells were shown to upregulate CCL-17 and CCL-22 secretion which will cause the infiltration of

Treg cells (Takegawa et al., 2008) and further impair host's immune responses.

1.1.3 Association of HL with members of tumor necrosis factor receptor family

Tumor necrosis factor receptor (TNFR) family members have been intensively studied in the context of HL (Skinnider and Mak, 2002). Among them, CD30 is expressed by most of the HRS cells and is one of the most studied diagnostic markers for HL. The function of CD30 on HL cells has not been elucidated at the current stage, and there is speculation that CD30 signaling

8 might not be effective in HRS cells (Hirsch et al., 2008). Nonetheless, plenty of individual reports have shown CD30 signaling in HRS cells. Notably, it was shown that CD30 can increase Nuclear Factor Kappa B (NF-κB) and cellular

FLICE inhibitory protein (cFLIP) activation in HRS cells (Boll et al., 2005) which will enhance HRS cell proliferation and protect them from Fas- mediated cell death (Dutton et al., 2004; Mathas et al., 2004). Interestingly, proliferation of HRS cells decreases due to deactivation of NF-κB and

Extracellular Signal-regulated Kinases (ERK) 1/2 pathway when CD30 is knocked down (Watanabe et al., 2011). In addition to that, CD30 can also inhibit T cell proliferation in co-culture experiments (Su et al., 2004). Due to high CD30 and HL association, different attempts were made to utilize CD30 as a potential therapeutic target to specifically kill CD30 + cells (Dietlein et al.,

2010; Gualberto, 2012).

Besides CD30, other TNFR family members like CD40, CD95, Receptor

activator of NF-κB (RANK), Transmembrane Activator and Calcium modulator and cyclophilin ligand Interactor (TACI), and B Cell Maturation

Antigen (BCMA) are also shown to be expressed by HRS cells (Chiu et al.,

2007; Fiumara et al., 2001; Kim et al., 2003). The finding of CD95 expression on HRS cells is very interesting, as CD95 is capable of inducing cell death by

CD95-mediated apoptosis upon ligation with CD95 ligand (CD95L) (Wajant,

2002). However, despite expressing high levels of CD95, HRS cells are resistant to CD95 mediated apoptosis (Metkar et al., 1999; Re et al., 2000).

This resistance might be due to high cFLIP expression in HRS cells (Dutton et al., 2004; Mathas et al., 2004). Besides that, HRS cells also express high levels

9 of CD95L, but co-expression of CD95 and CD95L cannot cause apoptosis on

HRS cells (Metkar et al., 1999; Verbeke et al., 2001). On the other hand,

CD95L expressed by HRS cells might induce apoptosis of infiltrating T cells which express CD95 upon activation.

CD40, TACI and BCMA expression on HRS cells were found to be functional and capable to increase proliferation of HRS cells upon stimulation (Carbone et al., 1995; Chiu et al., 2007). Besides that, CD40-CD40L interaction can induce rosette formation between T cells and HRS cells and cause an upregulation of Interferon regulatory factor 4 (IRF4) which might contribute to their pathogenesis (Aldinucci et al., 2010). Lastly, our group has shown that

CD137 is expressed ectopically by HRS cells (Figure 1).

Ectopic CD137 expression has a high correlation with HL with 86% of HL patients being positive for CD137 on HRS cells while CD137 expression was found to be absent in other subtypes of lymphoma as well as non-malignant reactive lymph nodes (Ho et al., 2012). Ectopic CD137 expression on HRS cells was also reported separately by another research group (Anderson et al.,

2012). Nonetheless, the function of CD137 and the mechanism of its upregulation in HRS cells have yet to be investigated. Interestingly, the EBV viral protein LMP1 was shown to upregulate the CD137 promoter activity in T cells and natural killer (NK) cells (Arai, 2009) but whether EBV causes the ectopic expression of CD137 in HRS cells has yet to be studied.

10 Therefore the main interest in this study is to investigate the effects of ectopic expression of CD137 on the survival of HRS cells. In the next section of

Introduction, the biology of CD137 and its involvement in disease state will be discussed.

Figure 1. CD137 expression on Hodgkin lymphoma Immunohistochemistry staining of paraffin-embedded Hodgkin lymphoma tissue for CD137 expression. High level of CD137 is expressed in cytoplasmic, Golgi and membrane area of HRS cells. A (400x), B (750x) and C. Isotype Staining (750x). (Taken from (Ho et al., 2012))

11 1.2 CD137 and CD137L

1.2.1 Expression and characteristic of CD137 and CD137L

CD137 is a member of TNFR superfamily and is also known as Induced by

Lymphocyte Activation (ILA) or 4-1BB (Kwon and Weissman, 1989;

Schwarz et al., 1993). CD137 is mainly expressed by T cells, dendritic cells

(DC) and NK cells, and its expression is strictly activation-dependent under normal physiological conditions (DeBenedette et al., 1995; Schwarz et al.,

1995). CD137 Ligand (CD137L), the only known binding partner of CD137, is a member of Tumor Necrosis Factor (TNF) superfamily and also known as

4-1BB ligand (4-1BBL) (Goodwin et al., 1993; Rabu et al., 2005). CD137L is expressed mainly by antigen presenting cells (APC) especially on monocyte,

B cells and hematopoietic progenitor cells (Jiang et al., 2008b; Zhou et al.,

1995).

Binding of CD137 to CD137L induces a bidirectional signalling whereby the signals will be conveyed downstream of CD137 and CD137L simultaneously into the cells they are expressed on (Figure 2). The combination of CD137-

CD137L signalling has been shown to be involved in the activation, maturation and differentiation of various immune cells and hematopoietic progenitor cells (Jiang et al., 2008a; Schwarz et al., 1996; Thum et al., 2009).

CD137 and CD137L are involved in the modulation of immune responses.

CD137 is expressed by T cells upon activation as a co-stimulatory molecule.

Upon stimulation, CD137 can further enhance activation of T cells and their proliferation, induce IFN-γ release and also the cytolytic activity of T cells

12 (Cannons et al., 2001). Besides that, CD137 can reverse anergy and activation induced cell death (AICD) of T cells, and make restore their cytotoxicity in an immunosuppressive environment (Hernandez-Chacon et al., 2011; Wilcox et al., 2004). CD137L expressing tumor cells are shown to induce higher T cell cytotoxicity and Th1 responses, and CD137L knock out mice were shown to have increased chances to develop malignant growth (Li et al., 2008;

Middendorp et al., 2009). All of these results have confirmed the roles of

CD137 in stimulating Th1 and cytotoxic T cell responses, and also in enhancing cell-mediated immunity of T cells. Due to its therapeutic potential,

CD137 has been studied extensively for its potential usage in tumor immunotherapy. The details will be discussed in next section.

On the other hand, CD137L is constitutively expressed by APC and the reverse signaling via CD137L is best described in myeloid cells. CD137L stimulation induces activation and proliferation of monocytes (Langstein et al.,

1998; Langstein et al., 1999; Tang et al., 2011). Apart from that, CD137L stimulation can also induce human monocyte differentiation into DC. The

CD137-differentiated DC are functional and have been shown to induce higher

T cell responses than conventional DC which are generated by Granulocyte

Macrophage Colony-Stimulating Factor (GM-CSF) + IL-4 (Kwajah and

Schwarz, 2010). However, the outcome of CD137L stimulation on human and murine monocytes seems to be different. CD137L stimulation on murine monocytes fails to induce differentiation into inflammatory DC, and this suggests that there might be a species difference in CD137L signaling (Tang et al., 2011).

13 CD137L has been found to be expressed by B cells but the function of

CD137L in B cells is poorly defined. Nonetheless, it is suggested that CD137L might be involved in the proliferation and also differentiation of B cells to antibody producing cells (Shao and Schwarz, 2011).

Moreover, stimulation of CD137L is able to promote the differentiation of both human and murine hematopoietic progenitor cells into myeloid lineage cells, mainly monocytes and macrophages. CD137L stimulation was also shown to be able to suppress G-CSF-induced granulocytic cell differentiation of hematopoietic progenitor cells (Jiang et al., 2008a; Jiang et al., 2008b; Jiang and Schwarz, 2010).

Antigen T cell Presenting cell

CD137 CD137L

Activation Act ivation Proliferation Proliferation Cytotoxicity Differentiation Prevent ion of anergy

Figure 2. CD137 and CD137L bidirectional signaling CD137 is mainly expressed by activated T cells while CD137L is predominantly expressed by APC (e.g. B cells and monocytes). When CD137 on T cells binds to CD137L on APC, co-stimulation will be induced which enhances T cell activation and proliferation. At the same time, a reverse signaling will also be initiated into APC which leads to the activation of APC and induction of differentiation, proliferation and survival of APC.

15 1.2.2 Targeting CD137 for immunotherapy

Due to the important function of CD137 in modulating immune responses, many researchers have tried to utilize CD137 as a means of immunotherapy.

Among all the approaches, the usage of agonistic CD137 antibodies to enhance T cell anti-tumor activity has attracted much attention. For instance, the effects of agonistic CD137 antibody on tumor rejection have been studied extensively in vivo using several kinds of tumor model for example melanoma,

liver cancer, lung cancer, myeloma and thymoma (Ju et al., 2005; Martinet et

al., 2002; Murillo et al., 2008; Murillo et al., 2009; Zhu et al., 2009). Despite

the differences in the models and methods used, they all showed that agonistic

CD137 antibody effectively suppresses the tumor growth in vivo . Agonistic

CD137 antibody is also currently in clinical trial to treat different kinds of cancer (Lynch, 2008; Wang et al., 2009).

TIL are the lymphocytes that migrate into the tumor mass from the bloodstream. TIL play an important role in tumor rejection and high TIL count is usually associated with good prognosis (Eerola et al., 2000). Many independent groups have reported that the TIL count is increased after anti-

CD137 antibody treatment (Ju et al., 2005; Wilcox et al., 2002) and interestingly, the majority of the TIL was found to be CD4 + and CD8 + T cells

(Ju et al., 2005; Martinet et al., 2002; Murillo et al., 2008; Wilcox et al., 2002).

However, the effects of anti-CD137 antibody treatment on tumor regression

were halted when CD8 + T cells were depleted or when the antigen presenting capacity of DC was affected. Anti-CD137 antibody treatment was also shown to induce the up-regulation of Natural-Killer Group 2, member D (NKG2D),

16 IL-2, IFN-γ and Tumor Necrosis Factor alpha (TNF-α) and the downregulation of IL-10, Transforming Growth Factor beta (TGF-β) and IL-4 production (Zhu et al., 2009). This indicates that a Th1 immune response is triggered after anti-CD137 antibody treatment while the Th2 response is suppressed.

On the other hand, secretion of IFN-γ was also found to be enhanced at the tumor site after anti-CD137 antibody treatment (Ju et al., 2005). IFN-γ was shown to increase the expression of Major histocompatibility complex (MHC) class I and class II in malignant cells (Ju et al., 2005; Murillo et al., 2008) which can lead to increased immunogenicity of tumor cells. This might be significant as TIL accumulation and a tumor suppression effect after anti-

CD137 antibody treatment cannot be observed in IFN-γ knock out mice (Ju et al., 2005; Murillo et al., 2008; Wilcox et al., 2002). This implies that IFN-γ is important in anti-CD137 antibody induced anti-tumor activity.

Furthermore, immunomodulating effects of agonistic CD137 antibody could only be observed if anti-CD137 antibody treatment was given after the mice were inoculated with cancer cell (Ju et al., 2005). This suggests that the agonistic CD137 antibody might act on T cells that had been primed with tumor antigen but were unable to eliminate the tumor cells due to the lack of proper co-stimulation, and that the CD137 antibody restored T cell anti-tumor activity.

17 1.2.3 CD137 and CD137L in malignant diseases

Despite being regulated strictly under normal circumstances, the expression of

CD137 and CD137L has been found to be dysregulated in many disease models. Ectopic expression of CD137 was found in follicular DC tumor, T cell lymphoma, lung cancer and osteosarcoma (Anderson et al., 2012; Lisignoli et al., 1998; Zhang et al., 2007). As discussed earlier, CD137 is also ectopically expressed in 80-90% of HL cases. Constitutive expression of CD137 on these tumor cells is extraordinary as CD137 expression is activation-dependent under physiological condition. Besides that, CD137 expression was also found on the vessel wall of malignant tumors (Broll et al., 2001) which strengthens the correlation of ectopic CD137 expression with malignant growth.

Soluble CD137 (sCD137), a product of differential splicing of CD137 messenger RNA (mRNA), has been considered as an antagonist in nature, and acting as a competitor of membrane bound CD137 (Shao et al., 2008). CD137 has been found to be upregulated in many autoimmune diseases, for example in systemic lupus erythematosus (SLE), pancreatitis and rheumatoid arthritis

(RA) (Michel et al., 1998; Shao et al., 2008; Shao et al., 2012). In addition to autoimmune disease, serum levels of sCD137 were also found to be upregulated in many hematological malignancies, especially in non-Hodgkin lymphoma (Furtner et al., 2005).

CD137L was found to be expressed by various cancer cells as well (Salih et al.,

2000; Wang et al., 2008). Surprisingly, these studies also showed that

CD137L expressed by these cancer cells is functional and can induce IFN-γ

18 release from T cells upon ligation with CD137, indicating increases of anti- tumor responses of T cells. Furthermore, CD137L expressed by multiple myeloma (MM) cells has been shown to induce apoptosis in MM cell lines upon stimulation (Gullo et al., 2010). The reason for these cancer cells expressing CD137L despite the potential disadvantage for tumor survival still remains unknown. One of the possibilities is that cancer cells might obtain some benefits from the downstream signaling of CD137L which has yet to be identified.

Notably, CD137L is expressed by different subtypes of non-Hodgkin B cell lymphoma but absent in HL (Zhao et al., 2012). This coincides with the increase of serum sCD137 in non-Hodgkin lymphoma (NHL), which is found to be absent in HL (Furtner et al., 2005). In contrast, ectopic CD137 is only expressed by HL and rarely in NHL (Anderson et al., 2012). This complicated relationship of CD137, CD137L and sCD137 expression in HL and NHL might suggest that these proteins have an inter-dependent relationship in the pathogenesis of HL and NHL.

1.3 Trogocytosis

Trogocytosis is an intercellular transfer of membrane protein together with a fragment of plasma membrane (Davis, 2007). Its function is to facilitate the transfer of membrane protein between cells that have strong cell to cell interaction (as further discussed below). The exact mechanism for trogocytosis remains unclear so far. Nonetheless, it has been shown that the occurrence of trogocytosis is dependent on both cell types and on the surface protein

19 expression on the cells which are involved in the trogocytosis (Aucher et al.,

2008; Hudrisier et al., 2007). Cell to cell contact is also required for trogocytosis to occur (Beum et al., 2008; Gary et al., 2012; Pham et al., 2011).

The force involved in the trogocytosis was shown to be equivalent to the force needed to break a strong protein-protein interaction (Bell, 1978). Thus, it is not surprising that trogocytosis is an active process that requires hydrolysis of

ATP, activation of downstream signal (eg: src kinase) and remodeling of actin filament (Tabiasco et al., 2002).

Transfer of cell surface material through trogocytosis was first described between DC and T cells, and the extent of this transfer is associated with DC-

T cells affinity and the T cell antigen specificity (Uzana et al., 2012).

Immature DC which express fewer MHC and costimulatory molecules were found to be less active in trogocytosis compared to mature DC (Gary et al.,

2012). It has been shown that transfer of CD86 from DC to CTLA-4 expressing Treg cells ceased when the CTLA-4 was neutralized (Qureshi et al.,

2011). In other words, this experiment suggests that the recipient cells must express the binding partner of the transferred material and the trogocytosis is ligand dependent. However, Gu et al. had opposed this theory by showing that

CD80 and CD86 transfer from mature DC to Treg cells is independent of

CD28 or CTLA-4 expression. Furthermore, transfer of OX40 ligand to OX40- negative cells was also demonstrated (Baba et al., 2001), which again suggested that ligand specificity is not necessary for the transfer of materials between cells through trogocytosis. However, this ligand-independent transfer

20 of cell protein could be due to the transfer of bystander proteins which are in near proximity to trogocytosis site (He et al., 2007).

The specific function of the trogocytosed protein is depended on the context of the transfer. However, most of the recent studies suggest that trogocytosis is more likely to be a negative feedback or a regulatory function (Helft et al.,

2008; Qureshi et al., 2011). For example, PD-L1 transfer from APC to T cells was shown to initiate apoptosis of PD-1 positive T cells (Gary et al., 2012). In addition, MHC class I molecules which co-transfer with MHC class II molecules from APC to CD4 + T cells causes the CD4 + T cells to become targets of cytotoxic T lymphocytes (CTL) cytolytic activity (Cox et al., 2007).

Interestingly, trogocytosis was also observed between cancerous cells and their surrounding cells. Transfer of Human Leucocytes Antigen G (HLA-G) from tumor cells to monocytes can inhibit PHA activated Peripheral Blood

Mononuclear Cell (PBMC) proliferation (HoWangYin et al., 2010). Similarly, transfer of material via trogocytosis was shown to upregulate N-cadherin on breast cancer cell lines, and potentially to enhance its invasiveness (Lis et al.,

2010). These findings suggest that trogocytosis might play a role in pathogenesis of malignant diseases. Nonetheless, the study of trogocytosis and its involvement in cancer is still at an early stage.

21 1.4 Research objectives

Above we discussed that ectopic CD137 expression is frequent in HL. CD137 which can enhance the anti-tumor effect of T cells is overexpressed by HRS cells, while HL has a Treg and Th2 predominant immunosuppressive microenvironment. Is it possible that ectopic CD137 expression on HRS cells competes with the CD137 expressed by T cells and hence reduces T cell anti- tumor activity? Besides that, CD137L which is commonly expressed by B cell lymphoma was also found missing in HL. All this evidence may indicate that ectopic CD137 expression plays a role in the pathogenesis of HL, and most likely benefits HL.

Hence, the main objective of this study is to investigate the correlation of

CD137 expression with the improved survival of HRS cells in HL, and to further examined the mechanisms behind this improved survival.

Specific objectives of the study are:

1. To elucidate the effects of ectopic CD137 expression on HRS cells and

whether it can enhance survival of HRS cells

2. If CD137 is beneficial to HRS cells survival, the mechanism of this

beneficial effect shall be determined

3. To confirm whether CD137 can induce CD137 signaling into HRS

cells

22 The results obtained in this study might provide new insights into the immune escape mechanisms of HL and other malignant diseases, and suggest a new regulatory mechanism of the CD137-CD137L system.

23 CHAPTER 2 MATERIALS AND METHODS

2.1 Cells and cell culture

2.1.1 Cell lines

HRS cell lines (HDLM-2, KM-H2, L-428, L-540, and L-1236) were obtained from DMSZ-German Collection of Microorganisms and Cell Cultures. CD137 expression on KM-H2 cells was knocked down by Ms Tan Teng Ee, NUS, using shRNA plasmid and Nucleofector Technologies (Lonza). CD137 knocked down KM-H2 cells were denominated as KM-H2-CD137 - cells. In addition, KMS-11 cells were kindly donated by Dr Cheng, National University

Health System (NUHS), Singapore.

These cell lines were cultured in RPMI-1640 medium (Sigma Aldrich) supplemented with 10-20% of fetal bovine serum (FBS, Biowest) and 2 mM of L-glutamine (Gibco) and were passaged every 3 to 4 days.

2.1.2 CD137 overexpressing cell lines

CD137 overexpressing L-428 and L-1236 cells were generated using lentiviral transfection. The CD137-pLenti6/V5-D-TOPO vector (Invitrogen) was constructed by Moh Mei Chung, NUS, while lentiviral transduction of L-428 and L-1236 cells was performed at the Phoenix laboratory of Gene and Cell therapy, NUS, with the help of Dr Gan Shu Uin. Selection of positive clones was done by culturing the transfected cells in complete medium supplemented with predetermined doses of Blasticidin (5 µg/ml for L-428 cells and 0.5

µg/ml for L-1236 cells). After the transfection, CD137 expression on L-428 and L-1236 cells was confirmed by flow cytometry. CD137 overexpressing L-

24 428 and L-1236 cells were denominated as L-428-CD137 and L-1236-CD137, respectively.

2.1.3 PBMC isolation

Isolation of PBMC was done using density gradient centrifugation. In brief, human buffy coats of healthy donors were purchased from Blood Donation

Center, National University Hospital (NUH), Singapore. The buffy coats were diluted with Phosphate Buffer Saline (PBS) and 2 mM EDTA before it was layered on top of Ficoll-Paque PLUS (GE Healthcare). Centrifugation for 30 min at 400 g pelleted the RBC below the Ficoll-paque layer while plasma and

PBMC, a white fluffy thin layer formed the upper layer of the Ficoll-pague as shown in the Figure 3. PBMC were then isolated carefully from the layer and any contamination of RBC in PBMC was removed by incubating PBMC in

Red Blood Cell (RBC) lysis buffer for 5 min. PBMC were then washed with

PBS and kept at room temperature for further use.

PBMC were cultured in RPMI1640 medium supplemented with 10% FBS, 2 mM of L-glutamine and penicillin (100 U/ml) and streptomycin (100 µg/ml).

Density gradient Plasma centrifugation Buffy coat PBMC 400g, 30 min Ficoll Ficoll RBC

Figure 3. PBMC isolation with density gradient centrifugation. A schematic diagram shows isolation of PBMC using density gradient centrifugation. A diluted buffy coat was gently layered on top of the Ficoll solution before centrifuged at 400 g for 30 min. After centrifugation, PBMC were found accumulated as a thin fluffy layer between plasma and Ficoll and cells were isolated carefully for downstream experiment.

2.1.4 T cells, B cells and Monocytes purification

Different subsets of PBMC were purified using magnetic activated cell sorting

(MACS) by Miltenyi Biotec. In general, microbead-conjugated antibodies

(anti-CD3 for T cells, anti-CD20 for B cells, and anti-CD14 for monocytes isolation) were diluted in MACS buffer (PBS with 0.1%BSA and 2 mM

ETDA) and incubated with PBMC for 15 min at 4 oC. The microbeads labeled

PBMC were then loaded into a magnetic column. Positively labeled cells will

be retained in the column by magnetic force while other non-specific cell

types will be removed from the column by washing with MACS buffer. The

26 positively labeled monocytes or B cells were later flushed out from the column and ready to be utilized.

2.1.5 Storage of cell lines and primary cells

Both cell lines and primary cells were occasionally frozen down for later use.

Cell lines were resuspended in RPMI with 20% Fetal Bovine Serum (FBS) and 10% Dimethyl Sulfoxide (DMSO) with minimum cell concentration of 4 x 10 6 cells/ml. On the other hand, primary cells were resuspended in RPMI with 40% FBS and 10% DMSO with minimum cell concentration of 2 x 10 7 cells/ml. DMSO was added dropwise to the cell suspension. The cell suspension was then transferred to CryoTube (Nunc) and into a freezing container (Nalgene). The freezing container was then stored in -80 oC freezer overnight, before the cryovials were transferred into liquid nitrogen tank.

2.2 Antibodies and reagents

An anti-CD3 antibody, clone Okt3, was purified from hybridoma supernatant by our previous colleague and was stored in -80 oC for further used. Anti-

CD137 antibody clone BBK-2 was purchased from Labvision while clone

JG1.6a was purchase from Novus Biologicals . On the other hand, anti-

CD137L antibodies, clone 41B436, clone 5F4, and clone C65-485 were

obtained from Adipogen, Biolegend and BD Pharmigen, respectively. Anti-

CD137L cytoplasmic domain antibody clone EPR1172Y was obtained from

Genetex.

27 Phycoerythrin (PE) labeled anti-CD137 antibody (clone 4B4-1) was purchased from BD Pharmigen while PE labeled anti-CD137L antibody (clone 5F4) was obtained from Biolegend. Anti-CD11b-APC antibody and anti-CD20-APC antibody were obtained from eBioscience. Mouse IgG1 (clone MOPC-21) which was used as isotype control was bought from Sigma-Aldrich and PE- conjugated mouse IgG1 κ was obtained from eBioscience. PE and APC labeled donkey anti-mouse secondary antibodies were purchased from eBioscience while Goat anti-mouse AF488 and Goat anti-rabbit AF597 secondary antibodies were obtained from Invitrogen. Monodansylcadaverine (MDC) was purchased from Sigma Aldrich. CFSE dye and PKH-26 dye were purchased from Invitrogen and Sigma-Aldrich, respectively.

2.3 Flow cytometry

Different combination of antibodies and isotypes were employed throughout this study. Nonetheless, basic protocol of flow cytometry was the same regardless of the experiment. First, cells were harvested and washed once with

Fluorescence-Activated Cell Sorting (FACS) staining buffer (PBS, 0.5% FBS and 0.02% sodium azide). Then the cells were stained with either unconjugated primary antibody or primary antibody conjugated with fluorochrome at 4 oC for 15 min followed by washing with FACS staining buffer. Cells stained with unconjugated primary antibody were then stained with secondary antibody conjugated with fluorochrome at 4 oC for 15 min. The

cells were then washed once with FACS staining buffer.

28 After the staining, the cells were resuspended in 400 µl of staining buffer with

1% Paraformaldehyde (PFA) as fixative. The stained cells were analyzed using CyAn ADP analytical counter (Beckman Coulter) or LSR Fortessa analyzer (BD Biosciences) depending on the combination of fluorochrome used. The data was acquired in flow cytometry standard 2.0 (FCS 2.0) format and analyzed with Summit analysis software (Dako).

2.4 Enzyme linked immunosorbent assay (ELISA)

ELISA for IFN-γ from R&D Systems was performed according to the manufacturer's protocols. In brief, the diluted (if necessary) samples were incubated for 2 h in wells pre-coated with capture antibody. This was then followed by incubation with detection antibody for 2 h, streptavidin-HRP for

30 min and eventually TMB substrate (BD bioscience). The plate was washed

3 times with washing buffer (PBS and 0.05% Tween-20) between each incubation step. The reaction was stopped with 2N H2SO 4 before analyzed using BioRad 680 microplate reader. Every sample was analyzed in duplicate.

2.5 Western Blot

2.5.1 Protein purification

Cell samples were pelleted down and washed with ice cold PBS. The cell pellets were then resuspended in RIPA buffer (50 mM Tris, 150 mM NaCl,

0.1% SDS, 0.5% sodium deoxycholate, 1% Triton-X and 1 mM of PMSF), incubated on ice for 10 min with gentle agitation and vortexed twice at maximum setting during the incubation period. After the incubation, the cell

29 lysates were centrifuged at 16,000 rpm for 10 min and the supernatants were aspirated and stored at -80 oC for further use.

2.5.2 Electrophoresis and electroblotting

Protein samples were first electrophoresed on an acrylamide gel (Appendix I)

using SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to

a PVDF membrane using wet transfer method. The apparatus used for

electrophoresis and electroblotting were obtained from Biorad. In brief, equal

amount of protein samples from each condition were boiled at 99 oC with

Laemmli Buffer (63 mM Tris, 10% glycerol, 2% SDS, 0.0025% bromophenol

blue, 1.25% of 2-mercaptoethanol) for 10 min. The denatured protein samples

were then loaded and electrophoresed on a 12% acrylamide gel at 150 V for

70 min. Precision Plus protein standard (Biorad) was used as molecular

marker. PVDF membrane was pre-activated with ethanol (30 sec in absolute

ethanol, 2 min in distilled water and 5 min with transfer buffer) before usage.

After electrophoresis, protein samples were transferred from polyacrylamide

gel to pre-activated PVDC membrane at 100 V, max. 400 mA, for 60 min .

2.5.3 Antibody probing and visualization

The PVDF membrane was blocked with blocking buffer (PBS, 5% skim milk,

0.05% Tween20) for 1 h at room temperature and probed with primary

antibody overnight at 4 oC. The membrane was then probed with secondary antibody for 1 h at room temperature. Antibodies were diluted in blocking buffer. Washing step was performed between each incubation step with adequate amount of washing buffer (PBS, 0.05% Tween20) for at least 3 x 10

30 min. The membrane was then visualized using SuperSignal West Pico or

SuperSignal West Pico substrate solution (Pierce) depending on the strength of the signal and exposed to CL-XPosure film (Pierce).

2.6 Co-immunoprecipitation (Co-IP)

CD137L and CD137 were co-immunoprecipitated by anti-CD137 antibody. In brief, cell samples were lysed under non-denaturing condition using non- denaturing lysis buffer (PBS, 1% Triton-X, 1 mM PMSF) for 10 min on ice with gentle agitation and vortex twice throughout the incubation. Cell lysates

ο were then collected by centrifugation for 10 min at 16,000 rpm at 4 C. Protein concentration in total cell lysate was determined using a Bradford protein assay (Biorad). For every 1 mg of protein sample, 50 µl of protein-G coated agarose beads (Pierce) and 2 µg of anti-CD137 antibody (clone BBK-2) were added and the mixture was incubated overnight at 4 oC on a rotator. On day 2,

agarose beads were harvested by centrifuge at 16,000 rpm for 10 min and

washed 4 times with complete lysis buffer. Protein was eluted by boiling the

agarose beads in Laemmli buffer for 10 min and the eluted samples were

subjected to Western Blot technique as described above for detection of

CD137L.

2.7 Confocal microscopy

Cell samples were harvested, washed with PBS and fixed with 1% PFA for 15

min at 37 oC. The samples were then permeabilized with 0.2% Triton-X for 10 min at room temperature. The samples were blocked with 1% BSA for 1 h at

ο room temperature. Primary antibody staining was done at 4 C overnight

31 followed by secondary staining which was performed at room temperature for

1 h. The samples were thoroughly washed with PBS for at least 3 times in between each step. The samples were counterstained with DAPI for 15 min at

ο 37 C and mounted on poly-lysine slide with ProLong Gold antifade reagent

(Invitrogen). The slides were air-sealed with nail-polish and images were

acquired using Olympus Fluoview FV1000 TIRF with the help from Ms Lee

Shu Ying, Confocal Microscopy Unit, NUHS.

2.8 Endocytosis inhibition assay

Endocytosis inhibitor, MDC, was reconstituted with DMSO to the stock

concentration (60 mg/ml) and was stored at -20 oC. Prior to treatment, the stock solution was diluted with DMSO to 100 times of the desired concentrations respectively and was added to cell samples at constant ratio of 1:100. This is to ensure that the addition of DMSO in each of the condition is constant so as to avoid the unspecific effect caused by DMSO. Cell viability assay was done to determine the toxic dose of MDC. DMSO alone was used as a negative control.

2.9 Cell viability assay

Cell viability was determined by 7-Aminoactinomycin (7AAD) staining.

7AAD solution was purchased from eBioscience. In brief, the samples were washed thoroughly with PBS before the addition of 7AAD solution according to the ratio recommended by manufacturer and incubated for 15 oC. The

percentage of 7AAD positive cells was then analyzed using flow cytometry

(CyAn ADP analyzer, Beckman Coulter).

32 2.10 Trogocytosis assay

PKH-26 membrane dye (Sigma) was used to quantify the transfer of plasma membrane material between cells. In brief, donor cells (L-428 cells and L-

1236 cells) were stained with 2 mM of PKH-26 for 3 min and the reaction was stopped by the addition of FBS. The cells were then washed 5 times using complete medium prior to use. At the same time, recipient cells (kms-11 cells)

ο were labeled with 3 mM of CFSE (Invitrogen) for 15 min at 37 C and the reaction was quenched by the addition of ice cold complete medium. The cells were then washed 3 times prior to use. Donor cells and recipient cells were then co-cultured at the ratio of 1:1 for 24 h and stained for CD137 and

CD137L expression as described previously. Transfer of PKH-26-labelled membrane cells between donor and recipient cells was measured using flow cytometry as described in flow cytometry section. CFSE gating was used to differentiate between donor and recipient cells.

2.11 Reverse-transcriptase polymerase chain reaction (RT-PCR)

2.11.1 RNA extraction

RNA was extracted using RNeasy Mini Kit (Qiagen). In brief, cell samples were harvested, resuspended in RLT buffer, and passed through 27 G needle 5 times in order to homogenize the cells. 70% ethanol was used to precipitate

RNA and the cell lysates were transfered to RNeasy mini-columns thereby trapping RNA in the columns. The columns were rinsed with RW1 buffer and centrifuged at 13,200 rpm for 1 min. The columns were then rinsed with RPE buffer and centrifuged at 13,200rpm for 1 min. RNA was eluted from the

33 columns using RNase-free water followed by centrifugation at 13,200 rpm for

1 min.

2.11.2 Reverse transcription

Total cDNA was generated using RevertAid™ First Strand cDNA Synthesis

Kit (Fermentas). 2 µg of total RNA and 0.5 µg of oligo dT primer were mixed with RNA free water to a total volume of 12 µl. The mixture was mixed gently and incubated at 65 oC for 5 min and chilled on ice. The following reagents were then added in the order as shown in Table 1 to a total volume of 20 µl.

The mixture was then mixed gently and incubated at 42 oC for 60 minutes followed by 70 oC for 5 min. The cDNA was stored at -20 oC for later usage.

Table 2. RT-PCR reaction mix RNA template 12 µl 5x Reaction Buffer 4 µl RiboLock RNase inhibitor (20 U/ µl) 1 µl 10mM dNTP 2 µl RevertAid M-MuLV Reverse Transcritptase (200 U/ µl) 1 µl Total 20 µl

34 2.11.3 Polymerase chain reaction (PCR)

5 µl of cDNA template from each sample was used for a PCR reaction. cDNA template was mixed with freshly prepared PCR master mix (Table 2) and topped up with nuclease free water to 50 µl. The primers used for this PCR were purchase from 1st Base Custom Oligos (FBCO) and their sequences are shown in Table 3. PCR was performed using Eppendorf automated thermal cycler with the setting as specified in Table 4. Products of PCR were then electrophoresed on a 3% agarose gel. The gel was stained with GelGreen

(Biotium) and visualized using Chemi-Smart 3000 Gel Doc Imaging System.

Table 3. Standard PCR reaction mix cDNA template 5 µl Mastermix: 10x Taq buffer 5 µl Forward primer, 25 µM 1 µl Reverse primer, 25 µM 1 µl dNTP, 10 µM 1 µl Taq DNA polymerase, 5U/ µl 0.5 µl Nuclease free water Top up to 50 µl

Table 4. Primers set for CD137L and cyclophilin PCR

Primer Sequence Size (bp)

CD137L forward 5’AGCTTTCGCCCGACGATCCC 400

CD137L reverse 5’GCACTCAGGTGCAGCAAGCG

Cyclophilin forward 5’GTCCAGCATTTGCCATGGAC 200

Cyclophilin reverse 5’GACAAGGTCCCAAAGACAGC

35 Table 5. PCR thermal cycling for CD137L amplification Initial denaturation 95 oC 5 min Denaturation 95 oC 30 sec Annealing 63 oC 45 sec Extension 72 oC 60 sec Repeating cycle 40 cycles Final extension 72 oC 10 min

2.12 Light microscopic examination

Light microscope was used to examine any morphological changes that happened throughout the experiment. Microscopic images were captured using

Zeiss Axiovert 40 inverted microscope attached with Canon digital camera.

2.13 Antibody treatment

2.13.1 CD137 and CD137L neutralization

To test the effects of CD137 neutralization, KM-H2 cells were treated with 5

µg/ml of anti-CD137 neutralization antibody (clone BBK-2) and incubated for

24 h prior to downstream experiments. In some of the experiments, CD137L was also neutralized using 5 µg/ml of anti-CD137L antibody (clone 5F4,

41B436 or C65-485).

2.13.2 Agonistic CD137 stimulation

To determine which anti-CD137 antibody can induce higher CD137

stimulation, both clone BBK-2 and JG1.6a antibodies were coated on 24 well

plate and PBMC suboptimally activated with 0.5 ng/ml of anti-CD3 were

36 subsequently seeded (clone Okt3). The supernatants were collected and tested for their IFN-γ using ELISA as describe previously in ELISA section.

Coating of anti-CD137 antibodies was done by diluting the antibody to 5

µg/ml with PBS. 250 µl of the diluted antibody was added into a 24 well plate.

The plate was then refrigerated at 4 oC overnight. This is to ensure that the

coating is sufficient. Any excess antibody was washed out from the well with

PBS. The coated well was then ready for PBMC or HRS cells seeding.

2.14 Statistics

Data were expressed as mean ± standard deviation (SD) and statistical analysis

was performed by 2 tailed Student’s t-test.

37 CHAPTER 3 RESULTS

Infiltration of non-cytotoxic T cells is one of the hallmarks of HL. These T cells, which primarily consist of Th2 and Treg cells, are unable to eradicate tumor cells in vivo (as discussed in the Introduction). Interestingly, CD137, one of the T cell activation markers, was found to be expressed ectopically by

HRS cells. Under normal circumstance, CD137 is mainly expressed by NK cells and cytotoxic T cells in a strictly activation dependent manner. In HL,

HRS cells introduce a huge amount of additional CD137 into the tumor microenvironment that might interfere with the normal CD137-CD137L response, which is otherwise tightly regulated. In particular, CD137 overexpressed by HRS cells might interfere with T cell activity which will lead to HRS cell survival its microenvironment.

A series of experiments have been designed to address this question. In the first part of this chapter, how CD137 interferes with T cell activity via suppression of CD137L expression on HRS cells was shown. Next, whether ectopic CD137 can cause a similar CD137L suppression on surrounding cells, especially monocytes will be presented. Finally, the underlying mechanisms which cause this immunosuppressive event will be discussed in details.

3.1 Screening and generation of HRS cell lines

HRS cells constitute only about 1% of the total tumor mass in HL, which makes it technically challenging to obtain and isolate them. Thus, the experiments in this study involve immortalized HRS cell lines. There are total of 5 HRS cell lines which are commercially available and widely used for in

38 vitro studies, namely KM-H2, HDLM-2, L-428, L-540 and L-1236. All HRS

cell lines were screened for CD137 expression by flow cytometry. Only KM-

H2 and HDLM-2 were positive for CD137 (Figure 4). This is not consistent

with the observation where the majority of primary HRS cells were positive

for CD137. Since the majority of HRS cell lines were established from

secondary metastasized lesions of patients and have adapted to in vitro

condition where the immune escape mechanism is unnecessary, I suppose

some HRS cells might have lost their CD137 expression during this process.

To generate a negative control for CD137-expressing HRS cell lines, one of

our lab members has attempted to knock down CD137 in KM-H2 and HDLM-

2 cells by transfecting an shRNA plasmid into these HRS cell line using

Nucleofector technique (Lonza). CD137 expression on KM-H2 cells was

successfully knocked down, and the KM-H2-CD137 - cells have a similar

phenotype and characteristic as the empty vector transfected KM-H2 cells

(KM-H2-control). KM-H2-CD137 - cells do not express CD137 as shown by flow cytometry (Figure 5).

In addition, L-428, L-1236 and, L-540 cells were transfected with CD137 using lentiviral transfection. 100% of the transfected cells express CD137 after transfection and selection (Figure 6). The untransfected L-428, L-1236 and, L-

540 cells were used as negative controls.

HDLM-2 KM-H2 60% 80%

L-540 L-428 L-1236

0% 0% 0%

Isotype

CD137

Figure 4. Expression of CD137 on Reed Sternberg cell lines Expression of CD137 on Reed Sternberg cell lines was determined by flow cytometry. 10 5 cells from each cell line were stained for CD137. Open histograms: isotype. Grey histograms: anti-CD137 antibody (clone 4B4-1). The percentages of positively stained cells are shown in the histograms.

CD137

KM -H2 -Control 85%

- KM -H2 - CD137 8% Isotype CD137

Figure 5. CD137 silencing on KM-H2 cells CD137 on KM-H2 cells was knocked down by using shRNA, and the expression of CD137 was tested by using flow cytometry. 10 5 cells from each cell line were stained for CD137. Open histograms: isotype. Grey histograms: anti-CD137 antibody (clone 4B4-1). The percentages of positively stained cells are shown in the histograms.

L1236-CD137

95%

Isotype CD137

Figure 6. CD137 expression on L428, L540 and L1236 cells after CD137 transfection L428, L540 and L1236 cells were stably transfected with CD137 using a lentiviral vector. Expression of CD137 on these cells after the transfection was determined by flow cytometry 10 5 cells from each cell line was stained for CD137. Open histograms: isotype. Grey histograms: anti-CD137 (clone 4B4- 1). The percentages of positively stained cells are shown in the histograms.

41 3.2 Ectopically expressed CD137 reduces the T cell stimulatory capacity of HRS cells

3.2.1 Co-culturing HRS cells with PBMC and T cells

HRS cells are found in a microenvironment comprising infiltrating lymphocytes, plasma cells and histiocytes. Therefore, it is interesting to examine the effect of ectopic CD137 by co-culturing HRS cells with these immune cells. PBMC comprising lymphocytes and monocytes are an easy source of these immune cells. Prior to the co-culture, PBMC were suboptimally activated with 0.5 ng/ml of anti-CD3 antibody (clone: Okt3, in order to activate the T cells for co-stimulation studies), and co-cultured them with either isotype-treated or CD137 neutralizing antibody-treated KM-H2 cells for 24 h. The supernatants were collected and tested for IFN-γ release,

which is an indicator for T cell activation. An increase in IFN-γ release

whenever CD137 was neutralized was observed, as compared to isotype-

treated KM-H2 cells (Figure 7).

To confirm this observation, KM-H2-control and KM-H2-CD137 - cells were

treated with anti-CD137 neutralizing antibody, prior to co-culture with PBMC.

Comparing this result with the previous result, there is an increased IFN-

γ release when KM-H2-CD137 - cells or CD137 neutralized KM-H2-control cells were used (Figure 8). Interestingly, no significant increase in IFN-

γ release was observed when KM-H2-CD137 - cells were treated with CD137 neutralizing antibody, probably because most of their CD137 has been knocked down. The slight increase in the IFN-γ could be due to the neutralization of residual CD137 on KM-H2-CD137 - cells after the knocked

42 down. As IFN-γ is predominantly secreted by activated T cells, the HRS cells

were also co-cultured with purified T cells, and obtained a similar result when

PBMC were used (Figure 8).

The induction of IFN-γ release could be due to a synergistic effect of both allogeneic responses and co-stimulatory activity, as the PBMC were co- cultured with haplotype-mismatched HRS cells. This is consistent with the observation of much lesser IFN-γ release in the PBMC alone condition (~500 pg/ml, Figure 30).

Interestingly, IFN-γ release from the T cells alone condition was lower compared to the release from PBMC. This could be due to the lack of APC in

T cells alone condition, as the presence of APC (which expresses co- stimulatory molecules) might be important for the proper activation and co- stimulation of the T cells.

IFN-γ promotes Th1 responses, antigen presentation and cytolytic activity. It is also one of the most important markers for measuring T cells responses, especially cytotoxic T cells and T helper cells type I. These results suggest that ectopically expressed CD137 on HRS cells might interfere with the T cell activity.

3500 *

3000

2500

2000

(pg/ml) (pg/ml) γ γ γ γ 1500 IFN- 1000

500

0 Isotype CD137 Neutralizing Ab

Figure 7. CD137 neutralization on KM-H2 cells induces higher IFN-γγγ release from PBMC KM-H2 cells were pre-treated with 5 µg/ml of isotype (clone MOPC-21) or CD137 neutralizing antibody (Clone BBK-2) for 24 h, before co-culturing them with PBMC suboptimally activated with 0.5 ng/ml of anti-CD3 (Clone Okt3). IFN-γ release was measured by ELISA 24 h after co-culture. These data are representative of 3 independent experiments. Depicted are means ± standard deviations. * p<0.05

- - KM -H2 -Control KM -H2 -CD137 KM -H2 -Control KM -H2 -CD137 PBMC T cells

Figure 8. Impaired CD137 expression on KM-H2 cells leads to higher IFN-γγγ release from PBMC and T cells 5 x 10 5 cells/ml of KM-H2-control and KM-H2-CD137 - cells were pre-treated with 5 µg/ml of isotype (clone MOPC-21) or CD137 neutralizing antibody (clone BBK-2) for 24 h, before co-culturing them with 10 6 PBMC/ml or T cells suboptimally activated with 0.5 ng/ml of anti-CD3 (Clone Okt3). IFN-γ release was by ELISA measured 24 h after co-culture. These data are representative of 3 independent experiments. Depicted are means ± standard deviations. * p<0.05

45 3.2.2 Ectopically expressed CD137 down-regulates CD137L expression on

HRS cells

CD137 expression on T cells is upregulated after their primary stimulation.

CD137 can induce co-stimulation in T cells upon crosslinking with CD137L and augment T cell proliferation and cytotoxicity. As CD137L is constitutively expressed by APC, we hypothesized that it may be also expressed on HRS cells which originate predominantly from B cells.

Therefore, the CD137L expression on HRS cells and the effects of ectopic

CD137 on this expression were examined. Surprisingly, CD137L expression on HRS cells increased about 2-fold, when CD137 was neutralized by antibody (Figure 9) or was knocked down by siRNA (Figure 10). This implies that CD137 and CD137L expression may have an inverse relationship, where ectopic expression of CD137 can down-regulate CD137L expression on HRS cells via an unknown mechanism.

CD137 CD137L

73% 33% Isotype

96% 52% -CD137 -CD137 α α α α

Figure 9. CD137 and CD137L expression in KM-H2 cells after CD137 neutralization KM-H2 cells were treated with 0.5 µg/ml of isotype (clone MOPC-21) or CD137 neutralizing antibody (clone BBK-2). After 24 h, the cells were stained for CD137 (clone BBK-2) or CD137L (clone 5F4). The same clone of CD137 antibody was used for neutralization and staining to avoid competition between two antibodies. Staining antibodies were added in excess to ensure that the staining was consistent and comparable between conditions. Open histograms: isotype. Grey histograms: anti-CD137 (clone BBK-2). These data are representative of 3 independent experiments. The percentages of positively stained cells are shown in the histograms.

CD137L

27% Control

54% CD137 Silenced

Figure 10. CD137L is upregulated in KM-H2 cells after CD137 silencing CD137 in KM-H2 cells were knocked down by using siRNA, and the expression of CD137L was measured by using flow cytometry. 10 5 cells from each cell line were harvested and stained for CD137L. Open histograms: isotype. Grey histograms: anti-CD137L (clone 5F4). The percentages of positively stained cells are shown in the histograms.

47 3.2.3 Induction of IFN-γγγ release is due to CD137L upregulation after

CD137 silencing

CD137L is primarily expressed by APC, and upon ligation with CD137 on T

cells, it acts as a co-stimulator of T cell activation. Therefore, it is

hypothesized that increase in IFN-γ release is indeed due to increase in

CD137L expression on KM-H2 cells after CD137 knock down or

neutralization. To test this hypothesis, CD137L expression on KM-H2-CD137 - cells was neutralized with various clones of CD137L neutralizing antibodies, before co-culturing them with PBMCs or T cells. As expected, KM-H2-

CD137 - cells induce higher IFN-γ release from both PBMC (Figure 11) and T cells (Figure 12) after co-culturing. Two out of three clones of CD137L neutralizing antibodies used in this experiment managed to abrogate the induction of IFN-γ release caused by the CD137 knock down. Although there

was low CD137L expression on KM-H2-control cells, it might not be

sufficient to induce a T cell response, and therefore neither of these CD137L

neutralizing antibodies significantly reduced IFN-γ release when the KM-H2- control cells were co-cultured with PBMC or T cells. Since the induction of

IFN-γ release after CD137 neutralization and knock down could be abrogated

using anti-CD137L neutralizing antibodies, it may be due to CD137L

upregulation on KM-H2 cells.

KM-H2-Control KM-H2-CD137 -

Figure 11. CD137L neutralizing antibodies abrogate the induction of IFN- γγγ release from PBMC during co-culture with KM-H2-CD137 - cells. 5 x 10 5 cells/ml of KM-H2-control and KM-H2-CD137 - cells were co-cultured with 10 6 cells/ml of suboptimally activated PBMC and their CD137L was neutralized with one of three clones of CD137L neutralizing antibodies, or their isotype control (5 µg/ml). Antibody 1 - clone 5F4, antibody 2 - clone 41B436, antibody 3 - clone C65-485. Supernatants were harvested 24 h later and tested for IFN-γ concentration by ELISA. Depicted are means ± standard deviations. * p<0.05

KM-H2-Control KM-H2-CD137 -

Figure 12. CD137L neutralizing antibodies abrogate the induction of IFN- γγγ release from T cells during co-culture with KM-H2-CD137 - cells. 5 x 10 5 cells/ml of KM-H2-control and KM-H2-CD137 - cells were co-cultured with 10 6 cells/ml of suboptimally activated T cells and their CD137L was neutralized with one of three clones of CD137L neutralizing antibodies, or their isotype control (5 µg/ml). Antibody 1 - clone 5F4, antibody 2 - clone 41B436, antibody 3 - clone C65-485. Supernatants were harvested 24 h later and tested for IFN-γ concentration by ELISA. Depicted are means ± standard deviations. * p<0.05

50 3.2.4 Summary

In this section, it has been shown that ectopic CD137 expression on HRS cells can reduce T cell co-stimulation by downregulating CD137L expression on

HRS cells. However, in the microenvironment of Hodgkin Lymphoma, there might be cell types other than HRS cells which express CD137L. For instance, both monocytes and B cells which constitutively express CD137L under physiological condition are also found in Hodgkin Lymphoma infiltrating cells.

Will ectopic CD137 cause a similar downregulation of CD137L on these cell types as well? This possibility will be explored in the next section. In addition, the mechanism which leads to the downregulation of CD137L upon ectopic

CD137 expression was demonstrated.

51 3.3 Mechanism of CD137L disappearance

We have confirmed that ectopic CD137 leads to a downregulation of CD137L.

This subsequently causes a reduction in the T cell response and potential immune evasion of HRS cells. But thus far, the results do not explain what leads to the disappearance of CD137L.

In this section, the underlying mechanism which leads to CD137L downregulation in CD137 expressing HRS cells was investigated. A few possible mechanisms of CD137 downregulation are as follows:

a) Ectopic CD137 may alter transcription or translation of CD137L, and

reduce the de novo synthesis of CD137L.

b) Ectopic CD137 may affect the turnover rate of CD137L, by increasing

internalization and breakdown of CD137L.

c) Ectopic CD137 blocks certain epitopes on CD137L, which neutralizes

its function and prevents the antibody to bind.

We decided to use KM-H2 cells as the model to study the mechanism of

CD137L downregulation, mainly due to KM-H2 cells expressing CD137

endogenously, so that it will be the best representation of in vivo condition.

52 3.3.1 CD137 neutralization does not affect CD137L mRNA expression

Firstly, whether ectopic CD137 expression affects transcription of CD137L was investigated. KM-H2 cells were treated with anti-CD137 neutralizing antibody and mRNA expression was quantified by using RT-PCR. As the increase in CD137L was observed after 24 h of CD137 neutralization by using flow cytometry, the same time point was used in this experiment. Interestingly, no changes in mRNA expression can be observed (Figure 13). This suggests that ectopic CD137 expression might not affect transcriptional activity of

CD137L.

1.4 ns A B 1.2

Anti- 1 Isotype CD137 0.8 CD137L 0.6

0.4 Cyclophilin CD137L:Cyclophilin 0.2

0 Anti- Isotype CD137

Figure 13. CD137 neutralization does not affect mRNA expression level of CD137L in KM-H2 cells KM-H2 cells were treated with 5 µg/ml of CD137 neutralizing antibody (clone BBK-2) for 24 h, and total mRNA of the cells was isolated as described in Materials and Methods. (A) mRNA expression of CD137L was determined by RT-PCR. Cyclophilin was used as internal control. (B) mRNA expression of CD137L was quantified (Image J) and normalized with cyclophilin expression. Depicted are means ± standard deviations of two individual experiments. ns: Not significant.

53 3.3.2 CD137L protein expression increases after CD137 neutralization

The interesting finding that CD137 neutralization does not affect CD137L transcription level led us to think whether CD137L expression on HRS cells was upregulated after CD137 neutralization. Is it possible for ectopic CD137 to bind on CD137L and hence masking/hindering its function and also its detection by flow cytometry? To address this question, KM-H2 cells were treated with CD137 neutralizing antibody as described in Materials and

Methods section, and the cells were lysed 24 h later. Denaturing RIPA buffer was used in cell lysis to ensure that protein-protein interactions are broken, so any bonding or epitope occupation would not interfere with detection of

CD137L by Western Blotting. Surprisingly, CD137 neutralization does upregulate total CD137L protein level in KM-H2 (Figure 14), which means there is in fact a true increase in CD137L protein levels.

Time course experiments were also performed, with samples being collected and lysed 24 h, 48 h, and 72 h after CD137 neutralization. The time course experiments have confirmed the result in Figure 14 and showed that CD137 neutralization upregulate CD137L level on KM-H2 cells at different time points (Figure 15). This result also shows that the effect of CD137 neutralization on upregulating CD137L is long lasting (up to 72 h after the treatment).

By observing upregulation of CD137L protein level and unchanged CD137L mRNA level, we are able to deduce that CD137 expression on HRS cells does

54 not alter transcriptional activity of CD137L but it plays a role in increasing turn over of CD137L protein in HRS cells.

KMH2

Isotype α-CD137

CD137L

GAPDH

Figure 14. Total CD137L protein in KM-H2 cells was upregulated after CD137 neutralization KM-H2 cells were treated with 0.5 µg/ml of CD137 neutralizing antibody (clone BBK-2), or its isotype control. After 24 h, the cells were lysed under denaturing conditions using RIPA buffer. 20 µg of protein from each condition were used in Western Blot analysis for CD137L detection. GAPDH was used as loading control.

24 h 72 h 96 h

Isotype α-CD137 Isotype α-CD137 Isotype α-CD137

Figure 15. Upregulation of CD137L protein in KM-H2 cells after CD137 neutralization KM-H2 cells were treated with 0.5 µg/ml of CD137 neutralizing antibody (clone BBK-2), or its isotype. The cells were harvested at different time points (24, 72, and 96 h) and were lysed under denaturing condition using RIPA buffer. 20ug of protein samples from each condition were used in Western Blot for CD137L detection. The samples were stored at -80 oC prior to usage. GAPDH was used as loading control.

55 3.3.3 CD137-CD137L co-localization in HRS cells

CD137L is the only known binding partner of CD137 (Herbert Schwarz, personal communication). So it would not be a surprise if ectopic CD137 on

HRS cells could interact with CD137L which is expressed by HRS cells as well. However, how strong is this CD137-CD137L interaction in HRS cells?

To answer this question, a co-immunoprecipitation on KM-H2 cell lysate was performed with anti-CD137 antibody, and a Western blot was performed to measure the amount of CD137L pulled down with CD137. As expected,

CD137L co-localized strongly with CD137 (Figure 16). This implies that

CD137 and CD137L are not only co-expressed on KM-H2 cells, but at the same time, there is a strong interaction between the two molecules. In previous experiments, it is shown that CD137 neutralization actually leads to an increase of CD137L expression on KM-H2 cells. Is it possible the strong

CD137-CD137L interaction plays a role in downregulation of CD137L on

HRS cells?

KM-H2 cells were treated with either isotype or CD137 neutralizing antibody for 24 h. The cells were then lysed and the extracts are subjected to co- immunoprecipitation as described in Materials and Methods section.

Interestingly, after neutralization of CD137, the KM-H2 cells showed less

CD137-CD137L interaction compared to the isotype treated control cells

(Figure 17). The CD137 neutralization was shown to upregulate CD137 and

CD137L expression on KM-H2 cells (Figure 9 and 10), therefore, the reduction of the CD137-CD137L complexes after CD137 neutralization is unlikely to be caused by reduced CD137L expression. Instead, it is more likely

56 that CD137 neutralization affects the potential CD137-CD137L interaction.

Some co-immunoprecipitation of CD137 and CD137L was observed in

CD137-neutralized mock-IP condition, which might be due to the binding of neutralizing antibody-CD137 complex to protein G.

Figure 16. CD137L co-immunoprecipitates with CD137 in KM-H2 cells. KM-H2 cells were lysed with 1% Triton-X and the lysate was immunoprecipitated with either isotype control antibody (Mock-IP) or anti- CD137 antibody (clone BBK-2, Co-IP). Precipitates were subjected to CD137L Western blot analysis. Equal amounts of protein lysate (1 µg) were used in each condition, and the light chain of the antibody was used as loading control.

Figure 17. CD137 neutralization reduces CD137 and CD137L interaction in KM-H2 cells. KM-H2 cells were treated with 5 µg/ml of anti-CD137 neutralizing antibody (clone BBK-2) or its isotype control. After 24 h, the cells were lysed with 1% Triton-X and the lysate was immunoprecipitated with either isotype control antibody (Mock-IP) or anti-CD137 antibody (clone BBK-2, Co-IP). Precipitates were subjected to CD137L Western blot analysis. Heavy chain of antibody was used as loading control. The slight increase in the CD137L pulled down in Mock-IP condition of CD137-neutralized KM-H2 cells is likely due to the binding of neutralizing antibody to the CD137-CD137L complex before the Co-IP was performed.

58 There is a strong CD137 and CD137L interaction in HRS cells, which can be abrogated by using CD137 neutralizing antibody. This led us to the next question, where are CD137 and CD137L colocalized in HRS cells? Confocal microscopy was used to answer this question. As expected, a strong colocalization of CD137 and CD137L in KM-H2 cells was observed. Most of this colocalization was found in the cytoplasm of KM-H2 cells, while some of it was scattered on the cell membrane (Figure 18). The colocalization of

CD137 and CD137L in KM-H2 cells looks similar to the clustering of CD40 upon stimulation, where CD40 forms a cap-like morphology and gets eventually internalized (Chen et al., 2006). It is possible that CD137L and

CD137 have a similar mechanism and get internalized after their ligation.

59 A DAPI Isotype

Merge

Isotype Phase contrast

50 µµµm

B DAPI ααα-CD137L

Merge

ααα-CD137 Phase contrast

Figure 18. Colocalization of CD137 and CD137L in KMH2 cells. KM-H2 cells were stained for CD137 (clone BBK-2) and CD137L (clone 5F4). Localization of the proteins was visualized by confocal microscopy. A. isotype staining. B. CD137 and CD137L staining. DAPI was used as a counterstain.

60 3.3.4 CD137 and CD137L are co-internalized via endocytosis

It is reported that members of TNF and TNFR superfamily can be co- internalized upon ligation. TNF-Related Apoptosis-Inducing Ligand (TRAIL) was shown to be endocytosed quickly with Death Receptor 4 (DR4) and DR5 upon ligation and hence reduces the effect of TRAIL induced cytotoxicity in tumor cells (Zhang and Zhang, 2008; Zhang et al., 2009). Moreover, endocytosis of CD95 and TNF-R1 were also observed (Eramo et al., 2004;

Kohlhaas et al., 2007). The outcome and downstream mechanisms involved were found to be complicated and dependent on the context of their stimulation.

Accumulation of CD137 in cytoplasm of HRS cells was found in clinical samples, although the function of CD137 in the cytoplasm remained unknown

(Anderson et al., 2012). Consistent with this observation, a strong CD137 and

CD137L colocalization was observed in the cytoplasm of KM-H2 cells

(Figure 18). The CD137-CD137L complex which was observed in the cytoplasm might be the result of CD137 and CD137L internalization from the plasma membrane after their ligation. By eliminating CD137L from the cell surface, HRS cells gain survival advantages by limiting T cell costimulation as described in earlier part of my result section.

To confirm this hypothesis, endocytosis in KM-H2 cells was inhibited using

MDC, an inhibitor of clathrin-mediated endocytosis. Clathrin-mediated endocytosis has been involved in internalization and recycling of many cell

61 surface receptors, including other member of TNFR superfamily, e.g. CD95,

DR4 and DR5 (Austin et al., 2006; Kohlhaas et al., 2007; Lee et al., 2006).

Since MDC is toxic at high concentration and its effect varies with different cell types, the lethal dose of MDC was first determined for KM-H2 cells. KM-

H2 cells were treated with different concentrations of MDC (12.5-200 µM)

and their cell viability were determined by using 7-AAD staining. At doses

higher than 100 µM, a dramatic reduction in cell viability can be observed

after 24 h (Figure 19). Hence the maximum MDC dose for KM-H2 should not

be more than 50 µM.

KM-H2 cells were treated with 50 µM of MDC for 24 h to inhibit their

endocytosis. As shown in Figure 20, both CD137 and CD137L on KM-H2

cells were upregulated after the 24 h of endocytosis inhibition. This suggests

that under normal circumstance, CD137 and CD137L were internalized

together via endocytosis, and MDC inhibits endocytosis and thereby restores

the CD137 and CD137L expression partially. Interestingly, whenever

endocytosis was inhibited simultaneously with CD137 neutralization, further

increases in CD137 and CD137L expression were not observed. This implies

CD137 neutralization and endocytosis inhibition might be redundant of each

other. From there, it can be deduced that anti-CD137 antibody can bind to

CD137, inhibit its interaction with CD137L, and hence reducing co-

internalization of the CD137-CD137L complex. Nonetheless, MDC treated

KM-H2 cells have lesser upregulation of CD137 and CD137L as compared to

62 CD137 neutralized KM-H2 cells. That is likely due to the toxic nature of

MDC which limits us to use more effective doses or longer treatment times.

MDC ( µµµg/ml) 0 12.5 25 50 100 3% 3% 2% 3% 7% 7AAD

Side Scatter

Figure 19. Toxicity titration of MDC on KM-H2 cells KM-H2 cells were treated with various concentrations of MDC as shown. After 24 h of treatment, the cells were harvested and stained with 7-AAD to determine their viability. 7-AAD + populations were gated and the percentages of positively stained cells are shown. These data are representative of 3 independent experiments.

A DMSO MDC

35% 43% Isotype

80% 81% -CD137 -CD137 α α α α

CD137L B DMSO MDC

20% 36% Isotype

78% 80% -CD137 -CD137 α α α α CD137

Figure 20. Inhibition of endocytosis increases CD137 and CD137L expression on KM-H2 cells KM-H2 cells were treated with 5 µg/ml of anti-CD137 neutralizing antibody, or its isotype control as described in Methods section. The cells were also treated with 50 µg/ml MDC or DMSO, respectively. The cells were harvested and stained for CD137 (A) or CD137L (B) 24 h later. Isotype staining is shown as open histograms and CD137 and CD137L staining are shown as grey histograms. The percentages indicate the positively stained populations. These data are representative of 3 independent experiments.

64 3.3.5 Trogocytosis mediates CD137 transfer which causes CD137L downregulation

Ectopic CD137 expression on HRS cells causes a reduction in T cell activation by increasing internalization of CD137L through endocytosis. Nonetheless, whether CD137 on HRS cells interacts with CD137L via cis (intracellular) or trans (intercellular) interaction remains unknown. It has been reported earlier that TNFR family members (e.g. OX40) can get transferred from a cell to the other through trogocytosis, which is a contact-dependent intercellular transfer of cell surface material accompanied by a transfer of plasma membrane fragments (Baba et al., 2001). This mechanism has been shown in various models to assist cells to cells acquisition of functional proteins or as a negative feedback mechanism to scavenge proteins which have been utilized. The extent of transfer through trogocytosis is often associated with a specific receptor-ligand interaction between cells involved in the trogocytosis (e.g.

TCR - MHC class I or TCR - MHC class II induced trogocytosis, as discussed in Introduction).

A strong interaction between CD137 and CD137L in HRS cells has been shown using co-immunoprecipitation (Figure 16). Hence, it is interesting to investigate whether this strong interaction can lead to the transfer of CD137 between HRS cells, and more importantly, cause a reduction of CD137L.

Donor cells, L-428 and L-1236 cells, were labelled with PKH-26, which is a lipophilic dye commonly used to monitor the transfer of plasma membrane fragments. The donor cells were then co-cultured with KMS-11 cells

65 overexpressing CD137L, which were used as recipient cells. The transfer of

CD137 from donor to recipient cells, and also the extent of trogocytosis were measured 24 h later. This co-culture experimental set up is needed primarily because it will be difficult to measure transfer of CD137 and trogocytosis by using one single cell line (e.g. using KM-H2 cells alone), as any transfer and trogocytosis will be bi-directional between cells, and any effect of transfer will be cancelled off each other.

After the co-culture, KMS-11 cells were found to be positively stained for

PKH-26 in all co-culturing conditions, regardless of CD137 or CD137L expression on L-428, L-1236 and KMS-11 cells respectively (Figure 21). This

PKH-26 transfer is most likely due to background trogocytosis between these cells, induced by some other strong protein-protein interactions ongoing between these cells. The residual and unbound PKH-26 dye on the donor cells which could contaminate the co-culture system might also lead to the increase of the PKH-26 background on KMS-11 cells. Interestingly, despite the high background of trogocytosis, a further increase in PKH-26 transfer was observed from both L-428-CD137 and L-1236-CD137 to KMS-11-CD137L cells, compared to KMS-11-control cells. This result implies that the level of trogocytosis is upregulated whenever CD137 and CD137L is present in between donor and recipient cells, and CD137 and CD137L interaction can increase cell-cell affinity, bringing two cells closer to each other, and allowing more trogocytosis to happen. This hypothesis is further strengthened when an increase in cell clumping was observed when L-428-CD137 or L-1236-CD137

66 cells were co-cultured with KMS-11-CD137L cells, as compared to other conditions which had little or no cell clumping (Figure 22).

A KMS -11 -control KMS -11 -CD137L 8%, MFI = 4 8%, MFI = 5 99% 44 98% 36 100% 60 99% 166

PKH-26 No co -culture L-428 -control L-428 -CD137

B KMS -11 -control KMS -11 -CD137L 8%, MFI = 4 8% MFI = 5 99% 86 99% 81 99% 87 99% 190

PKH-26 No co -culture L-1236 -control L-1236 -CD137

Figure 21. CD137 and CD137L induce trogocytosis between donor and recipient cells 5 x 10 5 cells/ml of PKH-26 labeled L-428 (A) or L-1236 (B) cells were co- cultured with 5 x 10 5 cells/ml of CFSE labelled KMS-11 cells. 24 h later, the cells were harvested and analyzed with flow cytometry. The population shown was gated on CFSE-positive (KMS-11) cells. The percentages indicate the number of positively stained cells as compared to background signal. MFI (mean fluorescent index) for the respective sample were also shown. These data are representative of 3 independent experiments.

A L-428-control L-428-CD137

KMS-11-control

B KMS-11-CD137L L-1236-control L-1236-CD137

KMS-11-control

KMS-11-CD137L

Figure 22. The presence of CD137 and CD137L induce L-428 and L-1236 cell aggregation with KMS-11 cells 5 x 10 5 cells/ml of L-428 (A) or L-1236 (B) cells were co-cultured with 5 x 10 5 cells/ml of KMS-11 cells. 24 h later, the images of cell clumping were captured under 100x magnification with an inverted microscope. Grayscale images are presented. These data are representative of 2 independent experiments.

To confirm whether CD137 transfers with trogocytosis, a similar experiment was performed and the cells were stained for CD137 and CD137L. As expected, an increase in CD137 transfer from L-428-CD137 (Figure 23A) and

L-1236-CD137 cells (Figure 24A) to KMS-11-CD137L cells was observed.

The transfer of CD137 is CD137L dependent since there can be no CD137 transfer observed when KMS-11-control cells were used. Moreover, this also confirmed that CD137 transfer from donor to recipient cells is instead through trogocytosis. What is more exciting is that CD137L expression on KMS-11-

CD137L cells was downregulated following the CD137 transfer (Figure 23B and 24B). The downregulation of CD137L is also CD137 dependent, as no

CD137L downregulation can be observed when KMS-11-CD137L cells were co-cultured with CD137 negative L-428 or L-1236 cells. The result also implies that the trans (intercellular) interaction between CD137 and CD137L

plays a role in the CD137L downregulation induced by ectopically expressed

CD137 on HRS cells. Nonetheless, this result does not rule out the possibility

that a cis interaction is also involved in the CD137-CD137L ligation and

CD137L downregulation, which will need to be further investigated.

Surprisingly, a strong transfer of CD137L was also observed from KMS-11-

CD137L cells to L-428-CD137 cells (Figure 25) and L-1236-CD137 (Figure

26) cells. The CD137L transfer is CD137 dependent, as no transfer can be observed from KMS-11-CD137L cells to L-428-control and L1236-control cells. This implies the transfer of CD137 and CD137L via trogocytosis might

69 be a bi-directional process, where both cells involved in the trogocytosis can be donor and recipient cells at the same time.

A KMS -11 -control KMS -11 -CD137L

7% 7% 8% 13% 8% 61%

CD137 B 37% 89% 37% 89% 13% 68%

CD137L No co -culture L-428 -control L-428 -CD137

Figure 23. Downregulation of CD137L on KMS-11 cells following CD137L dependent transfer of CD137 from L-428 cells to KMS-11 cells 5 x 10 5 cells/ml of PKH-26 labeled L-428 cells were co-cultured with 5 x 10 5 cells/ml of CFSE labeled KMS-11 cells. 24 h later, the cells were harvested and stained for (A) CD137 (clone BBK-2) or (B) CD137L (clone 41B436). The cells were then analyzed with flow cytometry. The population shown was gated on CFSE-positive (KMS-11) cells. Indicated are the percentages of positively stained cells as compared to isotype staining. This experiment was performed 3 times with similar results.

A KMS -11 -control KMS -11 -CD137L

7% 7% 10% 12% 55% 80%

CD137

B 34% 89% 40% 92% 12% 68%

CD137L No co -culture L-1236 -control L-1236 -CD137

Figure 24. Downregulation of CD137L on KMS-11 cells following CD137L dependent transfer of CD137 from L-1236 cells to KMS-11 cells 5 x 10 5 cells/ml of PKH-26 labeled L-1236 cells were co-cultured with 5 x 10 5 cells/ml of CFSE labeled KMS-11 cells. 24 h later, the cells were harvested and stained for (A) CD137 (clone BBK-2) or (B) CD137L (clone 41B436). The cells were then analyzed with flow cytometry. The population shown was gated on CFSE-positive (KMS-11) cells. Indicated are the percentages of positively stained cells as compared to isotype staining. This experiment was performed 3 times with similar results.

A L-428 -control L-428 -CD137

5% 99% 10% 99% 14% 99%

CD137

B 9% 3% 8% 6% 12% 91%

CD137L No co -culture Kms -11 -control Kms -11 -CD137L

Figure 25. CD137L transfer from KMS-11-CD137L cells to L-428-CD137 cells 5 x 10 5 cells/ml of PKH-26 labeled L-428 cells were co-cultured with 5 x 10 5 cells/ml of CFSE labeled KMS-11 cells. 24 h later, the cells were harvested and stained for (A) CD137 (clone BBK-2) or (B) CD137L (clone 41B436). The cells were then analyzed with flow cytometry. The population shown was gated on CFSE- negative (L-428) cells. Indicated are the percentages of positively stained cells as compared to isotype staining. This experiment was performed 3 times with similar results.

A L-1236 -control L-1236 -CD137

7% 98% 7% 99% 6% 99%

CD137

B 9% 4% 11% 5% 12% 65%

CD137L No co -culture Kms -11 -control Kms -11- CD137L

Figure 26. CD137L transfer from KMS-11-CD137L cells to L-1236- CD137 cells 5 x 10 5 cells/ml of PKH-26 labeled L-1236 cells were co-cultured with 5 x 10 5 cells/ml of CFSE labeled KMS-11 cells. 24 h later, the cells were harvested and stained for (A) CD137 (clone BBK-2) or (B) CD137L (clone 41B436). The cells were then analyzed with flow cytometry. The population shown was gated on CFSE- negative (L-1236) cells. Indicated are the percentages of positively stained cells as compared to isotype staining. This experiment was performed 3 times with similar results.

73 3.3.6 Summary

In this section, the underlying mechanism of CD137L disappearance on HRS cells induced by ectopic CD137 expression was defined. To summarize:

1. CD137 can downregulate total CD137L protein levels in HRS cells,

but does not influence CD137L transcription.

2. CD137 forms a strong interaction with CD137L, and co-localized in

the cytoplasm of HRS cells.

3. CD137 is internalized with CD137L by HRS cells, and the

internalization can be inhibited by inhibitors of the clathrin-dependent

internalization pathway.

4. CD137 neutralization can abrogate interaction and internalization of

CD137 and CD137L.

5. CD137 can get transferred from one cell to another through

trogocytosis. This process is CD137L dependent, and can also induce

downregulation of CD137L on the recipient cells.

With these results, it is now assured that CD137 ectopically expressed by HRS

cells can bind to CD137L on the surface of other cells via trans interaction,

and later become co-internalized into the cytoplasm. By downregulating

CD137L, HRS cells reduce T cell costimulation, and hence gain survival

advantages in their microenvironment.

74 3.4 Transfer of CD137 to surrounding monocytes and B cells

HRS cells are surrounded by infiltrating inflammatory cells which were shown to be crucial for HRS cells survival. These infiltrating cells provide cytokines and growth factors which are constantly needed by HRS cells. Interestingly,

APC and T cells which are part of these infiltrating cells fail to eliminate HRS cells effectively. It was shown that Treg dominating the HL microenvironment can undermine an effective cytotoxic T cells response (Re et al., 2005).

B cells and monocytes are APC which were shown to express CD137L constitutively (Shao and Schwarz, 2011). As described in the previous section,

CD137 can be transferred through trans interaction and causes the downregulation of CD137L in HRS cells. Hence, it will be interesting to find out if ectopic CD137 expression on HRS cells can also cause downregulation of CD137L in their surrounding APC.

In this section, transfer of ectopic CD137 from HRS cells to monocytes and B cells and whether this CD137 transfer causes the downregulation of CD137L was assessed. It was also investigated whether the downregulation of CD137L on APC affect T cell costimulation.

75 3.4.1 Ectopically expressed CD137 induces CD137L downregulation in monocytes

First of all, the transfer of CD137 from HRS cells to the surrounding monocytes or B cells which express CD137L was studied. KM-H2 and

HDLM-2 cells were co-cultured with PBMC which were sub-optimally activated with 1 ng/ml of anti-CD3 (clone Okt3). PBMC contain 10% of monocytes and B cells, respectively, hence they are an easy source and suitable for this co-culture experiment. After 24 h of co-culture, the cells were harvested and stained for CD11b or CD20 for the gating of monocytes or B cells. The cells were also stained for their CD137 and CD137L expression to observe the transfer of CD137 and its effect on CD137L.

Monocytes and B cells are negative for CD137 under physiological condition

(although there is a study reporting that B cells can upregulate their CD137 expression upon BCR stimulation (Zhang et al., 2010)). Therefore, any expression of CD137 on monocytes and B cells would most likely be due to the transfer from HRS cells.

It is possible that doublet cell formation between monocytes or B cells and

HRS cells might lead to an artifact of CD137 + monocytes or B cells. However,

with proper suspension of the cell mixture during the staining step, the

incident of the doublet cells should be minimal, and it is unlikely to

substantially affect the readout of CD137 transfer which was measured in this

experiment.

76 As shown in Figure 27 and Figure 28, monocytes and B cells have low or no

CD137 expression when they were not co-cultured with HRS cells.

Surprisingly, CD137 was transferred from HDLM-2 cells to monocytes and also from both HDLM-2 cells and KM-H2 cells to B cells. KM-H2 cells do not transfer or transfer only little CD137 to monocytes and B cells in contrast to HDLM-2 cells (Figure 27 and Figure 28). The difference in the CD137 transfer might be due to higher CD137 expression on HDLM-2 cells as compared to KM-H2 cells (Figure 4). This is supported by the observation that transfer of CD137 is both CD137 and CD137L dependent (Figure 23-24).

Hence, lesser CD137 is expected to get transferred from KM-H2 cells which have lower CD137 expression. Notably, one of my colleagues observed a much higher transfer of CD137 from CD137 overexpressing L1236 cells to both monocytes and B cells (Ho et al., 2012), which again confirms that the level of CD137 expression is positively associated with the CD137 transfer.

In addition, the fluorescent background of monocytes was also increased when the monocytes were co-cultured with HRS cells. The exact cause of this enhanced fluorescent background on monocytes is unknown, but it might be associated with the morphology and granularity changes on monocytes when they encountered HRS cells.

PBMC alone

HDLM-2 KM-H2

28% 3%

CD137

Figure 27. Transfer of CD137 from HRS cell lines to monocytes 5 x 10 5 cells/ml of KM-H2 or HDLM-2 cells were co-cultured with 10 6 of PBMC suboptimally activated with 1 ng/ml of anti-CD3 (Clone Okt3). 24 h later, the cells were harvested and stained for CD137 (clone BBK-2) and CD11b and analyzed using flow cytometry. The histograms shown are gated on CD11b + population only. Open histograms show the isotype staining of CD137. The percentages of positively stained cells are shown in the histograms. These data are representative of 2 independent experiments.

PBMC alone

HDLM-2 KM-H2

28% 20%

CD137

Figure 28. Transfer of CD137 from HRS cell lines to B cells 5 x 10 5 cells/ml of KM-H2 or HDLM-2 cells were co-cultured with 10 6 of PBMC suboptimally activated with 1 ng/ml of anti-CD3 (Clone Okt3). 24 h later, the cells were harvested and stained for CD137 (clone BBK-2) and CD20 and analyzed using flow cytometry. The histograms shown are gated on CD20 + population only. Open histograms show the isotype staining of CD137. The percentages of positively stained cells are shown in the histograms. These data are representative of 2 independent experiments.

79 Surprisingly, transfer of CD137 to monocytes or B cells does not induce a consistent downregulation of CD137L (as summarized in Table 5 and shown in Appendix IV). This observation is not tally with the data shown in Figure

23 and Figure 24, where CD137 transfer always caused CD137L downregulation on KMS-11 cells. The reason might be that CD137L is hardly detected on primary cells when analysed using flow cytometry, hence any change in CD137L expression might be harder to be measured by using flow cytometry.

To overcome this issue, Western blot instead of flow cytometry was used to observe CD137L downregulation after monocytes and B cells are co-cultured with HRS cell lines. The experiment was repeated by co-culturing L-428 cells with MACS-purified monocytes or B cells. After 24 h of co-culture, the monocytes and B cells were re-purified using MACS, and Western blot was done to determine CD137L protein levels in monocytes and B cells.

In most of the experiments, the number of re-purified B cells was too low to run a Western blot, and only little or no CD137L could be detected on B cells using Western blots. The reason for this observation is unknown, but it might be due to the influence of HRS cells on the viability of B cells (Figure 29).

Moreover, CD137L expression on L-428 cells was detected in this experiment, despite it was normally found to be negative or low expression with cell surface flow cytometry staining. Surprisingly, CD137L expressed by L-428 cells has a higher molecular weight than the CD137L expressed by primary monocytes (25 kD). This finding suggests that L-428 cells might express a

80 different form of the CD137L which can not be translocated to cell surface and hence only be found in the intracellular compartment. However, the functions of this alternative form of CD137L is unknown, and this hypothesis needs to be further investigated.

Are HRS cells expressing a different form of CD137L for their survival advantage? This question is not addressed here. But what is important is the difference in their molecular weight allows us to distinguish the CD137L originated from L-428 cells and the one originated from primary APC, since there is always a slight contamination of L-428 cells after the APC are re- purified from the co-culture.

From Figure 29, CD137L expressed by monocytes can be easily identified on the blot as bands which have the exact molecular weight of 25kD, while the bigger bands are from CD137L expressed by L-428 cells. Notably, CD137L on monocytes is downregulated whenever the cells are co-cultured with L-

428-CD137, but not L-428-control. This observation confirmed the hypothesis that CD137 transferred to monocytes causes the downregulation of CD137L on monocytes. Unfortunately, due to the low CD137L expression on B cells, this mechanism can not be confirmed for B cells by Western blot.

81 Table 6. Changes in CD137L expression on monocytes and B cells after co-culturing with HRS cell lines

Co-culture with Monocytes B cells

Experiment 1 HDLM-2 Decrease No change

KM-H2 Decrease No change

Experiment 2 HDLM-2 Decrease No change

KM-H2 Decrease Increase

Experiment 3 HDLM-2 No change No change

KM-H2 Increase Increase

*Individual flow cytometry data is shown in Appendix IV

Figure 29. HRS cells induce CD137L downregulation on monocytes 5 x 10 5 cells/ml of L-428 cells were co-cultured with 10 6 cells/ml of B cells or monocytes. 24 h after the co-culture, B cells and monocytes were re-purified from the co-culture by using magnetic column and then lysed using RIPA buffer. 40 µg of protein samples from each condition were used in Western blot for CD137L detection. GAPDH was used as loading control. These data are representative of 3 independent experiments.

82 3.4.2 CD137 overexpressing cell lines abrogate IFN-γγγ release from PBMC

HRS cells can transfer ectopic CD137 to both monocytes and B cells, and cause the downregulation of CD137L at least on monocytes, but what is the function of this mechanism? Will the CD137L downregulation on APC affect

T cell activation? To confirm this, L-428 or L-1236 cells were co-cultured with PBMC for 24 h, and IFN-γ release was measured as an indicator of T cell activation.

A significant decrease in IFN-γ release was observed when L-428-CD137 and

L-1236-CD137 cells were co-cultured with PBMC, as compared to L-428- control and L-1236-control cells (Figure 30). It was also shown in a parallel experiment that the T cells proliferation was reduced when they were co- cultured with L-1236-CD137 cells (Ho et al., 2013). This reduction in IFN-γ release and proliferation is likely due to the downregulation of CD137L on monocytes, which subsequently causes the reduction in T cell activation. This observation is similar to the KM-H2 cells and PBMC co-culture experiment in

Figure 7, but unlike KM-H2 cells, L-428 and L-1236 cells were stained negative for cell surface CD137L (Figure 25 and 26), hence any effect of

CD137L downregulation will be due to the transfer of CD137 from HRS cells to their surrounding cells in the coculture. Nonetheless, we can not rule out the possibility of other mechanisms which could also be involved in the downregulation of IFN-γ as we observed (e.g. the HRS cells forming aggregates with monocytes and preventing them from interacting with T cells).

Therefore, a more specific experiment like a monocyte-T cell co-culture is needed to further verify this hypothesis.

83 3.4.3 Summary

Besides causing downregulation of CD137L on HRS cells, ectopically expressed CD137 also might get transferred from HRS cells to its surrounding

APC, particularly monocytes. This causes the downregulation of CD137L on the recipient cells, and subsequently leads to reduced T cell response.

(pg/ml) γ γ γ γ IFN-

Figure 30. CD137 expressing L-428 and L-1236 cells inhibit IFN-γγγ releases from PBMC 5 x 10 5 cells/ml of L428 or L1236 cells were co-cultured with 10 6 cells/ml of PBMC suboptimally activated with 1 ng/ml of anti-CD3 (Clone Okt3). IFN-γ release was measured by ELISA 24 h after co-culture. Depicted are means ± standard deviations. * p<0.05. These data are representative of 2 independent experiments.

84 3.5 Potential of CD137 and CD137L signaling on HRS cells

Thus far, the potential beneficial effects of ectopic CD137 on HRS cells have been examined by testing its effect in the HRS cell microenvironment. HRS cells can reduce T cell activation by ectopically expressed CD137.

Nonetheless, CD137 might have more functions. Many members of TNFR superfamily overexpressed by HRS cells have been shown to be beneficial to

HRS cells by supporting downstream signaling. CD30 was shown to sustain constitutive activation of NF-κB in HRS cells (Boll et al., 2005). Another member of the TNFR superfamily, CD40, was shown to manipulate the cytokine response in HRS cells by upregulating IRF4 expression (Aldinucci et al., 2010). Hence, it will be interesting to look at whether stimulation of

CD137 on HRS cells can induce a signaling in HRS cells, and subsequently mediate survival advantages.

In this section, some preliminary experiments have been done to validate that

CD137 on HRS cells is functional in initiating signal transduction will be presented.

3.5.1 Agonistic anti-CD137 antibody

Like other TNFR superfamily members, activation of CD137 involves the crosslinking of surface CD137. In order to crosslink CD137, the antibody needs to be bond to a solid surface. Therefore, anti-CD137 antibodies (clone

BBK-2 or JG-1.6a) were coated onto a plate, and PBMC that had been suboptimally activated with 0.5 ng/ml of anti-CD3 were seeded (Clone Okt3).

IFN-γ release was measured 24 h later. Activated T cells express CD137 on

85 their cell surface and upon ligation with the anti-CD137 antibodies coated on the well plate, will get costimulated and increase IFN-γ production. This can

be used to determine which clone is optimal for CD137 stimulation. From

Figure 31, we can deduce that clone JG1.6a can induce more IFN-γ release

from PBMC than clone BBK-2, which indicates that JG1.6a induces a higher

CD137 response. Hence, clone JG1.6a is more agonistic for CD137, and this

antibody will be used for the rest of the experiments.

Figure 31. Anti-CD137 antibody clone JG1.6a induces more IFN-γ release from PBMC than the anti-CD137 antibody clone BBK-2 PBMC suboptimally activated with 0.5ng/ml of anti-CD3 (clone Okt3) were seeded in plates coated with 5 µg/ml of anti-CD137 antibodies (clone BBK-2 or JG1.6a) or their isotype. An untreated condition was included as a control. IFN-γ release was measured by ELISA 24 h after co-culture. Depicted are means ± standard deviations. * p<0.05. These data are representative of 3 independent experiments.

86 3.5.2 CD137 stimulation causes morphological changes in HRS cells

To test whether CD137 stimulation can affect the physiology of HRS cells, L-

428-control or L-428-CD137 cells were seeded into plates pre-coated with anti-CD137 antibody clone JG1.6a for 48 h. After the incubation, a portion of the L-428-CD137 cells had become more adhesive to the plate and became elongated shape (Figure 32). However, no morphological changes can be observed on L-428-control cells and isotype treated L-428 cells, which confirm that the morphological changes were caused by CD137 signaling.

Hence it is possible that agonistic anti-CD137 antibody does produce functional effect(s) on CD137 expressing L-428-CD137 cells and the ectopic

CD137 on HRS cells is potentially functional and might initiate signaling into

HRS cells.

Figure 32. The CD137 agonistic antibody causes morphological changes on CD137 expressing L-428 cells L-428 cells were seeded in a plate coated with 5 µg/ml of anti-CD137 antibody (clone JG1.6a) or its isotype control. The images were taken 48 h later using an inverted light microscope. Black arrows indicate the cells which become more adhesive and with elongated shape. 630x magnification. These data are representative of 3 independent experiments.

88 3.5.3 Involvement of the cytoplasmic domain of CD137L in HRS cell signaling

During optimization of the confocal microscopy experiment, an anti-CD137L antibody against the cytoplasmic domain of CD137L (cCD137L) has been used (clone: ERP1172Y). Interestingly, most of the signal produced by this antibody was localized in the nucleus of KM-H2 cells (Figure 33A). This is interesting as the staining of extracellular CD137L domain (clone 41B436) is solely found on the plasma membrane and in the cytoplasm (Figure 18). The reason why the extracellular domain and cytoplasmic domain of CD137L are found at different location in HRS cells is unknown, but that can be due to a proteolytic cleavage of CD137L. Nonetheless, the strong localization of the cCD137L in the nucleus suggests that a potential function of CD137L in regulating transcriptional activity. The localization of the cCD137L in the nucleus was also confirmed in a non-HRS cell line (Moh Mei Chung, personal communication).

Proliferating KM-H2 cells can be easily indentified by looking at the DNA condensation via DAPI staining. Whenever KM-H2 cells were proliferating, the cells become negative for cytoplasmic CD137L staining (Figure 33B).

This dramatic change and strong nucleus localization suggest that cytoplasmic

CD137L might be a part of the transcriptional network of the HRS cells, and provides the required survival signal to HRS cells.

A Mouse IgG Rabbit IgG

DAPI CD137 CD137L

Figure 33. The CD137L cytoplasmic domain is present in the nucleus of non-proliferating KM-H2 cells KM-H2 cells were stained with anti-CD137 antibody (clone BBK-2) and anti- cytoplasmic CD137L antibody (clone EPR1172Y) followed by appropriate secondary antibodies. The cells were counterstained with DAPI and visualized using confocal microscopy. Right panels were superimposed with DAPI staining to show the localization of the nucleus. A. Isotype staining. B. Non- proliferating KM-H2 cells. C. White arrow shows the proliferating cells with chromosome condensation. Magnification 400x. These data are representative of 3 independent experiments.

90 3.5.4 Summary

Ectopically expressed CD137 might support HRS cell survival in more complex ways than expected. Besides reducing T cell activation via downregulating CD137L, the CD137 also shows some potential of inducing signaling in HRS cells, which might enhance survival of HRS cells. That is something worth to be further explored in the future.

91 CHAPTER 4 DISCUSSION

The functions of CD137 have been studied mostly on activated T cells and NK cells. Agonistic anti-CD137 antibodies are currently in clinical trials, and plenty of articles describe tumor rejection by anti-CD137 antibodies. CD137 has also been reported to be expressed by other cell types like B cells and DC or during disease state like rheumatoid arthritis and osteosarcoma (as outlined in Introduction section). However, the number of studies on these autoimmune and malignant diseases is very limited and is often at an early stage. Thus, this study shows the tumor promoting effects of CD137 on HRS cells and the potential mechanism for this effect. This is also the first study documenting the transfer of CD137 to CD137L expressing cells, and a new mechanism of regulating CD137 and CD137L expression.

4.1 Ectopically expressed CD137 abrogates T cell activity

One of the emerging hallmarks of cancer is avoiding immune destruction.

Many types of tumor cells were shown to be less immunogenic or capable to suppress the immune system via different mechanism in order to evade from the host's immune system (Hanahan and Weinberg, 2011). In HL, studies have found that HRS cells express high level of MHC class II, transporter- associated proteins 1 (TAP-1) and TAP-2, while MHC class I was always positive in EBV + HRS cells (Murray et al., 1998; Oudejans et al., 1996; Tzardi

et al., 1996). These findings imply that HRS cells are indeed relatively

immunogenic as compared to other cancer cells which have a compromised

capability to antigen processing and presentation. Therefore, HRS cells must

be using other mechanisms to escape from immune surveillance.

92 In this study, ectopic CD137 expression on HRS cell lines was confirmed, and hence in vitro experiments were possible to study the beneficial effect of

ectopic CD137 expression in HL. As shown in Figure 7 and 8, when CD137-

silenced or CD137-neutralized KM-H2 cells were co-cultured with PBMC or

T cells, an increase in IFN-γ release from PBMC and T cells was observed.

This indicates an enhanced T cell activity. CD137-silenced and CD137-

neutralized KM-H2 were further characterized and it was revealed that

ectopically expressed CD137 leads to the downregulation of CD137L (Figure

9 and 10). CD137L is constitutively expressed by antigen presenting cells,

therefore, it is not surprising that HRS cells which originate from GC B cells

are also positive for CD137L. However, the expression of CD137L on

malignant cells causes negative effects for their survival, as CD137L is a co-

stimulatory molecule for T cells responses. Indeed, the downregulation of

CD137L on HRS cells was shown to play a role in the reduced T cell

stimulation (Figure 11 and 12). All of these observations imply that ectopic

CD137 expression on HRS cells can lead to the reduction of T cell activity.

Upon encounter of antigen presented on MHC molecules, T cells undergo

initial activation and co-stimulatory molecules like CD137 and CD28 are

upregulated. Proper co-stimulation is always required for full T cell activation

and prevents AICD and anergy in T cells. CD137 was shown to be a very

potent co-stimulator in enhancing further T cell activation, proliferation, and

in preventing T cells from becoming anergic (Cannons et al., 2001;

Hernandez-Chacon et al., 2011). Hence, the downregulation of CD137L

expression on HRS cells inevitably reduced their potential of inducing T cell

93 co-stimulation and T cell anti-tumor activity which constitutes a survival advantage to HRS cells. This may explain the reason why HRS cells are unable to stimulate proper T cell responses despite being immunogenic.

The reduction of T cell activity induced by ectopic CD137 expression might be strongly related to the immunosuppressive environment observed in HL.

Although infiltration of T cells into primary HL lesions is very common, most of the infiltrated T cells are either Th2 or Treg cells that show little capacity in anti-tumor responses (Koenecke et al., 2008; Marshall et al., 2004). Lack of the Th1 and CTL cells in HL microenvironment can be due to the absence of

CD137L which leads to anergy and AICD of these cytotoxic populations.

Interestingly, IFN-γ release from TIL is often downregulated in HL which is in line with the observation in this study (Marshall et al., 2004).

Although other co-stimulatory molecules like CD80 and CD86 were shown to be expressed by HRS cells, their capability to induce anti-tumor T cell activity in HL was not investigated (Bolognesi et al., 2000; Murray et al., 1995; Vooijs et al., 1997). Therefore, it is uncertain if HRS cells can induce effective anti- tumor responses via their CD80 and CD86 expression even when CD137L is downregulated. Nonetheless, stimulation of CD137 alone using either agonistic CD137 antibody (Ju et al., 2005; Martinet et al., 2002; Murillo et al.,

2008; Murillo et al., 2009; Zhu et al., 2009) or exogenous expression of

CD137L in tumor cells (Li et al., 2008) have been shown to eradicate established tumors in immunocompetent hosts, indicating that CD137 stimulation can enhance anti-tumor responses even in the presence of CD28

94 stimulation. CD137L knock-out mice were also shown to have a higher chance of developing GC B cell lymphoma throughout their life time (Middendorp et al., 2009). These findings again imply the involvement of CD137L and CD137 in immune surveillance, and that the downregulation of CD137L plays a role in the pathogenesis of HL.

In fact, CD137 inducing CD137L downregulation on HRS cells has already been suggested in vivo . Two consecutive studies done by Natkunam's group have shown that CD137 is highly expressed in the primary tumor of HL patients while CD137L expression is absent in all HL (Anderson et al., 2012;

Zhao et al., 2012). This inverse relationship of CD137 and CD137L expression is consistent with our finding in vitro . Interestingly, despite the absence of CD137L expression in HL, some of the subgroups of NHL were found to have high expression of CD137L while they were tested negative for

CD137 (Anderson et al., 2012; Zhao et al., 2012). This mutually exclusive relationship between CD137 and CD137L expression in primary tumors implies that CD137 inducing CD137L downregulation indeed occurs in in vivo.

However, high expression of CD137L on NHL cells is a clear survival disadvantage for NHL tumors as it will enhance T cell co-stimulation. NHL cells might have developed other mechanisms in order to counteract the disadvantage caused by the CD137L expression. One of the possibilities is via sCD137 that is often upregulated in serum of NHL patients (Furtner et al.,

2005). sCD137 can act as a competitor of membrane bound CD137 and hence, neutralize the effect of CD137L and diminish anti-tumor immune responses.

95 Considering that HRS cells comprise as little as 1% of the total HL mass, the effect of CD137L downregulation on HRS cells alone might not be sufficient in orchestrating the whole immunosuppressive environment that is observed in

HL. Thus, the effects of ectopically expressed CD137 on infiltrating cells in

HL were also examined in this study. Surprisingly, transfer of CD137 from

HRS cells to their surrounding cells and especially monocytes is possible as demonstrated in the co-culture experiment (Figures 27 and 28). The transfer of

CD137 causes CD137L downregulation on the monocytes which subsequently leads to reduced IFN-γ release from T cells when co-cultures were performed

(Figure 29 and 30). This finding confirms that the CD137L downregulating effect of CD137 is not only restricted to HRS cells but also causes a reduction of CD137L expression on monocytes in the HL microenvironment. This actually has a larger effect on establishing immunosuppressive conditions since the majority of the HL tumor mass is made up of APC like macrophages and B cells, and CD137L downregulation on these cells will produce a more substantial immunosuppressive effect than CD137L downregulation on HRS cells alone.

It is worth noting that a higher ratio of HRS cells was used in the current in

vitro study. However, the number of HRS cells in the primary HL lesion is

much lower, hence the total incidents of a HRS cell coming in contact with

another HRS cell or their surrounding APC will also be reduced. Therefore, it

is likely that the ectopic CD137 expression on HRS cells might induce a lesser

immunosuppressive effect in in vivo condition, as compared to the current

models used in this study. As a result, it is proposed that HRS cells might need

96 to adopt other mechanisms to suppress the host immune response further. For example, the release of soluble factor which can neutralize great amounts of

CD137L or CD137 in their vicinity, or the release of other immunosuppressive cytokines.

Figure 34 is a schematic illustration on how HL escapes from immune surveillance by ectopic CD137 expression. Under normal circumstances,

CD137 - malignant B cells will be removed by the T cell anti-tumor activity which is induced by binding of CD137 on T cells to CD137L expressed by both malignant B cells and their surrounding APC. However, by expressing ectopic CD137, HRS cells can induce downregulation of CD137L on both

HRS cells and APC (especially monocytes), and reduce T cell co-stimulation, hence protecting them from T cell cytotoxicity and obtain a survival advantage.

97 A

Figure 34. Ectopic CD137 promotes HRS cells survival via downregulation of CD137L expression (Ho et al., 2012) Under normal physiology conditions (A), Malignant B cells and APC (B cells and monocytes) express CD137L on their cell surface, thereby co-stimulating T cells and enhancing T cell cytotoxicity that leads to the removal of tumor cells. By overexpressing CD137 (B), HRS cells can downregulate CD137L expression on both HRS cells and APC through internalization of the CD137- CD137L complex, and gain survival advantages by reducing T cell co- stimulation and cytotoxicity.

98 4.2 CD137 trogocytosis

Intercellular transfer of membrane material emerged as an important topic recently, and the transfer was shown to be crucial in maintaining cell to cell communication and regulation of cellular function especially in the central nervous system and immune system. The most common mechanisms involved in the transfer of membrane material are nanotube formation, exosome secretion and trogocytosis. Among these mechanisms, trogocytosis actively regulates immune responses. However, the detail on how membrane material is transferred through trogocytosis is not fully understood thus far (Ahmed and

Xiang, 2011; Rechavi et al., 2009).

In this study, CD137 was found to be transferred from HRS cells to their surrounding cells, especially to monocytes or to another tumor cell line via trogocytosis. When CD137 binds to CD137L, the transfer of CD137 occurs via trogocytosis, followed by the internalization of the CD137-CD137L complex into the cytoplasm. Removal of CD137L from the cell surfaces reduces the amount of CD137L present in the HL microenvironment and leads to the reduction of T cell costimulation.

Trogocytosis of CD137 was solely described in the context of HL in this study, and how this process enhances survival advantages of HRS cells in their microenvironment. However, it is unlikely that trogocytosis of CD137 is only participating in the pathogenesis of HL. Rather, CD137 trogocytosis and

CD137L downregulation is hypothesized to be involved in the normal regulation of CD137-CD137L and in modulating immune responses, while

99 HRS cells just adopt this mechanism for their benefit. By extrapolating the observations in this study, there is a good possibility that trogocytosis of

CD137 serves as a negative feedback mechanism in order to remove CD137L from APC after sufficient T cell costimulation. However, this idea was not explored in this study, and further empirical evidence is needed to confirm this assumption.

In fact, some unpublished data from our group suggest the possibility of an involvement of CD137 trogocytosis in other conditions besides HL. For example, transfer of murine CD137 had been shown from the CD137- overexpressing A20 cell line (murine B cell lymphoma) to monocytes or other cancer cell lines through trogocytosis. The transfer of murine CD137 was also accompanied by the downregulation of CD137L expression in recipient cells.

CD137 knock out mice also express more CD137L compared to wild type mice (personal communication, Herbert Schwarz). This inverse relationship of

CD137 and CD137L expression in mice is comparable to the findings in HL, thus suggesting that CD137 might be involved in the turnover of CD137L in other systems.

Notably, Treg cells were also shown to acquire CD80 and CD86 from APC via trogocytosis and this leads to the downregulation of CD80 and CD86 in these APC. (Gu et al., 2012; Qureshi et al., 2011). The downregulation of

CD80 and CD86 on APC weakens the stimulation of T cells through CD28, and thus affects T cell activation. These findings are similar to the observations in HL in this study, where ectopic CD137 expression induces

100 CD137L downregulation on HRS cells and APC, and thus reduces T cell activity. Therefore, it is not surprising if CD137 trogocytosis is involved in the regulation of immune response as suggested.

4.2.1 CD137 and CD137L transfer

Rather than being a random event, CD137 transfer is a highly organized and

CD137L-dependent process as shown in this study. The transfer of CD137 was shown to occur from CD137 overexpressing HRS cells only to KMS-11-

CD137L cells but not to KMS-11-control cells. Both KMS-11-CD137L and

KMS-11-control cells have a high background of trogocytosis (as shown by high PKH-26 transfer in Figure 23) but only KMS-11-CD137L cells can induce the CD137 transfer from HRS cells. This implies that the transfer of

CD137 is dependent on CD137L expression on recipient cells regardless of other ongoing transfers of protein and membrane material between donor and recipient cells. The specificity in the CD137 transfer may ensure that the efficiency of the transfer is maximized by avoiding the transfer to non-specific cells.

The specificity in the transfer of membrane proteins is not restricted to CD137 but was also described in the transfer of CD80 and CD86 from APC to CTLA-

4 expressing T cells, and the transfer of Killer-cell Immunoglobulin-like

Receptor (KIR) from NK cells to MHC class I expressing cell (Qureshi et al.,

2011; Vanherberghen et al., 2004). Unspecific intercellular transfer of membrane proteins was also reported (Baba et al., 2001; Gu et al., 2012), however it is likely that those unspecific transfers are transfers of bystander

101 proteins. He et al. showed that MHC class I molecules are co-transferred as bystander proteins with MHC class II molecules when they share the same lipid rafts. This implies that a protein might be transferred if it is present in a close proximity of a trogocytosis site even if there is no specific binding partner for it. But in the case of CD137, no CD137 was transferred non- specifically from HRS cells to CD137 - KMS-11 cells, which suggests that

CD137 is not transferred as bystander protein at least in HRS cells.

Nonetheless, unpublished data from our group show murine CD137 might be transferred in a CD137L independent manner in B cell lymphoma models, thus suggesting that the CD137L dependency of CD137 transfer might vary depending on the species or circumstances (Pang Wan Lu, personal communication).

Besides CD137 transfer, it is surprising that CD137L also get transferred from

CD137L-expressing cells to CD137-expressing HRS cells, and that the transfer of CD137L is also strictly CD137 dependent (Figure 25 and 26). Bi- directional transfer of membrane protein is not restricted to CD137 and

CD137L since bidirectional transfer of KIR and MHC class I molecules was also reported (Carlin et al., 2001; Vanherberghen et al., 2004). However, the reason for this bi-directional transfer of CD137 and CD137L between cells has not been investigated in this study. The hypothesis is that CD137L acquired by

HRS cells might continue to induce CD137 signaling in HRS cells which will enhance their survival. Besides that, CD137L transfer might also cause the removal of CD137L from HRS cells and APC, and lead to reduced T cell co- stimulation.

102 4.2.2 Aggregation of CD137 and CD137L in the cytoplasm

Trogocytosis of CD137 is followed by the removal of CD137L from the cell membrane. Aggregation of the CD137-CD137L complex in the cytoplasm of

HRS cells was also observed, and it is most likely an event following the

CD137 trogocytosis, thereby removing the CD137-CD137L complex from the cell surface (Figure 18). Interestingly, the aggregation of CD137 was also shown by Anderson et al. in primary HRS cells in HL lesion. This is a strong verification of in vitro result in this study in in vivo condition. However, no indication of the reason for CD137 aggregation in HRS cells was provided in

Anderson et al. study.

Even though the co-internalization of the CD137-CD137L complex was not described in any system prior to this study, the internalization of other TNFR superfamily members has been shown extensively. For instance, CD40, CD95 and TNF-R1 were shown to be internalized quickly upon ligation of their respective ligands (Chen et al., 2006; Eramo et al., 2004; Kohlhaas et al., 2007;

Mosselmans et al., 1988; Reinehr and Haussinger, 2008). The internalization process was deemed to be critical for proper signaling of these TNFR members in most of the studies. Besides that, other TNFR family members,

TACI and BCMA, were found to colocalize with their respective ligands,

BAFF and APRIL, in the cytoplasm of HRS cells. The signals transduced by

TACI and BCMA were shown to enhance HRS cells survival by upregulating anti-apoptotic regulatory proteins (Chiu et al., 2007).

103 As discussed, internalization and clustering of members of TNFR superfamily is essential for their downstream signaling and proper function, thus, there is a possibility that the aggregation of CD137 in the cytoplasm of HRS cells has similar properties. Stimulation of CD137 was shown to activate NF-κB

pathway (Arch and Thompson, 1998; Jang et al., 1998), which is a survival

advantage for HRS cells. Therefore, it is possible that the aggregation of

CD137 in the cytoplasm leads to the continuous CD137 signaling in HRS cells

and partly contributes to the constitutive upregulation of NF-κB in HRS cells.

4.3 CD137 and CD137L induce signaling into HRS cells

Besides promoting tumor cell survival through modulating immune responses

in the HL microenvironment, CD137 was shown in this study to stimulate

morphological changes in HRS cells (Figure 32), which suggests the potential

of CD137 signaling into HRS cells. This result is still at a preliminary stage,

and the outcome of the signaling through CD137 is still unknown thus far.

However, it is likely that the signaling transduced by CD137 is favorable to

HRS cell growth.

HRS cells were shown to express many members of the TNFR superfamily on

their cell surface, and the expression of these TNFR superfamily members is

beneficial to HRS cells. Particularly, CD30, CD40 and TACI were shown to

activate NF-κB in HRS cells upon stimulation (Annunziata et al., 2000; Boll et al., 2005; Chiu et al., 2007). NF-κB has been shown to be an important transcription factor contributing to the pathogenesis of HL, and blockage of

NF-κB activity reduces proliferation of HRS cells (Younes et al., 2003). Apart

104 from that, CD137 was also shown to induce NF-κB activity (Arch and

Thompson, 1998; Jang et al., 1998). Since CD137 signaling in HRS cells is

functional, it is possible that constitutive expression of CD137 can be

contributed to the constant activation of NF-κB found in HL. However, this hypothesis needs to be further tested.

One of the most unexpected findings from this study is the presence of cCD137L in the nucleus of HRS cells while full length CD137L is primarily found in the cytoplasm and the cell membrane (Figure 18 and 33). This finding was confirmed in a non-HRS cell line by a colleague in my lab, hence ruling out the possibility of an artifact (Moh Mei Chung, personal communication). Whether cCD137L has a role in the pathogenesis of HRS is not elucidated in this study. However, based on its nuclear localization and disappearance during mitosis, it is speculated that cCD137L might be involved in the transcriptional activity of HRS cells. cCD137L may also participate in the signaling in non-malignant cells, rather than just in HL. Nevertheless, this hypothesis has yet to be investigated.

cCD137L is not the only cytoplasmic domain of a membrane bound protein which has been found to be translocated from the plasma membrane to the nucleus for its function. The other best known example is Notch. Upon stimulation, Notch is cleaved by γ-secretase and its cytoplasmic domain,

Notch intracellular domain (NICD), is translocated from the cell membrane to the nucleus. NICD has been shown to contain a nuclear localization sequence

105 (NLS), which facilitates the nuclear localization. NICD is able to modulate the transcriptional activity of the cells (Guruharsha et al., 2012).

It is not sure whether translocation of cCD137L to nucleus involves a similar pathway as Notch signaling since no NLS has been identified in CD137L thus far. Nonetheless, cCD137L is a relatively small protein (~15kD, Dr. Moh Mei

Chung, personal communication) therefore it might not be a transcription factor and interact directly with DNA. However, cCD137L might be capable of modulating transcription through the interaction with other nuclear proteins or nuclear factors.

4.4 Limitations

One of the most significant limitations of this study is that all of the experiments are performed using cell lines and in vitro . Although there are plenty of relevant correlations of this study to the clinical observations of HL, a confirmation in an in vivo system would be desirable. The main obstacle for

performing in vivo experiments is the lacking of proper animal models. Many

groups have tried to xenotransplant HL cell lines into immunocompromised

mice (Dewan et al., 2005), but the usages of these models is often limited to

drug testing and are not suitable for this current study, as the interaction of

HRS cells with a "human" microenvironment is necessary. An alternative to

the immunocompromised mice is to establish a humanized mouse model but

this will be both technically challenging and time consuming.

106 On the other hand, this study addresses the effect of CD137 on HL in a collective manner without considering the effect of CD137 in each individual subtype of HL. As detailed in the Introduction, different HL subtypes have distinct pathological features and these variations might cause a different involvement of the ectopically expressed CD137 in the pathogenesis of a particular HL subtype. For example in the lymphocyte depleted subtype of HL, histiocytes are the predominant infiltrating cells that surround HRS cells. The abundance of histiocytes might affect the downregulation of CD137L simply because the CD137L expressed by the histiocytes in the microenvironment is likely to outnumber the CD137 expressed by HRS cells. Therefore, it is possible that the ectopic CD137 expression might have different effects on each HL subtype. Therefore, further characterization of CD137 in each HL subtype should be performed.

4.5 Future works

In this current study, a novel mechanism which is used by HRS cells to escape host immunity has been described. This finding provides an insight which could be useful for understanding the pathogenesis of other malignancies and autoimmune diseases or for developing a more effective immunotherapy for

HL patients. Here, I would like to discuss a few potential future projects which will further strengthen the hypothesis in this study, and other potential branches resulting from this work which might be equally important.

First, the CD137 induced downregulation of CD137L should be confirmed in

vivo . However, there is currently no suitable animal model for Hodgkin

107 lymphoma as mentioned earlier. Although generation of a Hodgkin lymphoma model using NOG (NOD/Shi-scid/IL-2R γnull) mice has been reported earlier, the reproducibility of this murine model is unknown and might be unsuitable for our future works (Dewan et al., 2005). Hence the easier and more realistic alternative is to observe the effect of CD137 in primary HL lesions from HL patients. We should be able to observe the colocalization of CD137 and

CD137L in HRS cells and more importantly, to verify the downregulation of

CD137L in the HL microenvironment.

CD137 expressed on HRS cells is functional and the possibility that it may be inducing signal into HRS cells has been shown. In this study, a coated anti

CD137 antibody (clone JG1.6a) has been shown to induce morphological changes of HRS cells. However, the signaling mechanism leading to these morphological changes are still unknown. Hence, it is worth to investigate the advantages HRS cells acquire by this potential CD137 signalling. Does

CD137 signalling play a role in the constitutive activation of NF-κB in cells?

The effect of CD137 signalling on the proliferation and viability of HRS cells

should also be investigated.

Ectopic CD137 expression was also found in other cancers like Follicular DC

cancer, T cell lymphoma and osteosarcoma. Hence, it is important to

characterize the effects of ectopically expressed CD137 on these cancers.

Besides that, it is also important to confirm whether the ectopic CD137

expression leads to a similar immune escape mechanism in these cancers as

108 the one identified in HL. The new knowledge will lead to a better understanding of the pathogenesis of these malignant diseases.

As suggested in the previous section, intercellular CD137 trogocytosis and

CD137L downregulation is unlikely to be a mechanism solely used by HL for their survival. This mechanism is more likely involved in CD137-CD137L regulation under normal physiological conditions as well. Hence, the transfer of CD137 and CD137L downregulation should be confirmed in primary cells.

For example, the transfer of CD137 from activated T cells to APC and its functional consequences should be demonstrated.

Last but not least, the ultimate goal of this study is to make use of the current knowledge to improve the clinical outcome of patients. A neutralizing antibody might be useful to neutralize CD137 on HRS cells, and hence, enhance the elimination of HRS cells. However, the concern is that the anti-

CD137 neutralizing antibody might also be distributed through the blood stream and cause a systemic neutralization of CD137. This might eventually lead to the decrease of T cell and NK cell activity in patients, and worsen their condition. In order to overcome this issue, an anti-CD30/anti-CD137 bi- specific antibody can be developed. The bi-specific antibody will have a higher affinity towards CD30 + CD137 + HRS cells, than towards cells expressing only CD137 or CD30 on their cell surface. By adjusting the titre of the bi-specific antibody, we propose that it can specifically target HRS cells while leaving other immune cells untouched.

109 4.6 Conclusion

In this study, the effects of ectopic CD137 expression on survival of HRS cells have been elucidated. Ectopic CD137 expression has been shown to downregulate the expression of CD137L in both HRS cells and their surrounding cells which leads to a decreased T cell activity. CD137L downregulation is caused by interaction of CD137L with CD137 and the formation of a CD137-CD137L complex. This complex is then removed from the cell surface via endocytosis.

This study has described a new immune escape mechanism used by HRS cells for their survival advantage of HRS cells which will be important for the understanding of the pathogenesis of HL, and potentially the pathogenesis of other malignant diseases as well. In addition, this study also suggests a new regulatory mechanism of CD137 and CD137L that could be important in modulating immune responses in both a normal and a disease state.

Nonetheless, further studies are required to understand the involvement of the identified mechanism in vivo and under physiological condition, and whether

CD137 can be a potential therapeutic target and a prognostic marker of HL.

110 REFERENCES

Abe, M., E. Wachi, and H. Wakasa. 1988. Simultaneous expression of activated lymphoid cell-associated and granulocytic cell-associated antigens on Hodgkin's and Reed-Sternberg cells in Hodgkin's disease. Pathology, Research and Practice 183:418-424.

Ahmed, K.A., and J. Xiang. 2011. Mechanisms of cellular communication through intercellular protein transfer. Journal of Cellular and Molecular Medicine 15:1458-1473.

Aldinucci, D., B. Rapana, K. Olivo, D. Lorenzon, A. Gloghini, A. Colombatti, and A. Carbone. 2010. IRF4 is modulated by CD40L and by apoptotic and anti-proliferative signals in Hodgkin lymphoma. British Journal of Haematology 148:115-118.

Ambinder, R.F. 2007. Epstein-barr virus and hodgkin lymphoma. Hematology / the Education Program of the American Society of Hematology. American Society of Hematology. Education Program 204-209.

Anderson, M.W., S. Zhao, A.G. Freud, D.K. Czerwinski, H. Kohrt, A.A. Alizadeh, R. Houot, D. Azambuja, I. Biasoli, J.C. Morais, N. Spector, H.F. Molina-Kirsch, R.A. Warnke, R. Levy, and Y. Natkunam. 2012. CD137 is expressed in follicular dendritic cell tumors and in classical Hodgkin and T-cell lymphomas: diagnostic and therapeutic implications. The American Journal of Pathology 181:795-803.

Annunziata, C.M., Y.J. Safiran, S.G. Irving, U.N. Kasid, and J. Cossman. 2000. Hodgkin disease: pharmacologic intervention of the CD40-NF kappa B pathway by a protease inhibitor. Blood 96:2841-2848.

Arai, A.I., K.; Takahashi, M.; Naka, K.; Fukuda, T.; Fujiwara, T.; Miura, O. 2009. CD137 Expression Is Enhanced in EBV-Infected T or NK Cells by Viral Protein LMP1 and Mediates Anti-Apoptotic Intracellular Signaling through NF-κB Activation. In 51st ASH Annual Meeting and Exposition. American Society of Hematology New Orlean.

Arch, R.H., and C.B. Thompson. 1998. 4-1BB and Ox40 are members of a tumor necrosis factor (TNF)-nerve growth factor receptor subfamily that bind TNF receptor-associated factors and activate nuclear factor kappaB. Molecular and Cellular Biology 18:558-565.

Aucher, A., E. Magdeleine, E. Joly, and D. Hudrisier. 2008. Capture of plasma membrane fragments from target cells by trogocytosis requires signaling in T cells but not in B cells. Blood 111:5621-5628.

Austin, C.D., D.A. Lawrence, A.A. Peden, E.E. Varfolomeev, K. Totpal, A.M. De Maziere, J. Klumperman, D. Arnott, V. Pham, R.H. Scheller, and A. Ashkenazi. 2006. Death-receptor activation halts clathrin-dependent

111 endocytosis. Proceedings of the National Academy of Sciences of the United States of America 103:10283-10288.

Baba, E., Y. Takahashi, J. Lichtenfeld, R. Tanaka, A. Yoshida, K. Sugamura, N. Yamamoto, and Y. Tanaka. 2001. Functional CD4 T cells after intercellular molecular transfer of 0X40 ligand. Journal of Immunology 167:875-883.

Bechtel, D., J. Kurth, C. Unkel, and R. Kuppers. 2005. Transformation of BCR-deficient germinal-center B cells by EBV supports a major role of the virus in the pathogenesis of Hodgkin and posttransplantation lymphomas. Blood 106:4345-4350.

Bell, G.I. 1978. Models for the specific adhesion of cells to cells. Science 200:618-627.

Beum, P.V., D.A. Mack, A.W. Pawluczkowycz, M.A. Lindorfer, and R.P. Taylor. 2008. Binding of rituximab, trastuzumab, cetuximab, or mAb T101 to cancer cells promotes trogocytosis mediated by THP-1 cells and monocytes. Journal of Immunology 181:8120-8132.

Boll, B., H. Hansen, F. Heuck, K. Reiners, P. Borchmann, A. Rothe, A. Engert, and E. Pogge von Strandmann. 2005. The fully human anti-CD30 antibody 5F11 activates NF-{kappa}B and sensitizes lymphoma cells to bortezomib-induced apoptosis. Blood 106:1839-1842.

Bolognesi, A., L. Polito, P.L. Tazzari, R.M. Lemoli, C. Lubelli, M. Fogli, L. Boon, M. de Boer, and F. Stirpe. 2000. In vitro anti-tumour activity of anti-CD80 and anti-CD86 immunotoxins containing type 1 ribosome- inactivating proteins. British Journal of Haematology 110:351-361.

Broll, K., G. Richter, S. Pauly, F. Hofstaedter, and H. Schwarz. 2001. CD137 expression in tumor vessel walls. High correlation with malignant tumors. American Journal of Clinical Pathology 115:543-549.

Cannons, J.L., P. Lau, B. Ghumman, M.A. DeBenedette, H. Yagita, K. Okumura, and T.H. Watts. 2001. 4-1BB ligand induces cell division, sustains survival, and enhances effector function of CD4 and CD8 T cells with similar efficacy. Journal of Immunology 167:1313-1324.

Carbone, A., A. Gloghini, V. Gattei, D. Aldinucci, M. Degan, P. De Paoli, V. Zagonel, and A. Pinto. 1995. Expression of functional CD40 antigen on Reed-Sternberg cells and Hodgkin's disease cell lines. Blood 85:780-789.

Carlin, L.M., K. Eleme, F.E. McCann, and D.M. Davis. 2001. Intercellular transfer and supramolecular organization of human leukocyte antigen C at inhibitory natural killer cell immune synapses. The Journal of Experimental Medicine 194:1507-1517.

112 Chen, Y., J. Chen, Y. Xiong, Q. Da, Y. Xu, X. Jiang, and H. Tang. 2006. Internalization of CD40 regulates its signal transduction in vascular endothelial cells. Biochemical and Biophysical Research Communications 345:106-117.

Chiu, A., W. Xu, B. He, S.R. Dillon, J.A. Gross, E. Sievers, X. Qiao, P. Santini, E. Hyjek, J.W. Lee, E. Cesarman, A. Chadburn, D.M. Knowles, and A. Cerutti. 2007. Hodgkin lymphoma cells express TACI and BCMA receptors and generate survival and proliferation signals in response to BAFF and APRIL. Blood 109:729-739.

Cox, J.H., A.J. McMichael, G.R. Screaton, and X.N. Xu. 2007. CTLs target Th cells that acquire bystander MHC class I-peptide complex from APCs. Journal of Immunology 179:830-836.

Davis, D.M. 2007. Intercellular transfer of cell-surface proteins is common and can affect many stages of an immune response. Nature Reviews. Immunology 7:238-243.

DeBenedette, M.A., N.R. Chu, K.E. Pollok, J. Hurtado, W.F. Wade, B.S. Kwon, and T.H. Watts. 1995. Role of 4-1BB ligand in costimulation of T lymphocyte growth and its upregulation on M12 B lymphomas by cAMP. The Journal of Experimental Medicine 181:985-992.

Dewan, M.Z., M. Watanabe, S. Ahmed, K. Terashima, S. Horiuchi, T. Sata, M. Honda, M. Ito, T. Watanabe, R. Horie, and N. Yamamoto. 2005. Hodgkin's lymphoma cells are efficiently engrafted and tumor marker CD30 is expressed with constitutive nuclear factor-kappaB activity in unconditioned NOD/SCID/gammac(null) mice. Cancer Science 96:466-473.

Dietlein, M., S.M. Borner, T. Fischer, H. Hansen, R. Schnell, B. Zimmermanns, S. Tawadros, A. Engert, O. Staak, E. Pogge von Strandmann, C. Kobe, H. Schicha, and K. Schomacker. 2010. Development of anti-CD30 radioimmunoconstructs (RICs) for treatment of Hodgkin's lymphoma. Studies with cell lines and animal studies. Nuklearmedizin. Nuclear medicine 49:97-105.

Dutton, A., J.D. O'Neil, A.E. Milner, G.M. Reynolds, J. Starczynski, J. Crocker, L.S. Young, and P.G. Murray. 2004. Expression of the cellular FLICE-inhibitory protein (c-FLIP) protects Hodgkin's lymphoma cells from autonomous Fas-mediated death. Proceedings of the National Academy of Sciences of the United States of America 101:6611-6616.

Dykstra, M.L., R. Longnecker, and S.K. Pierce. 2001. Epstein-Barr virus coopts lipid rafts to block the signaling and antigen transport functions of the BCR. Immunity 14:57-67.

113 Eerola, A.K., Y. Soini, and P. Paakko. 2000. A high number of tumor- infiltrating lymphocytes are associated with a small tumor size, low tumor stage, and a favorable prognosis in operated small cell lung carcinoma. Clinical Cancer Research : An Official Journal of the American Association for Cancer Research 6:1875-1881.

Eramo, A., M. Sargiacomo, L. Ricci-Vitiani, M. Todaro, G. Stassi, C.G. Messina, I. Parolini, F. Lotti, G. Sette, C. Peschle, and R. De Maria. 2004. CD95 death-inducing signaling complex formation and internalization occur in lipid rafts of type I and type II cells. European Journal of Immunology 34:1930-1940.

Evens, A.M., M. Hutchings, and V. Diehl. 2008. Treatment of Hodgkin lymphoma: the past, present, and future. Nature Clinical Practice. Oncology 5:543-556.

Fiumara, P., V. Snell, Y. Li, A. Mukhopadhyay, M. Younes, A.M. Gillenwater, F. Cabanillas, B.B. Aggarwal, and A. Younes. 2001. Functional expression of receptor activator of nuclear factor kappaB in Hodgkin disease cell lines. Blood 98:2784-2790.

Furtner, M., R.H. Straub, S. Kruger, and H. Schwarz. 2005. Levels of soluble CD137 are enhanced in sera of leukemia and lymphoma patients and are strongly associated with chronic lymphocytic leukemia. Leukemia : Official Journal of the Leukemia Society of America, Leukemia Research Fund, U.K 19:883-885.

Gary, R., S. Voelkl, R. Palmisano, E. Ullrich, J.J. Bosch, and A. Mackensen. 2012. Antigen-specific transfer of functional programmed death ligand 1 from human APCs onto CD8+ T cells via trogocytosis. Journal of Immunology 188:744-752.

Goodwin, R.G., W.S. Din, T. Davis-Smith, D.M. Anderson, S.D. Gimpel, T.A. Sato, C.R. Maliszewski, C.I. Brannan, N.G. Copeland, N.A. Jenkins, and et al. 1993. Molecular cloning of a ligand for the inducible T cell gene 4-1BB: a member of an emerging family of cytokines with homology to tumor necrosis factor. European Journal of Immunology 23:2631-2641.

Gorczyca, W., J. Weisberger, Z. Liu, P. Tsang, M. Hossein, C.D. Wu, H. Dong, J.Y. Wong, S. Tugulea, S. Dee, M.R. Melamed, and Z. Darzynkiewicz. 2002. An approach to diagnosis of T-cell lymphoproliferative disorders by flow cytometry. Cytometry 50:177- 190.

Green, M.R., S. Rodig, P. Juszczynski, J. Ouyang, P. Sinha, E. O'Donnell, D. Neuberg, and M.A. Shipp. 2012. Constitutive AP-1 activity and EBV infection induce PD-L1 in Hodgkin lymphomas and posttransplant lymphoproliferative disorders: implications for targeted therapy.

114 Clinical Cancer Research : An Official Journal of the American Association for Cancer Research 18:1611-1618.

Gu, P., J.F. Gao, C.A. D'Souza, A. Kowalczyk, K.Y. Chou, and L. Zhang. 2012. Trogocytosis of CD80 and CD86 by induced regulatory T cells. Cellular & Molecular Immunology 9:136-146.

Gualberto, A. 2012. Brentuximab Vedotin (SGN-35), an antibody-drug conjugate for the treatment of CD30-positive malignancies. Expert Opinion on Investigational Drugs 21:205-216.

Gullo, C., L.K. Koh, W.L. Pang, K.T. Ho, S.H. Tan, and H. Schwarz. 2010. Inhibition of proliferation and induction of apoptosis in multiple myeloma cell lines by CD137 ligand signaling. PloS One 5:e10845.

Guruharsha, K.G., M.W. Kankel, and S. Artavanis-Tsakonas. 2012. The Notch signalling system: recent insights into the complexity of a conserved pathway. Nature reviews. Genetics 13:654-666.

Hanahan, D., and R.A. Weinberg. 2011. Hallmarks of cancer: the next generation. Cell 144:646-674.

He, T., S. Zong, X. Wu, Y. Wei, and J. Xiang. 2007. CD4+ T cell acquisition of the bystander pMHC I colocalizing in the same immunological synapse comprising pMHC II and costimulatory CD40, CD54, CD80, OX40L, and 41BBL. Biochemical and Biophysical Research Communications 362:822-828.

Helft, J., A. Jacquet, N.T. Joncker, I. Grandjean, G. Dorothee, A. Kissenpfennig, B. Malissen, P. Matzinger, and O. Lantz. 2008. Antigen-specific T-T interactions regulate CD4 T-cell expansion. Blood 112:1249-1258.

Hernandez-Chacon, J.A., Y. Li, R.C. Wu, C. Bernatchez, Y. Wang, J.S. Weber, P. Hwu, and L.G. Radvanyi. 2011. Costimulation through the CD137/4-1BB pathway protects human melanoma tumor-infiltrating lymphocytes from activation-induced cell death and enhances antitumor effector function. Journal of Immunotherapy 34:236-250.

Hirsch, B., M. Hummel, S. Bentink, F. Fouladi, R. Spang, R. Zollinger, H. Stein, and H. Durkop. 2008. CD30-induced signaling is absent in Hodgkin's cells but present in anaplastic large cell lymphoma cells. The American Journal of Pathology 172:510-520.

Ho, W.T., W.L. Pang, S.M. Chong, A. Castella, S. Al-Salam, T.E. Tan, M.C. Moh, L.K. Koh, S.U. Gan, C.K. Cheng, and H. Schwarz. 2012. Expression of CD137 on Hodgkin and Reed-Sternberg Cells Inhibits T cell Activation by Eliminating CD137 Ligand Expression. Cancer Research

115 Hodgkin. 1832. On some Morbid Appearances of the Absorbent Glands and Spleen. Medico-chirurgical Transactions 17:68-114.

HoWangYin, K.Y., E. Alegre, M. Daouya, B. Favier, E.D. Carosella, and J. LeMaoult. 2010. Different functional outcomes of intercellular membrane transfers to monocytes and T cells. Cellular and Molecular Life Sciences : CMLS 67:1133-1145.

Hudrisier, D., A. Aucher, A.L. Puaux, C. Bordier, and E. Joly. 2007. Capture of target cell membrane components via trogocytosis is triggered by a selected set of surface molecules on T or B cells. Journal of Immunology 178:3637-3647.

Jang, I.K., Z.H. Lee, Y.J. Kim, S.H. Kim, and B.S. Kwon. 1998. Human 4- 1BB (CD137) signals are mediated by TRAF2 and activate nuclear factor-kappa B. Biochemical and Biophysical Research Communications 242:613-620.

Jiang, D., Y. Chen, and H. Schwarz. 2008a. CD137 induces proliferation of murine hematopoietic progenitor cells and differentiation to macrophages. Journal of Immunology 181:3923-3932.

Jiang, D., P.S. Yue, D. Drenkard, and H. Schwarz. 2008b. Induction of proliferation and monocytic differentiation of human CD34+ cells by CD137 ligand signaling. Stem Cells 26:2372-2381.

Jiang, D., and H. Schwarz. 2010. Regulation of granulocyte and macrophage populations of murine bone marrow cells by G-CSF and CD137 protein. PloS One 5:e15565.

Ju, S.A., S.C. Lee, T.H. Kwon, S.K. Heo, S.M. Park, H.N. Paek, J.H. Suh, H.R. Cho, B. Kwon, B.S. Kwon, and B.S. Kim. 2005. Immunity to melanoma mediated by 4-1BB is associated with enhanced activity of tumour-infiltrating lymphocytes. Immunology and Cell Biology 83:344-351.

Kapp, U., W.C. Yeh, B. Patterson, A.J. Elia, D. Kagi, A. Ho, A. Hessel, M. Tipsword, A. Williams, C. Mirtsos, A. Itie, M. Moyle, and T.W. Mak. 1999. Interleukin 13 is secreted by and stimulates the growth of Hodgkin and Reed-Sternberg cells. The Journal of Experimental Medicine 189:1939-1946.

Kim, L.H., G.I. Eow, S.C. Peh, and S. Poppema. 2003. The role of CD30, CD40 and CD95 in the regulation of proliferation and apoptosis in classical Hodgkin's lymphoma. Pathology 35:428-435.

Koenecke, C., S.N. Ukena, A. Ganser, and A. Franzke. 2008. Regulatory T cells as therapeutic target in Hodgkin's lymphoma. Expert Opinion on Therapeutic Targets 12:769-782.

116 Kohlhaas, S.L., A. Craxton, X.M. Sun, M.J. Pinkoski, and G.M. Cohen. 2007. Receptor-mediated endocytosis is not required for tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)-induced apoptosis. The Journal of Biological Chemistry 282:12831-12841.

Kuppers, R., K. Rajewsky, M. Zhao, G. Simons, R. Laumann, R. Fischer, and M.L. Hansmann. 1994. Hodgkin disease: Hodgkin and Reed-Sternberg cells picked from histological sections show clonal immunoglobulin gene rearrangements and appear to be derived from B cells at various stages of development. Proceedings of the National Academy of Sciences of the United States of America 91:10962-10966.

Kuppers, R., M.L. Hansmann, and K. Rajewsky. 1998. Clonality and germinal centre B-cell derivation of Hodgkin/Reed-Sternberg cells in Hodgkin's disease. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO 9 Suppl 5:S17-20.

Kuppers, R., and K. Rajewsky. 1998. The origin of Hodgkin and Reed/Sternberg cells in Hodgkin's disease. Annual Review of Immunology 16:471-493.

Kwajah, M.M.S., and H. Schwarz. 2010. CD137 ligand signaling induces human monocyte to dendritic cell differentiation. European Journal of Immunology 40:1938-1949.

Kwon, B.S., and S.M. Weissman. 1989. cDNA sequences of two inducible T- cell genes. Proceedings of the National Academy of Sciences of the United States of America 86:1963-1967.

Lam, N., and B. Sugden. 2003. CD40 and its viral mimic, LMP1: similar means to different ends. Cellular Signalling 15:9-16.

Lamprecht, B., S. Kreher, I. Anagnostopoulos, K. Johrens, G. Monteleone, F. Jundt, H. Stein, M. Janz, B. Dorken, and S. Mathas. 2008. Aberrant expression of the Th2 cytokine IL-21 in Hodgkin lymphoma cells regulates STAT3 signaling and attracts Treg cells via regulation of MIP-3alpha. Blood 112:3339-3347.

Langstein, J., J. Michel, J. Fritsche, M. Kreutz, R. Andreesen, and H. Schwarz. 1998. CD137 (ILA/4-1BB), a member of the TNF receptor family, induces monocyte activation via bidirectional signaling. Journal of Immunology 160:2488-2494.

Langstein, J., J. Michel, and H. Schwarz. 1999. CD137 induces proliferation and endomitosis in monocytes. Blood 94:3161-3168.

Lee, K.H., C. Feig, V. Tchikov, R. Schickel, C. Hallas, S. Schutze, M.E. Peter, and A.C. Chan. 2006. The role of receptor internalization in CD95 signaling. The EMBO Journal 25:1009-1023.

117 Li, Q., J. Ai, Z. Song, J. Liu, and B. Shan. 2008. 4-1BB (CD137) ligand enhanced anti-tumor immune response against mouse forestomach carcinoma in vivo. Cellular & Molecular Immunology 5:379-384.

Lis, R., J. Capdet, P. Mirshahi, M. Lacroix-Triki, F. Dagonnet, C. Klein, M. Mirshahi, J.J. Fournie, A. Rafii, and M. Poupot. 2010. Oncologic trogocytosis with Hospicells induces the expression of N-cadherin by breast cancer cells. International Journal of Oncology 37:1453-1461.

Lisignoli, G., S. Toneguzzi, L. Cattini, C. Pozzi, and A. Facchini. 1998. Different expression pattern of cytokine receptors by human osteosarcoma cell lines. International Journal of Oncology 12:899-903.

Lynch, D.H. 2008. The promise of 4-1BB (CD137)-mediated immunomodulation and the immunotherapy of cancer. Immunological Reviews 222:277-286.

Mani, H., and E.S. Jaffe. 2009. Hodgkin lymphoma: an update on its biology with new insights into classification. Clinical Lymphoma & Myeloma 9:206-216.

Marshall, N.A., M.A. Vickers, and R.N. Barker. 2003. Regulatory T cells secreting IL-10 dominate the immune response to EBV latent membrane protein 1. Journal of Immunology 170:6183-6189.

Marshall, N.A., L.E. Christie, L.R. Munro, D.J. Culligan, P.W. Johnston, R.N. Barker, and M.A. Vickers. 2004. Immunosuppressive regulatory T cells are abundant in the reactive lymphocytes of Hodgkin lymphoma. Blood 103:1755-1762.

Martinet, O., C.M. Divino, Y. Zang, Y. Gan, J. Mandeli, S. Thung, P.Y. Pan, and S.H. Chen. 2002. T cell activation with systemic agonistic antibody versus local 4-1BB ligand gene delivery combined with interleukin-12 eradicate liver metastases of breast cancer. Gene Therapy 9:786-792.

Mathas, S., A. Lietz, I. Anagnostopoulos, F. Hummel, B. Wiesner, M. Janz, F. Jundt, B. Hirsch, K. Johrens-Leder, H.P. Vornlocher, K. Bommert, H. Stein, and B. Dorken. 2004. c-FLIP mediates resistance of Hodgkin/Reed-Sternberg cells to death receptor-induced apoptosis. The Journal of Experimental Medicine 199:1041-1052.

Metkar, S.S., K.N. Naresh, A.A. Redkar, C.S. Soman, S.H. Advani, and J.J. Nadkarni. 1999. Expression of Fas and Fas ligand in Hodgkin's disease. Leukemia & Lymphoma 33:521-530.

Michel, J., J. Langstein, F. Hofstadter, and H. Schwarz. 1998. A soluble form of CD137 (ILA/4-1BB), a member of the TNF receptor family, is released by activated lymphocytes and is detectable in sera of patients

118 with rheumatoid arthritis. European Journal of Immunology 28:290- 295.

Middendorp, S., Y. Xiao, J.Y. Song, V. Peperzak, P.H. Krijger, H. Jacobs, and J. Borst. 2009. Mice deficient for CD137 ligand are predisposed to develop germinal center-derived B-cell lymphoma. Blood 114:2280- 2289.

Mosselmans, R., A. Hepburn, J.E. Dumont, W. Fiers, and P. Galand. 1988. Endocytic pathway of recombinant murine tumor necrosis factor in L- 929 cells. Journal of Immunology 141:3096-3100.

Murillo, O., A. Arina, S. Hervas-Stubbs, A. Gupta, B. McCluskey, J. Dubrot, A. Palazon, A. Azpilikueta, M.C. Ochoa, C. Alfaro, S. Solano, J.L. Perez-Gracia, B.O. Oyajobi, and I. Melero. 2008. Therapeutic antitumor efficacy of anti-CD137 agonistic monoclonal antibody in mouse models of myeloma. Clinical cancer research : an official journal of the American Association for Cancer Research 14:6895- 6906.

Murillo, O., J. Dubrot, A. Palazon, A. Arina, A. Azpilikueta, C. Alfaro, S. Solano, M.C. Ochoa, C. Berasain, I. Gabari, J.L. Perez-Gracia, P. Berraondo, S. Hervas-Stubbs, and I. Melero. 2009. In vivo depletion of DC impairs the anti-tumor effect of agonistic anti-CD137 mAb. European Journal of Immunology 39:2424-2436.

Murray, P.G., J. Oates, G.M. Reynolds, J. Crocker, and L.S. Young. 1995. Expression of B7 (CD80) and CD40 antigens and the CD40 ligand in Hodgkin's disease is independent of latent Epstein-Barr virus infection. Clinical Molecular Pathology 48:M105-108.

Murray, P.G., C.M. Constandinou, J. Crocker, L.S. Young, and R.F. Ambinder. 1998. Analysis of major histocompatibility complex class I, TAP expression, and LMP2 epitope sequence in Epstein-Barr virus- positive Hodgkin's disease. Blood 92:2477-2483.

Niens, M., L. Visser, I.M. Nolte, G. van der Steege, A. Diepstra, P. Cordano, R.F. Jarrett, G.J. Te Meerman, S. Poppema, and A. van den Berg. 2008. Serum chemokine levels in Hodgkin lymphoma patients: highly increased levels of CCL17 and CCL22. British Journal of Haematology 140:527-536.

Oudejans, J.J., N.M. Jiwa, J.A. Kummer, A. Horstman, W. Vos, J.P. Baak, P.M. Kluin, P. van der Valk, J.M. Walboomers, and C.J. Meijer. 1996. Analysis of major histocompatibility complex class I expression on Reed-Sternberg cells in relation to the cytotoxic T-cell response in Epstein-Barr virus-positive and -negative Hodgkin's disease. Blood 87:3844-3851.

119 Pham, T., P. Mero, and J.W. Booth. 2011. Dynamics of macrophage trogocytosis of rituximab-coated B cells. PloS One 6:e14498.

Portis, T., P. Dyck, and R. Longnecker. 2003. Epstein-Barr Virus (EBV) LMP2A induces alterations in gene transcription similar to those observed in Reed-Sternberg cells of Hodgkin lymphoma. Blood 102:4166-4178.

Qureshi, O.S., Y. Zheng, K. Nakamura, K. Attridge, C. Manzotti, E.M. Schmidt, J. Baker, L.E. Jeffery, S. Kaur, Z. Briggs, T.Z. Hou, C.E. Futter, G. Anderson, L.S. Walker, and D.M. Sansom. 2011. Trans- endocytosis of CD80 and CD86: a molecular basis for the cell-extrinsic function of CTLA-4. Science 332:600-603.

Rabu, C., A. Quemener, Y. Jacques, K. Echasserieau, P. Vusio, and F. Lang. 2005. Production of recombinant human trimeric CD137L (4-1BBL). Cross-linking is essential to its T cell co-stimulation activity. The Journal of Biological Chemistry 280:41472-41481.

Rastelli, J., C. Homig-Holzel, J. Seagal, W. Muller, A.C. Hermann, K. Rajewsky, and U. Zimber-Strobl. 2008. LMP1 signaling can replace CD40 signaling in B cells in vivo and has unique features of inducing class-switch recombination to IgG1. Blood 111:1448-1455.

Re, D., A. Hofmann, J. Wolf, V. Diehl, and A. Staratschek-Jox. 2000. Cultivated H-RS cells are resistant to CD95L-mediated apoptosis despite expression of wild-type CD95. Experimental Hematology 28:31-35.

Re, D., R.K. Thomas, K. Behringer, and V. Diehl. 2005. From Hodgkin disease to Hodgkin lymphoma: biologic insights and therapeutic potential. Blood 105:4553-4560.

Rechavi, O., I. Goldstein, and Y. Kloog. 2009. Intercellular exchange of proteins: the immune cell habit of sharing. FEBS Letters 583:1792- 1799.

Reinehr, R., and D. Haussinger. 2008. CD95 ligation and intracellular membrane flow. The Biochemical Journal 413:e11-12.

Salih, H.R., S.G. Kosowski, V.F. Haluska, G.C. Starling, D.T. Loo, F. Lee, A.A. Aruffo, P.A. Trail, and P.A. Kiener. 2000. Constitutive expression of functional 4-1BB (CD137) ligand on carcinoma cells. Journal of Immunology 165:2903-2910.

Schmitz, R., J. Stanelle, M.L. Hansmann, and R. Kuppers. 2009. Pathogenesis of classical and lymphocyte-predominant Hodgkin lymphoma. Annual Review of Pathology 4:151-174.

120 Schreck, S., D. Friebel, M. Buettner, L. Distel, G. Grabenbauer, L.S. Young, and G. Niedobitek. 2009. Prognostic impact of tumour-infiltrating Th2 and regulatory T cells in classical Hodgkin lymphoma. Hematological Oncology 27:31-39.

Schwarz, H., J. Tuckwell, and M. Lotz. 1993. A receptor induced by lymphocyte activation (ILA): a new member of the human nerve- growth-factor/tumor-necrosis-factor receptor family. Gene 134:295- 298.

Schwarz, H., J. Valbracht, J. Tuckwell, J. von Kempis, and M. Lotz. 1995. ILA, the human 4-1BB homologue, is inducible in lymphoid and other cell lineages. Blood 85:1043-1052.

Schwarz, H., F.J. Blanco, J. von Kempis, J. Valbracht, and M. Lotz. 1996. ILA, a member of the human nerve growth factor/tumor necrosis factor receptor family, regulates T-lymphocyte proliferation and survival. Blood 87:2839-2845.

Shao, Z., F. Sun, D.R. Koh, and H. Schwarz. 2008. Characterisation of soluble murine CD137 and its association with systemic lupus. Molecular Immunology 45:3990-3999.

Shao, Z., and H. Schwarz. 2011. CD137 ligand, a member of the tumor necrosis factor family, regulates immune responses via reverse signal transduction. Journal of Leukocyte Biology 89:21-29.

Shao, Z., A. Schaffler, O. Hamer, J. Dickopf, A. Goetz, K. Landfried, M. Voelk, A. Kopp, H. Herfarth, T. Karrasch, T. Brunnler, L.K. Koh, C. Buechler, and H. Schwarz. 2012. Admission levels of soluble CD137 are increased in patients with acute pancreatitis and are associated with subsequent complications. Experimental and Molecular Pathology 92:1-6.

Skinnider, B.F., and T.W. Mak. 2002. The role of cytokines in classical Hodgkin lymphoma. Blood 99:4283-4297.

Su, C.C., H.H. Chiu, C.C. Chang, J.C. Chen, and S.M. Hsu. 2004. CD30 is involved in inhibition of T-cell proliferation by Hodgkin's Reed- Sternberg cells. Cancer Research 64:2148-2152.

Tabiasco, J., E. Espinosa, D. Hudrisier, E. Joly, J.J. Fournie, and A. Vercellone. 2002. Active trans-synaptic capture of membrane fragments by natural killer cells. European Journal of Immunology 32:1502-1508.

Takegawa, S., Z. Jin, T. Nakayama, T. Oyama, K. Hieshima, D. Nagakubo, A.K. Shirakawa, T. Tsuzuki, S. Nakamura, and O. Yoshie. 2008. Expression of CCL17 and CCL22 by latent membrane protein 1-

121 positive tumor cells in age-related Epstein-Barr virus-associated B-cell lymphoproliferative disorder. Cancer Science 99:296-302.

Tang, Q., D. Jiang, Z. Shao, J.M. Martinez Gomez, and H. Schwarz. 2011. Species difference of CD137 ligand signaling in human and murine monocytes. PloS One 6:e16129.

Tanijiri, T., T. Shimizu, K. Uehira, T. Yokoi, H. Amuro, H. Sugimoto, Y. Torii, K. Tajima, T. Ito, R. Amakawa, and S. Fukuhara. 2007. Hodgkin's reed-sternberg cell line (KM-H2) promotes a bidirectional differentiation of CD4+CD25+Foxp3+ T cells and CD4+ cytotoxic T lymphocytes from CD4+ naive T cells. Journal of Leukocyte Biology 82:576-584.

Teng, G., and F.N. Papavasiliou. 2007. Immunoglobulin somatic hypermutation. Annual Review of Genetics 41:107-120.

Thum, E., Z. Shao, and H. Schwarz. 2009. CD137, implications in immunity and potential for therapy. Frontiers in Bioscience : A Journal and Virtual Library 14:4173-4188.

Townsend, W., and D. Linch. 2012. Hodgkin's lymphoma in adults. Lancet 380:836-847.

Tzardi, M., C. Kouvidou, E. Papakonstantinou, G. Datseris, K. Darivianaki, E. Karidi, G. Eliopoulos, G. Delidis, D. Rontogianni, and P. Kanavaros. 1996. Major histocompatibility complex (MHC)-I and MHC-II expression in Hodgkin's disease in relation to the presence of Epstein- Barr Virus (EBV). Anticancer Research 16:827-831.

Uherova, P., R. Valdez, C.W. Ross, B. Schnitzer, and W.G. Finn. 2003. Nodular lymphocyte predominant Hodgkin lymphoma. An immunophenotypic reappraisal based on a single-institution experience. American Journal of Clinical Pathology 119:192-198.

Uzana, R., G. Eisenberg, Y. Sagi, S. Frankenburg, S. Merims, N. Amariglio, E. Yefenof, T. Peretz, A. Machlenkin, and M. Lotem. 2012. Trogocytosis is a gateway to characterize functional diversity in melanoma-specific CD8+ T cell clones. Journal of Immunology 188:632-640.

Vanherberghen, B., K. Andersson, L.M. Carlin, E.N. Nolte-'t Hoen, G.S. Williams, P. Hoglund, and D.M. Davis. 2004. Human and murine inhibitory natural killer cell receptors transfer from natural killer cells to target cells. Proceedings of the National Academy of Sciences of the United States of America 101:16873-16878.

Verbeke, C.S., U. Wenthe, R. Grobholz, and H. Zentgraf. 2001. Fas ligand expression in Hodgkin lymphoma. The American Journal of Surgical Pathology 25:388-394.

122 Vockerodt, M., S.L. Morgan, M. Kuo, W. Wei, M.B. Chukwuma, J.R. Arrand, D. Kube, J. Gordon, L.S. Young, C.B. Woodman, and P.G. Murray. 2008. The Epstein-Barr virus oncoprotein, latent membrane protein-1, reprograms germinal centre B cells towards a Hodgkin's Reed- Sternberg-like phenotype. The Journal of Pathology 216:83-92.

Vooijs, W.C., H.G. Otten, M. van Vliet, A.J. van Dijk, R.A. de Weger, M. de Boer, H. Bohlen, A. Bolognesi, L. Polito, and G.C. de Gast. 1997. B7- 1 (CD80) as target for immunotoxin therapy for Hodgkin's disease. British Journal of Cancer 76:1163-1169.

Wajant, H. 2002. The Fas signaling pathway: more than a paradigm. Science 296:1635-1636.

Wang, C., G.H. Lin, A.J. McPherson, and T.H. Watts. 2009. Immune regulation by 4-1BB and 4-1BBL: complexities and challenges. Immunological Reviews 229:192-215.

Wang, Q., P. Zhang, Q. Zhang, X. Wang, J. Li, C. Ma, W. Sun, and L. Zhang. 2008. Analysis of CD137 and CD137L expression in human primary tumor tissues. Croatian Medical Journal 49:192-200.

Watanabe, M., K. Nakano, T. Togano, M. Nakashima, M. Higashihara, M.E. Kadin, T. Watanabe, and R. Horie. 2011. Targeted repression of overexpressed CD30 downregulates NF-kappaB and ERK1/2 pathway in Hodgkin lymphoma cell lines. Oncology Research 19:463-469.

Wilcox, R.A., D.B. Flies, H. Wang, K. Tamada, A.J. Johnson, L.R. Pease, M. Rodriguez, Y. Guo, and L. Chen. 2002. Impaired infiltration of tumor- specific cytolytic T cells in the absence of interferon-gamma despite their normal maturation in lymphoid organs during CD137 monoclonal antibody therapy. Cancer Research 62:4413-4418.

Wilcox, R.A., K. Tamada, D.B. Flies, G. Zhu, A.I. Chapoval, B.R. Blazar, W.M. Kast, and L. Chen. 2004. Ligation of CD137 receptor prevents and reverses established anergy of CD8+ cytolytic T lymphocytes in vivo. Blood 103:177-184.

Yamamoto, R., M. Nishikori, T. Kitawaki, T. Sakai, M. Hishizawa, M. Tashima, T. Kondo, K. Ohmori, M. Kurata, T. Hayashi, and T. Uchiyama. 2008. PD-1-PD-1 ligand interaction contributes to immunosuppressive microenvironment of Hodgkin lymphoma. Blood 111:3220-3224.

Younes, A., A. Garg, and B.B. Aggarwal. 2003. Nuclear transcription factor- kappaB in Hodgkin's disease. Leukemia & Lymphoma 44:929-935.

Zhang, G.B., Q.M. Dong, J.Q. Hou, Y. Ge, S.G. Ju, B.F. Lu, and X.G. Zhang. 2007. Characterization and application of three novel monoclonal

123 antibodies against human 4-1BB: distinct epitopes of human 4-1BB on lung tumor cells and immune cells. Tissue Antigens 70:470-479.

Zhang, X., C.J. Voskens, M. Sallin, A. Maniar, C.L. Montes, Y. Zhang, W. Lin, G. Li, E. Burch, M. Tan, R. Hertzano, A.I. Chapoval, K. Tamada, B.R. Gastman, D.H. Schulze, and S.E. Strome. 2010. CD137 promotes proliferation and survival of human B cells. Journal of Immunology 184:787-795.

Zhang, Y., and B. Zhang. 2008. TRAIL resistance of breast cancer cells is associated with constitutive endocytosis of death receptors 4 and 5. Molecular Cancer Research : MCR 6:1861-1871.

Zhang, Y., T. Yoshida, and B. Zhang. 2009. TRAIL induces endocytosis of its death receptors in MDA-MB-231 breast cancer cells. Cancer Biology & therapy 8:917-922.

Zhao, S., H. Zhang, Y. Xing, and Y. Natkunam. 2012. CD137 Ligand Is Expressed in Primary and Secondary Lymphoid Follicles and in B-cell Lymphomas: Diagnostic and Therapeutic Implications. The American Journal of Surgical Pathology

Zhou, Z., S. Kim, J. Hurtado, Z.H. Lee, K.K. Kim, K.E. Pollok, and B.S. Kwon. 1995. Characterization of human homologue of 4-1BB and its ligand. Immunology Letters 45:67-73.

Zhu, B.Q., S.W. Ju, and Y.Q. Shu. 2009. CD137 enhances cytotoxicity of CD3(+)CD56(+) cells and their capacities to induce CD4(+) Th1 responses. Biomedicine & pharmacotherapy = Biomedecine & pharmacotherapie 63:509-516.

124 APPENDIX I ACRYLAMIDE GEL CASTING

For making of 10 ml of resolving gel:

Reagent (ml) 8% gel 10% gel 12% gel 15% gel

MilliQ water 4.6 4.0 3.3 2.3

30% Acrylamide mix (Bio-rad) 2.7 3.3 4.0 5.0

1.5M Tris (pH 8.8) 2.5 2.5 2.5 2.5

10% SDS (1st Base) 0.1 0.1 0.1 0.1

10% APS (Sigma-Aldrich) 0.1 0.1 0.1 0.1

TEMED (Sigma-Aldrich) 0.006 0.004 0.004 0.004

For making of 2 ml of stacking gel:

Reagent (ml)

MilliQ water 1.4

30% Acrylamide mix (Bio-rad) 0.33

1.5M Tris (pH 6.8) 0.25

10% SDS (1st Base) 0.02

10% APS (Sigma-Aldrich) 0.02

TEMED (Sigma-Aldrich) 0.002

125 APPENDIX II MEDIA AND BUFFERS

1. Cell culture

RPMI with 10% FBS

For preparing of 1 L:

Item Quantity Source

RPMI-1640 16.4 g Sigma-Aldrich

L-glutamine (200 mM) 10 ml Gibco

Sodium bicarbonate 2 g Sigma-Aldrich

FBS 100 ml Biowest

MilliQ water Top up to 1 L -

Adjusted to pH 7.0 and filtered with 0.22 µm filter unit (Nalgene).

RPMI, 10% FBS with Penicillin and streptomycin

For preparing of 100 ml:

Item Quantity Source

RPMI with 10% FBS 99 ml - Penicillin (10,0000 U) 1 ml Gibco /Streptomycin(10,000 µg/ml)

Selective medium (5 µµµg/ml Blasticidin)

For preparing of 100 ml:

Item Quantity Source

RPMI with 10% FBS 99.5 ml -

Blasticidin (1 mg/ml) 500 µl Sigma-Aldrich

126 Cell line freezing medium (20% FBS, 10% DMSO)

For preparing of 100 ml:

Item Quantity Source

RPMI with 10% FBS 77.8 ml -

FBS 12.2 ml Biowest

DMSO 10 ml Sigma-Aldrich

Primary cell freezing medium, Part A (40% FBS)

For preparing of 100 ml:

Item Quantity Source

RPMI with 10% FBS 66.7 ml -

FBS 33.3 ml Biowest

Primary cell freezing medium, Part B (40% FBS, 20% DMSO)

For preparing of 100 ml:

Item Quantity Source

RPMI with 10% FBS 44.4 ml -

FBS 35.6 ml Biowest

DMSO 20 ml Sigma-Aldrich

127 2. PBMC isolation and purification

10 x Phosphate buffer saline (PBS)

For preparing of 1 L:

Item Quantity Source

Sodium chloride (NaCl) 80 g Sigma-Aldrich Potassium chloride 2 g Sigma-Aldrich (KCl) Disodium hydrogen 14.4 g Sigma-Aldrich phosphate (Na 2HPO 4) Monopotassium 2.4 g Sigma-Aldrich phosphate (KH 2PO 4) MilliQ water Top up to 1 L -

Adjusted to pH 7.4.

1 x PBS

For preparing of 1 L:

Item Quantity Source

10 x PBS 100 ml -

MilliQ water Top up to 1 L -

Sterilized with autoclaving.

PBS with 2 mM EDTA

For preparing of 100 ml:

Item Quantity Source

1 x PBS 99.6 ml -

EDTA (0.5 M) 400 µl 1st Base

Sterilized with autoclaving.

128 RBC lysis buffer

For preparing of 1 L:

Item Quantity Source Ammonium chloride 8.29 g Sigma-Aldrich (NH 4Cl) Sodium bicarbonate 0.84 g Sigma-Aldrich (NaHCO 3) EDTA (0.5 M) 23 ml 1st Base

MilliQ water Top up to 1 L -

Filtered with 0.22 µm filter unit (Nalgene).

MASC buffer

For preparing of 1 L:

Item Quantity Source

10 x PBS 100 ml - Bovine serum albumin 5 g Science Werke (BSA) EDTA (0.5 M) 4 ml 1st Base

MilliQ water Top up to 1 L -

Stirred until dissolve and filtered with 0.22 µm filter unit (Nalgene).

129 3. Flow cytometry

FACS buffer

For preparing of 500 ml:

Item Quantity Source

10 x PBS 50 ml -

FBS 2.5 ml Biowest

20% Sodium azide 500 µl Sigma-Aldrich

MilliQ water Top up to 500 ml -

4. ELISA

PBS with 0.05% Tween-20 (PBST)

For preparing of 1 L:

Item Quantity Source

10 x PBS 100 ml -

Tween-20 500 µl NUMI store, NUS

MilliQ water Top up to 1 L -

Stirred until dissolve.

130 5. Co-Immunoprecipitation and Western blot

RIPA buffer

For preparing of 100 ml:

Item Quantity Source

Tris 606 mg 1st Base

NaCl 876 mg NUMI store, NUS

SDS (10% w/v) 1 ml 1st Base

Sodium deoxycholate 0.5 µg Sigma-Aldrich

Triton-X 1 ml Bio-Rad

PMSF (100mM) 1 ml* Sigma-Aldrich

MilliQ water Top up to 100 ml -

Stirred until dissolve and adjusted to pH 7.4. *PMSF was added prior to use.

Non-denaturing lysis buffer (PBS with 1 % Triton-X)

For preparing of 100 ml:

Item Quantity Source

1 x PBS 98 ml -

Triton-X 1 ml Bio-Rad

PMSF (100mM) 1 ml* Sigma-Aldrich

Stirred until dissolve. *PMSF was added prior to use.

131 5 x Tris-Glycine SDS Running Buffer

For preparing of 1 L:

Item Quantity Source

Tris 15.1 g 1st Base

Glycine 72 g 1st Base

SDS (10% w/v) 50 ml 1st Base

MilliQ water Top up to 1 L -

Stirred until dissolve.

1 x Tris-Glycine SDS Running Buffer

For preparing of 1 L:

Item Quantity Source

5 x Tris-Glycine Buffer 200 ml -

MilliQ water Top up to 1 L -

Transfer buffer

For preparing of 1 L:

Item Quantity Source

Tris 3.02 g 1st Base

Glycine 14.4 g 1st Base

Absolute Ethanol 200 ml NUMI store, NUS

MilliQ water Top up to 1 L -

Stirred until dissolve and adjusted to pH 8.0.

132 Western blot blocking buffer

For preparing of 100 ml:

Item Quantity Source

Skim milk powder 5 g Anlene

PBST Top up to 100 ml -

Stirred until dissolve.

6. Confocal microscopy

Fixing buffer (4% PFA in PBS)

For preparing of 100 ml:

Item Quantity Source

1 x PBS 100 ml -

Paraformaldehyde (PFA) 4 g Sigma-Aldrich

5 M sodium hydroxide (NaOH) was added dropwise until PFA was fully dissolved. Then adjusted to pH 7.2 with hydrochloric acid (HCl).

Permeabilization buffer (0.2% Triton-X in PBS)

For preparing of 100 ml:

Item Quantity Source

1 x PBS 100 ml -

Triton-X 200 µl Bio-Rad

Stirred until dissolve.

133 APPENDIX III ANTIBODIES LIST

Antigen Clone Company Usage Dilution Figure

CD137 4B4-1-PE BD Pharmigen Flow cytometry staining 1:10 4-6

BBK-2 Lab Vision CD137 neutralization 1:200 7-9, 14, 15, 17, 20

Flow cytometry staining 1:200 9, 20, 23-28

Co-immunoprecipitation 2 µg/1 mg of protein 16, 17

Confocal staining 1:100 18

JG1.6a Novus Biologicals CD137 stimulation 1:200 31, 32

CD137L 5F4-PE Biolegend Flow cytometry staining 1:10 9, 10, 20

Confocal staining 1:10 18

5F4 Biolegend CD137L neutralization 1:100 10, 11

134

41B436 Adipogen CD137L neutralization 1:200 10, 11

Western blot 1:2000 14-17, 29

Flow cytometry staining 1:200 23-26

C65-485 BD Pharmigen CD137L neutralization 1:100 10, 11

EPR1172Y Origene Confocal staining 33

CD11b ICRF44 eBioscience Flow cytometry staining 1:10 27

CD20 2H7 eBioscience Flow cytometry staining 1:10 28

135

135 APPENDIX IV CD137L DOWNREGULATION ON

PRIMARY MONOCYTES AND B CELLS

Experiment 1

A Monocytes

HDLM2 KMH2 70% 70% 41% 55%

CD137L

B B cells

HDLM2 KMH2 14% 14% 13% 15%

CD137L

PBMC alone Co-culture

CD137L expression on monocytes and B cells following co-culture with HRS cell lines 5 x 10 5 cells/ml of KM-H2 or HDLM-2 cells were co-cultured with 10 6 of PBMC suboptimally activated with 1 ng/ml of anti-CD3 (Clone Okt3). 24 h later, the cells were harvested and stained for CD137L (clone 41B436), CD11b (for monocytes) and CD20 (for B cells), and analyzed with flow cytometry. The histograms shown are gated on (A) CD11b + or (B) CD20 + population only. Open histograms show the PBMC alone condition, while grey histogram show PBMC co-culture with HRS cells condition. The percentages of positively stained cells are shown in the histograms.

136 Experiment 2

A Monocytes

PBMC alone

17%

HDLM2 KMH2

12% 13%

CD137L

B B cells

PBMC alone

17%

HDLM2 KMH2

19% 39%

CD137L

CD137L expression on monocytes and B cells following co-culture with HRS cell lines 5 x 105 cells/ml of KM-H2 or HDLM-2 cells were co-cultured with 10 6 of PBMC suboptimally activated with 1 ng/ml of anti-CD3 (Clone Okt3). 24 h later, the cells were harvested and stained for CD137L (clone 41B436), CD11b (for monocytes) and CD20 (for B cells), and analyzed with flow cytometry. The histograms shown are gated on (A) CD11b + or (B) CD20 + population only. Open histograms show the isotype staining of CD137L. The percentages of positively stained cells are shown in the histograms.

137 Experiment 3

A Monocytes PBMC alone

HDLM2 KMH2

16% 25%

CD137L

B B cells

PBMC alone

HDLM2 KMH2

8% 25%

CD137L

CD137L expression on monocytes and B cells following co-culture with HRS cell lines 5 x 10 5 cells/ml of KM-H2 or HDLM-2 cells were co-cultured with 10 6 of PBMC suboptimally activated with 1 ng/ml of anti-CD3 (Clone Okt3). 24 h later, the cells were harvested and stained for CD137L (clone 41B436), CD11b (for monocytes) and CD20 (for B cells), and analyzed with flow cytometry. The histograms shown are gated on (A) CD11b + or (B) CD20 + population only. Open histograms show the isotype staining of CD137L. The percentages of positively stained cells are shown in the histograms.

138