Molecular Mechanisms of Traf3- Deficient B Lymphomagenesis

MOLECULAR MECHANISMS OF TRAF3- DEFICIENT B LYMPHOMAGENESIS

by Shanique Katrice Elaine Edwards

A Dissertation submitted to the

Graduate School-New Brunswick Rutgers, the State University of New Jersey

and

The Graduate School of Biomedical Sciences

in partial fulfillment of the requirements

for the degree of

Doctor of Philosophy

Graduate Program in Cell and Developmental Biology

written under the direction of

Ping Xie

and approved by

______

New Brunswick, New Jersey

May 2015

ABSTRACT OF THE DISSERTATION

Molecular Mechanisms of TRAF3-deficient B Lymphomagenesis

By SHANIQUE KATRICE ELAINE EDWARDS

Dissertation Director:

Ping Xie

B cell neoplasms, including leukemias, lymphomas and myelomas, are a common type of cancer, but they remain difficult to treat. This outlines a need for a better understanding of the mechanisms by which malignant transformation occurs, in order to come up with better therapeutic strategies. Recently, TRAF3 has been shown to act as a tumor suppressor, as mice with this gene specifically deleted in B cells develop B lymphomas. TRAF3 deletion causes prolonged B cell survival, allowing other secondary oncogenic alterations to occur. To elucidate these secondary alterations, we performed microarray analyses to identify genes which are differentially expressed in mouse B lymphomas. Two such genes that I have investigated in my thesis research are MCC and Sox5, both of which are significantly upregulated specifically in malignant B cells. MCC, mutated in colorectal cancer, has been previously identified as a tumor suppressor in colorectal cancer. We discovered that in malignant B cells, MCC acts as an

oncogene to promote B cell survival and proliferation by modulating the signaling network centered at PARP1 and PHB1/2. The Sox5 gene encodes a transcription factor. Interestingly, we found that the Sox5 expressed in TRAF3-/- mouse B lymphomas represents a novel isoform of Sox 5, Sox5-BLM, which regulates malignant B cell proliferation by affecting the expression of p27 and -catenin.

These two genes have the potential to be used as diagnostic markers and therapeutic targets in B cell malignancies. Based on the understanding of TRAF3 signaling mechanisms, we also conducted translational studies using anti-cancer drugs to manipulate TRAF3 downstream signaling components. We have tested drugs targeting NF-B2 and PKC, and found that oridonin and AD198 exhibit potent anti-tumor activities on B cell neoplasms with TRAF3 deletions or mutations.

Furthermore, oridonin or AD 198 drastically potentiated the anti-cancer effects of bortezomib, an effective clinical drug for multiple myeloma. Taken together, our studies have gained new insights into the mechanisms underlying TRAF3 inactivation-initiated B lymphomagenesis, and have discovered novel therapeutic targets for B cell neoplasms. Our findings also provide a rationale for clinical evaluation of several drugs or drug combinations in the treatment of B cell malignancies.

iii

DEDICATION & ACKNOWLEDGEMENTS

DEDICATION

Sola Gratia, Soli Deo Gloria

To my family, especially Dr. Daddy, Mommy, Mika, Shaq and Sadé: you have been my support through everything! There aren’t enough words to say how much you all mean to me. I hope that I’m at least half as great a woman as you have believed me to be. Love wunna bad!

To my friends who have been there through it all, you are the best! Shamika, my frister, your realness has kept me from quitting many a time. Karlyn, thank you for your encouragement and commiseration through dissertation writing, and for your advice and our conversations. Ayodele Grace, your quiet but constant support has been everything to me. Javonni, your love and friendship have kept me sane throughout grad school. Ann, your excitement and energy have helped me to push through the tough days.

I am also very thankful for all those who have supported and encouraged me over these five years: James Street, GCF, NBUMC, LGC, and my IC fam, if you hadn’t believed in me, I might not have had the courage to believe in myself.

ACKNOWLEDGEMENTS

MY ADVISER & COMMITTEE

Many thanks to my dissertation adviser Dr. Ping Xie for her guidance and support over these years. I have learned so much from you about being a scientist and researcher. Thank you Dr. Lori Covey for your mentorship, support, guidance, encouragement, advice and listening ear. Thank you to Drs. Lisa Denzin, Bonnie Firestein & Ron Hart for pushing me to work harder, think more and be a better scientist.

XIE LAB

Thank you to everyone for the work that you have done to help me through this process, especially Yan, Carissa, Sukhdeep, Jackie, Anand, Ben and Punit. Without your help with experiments, I wouldn’t be here! Thank you to Almin, my lab mate for being there to bounce ideas off of, and to keep me on track.

DIVISION OF LIFE SCIENCES & MOLECULAR BIOSCIENCES

Thank you to the Division of Life Sciences for three wonderful years of financial support through the Teaching Assistant program! It has been such a pleasure working with Jana and everyone! Thanks also to the Molecular Biosciences Graduate Program for my first year, and summer support.

TABLE OF CONTENTS

Abstract of the dissertation ------ii

Dedication & Acknowledgements ------iv

Dedication ------iv

Acknowledgements ------iv

Table of Contents ------v

List of Figures ------x

List of Tables ------xii

Introduction ------1

B Cell Cancers: Prevalence & Unique Challenges ------1

TRAF Proteins and Immune Signaling ------2

TRAF3, a Novel Tumor Suppressor ------4

TRAF3 Signaling Pathways in B lymphocytes ------5

My Dissertation Research ------10

Chapter 1 Mutated in colorectal cancer (MCC) is a novel oncogene in B lymphocytes ------13

Abstract ------13

1.1 Background------14

1.2 Materials and Methods ------16

1.2.1 Mice, cell lines, and reagents ------16

1.2.2 Transcriptome microarray analysis ------17

1.2.3 Taqman assays of the transcript expression of identified genes ------18

1.2.4 Splenic B cell purification and stimulation ------18

1.2.5 Taqman copy number assay of the mouse MCC gene ------19

1.2.6 Chromatin immunoprecipitation assay ------19

1.2.7 Cloning of the full-length cDNA of the MCC gene from TRAF3-/- mouse B lymphomas and human MM cell lines ------20

1.2.8 Generation of lentiviral MCC expression and shRNA vectors ------21

1.2.9 Lentiviral packaging and transduction of human MM cells ------22

1.2.10 Growth curve determination, annexin V staining of apoptotic cells, cell cycle distribution and cell proliferation analyses ------22

1.2.11 Total protein lysates, fractionation of cytosol, mitochondria and microsomes (rich in ER), and immunoblot analysis ------23

1.2.12 Co-immunoprecipitation assay in whole cell lysates ------24

1.2.13 Co-immunoprecipitation assay in mitochondrial lysates------25

1.2.14 Mass spectrometry based-sequencing ------25

1.2.15 Statistics ------27

1.3 Results ------27

1.3.1 Transcriptome microarray analysis of TRAF3 /- mouse B lymphomas ------27

1.3.2 Striking up-regulation of MCC in TRAF3 /- mouse B lymphomas but not in premalignant TRAF3 /- B lymphocytes ------30

1.3.3 Aberrant up-regulation of MCC in human patient-derived MM cell lines with TRAF3 deletions or relevant mutations ------33

1.3.4 Subcellular localization of MCC in human MM cells and regulation of MCC during ER stress responses ------35

1.3.5 Lentiviral shRNA-mediated knockdown of MCC induced apoptosis and inhibited proliferation in human MM cells ------35

1.3.6 MCC signaling pathways in human MM cells ------39

1.3.7 MCC interacting proteins in human MM cells ------42

1.4 Discussion ------48

1.5 Conclusions ------51

Chapter 2 Expression and function of a novel isoform of Sox5 in malignant B cells ------53

Abstract ------53

2.1 Background------53

2.2 Materials and methods ------55

2.2.1 Mice and cell lines ------55

2.2.2 Antibodies and reagents ------56

2.2.3 Taqman assay of Sox5 mRNA expression ------57

2.2.4 Protein extraction and immunoblot analysis ------57

2.2.5 Cloning of the full-length cDNA of the Sox5 gene from TRAF3−/− mouse B lymphomas ------57

2.2.6 Generation of lentiviral Sox5 expression vectors ------58

2.2.7 Lentiviral packaging and transduction of human multiple myeloma cells ------58

2.2.8 Immunostaining and confocal imaging ------59

2.2.9 Statistics ------59

2.3 Results ------59

2.3.1 Striking up-regulation of Sox5 in TRAF3−/− B lymphomas ------59

2.3.2 A novel isoform of Sox5 was expressed in TRAF3−/− B lymphomas ------61

2.3.3 Overexpression of Sox5-BLM inhibited cell cycle progression in transduced human multiple myeloma cells ------68

2.3.4 Sox5-BLM was localized in the nucleus of TRAF3−/− mouse B lymphomas and transduced human multiple myeloma cells ------68

2.3.5 Overexpression of Sox5-BLM up-regulated the protein levels of p27 and β-catenin in transduced human multiple myeloma cells ------71

2.4 Discussion ------72

2.5 Conclusions ------77

Chapter 3. N-Benzyladriamycin-14-valerate (AD 198) exhibits potent anti-tumor activity on TRAF3-deficient mouse B lymphoma and human multiple myeloma ------78

Abstract ------78

3.1 Background------79

3.2 Materials and Methods ------80

3.2.1 Mice ------80

3.2.2 Cell lines and cell culture ------81

3.2.3 Antibodies and reagents ------82

3.2.4 MTT assay ------83

3.2.5 Measurement of apoptosis ------83

3.2.6 Lymphoma transplantation and drug treatment of NOD SCID mice ------84

3.2.7 Total protein lysates ------84

vii

3.2.8 Cytosolic and nuclear extracts ------85

3.2.9 Fractionation of cytosol, nuclei and membranes ------85

3.2.10 Immunoblot analysis ------86

3.2.11 Taqman assay of c-Myc mRNA expression ------86

3.2.12 Generation of lentiviral c-Myc expression vectors ------87

3.2.13 Lentiviral packaging and transduction of human MM cells ------88

3.2.14 Statistics ------88

3.3 Results ------89

3.3.1 Differential effects of AD 198 and PEP005 on TRAF3-/- tumor B cells ------89

3.3.2 AD 198 but not PEP005 induced apoptosis in TRAF3-/- tumor B cells ------93

3.3.3 AD 198 exhibited potent in vivo anti-tumor activity on TRAF3-/- mouse B lymphomas - 97

3.3.4 AD 198 induced PKC cleavage, while PEP005 induced PKC translocation in TRAF3-/- tumor B cells ------98

3.3.5 Differential effects of AD 198 and PEP005 on ERK, p38 and JNK activation in TRAF3-/- tumor B cells ------104

3.3.6 AD 198 rapidly suppressed c-Myc expression in TRAF3-/- tumor B cells ------107

3.3.7 AD 198 exhibited potent tumoricidal activity and rapidly suppressed c-Myc expression in TRAF3-sufficient B lymphoma cell lines------109

3.3.8 Lentiviral vector-mediated constitutive expression of c-Myc conferred partial resistance to the anti-tumor effects of AD 198 in human MM cell lines ------112

3.4 Discussion ------114

3.5 Conclusions ------119

Chapter 4 Signaling mechanisms of bortezomib in TRAF3-deficient mouse B lymphoma and human multiple myeloma cells------120

Abstract ------120

4.1 Background------121

4.2 Materials and Methods ------123

4.2.1 Cell lines, antibodies, and reagents ------123

4.2.2 MTT assay, cell cycle analyses, and caspase cleavage assays ------123

4.2.3 Protein extraction and immunoblot analysis ------123

viii

4.2.4 Statistics ------124

4.3 Results ------124

4.3.1 Potent tumoricidal activity of bortezomib on TRAF3-/- mouse B lymphoma and human MM cell lines ------124

4.3.2 Bortezomib did not inhibit canonical or non-canonical NF-B pathway ------129

4.3.3 Bortezomib induced accumulation of p21 and c-Myc ------131

4.3.4 Bortezomib induced up-regulation of Noxa and cleavage of Mcl1 ------133

4.3.5 Combined effects of bortezomib and oridonin or AD 198------137

4.4 Discussion ------140

4.5 Conclusion ------143

Summary and Significance ------144

Appendix 1: Tables ------150

Appendix 2: List of Abbreviations used ------182

Acknowledgement of Previous Publications ------183

Bibliography ------184

LIST OF FIGURES

Figure 1. Representative genes differentially expressed in TRAF3-/- mouse B lymphomas identified by the transcriptome microarray analysis...... 28

Figure 2. Striking up-regulation of MCC in TRAF3 /- mouse B lymphomas but not in premalignant TRAF3 /- B lymphocytes...... 31

Figure 3. Aberrant expression and subcellular localization of MCC in human MM cells...... 33

Figure 4. Lentiviral shRNA vector-mediated knockdown of MCC induced apoptosis in human MM cells...... 37

Figure 5. Knockdown of MCC induced apoptosis and inhibited proliferation in human MM cells. 38

Figure 6. MCC signaling pathways in human MM cells...... 41

Figure 7. The MCC-interactome in human MM cells identified by affinity purification followed by LC-MS/MS...... 44

Figure 8. The MCC isoform 2 was identified as an MCC-interacting protein by affinity purification and LC-MS/MS...... 46

Figure 9. PARP1 and PHB2 are two hubs of the MCC interaction network in human MM cells. .. 47

Figure 10. Up-regulation of Sox5 expression in TRAF3−/− mouse B lymphomas...... 60

Figure 11. A novel isoform of Sox5 was expressed in TRAF3−/− mouse B lymphomas...... 63

Figure 12. Distinct isoforms of the mouse Sox5 gene generated by alternative splicing...... 64

Figure 13. SOX5 is not expressed in human multiple myeloma cell lines 8226 and LP1...... 65

Figure 14. Overexpression of Sox5-BLM inhibited cell cycle progression in human multiple myeloma cells ...... 66

Figure 15. Predominant nuclear localization and signaling pathways of Sox5-BLM in malignant B cells ...... 69

Figure 16. Sox5 is primarily localized in the nucleus in malignant B cells...... 71

Figure 17. AD 198 exhibited potent anti-proliferative/survival-inhibitory effects on TRAF3-/- mouse B lymphoma and human MM cells...... 90

Figure 18. PEP005 exhibited differential effects on TRAF3-/- mouse B lymphoma and human MM cells...... 92

Figure 19. AD 198 induced apoptosis in TRAF3-/- tumor B cells...... 94

Figure 20. PEP005 did not induce apoptosis in TRAF3-/- tumor B cells...... 96

Figure 21. AD 198 and oridonin exhibited potent anti-tumor activity on transplanted TRAF3-/- B lymphomas in NOD SCID mice...... 97

Figure 22. AD 198 did not affect PKCδ nuclear translocation or NF-κB activation...... 100

Figure 23. Effects of PEP005 on the cytosolic and nuclear levels of PKCδ, NF-κB1 and NF-κB2 subunits, and c-Myc...... 101

Figure 24. Differential effects of AD 198 and PEP005 on PKCδ subcellular translocation and cleavage...... 102

Figure 25. AD 198 inhibited phosphorylation of ERK, p38 & JNK in TRAF3-/- tumor B cells. .... 106

Figure 26. AD 198 suppressed c-Myc expression in TRAF3-/- tumor B cells...... 108

Figure 27. AD 198 exhibited potent anti-tumor activity and suppressed c-Myc expression in TRAF3-sufficient mouse and human B lymphoma cell lines...... 110

Figure 28. Constitutive c-Myc expression counteracted the effects of AD 198 on the proliferation and survival of human MM cells...... 113

Figure 29. Bortezomib induced apoptosis in TRAF3 /- mouse B lymphoma and human MM cells...... 125

Figure 30. Bortezomib induced apoptosis in TRAF3 /- mouse B lymphoma and human MM cells...... 127

Figure 31. Bortezomib induced cleavage of caspases but did not affect NF-κB activation in TRAF3-/- malignant B cells...... 128

Figure 32. Bortezomib did not inhibit NF-κB pathways in TRAF3-/- malignant B cells...... 130

Figure 33. Bortezomib increased the accumulation of p21 and c-Myc in TRAF3-/- malignant B cells...... 132

Figure 34. Bortezomib increased the accumulation of p21 in TRAF3-/- malignant B cells...... 133

Figure 35. Bortezomib induced up-regulation of Noxa and cleavage of Mc1-1 in TRAF3-/- malignant B cells...... 135

Figure 36. Effects of bortezomib on the protein levels of cell survival regulators in TRAF3-/- malignant B cells...... 137

Figure 37. Oridonin or AD 198 potently augmented the cytotoxic activity of bortezomib on TRAF3 /- malignant B cells...... 139

Figure 38. MCC, a Novel Oncogene in B Lymphocytes...... 146

Figure 39. Complex mechanisms of TRAF3 inactivation-mediated B lymphomagenesis...... 148

LIST OF TABLES

Table 1. Genes differentially expressed in TRAF3-/- mouse splenic B lymphomas identified by the microarray analysis ...... 150

Table 2. The MCC-interactome in human MM cells identified by affinity purification followed by LC-MS/MS ...... 166

Table 3. Sox5 primers used in this study ...... 179

Table 4. Results of PCR cloning of the Sox5 cDNA from TRAF3-/- mouse B lymphomas ...... 180

Table 5. Sequence overlap extension PCRs to obtain full-length coding sequence of the Sox5- BLM cDNA ...... 181

Table 6. Results of PCR cloning of Sox5-V3 and S-Sox5 transcripts from TRAF3-/- mouse B lymphomas ...... 181

xii

INTRODUCTION

B CELL CANCERS: PREVALENCE & UNIQUE CHALLENGES

B cell neoplasms comprise a significant proportion of cancer diagnoses.

9.4% of new cancer cases in 2014 were leukemias, lymphomas or myelomas derived from B lymphocytes 1. These include acute B lymphoblastic leukemia, chronic B lymphocytic leukemia, lymphoma, multiple myeloma and Waldenström’s macroglobulinemia. Not only are these types of malignancies common, but they also contribute to 9.4% of all cancer deaths in the United States1-3. Despite their prevalence and mortality rate, B cell malignancies remain difficult to treat, and largely incurable. This highlights the need for further basic research into these types of malignancies to better understand the mechanisms of malignant transformation. This knowledge would identify new potential therapeutic targets, provide novel opportunities for the development of cancer treatments, and pinpoint new markers of B cell malignancies for early detection and treatment stratification.

The unique development and activation of B cells makes them particularly susceptible to oncogenic alteration4-7. Double stranded breaks in DNA and gene rearrangements are intrinsic parts of V(D)J recombination, somatic hypermutation

(SHM) and class switch recombination (CSR), but these also increase genomic instability and the likelihood of oncogenic alterations 4,6,8-11. Aberrations in any of these processes could result in unstable or oncogenic rearrangement in either Ig

genes or oncogenes such as Myc and proteins in the Bcl-2 family4,7-9,11-14. This contributes to a wide range of etiologies for B cell malignancies.

Accumulating evidence indicates that deletions/mutations of tumor suppressive proteins or up-regulation/activation of oncogenic proteins lead to prolonged survival or increased proliferation of B cells 15,16, resulting in B cell malignant transformation and lymphomagenesis. Identifying such proteins and elucidating their roles in B cell malignancies are important for the understanding of molecular mechanisms of B lymphomagenesis. These studies will also identify potential new therapeutic targets for treatment as well as new markers of malignancy for earlier detection of B cell neoplasms. Interestingly, recent studies identify members of the tumor necrosis factor receptor (TNF-R)-associated factor

(TRAF) family as candidate tumor suppressive or oncogenic proteins, whose deletions/mutations or up-regulation/aberrant expression of TRAF proteins have been documented in various B cell malignancies 17,18.

TRAF PROTEINS AND IMMUNE SIGNALING

The TRAF family of intracellular proteins were originally identified as signaling adaptors of the TNF-R superfamily19. TNF receptors are membrane bound proteins with no kinase activity in their cytoplasmic tail, so it was deduced that they must interact with other proteins with enzymatic activities to transduce signals19. The TRAFs were thus identified from yeast 2-hybrid analyses as adapter proteins which bind directly to the cytoplasmic tail of TNFR family proteins19. As such, TRAF proteins act as scaffolding molecules for kinases, cytokine receptors

and regulators of signal transduction, linking cell surface receptors with downstream signaling cascades20. Downstream pathways affected by TRAF proteins include NF-B, JNK and p3818.

Each member of the TRAF family contains a characteristic stretch of amino acid at the C-terminus, which has been deemed the TRAF domain19. This domain is subdivided into two further sub regions, TRAF-N (N-terminal side) and TRAF-C

(C-terminal side), the former of which is has some variability in comparison to the highly conserved TRAF-C sub region19,20. The TRAF domain enables TRAF proteins to homo- or hetero-dimerize in their association with each other and their respective receptors or signaling proteins20. Therefore, one important role of

TRAFs is to serve as adaptor proteins in the assembly of receptor-associated signaling complexes, linking upstream receptors to downstream effector enzymes.

All TRAFs also have a zinc finger motif, with one (TRAF1, TRAF7), five (TRAF2,

TRAF3, TRAF5, TRAF6) or seven (TRAF4) repeats, and with the exception of

TRAF1, all TRAFs also have an N-terminal RING finger19-21. In TRAF2, TRAF3,

TRAF5, and TRAF6, the RING finger has E3-ubiquitin ligase activity16,20. Thus,

TRAFs function as both adaptor proteins and E3 ubiquitin ligases to regulate signaling.

Many TRAF family proteins have been found to be altered in B cell malignancies. Two single nucleotide polymorphisms (SNPs) of TRAF1 are associated with chronic lymphocytic leukemia and small lymphocytic leukemia22.

Deletion or inactivation of TRAF2 has been implicated in multiple myeloma, mantle

cell lymphoma, diffuse large B cell lymphoma and B cell chronic lymphocytic leukemia23-29, and constitutive phosphorylation at S55 has been observed in

Hodgkin lymphoma30. Down-regulation of TRAF5 occurs in splenic marginal zone lymphoma (SMZL)31, and inactivating mutations of TRAF5 are linked to diffuse large B cell lymphoma and B cell chronic lymphocytic leukemia26,29. Of the TRAF proteins, TRAF3 is the most frequently and ubiquitously altered in B cell malignancies.

TRAF3, A NOVEL TUMOR SUPPRESSOR

TRAF3, the third member of the TRAF family to be discovered, was first identified as a cytoplasmic adaptor protein which binds to CD40 and LMP132-34, and is critical in both innate and adaptive immunity. In macrophages and dendritic cells, TRAF3 has been shown to play an important role in type I interferon induction in response to TLRs 3, 4, 7/8 and 9 35,36 by regulating NF-kB and IRF3/7 activation, acting downstream of TRIF or MyD88 37. TRAF3 also plays a role in host defense in vivo, as TRAF3 deficiency in humans has been described as a genetic predisposition to herpes simplex virus-1 (HSV-1) encephalitis, an infection of the human central nervous system38. These show the overall importance of TRAF3 in cell signaling and host defense.

TRAF3 is an essential part of physiological function, as TRAF3 null mice show early (but post-natal) lethality39. The early lethality of TRAF3 null mice made studying TRAF3 function very difficult until the creation of mouse models with conditional deletions of the traf3 gene. Mice with TRAF3 specifically deleted in B

cells (B-TRAF3-/- mice) showed enlarged spleens and lymph nodes, due to expanded B cell compartments In addition, despite TRAF3 being first identified by its interaction with CD40, TRAF3 is actually not required for CD40-mediated responses, and B-TRAF3 deficient mice do not display any altered CD40 signaling in vivo 16. 16,40. Furthermore, in B cells, TRAF3 is the only member of the TRAF family that has been shown to interact directly with BAFF-R, a receptor for a critical

B cell survival factor, B cell activating factor (BAFF) 41.

Recently, it has been discovered that TRAF3 plays a tumor suppressive role in B lymphocytes, given that specific deletion of TRAF3 in B cells of mice results prolonged cell survival, and subsequently the spontaneous development of splenic marginal zone and B1 lymphomas 15. This conclusion is corroborated by evidence from human patients, as deletions and mutations of the TRAF3 gene have been frequently detected in human non-Hodgkin lymphoma, including splenic marginal zone lymphoma, B cell chronic lymphocytic leukemia, mantle cell lymphoma, and diffuse large B cell lymphoma, as well as multiple myeloma (MM) and

Waldenström's macroglobulinemia 23,24,42,43.

TRAF3 SIGNALING PATHWAYS IN B LYMPHOCYTES

Since its discovery, TRAF3 has been identified as playing a role in two distinct pathways in B cells: NF-B signaling and PKCδ signaling (Fig. 1)16. TRAF3 deletion in B cells leads to constitutive NF-B2 activation resulting in prolonged cell survival 15. This is a similar observation to what is seen in constitutive BAFF-

R signaling, but adding BAFF to TRAF3 deficient B cells in vitro does not affect B cell survival16. More specifically, tumors from B-TRAF3-/- mice have higher nuclear levels of p52 and RelB, two subunits of NF-kB2 15. In fact, TRAF3-/- B cells treated with oridonin, an inhibitor of NF-kB, displayed a drastic decrease in viability, indicating that constitutive NF-kB2 activation drives the enhanced cell survival of

TRAF3-/- B cells 15. This is in contrast to TRAF3-/- tumor B cells treated with BMS-

345541, an inhibitor of IKK2, which did not show any significant tumoricidal activity, nor did other drugs used to clinically treat B lymphomas or leukemias: cyclophosphamide, doxorubicin, all trans-retinoic acid, and vincristine 15.

The proximal signaling events of how TRAF3 inhibits the activation of noncanonical NF-B, NF-B2, have been explicitly elucidated in the literature (Fig. i). It was found that in the absence of stimulation, TRAF3 and TRAF2 assemble a regulatory complex with cIAP1/2 and NIK. Assembly of this complex requires direct binding between TRAF3 and NIK, and also direct association between TRAF2 and cIAP1/2. TRAF3 and TRAF2 heteromeric interaction bring all 4 proteins into the complex. In this complex, the E3 ubiquitin ligase cIAP1/2 induces K48-linked polyubiquitination of NIK and targets NIK for proteasome-mediated degradation, thereby inhibiting NF-B2 activation. Therefore, in the absence of stimulation,

TRAF3 promotes cellular apoptosis in resting B cells by targeting NIK for degradation and inhibiting NF-B2 activation (Fig. iA). In response to stimulation by the B cell survival factor BAFF or the co-stimulatory ligand CD154, trimerized

BAFF-R or CD40 recruits the translocation of cytoplasmic TRAF3 and TRAF2 to

the receptor signaling complex in sphingolipid-enriched membrane rafts (Fig. iB).

This releases NIK from the TRAF3-TRAF2-cIAP1/2 complex, allowing NIK accumulation, which induces the phosphorylation of IKK and then the processing of NF-B2 from the inactive precursor p100 into the active p52. Processed p52 and RelB dimers subsequently translocate into the nucleus to induce the transcription of anti-apoptotic proteins of the Bcl2 family, including Bcl2, Bcl-xL and

Mcl1, thereby promoting B cell survival (Fig. iB). Similar to BAFF or CD40 stimulation, deletion of TRAF3 from B cells, caused by either biallelic deletions or inactivating mutations of the Traf3 gene, also releases NIK from the TRAF2- cIAP1/2 complex and thus allows NIK accumulation, leading to constitutive NF-

B2 activation and BAFF-independent survival of B cells (Fig. iC).

TRAF3-/- B cells also show decreased nuclear levels of PKCδ16, but the mechanism by which this decrease occurs has not been elucidated. Like B-

TRAF3-/- mice, PKCδ-/- mice also show expansion of the B cell compartment44,45 and BAFF also reduces nuclear PKCδ levels in splenic B cells18. In addition PKCδ, has been recently investigated as a novel therapeutic target because of its integral role in signal transduction pathways 46.

Figure i. TRAF3 signaling pathways in resting B cells and TRAF3-deficient B cells. (A) TRAF3 promotes apoptosis in resting B cells. In the absence of stimulation, TRAF3 inhibits NF- B2 activation while facilitating PKC nuclear translocation to promote B cell apoptosis. TRAF3 and TRAF2 constitutively form a complex with cIAP1/2 and NIK, targeting NIK for K48-linked polyubiquitination and degradation, thereby inhibiting NF-B2 activation in B cells. How TRAF3 facilitates PKC nuclear translocation remains to be determined (depicted by binding to an unknown protein or complex Factor Y in the figure). (B) BAFF stimulates B cell survival. Upon ligand engagement, trimerized BAFF-R recruits TRAF3-TRAF2-cIAP1/2 to membrane rafts, thus

allowing NIK accumulation and NF-B2 activation. Meanwhile, Factor Y is released from TRAF3 and may sequester PKC in the cytosol. NF-B2 activation together with reduced nuclear levels of PKC functions to induce the expression of anti-apoptotic proteins and thus B cell survival. (C) TRAF3 deficiency causes BAFF-independent B cell survival. Similar to BAFF stimulation, deletion of TRAF3 from B cells (caused by either biallelic deletion or inactivating mutations of the Traf3 gene) also releases NIK from the TRAF2-cIAP1/2 complex, causing constitutive NF-B2 activation. In the absence of TRAF3, Factor Y may also sequester PKC in the cytosol. Constitutive NF-B2 activation together with reduced nuclear levels of PKC leads to BAFF-independent, prolonged survival of TRAF3-/- B cells.

MY DISSERTATION RESEARCH

Although B-TRAF3-/- mice develop lymphomas, there is a latency period of about nine months before tumors develop 15, and TRAF3-/- B cells purified from young B-TRAF3-/- mice exhibit prolonged survival but do not autonomously proliferate. This indicates that TRAF3 inactivation and its downstream signaling pathways are not sufficient and that additional oncogenic alterations are required for B lymphomagenesis. To elucidate such alterations, we performed microarray analyses to identify genes with significant changes in expression in B lymphomas as compared with littermate control mice. Based on a stringent prioritization scheme, we selected two of the genes identified by microarray analyses, MCC and

Sox5, for further study in my thesis research.

MCC is mutated in colorectal cancer, which is a gene that encodes a protein with two isoforms. MCC has been shown to act as a tumor suppressor in colorectal cancer47-49, and a regulator of the cell cycle progression50,51, but has not been studied in B cells. Similarly, Sox5 encodes a transcription factor which plays a role in development52-57, but has not been looked at in B cell malignancies. Both genes were expressed much more highly in TRAF3-/- B lymphomas, and up-regulated at mRNA and protein levels in B lymphomas. Moreover, both MCC and Sox5 have been shown to be highly expressed in other types of cancers, although their role in TRAF3-/- B cell malignancies has not been studied. Thus we sought to better understand the role that MCC and Sox5 play in TRAF3 deficient B cell neoplasms, which would have implications for treatment and diagnosis of these types of

malignancies. In Chapter 1 I will describe our findings about the roles and mechanisms of MCC in our TRAF3-deficient B cell cancer model, including our identification of novel protein interactors for MCC. I will present the findings of our study of Sox5 in Chapter 2, the highlight of which is the novel isoform of Sox5 that we observed in B lymphomas spontaneously developed in B-TRAF3-/- mice.

These studies gained new insights into the mechanisms underlying TRAF3 inactivation-initiated B lymphomagenesis, and also discovered novel therapeutic targets for B cell neoplasms

We are also interested in applying our understanding of TRAF3 signaling mechanisms to develop new therapeutic strategies for the treatment of B cell neoplasms. In my thesis research, we have explored the possibility that restoration of TRAF3 downstream signaling components represents a line of new therapeutic avenues for B cell cancers. These translationally studies are described in the last two chapters. In Chapter 3, we sought to better understand the role of TRAF3 in

PKCδ signaling in B cells, by employing two pharmacological PKCδ activators, N- benzyladriamycin-14-valerate (AD 198), and ingenol-3-angelate (PEP005) to manipulate the PKCδ pathway in TRAF3-/- B cells. Interestingly, we found AD198 exhibit potent anti-tumor activities on B cell neoplasms via a novel, PKC- independent mechanism that specifically targets c-Myc. In Chapter 4, we investigated the therapeutic mechanisms of bortezomib, an effective clinical drug for multiple myeloma, in TRAF3 deficient mouse B lymphoma and human multiple myeloma models. We further examined the combined effects of bortezomib with

AD 198, as well as with oridonin (an NF-κB2 inhibitor), Our results provide a rationale for clinical evaluation of the combinations of bortezomib and oridonin (or other inhibitors of NF-B2) or AD 198 (or other drugs targeting c-Myc) in the treatment of B cell cancers, especially in patients containing TRAF3 deletions or relevant mutations..

CHAPTER 1 MUTATED IN COLORECTAL CANCER (MCC) IS A NOVEL ONCOGENE IN B LYMPHOCYTES

ABSTRACT

Identification of novel genetic risk factors is imperative for a better understanding of B lymphomagenesis and for the development of novel therapeutic strategies. TRAF3, a critical regulator of B cell survival, was recently recognized as a tumor suppressor gene in B lymphocytes. In the present study, we identified mutated in colorectal cancer (MCC) as a gene strikingly up-regulated in TRAF3-deficient mouse B lymphomas and human multiple myeloma (MM) cell lines. Aberrant up-regulation of MCC also occurs in a variety of primary human B cell malignancies, including non-Hodgkin lymphoma (NHL) and MM. In contrast,

MCC expression was not detected in normal or premalignant TRAF3-/- B cells even after treatment with B cell stimuli, suggesting that aberrant up-regulation of MCC is specifically associated with malignant transformation of B cells. In elucidating the functional roles of MCC in malignant B cells, we found that lentiviral shRNA- mediated knockdown of MCC induced apoptosis and inhibited proliferation in human MM cells. Experiments of knockdown and overexpression of MCC allowed us to identify several downstream targets of MCC in human MM cells, including

ERK, c-Myc, p27, cyclin B1, Mcl-1, caspases 8 and 3. Furthermore, we delineated the MCC-interactome in whole cells and mitochondria of human MM cells by employing affinity purification followed by mass spectrometry-based sequencing.

We identified 365 proteins (including 326 novel MCC-interactors) in the MCC

signaling complex, among which PARP1 and PHB2 were two hubs of MCC signaling pathways in human MM cells. Collectively, our results support the conclusion that MCC functions as an oncogene in B cells. Our findings suggest that MCC may serve as a diagnostic marker and therapeutic target in B cell malignancies, including NHL and MM.

1.1 BACKGROUND

Identification and validation of new genetic risk factors are imperative for a better understanding of B cell malignant transformation and for the development of new therapeutic strategies. Recent studies from our laboratory and others have identified TRAF3, a critical determinant of B cell survival, as a tumor suppressor gene in B lymphocytes. Homozygous deletions and inactivating mutations of the

TRAF3 gene have been identified in non-Hodgkin lymphoma (NHL), including splenic marginal zone lymphoma (MZL), B cell chronic lymphocytic leukemia (B-

CLL) and mantle cell lymphoma (MCL), as well as multiple myeloma (MM) and

Waldenström’s macroglobulinemia (WM)23,24,42,43. We recently reported that

TRAF3 deletion leads to spontaneous development of MZL and B1 lymphoma in mice15,16. Interestingly however, B lymphoma development in B cell-specific

TRAF3-deficient (B-TRAF3-/-) mice exhibits a long latency (about 9 months), indicating that additional oncogenic pathways are required for TRAF3-/- B lymphoma development. To identify such secondary oncogenic alterations in

TRAF3-/- B cells, we did a transcriptome microarray analysis using TRAF3-/- mouse

splenic B lymphomas. Surprisingly, our microarray analysis identified mutated in colorectal cancer (MCC), a tumor suppressor gene of colorectal cancer (CRC), as a strikingly up-regulated gene in TRAF3-/- B lymphomas.

The MCC gene was discovered in 1991 through its linkage to the region showing loss of heterozygosity (LOH) in familial adenomatous polyposis (FAP)47,58-

60. Subsequent studies revealed that the adenomatous polyposis coli (APC) gene and not MCC is responsible for FAP. However, the MCC gene is silenced through promoter methylation in approximately 50% of primary sporadic CRCs and 80% of serrated polyps, suggesting that the silencing of MCC is important in early colon carcinogenesis via the serrated neoplasia pathway49,61-63. Although an inactivating

MCC mutation in mice alone failed to induce any evident CRCs, the homozygous mice displayed a slightly higher proliferation rate of the epithelial crypt cells64,65.

Interestingly, an unbiased genetic screening of a mouse model of CRC implicated

MCC mutation as a key event in colorectal carcinogenesis66. Consistent with the genetic evidence, functional studies revealed that MCC blocks cell cycle progression in NIH3T3 fibroblasts and colorectal cancer cells50,67, inhibits cell proliferation and migration in colorectal cancer cells61,67-69, and is required for DNA damage response in colorectal cancer cells51. Collectively, the above genetic and functional evidence indicates that MCC functions as a tumor suppressor gene in

CRCs.

It has been shown that during development, MCC is expressed in diverse tissues derived from all three embryonic germ layers, including the developing gut

and central nervous system64. In adults, MCC is expressed in the surface epithelium of the colon and villi of the small intestines as well as other tissues, including the cerebellar cortex, kidney, pancreas, and liver50,70. However, expression of MCC was not reported in lymphocytes, and the function of MCC in lymphocytes or lymphomas has not been explored. In the present study, our unexpected finding that MCC was strikingly up-regulated in TRAF3-/- mouse B lymphomas prompts us to address this gap in knowledge. Our results demonstrate that in sharp contrast to its tumor suppressive function in CRCs, MCC is oncogenic in B lymphocytes.

1.2 MATERIALS AND METHODS

1.2.1 MICE, CELL LINES, AND REAGENTS

Mice and disease monitoring were as previously described15,16. Human patient-derived MM cell lines were generously provided by Dr. Leif Bergsagel

(Mayo Clinic, Scottsdale, AZ), including 8226 and KMS11 (contain biallelic TRAF3 deletions), LP1 and U266 (contain inactivating TRAF3 frameshift mutations), and

KMS28PE and KMS20 (contain biallelic cIAP1/2 deletions). The EBV-transformed human B lymphoblastoid cell line C3688 was provided by Dr. Lori Covey (Rutgers

University, Piscataway, NJ). All human B cell lines were cultured as described71.

Most antibodies and reagents used in this study were as previously described15,71,72. Mouse monoclonal Abs to MCC were purchased from Santa Cruz

Biotechnology (Santa Cruz, CA). DTT, thapsigargin, and lentiviral shRNA

constructs for human MCC were purchased from Sigma-Aldrich Corp. (St. Louis,

MO). The plasmid pENTR1A-NTAP-A containing the SBP tag sequence was purchased from Addgene (Cambridge, MA). PARP1 Abs were from eBioscience

(San Diego, CA), and PHB2 Abs were purchased from Bethyl Laboratories Inc.

(Montgomery, TX). Additional polyclonal rabbit Abs were from Cell Signaling

Technology (Beverly, MA).

1.2.2 TRANSCRIPTOME MICROARRAY ANALYSIS

Total RNA was extracted from splenocytes of LMC (mouse ID: 6983-6,

7041-9, and 7060-5) and tumor-bearing B-TRAF3-/- mice (mouse ID: 6983-2, 7041-

10, and 7060-8) using TRIzol reagent (Invitrogen, Carlsbad, CA) following the manufacturer’s instructions. RNA quality was assessed on an RNA Nano Chip using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). The mRNA was amplified with a TotalPrep RNA amplification kit with a T7-oligo(dT) primer according to the manufacturer’s instructions (Ambion), and microarray analysis was carried out with the Illumina Sentrix MouseRef-8 24K Array at the

Burnham Institute (La Jolla, CA). Results were extracted with Illumina

GenomeStudio v2011.1, background corrected and variance stabilized in

R/Bioconductor using the lumi package73,74 and modeled in the limma package75.

Microarray data are available from NIH GEO Accession GSE48818.

1.2.3 TAQMAN ASSAYS OF THE TRANSCRIPT EXPRESSION OF IDENTIFIED GENES

Total cellular RNA was extracted using TRIzol reagent (Invitrogen) according to the manufacturer’s protocol. Complementary DNA (cDNA) was prepared from RNA using High Capacity cDNA Reverse Transcription Kit (Applied

Biosystems, Carlsbad, CA). Quantitative real-time PCR of specific genes was performed using corresponding TaqMan Gene Assay kit (Applied Biosystems) as previously described76. Briefly, real-time PCR was performed using TaqMan primers and probes (FAM-labeled) specific for mouse MCC, Diras2, Tbc1d9,

Ccbp2, Btbd14a, Sema7a, Twsg1, Ppap2b, TCF4, Tnfrsf19, Zcwpw1, Abca3, or human MCC. Each reaction also included the probe (VIC-labeled) and primers for mouse or human β-actin mRNA, which served as endogenous control. Relative mRNA expression levels of each gene were analyzed using the Sequencing

Detection Software (Applied Biosystems) and the comparative Ct method (Ct) as previously described76.

1.2.4 SPLENIC B CELL PURIFICATION AND STIMULATION

Splenic B cells were purified using anti-mouse CD43-coated magnetic beads and a MACS separator (Miltenyi Biotec Inc.) following the manufacturer’s protocols as previously described16,76. The purity of isolated populations was monitored by FACS analysis and cell preparations of >95% purity were used for protein extraction. Purified B cells were cultured ex vivo in the absence or presence of B cell stimuli for 6, 24 or 48 hours as described previously16,76. B cell stimuli

examined include 2 μg/ml anti-CD40, 20 μg/ml LPS, 1 μg/ml anti-BCR, and 100 nM CpG2084, alone or in combination. Total cellular RNA was prepared at 6 or 24 hours after stimulation.

1.2.5 TAQMAN COPY NUMBER ASSAY OF THE MOUSE MCC GENE

Genomic DNA was prepared from splenocytes of LMC and tumor-bearing

B-TRAF3-/- mice as previously described15. Quantitative real-time PCR of the mouse MCC gene was performed using the TaqMan Copy Number Assay kit

(assay ID: Mm00490037_cn; Applied Biosystems) following the manufacturer’s protocols. Briefly, real-time PCR was performed using the TaqMan primers and probe (FAM-labeled) specific for the mouse MCC gene. Each reaction also included the probe (VIC-labeled) and primers specific for the mouse Trfc gene

(TaqMan Copy Number Reference Assay, Applied Biosystems), which served as reference control. Relative copy numbers of the MCC gene in genomic DNA samples were analyzed using the Sequencing Detection Software (Applied

Biosystems) and the comparative Ct method (Ct) following the manufacturer’s protocols.

1.2.6 CHROMATIN IMMUNOPRECIPITATION ASSAY

Chromatin immunoprecipitation (ChIP) assays were performed essentially as described previously77,78. Briefly, 20 x 106 cell equivalents of fragmented chromatins (containing DNA fragments < 400 bp) were immunoprecipitated with Ig isotype control or antibodies specific for histone modifications. We used the

following histone antibodies in the ChIP experiments: H3K27me3 (Millipore,

Billerica, MA) and H3K9/14ac (Diagenode, Denville, NJ). Immunoprecipitated DNA were purified, and then analyzed by quantitative real time PCR using primers specific for the promoter region of the mouse MCC or Diras2 gene. Primer used are: mMCC-U519F (5’- CAG GGA GGT TGG AGA GGA -3’) and mMCC-U446R

(5’- AAA CAT GCC CTG CCC TTG -3’); mDiras2-U559F (5’- GCA CAT GTG ACT

ACT ATT G -3’) and mDiras2-U480R (5’- AAT CTC TCC TCC CAC AAG -3’). PCR amplicons are in the 70–120 bp range. The enrichments of each gene promoter immunoprecipitated by histone marks were quantitated relative to the input DNA.

1.2.7 CLONING OF THE FULL-LENGTH CDNA OF THE MCC GENE FROM TRAF3-/- MOUSE B LYMPHOMAS AND HUMAN MM CELL LINES

Total cellular RNA was prepared from B lymphomas spontaneously developed in four individual B-TRAF3-/- mice (mouse ID: 6983-2, 7060-8, 105-8, and 115-6), and the corresponding cDNA samples were used as templates to clone the mouse MCC coding sequences using primers mMCC-F (5’- ATG AAT

TCT GGA GTT GCG GTG -3’) and mMCC-R (5’- TTA GAG TGA CGT TTC GTT

GGT G -3’). Similarly, human MCC coding sequences were cloned from human

MM cell lines LP1 and KMS11 cells using reverse transcription PCR. Primers used for the cloning of human MCC are hMCC-F (5’- TGC ATC ATG AAT TCC GGA GT

-3’), and hMCC-R (5’- TTA AAG CGA AGT TTC ATT GGT GTG -3’). The high fidelity polymerase Pfu UltraII (Santa Clara, CA) was used in these PCR reactions.

Sequences of the cloned mouse and human MCC were determined at GenScript

(Piscataway, NJ).

1.2.8 GENERATION OF LENTIVIRAL MCC EXPRESSION AND SHRNA VECTORS

The coding cDNA sequence of MCC cloned from the human MM cell line

LP1 cells was subcloned into the lentiviral expression vector pUB-eGFP-Thy1.179

(generously provided by Dr. Zhibin Chen, the University of Miami, Miami, FL) by replacing the eGFP coding sequence with the MCC coding sequence. To facilitate immunoprecipitation experiments, we engineered an N-terminal FLAG tag or a C- terminal SBP-6xHis tag 80 in frame with the MCC coding sequence, respectively.

We subsequently generated two lentiviral expression vectors of tagged hMCC, including pUB-FLAG-hMCC and pUB-hMCC-SBP-6xHis. Each lentiviral expression vector was verified by DNA sequencing.

Lentiviral shRNA vectors specific for human MCC (including hMCC shRNA

1332, 1388, 2284 and 2689; all in Torc1 vectors) or a scrambled shRNA vector were purchased from Sigma. To facilitate FACS analysis and cell sorting, we engineered an eGFP-expressing version of all the shRNA vectors by replacing the puromycin resistance gene of Torc1 with the eGFP coding sequence. All original and engineered lentiviral shRNA vectors were verified by DNA sequencing.

1.2.9 LENTIVIRAL PACKAGING AND TRANSDUCTION OF HUMAN MM CELLS

Lentiviruses of MCC shRNA vectors and a scrambled shRNA vector were packaged following the manufacturer’s protocol (Sigma) as previously described15,81. Lentiviruses of MCC expression vectors were packaged and titered as previously described71,72,79. Human MM cells were transduced with the packaged lentiviruses at a MOI of 1:5 (cell:virus) in the presence of 8 μg/ml polybrene15,79,81. Transduction efficiency of cells was analyzed by flow cytometry on day 3 post transduction. All shRNA vectors contain an eGFP expression cassette, and were directly analyzed using a flow cytometer. All pUB lentiviral expression vectors have an expression cassette of the marker Thy1.1, and thus allow the transduced cells to be analyzed by Thy1.1 immunofluorescence staining followed by flow cytometry. Transduced cells were subsequently sorted or directly analyzed for apoptosis, cell cycle distribution, proliferation, and protein expression.

1.2.10 GROWTH CURVE DETERMINATION, ANNEXIN V STAINING OF APOPTOTIC CELLS, CELL CYCLE DISTRIBUTION AND CELL PROLIFERATION ANALYSES

For MCC knockdown studies, on day 4 post-transduction, successfully transduced cells were sorted for GFP+ shRNA expressing populations, and then plated in 6-well plates for growth curve determination, annexin V staining, or cell cycle analysis. For growth curve determination, live and dead cells were differentiated using trypan blue staining, and counted using a hemocytometer. For analysis of apoptosis, cells were stained with annexin V and PI according to the

manufacturer’s protocol (Invitrogen), and analyzed by flow cytometry as previously described15. For cell cycle analysis, cells were fixed with ice-cold 70% ethanol. Cell cycle distribution was subsequently determined by propidium iodide (PI) staining followed by flow cytometry as previously described76,82. For cell proliferation analysis, cells were labeled with a cell proliferation dye eFluor® 670 (eBioscience) and dilution of the proliferation dye was analyzed by flow cytometry following the manufacturer’s protocol.

1.2.11 TOTAL PROTEIN LYSATES, FRACTIONATION OF CYTOSOL, MITOCHONDRIA AND MICROSOMES (RICH IN ER), AND IMMUNOBLOT ANALYSIS

For total protein lysates, cell pellets were lysed in 200 μl of 2X SDS sample buffer (0.0625 M Tris, pH6.8, 1% SDS, 15% glycerol, 2% β-mercaptoethanol and

0.005% bromophenol blue), sonicated for 30 pulses, and boiled for 10 minutes.

For biochemical fractionation, human MM cells (30 x 106 cells/condition) were cultured in the absence or presence of 250 μM DTT or 0.5 μM thapsigargin for 24 hours. Cytosol, mitochondria and microsomes (rich in ER) were fractionated from cells as previously described83-85. Briefly, cells were washed with ice-cold

PBS, swelled in 700 μl of Mitochondria Isolation Buffer (250 mM sucrose, 10 mM

HEPES, pH7.5, 10 mM KCl, 1 mM EDTA, and 0.1 mM EGTA with protease and phosphatase inhibitors) on ice for 10 minutes, and then homogenized in a Dounce homogenizer. Cell lysis was checked by trypan blue uptake, and homogenization stopped when 90% of cells were broken. Nuclei were pelleted by centrifugation at

1,000 g for 10 minutes at 4oC. The cleared lysates were then centrifuged at

10,000g for 25 minutes at 4°C to obtain the pellets of mitochondria. The supernatants were further centrifuged at 100,000 g for 2 hours to separate the pellets of microsomes (rich in ER) from cytosolic proteins (S100 fraction). One-fifth volume of 5X SDS sample buffer was added into each S100 fraction. The pellets of nuclei, mitochondria and microsomes (rich in ER) were lysed and sonicated in

200 μl of 2X SDS sample buffer, respectively. All protein samples were subsequently boiled for 10 minutes.

Total protein lysates, or cytosolic, ER, mitochondrial and nuclear proteins were separated by SDS-PAGE. Immunoblot analyses were performed using specific antibodies as previously described. Images of immunoblots were acquired using a low-light imaging system (LAS-4000 mini, FUJIFILM Medical Systems

USA, Inc., Stamford, CT).

1.2.12 CO-IMMUNOPRECIPITATION ASSAY IN WHOLE CELL LYSATES

Human MM cell line 8226 cells (5 x 107 cells/condition) transduced with pUB-FLAG-hMCC or pUB-hMCC-SBP-6xHis were lysed and sonicated in the

CHAPS lysis buffer86 (1% CHAPS, 20 mM Tris, pH 7.4, 150 mM NaCl, 50 mM β- glycerophosphate, and 5% glycerol with freshly added 1 mM DTT and EDTA-free

Mini-complete protease inhibitor cocktail). The insoluble pellets were removed by centrifugation at 10,000g for 20 minutes at 4°C. The CHAPS lysates were subsequently immunoprecipitated with anti-FLAG-agarose beads (Sigma; for

FLAG-hMCC), or streptavidin-sepharose beads (Pierce, Rockford, IL; for hMCC-

SBP-6xHis), separately. Immunoprecipitates were washed 5 times with the Wash

Buffer (0.5% CHAPS, 20 mM Tris, pH 7.4, 150 mM NaCl, 50 mM β- glycerophosphate, and 2.5% glycerol with freshly added 1 mM DTT and EDTA- free Mini-complete protease inhibitor cocktail). Immunoprecipitated proteins were resuspended in 2x SDS sample buffer, boiled for 10 minutes, and then separated on SDS-PAGE for mass spectrometry or immunoblot analyses.

1.2.13 CO-IMMUNOPRECIPITATION ASSAY IN MITOCHONDRIAL LYSATES

Human MM cell line 8226 cells (1.5 x 108 cells/condition) transduced with pUB-FLAG-hMCC or pUB-hMCC-SBP-6xHis were used for cytosol, mitochondria and ER fractionation as described above. Mitochondrial pellets were lysed and sonicated in the CHAPS lysis buffer86, and cleared by centrifugation at 10,000g for

20 minutes at 4°C. The mitochondrial lysates were subsequently immunoprecipitated with streptavidin-sepharose beads (Pierce, Rockford, IL; for hMCC-SBP-6xHis). Immunoprecipitates were washed 5 times with the Wash

Buffer, resuspended in 2x SDS sample buffer, boiled for 10 minutes, and then separated on SDS-PAGE for mass spectrometry or immunoblot analyses.

1.2.14 MASS SPECTROMETRY BASED-SEQUENCING

Whole cell lysates or mitochondrial lysates immunoprecipitated with streptavidin-sepharose beads were used for LC-MS/MS. LC-MS/MS was

performed in collaboration with Dr. David Perlman at Princeton Collaborative

Proteomics & Mass Spectrometry Center at Princeton University (Princeton, New

Jersey). The entire gel lanes for hMCC-SBP-6xHis complex and negative control

(FLAG-hMCC) were each sectioned into 15 continuous slices. The gel slice samples were subjected to thiol reduction by TCEP, alkylation with iodoacetamide, and digestion with sequencing-grade modified trypsin87,88. Peptides were eluted from the gel slices, desalted, and then subjected to reversed-phase nano-flow ultra-high performance capillary liquid chromatography (uPLC) followed by high- resolution/high-mass accuracy MS/MS analysis using an LC-MS platform consisting of an Eksigent Nano Ultra 2D Plus uPLC system hyphenated to a

Thermo Orbi Velos mass spectrometer. The MS/MS was set to operate in data dependent acquisition mode using a duty cycle in which the top 15 most abundant peptide ions in the full scan MS were targeted for MS/MS sequencing. Full scan

MS1 spectra were acquired at 100,000 resolving power and maintained mass calibration to within 2-3 ppm mass accuracy. LC-MS/MS data were searched against the human IPI and UniProt databases using the Mascot and Proteome

Discoverer search engines87,88. Protein assignments were considered highly confident using a stringent false discovery rate threshold of <1%, as estimated by reversed database searching, and requiring that ≥2 peptides per protein be unambiguously identified. Rough relative protein amounts were estimated using spectra counting values.

1.2.15 STATISTICS

Statistical analyses were performed using the Prism software (GraphPad,

La Jolla, CA). Statistical significance was assessed by Student t test. P values less than 0.05 are considered significant.

1.3 RESULTS

1.3.1 TRANSCRIPTOME MICROARRAY ANALYSIS OF TRAF3 /- MOUSE B LYMPHOMAS

To identify secondary oncogenic alterations required for B lymphoma development, we performed global gene expression profiling of TRAF3-/- mouse

B lymphomas by transcriptome microarray analysis. We used RNA samples of 3 representative TRAF3-/- splenic B lymphomas, in which B lymphomas are >70% of B cells. We identified 160 up-regulated genes and 244 down-regulated genes in

TRAF3-/- B lymphomas as compared to littermate control (LMC) spleens (cut-off fold of changes: 2-fold up or down, p < 0.05; Table 1) (NCBI GEO accession number: GSE48818). Selective examples of these identified genes are shown in the heatmap Fig. 1A. We verified the up-regulation of 12 genes using Taqman real time PCR, including Diras2, MCC, Tbc1d9, Ccbp2, Btbd14a, Sema7a, Twsg1,

Ppap2b, TCF4, Tnfrsf19, Zcwpw1, and Abca3 (Fig. 1B). Striking up-regulation of the transcripts was verified in splenic B lymphoma samples from six mice and also ascites from two mice, indicating the recurrent nature of up-regulation of these 12

genes. We next surveyed the public gene expression database of microarray data of human cancers (http://www.oncomine.org). Interestingly, among the up- regulated genes identified in our microarray analysis, MCC is most consistently up-regulated in a variety of primary human B cell malignancies, including primary effusion lymphoma (PEL), centroblastic lymphoma (CBL), diffuse large B-cell lymphoma (DLBCL), Burkitt's lymphoma (BL), and MM89-92. We thus selected MCC for further study.

Figure 1. Representative genes differentially expressed in TRAF3-/- mouse B lymphomas identified by the transcriptome microarray analysis. (A) Heatmap of representative microarray data. The mRNA expression profiles of splenocytes from 3 pairs of LMC and tumor-bearing B-TRAF3-/- mice were analyzed by microarray analysis. cRNA was hybridized to the Illumina Sentrix MouseRef-8 24K Array (Illumina). Genes shown in the heatmap are selected from 160 up-regulated and 244 down-regulated genes: red indicates overexpression, blue indicates underexpression, and black indicates median expression. Values under the color key indicates Log2 (fold of change). Dendrogram on top of the heatmap shows the relatedness between samples as determined by hierarchical clustering. (B) Verification of transcript up-regulation of genes identified by the microarray analysis using quantitative real time PCR. Total cellular RNA was prepared from splenocytes of LMC mice, or splenic B lymphomas (spl) and ascites (asc) of diseased B-TRAF3-/- mice, and cDNA was synthesized by reverse transcription. Real time PCR was performed using TaqMan primers and probes (FAM-labeled) specific for mouse Diras2, MCC, Tbc1d9, Ccbp2, Btbd14a, Sema7a, Twsg1, Ppap2b, TCF4, Tnfrsf19, Zcwpw1, and Abca3. Each reaction also included the probe (VIC-labeled) and primers for mouse -actin mRNA, which served as endogenous control. Relative mRNA expression levels of each gene were analyzed using the Sequencing Detection Software (Applied Biosystems) and the comparative Ct (Ct) method. Graphs depict the results of two experiments with duplicate reactions in each experiment (mean ± S.D.), and mouse ID of each sample is indicated in the graphs.

A B Diras2 212 2 10 2 8 2 6 2 4 2 2

2 0 Relative mRNA level mRNA Relative MCC 212 2 10 2 8 2 6 2 4 2 2

2 0 Relative mRNA level mRNA Relative Tbc1d9 2 7 2 6 2 5 2 4 2 3 2 2 2 1 2 0

LMC B lymphomas level mRNA Relative

5-5 spl

7060-5

7060-6

7041-2

7041-9

6983-2 spl

7060-8 spl 7140-7 spl

-3 -2 -1 0 1 2 3 7140-3 spl

7060-8 asc

7140-3 asc 7041-10 spl

LMC TRAF3-/- lymphoma

Ccbp2 Btbd14a Sema7a 2 8 2 10 212 8 2 10 2 6 2 2 6 2 8 2 4 2 6 2 4 2 4 2 2 2 2 2 2 2 0 2 0 2 0

Relative mRNA level mRNA Relative

10 Twsg1 Ppap2b TCF4 2 212 212 2 8 2 10 2 10 8 8 2 6 2 2 2 6 2 6 2 4 2 4 2 4 2 2 2 2 2 2 2 0 2 0 2 0

Relative mRNA level mRNA Relative

Tnfrsf19 Zcwpw1 6 Abca3 212 212 2 5 210 210 2 28 28 24 26 26 23 24 24 22 22 22 21 20 20 20

Relative mRNA level mRNA Relative

5-5 spl

7060-5

7060-6

7041-2 7041-9

5-5 spl 5-5 spl

7060-5

7060-6

7041-2

7041-9

7060-5

7060-6

7041-2

7041-9

6983-2 spl

7060-8 spl

7140-7 spl 7140-3 spl

6983-2 spl

7060-8 spl

7140-7 spl

7140-3 spl

6983-2 spl

7060-8 spl

7140-7 spl

7140-3 spl

7060-8 asc 7140-3 asc

7060-8 asc

7140-3 asc

7060-8 asc

7140-3 asc 7041-10 spl

7041-10 spl 7041-10 spl LMC TRAF3-/- lymphoma LMC TRAF3-/- lymphoma LMC TRAF3-/- lymphoma

1.3.2 STRIKING UP-REGULATION OF MCC IN TRAF3 /- MOUSE B LYMPHOMAS BUT NOT IN PREMALIGNANT TRAF3 /- B LYMPHOCYTES

We first verified the up-regulation of MCC protein in TRAF3-/- splenic B lymphomas and ascites, and also that MCC protein expression was not detected in LMC splenic B cells (Fig. 2A). The expression of MCC is gradually up-regulated during differentiation of PC12 cells induced by NGF70. We thus investigated the expression of MCC during the proliferation and differentiation of B lymphocytes induced by a variety of B cell stimuli. We purified splenic B cells from LMC or tumor- free young B-TRAF3 /- mice (10-12 weeks). We found that the transcript of MCC was neither up-regulated nor detected in LMC or premalignant TRAF3-/- splenic B cells after treatment with any of the stimuli examined, including CD40, LPS, B cell receptor (BCR), and CpG2084, alone or in combination (Fig. 2B). In contrast, the expression of Zcwpw1, another gene identified by our microarray analysis, was robustly up-regulated in B cells after treatment with LPS or CpG in combination with CD40. Thus, MCC up-regulation appears to be selectively associated with B cell malignant transformation.

We then cloned the full-length coding sequence of the MCC transcripts from primary splenic B lymphomas of 4 different individual B-TRAF3-/- mice, and found that the MCC gene expressed in TRAF3-/- B lymphomas is predicted to encode a protein of 828 amino acids and contains no mutations. To understand how MCC is up-regulated in TRAF3-/- B lymphomas, we found that the copy number of the

MCC gene was not changed in TRAF3-/- mouse B lymphomas (Fig. 2C). The

activating H3K9/14 acetylation of the promoter region of the MCC gene was not obviously changed in TRAF3-/- B lymphoma cells either (Fig. 2D). In contrast, the repressive H3K27 trimethylation of the MCC promoter region was almost completely abolished in TRAF3-/- B lymphoma cells, and also slightly decreased in premalignant TRAF3-/- splenic B cells as compared to LMC B cells (Fig. 2D).

Such changes in H3K27 trimethylation were not observed in the promoter region of Diras2 (Fig. 2D), another gene identified in our microarray as being strikingly up-regulated in TRAF3/-/ B lymphomas. Our results suggest that alterations in epigenetic modifications (such as H3K27Me3) of the MCC promoter contribute at least partially to the up-regulation of MCC observed in TRAF3-/- B lymphomas.

Figure 2. Striking up-regulation of MCC in TRAF3 /- mouse B lymphomas but not in premalignant TRAF3 /- B lymphocytes. (A) Western blot analysis of the MCC protein. Total cellular proteins were prepared from purified LMC splenic B cells or splenic B lymphomas (spl) or ascites (asc) of different individual B-TRAF3- /- mice. Proteins were immunoblotted for MCC, followed by actin. (B) Regulation of the expression of MCC and Zcwpw1 in response to B cell stimuli. Splenic B cells were purified from 10- to 12- week-old LMC and tumor-free B-TRAF3-/- mice. Purified B cells were cultured ex vivo in the absence or presence of stimuli of B cell proliferation, differentiation and activation for 6 or 24 hours. B cell stimuli examined include: 2 g/ml anti-CD40, 20 g/ml LPS, 1 g/ml anti-BCR, and 100 nM CpG2084, alone or in combination. RNA samples of primary TRAF3-/- mouse B lymphomas (mouse ID: 7060-8) were used as positive control of the MCC and Zcwpw1 transcripts in the experiments. Taqman assays of MCC and Zcwpw1 were performed as described in Figure 1. Graphs depict the results of three independent experiments with duplicate reactions in each experiment (mean ± S.D.). (C) Normal copy number of the MCC gene in TRAF3-/- mouse B lymphomas. Copy number of the mouse MCC gene in genomic DNA samples prepared from LMC splenocytes or TRAF3-/- B lymphomas was determined using the TaqMan Copy Number Assay kit. Mouse ID was indicated in the figure. Graphs depict the results of three experiments with duplicate reactions in each experiment (mean ± S.D.). (D) Histone modifications of the promoter region of the MCC and Diras2 genes. Chromatin was prepared from splenic B cells purified from 10- to 12- week-old LMC and tumor-free B-TRAF3-/- mice (young TRAF3-/-), or TRAF3-/- B lymphomas. Fragmented chromatin was immunoprecipitated using antibodies specific for histone marks (H3K27me3 or H3K9/14ac) or non-specific rabbit Ig. Immunorecipitated DNA was quantified by quantitative real time PCR using primer pairs specific for the promoter region of the MCC and Diras2 genes, respectively. Quantity of immunoprecipitated DNA is presented as percentage of input DNA in graphs. The graphs depict

the results of three independent experiments with duplicate reactions in each experiment (mean ± S.D.). *P < 0.05 and **P < 0.0001 by Student’s t test.

A TRAF3-/- B lymphomas

spl

asc

10 10 spl

spl

asc

spl

2 2

8 8

spl

8 8 8 8

LMC splenic B 6

3 3

105

6983

105

115

69 77041 1 2 3 4 5 7060

MCC

Actin

Zcwpw1, 24h B MCC, 6h MCC, 24h 212 212 26 210 210 LMC 28 28 B-TRAF3-/- 24 26 26 24 24 22 22 22

Relative mRNA level mRNA Relative 0 20

20 2

Un.

Un. LPS

LPS

CD40 CD40

CD40

CD40+LPS

CpG+CD40

CD40+BCR

CD40+LPS

CpG+CD40 CD40+BCR

CD40+LPS B lymphoma

CpG+CD40

CD40+BCR B lymphoma

B lymphoma

CD40+BCR+LPS CD40+BCR+LPS

CD40+BCR+LPS

C MCC 3

Gene copy number copy Gene 0

27-9 33-3 69-3

284-2 288-1

105-8 115-6

7041-2 7126-3

6983-2

7060-8

7041-10 LMC TRAF3-/- lymphoma

D MCC * Diras2 25

LMC 20 ** 6 Young TRAF3-/- 15 B lymphoma 4 10 2

% of Input DNA of Input %

5 DNA of Input %

0 0 Rabbit Ig H3K9/14Ac H3K27Me3 Rabbit Ig H3K9/14Ac H3K27Me3

1.3.3 ABERRANT UP-REGULATION OF MCC IN HUMAN PATIENT- DERIVED MM CELL LINES WITH TRAF3 DELETIONS OR RELEVANT MUTATIONS

Confirming the clinical relevance, we found that MCC is aberrantly up- regulated at both the transcript and protein levels in 6 human patient-derived MM cell lines with TRAF3 deletions or relevant mutations and an EBV-transformed B lymphoblastoid cell line93, which all exhibit constitutive NF-κB2 activation (Fig. 3A and 3B). We further cloned the full-length coding sequence of the MCC transcripts from human MM cell lines LP1 and KMS11 cells, and found that the cloned MCC are predicted to encode a protein of 829 amino acids and contains no mutations.

Therefore, MCC is aberrantly expressed in both mouse and human transformed B cells, but not expressed in normal B lymphocytes.

Figure 3. Aberrant expression and subcellular localization of MCC in human MM cells. (A) Taqman assay of the MCC transcript. Total cellular RNA was prepared from normal human blood B lymphocytes (Normal B), an EBV-transformed B lymphoblastoid cell line C3688, or human MM cell lines with TRAF3 deletions or relevant mutations. The human MM cell lines include 8226 and KMS11 (with biallelic TRAF3 deletions), LP1 and U266 (with inactivating TRAF3 frameshift mutations), and KMS28PE and KMS20 (with biallelic cIAP1/2 deletions). Real-time PCR was performed using TaqMan primers and probes (FAM-labeled) specific for human MCC. Each reaction also included the probe (VIC-labeled) and primers for human β-actin mRNA, which served as an endogenous control. Relative mRNA expression levels of the MCC gene were analyzed using the Sequencing Detection Software (Applied Biosystems) and the comparative Ct (Ct) method. Graphs depict the results of three independent experiments with duplicate reactions in each experiment (mean ± S.D.). (B) Western blot analysis of the MCC protein. Total cellular proteins were prepared from normal B lymphocytes, C3688 cells, or human MM cell lines with TRAF3 deletions or relevant mutations. Proteins were immunoblotted for MCC, followed by TRAF3 and actin. Data shown are representative of 3 experiments. (C) Subcellular localization of MCC and its regulation during ER stress responses. LP1 cells were cultured in the absence (control) or presence of ER stress inducers DTT or thapsigargin (Thg). After treatment for 24 hours, cytosol (S100), ER, mitochondria (Mito) and nuclei (Nuc) were biochemically fractionated from cells, and an aliquot of cells was used for total protein lysates (Total). Proteins in each fraction or total lysates were immunoblotted for MCC, Rhbdf1 (an ER protein), cIAP1, cIAP2, cyclin D2, COX IV (a mitochondrial protein), and actin. Results shown are representative of 3 independent experiments.

A MCC B 2 7

2 6

KMS28PE

KMS11

KMS20

Normal B Normal

C3688

U266

8226 LP1 2 5

4 2 MCC 2 3 2 2

Relative mRNA level mRNA Relative 2 1 TRAF3 2 0

LP1

8226 U266 Actin

C3688

KMS11

KMS20

Normal B

KMS28PE

C S100 ER mito Nuc Total

Control

DTT

Thg

Control

DTT

Thg

Control

Thg

Control

DTT

Thg

Control

DTT

Thg DTT

MCC

Rhbdf1

cIAP1

cIAP2

Cleaved cIAP2

Cyclin D2

Cox IV

Actin

1.3.4 SUBCELLULAR LOCALIZATION OF MCC IN HUMAN MM CELLS AND REGULATION OF MCC DURING ER STRESS RESPONSES

To elucidate where MCC functions in malignant B cells, we examined the subcellular localization of MCC by biochemical fractionation in the absence or presence of two ER stress inducers, DTT (a reducing agent) and thapsigargin (an inhibitor of ER Ca2+ transport). Our results revealed that MCC protein was primarily localized in the mitochondria, but also detectable in the ER, cytosol and nucleus in human MM cells (Fig. 3C). Interestingly, DTT and thapsigargin markedly decreased the protein level of MCC, but did not change its subcellular localization in human MM cells. The ER stress responses induced by DTT and thapsigargin were evident as demonstrated by the cleavage of cIAP2 and the induction of apoptosis in human MM cells. These results suggest that MCC, primarily localized in mitochondria, may regulate malignant B cell survival.

1.3.5 LENTIVIRAL SHRNA-MEDIATED KNOCKDOWN OF MCC INDUCED APOPTOSIS AND INHIBITED PROLIFERATION IN HUMAN MM CELLS

To explore the roles of MCC, we employed genetic means to manipulate the expression levels of MCC in human MM cells. In contrast to the cell cycle blocking and proliferation inhibitory effects of MCC overexpression reported in fibroblasts and CRCs 50,51,61,67, overexpression of MCC in human MM cell line 8226 cells (containing modest levels of endogenous MCC) did not affect cell cycle progression or proliferation (data not shown). Although expression of ectopically transfected MCC was strongly reduced in CRCs after several passages,

overexpression of MCC was maintained in human MM 8226 cells after >10 passages (data not shown). Our results of overexpression studies thus argue against a negative role for MCC in the survival or proliferation of malignant B cells.

We then found that lentiviral shRNA vector-mediated knockdown of MCC inhibited growth and induced apoptosis in human MM cells (Fig. 4). Interestingly, MCC shRNA 1332 resulted in a greater decrease in MCC protein level and was most effective at inducing apoptosis in human MM cells. In contrast, MCC shRNAs 1388 and 2284, which did not markedly knock down MCC protein level, did not drastically induce apoptosis in human MM cells (Fig. 4). Our cell cycle analyses demonstrated that transduction of MCC shRNA 1332 or 2689 resulted in a dramatic decrease in the proliferating cell population (2n < DNA content ≤ 4n) and an increase in the apoptotic cell population (DNA content < 2n) in human MM cells

(Fig. 5A). MCC shRNA 1332 was also shown to be more potent than shRNA 2689 in these experiments (Fig. 5A). Results of proliferation dye labeling experiments further confirmed that knockdown of MCC by shRNA 1332 remarkably inhibited the proliferation of human MM cells (Fig. 5B). Together, our results reveal that

MCC is required for the survival and proliferation of human MM cells.

MCC shRNA

1388

2284

2689

Scr 1332

MCC

Actin

B Torc.Scr hMCC shRNA 1332 0.8 100 hMCC shRNA 1388 Live hMCC shRNA 2284 0.6 80 Dead hMCC shRNA 2689 60

cells/ml) 0.4

6 40

(x10 0.2

20 Percentage of cells Percentage

Live cell concentration cell Live 0.0 0 0 1 2 3 4 Scr 1332 1388 2284 2689 Days MCC shRNA

C MCC shRNA

Scr 1332 1388 2284 2689

Figure 4. Lentiviral shRNA vector-mediated knockdown of MCC induced apoptosis in human MM cells. The human MM cell line LP1 cells were transduced with lentiviruses expressing MCC shRNAs (including 1332, 1388, 2284 or 2689) or a scrambled shRNA (Scr). (A) Knockdown of the MCC protein determined by immunoblot analysis. Total cellular proteins were prepared from LP1 cells successfully transduced with a scrambled shRNA (Scr), or MCC shRNA 1332, 1388, 2284 or 2689 on day 5 post-transduction. Proteins were immunoblotted for MCC, then actin (a loading control). Results shown are representative of 3 independent experiments, and similar results were obtained in KMS11 cells. (B) Growth curves of live cells and percentage of live versus dead cells determined by Trypan blue-stained cell counting. On day 4 post transduction, successfully transduced LP1 cells were cultured in a 6-well plate for growth curve determination. Live and dead cells were counted daily for 4 days using Trypan blue staining and a hemocytometer. The graphs depict the results of 3 independent experiments (mean ± S.D.). Similar results were obtained in KMS11 cells. (C) Representative FACS profiles of transduced cells analyzed by annexin V and PI staining. On day 6 post-transduction, transduced LP1 cells were stained with annexin V and PI, and then analyzed by a flow cytometer. In the FACS profiles, apoptotic cells were identified as annexin V+ PI-, dead cells were annexin V+ PI+, and live cells were annexin V- PI-. Data shown are

representative of 3 independent experiments, and similar results were obtained in the human MM cell line KMS11 cells.

MCC shRNA A

Scr 1332 2689

# cells#

Day 0 after sorting 0Day after

# cells#

# cells# Day 1 after sorting 1Day after

DNA content

B Profile Overlay

Scr 1332 GFP- GFP+

Day 0 Day

670

eFluor

- Day 2 Day

Scr.1

Labeling with a proliferation dye a proliferationwith Labeling Scr.2

Day 4 Day 1332.1 1332.2

Figure 5. Knockdown of MCC induced apoptosis and inhibited proliferation in human MM cells. (A) Cell cycle distribution analyzed by PI staining and flow cytometry. The human MM cell line LP1 cells were transduced with lentiviruses expressing MCC shRNAs (1332 or 2689) or a scrambled shRNA (Scr). Successfully transduced (GFP+) cells were sorted by a FACS sorter on day 4 post transduction. Top panel shows the FACS histograms of cells right after sorting (Day 0), and bottom panel shows the FACS histograms of sorted cells after cultured for 24 hours (Day 1). Gated populations in the FACS histograms indicate apoptotic cells (DNA content < 2n) or proliferating

cells (2n < DNA content ≤ 4n). Results shown are representative of 3 independent experiments, and similar results were also obtained in the human MM cell line KMS11 cells. (B) Cell proliferation analyzed by dilution of the proliferation dye (eFluor 670)-labeling and flow cytometry. On day 4 post transduction, LP1 cells transduced with MCC shRNA 1332 (1332) or a scrambled shRNA (Scr) were labeled with a proliferation dye eFluor 670, which binds to any cellular protein containing primary amines. As cells divide, the dye is distributed equally between daughter cells that can be measured as successive halving or dilution of the fluorescence intensity of the dye. Day 0 FACS profiles and overlay histograms show the eFluor 670 signals of freshly labeled cells, while those of Day 2 and Day 4 show diluted eFluor 670 signals of cells after cultured for 2 and 4 days, respectively. Inhibited proliferation of MCC shRNA 1332-transduced cells was demonstrated by the hampered dilution of the proliferation dye as compared to those observed in both the untransduced (GFP-) cells and the scrambled shRNA (Scr)-transduced cells. Results shown are representative of 3 independent experiments with duplicate samples in each experiment. Similar results were also obtained in the human MM cell line KMS11 cells.

1.3.6 MCC SIGNALING PATHWAYS IN HUMAN MM CELLS

To understand how MCC functions in human MM cells, we first investigated the known targets of MCC, including NF-κB and β-catenin pathways identified in

CRCs and hepatocellular carcinoma61,67,94,95. Our results showed that knockdown or overexpression of MCC did not change any of the known MCC targets in human

MM cells, including IκBα, IκBβ, RelA, phospho-β-catenin and β-catenin (Fig. 6).

Interestingly, we found that as opposed to what was observed in CRCs 51, knockdown of MCC markedly decreased the level of phospho-histone H3 (P-HH3, a marker of mitosis) in human MM cells, confirming the proliferation-promoting function of MCC in malignant B cells. We subsequently measured the levels or activation of a number of regulators of cell apoptosis and proliferation, including caspases, the Bcl-2 family proteins, cyclins, and kinases. Our results demonstrated that knockdown of MCC led to activation of both caspases 8 and 3

(Fig. 6). Knockdown of MCC, interestingly, decreased the levels of Mcl1, cyclin B1,

phospho-ERK, and c-Myc and in contrast, its overexpression increased the levels of these proteins in human MM cells (Fig. 6). Moreover, knockdown of MCC led to specific up-regulation of protein levels of the cell cycle inhibitor p27, which was modestly down-regulated by overexpression of MCC in human MM cells. However,

MCC knockdown or overexpression did not affect the protein levels of other examined proteins, including Bcl-xL, Bcl2, Bim, Bad, Bid, Bik, cyclin D1, cyclin D2, p21, E2F1, p53, p38, JNK and Akt. Our findings suggest that MCC inhibits apoptosis and induces proliferation by up-regulating Mcl1, cyclin B1, ERK activation and c-Myc, and by inhibiting cleavage of caspases 8 and 3 and down- regulating p27 in human MM cells.

LP1 8226 LP1 8226 LP1 8226

1332 1332 1332

hMCC hMCC Thy1.1 Thy1.1 Thy1.1 hMCC

------

pUB pUB SCR shRNA pUB SCR shRNA pUB SCR shRNA pUB pUB

IB P-Bad P-p38

IB Bad p38

Rel A Bid P-ERK

P--catenin

Bik ERK -catenin

Cyclin D1 P-JNK P-HH3

Cyclin D2 JNK

Caspase 8

Cyclin B1 P-Akt

Cyclin A Akt Caspase 3

p21 c-Myc

Bcl-xL

p27 MCC Bcl2

Actin Mcl1 E2F1

p53

Bim

MCC

MCC Actin

Actin

Figure 6. MCC signaling pathways in human MM cells. The human MM cell line LP1 cells, which express high levels of endogenous MCC, were transduced with lentiviruses expressing MCC shRNA 1332 or a scrambled shRNA (SCR). Another human MM cell line 8226 cells, which express modest level of endogenous MCC, were transduced with lentiviruses expressing hMCC (pUB-hMCC) or Thy1.1 only (pUB-Thy1.1). Total cellular lysates were prepared and immunoblotted for known targets of MCC and regulators of apoptosis or cell cycle. Known target proteins of MCC examined include phosphorylated β-catenin (P-β-catenin), β- catenin, IκBα, IκBβ, RelA, and phosphorylated histone H3 (P-HH3). Regulators of apoptosis

examined include caspase 8, caspase 3, Bcl-xL, Bcl2, Mcl1, Bim, P-Bad, Bad, Bid, and Bik. Regulators of cell cycle examined include cyclin D1, cyclin D2, cyclin B1, cyclin A, p21, p27, E2F1, p53, phosphorylated-p38 (P-p38), p38, P-ERK, ERK, P-JNK, JNK, P-Akt, Akt, and c-Myc. The MCC blot was used as a control for MCC shRNA 1332 or pUB-hMCC transduction, and the actin blot was used as a loading control. Proteins that are changed by knockdown of MCC are highlighted in blue color, and those that are changed by both knockdown and overexpression of MCC are highlighted in red color. Results shown are representative of 3 independent experiments, and similar results were obtained in the human MM cell line KMS11 cells.

1.3.7 MCC INTERACTING PROTEINS IN HUMAN MM CELLS

To gain further insight into the mechanisms of MCC, we sought to identify

MCC interacting proteins. We first examined a number of previously known MCC interacting proteins identified in CRCs or 293T cells, including β-catenin, Mst3,

VCP, PP2A, DFFA, VHL, VDAC, scribble, and myosin IIb 61,67-69,96. None of these proteins were detected in co-immunoprecipitates with either FLAG-MCC or MCC-

SBP-6xHis in human MM cells (data not shown). We next employed an unbiased strategy: affinity purification followed by mass spectrometry-based sequencing.

We identified 365 proteins comprising the MCC-interactome of human MM cells:

333 proteins from whole cell lysates, and 207 proteins specifically in purified mitochondrial fractions (Fig. 7A and Table 2). Among these, 175 MCC interactors were identified in both whole cell lysates and purified mitochondria. However, only

39 proteins from our study were previously known MCC interactors identified in

CRCs or 293T cells (Fig. 7A and Table 2)61,67-69,94,96. Thus, 326 proteins identified in our study are novel MCC-interacting partners.

Disease association analyses using Ingenuity (http://www.ingenuity.com) identified cancer as the top disease associated with the MCC-interactome, with

195 MCC-interactors from whole cell lysates and 91 MCC-interactors from

mitochondria. Functional annotation and clustering analyses by Ingenuity revealed that mitochondrial MCC-interactors are mainly regulators of cell death and survival

(81 out of 207) (Fig. 7B). Similarly, MCC-interactors in whole cell lysates are mainly regulators of cell death and survival (123/333) or cellular growth and proliferation

(120/333) (Fig. 7C). Together, the MCC-interactome identified in our study is consistent with the prominent roles of MCC in promoting survival and proliferation in human MM cells (Figs. 4 and 5).

A B Previously known MCC interactors : 255 MCC interactors in whole lysates: 333

Cell Death and Survival (123) 18 Cellular Growth and Proliferation (120) 21 MCC interactors MCC interactors DNA Replication and Repair (39) in whole lysates : 333 in mitochondria: 207 Molecular Transport (34) 175 Protein Trafficking (17)

Extracted from: U:\RawDataArchive\Orbi Elite\2013-12\BRIMS\XiePing\From Sarracino Drive\MT-FLAG1-08.raw #11123 RT: 119.60 ITMS, [email protected], z=+2, Mono m/z=1113.07764 Da, MH+=2225.14800 Da, Match Tol.=1.2 Da

12 C y₁₀⁺ 1222.57

MCC interactors 10 in mitochondria: 207

Cell Death and Survival (81)

8 DNA Replication and Repair (39)

) b₁₀⁺-NH₃

3 Cell Cycle (38) ^ 0 1099.91

(

]

Molecular Transport (33) o 6

[

t y₁₁⁺

n 1321.62

Protein Trafficking (16) t

n I b₉⁺ y₇⁺ 1003.52 y₆⁺ 882.38 754.36 y₈⁺ 4 995.46 b₈⁺ y₅⁺ 904.57 b₁₁⁺ b₁₂⁺ 639.32 1230.75 1343.69 y₄⁺ y₁₂⁺ 510.21 1434.77 y₁₃⁺ y₁₄⁺ 2 1548.73 b₁₁²⁺ b₁₃⁺ 1663.84 b₇⁺ b₁₅⁺, y₁₅⁺-H₂O, y₁₅⁺-NH₃ 615.50 [M+2H]²⁺ 1471.85 b₁₄⁺ y₃⁺ 791.49 1716.87 y₁₂²⁺-NH₃ 1112.35 1587.87 423.31 b₉²⁺ b₅⁺ 710.40 b₁₇⁺-NH₃ y₆²⁺ 503.21 562.33 1934.10 378.15 0 E 400 600 800 1000 1200 1400 1600 1800 2000 m/z PHB2_Human: 64.21% coverage

Figure 7. The MCC-interactome in human MM cells identified by affinity purification followed by LC-MS/MS. The human MM cell line 8226 cells were transduced with lentiviruses expressing hMCC-SBP-6xHis or FLAG-hMCC. Immunoprecipitates of hMCC-SBP-6xHis by streptavidin-sepharose beads from whole cell lysates and purified mitochondria of 8226 cells were analyzed by high resolution LC- MS/MS, respectively. Immunoprecipitates of FLAG-hMCC by streptavidin-sepharose beads were used as negative control in these experiments. (A) Venn diagram of MCC-interactors identified in the present study and previous studies. The diagram depicts the number of previously known MCC- interactors or MCC-interactors identified in whole cell lysates or mitochondria of human MM cells in our study, as well as the overlap between each category. (B and C) Functional clustering and annotation of MCC-interactors identified in whole cell lysates (B) or mitochondria (C) of human MM cells. The number of proteins of each functional cluster is shown. (D) Representative mass spectral profile of a peptide of PHB2 identified by LC-MS/MS. The deduced peptide sequence is shown. (E) Schematic diagram of peptide sequences of PHB2 identified by LC-MS/MS. The peptide sequences detected by LC-MS/MS are highlighted in green color.

There are two isoforms of MCC, isoform 1 (828 amino acids in mouse and

829 amino acids in human) and isoform 2 (1019 amino acids), which differ at their extreme N-terminus due to alternative promoter usage. The MCC that we cloned from LP1 cells and used to generate FLAG-hMCC and hMCC-SBP-6xHis is isoform 1 (829 aa). Interestingly, our LC-MS/MS data identified 6 unique peptides of the isoform 2 of MCC (1019 aa) in the immunoprecipitates of hMCC-SBP-6xHis in both whole lysates and mitochondria (Fig. 8 and Table 2). Thus, the two isoforms of MCC form hetero-dimers or hetero-oligomers in human MM cells.

From the top 10 MCC-interactors, we selected two important regulators of survival and proliferation, PARP1 and prohibitin-2 (PHB2, Fig. 7D and 7E) 97-100 to further verify their interaction with MCC. Our results confirmed the interactions of

MCC with PARP1 and PHB2 in both whole cell lysates and purified mitochondria of human MM cells (Fig. 8A and 8B). We further performed an interaction network analysis using the top 100 MCC-interactors of human MM cells and the STRING analysis tool (http://www.string-db.org)101. As shown in Fig. 8C, 61 proteins of the top 100 MCC-interactors form an interaction network centered upon PARP1 and

PHB2. Interestingly, PARP1 and PHB2 have also been previously shown to directly or indirectly interact with and/or regulate MCC targets identified by knockdown and overexpression of MCC (Fig. 6), including phosphor-ERK, c-Myc, p27, cyclin B1, Mcl-1, caspase 8, and caspase 397-100. Collectively, our findings suggest that MCC promotes survival and proliferation by associating with and

modulating the interaction network centered at PARP1 and PHB2 in malignant B cells.

Figure 8. The MCC isoform 2 was identified as an MCC-interacting protein by affinity purification and LC-MS/MS. The two isoforms of MCC proteins differ at their extreme N-terminus due to alternative promoter usage. Both pUB-FLAG-hMCC and pUB-hMCC-SBP-6xHis are cloned from MCC isoform 1 (829 aa). (A) Schematic diagram of peptide sequences of MCC isoform 1 identified by LC-MS/MS. (B) Schematic diagram of peptide sequences of MCC isoform 2 (1019 aa) identified by LC- MS/MS. The peptide sequences identified by LC-MS/MS are highlighted in green. The unique region of MCC isoform 2 is marked with a dashed black box.

A Whole cell lysates B Mitochondria

Input SA IP Input SA IP

HIs HIs HIs HIs

- - - -

MCC MCC MCC MCC

- - - -

SBP SBP SBP SBP

- - - -

FLAG MCC FLAG MCC FLAG MCC FLAG MCC

PARP1 PARP1

PHB2 PHB2

SBP SBP

MCC-SBP-His MCC-SBP-His FLAG-MCC FLAG-MCC

MCC

Figure 9. PARP1 and PHB2 are two hubs of the MCC interaction network in human MM cells. The human MM cell line 8226 cells were transduced with pUB-hMCC-SBP-6xHis or pUB-FLAG- hMCC. Immunoprecipitation was performed as described in Figure 7. (A and B) Immunoblot analyses of proteins pulled down by streptavidin-sepharose beads (SA IP) from whole cell lysates (A) or purified mitochondria (B). Proteins were immunoblotted for PARP1, PHB2, SBP, and followed by MCC. Blots of an aliquot of whole cell or mitochondrial lysates before immunoprecipitation were used as the input control. (C) The MCC interaction network in human MM cells. We performed an interaction network analysis using the top 100 MCC-interactors and the STRING analysis tool (http://www.string-db.org). The top 100 MCC-interactors identified in our study and used for the interaction network analysis include 60 proteins identified in both whole cell lysates and mitochondria, 10 proteins identified in mitochondria only, and 30 proteins identified in whole cell lysates only. Sixty-one proteins of the top 100 MCC-interactors of human MM cells form an interaction network centered at PARP1 and PHB2.

1.4 DISCUSSION

Contrary to the striking up-regulation of MCC in B cell neoplasms identified in the present study, the MCC gene is frequently deleted, mutated, or silenced in other human cancers, including colorectal cancer47,49,58-61,63, lung cancer47,102, gastric carcinoma103, esophageal cancer104 and hepatocellular carcinoma 95,105.

Functional evidence and signaling pathway studies indicate that MCC acts as a tumor suppressor gene by inhibiting the oncogenic NF-κB and β-catenin pathways in CRCs and hepatocellular carcinoma50,51,61,67-69,94,95. We also previously observed regulation of β-catenin by Sox5, another gene identified in our microarray analysis, in human MM cells72. In our mouse B lymphoma cell lines, we were unable to successfully transduce them with lentiviral vectors in order to observe the effects of MCC knockdown or overexpression. However, NF-κB and β-catenin levels were not altered by knockdown or overexpression of MCC in human MM cells. Thus, MCC plays distinct roles via different signaling mechanisms in B cell malignancies versus other human cancers.

In CRCs, MCC mainly localizes in the cytoplasm, and is induced to shuttle into the nucleus in response to DNA damage51. Interestingly, we found that MCC is primarily localized at mitochondria in human MM cells. MCC does not contain any mitochondrial targeting motifs or transmembrane domains. Through delineation of the mitochondrial MCC-interactome, we identified a number of mitochondrial proteins that are associated with MCC in human MM cells, including

PHB2, prohibitin (PHB), ECHA, VDAC3, ADT1, ADT2, and ADT3. All of these mitochondrial proteins are known as critical regulators of cell survival and apoptosis (http://www.ingenuity.com). Therefore, MCC is primarily localized at mitochondria to promote survival by interacting with multiple mitochondrial proteins in malignant B cells.

PARP1 is the top novel MCC-interactor identified in our study. PARP1 catalyzes post-translational modification of proteins by polyADP-ribosylation, which affects protein-protein and protein-DNA interactions. In additional to its pivotal role in DNA repair, PARP1 also critically regulates transcription, cell survival, and proliferation97,99. PARP-1 can act in both pro- and anti-tumor manner depending on the context97. Here we identified PARP1 as a signaling hub of the

MCC-interactome in human MM cells. It is known to interact with numerous MCC interactors identified in our study, including CDK1, MCM4, XRCC6, RFC1, CCT2,

SSRP1, TOP2B, HIST1H1C, and SUPT16H (http://www.string-db.org).

Interestingly, PARP1 has been shown to directly or indirectly regulate the activation or expression of multiple MCC targets identified by our knockdown and overexpression experiments, including caspase 8, ERK, c-Myc, p27, cyclin B1, and

Mcl-1. Indeed, PARP-1 can induce the PARylation of caspase-8, thereby inhibiting apoptosis97. PARP1 directly interacts with phosphorylated ERK2 to mediate cell proliferation97,99. PARP-1 also down-regulates the expression of MKP-1, which dephosphorylates ERK97. Through modifications of chromatin or interactions with gene specific promoters/transcription factors, PARP1 regulates in total 3.5% of the

transcriptome, including increasing c-Myc97,99 and repressing p27106. Furthermore,

PARP1 can indirectly regulate the degradation of Mcl-1 through interaction with

CDK1/cyclin B1107,108. Therefore, PARP1 plays a central role in mediating the oncogenic effects of MCC in malignant B cells.

PHB2 is the top protein of the 39 previously known MCC-interactors. PHB2 is mainly localized in mitochondria by forming heteromeric complex with PHB, but also present in the cytosol, nucleus and plasma membrane98,100,109. PHB2 plays crucial roles in regulating mitochondrial function, cell survival, and proliferation98,100,109. Silencing of PHB2 induces apoptosis and reduces proliferation in a variety of cancer cells98,100,109. Here we identified PHB2 as another hub of the MCC interaction network. Association has been previously demonstrated between PHB2 and multiple MCC-interactors identified in our study, including PHB, CSNK1D (CK1δ), CSNK1E (CK1ε), ATP5B, NUMA1, and

SLC25A6 (http://www.string-db.org)98,110. PHB2 can also directly or indirectly regulate the activity or expression of multiple MCC targets identified in our study, including caspase 3, ERK, c-Myc, p27, and cyclin B1. For example, overexpression of PHB2 inhibits caspase-3 activation and cell apoptosis98,100,109, whereas down-regulation of PHB2 is associated with increased caspase-3 expression and cell apoptosis. PHB and PHB2 interact with c-Raf to induce the phosphorylation of ERK1/2 via MEK198. Through activation of ERK1/2, PHB2 may also indirectly regulate the expression levels of downstream targets of the ERK pathway, including c-Myc, p27 and cyclin B198,100,109,111. Thus, similar to PARP1,

PHB2 is likely to play essential roles in mediating the oncogenic effects of MCC in malignant B cells.

Interestingly, both PARP1 and PHB2/PHB are excellent therapeutic targets in cancer97,98. In preclinical studies, PARP inhibitors (such as olaparib) exhibit potent tumoricidal activities on breast cancer, ovarian cancer, pancreatic cancer, prostate cancer, and small cell lung carcinoma, among others. The efficacy of olaparib on BRCA-mutated breast cancer has been confirmed in early phase clinical trials with only mild adverse side effects97. Notably, PHB ligands (such as flavaglines and capsaicin) display robust cytotoxicity on cancer cells, but have cytoprotective activities on normal cells (e.g. neurons and cardiomyocytes), particularly against oxidative stress98. PHB1 and PHB2 were recently identified as the direct targets of flavaglines, natural products isolated from medicinal plants that show significant anticancer effects but no sign of toxicity in mice98. Capsaicin, a component of hot chili peppers, binds to PHB2 to induce apoptosis in human myeloid leukemia cells98. Identification of PARP1 and PHB2/PHB as hubs of the signaling pathways of MCC implicates potential use of PARP inhibitors and PHB ligands in the treatment of B cell malignancies involving aberrant expression of

MCC.

1.5 CONCLUSIONS

In summary, we identified MCC as a novel oncogene in B lymphocytes and provided insights into its signaling mechanisms in human MM cells. In the unique

cellular context of malignant B cells, MCC forms an interaction network centered at PARP1 and PHB2 to promote cellular survival and proliferation. The lack of expression of MCC in normal or premalignant B cells, coupled with its ubiquitous up-regulation in tumor B cells suggests that MCC may be a useful diagnostic marker for B cell neoplasms. Our finding that knockdown of MCC induced apoptosis and inhibited proliferation in human MM cells suggests that MCC may also serve as a therapeutic target in B cell malignancies. Furthermore, the central role of PARP1 and PHB2 in the MCC interaction network implies that PARP1 inhibitors and PHB ligands may have therapeutic application in B cell neoplasms, including NHL and MM.

CHAPTER 2 EXPRESSION AND FUNCTION OF A NOVEL ISOFORM OF SOX5 IN MALIGNANT B CELLS

ABSTRACT

Using a mouse model with the tumor suppressor TRAF3 deleted from B cells, we identified Sox5 as a gene strikingly up-regulated in B lymphomas. Sox5 proteins were not detected in normal or premalignant TRAF3−/−B cells even after treatment with B cell stimuli. The Sox5 expressed in TRAF3−/− B lymphomas represents a novel isoform of Sox5, and was localized in the nucleus of malignant

B cells. Overexpression of Sox5 inhibited cell cycle progression, and up-regulated the protein levels of p27 and β-catenin in human multiple myeloma cells. Together, our findings indicate that Sox5 regulates the proliferation of malignant B cells.

2.1 BACKGROUND

Identification and validation of new genetic risk factors are imperative for a better understanding of B lymphomagenesis and for the development of new therapeutic strategies. TRAF3 was recently identified as a novel tumor suppressor in human non-Hodgkin lymphoma, including splenic marginal zone lymphoma, B cell chronic lymphocytic leukemia, mantle cell lymphoma, as well as multiple myeloma (MM) and Waldenström's macroglobulinemia23,24,42,43,112. Using a new mouse model with the TRAF3 gene specifically deleted in B cells (B-

TRAF3−/− mice), we recently demonstrated that TRAF3 deletion causes prolonged survival of mature B cells, which eventually leads to B lymphoma development at

the age of 9–18 months15,16. The long latency of B lymphoma development observed in B-TRAF3−/− mice suggests that additional oncogenic alterations are required for B lymphomagenesis. To delineate such oncogenic alterations in

TRAF3−/− B lymphomas, we performed a microarray analysis and identified Sox5 as a gene strikingly up-regulated in B lymphomas spontaneously developed in different individual B-TRAF3−/− mice.

Sox5 is a member of the Sox family of transcription factors that contain a highly conserved sex-determining region (Sry)-related high-mobility-group (HMG) box, which mediates the binding to the minor groove of target DNA sequences 113.

Sox5 together with Sox6 and Sox13 constitute group D of Sox (SoxD) genes, which are expressed into several isoforms by alternative splicing 113. As a consequence, SoxD proteins exist in long and short isoforms. Only the long isoforms contain a characteristic N-terminal domain, including a leucine zipper, coiled-coil domains and a glutamine-rich region (Q-box), which allows them to homo-dimerize or hetero-dimerize with other SoxD proteins113 . The short isoform of Sox5 (S-Sox5) is expressed in the testis, brain, and lung, and likely plays a specialized role in testis development and ciliogenesis114,115. In contrast, the long isoform of Sox5 (L-Sox5) is expressed in multiple tissues, including the cartilage, heart, brain, kidney, lung, and skeletal muscle53,55,116,117. It has been shown that L-

Sox5 plays important roles in regulating diverse processes of embryonic development and cell fate determination, including chondrogenesis 118, notochord and joint development 56,119, neural crest generation 120, oligodendrogenesis 121,

melanogenesis 122, lung development 117, and the sequential generation of cortical neurons 123,124. Sox5 proteins achieve these developmental roles by modulating cell proliferation, survival, differentiation, or terminal maturation in different cell lineages 113.

Interestingly, corroborating our finding that Sox5 is strikingly up-regulated in TRAF3−/− mouse B lymphomas, up-regulation of SOX5 mRNA has also been identified in human memory B cells125, in clonal B cells of patients with hepatitis C virus (HCV)-associated B cell lymphoproliferative disorders mixed cryoglobulinemia 126, and in human patients with follicular lymphoma127. However, the role of Sox5 in B cell physiology and pathology remains unclear. The present study aimed to address this gap in knowledge. Specifically, the objectives of this study are: (1) to determine the expression of Sox5 in TRAF3−/− B lymphocytes and

B lymphomas; (2) to identify which isoform of Sox5 is expressed in TRAF3−/− B lymphomas; (3) to elucidate the function and mechanism of Sox5 in B cell malignancies.

2.2 MATERIALS AND METHODS

2.2.1 MICE AND CELL LINES

TRAF3flox/floxCD19+/Cre (B-TRAF3−/−) and TRAF3flox/flox (littermate control,

LMC) mice were generated as previously described. All mice were kept in specific pathogen-free conditions in the Animal Facility at Rutgers University, and were

used in accordance with NIH guidelines and under an animal protocol (Protocol #

08-048) approved by the Animal Care and Use Committee of Rutgers University.

Human MM cell lines 8226 (contains biallelic TRAF3 deletions) and LP1 (contains inactivating TRAF3 frameshift mutations) were generously provided by Dr. Leif

Bergsagel (Mayo Clinic, Scottsdale, AZ), and were cultured as previously described23. Mouse splenic B lymphocytes were prepared as previously described16.

2.2.2 ANTIBODIES AND REAGENTS

Rabbit Sox5 antibody (Ab) was purchased from Abcam (Cambridge, MA).

Rabbit Abs against p27, p21, cyclin D1, cyclin D2, c-Myc, Bcl-xL, Mcl1, and phosphorylated or total β-catenin or Akt were from Cell Signaling Technology

(Beverly, MA). Polyclonal rabbit Abs against TRAF1, p53 and HDAC1 were from

Santa Cruz Biotechnology (Santa Cruz, CA). Anti-actin Ab was from Chemicon

(Temecula, CA). HRP-labeled secondary Abs and affinity-purified (Fab′)2 goat anti- mouse IgM (μ-chain specific, anti-BCR) Abs were purchased from Jackson

ImmunoResearch Laboratories, Inc. (West Grove, PA). Hamster anti-mouse CD40

Abs were obtained from eBioscience (San Diego, CA). DNA oligonucleotide primers and CpG oligonucleotide 2084 (T*C*C*T*G*A*C*G*T*T*G*A*A*G*T; * denotes phosphorothioate bond) were obtained from Integrated DNA

Technologies (Coralville, IA). Lipopolysaccharides (LPS) and propidium iodide (PI) were purchased from Sigma–Aldrich Corp. (St. Louis, MO). Tissue culture

supplements including stock solutions of sodium pyruvate, l-glutamine, and non- essential amino acids and HEPES (pH 7.55) were from Invitrogen (Carlsbad, CA).

Pfu UltraII was purchased from Agilent (Santa Clara, CA).

2.2.3 TAQMAN ASSAY OF SOX5 MRNA EXPRESSION

Total cellular RNA was extracted using TRIzol reagent (Invitrogen) according to the manufacturer's protocol. Complementary DNA (cDNA) was prepared from RNA using High Capacity cDNA Reverse Transcription Kit (Applied

Biosystems). Quantitative real-time PCR of Sox5 was performed using TaqMan

Gene Assay Kit (Applied Biosystems) as previously described76.

2.2.4 PROTEIN EXTRACTION AND IMMUNOBLOT ANALYSIS

Total protein lysates, cytosolic and nuclear extracts were prepared as previously described16. Immunoblot analysis was performed using various antibodies as described128. Images of immunoblots were acquired using a low-light imaging system (LAS-4000 mini, FUJIFILM Medical Systems USA, Inc., Stamford,

CT).

2.2.5 CLONING OF THE FULL-LENGTH CDNA OF THE SOX5 GENE FROM TRAF3−/− MOUSE B LYMPHOMAS

Total cellular RNA was prepared from B lymphomas spontaneously developed in four individual B-TRAF3−/−mice, and the corresponding cDNA samples were used as templates to clone the Sox5 coding sequences as detailed in Supplementary Materials and Methods.

2.2.6 GENERATION OF LENTIVIRAL SOX5 EXPRESSION VECTORS

The Sox5 coding cDNA sequence cloned from TRAF3−/− mouse B lymphomas (Sox5-BLM) and the L-Sox5 cDNA expressed in other tissues

(purchased from Open BioSystems, Pittsburgh, PA) were subcloned into the lentiviral expression vector pUB-eGFP-Thy1.1109 (kindly provided by Dr. Zhibin

Chen, the University of Miami, Miami, FL) by replacing the eGFP coding sequence with the Sox5 coding sequences, respectively. For subcloning into pUB-eGFP-

Thy1.1 digested with BamH1 and XbaI, the BamH1 and XbaI restriction enzyme sites were added to flank the Sox5-BLM or L-Sox5 coding sequence by PCR using primers mSox5 5′ BamH1 and mSox5-R-XbaI (sequences detailed in Table 1).

Each lentiviral vector was verified by DNA sequencing.

2.2.7 LENTIVIRAL PACKAGING AND TRANSDUCTION OF HUMAN MULTIPLE MYELOMA CELLS

Lentiviruses of Sox5 expression vectors were packaged and titered as previously described109,129. Human multiple myeloma cells were transduced with the packaged lentiviruses at a MOI of 1:5 (cell:virus) in the presence of 8 μg/ml polybrene109. At 96 h post transduction, the transduction efficiency of cells was analyzed by Thy1.1 immunofluorescence staining and flow cytometry. Transduced cells were subsequently analyzed for Sox5 protein expression. Cell cycle distribution of transduced cells was determined by propidium iodide (PI) staining followed by flow cytometry as previously described16,82.

2.2.8 IMMUNOSTAINING AND CONFOCAL IMAGING

Immunofluorescence staining of Sox5 and confocal imaging were performed as described in Supplementary Materials and Methods.

2.2.9 STATISTICS

For cell cycle analysis experiments, statistical significance was assessed by Student's t-test. P values less than 0.05 are considered significant.

2.3 RESULTS

2.3.1 STRIKING UP-REGULATION OF SOX5 IN TRAF3−/− B LYMPHOMAS

To delineate secondary oncogenic alterations in TRAF3−/− mouse B lymphomas, we performed a microarray analysis130 and identified Sox5 as a strikingly up-regulated gene. We first verified the transcriptional up-regulation of

Sox5 in splenic B lymphomas and ascites spontaneously developed in 6 different individual B-TRAF3−/− mice using TaqMan gene expression assay (Fig. 10A). We also verified the up-regulation of Sox5 at the protein level using Western blot analysis (Fig. 10B). Interestingly, only the long isoform of the Sox5 protein (MW:

∼80 kDa), but not the short isoform (MW: ∼48 kDa), was detected and up- regulated in TRAF3−/− B lymphomas.

Figure 10. Up-regulation of Sox5 expression in TRAF3−/− mouse B lymphomas. (A) Quantitative real time PCR analyses of the Sox5 transcript. Total cellular RNA was prepared from splenocytes of LMC mice, or splenic B lymphomas and ascites of diseased B-TRAF3−/− mice. Real time PCR was performed using TaqMan primers and probes (FAM-labeled) specific for mouse Sox5. Each reaction also included the probe (VIC-labeled) and primers for β-actin, an endogenous control. Graphs depict the results of two independent experiments with duplicate reactions in each experiment (mean ± S.D.). (B) Western blot analysis of the Sox5 protein. Total cellular proteins were prepared from purified LMC splenic B cells or splenic B lymphomas or ascites of different individual B-TRAF3−/− mice. Proteins were immunoblotted for Sox5, followed by actin. (C and D) Potential regulation of the expression of the Sox5 gene in response to B cell stimuli. Splenic B cells were purified from 10- to 12-week-old LMC and tumor-free B-TRAF3−/− mice. Purified B cells were cultured ex vivo in the absence/ presence of stimuli of B cell survival, proliferation, and activation, including: 2 μg/ml anti-CD40, 20 μg/ml LPS, 1 μg/ml anti-BCR, and 100 nM CpG2084, alone or in combination. RNA and protein samples of primary TRAF3−/− mouse B lymphomas (mouse ID: 7060-8 and 6983-2) were used as positive control of Sox5 mRNA and protein. (C) Total cellular RNA was prepared 6 h after treatment. Taqman assay of Sox5 was performed as described in (A). Graphs depict the results of two independent experiments with duplicate reactions in each (mean ± S.D.). (D) Total cellular proteins were prepared 24 h after treatment, and immunoblotted for Sox5, followed by TRAF1 (successful stimulation control) and actin (loading control).

We then investigated the potential involvement of Sox5 up-regulation in the survival, proliferation and activation of B lymphocytes. Splenic B cells were purified from LMC and tumor-free young B-TRAF3−/− mice (age: 10–12 weeks), and then stimulated with a variety of B cell stimuli. These include agonistic anti-CD40 Abs,

LPS (TLR4 agonist), anti-B cell receptor (BCR) crosslinking Abs, and CpG2084

(TLR9 agonist), alone or in combination. We found that the transcript of Sox5 was modestly up-regulated by the combined treatment with CpG and CD40 in premalignant TRAF3−/− B cells, but not induced in LMC B cells or by other treatment (Fig. 10C). Interestingly, Sox5 proteins were not detectable in normal

LMC or premalignant TRAF3−/− B cells after treatment with any examined B cell stimuli, although TRAF1 proteins were potently induced by these stimuli (Fig. 10D).

Thus, Sox5 protein was only up-regulated and detected in TRAF3−/− B lymphoma cells.

2.3.2 A NOVEL ISOFORM OF SOX5 WAS EXPRESSED IN TRAF3−/− B LYMPHOMAS

Three different variants of mouse L-Sox5 transcripts have been reported in the literature and GenBank databases53,114,116. To identify which isoform of Sox5 was expressed in TRAF3−/− mouse B lymphomas, we cloned the full-length Sox5 coding cDNA from B lymphomas of 4 different individual B-TRAF3−/− mice using reverse transcription and PCR as described (Tables 3–5). Surprisingly, our sequencing data revealed that the Sox5 cDNA cloned from TRAF3−/− mouse B lymphomas represents a novel isoform of mouse Sox5 (Sox5-BLM), which is

distinct from previously reported mouse Sox5 isoforms (Fig. 11). We thus submitted the sequence of Sox5-BLM to GenBank database (accession number:

KF793916). Sox5-BLM contains a 35 amino acid (aa) deletion in the N-terminal region in front of the leucine zipper domain. Although a similar 35 aa deletion is also present in Sox5 variant 3 (Sox5-V3), the latter has an additional deletion of

49 aa between the first and the second coiled-coil domains. Examination of the exon and intron structure of the mouse Sox5 gene revealed that this novel isoform,

Sox5-BLM, is likely generated by alternative splicing (Fig. 12).

Figure 11. A novel isoform of Sox5 was expressed in TRAF3−/− mouse B lymphomas. The protein sequence encoded by the Sox5 cDNA cloned from TRAF3−/− mouse B lymphomas (Sox5-BLM, 728 aa) is aligned with those of the long isoform of Sox5 (L-Sox5, 763 aa, accession no.: NM_011444.3), the short isoform of Sox5 (S-Sox5, 392 aa, accession no.: X65657.1), Sox5 variant 2 (Sox5 V2, 715 aa, accession no.: NM_001113559.2), and Sox5 variant 3 (Sox5 V3, 679 aa, accession no.: NM_001243163.1). Similar to Sox5 variant 3, Sox5-BLM contains a 35 aa deletion in the N-terminal region in front of the leucine zipper domain. However, Sox5 variant 3, but not Sox5-BLM, has an additional 49 aa deletion between the two coiled-coiled domains. The first coiled-coil domain (1st CC) is indicated with a solid black box, the second coiled-coil domain (2nd CC) is indicated with a dotted black box, the high mobility box (HMG) domain is marked with a dashed black box, the leucine zipper (LZ) domain is shown in a dashed gray box, and the glutamine-rich domain (Q box) is highlighted with a solid gray box.

Figure 12. Distinct isoforms of the mouse Sox5 gene generated by alternative splicing. The mouse Sox5 gene locus consists of 15 exons spanning ~383 kb of genomic DNA, which gives rise to several splice variants. There are four Sox5 transcription products described in the literature and GenBank database, including the long isoform of Sox5 (L-Sox5, NM_011444.3), Sox5 variant 2 (Sox5 V2, NM_001113559.2), Sox5 variant 3 (Sox5 V3, NM_001243163.1), and the short isoform of Sox5 (S-Sox5, X65657.1). All 15 exons comprise the L-Sox5 transcript. The Sox5 V2 transcript is almost identical to that of L-Sox5, except that it lacks exon 9. The Sox5 V3 transcript is almost identical to Sox5 V2, except that it lacks 105 bp of exon 2 due to usage of an alternative splicing site within that exon. The transcript of the novel isoform of Sox5, Sox5-BLM cloned from TRAF3-/- mouse B lymphomas in the present study, is almost identical to Sox5 V3, except that it contains exon 9. The S-Sox5 transcript contains exons 8, and 10-15, and an alternative exon 1 (exon 1S), located in the introns between exons 7 and 8.

To further determine whether other known Sox5 transcript variants were present in TRAF3−/− mouse B lymphomas, we designed multiple pairs of PCR primers flanking the alternative splice sites of Sox5 isoforms (Table 3). We did not detect any transcript expression of L-Sox5, Sox5-V2, or S-Sox5 by PCR (Tables 4 and 6). Interestingly, we observed low level of expression of the Sox5-V3 transcript in TRAF3−/− mouse B lymphomas (Table 6). Thus, our results demonstrated that although Sox5-V3 transcript is also present, the novel isoform (Sox5-BLM) is the predominant transcript expressed in TRAF3−/− mouse B lymphomas.

To generate research tools for transduction of human B cell lines, we constructed lentiviral expression vectors using the Sox5-BLM cDNA cloned from

TRAF3−/− mouse B lymphomas and the L-Sox5 cDNA expressed in other tissues, respectively. We use these vectors to transduce human patient-derived multiple myeloma cell lines 8226 and LP1 cells, which contain TRAF3 deletions or inactivating mutations and do not express endogenous Sox5 proteins (Fig. 13). Figure 13. SOX5 is not expressed in human multiple myeloma cell lines 8226 and LP1. (A) Quantitative real time PCR analyses of the human SOX5 transcript. Total cellular RNA was prepared from human multiple myeloma cell lines 8826 and LP1, or the human neuroblastoma cell line SH-SY5Y. Total cellular RNA of human induced pluripotent stem cells (iPS) was generously provided by Dr. Ronald Hart. Real time PCR was performed using TaqMan primers and probes (FAM-labeled) for human SOX5. Each reaction also included the probe (VIC-labeled) and primers for human β-actin (endogenous control). Graphs depict the results of two independent experiments with duplicate reactions in each (mean = S.D.) (B) Western blot analysis of the SOX5 protein. Total cellular proteins were prepared from 8226, LP1 or SH-Sy5Y cells. Proteins were immunoblotted for SOX5, then actin. Results shown represent two independent experiments.

Our flow cytometric data demonstrated that the lentiviral transduction efficiency

was >80% in human multiple myeloma cells (Fig. 14A). We found that the cloned

Sox5-BLM was expressed into a protein of ∼80 kDa, a size identical to that

detected in primary TRAF3−/− mouse B lymphomas but smaller than that of L-Sox5

(Fig. 14B). Thus, the transduction data further verified that the novel isoform, Sox5-

BLM, is the predominant isoform expressed in TRAF3−/− mouse B lymphomas.

Figure 14. Overexpression of Sox5-BLM inhibited cell cycle progression in human multiple myeloma cells 8226 cells (HMCL) were transduced with lentiviral expression vectors of Sox5-BLM (pUB-Sox5-BLM) or L-Sox5 (pUB-L-Sox5). Cells transduced with an empty lentiviral expression vector (pUB-Thy1.1) were a control. (A) Transduction efficiency analyzed by Thy1.1 staining and flow cytometry. Gated population (Thy1.1+) indicates successfully transduced cells. (B) Expression of the transduced Sox5 proteins analyzed by Western blot analysis. Total cellular proteins were prepared 4 days post transduction, and immunoblotted for Sox5, followed by actin (loading control). MW of Sox5 proteins expressed in transduced 8226 cells were compared to those expressed in primary TRAF3−/− mouse B lymphomas. (C) Representative histograms of cell cycle analysis. Cell cycle distribution of transduced 8226 cells was analyzed by PI staining and flow cytometry at 5 days post transduction. Gated populations indicate apoptotic cells (sub-G0: DNA content < 2n) and proliferating cells (S/G2/M phase: 2n < DNA content ≤ 4n). (D) Statistical analysis of PI staining results. The graph depicts the results of 3 independent experiments (mean ± S.D.). p values were analyzed using Student's t test, * significantly different from pUB-Thy1.1 control (t test, p < 0.05). Similar results were obtained in LP1 cells, another human multiple myeloma cell line.

2.3.3 OVEREXPRESSION OF SOX5-BLM INHIBITED CELL CYCLE PROGRESSION IN TRANSDUCED HUMAN MULTIPLE MYELOMA CELLS

To explore the functional roles of Sox5 in the survival and proliferation of malignant B cells, we analyzed the cell cycle distribution of human multiple myeloma 8226 and LP1 cells transduced with lentiviral expression vectors of Sox5-

BLM or L-Sox5 using PI staining and FACS analysis. Cells transduced with an empty lentiviral expression vector (pUB-Thy1.1) were used as control in these experiments. We found that overexpression of either Sox5-BLM or L-Sox5 resulted in a significant decrease in the proliferating cell population (S/G2/M phase:

2n < DNA content ≤ 4n) and an increase in the apoptotic cell population (DNA content < 2n) of transduced 8226 or LP1 cells (Fig. 14C and D, and data not shown). These results indicate that Sox5 plays a role in regulating cell cycle progression in malignant B cells.

2.3.4 SOX5-BLM WAS LOCALIZED IN THE NUCLEUS OF TRAF3−/− MOUSE B LYMPHOMAS AND TRANSDUCED HUMAN MULTIPLE MYELOMA CELLS

Previous studies have shown that Sox5 is localized in the nucleus of chondrocytes, neurons, and oligodendrocytes113. To verify whether this is case and to elucidate where Sox5 exerts its roles in regulating cell cycle progression in malignant B cells, we examined the subcellular localization of Sox5 using a biochemical fractionation method. Cytosolic and nuclear extracts were prepared from mouse splenic B cells, TRAF3−/− mouse B lymphomas, or transduced human

multiple myeloma cells. We observed that Sox5-BLM protein was primarily localized in the nucleus of TRAF3−/− mouse B lymphomas (Fig. 15A).

Figure 15. Predominant nuclear localization and signaling pathways of Sox5-BLM in malignant B cells (A) Nuclear localization of endogenous Sox5 in primary TRAF3−/− mouse B lymphomas. Splenic B cells were purified from LMC (samples 1–3) or tumor-free young B-TRAF3−/− mice (samples 4– 6), or splenic B lymphomas of diseased B-TRAF3−/− mice (samples 7–8). (B) Nuclear localization of transduced Sox5 in human multiple myeloma cells. 8226 cells (HMCL) were transduced with lentiviral expression vectors of Sox5 (pUB-Sox5-BLM or pUB-L-Sox5) or a control vector (pUB- Thy1.1). Cytosolic and nuclear extracts were fractionated using a biochemical method, and immunoblotted for Sox5, followed by HDAC1 (nuclear protein loading control) and actin (cytosolic protein loading control). (C) Up-regulation of p27 and β-catenin protein levels by Sox5 overexpression in human multiple myeloma cells. Human multiple myeloma cell line 8226 cells were transduced with lentiviral expression vectors of Sox5 (pUB-Sox5-BLM or pUB-L-Sox5) or a control lentiviral expression vector (pUB-Thy1.1). Total cellular lysates were prepared and immunoblotted for p27, p21, cyclin D1, cyclin D2, phosphorylated β-catenin (Phospho-β-catenin), β-catenin, phosphorylated-Akt (Phospho-Akt), Akt, c-Myc, Bcl-xL, and Mcl1, followed by Sox5 (transduction control) and actin (loading control). Results shown are representative of two independent experiments, and similar results were also observed in LP1 cells.

Similarly, both Sox5-BLM and L-Sox5 were also predominantly localized in the nucleus of transduced human multiple myeloma 8226 and LP1 cells (Fig. 15B and data not shown). We also used an immunostaining method to verify the predominant nuclear localization of Sox5-BLM and L-Sox5 in transduced human multiple myeloma cells (Fig. 16). These data suggest that Sox5 may function as a transcription factor or transcriptional regulator in malignant B cells.

Figure 16. Sox5 is primarily localized in the nucleus in malignant B cells. LP1 (HMCL) cells were transduced with lentiviral expression vectors of Sox5 (pUB-Sox5-BLM or pUB-L-Sox5) or a control vector (pUB-Thy1.1). Transduced cells were cultured on poly-L-Lysine coated coverslips. Immunostaining and confocal imaging were performed. Green fluorescence (FITC) shows immunostaining of Sox5, red fluorescence of actin by rhodamine phalloidin, and cell nuclei are counterstained with Hoechst 33342, which gives blue fluorescence. Images shown are representative of two independent experiments with duplicate samples in each experiment, and similar results were also observed in 8226 cells, another human multiple myeloma cell line.

2.3.5 OVEREXPRESSION OF SOX5-BLM UP-REGULATED THE PROTEIN LEVELS OF P27 AND Β-CATENIN IN TRANSDUCED HUMAN MULTIPLE MYELOMA CELLS

To understand the mechanisms of Sox5-mediated regulation of cell cycle progression, we measured the protein levels of a number of cell cycle and survival

regulators in human multiple myeloma 8226 and LP1 cells transduced with Sox5-

BLM or L-Sox5. We found that overexpression of either Sox5-BLM or L-Sox5 led to specific up-regulation of the protein levels of the cell cycle inhibitor p27, but not other cell cycle and survival regulators examined, including p21, cyclin D1, cyclin

D2, cyclin B1, c-Myc, p53, Bcl-xL, Mcl1, Bcl2, Bim, and Bad (Fig. 15C and data not shown). In light of the evidence that Sox5 regulated p27 protein levels via the Akt or β-catenin pathway 120,131, we further examined the levels of phosphorylated and total Akt and β-catenin in transduced human multiple myeloma cells. We found that overexpression of either Sox5-BLM or L-Sox5 caused robust up-regulation of the total protein levels of β-catenin, but not Akt (Fig. 15C). Thus, Sox5-BLM and

L-Sox5 appear to function equivalently in malignant B cells. Furthermore, our findings suggest that Sox5 inhibits cell cycle progression by up-regulating p27 and

β-catenin protein levels in human multiple myeloma cells.

2.4 DISCUSSION

Intragenic and complete deletions of the SOX5 gene, either inherited or de novo, have recently been identified as causal genetic alterations in human patients with developmental delay, intellectual disability, and behavior abnormality132-134. In contrast, amplification of the SOX5 gene has been reported in human prostate cancer 135 and testicular seminomas136. Additionally, up-regulation of SOX5 has been shown in several other human cancer types, including nasopharyngeal

carcinoma137 and melanoma138. Contradictory up-regulation 139 and down- regulation131,140 and of SOX5 expression have been documented in human glioma.

In the present study, we demonstrated the striking up-regulation of Sox5 expression in primary TRAF3−/−mouse B lymphomas. Expression of Sox5 proteins was not detected in normal splenic B cells purified from LMC mice or premalignant

B cells purified from tumor-free young B-TRAF3−/− mice, even after treatment with a variety of stimuli of B cell survival, proliferation and activation. This suggests that aberrant up-regulation of Sox5 is associated with B lymphomagenesis. However, we did not detect SOX5 protein expression in human multiple myeloma cell lines

8226 and LP1, which contain bi-allelic deletions and frameshift mutations of the

TRAF3 gene respectively. The striking up-regulation of Sox5 in TRAF3−/− mouse

B lymphomas together with the absence of SOX5 in human multiple myeloma cell lines with TRAF3 deletions/mutations prompted us to further examine SOX5 expression in other malignant murine and human B cells. We thus surveyed the gene expression omnibus (GEO) database and the public gene expression database of human cancers (http://www.oncomine.org), and found that SOX5 expression is also aberrantly and significantly up-regulated in a variety of other murine and human B cell malignancies. These include diffuse large B cell lymphoma developed in Brd2-transgenic mice (GEO accession number: GSE6136)141, multiple myeloma developed in XBP-1 transgenic mice

(GEO accession number: GSE6980)142, human follicular lymphoma143, and human hairy cell leukemia (a sub-type of chronic lymphoid leukemia of B lymphocytes)89.

Hence, up-regulation of SOX5 appears to be frequently associated with B cell malignant transformation. How Sox5 is up-regulated during TRAF3−/− B lymphomagenesis remains unclear. Using Taqman gene copy number assay, we found that the Sox5 gene copy number was not altered in TRAF3−/− mouse B lymphomas (data not shown). Consistent with our observation, the SOX5 gene copy number is not changed in human multiple myeloma144,145 and marginal zone lymphoma either146. Interestingly, up-regulation of SOX5 expression in a case of human follicular lymphoma is due to a chromosomal translocation t(X;12), which fuses the promoter region of the G-protein coupled purinergic receptor P2Y8 gene with the SOX5 coding sequence127. Future study is required to investigate whether similar chromosomal translocation involving Sox5 occurs in TRAF3−/− mouse B lymphomas. Alternatively, transcriptional activation of Sox5 may be acquired by activation or up-regulation of its upstream transcription factors, inactivation or down-regulation of its upstream transcriptional repressors, or alterations in epigenetic modifications during TRAF3−/− B lymphomagenesis. Such mechanisms need to be further elucidated.

By cloning the coding sequence of the Sox5 cDNA expressed in

TRAF3−/− mouse B lymphomas, we found that it represents a novel isoform of

Sox5, Sox5-BLM. Using subcellular fractionation, we found that this new isoform of Sox5 was localized in the nucleus of TRAF3−/− mouse B lymphoma cells and transduced human multiple myeloma cells. Interestingly, in contrary to an oncogenic role as predicted by its striking up-regulation in TRAF3−/− mouse B

lymphomas, we found that overexpression of Sox5-BLM or L-Sox5 inhibited cell cycle progression in transduced human multiple myeloma cells. Furthermore, we found that overexpression of Sox5-BLM or L-Sox5 led to up-regulation of the cell cycle inhibitor p27 and another cell cycle regulator β-catenin in human multiple myeloma cells. These results suggest that Sox5 may play a tumor suppressive role in malignant B cells and that Sox5 expression may be up-regulated as a counteracting mechanism during B lymphomagenesis. Alternatively, it remains possible that Sox5 is oncogenic in B lymphomagenesis, but human multiple myeloma cells do not provide permissive cellular context to allow Sox5 to exert its transforming activity. Similar phenomena have been previously observed for the potent oncogene c-Myc147-149. When introduced into non-permissive cells, c-Myc induces cell cycle arrest, senescence, or apoptosis147-149. A third possibility is that murine Sox5 may not be able to interact with human SOX5-interacting factors in a similar way as with murine factors, or murine Sox5 may even act as dominant negative factor in the human system, thereby causing different functional consequence due to inter-species incompatibilities. In this regard, we noticed that the protein sequence of murine L-Sox5 is 97% (740/763 aa) but not 100% identical to that of human L-SOX5. Nonetheless, our findings suggest that Sox5 plays a role in regulating the proliferation of malignant B cells, which could be negative or positive depending on the cellular context.

Corroborating our findings, both oncogenic and tumor suppressive roles have been reported for SOX5. In nasopharyngeal carcinoma, up-regulation of

SOX5 appears to promote tumor progression by down-regulating the expression of the tumor suppressor gene SPARC137. In contrast, Sox5 suppresses PDGFB- induced glioma development in Ink4a−/− mice131. Interestingly, increased levels of p27 upon Sox5 expression were also observed in PDGFB-induced glioma of

Ink4a−/− mice and in chicken developing neurons120,131. However, overexpression of Sox5 results in a decrease in the levels of the dephosphorylated active form of

β-catenin in chicken developing neurons 120, but an increase in the levels of β- catenin in human multiple myeloma cells (Fig. 15C). Other cell cycle regulators downstream of Sox5 identified in previous studies, including cyclin D1, cyclin D2,

Akt, and Myc120,131, are not changed by Sox5 overexpression in human multiple myeloma cells. Future studies are required to decipher how Sox5 up-regulates the protein levels of p27 and β-catenin in malignant B cells. Sox5 does not have a transactivation domain113. It has been shown that Sox5 may promote or repress gene expression by interacting with SoxE proteins (such as Sox9 and Sox10), by recruiting transcriptional repressors (such as CtBP and HDAC1), by competing with other Sox proteins (such as Sox4 and Sox11) for target sites, or indirectly by regulating chromatin architecture113,150,151. Regardless of the exact mechanisms, p27 and β-catenin appear to be common downstream targets of Sox5 in regulating cell cycle progression of neurons, glioma, and multiple myeloma.

2.5 CONCLUSIONS

In summary, we found that a novel isoform of Sox5, Sox5-BLM, was aberrantly expressed in B lymphomas spontaneously developed in different individual B-TRAF3−/− mice. This new isoform of Sox5 was localized in the nucleus of TRAF3−/− mouse B lymphomas and transduced human multiple myelomas, and able to inhibit the proliferation of human multiple myeloma cells by up-regulating the protein levels of p27 and β-catenin.

CHAPTER 3. N-BENZYLADRIAMYCIN-14-VALERATE (AD 198) EXHIBITS POTENT ANTI-TUMOR ACTIVITY ON TRAF3-DEFICIENT MOUSE B LYMPHOMA AND HUMAN MULTIPLE MYELOMA

ABSTRACT

TRAF3, a new tumor suppressor identified in human non-Hodgkin lymphoma (NHL) and multiple myeloma (MM), induces PKCδ nuclear translocation in B cells. In this study we aimed to evaluate the therapeutic potential of two PKCδ activators, N-Benzyladriamycin-14-valerate (AD 198) and ingenol-3-angelate

(PEP005), on NHL and MM. In vitro tumoricidal activities of AD 198 and PEP005 were determined using TRAF3-/- mouse B lymphoma and human patient-derived

MM cell lines as model systems. In vivo therapeutic effects of AD 198 were assessed using NOD SCID mice transplanted with TRAF3-/- mouse B lymphoma cells. Biochemical studies were performed to investigate signaling mechanisms induced by AD 198 or PEP005, including subcellular translocation of PKCδ. We found that AD 198 exhibited potent in vitro and in vivo tumoricidal activity on

TRAF3-/- tumor B cells, while PEP005 displayed contradictory anti-tumor or pro- tumor activities on different cell lines. Detailed mechanistic investigation revealed that AD 198 did not affect PKCδ nuclear translocation, but strikingly suppressed c-

Myc expression and inhibited the phosphorylation of ERK, p38 and JNK in TRAF3-

/- tumor B cells. In contrast, PEP005 activated multiple signaling pathways in these cells, including PKCα, PKCδ, PKCε, NF-κB1, ERK, JNK, p38, and Akt.

Furthermore, we found that AD198 also potently inhibited the viability and

suppressed c-Myc expression in TRAF3-sufficient mouse and human B lymphoma cell lines. Our findings uncovered a novel, PKCδ-independent mechanism of the anti-tumor effects of AD 198, and suggest that AD 198 has therapeutic potential for the treatment of NHL and MM involving TRAF3 inactivation or c-Myc up- regulation.

3.1 BACKGROUND

TRAF3, a member of the TRAF family of cytoplasmic adaptor proteins was first identified as an interacting protein shared by CD40 (a receptor pivotal for B cell activation) and LMP1 (an Epstein-Barr virus-encoded oncogenic protein) 17.

TRAF3 also binds to receptors for the critical B cell survival factor BAFF, including

BAFF-R, TACI and BCMA. By characterizing mice that have the Traf3 gene specifically deleted in B lymphocytes (B-TRAF3-/- mice), we found that resting splenic B cells from these mice show increased levels of active NF-κB2 but decreased levels of nuclear PKCδ 16,40. Using B lymphoma cells derived from B-

TRAF3-/- mice as model systems, we demonstrated that oridonin, a pharmacological inhibitor of NF-κB, and lentiviral vectors of NF-κB2 shRNAs induced apoptosis in cultured TRAF3-/- B lymphoma cells 15. These studies identified constitutive NF-κB2 activation as one oncogenic pathway in TRAF3-/- B cells.

Interestingly, available evidence suggests that the second signaling pathway downstream of TRAF3, the reduced PKCδ nuclear translocation, may also contribute to prolonged B cell survival. First, the splenic B cell compartment of PKCδ-/- mice is greatly expanded 44,45, similar to that observed in B-TRAF3-/- mice 16,40, and BAFF or NF-κB2 transgenic mice 152,153. Second, the physiological

B cell survival factor, BAFF, also reduces PKCδ nuclear levels in splenic B cells154.

In light of these observations, we sought to evaluate the therapeutic potential of

PKCδ activation in TRAF3-/- tumor B cells using two pharmacological activators of

PKCδ, N-Benzyladriamycin-14-valerate (AD 198) and ingenol-3-angelate

(PEP005)155-160.

3.2 MATERIALS AND METHODS

3.2.1 MICE

TRAF3flox/floxCD19+/Cre (B-TRAF3-/-) and TRAF3flox/flox (littermate control,

LMC) mice were generated as previously described16. NOD SCID mice (Jackson

Laboratory, stock number: 001303, strain name: NOD.CB17-Prkdcscid/J) were used as recipients in B lymphoma transplantation and in vivo drug treatment experiments. All mice were kept in specific pathogen-free conditions in the Animal

Facility at Rutgers University, and were used in accordance with NIH guidelines and under an animal protocol (Protocol # 08–048) approved by the Animal Care and Use Committee of Rutgers University.

3.2.2 CELL LINES AND CELL CULTURE

Human MM cell lines 8226 (contains bi-allelic TRAF3 deletions), KMS11

(contains bi-allelic TRAF3 deletions) and LP1 (contains inactivating TRAF3 frameshift mutations) were generously provided by Dr. Leif Bergsagel (Mayo

Clinic, Scottsdale, AZ). Human B lymphoma cell lines Daudi (Burkitt’s lymphoma),

Ramos (Burkitt’s lymphoma), and JeKo-1 (mantle cell lymphoma) were obtained from American Type Culture Collection (ATCC, Manassas, VA). All human MM and B lymphoma cell lines were cultured as previously described23. Mouse B lymphoma cell lines A20.2 J and CH12.LX were generously provided by Dr. Gail

Bishop (University of Iowa, Iowa City, IA), and m12.4.1 was purchased from ATCC.

All mouse B lymphoma cell lines were cultured as we described 15. Generation of

TRAF3−/− mouse B lymphoma cell line 27–9.5.3 was described previously 15.

Mouse B lymphoma cell line 105–8.1B6 was generated from ascites harvested from a B-TRAF3−/− mouse (mouse ID: 105–8) 15. Briefly, ascitic cells (5

× 105 plated in 24-well plates in mouse B cell medium 16 containing 10% fetal bovine serum. After being cultured for 2 months, 4 actively proliferating clones were expanded, passaged, and frozen. The 105–8.1B6 clone had been cultured for 5 months without obvious changes in morphology or growth rate, and was used for drug treatment experiments.

Mouse B lymphoma cell line 115–6.1.2 was derived from splenic B lymphoma of another B-TRAF3−/− mouse (mouse ID: 115–6) 15. Briefly, Primary splenic B lymphoma cells harvested from mouse 115–6 were serially passaged in

NOD SCID mice twice. B lymphoma cells (5 × 105 plates in mouse B cell media containing 10% fetal bovine serum. After being cultured for 1 month, 8 actively proliferating clones were expanded, passaged, and frozen. The 115–6.1.2 clone had been cultured for 5 months without obvious changes in morphology or growth rate, and was used for drug treatment experiments.

3.2.3 ANTIBODIES AND REAGENTS

Polyclonal rabbit Abs against RelB, NF-κB1, RelA, c-Rel, HDAC1, and

PKCδ were purchased from Santa Cruz Biotechnology (Santa Cruz, CA).

Polyclonal rabbit Abs specific for NF-κB2, c-Myc, phosphorylated PKCδ, PKCα,

PKCε, caspase 3, and COX IV, and Abs against total or phosphorylated ERK, p38,

JNK, and Akt, were from Cell Signaling Technology (Beverly, MA). Anti-actin Ab was from Chemicon (Temecula, CA). HRP-labeled secondary Abs were purchased from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA).

Tissue culture supplements including stock solutions of sodium pyruvate, L- glutamine, and non-essential amino acids and Hepes (pH 7.55) were from

Invitrogen (Carlsbad, CA). Oridonin was purchased from CalBiochem (EMD

Chemicals, Gibbstown, NJ). AD 198, PEP005, 3-(4,5-dimethylthiazol-2-yl)-2,5- diphenyltetrazolium bromide (MTT), propidium iodide (PI), hexadimethrine bromide (polybrene), and rabbit anti-FLAG Abs were purchased from Sigma-

Aldrich Corp. (St. Louis, MO). Allophycocyanin (APC)-conjugated-anti-Thy1.1 Ab was obtained from eBioscience (San Diego, CA). TRIzol reagent was from

Invitrogen, and the High Capacity cDNA Reverse Transcription Kit was purchased

from Applied Biosystems (Carlsbad, CA). DNA oligonucleotide primers were obtained from Integrated DNA Technologies (Coralville, IA). Pfu UltraII was purchased from Agilent (Santa Clara, CA).

3.2.4 MTT ASSAY

For primary TRAF3−/− B lymphomas, cells were cultured for 4 days to obtain cleaner tumor cell populations for MTT assays. Tumor cells (1 × 105 cells/well) were plated in 24-well plates in the absence or presence of AD 198 or

PEP005 of various concentrations. Twenty-four hours later, total viable cell numbers were measured using the MTT assay as described15,161. Possible influences caused by direct MTT-drug interactions were excluded by studies in a cell-free system. Wells with untreated cells or with medium alone were used as positive and negative controls, respectively. Total viable cell number curves were plotted as a percentage of untreated control cells15,159,162.

3.2.5 MEASUREMENT OF APOPTOSIS

Cell apoptosis was assessed by both cell cycle analyses of the sub-G1 population and caspase 3 cleavage assays. For cell cycle analyses, cells

(3×105 cells/well) were cultured in 24-well plates in the absence or presence of appropriate doses of AD 198 or PEP005 for 24 hours, and fixed with ice-cold 70% ethanol. Cell cycle distribution was determined by propidium iodide (PI) staining followed by flow cytometry as previously described16,82. For caspase 3 cleavage assays, total cellular proteins were prepared from cells at different time points after

treatment with AD 198, and cleavage of caspase 3 was subsequently examined by immunoblot analysis.

3.2.6 LYMPHOMA TRANSPLANTATION AND DRUG TREATMENT OF NOD SCID MICE

TRAF3-/- mouse B lymphoma cell line 27–9.5.3 cells (3 × 106 cells per mouse) were i.p. injected into NOD SCID mice. On day 2 post transplantation, mice were divided into 3 cohorts for administration with drugs or with vehicle. Mice were i.p. injected with 150 μl (for a 20 g mouse) of AD 198 (5 mg/kg), oridonin (7.5 mg/kg), or vehicle (90% PBS and 10% DMSO). Drug or vehicle injections were carried out three times a week for 2 weeks. Transplanted NOD SCID mice were monitored daily for tumor development, or signs of illness or discomfort, including weight loss, enlarged lymph nodes or abdomen, labored breathing, hunched posture, and paralysis163. Histopathological examination of diseased mice was performed as previously described15,163.

3.2.7 TOTAL PROTEIN LYSATES

Cells (for mouse B lymphoma, 10 × 106 cells per condition; for human MM,

3.5 × 106 cells per condition) were treated with appropriate doses of AD 198 or

PEP005 for different time periods. Cell pellets were lysed in 200 μl of 2X SDS sample buffer (0.0625 M Tris, pH6.8, 1% SDS, 15% glycerol, 2% β- mercaptoethanol and 0.005% bromophenol blue), sonicated for 30 pulses, and boiled for 10 minutes.

3.2.8 CYTOSOLIC AND NUCLEAR EXTRACTS

Cells (for mouse B lymphoma, 10 × 106 per condition; for human MM, 5 ×

106 per condition) were treated with appropriate doses of AD 198 or PEP005 for different time periods. Cytosolic and nuclear extracts were prepared from the cells as described15,16. Briefly, cells were washed with ice-cold PBS, swelled in 500 μl of hypotonic Buffer A (10 mM HEPES, pH7.5, 10 mM KCl, 0.1 mM EDTA and 0.1 mM EGTA with protease and phosphatase inhibitors) for 15 minutes, and then lysed by addition of 31.5 μl of 10% NP-40. Lysates were centrifuged at 13,000 rpm for 5 minutes, and the supernatants were harvested as cytosolic extracts. The pellets were incubated with 100 μl of hypertonic Buffer C (20 mM HEPES, pH7.5,

0.4 M NaCl, 1 mM EDTA and 1 mM EGTA with protease and phosphatase inhibitors), vigorously agitated at 4°C for 45 minutes, and centrifuged at 13,000 rpm for 10 minutes at 4°C. The resulting supernatants in Buffer C were harvested as nuclear extracts. One-fifth volume of 5× SDS sample buffer was added into each cytosolic or nuclear extracts, which were subsequently boiled for 10 minutes.

3.2.9 FRACTIONATION OF CYTOSOL, NUCLEI AND MEMBRANES

Cells (for mouse B lymphoma, 30 × 106 cells per condition; for human MM,

15 × 106 cells per condition) were treated with appropriate doses of AD 198 or

PEP005 for 5, 10 or 30 minutes. Cytosol, nuclei and membranes were fractionated from cells as previously described158,159. Briefly, cells were washed with ice-cold

PBS, swelled in 700 μl of hypotonic Buffer M (10 mM HEPES, pH7.5, 10 mM KCl,

1 mM EDTA, 0.1 mM EGTA and 250 mM sucrose with protease and phosphatase

inhibitors) on ice for 10 minutes, and then homogenized in a Dounce homogenizer.

Cell lysis was checked by trypan blue uptake. Nuclei were isolated by centrifugation at 2,000 rpm for 10 minutes at 4°C. The supernatants were transferred to new tubes, and centrifugation at 13,000 rpm for 45 minutes was used to separate cytosol (supernatant) from membrane (pellet) fractions. One-fifth volume of 5× SDS sample buffer was added into the cytosol fraction. The pellets of nuclei and membranes were lysed in 200 μl of 2× SDS sample buffer respectively, and sonicated for 10 pulses. All protein samples were subsequently boiled for 10 minutes.

3.2.10 IMMUNOBLOT ANALYSIS

Aliquots of total protein lysates, cytosolic and nuclear extracts, or fractions of cytosol, nuclei and membranes were separated by SDS-PAGE, and electroblotted onto nitrocellulose membranes (ProTran, Schleicher & Schuell

BioScience, Keene, NH). Immunoblot analysis was performed using various antibodies as previously described128. Images of immunoblots were acquired using a low-light imaging system (LAS-4000 mini, FUJIFILM Medical Systems USA, Inc.,

Stamford, CT).

3.2.11 TAQMAN ASSAY OF C-MYC MRNA EXPRESSION

Cells (for mouse B lymphoma, 10 × 106 cells per condition; for human MM,

5 × 106 cells per condition) were treated with appropriate doses of AD 198 for different time periods. Total cellular RNA was extracted using TRIzol reagent

(Invitrogen) according to the manufacturer’s protocol. Complementary DNA

(cDNA) was prepared from RNA using High Capacity cDNA Reverse Transcription

Kit (Applied Biosystems). Quantitative real-time PCR was performed using

TaqMan Gene Assay kit (Applied Biosystems). TaqMan primers and probes (FAM- labeled) specific for mouse or human c-Myc were used in the PCR reaction to detect c-Myc mRNA in mouse B lymphoma and human MM cells, respectively.

Each reaction also included primers and the probe (VIC-labeled) specific for mouse or human β-actin mRNA, which served as endogenous control. Relative mRNA expression levels of c-Myc were analyzed using the Sequence Detection

Software (Applied Biosystems) and the comparative Ct (ΔΔCt) method following the manufacturer’s procedures. For each biological sample, duplicate PCR reactions were performed.

3.2.12 GENERATION OF LENTIVIRAL C-MYC EXPRESSION VECTORS

The c-Myc coding cDNA sequence was cloned from the human MM cell line

8226 cells by reverse transcription and high-fidelity PCR using primers human c-

Myc-F (5′- ACG ATG CCC CTC AAC GTT AG -3′) and human c-Myc-R (5′- TCC

TTA CGC ACA AGA GTT CC -3′). The high-fidelity polymerase Pfu UltraII (Agilent) was used in the PCR reaction. The c-Myc coding cDNA sequence was subcloned into the lentiviral expression vector pUB-eGFP-Thy1.1 79 (kindly provided by Dr.

Zhibin Chen, the University of Miami, Miami, FL) by replacing the eGFP coding sequence with the c-Myc coding sequence. To facilitate the differentiation of

transduced c-Myc from endogenous c-Myc, we engineered an N-terminal FLAG tag in frame with the c-Myc coding sequence, and generated a lentiviral expression vector of FLAG-tagged c-Myc (pUB-FLAG-c-Myc-Thy1.1). The human c-Myc coding sequence and the lentiviral expression vectors were verified by DNA sequencing.

3.2.13 LENTIVIRAL PACKAGING AND TRANSDUCTION OF HUMAN MM CELLS

Lentiviruses of the pUB-FLAG-c-Myc-Thy1.1 vector or an empty lentirival vector pUB-Thy1.1 were packaged and titered as previously described79,129. The pUB lentiviral expression vectors have an expression cassette of the marker

Thy1.1, and thus allow the transduced cells to be analyzed by Thy1.1 immunofluorescence staining followed by flow cytometry. Human MM cell lines

8226 and LP1 cells were transduced with the packaged lentiviruses at a multiplicity of infection (MOI) of 1:5 (cell:virus) in the presence of 8 μg/ml polybrene 79. On day

4 post transduction, the transduction efficiency of cells was analyzed by flow cytometry. Transduced cells were subsequently analyzed for c-Myc protein expression and responses to treatment with AD 198.

3.2.14 STATISTICS

Statistical analyses were performed using the Prism software (GraphPad,

La Jolla, CA). Survival curves were generated using the Kaplan-Meier method, and were compared using a log-rank (Mantel-Cox) test to determine whether

differences are significant. For other experiments, statistical significance was assessed by Student t test. P values less than 0.05 are considered significant.

3.3 RESULTS

3.3.1 DIFFERENTIAL EFFECTS OF AD 198 AND PEP005 ON TRAF3-/- TUMOR B CELLS

Decreased PKCδ nuclear translocation is a feature of both premalignant

TRAF3-/- B cells and primary TRAF3-/- B lymphomas derived from B-TRAF3-/- mice

15,16. PKCδ-/- mice have also been shown to have greatly increased numbers of B cells, and inhibition of PKCδ nuclear translocation contributes to BAFF-mediated survival of peripheral B cells 44,45,154. These previous studies suggest that PKCδ nuclear translocation is a downstream signaling event of TRAF3 and induces apoptosis in B cells. This prompted us to test the possibility that restoration of

PKCδ nuclear translocation may have therapeutic potential in TRAF3-/- tumor B cells. We thus evaluated the effects of two pharmacological activators of PKCδ,

AD 198 and PEP005, on primary TRAF3-/- B lymphomas harvested from diseased

B-TRAF3-/- mice. Our results of MTT assays demonstrated that AD 198 exhibited potent tumoricidal activity on primary TRAF3-/- B lymphoma cells in a dose- dependent manner (effective doses: 0.1 to 1 μM) (Fig. 17A and 17B). In sharp contrast, PEP005 stimulated the proliferation of primary TRAF3-/- B lymphoma cells at the dose range (10 to 100 nM) that has been previously shown to display anti-tumor activity on myeloid leukemia cells 158,159 (Fig. 18A and 18B).

Figure 17. AD 198 exhibited potent anti-proliferative/survival-inhibitory effects on TRAF3- /- mouse B lymphoma and human MM cells. Tumor B cells were treated with various concentrations of AD 198 for 24 h. Total viable cell numbers were subsequently determined by MTT assay. (A and B) Effects on primary splenic B lymphoma cells harvested from diseased B-TRAF3-/- mice. Panel A shows the activity of AD 198 examined with a wide range of doses (1:10 serial dilutions). Panel B shows refined dose- dependent effects of AD 198 (examined at 1:2 serial dilutions). Similar results were also obtained with primary B lymphoma cells purified from ascites, cervical and mesenteric LNs of several individual B-TRAF3-/- mice with spontaneous tumors. (C) Effects on TRAF3-/- B lymphoma cell lines derived from 3 individual B-TRAF3-/- mice, including 105–8.1B6 (105–8), 115–6.1.2 (115–6), and 27–9.5.3 (27–9). (D) Effects on human patient-derived MM cell lines with TRAF3 bi-allelic deletions or frameshift mutations, including 8226, KMS11 and LP1. The graphs depict the results of three independent experiments with duplicate samples in each experiment (mean ± SEM).

We then compared the effects of AD 198 and PEP005 on three TRAF3-/- B lymphoma cell lines derived from different individual B-TRAF3-/- mice, including

105-8.1B6, 115-6.1.2 and 27-9.5.315. These 3 cell lines differ in their Ig VDJ sequences, malignant states, and metastatic capabilities. The 105-8.1B6 cell line is IgM+, does not contain somatic hypermutation (SHM) in the Ig VDJ region, and develops peritoneal and splenic B lymphomas within 4 months when transplanted into NOD SCID recipient mice. The 115-6.1.2 cell line is IgM+, contains SHM in the Ig VDJ region, and develops peritoneal and splenic B lymphomas within 2 months when transplanted into NOD SCID recipient mice, which occasionally metastasize to the kidney and liver. The 27-9.5.3 cell line is IgG+, contains SHM in the Ig VDJ region, and develops peritoneal and splenic B lymphomas, which often metastasize to the kidney, liver and lung within 3 weeks when transplanted into NOD SCID recipient mice. We found that AD 198 consistently exhibited potent tumoricidal activity on all 3 TRAF3-/- B lymphoma cell lines in a dose-dependent manner (effective doses: 0.25 to 4 μM) (Fig. 17C). In contrast, PEP005 displayed contradictory effects on these 3 cell lines. PEP005 induced the proliferation of 105-

8.1B6 cells, killed 115-6.1.2 cells, and did not affect 27-9.5.3 cells at the dose range of 12.5 to 100 nM (Fig. 18C).

Figure 18. PEP005 exhibited differential effects on TRAF3-/- mouse B lymphoma and human MM cells. Total viable cell numbers were determined by MTT assay at 24 h after PEP005 treatment. (A and B) Effects on primary splenic B lymphoma cells harvested from diseased B-TRAF3-/- mice. Similar results were also obtained with primary B lymphoma cells purified from ascites, cervical and mesenteric LNs of several individual B-TRAF3-/- mice with tumors. (C) Effects on TRAF3-/- mouse B lymphoma cell lines. (D) Effects on human MM cell lines. The graphs depict the results of three independent experiments with duplicate samples in each experiment (mean ± SEM).

To extend the clinical relevance of our findings, we further examined the effects of AD 198 and PEP005 on three human patient-derived MM cell lines with

TRAF3 deletions or mutations: 8226, KMS11 and LP1. Both 8226 and KMS11 cell lines contain bi-allelic deletions of the Traf3 gene, while LP1 cell line has frameshift mutations of Traf3 15. These cell lines represent naturally occurring TRAF3-/- tumor

B cells. Our results of MTT assays showed that the responses of the 3 lines to AD

198 and PEP005 recapitulated those of mouse TRAF3-/- B lymphoma cell lines

(Figure 17D and 18D). Together, these data indicate that AD 198 exhibits potent tumoricidal activity, whereas PEP005 displays divergent effects on TRAF3-/- mouse B lymphoma cells and human MM cells.

3.3.2 AD 198 BUT NOT PEP005 INDUCED APOPTOSIS IN TRAF3-/- TUMOR B CELLS

To understand the mechanism of AD 198 and PEP005, we first performed cell cycle analysis using PI staining followed by flow cytometry. We found that AD

198 induced TRAF3-/- mouse B lymphoma cells and human MM cells to undergo apoptosis, as demonstrated by the drastic increase of the sub G0/G1 population with DNA content < 2n (Fig. 19A and 19B). AD 198 also inhibited the proliferation of TRAF3-/- tumor B cells, as shown by the marked decrease of the population at the S/G2/M phase (2n < DNA content ≤ 4n). In contrast, PEP005 increased the population at the S/G2/M phase in mouse 105-8.1B6 and human 8226 cells, but induced the apoptotic population and decreased the population at the S/G2/M phase in mouse 115-6.1.2 cells. PEP005 did not have significant effects on the cell cycle distribution in mouse 27-9.5.3 as well as human KMS11 and LP1 cell lines

(Fig. 20A and 20B). We next determined whether AD 198 induced the activation

of the key effector caspase, caspase 3, involved in apoptosis. We found that AD

198 induced the rapid activation of caspase 3, as evidenced by the cleavage of caspase 3 as early as three hours after treatment with AD 198 in TRAF3-/- mouse

B lymphoma and human MM cell lines (Fig. 19C). Collectively, our data demonstrate that AD 198 but not PEP005 induces rapid apoptosis in TRAF3-/- tumor B cells.

Figure 19. AD 198 induced apoptosis in TRAF3-/- tumor B cells. (A and B) Cell cycle distribution determined by PI staining and flow cytometry. TRAF3-/- mouse B lymphoma cell lines (A) or human MM cell lines (B) were cultured in the absence or presence of AD 198 of indicated concentration for 24 h, and then fixed. Fixed cells were stained with PI and analyzed by FACS. Representative histograms of PI staining are shown, and percentage of apoptotic cells (DNA content < 2n; sub-G1) and proliferating cells (2n < DNA content ≤ 4n; S/G2/M) are indicated. Results are representative of 3 independent experiments. (C) Western blot analysis of caspase 3 cleavage. Cells were cultured in the absence or presence of AD 198 of indicated concentration. Total cellular lysates were prepared at indicated time points, and then immunoblotted for caspase 3, followed by actin. Immunoblots of actin were used as loading control. Results are representative of three independent experiments. Similar results were also obtained with other cell lines examined.

Figure 20. PEP005 did not induce apoptosis in TRAF3-/- tumor B cells. Cell cycle distribution was determined by PI staining and flow cytometry. TRAF3-/- mouse B lymphoma cell lines (A) or human MM cell lines (B) were cultured in the absence or presence of PEP005 of indicated concentration for 24 h before PI staining. Representative histograms of PI staining are shown, and percentage of apoptotic cells (DNA content < 2n) and proliferating cells (2n < DNA content ≤ 4n) are indicated. Results are representative of three independent experiments.

3.3.3 AD 198 EXHIBITED POTENT IN VIVO ANTI-TUMOR ACTIVITY ON TRAF3-/- MOUSE B LYMPHOMAS

The potent in vitro tumoricidal activity of AD 198 led us to further assess its in vivo therapeutic potential. We recently reported that B-TRAF3-/- mice display a long and varied latency (9~18 months) in developing B lymphomas 15. Therefore,

B-TRAF3-/- mice are not ideal for drug treatment experiments. In this study, we used NOD SCID mice transplanted with the highly malignant TRAF3-/- B lymphoma cell line 27-9.5.3 as model systems for in vivo drug treatment experiments. We also included the study of oridonin, an inhibitor of NF-κB2 and NF-κB1 activation, which also exhibits robust in vitro tumoricidal activity on primary TRAF3-/- B lymphomas harvested from diseased B-TRAF3-/- mice 15.

Figure 21. AD 198 and oridonin exhibited potent anti-tumor activity on transplanted TRAF3-/- B lymphomas in NOD SCID mice. TRAF3-/- mouse B lymphoma cell line 27–9.5.3 cells (3 x 106 per mouse) were i.p. injected into NOD SCID recipient mice. On day 2 post-transplantation, mice were divided into 3 cohorts for administration with drugs or vehicle. Mice were i.p. injected with 150 μl (for a 20 g mouse) of AD 198 (5 mg/kg, n = 8), oridonin (7.5 mg/kg, n = 6), or vehicle (90% PBS and 10% DMSO; CTL, n = 12). Drug or vehicle injections were carried out three times a week for 2 weeks. Transplanted NOD SCID mice were monitored daily for tumor development. Survival curves of mice were generated using the Kaplan-Meier method. P value determined by the Mantel-Cox log-rank test.

In the absence of drug treatment, transplantation of 27-9.5.3 cells (3 x 106 cells/mouse) caused rapid B lymphoma development in NOD SCID mice, which killed the mice at 23 ± 3 (Mean ± SD) days post-transplantation (Fig. 21). Necropsy revealed that B lymphomas were not only developed in the peritoneal cavity and spleen, but also often metastasized to the kidney, liver and lung. Treatment of transplanted NOD SCID mice with oridonin significantly prolonged the survival of mice to 30 ± 3.7 days post-transplantation (p = 0.0043, determined by the Mantel-

Cox log-rank test). However, metastasis to the kidney, liver and lung was also common in oridonin-treated mice. Interestingly, administration of AD 198 into transplanted NOD SCID mice vastly extended the survival of mice to 46 ± 12 days post-transplantation (p < 0.0001, determined by the Mantel-Cox log-rank test).

Furthermore, B lymphomas were typically localized in the peritoneal cavity, and metastasis to the kidney, liver and lung was rare (2 out of 8) in AD 198-treated mice. Thus, our results demonstrate that both AD 198 and oridonin exhibit in vivo anti-tumor activity on TRAF3-/- mouse B lymphomas.

3.3.4 AD 198 INDUCED PKC CLEAVAGE, WHILE PEP005 INDUCED PKC TRANSLOCATION IN TRAF3-/- TUMOR B CELLS

Both AD 198 and PEP005 have been previously shown to induce the subcellular translocation of PKCδ in myeloid leukemia cells, which mediates the anti-leukemic effects of these two drugs 155,156,158,159. PKCδ nuclear translocation also regulates B cell apoptosis44,45,154. We thus compared the effects of AD 198 and PEP005 on PKCδ nuclear translocation using cytosolic and nuclear extracts,

which were prepared by the same method described in our previous studies 15,16.

Surprisingly, neither AD 198 nor PEP005 increased the nuclear levels of PKCδ at

6 hours after treatment in TRAF3-/- tumor B cells (Fig. 22A and Fig. 23).

Interestingly, we noticed that AD 198 but not PEP005 induced the cleavage of cytosolic PKCδ from the 78 kDa holoenzyme to the 40 kDa catalytic fragment at 6 hours after treatment in a dose dependent manner (Fig. 22A and Fig. 23). We next determined the time-course effects of AD 198 and PEP005 on PKCδ nuclear translocation or cleavage. We found that neither AD 198 nor PEP005 increased the nuclear levels or cleavage of PKCδ at 5 to 60 minutes after treatment (Fig. 22B and Fig. 23). In these experiments, we also examined the effects of AD 198 and

PEP005 on the other oncogenic pathways that we recently identified, NF-κB2 and

NF-κB1 activation 15. It has been shown that AD 198 and PEP005 promote NF-

κB1 activation in breast cancer and primary acute myeloid leukemia cells157,164,165.

No obvious effects of AD 198 were observed on nuclear levels of NF-κB2

(p52/p100) or NF-κB1 subunits (p50, c-Rel and RelA), except for a moderate inhibition of RelB levels by the highest dose of AD 198 examined (Fig. 20 and Fig.

23). The only noticeable effect of PEP005 was the increase of nuclear RelA and c-Rel levels at 15 to 60 minutes after treatment in mouse 105-8 cells (Fig. 23).

100

Figure 22. AD 198 did not affect PKCδ nuclear translocation or NF-κB activation. (A) Dose-dependent effects of AD 198. Mouse or human tumor B cells were cultured in the absence or presence of various concentrations of AD 198 for 6 h. (B) Time-dependent effects of AD 198. Mouse or human tumor B cells were cultured in the absence or presence of AD 198 (1 μM for 105–8 cells and 2 μM for 8226 cells) for indicated time periods. Cytosolic and nuclear extracts were prepared as described. Proteins were immunoblotted for PKCδ, NF-κB2 (p100 – p52), RelB, NF-κB1 (p50), c-Rel, and RelA, followed by HDAC1 (nuclear protein loading control) and actin (cytosolic protein loading control). Results represent three independent experiments. Similar results were also obtained with other cell lines examined.

101

Figure 23 . Effects of PEP005 on the cytosolic and nuclear levels of PKCδ, NF-κB1 and NF- κB2 subunits, and c-Myc. (A) Dose-dependent effects of PEP005. Mouse or human tumor B cells were cultured with various concentrations of PEP005 for 6 h. (B) Time-dependent effects of PEP005. Mouse or human tumor B cells were cultured in the absence or presence of PEP005 for indicated time periods. Cytosolic and nuclear extracts were prepared as described in the Methods. Proteins were immunoblotted for PKCδ, NF-κB2 (p100 – p52), RelB, NF-κB1 c-Rel, RelA, c-Myc, followed by HDAC1 and actin. Results are representative of three independent experiments. Similar results were also obtained with other TRAF3-/- cell lines.

102

Our unexpected results of PKCδ nuclear translocation prompted us to compare our biochemical method of nuclear extraction with the previous method used to study AD 198 and PEP005 in myeloid leukemia cells. We identified one major difference in the methods: our nuclear extracts did not include proteins localized at or associated with the nuclear membrane, but the nuclei fraction prepared in previous studies of AD 198 and PEP005 did. We thus prepared subcellular fractions of cytosol, nuclei, and membrane (including mitochondria) using the method that was previously employed in the study of AD 198 and

PEP005 in myeloid leukemia cells158,159. As shown in Fig. 24A, our results clearly demonstrated that PEP005 induced the rapid translocation of PKCα, PKCε and

PKCδ from the cytosol to the nuclei and membranes (including mitochondria) in

TRAF3-/- mouse B lymphoma cells. Similarly, PEP005 induced the rapid translocation of PKCδ from the cytosol to the nuclei and membranes in TRAF3-/- human MM cells (Fig. 24A). However, in sharp contrast, AD 198 did not affect the

Figure 24. Differential effects of AD 198 and PEP005 on PKCδ subcellular translocation and cleavage. (A) Subcellular translocation of PKC isoforms. Mouse or human tumor B cells were cultured in the absence or presence of AD 198 or PEP005 for 10 minutes. Drug concentrations used: 1 μM AD 198, 0.1 μM PEP005 for 105–8 cells; 4 μM AD 198, 0.2 μM PEP005 for 8226 cells. Biochemical fractionation of cytosol, nuclei and membrane was performed and proteins were immunoblotted for PKCδ, phosphorylated-PKCδ (P-PKCδ), PKCϵ, and PKCα, followed by COXIV (membrane control), HDAC1 (nuclear control), and actin (cytosolic control). Similar results were obtained at 5 or 30 minutes after treatment, and with another TRAF3-/- mouse B lymphoma cell line 27–9.5.3 and human MM cell line 8226 (with barely detectable levels of PKCα in 8226 cells). (B) Phosphorylation and cleavage of PKCδ. Cells were cultured in the absence/presence of AD 198 of indicated concentration. Total cellular lysates were prepared at indicated time points, then immunoblotted for phosphorylated-PKCδ (P-PKCδ), PKCδ, followed by actin. Similar results also obtained with other cell lines. Results represent three independent experiments.

103

subcellular distribution of PKCα, PKCε or PKCδ in any TRAF3-/- tumor B cell lines examined in this study (Fig. 24A).

Considering that PKCδ activation can also be modulated by phosphorylation and cleavage166,167, we further assessed the effects of AD 198 on

104

the phosphorylation and cleavage of PKCδ in TRAF3-/- tumor B cell lines. We found that AD 198 did not induce the phosphorylation of PKCδ from 10 minutes up to 6 hours after treatment in any TRAF3-/- tumor B cell lines examined in this study (Fig.

24A and 24B). Interestingly, AD 198 did induce the cleavage of PKCδ at 6 hours after treatment in TRAF3-/- tumor B cells (Fig. 22A and 24B). However, the induction of PKCδ cleavage occurred relatively late, and was preceded by caspase

3 activation, which was detected at 3 hours after AD 198 treatment (Fig. 19C). It has been previously shown that PKCδ is a substrate of caspase 3, which cleaves the 78 kDa holoenzyme of PKCδ to generate the 40 kDa catalytic fragment of

PKCδ 168. Therefore, it is very likely that PKCδ cleavage induced by AD 198 is a consequence of caspase 3 activation in TRAF3-/- tumor B cells, and is not the initiating signal that triggers the apoptosis. Taken together, our findings suggest that AD 198 induces the apoptosis of TRAF3-/- tumor B cells not through the translocation or activation of its known target PKCδ, but through a distinct novel mechanism.

3.3.5 DIFFERENTIAL EFFECTS OF AD 198 AND PEP005 ON ERK, P38 AND JNK ACTIVATION IN TRAF3-/- TUMOR B CELLS

To gain further insights into the molecular mechanisms underlying the anti- tumor effect of AD 198 and the divergent effects of PEP005, we next sought to investigate key signaling pathways that are known to play important roles in regulating B cell survival and proliferation, including the activation of ERK, p38,

JNK, and Akt. We found that AD 198 markedly and rapidly decreased the

105

phosphorylation levels of ERK1 (p44), ERK2 (p42), and p38 in TRAF3-/- mouse B lymphoma and human MM cells (Fig. 25A and 25B). Inhibition of ERK and p38 phosphorylation was detected as early as 5 minutes after AD 198 treatment. AD

198 also inhibited JNK activation in TRAF3-/- human MM cells. Interestingly, although Akt activation is considered as a general survival pathway, AD 198 increased the Ser473 phosphorylation and thus activation of Akt in TRAF3-/- mouse

B lymphoma and human MM cells. In contrast, PEP005 induced the activation of

ERK, JNK and Akt in TRAF3-/- mouse B lymphoma cells, and also induced ERK and Akt activation in human MM cells (Fig. 25A). Taken together, our results suggest that the differential effects of AD 198 and PEP005 on tumor B cells are mediated by their distinct effects on multiple signaling pathways, including PKCδ,

PKCα, and PKCε translocation (Fig. 24A), and ERK, p38 and JNK phosphorylation

(Fig. 25A).

106

Figure 25 . AD 198 inhibited phosphorylation of ERK, p38 & JNK in TRAF3-/- tumor B cells. (A) Phosphorylation of MAPK and Akt analyzed by immunoblot analysis. Mouse or human tumor B cells were cultured in the absence/presence of AD 198 or PEP005 for indicated time periods. Drug concentrations used: 1 μM AD 198, 0.1 μM PEP005 for 105–8 cells; 4 μM AD 198, 0.2 μM PEP005 for 8226 cells. Total cellular lysates were then immunoblotted for phosphorylated (P-) or total ERK, p38, JNK, and Akt, followed by actin. (B) Quantitation of ERK and p38 phosphorylation. Phosphorylated and total ERK1, ERK2, and p38 bands on immunoblots were quantitated using a low-light imaging system, and the results are presented graphically. The amount of P-ERK1, P- ERK2, or P-p38 in each lane was normalized to the intensity of the corresponding total ERK1, ERK2 or p38 band. The graphs depict ERK and p38 phosphorylation observed in three independent experiments (mean ± SD).

107

3.3.6 AD 198 RAPIDLY SUPPRESSED C-MYC EXPRESSION IN TRAF3- /- TUMOR B CELLS

One known target gene of ERK, p38 and JNK signaling pathways that is especially essential for B cell survival and proliferation is c-Myc169-171. In light of our evidence that AD 198 inhibited ERK, p38 and JNK signaling pathways, we further investigated the effects of AD 198 on c-Myc protein levels. We found that

AD 198 potently decreased protein levels of c-Myc, which was primarily localized in the nucleus, in a dose-dependent manner at 6 hours after treatment in TRAF3-

/- mouse B lymphoma and human MM cells (Fig. 26A). We next found that AD 198 vastly inhibited c-Myc protein levels as early as 1 hour after treatment in all TRAF3-

/- tumor B cell lines examined in this study (Fig. 26B and data not shown). In contrast, PEP005 did not inhibit c-Myc protein levels in any tumor B cell lines examined.

108

Figure 26. AD 198 suppressed c-Myc expression in TRAF3-/- tumor B cells. (A) Dose -dependent effects of AD 198 on c-Myc protein levels. Mouse or human tumor B cells were cultured in the absence/presence of various concentrations of AD 198 for 6h. Cytosolic and nuclear extracts were prepared. (B) Time-dependent effects of AD 198 on c-Myc protein levels. Mouse or human tumor B cells were cultured in the absence/presence of AD 198. Total cellular lysates were prepared at indicated time points. Proteins were immunoblotted for c-Myc, followed by actin (loading control). Results of A and B represent three independent experiments. Similar results were also obtained with other TRAF3-/- cell lines. (C) c-Myc transcript levels analyzed by real time PCR. Cells were treated with AD 198 for indicated time periods. Drug concentrations used: 1 μM AD 198 for 105–8 cells; 2 μM AD 198 for 27–9 cells; 4 μM AD 198 for 8226 and KMS11 cells. Total cellular RNA was extracted and reverse transcribed. Quantitative real-time PCR was performed using TaqMan primers and probe (FAM-labeled) specific for c-Myc. Each PCR reaction also included primers and the probe for β-actin mRNA, an endogenous control. Graphs depict three independent experiments with duplicate samples in each (mean ± SEM).

109

To understand the mechanism of AD198-mediated suppression of c-Myc protein levels, we examined the mRNA levels of c-Myc by reverse transcription and quantitative real time PCR analyses. As shown in Fig. 26C, AD 198 dramatically and rapidly inhibited the mRNA levels of c-Myc in TRAF3-/- mouse B lymphoma and human MM cells. Decrease in c-Myc mRNA levels was detected as early as 10 minutes after AD 198 treatment, and could completely account for the decrease in c-Myc protein levels observed in these cells. These results indicate that AD 198 potently suppresses c-Myc mRNA and protein expression in TRAF3-

/- tumor B cells.

3.3.7 AD 198 EXHIBITED POTENT TUMORICIDAL ACTIVITY AND RAPIDLY SUPPRESSED C-MYC EXPRESSION IN TRAF3-SUFFICIENT B LYMPHOMA CELL LINES

Considering that elevated expression of c-Myc is associated with many B cell malignancies 172, we further tested the therapeutic effects of AD 198 on

TRAF3-sufficient B lymphoma cell lines. These include mouse B lymphoma cell lines A20.2J, m12.4.1, and CH12.LX, and human B lymphoma cell lines Daudi

(Burkitt’s lymphoma), Ramos (Burkitt’s lymphoma), and JeKo-1 (mantle cell lymphoma). Our results of MTT assay demonstrated that AD 198 also exhibited potent tumoricidal activity (effective dose: 0.25 to 2 μM) on all the TRAF3-sufficient mouse and human B lymphoma cell lines examined in this study (Fig. 27A). To determine whether c-Myc is also the principal target of AD 198 in TRAF3-sufficient

B lymphoma cells, we examined the effects of AD 198 on c-Myc protein levels. We

110

found that AD 198 strikingly inhibited c-Myc protein levels as early as 1 hour after

treatment in all TRAF3-sufficient B lymphoma cell lines examined in this study (Fig.

27B). AD 198 also induced the cleavage and activation of caspase 3 at 3 hours

after treatment in these cell lines. It should be noted that c-Myc suppression

precedes caspase 3 activation, suggesting that c-Myc is not the consequence but

may be the trigger of apoptosis. Together, these results indicate that AD 198 also

has therapeutic potential and targets c-Myc in TRAF3-sufficient B lymphomas.

Figure 27. AD 198 exhibited potent anti-tumor activity and suppressed c-Myc expression in TRAF3-sufficient mouse and human B lymphoma cell lines. (A) Anti-tumor activity of AD 198. Mouse B lymphoma cell lines A20.2 J, m12.4.1 and CH12.LX, or human B lymphoma cell lines Daudi, Ramos and JeKo-1 were treated with various concentrations of AD 198 for 24 h. Total viable cell numbers were subsequently determined by MTT assay. The graphs depict the results of three independent experiments with duplicate samples in each experiment (mean ± SEM). (B) Time-dependent effects of AD 198 on c-Myc protein levels. Mouse or human B lymphoma cells were cultured in the absence or presence of AD 198 of indicated concentrations. Total cellular lysates were prepared at indicated time points. Proteins were immunoblotted for c-Myc, followed by caspase 3 and actin. Immunoblots of actin were used as loading control. Results are representative of three independent experiments.

111

112

3.3.8 LENTIVIRAL VECTOR-MEDIATED CONSTITUTIVE EXPRESSION OF C-MYC CONFERRED PARTIAL RESISTANCE TO THE ANTI- TUMOR EFFECTS OF AD 198 IN HUMAN MM CELL LINES

To further investigate whether c-Myc suppression contributes to the anti- tumor effects of AD 198 in malignant B cells, we performed reconstitution of c-Myc expression experiments. We generated a lentiviral expression vector of FLAG- tagged human c-Myc, pUB-FLAG-c-Myc-Thy1.1, in which constitutive expression of c-Myc is driven by the ubiquitin promoter. Human MM cell lines 8226 and LP1 cells were transduced with this vector or an empty lentiviral expression vector

(pUB-Thy1.1), and then analyzed for their responses to AD 198 treatment.

Transduction efficiency of the lentiviral vectors was over 90% in human MM cells, as demonstrated by immunofluorescence staining and flow cytometry

(Figure 28A). Following treatment with AD 198, although endogenous c-Myc protein levels were strikingly decreased, the transduced FLAG-c-Myc protein levels were not suppressed by AD 198 as shown in the immunoblots of both FLAG and c-Myc (Figure 28B). These results demonstrated that expression of the transduced FLAG-c-Myc driven by the ubiquitin promoter was not suppressed by

AD 198, suggesting that AD 198 inhibits the transcription of endogenous c-Myc via its effects on the c-Myc promoter. Interestingly, we further found that constitutive expression of FLAG-c-Myc substantially counteracted the effects of AD 198 on the proliferation and survival of human MM cells (Figure 28C). Thus, our results indicate that c-Myc suppression is a major contributing factor to the anti-tumor effects of AD 198 in human MM cells.

113

Figure 28. Constitutive c-Myc expression counteracted the effects of AD 198 on the proliferation and survival of human MM cells. The human MM cell line 8226 cells were transduced with a lentiviral expression vector of FLAG tagged c-Myc (pUB-FLAG-c-Myc-Thy1.1, labeled as pUB-FLAG-Myc in the figure). Cells transduced with an empty vector (pUB-Thy1.1) were the control. (A) Transduction efficiency analyzed by Thy1.1 immunofluorescence staining and flow cytometry. Gated population (Thy1.1+) indicates cells successfully transduced. Untransduced 8226 cells were the negative control. (B) Time-dependent effects of AD 198 on endogenous c-Myc and transduced FLAG-c- Myc protein levels. Transduced 8226 cells (8226/pUB-Thy1.1 or 8226/pUB-FLAG-Myc) were cultured in the absence/presence of 4μM AD 198. Total cellular lysates were prepared at indicated time points, and immunoblotted for FLAG, c-Myc, followed by actin (loading control). Results represent three independent experiments. (C) Decreased anti-proliferative/survival-inhibitory effects of AD 198 on human MM cells with constitutive c-Myc expression. Transduced 8226 cells (8226/pUB-Thy1.1 or 8226/pUB-FLAG-Myc) were treated with AD 198 for 24h. Total viable cell numbers were determined by MTT assay. The graph depicts three independent experiments with duplicate samples in each (mean ± SEM). Similar results were also observed in LP1, another human MM cell line.

114

3.4 DISCUSSION

We previously showed that premalignant TRAF3-/- B cells and TRAF3-/- B lymphomas have decreased nuclear levels of PKCδ15,16. This, together with the evidence that decreased nuclear translocation of PKCδ promotes B cell survival

44,45,154, prompted us to evaluate the therapeutic potential of two PKCδ activators,

AD 198 and PEP005, in TRAF3-/- mouse B lymphomas and human MM cells. In the present study, we report that AD 198 exhibited potent in vitro and in vivo tumoricidal activity on TRAF3-/- tumor B cells, while PEP005 displayed contradictory anti-tumor or pro-tumor activities on different cell lines. AD 198 and

PEP005 induced differential effects on TRAF3-/- tumor B cells through distinct biochemical mechanisms. Our detailed mechanistic study uncovered a novel

PKCδ-independent mechanism of the anti-tumor effect of AD 198 that involves c-

Myc suppression. Furthermore, we found that AD 198 also exhibited potent anti- tumor effects and targeted c-Myc in TRAF3-sufficient mouse and human B lymphoma cell lines. Our findings suggest that AD 198 has therapeutic potential for the treatment of NHL and MM involving TRAF3 inactivation or c-Myc up- regulation.

Ingenol-3-angelate (PEP005) is an active ingredient of the sap from

Euphorbia peplus, which has been used for centuries in the U.K. and Australia as a traditional treatment for skin conditions, including warts, corns and skin cancers.

PEP005 has now entered phase II clinical trials as a topical treatment for non- melanoma skin cancers and actinic keratoses. PEP005 is also being developed as

115

a systemic treatment for acute myeloid leukemia in preclinical models. Anti-tumor effects of PEP005 have also been demonstrated in s.c. inoculated melanoma, lung carcinoma, prostate cancer, cervical carcinoma, and bladder cancer173,174.

PEP005 is structurally closely related to phorbols, and is a potent activator of novel

(δ, ε, η, θ) and classical (α, β, γ) isoenzymes of PKC at lower concentrations (10 to 100 ng/ml)175. However, PKCδ is the isoform that mediates the pro-apoptotic effects of PEP005 in myeloid leukemia and colon cancer cells158,159,176. In these cells, PEP005 induces PKCδ translocation from the cytoplasm to the plasma membrane, nuclear membrane and mitochondrial membrane.

Interestingly, we detected PEP005-induced nuclear and membrane translocation of PKCδ, PKCα and PKCε in TRAF3-/- tumor B cells (Fig. 24A). In cancer, PKCα and PKCε are generally linked to proliferation or survival and thus considered as oncogenes. In contrast, PKCδ has a pro-apoptotic function in a variety of cancer cells157,159,177. Activation of PKC isoforms signals further downstream events, such as the activation of p38, ERK, JNK or NF-κB in melanoma, myeloid leukemia and colon cancer cells 160,162,177-179, which were all observed in our study of tumor B cells. In colon cancer cells, inhibition of Akt phosphorylation via a PKC-independent mechanism also contributes to the apoptotic effects of PEP005159,176. In contrast, we found that PEP005 induced Akt phosphorylation in TRAF3-/- tumor B cells. Therefore, the ultimate effect of cell proliferation or apoptosis induction by PEP005 depends on the balance between the levels and activities of pro-apoptotic (PKCδ) and anti-apoptotic (PKCα and

116

PKCε) isoforms of PKC as well as their crosstalk with different signaling pathways

(MAPKs, NF-κB, and Akt) in each tumor B cell line. Indeed, we detected varying levels of different PKC isoforms (α, δ and ε) in different tumor B cell lines, and this may contribute to the observed divergent responses of the cells to PEP005. Our findings thus provide new insights into the complexity of the signaling pathways controlled by PEP005 in TRAF3-/- tumor B cells.

N-Benzyladriamycin-14-valerate (AD 198) is a novel semisynthetic, lipophilic anthracycline analogue with experimental anti-tumor activity superior to that of doxorubicin (DOX)157,180,181. Nuclear-targeted anthracycline antibiotics, such as DOX, are known to exert their anti-tumor effects primarily via DNA intercalation and topoisomerase II inhibition, thereby causing double-stranded

DNA breaks and promoting apoptosis. The extranuclear-targeted AD 198, unlike

DOX, only weakly binds DNA, and does not inhibit topoisomerase II157,180-182. AD

198 is structurally similar to commonly accepted ligands for the C1 regulatory domain of PKC, and binds to the diacylglycerol-binding C1b domain of classical and novel PKC isoforms165,181,183. The interaction between AD 198 and the C1b domain leads to the activation of PKC isoforms, especially PKCδ, and thereby promoting rapid apoptosis in a variety of cancer cells, including myeloid leukemia, breast cancer, prostate cancer, melanoma, colon cancer, and ovarian carcinoma

155,157,165,183. PKCδ was identified as the principal target of AD 198 that mediates the apoptotic effects of AD 198155,157,182.

117

In myeloid leukemia cells, AD 198 induces the mitochondrial and membrane translocation of PKCδ, PKCα and PKCε155,156,165,182. In the present study, we found that AD 198 induced caspase 3 activation, PKCδ cleavage, and DNA fragmentation in TRAF3-/- tumor B cells as previously observed in myeloid leukemia cells. However, AD 198-induced apoptosis of tumor B cells was not mediated through the activation of PKCδ, as translocation or phosphorylation of

PKCδ was not detected. Additionally, translocation of PKCα or PKCε was not induced by AD 198 in tumor B cells either. Thus, the anti-proliferative and apoptotic effects of AD 198 on TRAF3-/- tumor B cells are mediated through a novel, PKCδ- independent mechanism.

We identified c-Myc as the principal target of AD 198 in TRAF3-/- and

TRAF3-sufficient malignant B cells. We found that both the mRNA and protein levels of c-Myc were strikingly and rapidly suppressed by AD 198. c-Myc protein is a central regulator of B cell survival and proliferation, and has a short half-life

(about 20 - 30 minutes)184,185. It has been previously shown that the promoter regions of both human and mouse c-Myc genes contain binding sites for AP-1, a transcription factor directly activated by ERK, p38 and JNK signaling pathways169-

171. AP-1 is also indirectly inhibited by Akt activity169. Interestingly, we found that

AD 198 inhibited ERK, p38 and JNK activation, but promoted Akt activation in

TRAF3-/- tumor B cells. In this context, our results suggest that AD 198 targets c-

Myc by inhibiting c-Myc transcription in tumor B cells, which is mediated through inhibition of ERK, p38 and JNK pathways as well as activation of the Akt pathway.

118

However, we could not exclude additional mechanisms, as it has been shown that

AD 198 inhibits E. coli RNA polymerase or chicken leukemic RNA polymerase activity through drug-template interaction or enzyme inactivation, respectively 186.

Regardless of the exact mechanisms, considering that elevated expression of c-

Myc is ubiquitously observed in many B cell malignancies 172, our findings suggest that AD 198 may have wide therapeutic application in B cell neoplasms.

It has been shown that AD 198 has anti-tumor activity superior to DOX in breast cancer, ovarian carcinoma and melanoma models187,188, which was recapitulated in our TRAF3-/- mouse B lymphomas. We previously showed that

DOX did not exhibit tumoricidal activity on primary B lymphoma cells derived from

B-TRAF3-/- mice. Here we report that AD 198 has potent anti-tumor effects on

TRAF3-/- mouse B lymphomas and human MM. AD 198 can also override multiple mechanisms of DOX resistance, including those mediated by p53 dysfunction, or by overexpression of the multidrug transporters (P-glycoprotein and multidrug resistance protein) or the anti-apoptotic proteins (Bcl-2, Bcl-xL and NF-κB)

155,165,182,189. Importantly, AD 198 is also pharmacologically superior to DOX in terms of its decreased cardiotoxicity, low hematotoxicity, and the rapid rate of intracellular uptake178,180,187,190. The use of DOX is limited by its dose-dependent, and often irreversible cardiotoxicity. However, AD 198 does not exhibit significant cardiotoxicity or other organ toxicities at therapeutic doses, and is cardioprotective in rodent models178,180,187,190. In support of this notion, we demonstrated that in

NOD SCID mice transplanted with TRAF3-/- mouse B lymphomas, administration

119

of AD 198 drastically extended the survival of mice and inhibited the growth and metastasis of B lymphomas. In fact, AD 198 demonstrated a higher in vivo potency than oridonin, an inhibitor of both NF-κB2 and NF-κB1 pathways15. In summary, our findings reveal a novel PKCδ-independent mechanism of AD 198 that targets c-Myc in malignant B cells, and support further clinical studies of AD 198 as an anti-cancer agent for NHL and MM.

3.5 CONCLUSIONS

The present study evaluated the therapeutic potential of two PKCδ activators, AD 198 and PEP005, on TRAF3-/- mouse B lymphomas and human MM cells. Surprisingly, we found that AD 198 and PEP005 have differential effects on

TRAF3-/- tumor B cells through distinct biochemical mechanisms. Our results uncovered a novel, PKCδ-independent mechanism of the anti-tumor effects of AD

198 that strikingly targets c-Myc and inhibits the phosphorylation of ERK, p38 and

JNK in TRAF3-/- tumor B cells. In contrast, PEP005 activates multiple signaling pathways in these cells, including PKCδ, PKCα, PKCε, NF-κB1, ERK, JNK, p38, and Akt. Furthermore, we extended the investigation of AD 198 to TRAF3-sufficient malignant B cells, and found that AD198 also potently inhibits the viability and suppresses c-Myc expression in TRAF3-sufficient mouse and human B lymphoma cell lines. Taken together, our findings suggest that AD 198 has therapeutic potential for the treatment of NHL and MM involving TRAF3 inactivation or c-Myc up-regulation.

120

CHAPTER 4 SIGNALING MECHANISMS OF BORTEZOMIB IN TRAF3-DEFICIENT MOUSE B LYMPHOMA AND HUMAN MULTIPLE MYELOMA CELLS

ABSTRACT

Bortezomib, a clinical drug for multiple myeloma (MM) and mantle cell lymphoma, exhibits complex mechanisms of action, which vary depending on the cancer type and the critical genetic alterations of each cancer. Here we investigated the signaling mechanisms of bortezomib in mouse B lymphoma and human MM cells deficient in a new tumor suppressor gene, TRAF3. We found that bortezomib consistently induced up-regulation of the cell cycle inhibitor p21

(WAF1) and the pro-apoptotic protein Noxa as well as cleavage of the anti- apoptotic protein Mcl-1. Interestingly, bortezomib did not inhibit NF-κB1 and NF-

κB2 pathways, but induced accumulation of the oncoprotein c-Myc. Furthermore, we demonstrated that oridonin (an inhibitor of NF-κB1 and NF-κB2) or AD 198 (a drug targeting c-Myc) drastically potentiated the anti-cancer effects of bortezomib in TRAF3-deficient malignant B cells. Taken together, our findings increase the understanding of the mechanisms of action of bortezomib, which would aid the design of novel bortezomib-based combination therapies. Our results also provide a rationale for clinical evaluation of the combinations of bortezomib and oridonin

(or other inhibitors of NF-κB1/2) or AD 198 (or other drugs targeting c-Myc) in the treatment of lymphoma and MM, especially in patients containing TRAF3 deletions or relevant mutations.

121

4.1 BACKGROUND

Bortezomib (VELCADE, also called PS-341), the first proteasome inhibitor to enter the clinic, is an U.S. Food and Drug Administration (FDA) approved drug for the treatment of newly diagnosed multiple myeloma (MM), relapsed or refractory MM, and mantle cell lymphoma (MCL)191-197. Its specific 26S proteasome inhibitory function has been attributed to the reversible but strong covalent bond formed between its dipeptidyl boronic acid moiety and the threonine proteases of the 20S subunit191-197. Bortezomib thus effectively inhibits the ubiquitin- proteasome pathway (UPP), which is essential in regulating the homeostasis of proteins controlling cell cycle progression, survival and apoptosis191-196. Evidence accumulated in the literature reveals that bortezomib exhibits complex mechanisms of action in cancer cells. These include inhibition of canonical NF-κB activation, suppression of cyclins and c-Myc, up-regulation of inhibitors of cell cycle progression (such as p21, p27, and p53), down-regulation or cleavage of anti- apoptotic proteins of the Bcl-2 family (such as Bcl-2, Bcl-xL, Mcl1, and A1/bfl), and induction of pro-apoptotic proteins of the Bcl-2 family (such as Bim, Bid, Noxa, and

Puma) 191-196. Notably, bortezomib may target different oncogenic pathways in different cellular context, depending on the cancer cell type and the critical genetic alterations of each specific cancer191-196.

Biallelic deletions or inactivating mutations of TRAF3, a novel tumor suppressor gene of B lymphocytes, were recently identified as a critical genetic alteration in human non-Hodgkin lymphoma (NHL), including splenic marginal

122

zone lymphoma (MZL), B cell chronic lymphocytic leukemia (B-CLL), MCL, as well as MM and Waldenström’s macroglobulinemia23,24,42,43,112. Indeed, specific deletion of TRAF3 from B lymphocytes causes prolonged survival of mature B cells, which eventually leads to B lymphoma development in mice15,16. Detailed mechanistic investigation elucidated that TRAF3 inactivation results in constitutive activation of the non-canonical NF-κB (NF-κB2) pathway, which is pivotal for the survival of B cells16,23,40,198,199. Interestingly, human MM patients with TRAF3 deletions or mutations are resistant to the chemotherapy agent dexamethasone but are sensitive to bortezomib23. However, the mechanisms of action of bortezomib in TRAF3-deficient NHL or MM remain unclear.

In the present study, we aimed to investigate the signaling mechanisms of bortezomib using TRAF3-/- mouse B lymphoma cell lines and human MM cell lines with TRAF3 deletions or mutations as model systems. Specifically, the objectives of this study are: (1) to determine the effects of bortezomib on canonical and non- canonical NF-κB pathways in TRAF3-deficient mouse B lymphoma and human

MM cells; (2) to identify which cell cycle regulators are targeted by bortezomib in

TRAF3-deficient malignant B cells; (3) to delineate pro- or anti- apoptotic proteins of the Bcl-2 family that are specifically changed by bortezomib in the context of

TRAF3 inactivation. Information gained from these experiments will be helpful in rational design of combination therapy involving bortezomib and other drugs with distinct mechanisms of action for the treatment of NHL and MM patients containing

TRAF3 deletions or mutations.

123

4.2 MATERIALS AND METHODS

4.2.1 CELL LINES, ANTIBODIES, AND REAGENTS

Human MM cell lines 8226 (contains bi-allelic TRAF3 deletions), KMS11

(contains bi-allelic TRAF3 deletions), and LP1 (contains inactivating TRAF3 frameshift mutations) were kindly provided by Dr. Leif Bergsagel (Mayo Clinic,

Scottsdale, AZ), and were cultured as previously described71. TRAF3-/- mouse B lymphoma cell lines 27-9.5.3, 105-8.1B6, and 115-6.1.2 were generated and cultured as described previously15,71. Bortezomib was generously supplied by

Millennium Pharmaceuticals (Cambridge, MA). Rabbit polyclonal antibody against

Noxa was purchased from AXXORA (Farmingdale, NY). Other antibodies and reagents used in this study were described previously15,71,72,130.

4.2.2 MTT ASSAY, CELL CYCLE ANALYSES, AND CASPASE CLEAVAGE ASSAYS

MTT assay and cell cycle analyses were performed as previously described15,71,72,130. For caspase cleavage assays, total protein lysates were prepared from cells at different time points after treatment with bortezomib, and cleavage of caspases 3, 8, and 9 was subsequently examined by immunoblot analysis.

4.2.3 PROTEIN EXTRACTION AND IMMUNOBLOT ANALYSIS

Total protein lysates, cytosolic and nuclear extracts were prepared as previously described15,16. Immunoblot analysis was performed using various

124

antibodies as described15,71,72,130. Images of immunoblots were acquired using a low-light imaging system (LAS-4000 mini, FUJIFILM Medical Systems USA, Inc.,

Stamford, CT).

4.2.4 STATISTICS

Statistical significance was assessed by Student T test. P values less than

0.05 are considered significant.

4.3 RESULTS

4.3.1 POTENT TUMORICIDAL ACTIVITY OF BORTEZOMIB ON TRAF3-/- MOUSE B LYMPHOMA AND HUMAN MM CELL LINES

It has been previously shown that human MM patients with TRAF3 deletions or mutations are sensitive to bortezomib23. This prompted us to test the tumoricidal activity of bortezomib on TRAF3-/- mouse B lymphoma cell lines and human MM cell lines with TRAF3 deletions or mutations. The results of our MTT assays demonstrated that bortezomib exhibited potent anti-proliferative/survival-inhibitory effects on all the examined TRAF3-/- mouse B lymphoma and human MM cell lines in a dose-dependent manner (Fig. 29A). To understand the mechanism of bortezomib, we performed cell cycle analysis using propidium iodide (PI) staining followed by flow cytometry. We found that bortezomib induced TRAF3-/- mouse B lymphoma and human MM cells to undergo apoptosis, as demonstrated by the drastic increase of the sub-G1 population with DNA content < 2n (Fig. 29B).

125

Bortezomib also inhibited the proliferation of TRAF3-/- tumor B cells, as shown by the marked decrease of the population at the S/G2/M phase (2n < DNA content ≤

4n) (Fig. 29B).

Figure 29. Bortezomib induced apoptosis in TRAF3 /- mouse B lymphoma and human MM cells. TRAF3-/- mouse B lymphoma cell lines examined include 27-9.5.3 (27-9), 105-8.1B6 (105-8), and 115-6.1.2 (115-6). Human patient-derived MM cell lines examined include 8226 (with TRAF3 biallelic deletions), KMS11 (with TRAF3 bi-allelic deletions), and LP1 (with TRAF3 frameshift mutations). (A) Viable cell curves analyzed by MTT assay. Cells were treated with various concentrations (1:2 serial dilutions) of bortezomib for 24 h. Total viable cell numbers were subsequently determined by MTT assay. The graphs depict the results of three independent experiments with duplicate samples in each experiment (mean ± SEM). (B) Cell cycle distribution determined by PI staining and flow cytometry. Cells were cultured in the absence or presence of bortezomib for 24 h. Concentrations of bortezomib used: 10 nM for 27-9, 5 nM for 105-8, 10 nM for 115-6, 50 nM for 8226, 20 nM for KMS11, and 50 nM for LP1. Cells were fixed, and then stained

126

with PI. Stained cells were subsequently analyzed by FACS. Representative FACS histograms of PI staining are shown, and percentages of apoptotic cells (DNA content < 2n; sub-G1) and proliferating cells (2n < DNA content ≤ 4n; S/G2/M) are indicated. Results are representative of three independent experiments.

We verified bortezomib-induced apoptosis using annexin V staining, which showed phosphatidylserine exposure (Fig. 30A). We further demonstrated that bortezomib triggered mitochondrial membrane permeabilization as analyzed by MitoProbe JC-

1 staining (Fig. 30B). We next determined whether bortezomib induced the activation of key caspases involved in apoptosis. We found that bortezomib induced rapid activation of caspases 9, 8, and 3, as evidenced by the cleavage of these caspases after treatment with bortezomib in TRAF3-/- mouse B lymphoma and human MM cell lines (Fig. 31A). Collectively, our data demonstrate that bortezomib induces caspase-mediated apoptosis via the mitochondrial apoptotic pathway in TRAF3-/- malignant B cells.

127

Figure 30. Bortezomib induced apoptosis in TRAF3 /- mouse B lymphoma and human MM cells. Cells were cultured in the absence or presence of bortezomib for 24 h. Concentrations of bortezomib used: 10 nM for 27-9, 5 nM for 105-8, 10 nM for 115-6, 50 nM for 8226, 20 nM for KMS11, and 50 nM for LP1. (A) Cell apoptosis analyzed by annexin V staining. Cells were stained with annexin V and then analyzed by a flow cytometer. Representative FACS histograms are shown, and gated populations indicate annexin V+ apoptotic cells. (B) Mitochondria membrane potential change induced by bortezomib. Mitochondria membrane permeabilization was measured using a MitoProbe JC-1 Assay Kit (Molecular Probes), and then analyzed by FACS. Representative FACS profiles are shown. Gated populations in H indicate cells with heathy mitochondria, and gated populations in P indicate cells with permeabilized mitochondria. Results (A and B) are representative of three independent experiments.

128

Figure 31. Bortezomib induced cleavage of caspases but did not affect NF-κB activation in TRAF3-/- malignant B cells. (A) Cleavage of caspases. Cells were cultured in the absence or presence of bortezomib for indicated time periods. Concentrations of bortezomib used: 10 nM for 27-9, 5 nM for 105-8, 10 nM for 115-6, 50 nM for 8226, 20 nM for KMS11, and 50 nM for LP1. Total cellular lysates were prepared, and then immunoblotted for caspase 9, caspase 8, and caspase 3, followed by actin. Immunoblots of actin were used as loading control. (B) Cytosolic and nuclear levels of NF-κB subunits. Cells were cultured in the absence or presence of bortezomib for indicated time periods. Cytosolic and nuclear proteins were extracted, and then immunoblotted for NF-κB2 (p100 – p52), RelB, NF-κB1 (p50), c-Rel, and RelA, followed by HDAC1 and actin. Immunoblots of actin and HDAC1 were used as loading control for cytosolic and nuclear proteins, respectively. Similar results

129

were also obtained with other cell lines examined. Results (A and B) are representative of three independent experiments.

4.3.2 BORTEZOMIB DID NOT INHIBIT CANONICAL OR NON- CANONICAL NF-B PATHWAY

From the beginning of its clinical use, the principal mechanism of bortezomib was believed to be the inhibition of the NF-B pathway, which promotes cell survival by activating anti-apoptotic genes192-195. However, contradictory effects of bortezomib on the canonical NF-B pathway have been reported in NHL and MM194,200,201. Ma et al. showed that bortezomib substantially inhibits the canonical NF-B pathway in chemosensitive and chemoresistant MM cell lines200. Inhibition of the canonical NF-B pathway has also been demonstrated in MCL and primary effusion lymphomas (PEL)193. In contrast,

Hideshima et al. reported that bortezomib induces canonical NF-B activation in

MM cell lines and primary tumor from MM patients201. In addition, Markovina et al. found that constitutive NF-B activity in primary tumor cells from MM patients is refractory to inhibition by bortezomib202. We recently demonstrated that TRAF3-/- mouse B lymphomas exhibit constitutive activation of both canonical and non- canonical NF-B pathways15. It is known that both NF-B pathways require proteasome-dependent activity, including IB degradation of the canonical pathway and processing of p100 NF-B2 of the non-canonical pathway23,194,201,203.

In this context, we examined and compared the cytosolic and nuclear levels of proteins of both canonical and non-canonical NF-B pathways in TRAF3-/-

130

malignant B cells after treatment with bortezomib. Our results demonstrated that bortezomib did not change the cytosolic or nuclear levels of any proteins of both

NF-B pathways in TRAF3-/- mouse B lymphoma and human MM cell lines (Fig.

31B, and 32). Thus, the anti-proliferative/survival-inhibitory effects of bortezomib are independent of both NF-B pathways in TRAF3-/- malignant B cells.

Figure 32. Bortezomib did not inhibit NF-κB pathways in TRAF3-/- malignant B cells. Cells were cultured in the absence or presence of bortezomib of indicated concentrations for 6h. Cytosolic and nuclear proteins were extracted, and then immunoblotted for NF-κB2 (p100 – p52), RelB, NF-κB1 (p50), c-Rel, and RelA, followed by HDAC1 and actin. Immunoblots of actin and HDAC1 were used as loading control for cytosolic and nuclear proteins, respectively. Similar results were also obtained with other cell lines examined. Results are representative of three independent experiments.

131

4.3.3 BORTEZOMIB INDUCED ACCUMULATION OF P21 AND C-MYC

Among the proteins degraded by the ubiquitin-proteasome pathway are cyclins (A, B, D and E), cyclin-dependent kinase inhibitors such as p21 and p27, the tumor suppressor p53 and the oncoprotein c-Myc192,193,196,204. Decreased degradation and increased accumulation of these proteins may disrupt cell cycle progression192,193,196,204. To understand its mechanisms of action, we then investigated the effects of bortezomib on the protein levels of these cell cycle regulators. We found that treatment with bortezomib led to a time-dependent increase in total cellular levels of the cyclin-dependent kinase inhibitor p21 (WAF1) in all examined TRAF3-/- mouse B lymphoma and human MM cell lines (Fig. 33A).

Interestingly, accumulation of c-Myc was induced by bortezomib in all examined cell lines except 105-8 (Fig. 33A). Although total cellular levels of cyclin D2 were decreased and those of cyclin B1 were increased by bortezomib in all TRAF3-/- mouse B lymphoma cell lines, these observations were not mirrored in human MM cell lines (Fig. 33A). In contrast, bortezomib did not consistently affect the protein levels of other cell cycle regulators in different cell lines examined in our study, including p27, cyclin D1, cyclin A, and p53 (Fig. 33A). In light of the consistent accumulation of p21 in all examined cell lines, we further determined the subcellular distribution of p21 in TRAF3-/- malignant B cells after treatment with bortezomib. Our results showed that bortezomib increased p21 protein levels in both the cytosol and nucleus in a dose-dependent and time-dependent manner

(Fig. 33B and 34). Our findings thus suggest that bortezomib-induced

132

accumulation of the cell cycle inhibitor p21 is associated with its anti-proliferative effects in TRAF3-/- malignant B cells.

Figure 33. Bortezomib increased the accumulation of p21 and c-Myc in TRAF3-/- malignant B cells. (A) Time-dependent effects of bortezomib on the protein levels of cell cycle regulators. Cells were cultured in the absence or presence of bortezomib for indicated time periods. Concentrations of bortezomib used: 10 nM for 27-9, 5 nM for 105-8, 10 nM for 115-6, 50 nM for 8226, 20 nM for KMS11, and 50 nM for LP1. Total cellular lysates were prepared, and then immunoblotted for p21, p27, cyclin D1, cyclin D2, cyclin B1, cyclin A, p53, and c-Myc, followed by actin. Immunoblots of actin were used as loading control. (B) Dose-dependent effects of bortezomib on p21 protein levels in the cytosol and nucleus. Cells were cultured in the absence or presence of bortezomib of indicated concentrations for 6 h. Cytosolic and nuclear proteins were extracted, and then

133

immunoblotted for p21, followed by HDAC1 and actin. Immunoblots of actin and HDAC1 were used as loading control for cytosolic and nuclear proteins, respectively. Similar results were also obtained with other cell lines examined. Results (A and B) are representative of three independent experiments.

Figure 34. Bortezomib increased the accumulation of p21 in TRAF3-/- malignant B cells. Cells were cultured in the absence or presence of bortezomib for indicated time periods. Cytosolic and nuclear proteins were extracted, and then immunoblotted for p21, followed by HDAC1 and actin. Immunoblots of actin and HDAC1 were used as loading control for cytosolic and nuclear proteins, respectively. Similar results were also obtained with other cell lines examined. Results are representative of three independent experiments.

4.3.4 BORTEZOMIB INDUCED UP-REGULATION OF NOXA AND CLEAVAGE OF MCL1

Pro- and anti-apoptotic proteins of the Bcl-2 family are also known substrates of the proteasome. Increasing evidence indicates that proteasome inhibition-dependent regulation of the Bcl-2 family is required for the anti-tumor activity of bortezomib191-194. In particular, stabilization of BH3-only proteins Bik,

Noxa and Bim, appears to be essential for effecting Bax/Bak-dependent apoptosis triggered by bortezomib191-194. However, the effects of bortezomib on the Bcl-2 family proteins are variable and appear to be context-dependent191-194. We thus examined the regulation of the Bcl-2 family proteins by bortezomib in TRAF3-/- malignant B cells. We found that bortezomib induced striking up-regulation of the

134

pro-apoptotic protein Noxa in all examined TRAF3-/- mouse B lymphoma and human MM cell lines (Fig. 35A). Interestingly, bortezomib also induced the phosphorylation and cleavage of the anti-apoptotic protein Mcl-1 in all examined cell lines (Fig. 35A). In contrast, although increases of Bax, Bak and Bim as well as decreases in Bcl-2, Bcl-xL and A1/Bfl have been previously observed in NHL or

MM after treatment with bortezomib191-194, we did not detect any effects of bortezomib on these proteins in all examined TRAF3-/- cell lines (Fig. 35A and

36A).

135

Figure 35. Bortezomib induced up-regulation of Noxa and cleavage of Mc1-1 in TRAF3-/- malignant B cells. (A) Time-dependent effects of bortezomib on the protein levels of cell survival regulators of the Bcl- 2 family. Cells were cultured in the absence or presence of bortezomib for indicated time periods. Concentrations of bortezomib used: 10 nM for 27-9, 5 nM for 105-8, 10 nM for 115-6, 50 nM for 8226, 20 nM for KMS11, and 50 nM for LP1. Total cellular lysates were prepared, and then immunoblotted for Noxa, phosphorylated Mcl-1, Mcl-1, Bcl-xL, Bcl2, Bik, Bax, and Bak, followed by actin. Immunoblots of actin were used as loading control. (B) Dose-dependent effects of bortezomib on the protein levels of Noxa and Mcl-1. Cells were cultured in the absence or presence of

136

bortezomib of indicated concentrations for 6 h. Total cellular lysates were prepared, and then immunoblotted for Noxa, phosphorylated Mcl-1, and Mcl-1, followed by actin. Similar results were also obtained with other cell lines examined. Results (A and B) are representative of three independent experiments.

Bid cleavage was induced by bortezomib in two TRAF3-/- mouse B lymphoma cell lines (27-9 and 115-6), but this effect was not recapitulated in human MM cell lines

(Fig. 36A). Similarly, Bik was decreased by bortezomib in TRAF3-/- mouse B lymphoma cell lines, but slightly increased by bortezomib in human MM cell lines

(Fig. 35A). Therefore, the most consistent effects of bortezomib on the Bcl-2 family are the induction of Noxa up-regulation and Mcl-1 cleavage in TRAF3-/- malignant

B cells. Given the previous evidence that Noxa induction and Mcl-1 cleavage are crucial for the cytotoxic effects of bortezomib on MM191-194, we further verified that bortezomib induced Noxa up-regulation and Mcl-1 cleavage in a dose-dependent and time-dependent manner in TRAF3-/- malignant B cells (Fig. 35B). Taken together, our results highlight the importance of Noxa and Mcl-1 in bortezomib- mediated tumoricidal activities in TRAF3-/- malignant B cells.

137

Figure 36. Effects of bortezomib on the protein levels of cell survival regulators in TRAF3-/- malignant B cells. Cells were cultured in the absence or presence of bortezomib for indicated time periods. Concentrations of bortezomib used: 10 nM for 27-9, 5 nM for 105-8, 10 nM for 115-6, 50 nM for 8226, 20 nM for KMS11, and 50 nM for LP1. Total cellular lysates were prepared, and then immunoblotted for A1/Bfl, Bcl2l10, Bim, Bad, Bid, and Bif1, followed by actin. Immunoblots of actin were used as loading control. Similar results were also obtained with other cell lines examined. Results are representative of three independent experiments.

4.3.5 COMBINED EFFECTS OF BORTEZOMIB AND ORIDONIN OR AD 198 Detailed elucidation of its mechanisms of action will not only provide better understanding of the molecular determinants of sensitivity of malignant B cells to bortezomib, but also suggest new, rational combination therapies with potential to improve outcome in patients. Our evidence that bortezomib did not inhibit canonical and non-canonical NF-B pathways prompted us to test the hypothesis that an NF-B inhibitor may enhance the anti-tumor activity of bortezomib on

138

TRAF3-/- malignant B cells. We recently found that oridonin, an active diterpenoid isolated from a traditional Chinese herbal medicine, potently inhibits both canonical and non-canonical NF-B pathways in TRAF3-/- mouse B lymphoma cells 15. We thus compared the cytotoxic effects of bortezomib and oridonin, alone or in combination, on TRAF3-/- mouse B lymphoma and human MM cell lines. Our results demonstrated that bortezomib and oridonin exhibited additive or synergistic tumoricidal activities on all examined cell lines except 8226 (Fig. 37A).

Furthermore, our finding that bortezomib induced the accumulation of c-Myc in most examined cell lines suggests that simultaneous targeting of the proteasome and c-Myc may be more effective in the treatment of TRAF3-deficient B cell neoplasms. We thus examined the effects of bortezomib in combination with AD

198, a new drug that we recently found to specifically target c-Myc in malignant B cells71. We found that AD 198 potently enhanced the cytotoxic activities of bortezomib in all examined TRAF3-/- cell lines (Fig. 37B). Taken together, our results provide a rationale for further clinical evaluation of the combinations of bortezomib and oridonin or AD 198 in the treatment of NHL and MM patients containing TRAF3 deletions or relevant mutations.

139

Figure 37. Oridonin or AD 198 potently augmented the cytotoxic activity of bortezomib on TRAF3 /- malignant B cells. Cells were treated with indicated concentrations of bortezomib (BTZ), oridonin (Ori) or AD 198 (AD), alone or in combination, for 24 h. Total viable cell numbers were subsequently determined by MTT assay. (A) Combined effects of bortezomib and oridonin. (B) Combined effects of bortezomib and AD 198. The graphs depict the results of three independent experiments with duplicate samples in each experiment (mean ± SEM). *, significantly different from single agent treatment (t test, p < 0.05); **, very significantly different from single agent treatment (t test, p < 0.005); ***, highly significantly different from single agent treatment (t test, p < 0.001).

140

4.4 DISCUSSION

During the past decade, the clinical success of the proteasome inhibitor bortezomib as a front-line and second-line regimen has dramatically improved the outcome of patients with MM and MCL196,197,205. However, it should be noted that bortezomib also has side-effects, including peripheral neuropathy, diarrhea, vomiting, fatigue, dizziness, anorexia, anemia, and thrombocytopenia192,193,195,196,204,206. Moreover, some patients do not benefit from bortezomib treatment, about 60% of responding patients eventually acquire resistance, and relapse is common in patients that initially responded to bortezomib195,205,206. In this context, although bortezomib has demonstrated efficacy as monotherapy, combination therapies are expected to maximize benefit while minimizing toxicity194,197,205,206. Understanding of detailed signaling mechanisms of bortezomib provides a basis for rational design of novel combination therapies that target non-overlapping pathways. Given the significant heterogeneity in the outcomes despite similar clinical stage classifications as seen in MM patients, it is now recognized to be crucial to consider genetic risk factors for optimum therapeutic recommendation for individual groups of patients196. In the present study, we investigated detailed signaling mechanisms of bortezomib in mouse B lymphoma and human MM cells deficient in a new tumor suppressor gene, TRAF3.

We identified the most striking and consistent signaling events of bortezomib in TRAF3-deficient malignant B cells, including induction of the cell

141

cycle inhibitor p21 (WAF1) and the pro-apoptotic protein Noxa as well as cleavage of the anti-apoptotic protein Mcl-1. Our findings corroborate several previous studies of p21, Noxa, and Mcl-1 in MM cells207-209. Induction of p21 is known to inhibit proliferation and induce cell cycle arrest predominantly at the G1/S checkpoint192,193,196,204. Noxa promotes apoptosis by directly interacting with Mcl-

1, thereby releasing the apoptotic protein Bak to trigger mitochondrial depolarization195,208,210,211. Furthermore, the cleaved Mcl-1 fragment (28 kDa) has strong pro-apoptotic activity by enhancing its turnover and impairing its ability to interact with BH3-only pro-apoptotic proteins208,209. Indeed, silencing of p21 reverses the anti-proliferative effects of bortezomib212-214, knockdown of Noxa protects against apoptosis induced by bortezomib, and overexpression of cleaved

Mcl-1 sensitize MM cells to bortezomib-induced apoptosis208,210,211. Collectively, up-regulation of p21 and Noxa together with cleavage of Mcl-1 appears to be key mediators of the cytotoxic effects of bortezomib in TRAF3-/- malignant B cells.

Therefore, to improve the efficacy or prevent relapse, we recommend the combination of bortezomib with agents independent of p21, Noxa and Mcl-1.

Although NF-B inhibition was initially believed to be responsible for the anti-cancer activity of bortezomib193,195,196,204,205, we did not detect any inhibition of

NF-B1 or NF-B2 pathways in TRAF3-/- malignant B cells. Our results reinforce recent evidence that argues against a critical role for NF-B inhibition in the mechanisms of action of bortezomib in MM201,202,205. Notably, NF-κB1 and NF-κB2

142

control the expression of a variety of target genes crucial for cell survival (e.g., Bcl- xL and Mcl-1), proliferation (e.g., cyclin D1 and c-Myc), adhesion (e.g., CD54 and

CD106), and cytokine signaling (e.g., IL-6 and BAFF secretion)193,195,196,204,205.

Frequent genetic abnormalities of the NF-B2 pathway in MM further highlight the importance of NF-B2 signaling in the pathogenesis of MM23,24. We recently demonstrated that knockdown of NF-B2 induced apoptosis in TRAF3-/- mouse B lymphoma cells15. Based on the above evidence, we tested the combined effects of bortezomib and an inhibitor of both NF-κB1 and NF-κB2, oridonin15. We found that co-treatment with bortezomib and oridonin remarkably potentiated the anti- cancer effects of bortezomib in all examined TRAF3-/- cell lines except 8226 (which might have acquired additional mutations and NF-B-independent oncogenic pathways). Similarly, Fabre et al. recently reported that another inhibitor of NF-κB1 and NF-κB2, PBS-1086, markedly enhanced the cytotoxicity of bortezomib on MM cells215. Therefore, combination of bortezomib and oridonin (or other inhibitors of both NF-κB1 and NF-κB2) represent a rational therapeutic regimen for the treatment of NHL and MM, especially in patients that are refractory to bortezomib- mediated inhibition of NF-κB pathways.

Interestingly, bortezomib induced the accumulation of c-Myc, an oncoprotein of B cells170, in most examined TRAF3-/- cell lines. We thus also examined the combined effects of bortezomib and a drug specifically targeting c-

Myc, AD 19871. We found that this combination killed TRAF3-/- malignant B cells to

143

a much greater extent than either drug alone. Given the ubiquitously elevated expression of c-Myc in B cell malignancies172, our findings suggest that combination of bortezomib and AD 198 (or other drugs targeting c-Myc) is another promising therapy for NHL and MM.

4.5 CONCLUSION

In summary, our study provides a better understanding of the mechanistic basis for the anti-cancer effects of bortezomib in NHL and MM, which would aid the design of novel bortezomib-based combination regimens. Here we found that combinations of bortezomib with oridonin or AD 198 allow the use of lower doses of bortezomib to reduce side-effects while enhancing anti-cancer efficacy in

TRAF3-/- malignant B cells through complementary mechanisms of action. Our findings thus provide the preclinical framework for clinical evaluation of the combinations of bortezomib and oridonin or AD 198, which would expand the therapeutic armamentarium for NHL and MM.

SUMMARY AND SIGNIFICANCE

Identification and validation of novel genetic risk factors and oncogenic signaling pathways for B cell neoplasms are necessary for the development of new therapeutic strategies. Inactivation of TRAF3, a recently identified tumor suppressor in B cells, results in prolonged survival of mature B cells, which eventually leads to spontaneous development of B lymphomas in mice 15,16. In addition, TRAF3 deletions and mutations frequently occur in human B cell chronic lymphocytic leukemia, splenic marginal zone lymphoma, mantle cell lymphoma, multiple myeloma, Waldenström’s macroglobulinemia, and Hodgkin lymphoma

23,24,42,43,112. In this context, we have been investigating TRAF3 signaling mechanisms in B cells, and are developing new therapeutic strategies to target

TRAF3 downstream signaling pathways in B cell neoplasms. We have obtained translational data that demonstrate the therapeutic potential of targeting TRAF3 downstream signaling pathways in B lymphoma and multiple myeloma.

In our study of the signaling mechanisms at work in TRAF3-deficient B cell cancers, we have shown that MCC appears to be oncogenic in these types of malignancies, which is in direct contrast to the tumor suppressive functions of MCC in colorectal cancer, where it was first identified 47,49,58,62,216. As we observed from our microarray and preliminary data, MCC is significantly upregulated in mouse

TRAF3-deficient B lymphomas, as well as in human patient-derived multiple myeloma cell lines, and primary human blood cancers. Our study of MCC showed that it affects different signaling pathways from those targeted by MCC in colorectal

145

cancer, giving it an overall oncogenic effect in TRAF3-deficient B cell neoplasms.

Our overexpression and knockdown studies in human multiple myeloma cell lines showed the downstream targets of MCC involved in cell signaling, cell survival, and cell proliferation. These include Mcl-1, caspase 8, caspase 3, p27, cyclin B1, phospho-ERK and c-Myc, which strengthen the conclusion that MCC is a mediator of prolonged cell survival and a promoter of cell proliferation in B cell cancers. More proximally, we have identified two interacting partners of MCC, prohibitin 2 and

PARP1, which have been shown to affect the aforementioned downstream targets of MCC (Fig. 38). Prohitibin 2 and PARP1 therefore have the potential to be therapeutic targets for the treatment of TRAF3-deficient B cell malignancies.

In our microarray, we also identified Sox5 as being significantly upregulated in mouse B lymphoma at both the mRNA and protein levels. Sox5 encodes a transcription factor and a member of the SoxD family of genes, all of which express long and short isoforms due to alternative splicing 113. Sox5 proteins modulate cell proliferation, survival, differentiation, or terminal maturation in different cell lineages, and in doing so play a role in development and cell fate determination.

Upregulation of Sox5 mRNA has been identified in human memory B cells, follicular lymphoma and other diseases125-127. In mouse B lymphomas, we identified a novel isoform of Sox5, Sox5-BLM, a nuclear protein containing a 35 amino acid deletion in front of the leucine zipper region. Overexpression of Sox5 in human multiple myeloma cell lines up-regulated p27 and β-catenin expression, and inhibited cell cycle progression, indicating that it is not oncogenic in this type

146

of B cell malignancy, but regulates the proliferation of malignant B cells and has the potential to be used as a marker in B cell neoplasms.

Figure 38. MCC, a Novel Oncogene in B Lymphocytes. In the context of TRAF3 deletion or inactivating mutation (which results in prolonged B cell survival), MCC interacts with PARP1 and PHB1/2, which have downstream effects on several proteins that promote cell survival and/or cell proliferation, including Mcl1, caspase 8, ERK, p27, c- Myc, Cyclin B1 and caspase 3. We propose that the signaling cascades of TRAF3 inactivation and MCC in combination result in B cell malignant transformation.

Having identified several new potential targets and/or markers of B cell cancers, including TRAF3, PKCδ, NF-B, MCC, PHB2, and PARP1, we can exploit their therapeutic potential for treatment of these kinds of cancers. It is already known that TRAF3-deletion prevents PKCδ nuclear translocation, one of the mechanisms by which TRAF3-deficient B cells have prolonged cell survival. We sought to restore this pathway using two PKCδ activators, AD 198 and PEP005.

147

PEP005 showed contradictory pro- and anti-tumor effects in our TRAF3-deficient

B cell cancer cell lines. However, we found that AD 198 did exhibit potent anti- tumor effects on TRAF3-deficient mouse B lymphomas and human multiple myelomas, but did so independently of PKCδ, instead suppressing c-Myc expression to cause cell death. Nevertheless, the potent nature of AD 198 suggests that it has great potential for treatment of B cell cancers involving TRAF3 inactivation or c-Myc up-regulation.

One of the newer methods of cancer treatment is multi-drug therapies, which use combinations of drugs that have different targeted effects to attack cancer cells217,218. We used this strategy to do some in vitro work using a drug combination therapy combining AD 198 or oridonin with bortezomib, a proteasome inhibitor clinically used for the treatment of multiple myeloma. Oridonin is an NF-

B2 inhibitor that also exhibits very effective anti-tumor effects on TRAF3-deficient

B cell cancers as shown in a previous study of our laboratory 15. In our TRAF3- deficient B cell cancer lines, we saw that bortezomib displayed an anti-tumor effect, up-regulating p21 (a cell cycle inhibitor) and Noxa (a pro-apoptotic protein) as well as inducing Mcl-1 cleavage, all of which lead to apoptosis. However, unlike in other studies, bortezomib did not inhibit the NF-kB1 nor the NFkB2 pathway, and bortezomib treatment resulted in c-Myc accumulation. When we combined bortezomib with oridonin or AD 198 (a c-Myc repressing drug identified in this study), we saw that the total tumoricidal effect of the combined drugs significantly potentiated that of bortezomib alone, and in a manner that was much more drastic

148

than the expected additive effect of two drugs. This indicates that combined drug therapy may be a viable and efficacious option for the treatment of B cell cancers.

Figure 39. Complex mechanisms of TRAF3 inactivation-mediated B lymphomagenesis. Based on previous understanding of signaling mechanisms of TRAF3 in B cells, we tested drugs targeting NF-kB2 and PKCδ, two downstream signaling components of TRAF3. We found that both oridonin and AD 198 exhibit potent anti-tumor activities on TRAF3-/- malignant B cells. Oridonin acts primarily as an inhibitor of NF-κB2. Although we originally tested AD 198 as an activator of PKCδ, our results revealed that AD 198 does not induce PKCδ nuclear translocation but specifically targets c-Myc in malignant B cells. Therefore, further studies (using genetic means or more specific PKCδ activators) are required to evaluate the therapeutic potential of activating PKCδ nuclear translocation. B cells purified from young B-TRAF3-/- mice exhibit prolonged survival but do not proliferate autonomously, and are therefore premalignant B cells. This suggests that TRAF3 inactivation and its downstream signaling pathways are not sufficient for B lymphomagenesis and that additional oncogenic alterations are required. To identify such alterations, we performed a microarray analysis and identified 160 up-regulated genes and 244 down-regulated genes in TRAF3-/- B lymphomas, including aberrant expression of MCC and a novel isoform of Sox5 (Sox5- BLM). Our proteomic data identified PARP1 and PHB2 as two signaling hubs of the novel oncogene MCC in human MM cells, we are currently testing the therapeutic efficacy of PARP1 inhibitors and PHB2 ligands in B cell cancers with TRAF3 inactivation or aberrant MCC expression. Furthermore, our studies also suggest that combination therapy using the clinical drug bortezomib and drugs targeting TRAF3 or MCC downstream signaling components may greatly improve the therapeutic efficacy of bortezomib in B cell cancers. Indeed, we have obtained preclinical evidence demonstrating the synergistic effects of bortezomib in combination with oridonin or AD 198 on killing TRAF3-deficient B cell cancers.

149

In summary, my dissertation research has identified aberrant expression of

MCC or Sox5 as candidate diagnostic markers for B cell malignancies. We also discovered a number of candidate therapeutic targets for the treatment of B cell cancers, including MCC, PARP1, and PHB2. Based on previous understanding of

TRAF3 signaling mechanisms of TRAF3 in B cells, we tested drugs targeting PKCδ

(a downstream signaling component of TRAF3), and found that AD198 exhibits potent anti-tumor activities on B cell cancers with TRAF3 deletions or mutations.

Although we originally tested AD 198 as an activator of PKCδ, our results revealed that AD 198 does not affect PKCδ but specifically targets c-Myc in malignant B cells. Of particular interest, our studies suggest the promising potential of combination therapy to treat TRAF3-deficient B cell tumors, as evidenced by our preclinical data of dual drug treatments using bortezomib with either oridonin or

AD 198 (Fig. 39). Since our new proteomic data identified PARP1 and PHB2 as two signaling hubs of the novel oncogene MCC in human MM cells, our future work will test the therapeutic efficacy of PARP1 inhibitors and PHB2 ligands in B cell cancers with TRAF3 inactivation or aberrant MCC expression (Fig. ii). Altogether, our findings indicate that restoring TRAF3 downstream signaling pathways or targeting MCC downstream signaling components represent important therapeutic opportunities for B cell cancers.

APPENDIX 1: TABLES

Table 1. Genes differentially expressed in TRAF3-/- mouse splenic B lymphomas identified by the microarray analysis # GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 1 Zcwpw1 zinc finger, CW type with PWWP domain 1 3.36 9.50 14.20 4.07E-07 0.000126365 2 Diras2 DIRAS family, GTP-binding RAS-like 2 3.35 10.56 14.57 3.32E-07 0.000126365 3 Serpina3f serine (or cysteine) peptidase inhibitor, clade A, member 3.03 10.87 9.99 6.54E-06 0.000571 3F 4 Sox5 SRY-box containing gene 5 2.93 9.48 13.97 4.65E-07 0.000130244 5 C130026I21Rik RIKEN cDNA C130026I21 gene 2.87 9.95 5.58 0.000456113 0.008256021 6 Tnfrsf19 tumor necrosis factor receptor superfamily, member 19 2.87 9.00 13.59 5.78E-07 0.000148597 7 Mcc mutated in colorectal cancers 2.76 8.68 31.01 6.98E-10 9.15E-06 8 Slamf9 SLAM family member 9 2.75 10.73 4.75 0.00131062 0.016097166 9 Fah fumarylacetoacetate hydrolase 2.63 11.18 7.83 4.13E-05 0.00178595 10 Rdh12 retinol dehydrogenase 12 2.61 11.29 12.81 9.27E-07 0.000205961 11 Ahnak2 AHNAK nucleoprotein 2 2.54 8.93 12.49 1.13E-06 0.000226433 12 Chst7 carbohydrate (N-acetylglucosamino) sulfotransferase 7 2.42 9.23 7.85 4.05E-05 0.001759595 13 Vars valyl-tRNA synthetase 2.30 11.46 16.48 1.23E-07 8.74E-05 14 Twsg1 twisted gastrulation homolog 1 (Drosophila) 2.30 10.96 10.51 4.41E-06 0.000441575 15 Tbc1d9 TBC1 domain family, member 9 2.29 9.53 17.05 9.34E-08 7.20E-05 16 Vars valyl-tRNA synthetase 2.13 9.29 17.17 8.83E-08 7.20E-05 17 Cd59a CD59a antigen 2.11 10.04 5.72 0.000385181 0.007336917 18 Gnb3 guanine nucleotide binding protein (G protein), beta 3 2.09 8.73 5.99 0.000283427 0.006099042 19 Sspn sarcospan 2.08 10.14 4.98 0.000958982 0.013242851 20 Lacc1 laccase (multicopper oxidoreductase) domain containing 2.08 8.98 19.86 2.70E-08 5.06E-05 1 21 Gbp1 guanylate binding protein 1 2.06 9.78 4.19 0.00279695 0.02699119 22 Vars valyl-tRNA synthetase 2.06 12.05 21.24 1.56E-08 3.41E-05 23 Ccbp2 chemokine binding protein 2 1.96 9.96 7.62 5.07E-05 0.002000536

150

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 24 Gng13 guanine nucleotide binding protein (G protein), gamma 13 1.94 8.60 7.33 6.70E-05 0.00236437 25 Plscr1 phospholipid scramblase 1 1.94 10.35 6.13 0.000239207 0.005519038 26 Clic4 chloride intracellular channel 4 (mitochondrial) 1.93 12.48 7.57 5.33E-05 0.002054453 27 Sema7a sema domain, immunoglobulin domain (Ig), and GPI 1.91 9.22 9.70 8.22E-06 0.000643485 membrane anchor, (semaphorin) 7A 28 Chrnb1 cholinergic receptor, nicotinic, beta polypeptide 1 1.89 9.84 8.35 2.55E-05 0.001291342 (muscle) 29 Ppap2b phosphatidic acid phosphatase type 2B 1.86 10.46 18.00 6.01E-08 7.20E-05 30 Ebi3 Epstein-Barr virus induced gene 3 1.82 9.85 10.82 3.51E-06 0.00039538 31 4930539E08Rik RIKEN cDNA 4930539E08 gene 1.81 8.92 8.10 3.22E-05 0.001536721 32 Nacc2 nucleus accumbens associated 2, BEN and BTB (POZ) 1.79 9.23 12.76 9.57E-07 0.000209017 domain containing 33 Fcrl5 Fc receptor-like 5 1.79 9.62 8.79 1.74E-05 0.001046831 34 Sel1l3 sel-1 suppressor of lin-12-like 3 (C. elegans) 1.76 8.30 7.70 4.70E-05 0.001893452 35 Dnajc7 DnaJ (Hsp40) homolog, subfamily C, member 7 1.76 10.73 6.73 0.000124642 0.003589973 36 Pdlim1 PDZ and LIM domain 1 (elfin) 1.74 11.75 6.79 0.000116482 0.003392213 37 Abca3 ATP-binding cassette, sub-family A (ABC1), member 3 1.71 10.86 11.21 2.66E-06 0.0003633 38 Rassf4 Ras association (RalGDS/AF-6) domain family member 4 1.66 11.54 10.68 3.89E-06 0.000397578 39 Kynu kynureninase (L-kynurenine hydrolase) 1.65 10.56 14.43 3.57E-07 0.000126365 40 Cd274 CD274 antigen 1.64 11.38 16.17 1.43E-07 8.74E-05 41 Cd80 CD80 antigen 1.64 8.83 12.33 1.25E-06 0.00023571 42 Ccdc28b coiled coil domain containing 28B 1.64 10.34 7.19 7.74E-05 0.002628686 43 Rassf4 Ras association (RalGDS/AF-6) domain family member 4 1.63 11.53 11.04 3.00E-06 0.000376611 44 Sp140 Sp140 nuclear body protein 1.63 9.80 11.51 2.15E-06 0.000322971 45 AF067061 cDNA sequence AF067061 1.63 8.54 4.77 0.001275097 0.015838995 46 Ero1lb ERO1-like beta (S. cerevisiae) 1.63 11.22 12.45 1.16E-06 0.000226433 47 Nfatc1 nuclear factor of activated T cells, cytoplasmic, calcineurin 1.61 9.89 8.27 2.76E-05 0.001378762 dependent 1 48 Hdac4 histone deacetylase 4 1.61 8.78 8.11 3.17E-05 0.001523759 49 Tubb2b tubulin, beta 2B class IIB 1.60 10.65 5.22 0.000712545 0.010897018

151

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 50 Trim40 tripartite motif-containing 40 1.59 8.56 8.79 1.74E-05 0.001046831 51 Plscr1 phospholipid scramblase 1 1.57 9.33 6.70 0.000129091 0.003645988 52 Arhgap24 Rho GTPase activating protein 24 1.52 10.13 11.91 1.65E-06 0.000270244 53 D13Ertd608e DNA segment, Chr 13, ERATO Doi 608, expressed 1.52 8.97 10.25 5.35E-06 0.0005051 54 Hn1l hematological and neurological expressed 1-like 1.52 9.27 15.10 2.49E-07 0.000112734 55 Ccdc28b coiled coil domain containing 28B 1.50 9.62 9.68 8.31E-06 0.000643485 56 Tcf4 transcription factor 4 1.50 12.04 9.19 1.24E-05 0.000847921 57 Zbtb32 zinc finger and BTB domain containing 32 1.49 8.54 9.39 1.05E-05 0.000754882 58 Gbp2 guanylate binding protein 2 1.48 11.56 7.68 4.79E-05 0.001904754 59 Caln1 calneuron 1 1.48 8.43 6.18 0.000228083 0.005366295 60 Gpr34 G protein-coupled receptor 34 1.48 8.77 6.37 0.000182843 0.004607992 61 Serpina3g serine (or cysteine) peptidase inhibitor, clade A, member 1.47 12.85 8.47 2.30E-05 0.001206517 3G 62 Rbm47 RNA binding motif protein 47 1.45 9.83 11.02 3.04E-06 0.000376611 63 Fgd2 FYVE, RhoGEF and PH domain containing 2 1.45 11.91 5.76 0.000368771 0.007141322 64 Fgd6 FYVE, RhoGEF and PH domain containing 6 1.44 9.42 8.42 2.41E-05 0.001236456 65 Neurod4 neurogenic differentiation 4 1.43 8.19 7.07 8.74E-05 0.002845238 66 Rassf4 Ras association (RalGDS/AF-6) domain family member 4 1.42 10.14 7.45 5.94E-05 0.002191762 67 Rhbdf1 rhomboid family 1 (Drosophila) 1.42 10.30 8.74 1.81E-05 0.001056366 68 Gstt3 glutathione S-transferase, theta 3 1.40 9.14 6.47 0.000164429 0.004282628 69 Pafah1b3 platelet-activating factor acetylhydrolase, isoform 1b, 1.40 10.94 5.50 0.000501117 0.00873186 subunit 3 70 Asph aspartate-beta-hydroxylase 1.37 8.04 5.64 0.000426167 0.007789294 71 Nid1 nidogen 1 1.37 8.57 7.77 4.37E-05 0.001817842 72 Sgk3 serum/glucocorticoid regulated kinase 3 1.36 9.50 10.11 5.97E-06 0.000539159 73 Gas7 growth arrest specific 7 1.35 8.81 6.41 0.000176165 0.004524439 74 NA NA 1.34 10.20 11.26 2.56E-06 0.000352915 75 Gm14137 predicted gene 14137 1.34 8.18 17.51 7.51E-08 7.20E-05 76 Prps2 phosphoribosyl pyrophosphate synthetase 2 1.33 9.69 8.55 2.13E-05 0.001175007

152

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 77 Sox5 SRY-box containing gene 5 1.32 8.42 11.10 2.87E-06 0.000365287 78 Hsp90ab1 heat shock protein 90 alpha (cytosolic), class B member 1.30 11.59 7.12 8.32E-05 0.002740232 1 79 Eps15 epidermal growth factor receptor pathway substrate 15 1.30 9.95 9.51 9.50E-06 0.000711445 80 Cybasc3 cytochrome b, ascorbate dependent 3 1.29 10.90 11.68 1.92E-06 0.000303211 81 Fbxw13 F-box and WD-40 domain protein 13 1.29 8.28 9.97 6.61E-06 0.000572834 82 Recql5 RecQ protein-like 5 1.28 9.46 7.40 6.26E-05 0.002267958 83 Kcnk5 potassium channel, subfamily K, member 5 1.28 10.16 10.71 3.81E-06 0.00039567 84 Ptpn22 protein tyrosine phosphatase, non-receptor type 22 1.27 11.57 8.75 1.80E-05 0.001056366 (lymphoid) 85 Pdia4 protein disulfide isomerase associated 4 1.27 12.33 7.50 5.68E-05 0.002143803 86 Ccnd2 cyclin D2 1.27 10.35 7.56 5.37E-05 0.002064785 87 Plac1l placenta-specific 1-like 1.27 8.63 7.49 5.74E-05 0.002156914 88 Gpr137b G protein-coupled receptor 137B 1.26 8.68 8.07 3.31E-05 0.001539334 89 R74862 expressed sequence R74862 1.26 8.57 8.55 2.13E-05 0.001175007 90 Cfp complement factor properdin 1.25 13.35 8.51 2.22E-05 0.001185616 91 Plp2 proteolipid protein 2 1.24 10.88 5.00 0.000937488 0.013156142 92 Blm Bloom syndrome, RecQ helicase-like 1.23 10.01 10.89 3.33E-06 0.000389533 93 Slc29a3 solute carrier family 29 (nucleoside transporters), member 1.23 9.84 11.50 2.17E-06 0.000322971 3 94 Neo1 neogenin 1.23 8.29 5.97 0.00028734 0.006114237 95 Gpr34 G protein-coupled receptor 34 1.23 8.84 3.81 0.004793627 0.038587517 96 Lysmd2 LysM, putative peptidoglycan-binding, domain containing 1.22 8.80 9.68 8.32E-06 0.000643485 2 97 Optn optineurin 1.22 8.75 8.54 2.16E-05 0.001177646 98 D10Wsu102e DNA segment, Chr 10, Wayne State University 102, 1.22 9.14 10.67 3.91E-06 0.000397578 expressed 99 Igsf9 immunoglobulin superfamily, member 9 1.22 8.76 8.60 2.06E-05 0.001143744 100 Hmgn3 high mobility group nucleosomal binding domain 3 1.21 11.16 4.63 0.001537694 0.018154491 101 Man1a mannosidase 1, alpha 1.21 9.81 7.55 5.42E-05 0.002069558 102 Robo1 roundabout homolog 1 (Drosophila) 1.21 8.23 5.95 0.000295706 0.006190469

153

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 103 Hmgn3 high mobility group nucleosomal binding domain 3 1.21 11.28 3.69 0.005709615 0.043452096 104 Apoe apolipoprotein E 1.20 9.14 5.38 0.000582464 0.009505843 105 Oosp1 oocyte secreted protein 1 1.20 10.03 3.57 0.006900257 0.049987767 106 Nsf N-ethylmaleimide sensitive fusion protein 1.20 10.71 10.25 5.36E-06 0.0005051 107 Rap1gap2 RAP1 GTPase activating protein 2 1.19 9.90 5.20 0.000724814 0.011044987 108 Camk2d calcium/calmodulin-dependent protein kinase II, delta 1.19 9.22 8.84 1.66E-05 0.001025724 109 Il10 interleukin 10 1.19 8.19 11.29 2.51E-06 0.000352016 110 Ticam2 toll-like receptor adaptor molecule 2 1.19 8.59 14.17 4.14E-07 0.000126365 111 NA NA 1.18 8.92 7.95 3.69E-05 0.001646747 112 Sgk3 serum/glucocorticoid regulated kinase 3 1.18 9.77 14.40 3.65E-07 0.000126365 113 Tcstv1 2-cell-stage, variable group, member 1 1.18 8.27 4.56 0.001684377 0.019177893 114 Rilpl2 Rab interacting lysosomal protein-like 2 1.18 10.26 4.87 0.001106368 0.014432138 115 Dnase1l3 deoxyribonuclease 1-like 3 1.17 9.18 7.37 6.46E-05 0.002318592 116 Suclg2 succinate-Coenzyme A ligase, GDP-forming, beta subunit 1.17 10.23 3.86 0.004475495 0.036934103 117 NA NA 1.17 9.63 4.38 0.002134617 0.022487262 118 Ivns1abp influenza virus NS1A binding protein 1.16 10.99 5.23 0.000701269 0.010805477 119 Rtn4ip1 reticulon 4 interacting protein 1 1.16 8.44 8.02 3.46E-05 0.001575493 120 Gpm6a glycoprotein m6a 1.16 8.19 13.33 6.77E-07 0.000167465 121 Tmem154 transmembrane protein 154 1.15 10.03 16.07 1.50E-07 8.74E-05 122 Cd59b CD59b antigen 1.15 8.70 5.23 0.0007025 0.010805477 123 NA NA 1.14 8.20 11.11 2.86E-06 0.000365287 124 Tbc1d5 TBC1 domain family, member 5 1.13 10.54 7.82 4.17E-05 0.00179651 125 Lrrc8a leucine rich repeat containing 8A 1.13 8.66 5.80 0.000350171 0.006880043 126 Optn optineurin 1.12 8.65 7.60 5.16E-05 0.002011249 127 Timd2 T cell immunoglobulin and mucin domain containing 2 1.12 8.33 8.74 1.82E-05 0.001056366 128 Cobll1 Cobl-like 1 1.11 9.58 7.97 3.61E-05 0.001626306 129 Gbp3 guanylate binding protein 3 1.10 11.00 7.46 5.93E-05 0.002191762 130 Ftsjd2 FtsJ methyltransferase domain containing 2 1.10 12.78 7.16 7.94E-05 0.002664622

154

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 131 Kctd12 potassium channel tetramerisation domain containing 12 1.10 9.91 7.79 4.30E-05 0.00181213 132 Chrnb1 cholinergic receptor, nicotinic, beta polypeptide 1 1.09 8.73 7.03 9.11E-05 0.002946967 (muscle) 133 Psmd14 proteasome (prosome, macropain) 26S subunit, non- 1.09 10.25 8.53 2.17E-05 0.001177646 ATPase, 14 134 Irak2 interleukin-1 receptor-associated kinase 2 1.09 10.08 12.03 1.52E-06 0.000258979 135 Fcho2 FCH domain only 2 1.08 11.25 9.05 1.40E-05 0.000906822 136 Tnfsf4 tumor necrosis factor (ligand) superfamily, member 4 1.07 8.38 5.71 0.000389426 0.007374903 137 Hsp90ab1 heat shock protein 90 alpha (cytosolic), class B member 1.07 11.52 6.39 0.000179522 0.004550005 1 138 Fut8 fucosyltransferase 8 1.06 8.78 7.94 3.74E-05 0.001653635 139 Ryk receptor-like tyrosine kinase 1.06 9.24 7.63 5.01E-05 0.001984949 140 Myadm myeloid-associated differentiation marker 1.05 10.34 6.12 0.000242092 0.005562299 141 Tgif1 TGFB-induced factor homeobox 1 1.05 9.59 7.81 4.20E-05 0.001800796 142 Cobll1 Cobl-like 1 1.05 9.11 10.81 3.53E-06 0.00039538 143 Il15 interleukin 15 1.05 8.97 3.89 0.004302602 0.036006133 144 Ints4 integrator complex subunit 4 1.04 11.45 13.91 4.82E-07 0.000130244 145 Camk2n1 calcium/calmodulin-dependent protein kinase II inhibitor 1 1.04 8.74 4.50 0.001832528 0.02028317 146 Fert2 fer (fms/fps related) protein kinase, testis specific 2 1.04 8.04 13.68 5.50E-07 0.000144109 147 Gbp2 guanylate binding protein 2 1.03 8.47 7.78 4.32E-05 0.00181213 148 Aifm2 apoptosis-inducing factor, mitochondrion-associated 2 1.03 8.85 8.25 2.80E-05 0.00138606 149 F9 coagulation factor IX 1.03 8.67 9.68 8.30E-06 0.000643485 150 Ppfibp2 PTPRF interacting protein, binding protein 2 (liprin beta 2) 1.03 8.92 7.87 3.98E-05 0.001737923 151 Marcks myristoylated alanine rich protein kinase C substrate 1.03 11.22 6.39 0.000179412 0.004550005 152 Sh3bgrl2 SH3 domain binding glutamic acid-rich protein like 2 1.03 8.48 4.36 0.002212342 0.023065033 153 Rab31 RAB31, member RAS oncogene family 1.02 10.80 4.78 0.001244962 0.015627613 154 Unc119 unc-119 homolog (C. elegans) 1.02 9.44 5.65 0.000422166 0.007726939 155 Cbfa2t3 core-binding factor, runt domain, alpha subunit 2, 1.01 9.74 6.72 0.000125996 0.003600003 translocated to, 3 (human) 156 Ly96 lymphocyte antigen 96 1.01 9.80 5.08 0.000853268 0.012383256

155

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 157 NA NA 1.01 10.56 8.60 2.06E-05 0.001143744 158 Irak2 interleukin-1 receptor-associated kinase 2 1.01 9.42 12.27 1.30E-06 0.000236954 159 Cyp51 cytochrome P450, family 51 1.01 9.19 5.56 0.000465437 0.008365629 160 Cyb5r2 cytochrome b5 reductase 2 1.00 8.00 4.16 0.002895417 0.027587193 161 Tmem38b transmembrane protein 38B -1.00 8.78 -8.49 2.25E-05 0.001195989 162 Sit1 suppression inducing transmembrane adaptor 1 -1.01 8.65 -8.52 2.20E-05 0.001183798 163 Fam117a family with sequence similarity 117, member A -1.02 10.61 -4.07 0.003303844 0.030200374 164 Abhd8 abhydrolase domain containing 8 -1.02 8.87 -9.44 1.01E-05 0.00074318 165 Rgcc regulator of cell cycle -1.02 9.66 -4.43 0.001991212 0.021459569 166 P2ry6 pyrimidinergic receptor P2Y, G-protein coupled, 6 -1.03 8.68 -6.88 0.000106007 0.003202773 167 Cyp27a1 cytochrome P450, family 27, subfamily a, polypeptide 1 -1.03 8.45 -6.08 0.000253691 0.005670363 168 Fcgrt Fc receptor, IgG, alpha chain transporter -1.03 9.14 -4.21 0.002699974 0.026333605 169 Rnf144a ring finger protein 144A -1.03 8.65 -6.75 0.000121862 0.003541023 170 Gm5483 predicted gene 5483 -1.04 8.51 -4.12 0.003099077 0.028885777 171 Gata3 GATA binding protein 3 -1.04 8.38 -12.33 1.26E-06 0.00023571 172 Ddit4 DNA-damage-inducible transcript 4 -1.04 10.01 -5.42 0.000556829 0.009225337 173 Mgst1 microsomal glutathione S-transferase 1 -1.05 8.54 -3.57 0.006872263 0.049887542 174 Nrp1 neuropilin 1 -1.05 8.57 -6.12 0.000243368 0.005575765 175 Frat2 frequently rearranged in advanced T cell lymphomas 2 -1.05 8.43 -11.89 1.67E-06 0.000270244 176 Brwd1 bromodomain and WD repeat domain containing 1 -1.05 9.29 -3.86 0.004465454 0.036920991 177 P2ry14 purinergic receptor P2Y, G-protein coupled, 14 -1.05 8.76 -9.39 1.05E-05 0.000754882 178 Gmfg glia maturation factor, gamma -1.06 11.65 -4.91 0.001051831 0.013994156 179 Akap12 A kinase (PRKA) anchor protein (gravin) 12 -1.06 8.18 -7.97 3.64E-05 0.001633462 180 Gpr83 G protein-coupled receptor 83 -1.06 8.47 -9.72 8.06E-06 0.000643485 181 Cd27 CD27 antigen -1.06 8.66 -12.95 8.50E-07 0.000191956 182 Apol7c apolipoprotein L 7c -1.06 8.64 -6.23 0.000214469 0.005176085 183 Socs3 suppressor of cytokine signaling 3 -1.06 10.41 -6.32 0.00019337 0.004772354 184 Serpinb6a serine (or cysteine) peptidase inhibitor, clade B, member -1.07 9.46 -8.26 2.78E-05 0.00138606 6a

156

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 185 Hpgd hydroxyprostaglandin dehydrogenase 15 (NAD) -1.07 8.13 -5.49 0.000510547 0.008795466 186 Xdh xanthine dehydrogenase -1.07 8.70 -9.00 1.45E-05 0.000923744 187 Xcl1 chemokine (C motif) ligand 1 -1.07 8.29 -13.93 4.75E-07 0.000130244 188 Zap70 zeta-chain (TCR) associated protein kinase -1.08 8.76 -7.71 4.62E-05 0.001888593 189 Lrig1 leucine-rich repeats and immunoglobulin-like domains 1 -1.08 8.39 -10.76 3.66E-06 0.00039567 190 Klhl6 kelch-like 6 -1.08 12.72 -4.80 0.001218916 0.015444675 191 Pilrb1 paired immunoglobin-like type 2 receptor beta 1 -1.08 8.38 -9.71 8.15E-06 0.000643485 192 Klra7 killer cell lectin-like receptor, subfamily A, member 7 -1.09 8.24 -6.95 9.86E-05 0.003077668 193 Stat4 signal transducer and activator of transcription 4 -1.09 8.58 -7.78 4.35E-05 0.001817277 194 Nsg2 neuron specific gene family member 2 -1.09 8.21 -5.72 0.000388558 0.007374903 195 Sepp1 selenoprotein P, plasma, 1 -1.09 9.20 -7.46 5.92E-05 0.002191762 196 Prkcq protein kinase C, theta -1.09 8.51 -9.26 1.17E-05 0.000820031 197 Gstk1 glutathione S-transferase kappa 1 -1.09 8.52 -10.55 4.26E-06 0.000429447 198 Arrdc4 arrestin domain containing 4 -1.09 8.62 -9.97 6.64E-06 0.000572834 199 Stard10 START domain containing 10 -1.10 9.46 -5.06 0.000867359 0.012512813 200 Fam189b family with sequence similarity 189, member B -1.10 8.64 -5.67 0.000407926 0.007601808 201 Acss2 acyl-CoA synthetase short-chain family member 2 -1.10 9.26 -4.55 0.001709577 0.019280562 202 Tnfaip8l2 tumor necrosis factor, alpha-induced protein 8-like 2 -1.10 9.45 -14.09 4.34E-07 0.000126365 203 Rab3d RAB3D, member RAS oncogene family -1.10 9.20 -6.95 9.83E-05 0.003077254 204 Lpl lipoprotein lipase -1.10 8.47 -7.03 9.09E-05 0.002946967 205 Smad1 SMAD family member 1 -1.10 8.66 -6.20 0.000221278 0.005282699 206 Cdk5r1 cyclin-dependent kinase 5, regulatory subunit 1 (p35) -1.11 9.32 -4.76 0.001283768 0.015886479 207 Apobr apolipoprotein B receptor -1.11 8.51 -6.82 0.00011301 0.003305794 208 Nrn1 neuritin 1 -1.11 8.48 -5.92 0.000306885 0.006333433 209 Cd14 CD14 antigen -1.12 8.42 -7.59 5.21E-05 0.002014325 210 Clec4b1 C-type lectin domain family 4, member b1 -1.12 8.45 -7.25 7.28E-05 0.00250334 211 Smad1 SMAD family member 1 -1.13 8.63 -8.82 1.70E-05 0.001044514 212 Slc11a1 solute carrier family 11 (proton-coupled divalent metal ion -1.13 8.94 -5.50 0.000504008 0.008759984 transporters), member 1

157

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 213 Prkd3 protein kinase D3 -1.13 8.79 -5.06 0.000872332 0.012548754 214 Krt10 keratin 10 -1.13 8.55 -14.30 3.85E-07 0.000126365 215 Vopp1 vesicular, overexpressed in cancer, prosurvival protein 1 -1.13 9.30 -5.65 0.000419539 0.007700368 216 Nkg7 natural killer cell group 7 sequence -1.14 11.33 -4.16 0.002918159 0.027732032 217 Tcf7 transcription factor 7, T cell specific -1.14 8.18 -5.88 0.000321124 0.00652455 218 Klk1 kallikrein 1 -1.14 8.12 -4.59 0.001608869 0.018579935 219 Hcst hematopoietic cell signal transducer -1.15 10.76 -5.30 0.00064044 0.010185633 220 Lck lymphocyte protein tyrosine kinase -1.15 11.14 -5.56 0.000467913 0.008365629 221 LOC547323 uncharacterized LOC547323 -1.15 8.57 -12.71 9.89E-07 0.000212413 222 NA NA -1.16 8.67 -10.00 6.49E-06 0.000571 223 Tmem66 transmembrane protein 66 -1.16 9.48 -7.35 6.56E-05 0.002342512 224 Pygl liver glycogen phosphorylase -1.17 9.54 -3.58 0.006723284 0.049085591 225 Cst3 cystatin C -1.17 12.82 -6.51 0.000156878 0.004170149 226 Kmo kynurenine 3-monooxygenase (kynurenine 3- -1.17 8.51 -7.15 8.07E-05 0.002684506 hydroxylase) 227 Ear10 eosinophil-associated, ribonuclease A family, member 10 -1.18 8.48 -7.67 4.80E-05 0.001904754 228 Hist1h1c histone cluster 1, H1c -1.18 10.21 -10.20 5.56E-06 0.000513086 229 Acpl2 acid phosphatase-like 2 -1.18 8.76 -7.89 3.91E-05 0.001713178 230 Pi16 peptidase inhibitor 16 -1.19 8.55 -6.92 0.000102002 0.003130533 231 NA NA -1.20 8.87 -15.10 2.49E-07 0.000112734 232 Prkd3 protein kinase D3 -1.21 9.48 -4.03 0.003499375 0.031290611 233 Nrp1 neuropilin 1 -1.21 8.85 -5.06 0.000867924 0.012512813 234 Anxa1 annexin A1 -1.21 8.46 -4.22 0.002662353 0.026134882 235 Il27ra interleukin 27 receptor, alpha -1.22 11.04 -9.74 7.94E-06 0.000643485 236 Selplg selectin, platelet (p-selectin) ligand -1.22 10.59 -7.42 6.13E-05 0.002226548 237 Lpar6 lysophosphatidic acid receptor 6 -1.22 10.65 -12.22 1.35E-06 0.000242138 238 Ppic peptidylprolyl isomerase C -1.23 8.33 -4.05 0.003389681 0.030741707 239 Sh2d2a SH2 domain protein 2A -1.23 10.75 -5.55 0.000476 0.008464011 240 Ear12 eosinophil-associated, ribonuclease A family, member 12 -1.23 8.60 -10.75 3.69E-06 0.00039567

158

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 241 Hvcn1 hydrogen voltage-gated channel 1 -1.24 12.02 -5.22 0.000712795 0.010897018 242 Sun2 Sad1 and UNC84 domain containing 2 -1.25 11.74 -10.93 3.24E-06 0.000384201 243 Fyb FYN binding protein -1.25 9.43 -9.69 8.24E-06 0.000643485 244 Fxyd5 FXYD domain-containing ion transport regulator 5 -1.25 10.04 -7.25 7.25E-05 0.00250152 245 Lrrk2 leucine-rich repeat kinase 2 -1.26 9.33 -3.86 0.004483801 0.036956105 246 Neurl3 neuralized homolog 3 homolog (Drosophila) -1.27 8.57 -9.06 1.38E-05 0.000901477 247 NA NA -1.28 9.30 -14.09 4.34E-07 0.000126365 248 Sit1 suppression inducing transmembrane adaptor 1 -1.28 8.95 -10.25 5.34E-06 0.0005051 249 Foxp3 forkhead box P3 -1.28 8.88 -11.28 2.52E-06 0.000352016 250 Klf13 Kruppel-like factor 13 -1.29 10.80 -7.22 7.52E-05 0.002573373 251 Cd160 CD160 antigen -1.29 8.72 -8.53 2.17E-05 0.001177646 252 Fas Fas (TNF receptor superfamily member 6) -1.29 8.97 -9.43 1.02E-05 0.00074318 253 Ifngr1 interferon gamma receptor 1 -1.29 9.89 -7.94 3.72E-05 0.001650647 254 Fam78a family with sequence similarity 78, member A -1.30 11.06 -6.07 0.000258382 0.005742531 255 Osm oncostatin M -1.30 9.04 -17.16 8.87E-08 7.20E-05 256 Asb2 ankyrin repeat and SOCS box-containing 2 -1.30 9.77 -7.78 4.33E-05 0.00181213 257 Tyrobp TYRO protein tyrosine kinase binding protein -1.31 9.68 -8.30 2.69E-05 0.001355341 258 Casp1 caspase 1 -1.32 11.79 -5.17 0.000758776 0.01139033 259 Bcl7a B cell CLL/lymphoma 7A -1.32 9.92 -9.88 7.12E-06 0.000602323 260 Cd247 CD247 antigen -1.32 9.05 -12.17 1.39E-06 0.0002457 261 Zfp36 zinc finger protein 36 -1.33 11.52 -10.17 5.67E-06 0.000519334 262 Myo1f myosin IF -1.33 9.89 -6.64 0.000137512 0.003809939 263 Gpr68 G protein-coupled receptor 68 -1.34 8.89 -12.97 8.39E-07 0.000191956 264 Ephx1 epoxide hydrolase 1, microsomal -1.34 10.81 -8.48 2.28E-05 0.001199946 265 Tacstd2 tumor-associated calcium signal transducer 2 -1.35 8.54 -15.12 2.46E-07 0.000112734 266 Itgad integrin, alpha D -1.35 8.45 -6.40 0.000178721 0.004550005 267 Tcf7 transcription factor 7, T cell specific -1.36 8.57 -6.99 9.46E-05 0.003022162 268 Rgs10 regulator of G-protein signalling 10 -1.36 9.95 -13.03 8.11E-07 0.000189903

159

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 269 Sep9 septin 9 -1.37 8.95 -8.86 1.63E-05 0.001013828 270 Cd8b1 CD8 antigen, beta chain 1 -1.37 9.16 -3.66 0.005982465 0.045057589 271 Sema4a sema domain, immunoglobulin domain (Ig), -1.37 10.37 -9.41 1.04E-05 0.000750419 transmembrane domain (TM) and short cytoplasmic domain, (semaphorin) 4A 272 Csf3r colony stimulating factor 3 receptor (granulocyte) -1.37 8.57 -10.39 4.81E-06 0.00047416 273 F13a1 coagulation factor XIII, A1 subunit -1.37 8.55 -6.71 0.000126752 0.003600003 274 Tmem51 transmembrane protein 51 -1.38 9.54 -5.82 0.000341886 0.006791276 275 Gpr171 G protein-coupled receptor 171 -1.38 9.82 -17.78 6.65E-08 7.20E-05 276 Glipr2 GLI pathogenesis-related 2 -1.38 10.42 -8.98 1.48E-05 0.000927095 277 Csrp2 cysteine and glycine-rich protein 2 -1.39 8.74 -14.65 3.17E-07 0.000126365 278 Dusp2 dual specificity phosphatase 2 -1.39 9.31 -9.72 8.02E-06 0.000643485 279 Sostdc1 sclerostin domain containing 1 -1.41 8.64 -4.41 0.002053933 0.021870864 280 Ccr6 chemokine (C-C motif) receptor 6 -1.42 9.02 -6.59 0.000144894 0.003946706 281 Igfbp4 insulin-like growth factor binding protein 4 -1.42 8.50 -5.22 0.000711049 0.010897018 282 Sort1 sortilin 1 -1.42 9.14 -9.58 8.99E-06 0.000680709 283 Fpr2 formyl peptide receptor 2 -1.43 11.21 -4.16 0.002929021 0.027754753 284 Fxyd5 FXYD domain-containing ion transport regulator 5 -1.44 11.88 -7.42 6.12E-05 0.002226548 285 Klra4 killer cell lectin-like receptor, subfamily A, member 4 -1.45 8.77 -14.12 4.26E-07 0.000126365 286 Cmc1 COX assembly mitochondrial protein 1 -1.45 10.24 -8.41 2.43E-05 0.001241645 287 Myl4 myosin, light polypeptide 4 -1.45 10.47 -6.37 0.000183946 0.004612847 288 Dennd3 DENN/MADD domain containing 3 -1.46 10.21 -3.97 0.003801274 0.033188335 289 Hsd11b1 hydroxysteroid 11-beta dehydrogenase 1 -1.46 9.49 -14.72 3.06E-07 0.000126365 290 Csf3r colony stimulating factor 3 receptor (granulocyte) -1.47 8.43 -17.60 7.20E-08 7.20E-05 291 F2r coagulation factor II (thrombin) receptor -1.48 10.49 -7.14 8.10E-05 0.002688196 292 Fxyd5 FXYD domain-containing ion transport regulator 5 -1.48 10.46 -9.09 1.34E-05 0.000886427 293 Klre1 killer cell lectin-like receptor family E member 1 -1.48 8.64 -9.52 9.47E-06 0.000711445 294 Trib2 tribbles homolog 2 (Drosophila) -1.50 9.82 -8.80 1.72E-05 0.001046831 295 Neurl3 neuralized homolog 3 homolog (Drosophila) -1.50 8.63 -8.74 1.81E-05 0.001056366

160

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 296 Ctnna1 catenin (cadherin associated protein), alpha 1 -1.51 9.75 -15.87 1.67E-07 8.74E-05 297 Bcl11b B cell leukemia/lymphoma 11B -1.52 10.28 -5.81 0.000345715 0.006826447 298 Tmem108 transmembrane protein 108 -1.52 8.48 -10.23 5.44E-06 0.000509109 299 Cyp27a1 cytochrome P450, family 27, subfamily a, polypeptide 1 -1.52 9.00 -10.72 3.77E-06 0.00039567 300 Csf1r colony stimulating factor 1 receptor -1.52 9.68 -4.30 0.002397836 0.024405413 301 Gpc1 glypican 1 -1.52 9.28 -10.92 3.25E-06 0.000384201 302 Fcer2a Fc receptor, IgE, low affinity II, alpha polypeptide -1.55 8.61 -7.02 9.20E-05 0.002950886 303 Gcnt2 glucosaminyl (N-acetyl) transferase 2, I-branching -1.55 9.28 -9.91 6.92E-06 0.000592709 enzyme 304 Zap70 zeta-chain (TCR) associated protein kinase -1.56 10.13 -12.46 1.15E-06 0.000226433 305 Chst1 carbohydrate (keratan sulfate Gal-6) sulfotransferase 1 -1.56 9.18 -4.94 0.001016143 0.013686077 306 Hvcn1 hydrogen voltage-gated channel 1 -1.56 11.55 -9.26 1.16E-05 0.000820031 307 Cdc42ep3 CDC42 effector protein (Rho GTPase binding) 3 -1.57 9.86 -10.46 4.57E-06 0.000453629 308 Cxcr3 chemokine (C-X-C motif) receptor 3 -1.59 9.84 -8.07 3.31E-05 0.001539334 309 Leprotl1 leptin receptor overlapping transcript-like 1 -1.61 10.19 -12.28 1.30E-06 0.000236954 310 Mmp9 matrix metallopeptidase 9 -1.61 9.03 -4.50 0.001814258 0.020131968 311 Gpr114 G protein-coupled receptor 114 -1.61 9.43 -7.29 6.97E-05 0.002424401 312 Ffar2 free fatty acid receptor 2 -1.61 9.67 -4.83 0.001168508 0.015027774 313 Pacsin1 protein kinase C and casein kinase substrate in neurons -1.62 9.02 -7.53 5.52E-05 0.002097544 1 314 Lyz1 lysozyme 1 -1.62 14.55 -9.75 7.87E-06 0.000643485 315 Clec7a C-type lectin domain family 7, member a -1.62 8.95 -14.82 2.89E-07 0.000126365 316 Clec4d C-type lectin domain family 4, member d -1.64 8.80 -7.18 7.80E-05 0.002639989 317 Ngfrap1 nerve growth factor receptor (TNFRSF16) associated -1.64 9.86 -8.08 3.27E-05 0.001539334 protein 1 318 Rab32 RAB32, member RAS oncogene family -1.64 9.66 -10.97 3.15E-06 0.000382174 319 Mt1 metallothionein 1 -1.65 10.74 -4.34 0.002276487 0.023602341 320 Cyp27a1 cytochrome P450, family 27, subfamily a, polypeptide 1 -1.65 8.74 -9.01 1.44E-05 0.00092333 321 Sgk1 serum/glucocorticoid regulated kinase 1 -1.65 10.28 -7.75 4.46E-05 0.001845096 322 Lgmn legumain -1.65 10.49 -7.38 6.39E-05 0.002306549

161

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 323 Itk IL2 inducible T cell kinase -1.66 9.64 -7.55 5.41E-05 0.002069558 324 Tmem66 transmembrane protein 66 -1.66 11.10 -6.92 0.000101909 0.003130533 325 Sort1 sortilin 1 -1.66 9.30 -10.09 6.06E-06 0.000540025 326 Tmem66 transmembrane protein 66 -1.67 11.01 -4.80 0.001212497 0.015397073 327 Arap3 ArfGAP with RhoGAP domain, ankyrin repeat and PH -1.67 9.00 -8.54 2.17E-05 0.001177646 domain 3 328 Ramp1 receptor (calcitonin) activity modifying protein 1 -1.69 8.75 -7.93 3.77E-05 0.001662267 329 St6galnac2 ST6 (alpha-N-acetyl-neuraminyl-2,3-beta-galactosyl-1,3)- -1.69 9.92 -13.38 6.54E-07 0.000164898 N-acetylgalactosaminide alpha-2,6-sialyltransferase 2 330 Pglyrp1 peptidoglycan recognition protein 1 -1.69 10.42 -4.91 0.001058114 0.014063476 331 Serpinb1a serine (or cysteine) peptidase inhibitor, clade B, member -1.70 9.57 -9.08 1.35E-05 0.000886427 1a 332 Ifitm3 interferon induced transmembrane protein 3 -1.72 11.17 -6.49 0.000160495 0.004214998 333 Hp haptoglobin -1.72 9.14 -4.28 0.002457471 0.024849662 334 Lmo2 LIM domain only 2 -1.72 11.85 -14.32 3.80E-07 0.000126365 335 Cd3e CD3 antigen, epsilon polypeptide -1.74 10.70 -10.78 3.61E-06 0.00039567 336 Tgfbi transforming growth factor, beta induced -1.74 9.71 -11.93 1.62E-06 0.000270244 337 Xlr4a X-linked lymphocyte-regulated 4A -1.74 11.46 -12.47 1.15E-06 0.000226433 338 Dok2 docking protein 2 -1.79 9.77 -7.16 7.95E-05 0.002664622 339 Zfp608 zinc finger protein 608 -1.79 8.70 -9.44 1.01E-05 0.00074318 340 Cd6 CD6 antigen -1.80 9.99 -13.13 7.62E-07 0.00018156 341 NA NA -1.80 9.99 -8.21 2.90E-05 0.001422116 342 Tnfrsf4 tumor necrosis factor receptor superfamily, member 4 -1.81 10.02 -7.35 6.59E-05 0.002347302 343 Zyx zyxin -1.81 12.54 -12.12 1.44E-06 0.000248712 344 Hsd11b1 hydroxysteroid 11-beta dehydrogenase 1 -1.82 9.90 -18.01 5.98E-08 7.20E-05 345 Dusp2 dual specificity phosphatase 2 -1.82 10.38 -11.63 2.00E-06 0.000307633 346 Dgka diacylglycerol kinase, alpha -1.82 10.89 -7.18 7.82E-05 0.002642851 347 Lat linker for activation of T cells -1.83 10.61 -16.23 1.39E-07 8.74E-05 348 Tiam1 T cell lymphoma invasion and metastasis 1 -1.83 9.29 -10.33 5.03E-06 0.000491624 349 Trf transferrin -1.85 9.00 -6.84 0.000111008 0.003276496

162

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 350 NA NA -1.87 8.91 -8.06 3.35E-05 0.001544265 351 Gadd45g growth arrest and DNA-damage-inducible 45 gamma -1.90 10.08 -9.22 1.21E-05 0.000832985 352 Bcl6 B cell leukemia/lymphoma 6 -1.91 10.44 -8.08 3.29E-05 0.001539334 353 Dgka diacylglycerol kinase, alpha -1.91 11.54 -6.59 0.000144058 0.003943894 354 Il18r1 interleukin 18 receptor 1 -1.93 9.89 -17.37 8.04E-08 7.20E-05 355 C1qc complement component 1, q subcomponent, C chain -1.93 9.22 -11.44 2.27E-06 0.000333942 356 Fcgr3 Fc receptor, IgG, low affinity III -1.93 9.07 -5.93 0.000302045 0.006253232 357 Cd6 CD6 antigen -1.94 10.43 -11.40 2.32E-06 0.000338092 358 Hsd11b1 hydroxysteroid 11-beta dehydrogenase 1 -1.96 10.48 -23.34 7.23E-09 1.90E-05 359 Il7r interleukin 7 receptor -1.96 10.42 -9.99 6.53E-06 0.000571 360 Cd3g CD3 antigen, gamma polypeptide -1.98 11.13 -11.02 3.05E-06 0.000376611 361 Pxdc1 PX domain containing 1 -1.98 9.32 -9.14 1.29E-05 0.000864652 362 Tmem176b transmembrane protein 176B -2.00 9.68 -12.43 1.17E-06 0.000226433 363 Cd8b1 CD8 antigen, beta chain 1 -2.02 12.18 -5.97 0.000289485 0.006114237 364 Igfbp4 insulin-like growth factor binding protein 4 -2.04 8.88 -7.02 9.21E-05 0.002950886 365 Cd6 CD6 antigen -2.05 10.96 -9.43 1.02E-05 0.00074318 366 Emb embigin -2.06 9.96 -15.10 2.49E-07 0.000112734 367 Ear4 eosinophil-associated, ribonuclease A family, member 4 -2.07 9.25 -8.97 1.48E-05 0.000927095 368 Fcer2a Fc receptor, IgE, low affinity II, alpha polypeptide -2.07 9.04 -9.09 1.35E-05 0.000886427 369 Cd27 CD27 antigen -2.08 11.11 -17.27 8.41E-08 7.20E-05 370 Klk8 kallikrein related-peptidase 8 -2.09 9.80 -8.60 2.05E-05 0.001143744 371 Ifitm6 interferon induced transmembrane protein 6 -2.10 9.77 -5.07 0.000859724 0.012443619 372 C1qb complement component 1, q subcomponent, beta -2.10 9.58 -11.12 2.82E-06 0.000365287 polypeptide 373 Lyz2 lysozyme 2 -2.12 12.10 -15.93 1.62E-07 8.74E-05 374 Ly6c1 lymphocyte antigen 6 complex, locus C1 -2.14 10.49 -6.17 0.000228945 0.005376928 375 Lgals3 lectin, galactose binding, soluble 3 -2.15 11.95 -11.17 2.73E-06 0.000364587 376 Stat4 signal transducer and activator of transcription 4 -2.17 10.16 -10.82 3.52E-06 0.00039538

163

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 377 B3gnt8 UDP-GlcNAc:betaGal beta-1,3-N- -2.17 10.18 -14.13 4.23E-07 0.000126365 acetylglucosaminyltransferase 8 378 Cd3d CD3 antigen, delta polypeptide -2.18 10.96 -13.89 4.87E-07 0.000130244 379 Sirpb1a signal-regulatory protein beta 1A -2.19 9.88 -16.08 1.50E-07 8.74E-05 380 Ccr6 chemokine (C-C motif) receptor 6 -2.19 10.34 -5.75 0.000374546 0.007176066 381 Il4i1 interleukin 4 induced 1 -2.22 11.72 -7.69 4.75E-05 0.00190209 382 Thy1 thymus cell antigen 1, theta -2.24 10.76 -10.21 5.49E-06 0.000510469 383 Ifitm2 interferon induced transmembrane protein 2 -2.25 10.80 -10.10 6.01E-06 0.000539159 384 Ccl9 chemokine (C-C motif) ligand 9 -2.27 9.39 -13.17 7.42E-07 0.000180048 385 Dpp4 dipeptidylpeptidase 4 -2.28 10.18 -11.69 1.91E-06 0.000303211 386 Ctsw cathepsin W -2.28 10.26 -14.53 3.40E-07 0.000126365 387 Klrd1 killer cell lectin-like receptor, subfamily D, member 1 -2.29 9.84 -10.70 3.83E-06 0.00039567 388 Il1b interleukin 1 beta -2.31 9.62 -12.15 1.41E-06 0.000246908 389 Axl AXL receptor tyrosine kinase -2.33 9.65 -9.10 1.33E-05 0.000885286 390 Lyz2 lysozyme 2 -2.33 13.02 -10.27 5.28E-06 0.0005051 391 Hp haptoglobin -2.34 9.83 -5.82 0.000344727 0.006824249 392 S100a9 S100 calcium binding protein A9 (calgranulin B) -2.35 12.85 -4.01 0.003592089 0.031897675 393 Ear2 eosinophil-associated, ribonuclease A family, member 2 -2.44 9.87 -10.97 3.14E-06 0.000382174 394 Prg2 proteoglycan 2, bone marrow -2.50 9.43 -26.42 2.61E-09 1.14E-05 395 Vcam1 vascular cell adhesion molecule 1 -2.53 9.44 -10.93 3.24E-06 0.000384201 396 Alox5ap arachidonate 5-lipoxygenase activating protein -2.53 10.63 -12.50 1.12E-06 0.000226433 397 Chchd10 coiled-coil-helix-coiled-coil-helix domain containing 10 -2.60 11.67 -23.74 6.28E-09 1.90E-05 398 Chi3l3 chitinase 3-like 3 -2.63 10.46 -3.92 0.004099802 0.034940263 399 Fcna ficolin A -2.74 9.51 -14.52 3.40E-07 0.000126365 400 Fcer2a Fc receptor, IgE, low affinity II, alpha polypeptide -2.90 9.99 -9.64 8.61E-06 0.000659489 401 Satb1 special AT-rich sequence binding protein 1 -2.92 11.26 -9.26 1.17E-05 0.000820031 402 Slc40a1 solute carrier family 40 (iron-regulated transporter), -2.98 10.90 -16.03 1.54E-07 8.74E-05 member 1 403 Vpreb3 pre-B lymphocyte gene 3 -3.59 10.07 -26.45 2.58E-09 1.14E-05

164

# GeneSymbol Gene Name LogFC AveExpr t P. Value Adj. P. Val. 404 Slpi secretory leukocyte peptidase inhibitor -3.67 11.58 -6.89 0.000105045 0.003186622

165

Table 2. The MCC-interactome in human MM cells identified by affinity purification followed by LC-MS/MS

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor NA P23508 Colorectal mutant cancer protein CRCM_HUMAN 415 1223 Not applicable 1 P23508-2 Isoform 2 of Colorectal mutant cancer protein CRCM_HUMAN 395.5 1172.5 No 2 P09874 Poly [ADP-ribose] polymerase 1 PARP1_HUMAN 72 43.5 No 3 O15020-2 Isoform 2 of Spectrin beta chain, non-erythrocytic 2 SPTN2_HUMAN 34 3 No 4 P40939 Trifunctional enzyme subunit alpha, mitochondrial ECHA_HUMAN 31 24 No 5 P05141 ADP/ATP translocase 2 ADT2_HUMAN 24 33.5 No 6 P12236 ADP/ATP translocase 3 ADT3_HUMAN 20.5 27.5 No 7 P06576 ATP synthase subunit beta, mitochondrial ATPB_HUMAN 18.5 19.5 No 8 Q99623 Prohibitin-2 PHB2_HUMAN 17 23.5 Yes (Ewing, 2007) 9 Q13576 Ras GTPase-activating-like protein IQGAP2 IQGA2_HUMAN 16 50 No 10 P12235 ADP/ATP translocase 1 ADT1_HUMAN 15.5 24 No 11 Q9Y277 Voltage-dependent anion-selective channel protein 3 VDAC3_HUMAN 15.5 2.5 No 12 Q02880-2 Isoform Beta-1 of DNA topoisomerase 2-beta TOP2B_HUMAN 14.5 12.5 No O60264 SWI/SNF-related matrix-associated actin-dependent regulator of 13.5 2 No 13 chromatin subfamily A member 5 SMCA5_HUMAN 14 P35232 Prohibitin PHB_HUMAN 13 3.5 No 15 O00571 ATP-dependent RNA helicase DDX3X DDX3X_HUMAN 12 17.5 No 16 P13674 Prolyl 4-hydroxylase subunit alpha-1 P4HA1_HUMAN 12 9 No 17 Q9Y2X3 Nucleolar protein 58 NOP58_HUMAN 11.5 10.5 No 18 P13674-2 Isoform 2 of Prolyl 4-hydroxylase subunit alpha-1 P4HA1_HUMAN 11 8.5 No P16403 Histone H1.2 9.5 8.5 Yes (Sigglekow, 19 H12_HUMAN 2012) 20 P13796 Plastin-2 PLSL_HUMAN 9.5 7.5 No 21 P42166 Lamina-associated polypeptide 2, isoform alpha LAP2A_HUMAN 9 19.5 No 22 P46940 Ras GTPase-activating-like protein IQGAP1 IQGA1_HUMAN 8 31 No 23 P47756-2 Isoform 2 of F-actin-capping protein subunit beta CAPZB_HUMAN 7.5 6.5 Yes (Ewing, 2007) 24 P06493 Cyclin-dependent kinase 1 CDK1_HUMAN 7 8 No 25 P12956 X-ray repair cross-complementing protein 6 XRCC6_HUMAN 7 6.5 No 26 P35251-2 Isoform 2 of Replication factor C subunit 1 RFC1_HUMAN 7 3.5 No

166

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 27 Q14974 Importin subunit beta-1 IMB1_HUMAN 6.5 18 No 28 P78371 T-complex protein 1 subunit beta TCPB_HUMAN 6.5 12 No 29 Q9Y5B9 FACT complex subunit SPT16 SP16H_HUMAN 6.5 6 Yes (Ewing, 2007) 30 Q86UE4 Protein LYRIC LYRIC_HUMAN 6.5 3.5 No 31 Q9BVP2-2 Isoform 2 of Guanine nucleotide-binding protein-like 3 GNL3_HUMAN 6.5 1 No 32 Q5T4S7-3 Isoform 3 of E3 ubiquitin-protein ligase UBR4 UBR4_HUMAN 6 69 No 33 Q96ER9 Coiled-coil domain-containing protein 51 CCD51_HUMAN 6 17 No 34 P31689 DnaJ homolog subfamily A member 1 DNJA1_HUMAN 6 14.5 No P63092 Guanine nucleotide-binding protein G(s) subunit alpha isoforms 6 5.5 No 35 short GNAS2_HUMAN 36 P53621 Coatomer subunit alpha COPA_HUMAN 6 5 No 37 P22732 Solute carrier family 2, facilitated glucose transporter member 5 GTR5_HUMAN 6 4 No P63092-2 Isoform Gnas-2 of Guanine nucleotide-binding protein G(s) subunit 5.5 5.5 No 38 alpha isoforms short GNAS2_HUMAN P63092-3 Isoform 3 of Guanine nucleotide-binding protein G(s) subunit alpha 5.5 5 No 39 isoforms short GNAS2_HUMAN 40 Q14839 Chromodomain-helicase-DNA-binding protein 4 CHD4_HUMAN 5.5 4.5 No 41 Q8NI36 WD repeat-containing protein 36 WDR36_HUMAN 5.5 4 No 42 P17858 6-phosphofructokinase, liver type K6PL_HUMAN 5 6.5 No 43 P30041 Peroxiredoxin-6 PRDX6_HUMAN 5 6 No Q9P035 Very-long-chain (3R)-3-hydroxyacyl-[acyl-carrier protein] 5 6 No 44 dehydratase 3 HACD3_HUMAN 45 P18754 Regulator of chromosome condensation RCC1_HUMAN 5 4.5 No 46 Q9BSD7 Cancer-related nucleoside-triphosphatase NTPCR_HUMAN 5 4 No 47 P62873 Guanine nucleotide-binding protein G(I)/G(S)/G(T) subunit beta-1 GBB1_HUMAN 5 3.5 Yes (Ewing, 2007) 48 O94874 E3 UFM1-protein ligase 1 UFL1_HUMAN 5 3 No O95299 NADH dehydrogenase [ubiquinone] 1 alpha subcomplex subunit 10, 5 3 No 49 mitochondrial NDUAA_HUMAN 50 Q9UQE7 Structural maintenance of chromosomes protein 3 SMC3_HUMAN 4.5 24 Yes (Ewing, 2007) 51 P62136 Serine/threonine-protein phosphatase PP1-alpha catalytic subunit PP1A_HUMAN 4.5 8.5 No 52 Q14157 Ubiquitin-associated protein 2-like UBP2L_HUMAN 4.5 7 No 53 O75534-2 Isoform Short of Cold shock domain-containing protein E1 CSDE1_HUMAN 4.5 5 Yes (Ewing, 2007)

167

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 54 Q00013 55 kDa erythrocyte membrane protein EM55_HUMAN 4.5 2.5 No 55 P17480-2 Isoform UBF2 of Nucleolar transcription factor 1 UBF1_HUMAN 4.5 2.5 No 56 Q01813 6-phosphofructokinase type C K6PP_HUMAN 4.5 2 Yes (Ewing, 2007) 57 Q9UIG0-2 Isoform 2 of Tyrosine-protein kinase BAZ1B BAZ1B_HUMAN 4.5 2 No 58 P33991 DNA replication licensing factor MCM4 MCM4_HUMAN 4 27 Yes (Ewing, 2007) 59 Q92973-2 Isoform 2 of Transportin-1 TNPO1_HUMAN 4 15.5 No 60 Q08945 FACT complex subunit SSRP1 SSRP1_HUMAN 4 9.5 Yes (Ewing, 2007) 61 Q13242 Serine/arginine-rich splicing factor 9 SRSF9_HUMAN 4 5 No 62 P23528 Cofilin-1 COF1_HUMAN 4 4 No 63 Q92922 SWI/SNF complex subunit SMARCC1 SMRC1_HUMAN 4 3.5 No 64 Q14699 Raftlin RFTN1_HUMAN 4 3 No 65 O94826 Mitochondrial import receptor subunit TOM70 TOM70_HUMAN 4 2.5 No 66 P51531-2 Isoform Short of Probable global transcription activator SNF2L2 SMCA2_HUMAN 4 2.5 No 67 Q9UQ80 Proliferation-associated protein 2G4 PA2G4_HUMAN 4 2 No 68 O43396 Thioredoxin-like protein 1 TXNL1_HUMAN 3.5 15.5 No 69 O60884 DnaJ homolog subfamily A member 2 DNJA2_HUMAN 3.5 10 No 70 P36873 Serine/threonine-protein phosphatase PP1-gamma catalytic subunit PP1G_HUMAN 3.5 8.5 No 71 P53985 Monocarboxylate transporter 1 MOT1_HUMAN 3.5 8.5 No 72 P40938 Replication factor C subunit 3 RFC3_HUMAN 3.5 4.5 No 73 Q53HL2 Borealin BOREA_HUMAN 3.5 4 No P21912 Succinate dehydrogenase [ubiquinone] iron-sulfur subunit, 3.5 3.5 Yes (Ewing, 2007) 74 mitochondrial DHSB_HUMAN 75 Q86WU2-2 Isoform 2 of Probable D-lactate dehydrogenase, mitochondrial LDHD_HUMAN 3.5 3.5 No 76 Q96JB5 CDK5 regulatory subunit-associated protein 3 CK5P3_HUMAN 3.5 3 No 77 O95563 Mitochondrial pyruvate carrier 2 MPC2_HUMAN 3.5 2.5 No 78 Q96QK1 Vacuolar protein sorting-associated protein 35 VPS35_HUMAN 3.5 2.5 Yes (Ewing, 2007) 79 Q9BZQ8 Protein Niban NIBAN_HUMAN 3.5 1.5 No 80 O14773-2 Isoform 2 of Tripeptidyl-peptidase 1 TPP1_HUMAN 3.5 1.5 No 81 Q92945 Far upstream element-binding protein 2 FUBP2_HUMAN 3 7.5 No 82 Q9P0L0 Vesicle-associated membrane protein-associated protein A VAPA_HUMAN 3 6 No 83 Q15181 Inorganic pyrophosphatase IPYR_HUMAN 3 5.5 No

168

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 84 Q12788 Transducin beta-like protein 3 TBL3_HUMAN 3 5 No 85 O95573 Long-chain-fatty-acid--CoA ligase 3 ACSL3_HUMAN 3 4.5 No 86 Q02978 Mitochondrial 2-oxoglutarate/malate carrier protein M2OM_HUMAN 3 3 No Q9UKM7 Endoplasmic reticulum mannosyl-oligosaccharide 1,2-alpha- 3 3 No 87 mannosidase MA1B1_HUMAN 88 Q7Z3B4 Nucleoporin p54 NUP54_HUMAN 3 2 No 89 P42224 Signal transducer and activator of transcription 1-alpha/beta STAT1_HUMAN 3 2 No 90 Q99733 Nucleosome assembly protein 1-like 4 NP1L4_HUMAN 3 2 No 91 P35998 26S protease regulatory subunit 7 PRS7_HUMAN 2.5 10 No 92 P12004 Proliferating cell nuclear antigen PCNA_HUMAN 2.5 8.5 Yes (Ewing, 2007) 93 O00299 Chloride intracellular channel protein 1 CLIC1_HUMAN 2.5 8.5 Yes (Ewing, 2007) 94 O75083 WD repeat-containing protein 1 WDR1_HUMAN 2.5 5.5 No 95 Q9Y3F4 Serine-threonine kinase receptor-associated protein STRAP_HUMAN 2.5 5 No 96 P43358 Melanoma-associated antigen 4 MAGA4_HUMAN 2.5 5 No 97 Q08752 Peptidyl-prolyl cis-trans isomerase D PPID_HUMAN 2.5 5 No 98 O43175 D-3-phosphoglycerate dehydrogenase SERA_HUMAN 2.5 4.5 No 99 Q9GZS3 WD repeat-containing protein 61 WDR61_HUMAN 2.5 4 No 100 P53701 Cytochrome c-type heme lyase CCHL_HUMAN 2.5 4 No 101 Q14258 E3 ubiquitin/ISG15 ligase TRIM25 TRI25_HUMAN 2.5 3.5 No 102 P35249 Replication factor C subunit 4 RFC4_HUMAN 2.5 3 No 103 P51159 Ras-related protein Rab-27A RB27A_HUMAN 2.5 3 No 104 P07741 Adenine phosphoribosyltransferase] APT_HUMAN 2.5 2.5 Yes (Ewing, 2007) 105 Q96HE7 ERO1-like protein alpha ERO1A_HUMAN 2.5 2.5 No 106 Q8WYP5 Protein ELYS ELYS_HUMAN 2.5 2.5 No 107 Q9H0U3 Magnesium transporter protein 1 MAGT1_HUMAN 2.5 1.5 No 108 O15400-2 Isoform 2 of Syntaxin-7 STX7_HUMAN 2.5 1 No 109 Q9Y5X1 Sorting nexin-9 SNX9_HUMAN 2.5 1 No 110 Q969V3-2 Isoform 2 of Nicalin NCLN_HUMAN 2.5 1 No 111 Q9BQG0 Myb-binding protein 1A MBB1A_HUMAN 2.5 1 No 112 P38117 Electron transfer flavoprotein subunit beta ETFB_HUMAN 2 14 No 113 Q9BYG3 MKI67 FHA domain-interacting nucleolar phosphoprotein MK67I_HUMAN 2 9 No

169

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 114 Q9NX58 Cell growth-regulating nucleolar protein LYAR_HUMAN 2 9 No 115 Q9BPW8 Protein NipSnap homolog 1 NIPS1_HUMAN 2 8 No 116 Q8IYB3 Serine/arginine repetitive matrix protein 1 SRRM1_HUMAN 2 8 No 117 P69849 Nodal modulator 3 NOMO3_HUMAN 2 7.5 No 118 Q14203-5 Isoform 5 of Dynactin subunit 1 DCTN1_HUMAN 2 6.5 No 119 Q6DD88 Atlastin-3 ATLA3_HUMAN 2 6.5 No 120 P53004 Biliverdin reductase A BIEA_HUMAN 2 6.5 No 121 O60763 General vesicular transport factor p115 USO1_HUMAN 2 5 No 122 P13010 X-ray repair cross-complementing protein 5 XRCC5_HUMAN 2 5 No 123 Q16836 Hydroxyacyl-coenzyme A dehydrogenase, mitochondrial HCDH_HUMAN 2 4.5 Yes (Ewing, 2007) 124 Q09028-3 Isoform 3 of Histone-binding protein RBBP4 RBBP4_HUMAN 2 4 No 125 Q15126 Phosphomevalonate kinase PMVK_HUMAN 2 3.5 No 126 P18669 Phosphoglycerate mutase 1 PGAM1_HUMAN 2 3.5 No 127 P23258 Tubulin gamma-1 chain TBG1_HUMAN 2 3.5 No 128 Q9HC21 Mitochondrial thiamine pyrophosphate carrier TPC_HUMAN 2 3.5 No 129 P61160 Actin-related protein 2 ARP2_HUMAN 2 3 No 130 Q12769 Nuclear pore complex protein Nup160 NU160_HUMAN 2 2.5 No 131 Q96HY6 DDRGK domain-containing protein 1 DDRGK_HUMAN 2 2 No 132 P48960-2 Isoform 2 of CD97 antigen CD97_HUMAN 2 2 No 133 Q13610 Periodic tryptophan protein 1 homolog PWP1_HUMAN 2 1.5 No 134 Q9H7Z7 Prostaglandin E synthase 2 PGES2_HUMAN 2 1.5 No 135 Q14008-2 Isoform 2 of Cytoskeleton-associated protein 5 CKAP5_HUMAN 2 1 No Q14C86-4 Isoform 4 of GTPase-activating protein and VPS9 domain- 1.5 29.5 No 136 containing protein 1 GAPD1_HUMAN 137 P62937 Peptidyl-prolyl cis-trans isomerase A PPIA_HUMAN 1.5 14.5 No 138 Q99798 Aconitate hydratase, mitochondrial ACON_HUMAN 1.5 8.5 No 139 O43291 Kunitz-type protease inhibitor 2 SPIT2_HUMAN 1.5 6 No 140 Q02790 Peptidyl-prolyl cis-trans isomerase FKBP4 FKBP4_HUMAN 1.5 5 No 141 Q15397 Pumilio domain-containing protein KIAA0020 K0020_HUMAN 1.5 5 No 142 Q96SB4 SRSF protein kinase 1 SRPK1_HUMAN 1.5 5 No

170

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor P46977 Dolichyl-diphosphooligosaccharide--protein glycosyltransferase 1.5 4.5 No 143 subunit STT3A STT3A_HUMAN 144 P07686 Beta-hexosaminidase subunit beta HEXB_HUMAN 1.5 4 No 145 Q6P1M0 Long-chain fatty acid transport protein 4 S27A4_HUMAN 1.5 3.5 No 146 P11532-3 Isoform 2 of Dystrophin DMD_HUMAN 1.5 2.5 No 147 Q9NQS7-2 Isoform 2 of Inner centromere protein INCE_HUMAN 1.5 2.5 No 148 O96008 Mitochondrial import receptor subunit TOM40 homolog TOM40_HUMAN 1.5 2 No 149 O60488-2 Isoform Short of Long-chain-fatty-acid--CoA ligase 4 ACSL4_HUMAN 1.5 2 No 150 Q9UL25 Ras-related protein Rab-21 RAB21_HUMAN 1.5 2 Yes (Ewing, 2007) 151 Q9Y2W1 Thyroid hormone receptor-associated protein 3 TR150_HUMAN 1 25 No 152 Q9Y3T9 Nucleolar complex protein 2 homolog NOC2L_HUMAN 1 7 No 153 P07195 L-lactate dehydrogenase B chain LDHB_HUMAN 1 6.5 No 154 Q13243 Serine/arginine-rich splicing factor 5 SRSF5_HUMAN 1 6.5 No 155 P22061 Protein-L-isoaspartate(D-aspartate) O-methyltransferase PIMT_HUMAN 1 5.5 Yes (Ewing, 2007) Q5JTV8 Torsin-1A-interacting protein 1 OS=Homo sapiens GN=TOR1AIP1 1 5.5 No 156 PE=1 SV=2 - [TOIP1_HUMAN] TOIP1_HUMAN 157 O75494-5 Isoform 5 of Serine/arginine-rich splicing factor 10 SRS10_HUMAN 1 5.5 No 158 P35914 Hydroxymethylglutaryl-CoA lyase, mitochondrial HMGCL_HUMAN 1 5 No 159 Q9H9B4 Sideroflexin-1 SFXN1_HUMAN 1 4.5 Yes (Ewing, 2007) 160 O00487 26S proteasome non-ATPase regulatory subunit 14 PSDE_HUMAN 1 4.5 Yes (Ewing, 2007) 161 O43684-2 Isoform 2 of Mitotic checkpoint protein BUB3 BUB3_HUMAN 1 4 No 162 Q9UNQ2 Probable dimethyladenosine transferase DIM1_HUMAN 1 4 No 163 Q8ND30 Liprin-beta-2 LIPB2_HUMAN 1 4 No 164 Q96F07-2 Isoform 2 of Cytoplasmic FMR1-interacting protein 2 CYFP2_HUMAN 1 4 No 165 Q9NTK5 Obg-like ATPase 1 OLA1_HUMAN 1 4 No 166 P62879 Guanine nucleotide-binding protein G(I)/G(S)/G(T) subunit beta-2 GBB2_HUMAN 1 3.5 No 167 Q9UKG1 DCC-interacting protein 13-alpha DP13A_HUMAN 1 3.5 No 168 Q9Y5K5-2 Isoform 2 of Ubiquitin carboxyl-terminal hydrolase isozyme L5 UCHL5_HUMAN 1 3 No 169 P61964 WD repeat-containing protein 5 WDR5_HUMAN 1 3 Yes (Ewing, 2007) 170 Q8NE71-2 Isoform 2 of ATP-binding cassette sub-family F member 1 ABCF1_HUMAN 1 3 No 171 Q9BQ75-2 Isoform 2 of Protein CMSS1 CMS1_HUMAN 1 2.5 No

171

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 172 Q14764 Major vault protein MVP_HUMAN 1 2.5 No 173 Q96N66-2 Isoform 2 of Lysophospholipid acyltransferase 7 MBOA7_HUMAN 1 2.5 No 174 O75477 Erlin-1 ERLN1_HUMAN 1 2.5 No 175 Q9Y679-2 Isoform Short of Ancient ubiquitous protein 1 AUP1_HUMAN 1 2 No 176 Q15149-4 Isoform 4 of Plectin PLEC_HUMAN 61.5 0 No 177 P33527-4 Isoform 4 of Multidrug resistance-associated protein 1 MRP1_HUMAN 9.5 0 No 178 Q8N884 Cyclic GMP-AMP synthase CGAS_HUMAN 9 0.5 No 179 P43121 Cell surface glycoprotein MUC18 MUC18_HUMAN 6.5 0 No 180 P31946-2 Isoform Short of 14-3-3 protein beta/alpha 1433B_HUMAN 6 0 No 181 O00161-2 Isoform SNAP-23b of Synaptosomal-associated protein 23 SNP23_HUMAN 5.5 0.5 No 182 Q9BUR5 Apolipoprotein O APOO_HUMAN 5.5 0.5 No 183 P46459 Vesicle-fusing ATPase NSF_HUMAN 5 0 No 184 Q04917 14-3-3 protein eta 1433F_HUMAN 4.5 0.5 No 185 Q7Z403 Transmembrane channel-like protein 6 TMC6_HUMAN 4.5 0 No 186 Q92508 Piezo-type mechanosensitive ion channel component 1 PIEZ1_HUMAN 4.5 0 No 187 Q8NC56 LEM domain-containing protein 2 LEMD2_HUMAN 4 0.5 No 188 P31947-2 Isoform 2 of 14-3-3 protein sigma 1433S_HUMAN 4 0 No 189 Q9C0B5-2 Isoform 2 of Palmitoyltransferase ZDHHC5 ZDHC5_HUMAN 3.5 0 No 190 Q8IV63-3 Isoform 3 of Inactive serine/threonine-protein kinase VRK3 VRK3_HUMAN 3.5 0 No 191 Q8NBJ5 Procollagen galactosyltransferase 1 GT251_HUMAN 3.5 0 No 192 O96000 NADH dehydrogenase [ubiquinone] 1 beta subcomplex subunit 10 NDUBA_HUMAN 3 0.5 No 193 Q9Y5Y6 Suppressor of tumorigenicity 14 protein ST14_HUMAN 3 0 No 194 Q96S97 Myeloid-associated differentiation marker MYADM_HUMAN 3 0 No 195 Q92485 Acid sphingomyelinase-like phosphodiesterase 3b ASM3B_HUMAN 3 0 No 196 Q13185 Chromobox protein homolog 3 CBX3_HUMAN 3 0 No 197 Q9H078-2 Isoform 2 of Caseinolytic peptidase B protein homolog CLPB_HUMAN 2.5 0.5 No 198 Q8NBI5 Solute carrier family 43 member 3 S43A3_HUMAN 2.5 0 No 199 O94776 Metastasis-associated protein MTA2 MTA2_HUMAN 2.5 0 No 200 Q13425-2 Isoform 2 of Beta-2-syntrophin SNTB2_HUMAN 2.5 0 No 201 Q13951-2 Isoform 2 of Core-binding factor subunit beta PEBB_HUMAN 2.5 0 No 202 Q99595 Mitochondrial import inner membrane translocase subunit Tim17-A TI17A_HUMAN 2.5 0 No

172

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 203 Q8TB52 F-box only protein 30 FBX30_HUMAN 2 0 No 204 Q6PJF5-2 Isoform 2 of Inactive rhomboid protein 2 RHDF2_HUMAN 2 0 No 205 Q9H4I3 TraB domain-containing protein TRABD_HUMAN 2 0 No 206 Q15599-2 Isoform 2 of Na(+)/H(+) exchange regulatory cofactor NHE-RF2 NHRF2_HUMAN 2 0 No 207 Q9H061 Transmembrane protein 126A T126A_HUMAN 2 0 No 208 P12270 Nucleoprotein TPR TPR_HUMAN 0 22 No 209 Q9H307 Pinin PININ_HUMAN 0 18.5 No 210 Q9NYF8-2 Isoform 2 of Bcl-2-associated transcription factor 1 BCLF1_HUMAN 0 18 No 211 P49674 Casein kinase I isoform epsilon KC1E_HUMAN 0 15.5 Yes (Ewing, 2007) 212 Q9UQ35 Serine/arginine repetitive matrix protein 2 SRRM2_HUMAN 0 15.5 No 213 Q14980-2 Isoform 2 of Nuclear mitotic apparatus protein 1 NUMA1_HUMAN 0.5 14.5 No 214 Q86V48-2 Isoform 2 of Leucine zipper protein 1 LUZP1_HUMAN 0 14.5 No 215 P48730-2 Isoform 2 of Casein kinase I isoform delta KC1D_HUMAN 0 14.5 Yes (Ewing, 2007) 216 Q9UKV3 Apoptotic chromatin condensation inducer in the nucleus ACINU_HUMAN 0.5 14 No 217 Q14978-3 Isoform 3 of Nucleolar and coiled-body phosphoprotein 1 NOLC1_HUMAN 0 13 No 218 Q14978 Nucleolar and coiled-body phosphoprotein 1 NOLC1_HUMAN 0 13 No 219 P02788-2 Isoform DeltaLf of Lactotransferrin TRFL_HUMAN 0 11.5 No 220 P27816-6 Isoform 6 of Microtubule-associated protein 4 MAP4_HUMAN 0.5 10 No 221 Q8N163 DBIRD complex subunit KIAA1967 K1967_HUMAN 0 10 No 222 O75828 Carbonyl reductase [NADPH] 3 CBR3_HUMAN 0 9.5 No 223 P05109 Protein S100-A8 S10A8_HUMAN 0.5 9 No 224 Q9H2U2 Inorganic pyrophosphatase 2, mitochondrial IPYR2_HUMAN 0.5 8.5 No 225 O75691 Small subunit processome component 20 homolog UTP20_HUMAN 0 8.5 No 226 P62191 26S protease regulatory subunit 4 PRS4_HUMAN 0 8 No 227 Q96R06 Sperm-associated antigen 5 SPAG5_HUMAN 0 8 No 228 Q15154 Pericentriolar material 1 protein PCM1_HUMAN 0 8 No 229 Q9BZQ6 ER degradation-enhancing alpha-mannosidase-like protein 3 EDEM3_HUMAN 0.5 7.5 No 230 Q13595 Transformer-2 protein homolog alpha TRA2A_HUMAN 0.5 7 No 231 Q9NZ01 Very-long-chain enoyl-CoA reductase TECR_HUMAN 0 7 No 232 Q9NY61 Protein AATF AATF_HUMAN 0 7 No 233 P46109 Crk-like protein CRKL_HUMAN 0 6.5 No

173

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 234 O75663 TIP41-like protein TIPRL_HUMAN 0 6.5 No 235 Q06323 Proteasome activator complex subunit 1 PSME1_HUMAN 0 6.5 No 236 P29590-14 Isoform PML-14 of Protein PML PML_HUMAN 0 6.5 No 237 Q9P0J7 E3 ubiquitin-protein ligase KCMF1 KCMF1_HUMAN 0 6.5 No 238 Q8IZT6-2 Isoform 2 of Abnormal spindle-like microcephaly-associated protein ASPM_HUMAN 0 6.5 No P49821-2 Isoform 2 of NADH dehydrogenase [ubiquinone] flavoprotein 1, 0.5 6 No 239 mitochondrial NDUV1_HUMAN 240 P45974-2 Isoform Short of Ubiquitin carboxyl-terminal hydrolase 5 UBP5_HUMAN 0.5 6 No 241 Q5SRE5-2 Isoform 2 of Nucleoporin NUP188 homolog NU188_HUMAN 0.5 6 No 242 P51114 Fragile X mental retardation syndrome-related protein 1 FXR1_HUMAN 0 6 No 243 Q9H9E3 Conserved oligomeric Golgi complex subunit 4 COG4_HUMAN 0 6 No 244 P29353-2 Isoform p52Shc of SHC-transforming protein 1 SHC1_HUMAN -0.5 6 No 245 Q92597 Protein NDRG1 NDRG1_HUMAN -0.5 6 No 246 Q9BRK5 45 kDa calcium-binding protein CAB45_HUMAN 0.5 5.5 No 247 P50897 Palmitoyl-protein thioesterase 1 PPT1_HUMAN 0.5 5.5 No 248 Q6NVY1 3-hydroxyisobutyryl-CoA hydrolase, mitochondrial HIBCH_HUMAN 0.5 5.5 No 249 O00170 AH receptor-interacting protein AIP_HUMAN 0 5.5 No 250 Q9NVN8 Guanine nucleotide-binding protein-like 3-like protein GNL3L_HUMAN 0 5.5 No 251 Q00534 Cyclin-dependent kinase 6 CDK6_HUMAN 0 5.5 No 252 Q96FW1 Ubiquitin thioesterase OTUB1 OTUB1_HUMAN 0 5 Yes (Ewing, 2007) 253 Q7LBC6 Lysine-specific demethylase 3B KDM3B_HUMAN 0 5 No 254 Q5VT06 Centrosome-associated protein 350 CE350_HUMAN 0 5 No 255 Q8TC07-2 Isoform 2 of TBC1 domain family member 15 TBC15_HUMAN 0 5 Yes (Ewing, 2007) 256 P54819-5 Isoform 5 of Adenylate kinase 2, mitochondrial KAD2_HUMAN 0.5 4.5 Yes (Ewing, 2007) 257 Q6FI81-3 Isoform 3 of Anamorsin CPIN1_HUMAN 0.5 4.5 No 258 Q8WUM0 Nuclear pore complex protein Nup133 NU133_HUMAN 0.5 4.5 No 259 O75832 26S proteasome non-ATPase regulatory subunit 10 PSD10_HUMAN 0 4.5 No 260 O00233 26S proteasome non-ATPase regulatory subunit 9 PSMD9_HUMAN 0 4.5 No 261 O60711 Leupaxin LPXN_HUMAN 0 4.5 No 262 P61289 Proteasome activator complex subunit 3 PSME3_HUMAN 0 4.5 Yes (Ewing, 2007) 263 P55735-2 Isoform 2 of Protein SEC13 homolog SEC13_HUMAN 0 4.5 Yes (Ewing, 2007)

174

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 264 Q8IYS1 Peptidase M20 domain-containing protein 2 P20D2_HUMAN 0 4.5 No Q8IUF8-4 Isoform 4 of Bifunctional lysine-specific demethylase and histidyl- 0 4.5 No 265 hydroxylase MINA MINA_HUMAN 266 Q9UHD8-7 Isoform 7 of Septin-9 SEPT9_HUMAN 0 4.5 No 267 Q9ULX6 A-kinase anchor protein 8-like AKP8L_HUMAN 0 4.5 No 268 Q8N3U4 Cohesin subunit SA-2 STAG2_HUMAN 0 4.5 No 269 P59045-3 Isoform 3 of NACHT, LRR and PYD domains-containing protein 11 NAL11_HUMAN 0 4.5 No 270 O75330-4 Isoform 4 of Hyaluronan mediated motility receptor HMMR_HUMAN 0 4.5 No 271 Q6RFH5 WD repeat-containing protein 74 WDR74_HUMAN 0.5 4 No O75569-3 Isoform 3 of Interferon-inducible double stranded RNA-dependent 0.5 4 No 272 protein kinase activator A PRKRA_HUMAN 273 Q9Y5Y2 Cytosolic Fe-S cluster assembly factor NUBP2 NUBP2_HUMAN 0.5 4 No 274 Q9Y3C1 Nucleolar protein 16 NOP16_HUMAN 0.5 4 No 275 Q8NEJ9-2 Isoform 2 of Neuroguidin NGDN_HUMAN 0.5 4 No 276 Q92785 Zinc finger protein ubi-d4 REQU_HUMAN 0.5 4 No 277 Q9GZP4-2 Isoform 2 of PITH domain-containing protein 1 PITH1_HUMAN 0 4 No 278 P61758 Prefoldin subunit 3 PFD3_HUMAN 0 4 No 279 Q9UBV8 Peflin PEF1_HUMAN 0 4 No 280 P68400 Casein kinase II subunit alpha CSK21_HUMAN 0 4 No 281 Q9H501 ESF1 homolog ESF1_HUMAN 0 4 No 282 Q7Z2W4 Zinc finger CCCH-type antiviral protein 1 ZCCHV_HUMAN 0 4 No Q15628 Tumor necrosis factor receptor type 1-associated DEATH domain 0.5 3.5 No 283 protein TRADD_HUMAN 284 Q86YT6 E3 ubiquitin-protein ligase MIB1 MIB1_HUMAN 0.5 3.5 No 285 P25787 Proteasome subunit alpha type-2 PSA2_HUMAN 0 3.5 No 286 P09525 Annexin A4 ANXA4_HUMAN 0 3.5 No 287 P40925 Malate dehydrogenase, cytoplasmic MDHC_HUMAN 0 3.5 No Q16740 Putative ATP-dependent Clp protease proteolytic subunit, 0 3.5 No 288 mitochondrial CLPP_HUMAN 289 O95861 3'(2'),5'-bisphosphate nucleotidase 1 BPNT1_HUMAN 0 3.5 No 290 P23526-2 Isoform 2 of Adenosylhomocysteinase SAHH_HUMAN 0 3.5 Yes (Ewing, 2007) 291 Q15054 DNA polymerase delta subunit 3 DPOD3_HUMAN 0 3.5 No

175

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 292 O15355 Protein phosphatase 1G PPM1G_HUMAN 0 3.5 No 293 Q9UJU6 Drebrin-like protein DBNL_HUMAN 0 3.5 No 294 Q8IWC1-2 Isoform 2 of MAP7 domain-containing protein 3 MA7D3_HUMAN 0 3.5 No 295 Q9H3P7 Golgi resident protein GCP60 GCP60_HUMAN 0 3.5 No 296 O00267-2 Isoform 2 of Transcription elongation factor SPT5 SPT5H_HUMAN 0 3.5 No 297 Q6ZU80 Centrosomal protein of 128 kDa CE128_HUMAN 0 3.5 No 298 Q8IWJ2 GRIP and coiled-coil domain-containing protein 2 GCC2_HUMAN 0 3.5 No 299 Q9H9E3-3 Isoform 3 of Conserved oligomeric Golgi complex subunit 4 COG4_HUMAN 0 3.5 No P36957 Dihydrolipoyllysine-residue succinyltransferase component of 2- 0.5 3 No 300 oxoglutarate dehydrogenase complex, mitochondrial ODO2_HUMAN 301 P41227-2 Isoform 2 of N-alpha-acetyltransferase 10 NAA10_HUMAN 0 3 Yes (Ewing, 2007) 302 Q96J01 THO complex subunit 3 THOC3_HUMAN 0 3 No 303 Q96P16 Regulation of nuclear pre-mRNA domain-containing protein 1A RPR1A_HUMAN 0 3 No 304 P35520 Cystathionine beta-synthase CBS_HUMAN 0 3 No 305 Q9UJU6-2 Isoform 2 of Drebrin-like protein DBNL_HUMAN 0 3 No 306 P23458 Tyrosine-protein kinase JAK1 JAK1_HUMAN 0 3 No 307 Q9HA64 Ketosamine-3-kinase KT3K_HUMAN 0 3 No 308 Q6PJT7-4 Isoform 4 of Zinc finger CCCH domain-containing protein 14 ZC3HE_HUMAN 0 3 No 309 Q86VM9 Zinc finger CCCH domain-containing protein 18 ZCH18_HUMAN 0 3 No 310 Q9Y2X9 Zinc finger protein 281 ZN281_HUMAN 0 3 No 311 Q9BUQ8 Probable ATP-dependent RNA helicase DDX23 DDX23_HUMAN 0 3 No 312 O14976 Cyclin-G-associated kinase GAK_HUMAN 0 3 No 313 Q53H96 Pyrroline-5-carboxylate reductase 3 P5CR3_HUMAN 0 3 No 314 P52732 Kinesin-like protein KIF11 KIF11_HUMAN 0 3 No Q96KA5-2 Isoform 2 of Cleft lip and palate transmembrane protein 1-like 0 3 No 315 protein CLP1L_HUMAN 316 P55769 NHP2-like protein 1 NH2L1_HUMAN 0.5 2.5 No 317 Q15691 Microtubule-associated protein RP/EB family member 1 MARE1_HUMAN 0.5 2.5 Yes (Ewing, 2007) Q15653-2 Isoform 2 of NF-kappa-B inhibitor beta 0 2.5 Yes (Bouwmeester, 318 IKBB_HUMAN 2004) 319 Q9NUQ9 Protein FAM49B FA49B_HUMAN 0 2.5 No

176

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 320 P62993 Growth factor receptor-bound protein 2 GRB2_HUMAN 0 2.5 Yes (Ewing, 2007) 321 Q96EY8 Cob(I)yrinic acid a,c-diamide adenosyltransferase, mitochondrial MMAB_HUMAN 0 2.5 No P51553-2 Isoform 2 of Isocitrate dehydrogenase [NAD] subunit gamma, 0 2.5 No 322 mitochondrial IDH3G_HUMAN 323 Q9BTE7 DCN1-like protein 5 DCNL5_HUMAN 0 2.5 No 324 Q16637-4 Isoform SMN-delta57 of Survival motor neuron protein SMN_HUMAN 0 2.5 No 325 Q9H8H0 Nucleolar protein 11 NOL11_HUMAN 0 2.5 No 326 Q9BV38 WD repeat-containing protein 18 WDR18_HUMAN 0 2.5 No 327 P08758 Annexin A5 ANXA5_HUMAN 0 2.5 No 328 Q6JBY9 CapZ-interacting protein CPZIP_HUMAN 0 2.5 No 329 Q9Y4R8 Telomere length regulation protein TEL2 homolog TELO2_HUMAN 0 2.5 No 330 Q92540-2 Isoform 2 of Protein SMG7 SMG7_HUMAN 0 2.5 No 331 Q76FK4-4 Isoform 4 of Nucleolar protein 8 NOL8_HUMAN 0 2.5 No 332 Q9UJC3 Protein Hook homolog 1 HOOK1_HUMAN 0 2.5 No 333 P11802 Cyclin-dependent kinase 4 CDK4_HUMAN 0 2.5 Yes (Ewing, 2007) 334 Q00535-2 Isoform 2 of Cyclin-dependent kinase 5 CDK5_HUMAN 0 2.5 No P04083 Annexin A1 0 2.5 Yes (Sigglekow, 335 ANXA1_HUMAN 2012) 336 P43487 Ran-specific GTPase-activating protein RANG_HUMAN 0.5 2 Yes (Ewing, 2007) 337 P54920 Alpha-soluble NSF attachment protein SNAA_HUMAN 0.5 2 No 338 Q6UXN9 WD repeat-containing protein 82 WDR82_HUMAN 0.5 2 No 339 Q14790-8 Isoform 8 of Caspase-8 CASP8_HUMAN 0.5 2 No 340 Q96T51 RUN and FYVE domain-containing protein 1 RUFY1_HUMAN 0.5 2 No 341 Q86Y82 Syntaxin-12 STX12_HUMAN 0.5 2 No 342 Q9NXW2 DnaJ homolog subfamily B member 12 DJB12_HUMAN 0.5 2 No 343 P16591-3 Isoform 3 of Tyrosine-protein kinase Fer FER_HUMAN 0.5 2 No 344 Q9BVC4 Target of rapamycin complex subunit LST8 LST8_HUMAN 0.5 2 No 345 Q8IZP0-10 Isoform 10 of Abl interactor 1 ABI1_HUMAN 0 2 No 346 P48556 26S proteasome non-ATPase regulatory subunit 8 PSMD8_HUMAN 0 2 Yes (Ewing, 2007) 347 Q9H6Y2 WD repeat-containing protein 55 WDR55_HUMAN 0 2 No 348 Q9NQT4 Exosome complex component RRP46 EXOS5_HUMAN 0 2 No

177

Average spectral count difference (hMCC-SBP-6xHis - FLAG- hMCC) Previously known # Accession Description SwissProt ID Mitochondria Whole lysates MCC-interactor 349 Q15003 Condensin complex subunit 2 CND2_HUMAN 0 2 No 350 P38432 Coilin COIL_HUMAN 0 2 No 351 Q86X83 COMM domain-containing protein 2 COMD2_HUMAN 0 2 No 352 Q6IQ49-2 Isoform 2 of Protein SDE2 homolog SDE2_HUMAN 0 2 No 353 P42574 Caspase-3 CASP3_HUMAN 0 2 No 354 Q3KQU3-2 Isoform 2 of MAP7 domain-containing protein 1 MA7D1_HUMAN 0 2 No 355 P39687 Acidic leucine-rich nuclear phosphoprotein 32 family member A AN32A_HUMAN 0 2 No 356 P78417-3 Isoform 3 of Glutathione S-transferase omega-1 GSTO1_HUMAN 0 2 No 357 O00273 DNA fragmentation factor subunit alpha DFFA_HUMAN 0 2 Yes (Ewing, 2007) 358 Q7Z460-2 Isoform 2 of CLIP-associating protein 1 CLAP1_HUMAN 0 2 No 359 Q5JSZ5 Protein PRRC2B PRC2B_HUMAN 0 2 No 360 O15118 Niemann-Pick C1 protein NPC1_HUMAN 0 2 No 361 Q9NWH9 SAFB-like transcription modulator SLTM_HUMAN 0 2 No 362 Q9Y263 Phospholipase A-2-activating protein PLAP_HUMAN 0 2 No 363 P15927 Replication protein A 32 kDa subunit RFA2_HUMAN 0 2 No 364 Q9Y697-2 Isoform Cytoplasmic of Cysteine desulfurase, mitochondrial NFS1_HUMAN 0 2 No 365 Q9H3N1 Thioredoxin-related transmembrane protein 1 TMX1_HUMAN 0 2 Yes (Ewing, 2007)

178

179

Table 3. Sox5 primers used in this study Primer name* Sequence Orientation mSox5-201F 5'- ATG CTT ACT GAC CCT GAT -3' Forward mSox5-153F 5'- CCG ATC ATA GGT GGC CGC TG -3' Forward Sox5-2492R 5'- TCA GTT GGC TTG TCC CGC -3' Reverse mSox5-2567R 5'- TGC TTG GCC ACT GGG AAG GAT -3' Reverse mSox5-193F 5'- TTT CCA GCA TGC TTA CTG AC -3' Forward mSox5-1246R 5'- ATT GCG GCT AGC TGG GCA -3' Reverse mSox5-346F 5'- CCT TTC ACC TTC CCT TAC ATG -3' Forward mSox5-512F 5'- CAA CTC ATC TAC CTC ACC TCA G -3' Forward mSox5-1178R 5'- AGC TGC CAT AGT GGT TG -3' Reverse mSox5-968R 5'- CAG AAG CTG TTG -3' Reverse mSox5-899F 5'- ACA GCG TCA GCA GAT GGA G -3' Forward mSox5-1535R 5'- GCT AAC TCT TGC AGA AGG AC -3' Reverse mSox5-1426F 5'- CTG CAT CAC CCA CCT CTC -3' Forward

mSox5-1960R 5'- CTG ATG TTG GAA TTG TGC ATG -3' Reverse mSox5-1868F 5'- AAA GCG TCC AAT GAA TGC -3' Forward mSox5-2520R 5'- TCG GCT CAC AGT CTG -3' Reverse mSox5-195F 5'- TCC AGC ATG CTT ACT GAC CCT -3' Forward mSox5-536R 5'- CTT CTG AGG TGA GGT AGA TG -3' Reverse mSox5 5' BamHI 5'- GGC GGA TCC ATG CTT ACT GAC CCT GAT -3' Forward mSox5-R-XbaI 5'- CTA GTC TAG ACT TAT CAG TTG GCT TGT CC Reverse -3' mSox5-1030F 5'- CAT TGA TTC CCG TGT TC -3' Forward mSox5-1181F 5'- TGC AGC AAC ACC AGG CTT AG -3' Forward mSox5-1412R 5'- AGA GGT CTT GGG CTT AGC TG -3' Reverse mSox5-1463R 5'- TCT CAG AGC TGG CAT GTG AG -3' Reverse mS-Sox5-12Fs 5'- AGA AGC TAG GAG ATG AAG CAG -3' Forward mS-Sox5-42Fs 5'- AGC CTC TTT CTT CAG CGT AC -3' Forward * Primers are named based on their position at the L-Sox5 transcript (NM_011444.3) The mouse L-Sox5 start codon "ATG" is at 201-203, and stop codon "TGA" is at 2490-2492. S The two primers specific for the short isoform of Sox5 are named based on their position at the S-Sox5 transcript (X65657.1).

180

Table 4. Results of PCR cloning of the Sox5 cDNA from TRAF3-/- mouse B lymphomas

Predicted Primer pair Template cDNA size Forward Reverse (bp) 6983-2 115-6 33-3 7060-8 spleen spleen spleen ascites mSox5-201F mSox5-2492R 2292 No No No No mSox5-201F mSox5-2567R 2367 No No No No mSox5-153F mSox5-2492R 2340 No No No No mSox5-153F mSox5-2567R 2415 No No No No mSox5-193F mSox5-1246R 1054 No No No No mSox5-153F mSox5-1178R 1026 No No No No mSox5-201F mSox5-1178R 978 Wrong Wrong Wrong Wrong mSox5-201F mSox5-968R 768 Wrong Wrong Wrong Wrong mSox5-346F mSox5-968R 623 Yesa Yesa Yesa Yesa mSox5-346F mSox5-1178R 833 Wrong Wrong Wrong Wrong mSox5-346F mSox5-1535R 1190 No No No No mSox5-346F mSox5-2520R 2175 No No No No mSox5-512F mSox5-968R 457 Yes Yes Yes Yes mSox5-512F mSox5-1178R 667 Yes Yes Yes Yes mSox5-899F mSox5-1535R 637 Yes Yes Yes Yes mSox5-1426F mSox5-1960R 535 Yes Yes Yes Yes mSox5-1426F mSox5-2520R 1095 Yes Yes Yes Yes mSox5-1868F mSox5-2520R 653 Yes Yes Yes Yes mSox5-153F mSox5-2520R 2368 No No No No mSox5-201F mSox5-2520R 2320 Wrong Wrong Wrong Wrong mSox5-512F mSox5-1535R 1024 No No No No mSox5-346F mSox5-1535R 1190 No No No No mSox5-346F mSox5-2520R 2175 Wrong Wrong Wrong Wrong mSox5-195F mSox5-536R 342 Yesb Yesb Yesb Yesb a: the PCR products of (346F+968R) contain a deletion of 105 bp (position: 366-470 bp of L- Sox5) when compared to the L-Sox5 transcript (NM_011444.3), but are 100% identical to the corresponding region of the Sox5 variant 3 (Sox5 V3, NM_001243163.1). b: the PCR products of (195F+536R) contain a deletion of 105 bp (position: 366-470 bp of L- Sox5) when compared to the L-Sox5 transcript (NM_011444.3), but are 100% identical to the corresponding region of the Sox5 variant 3 (Sox5 V3, NM_001243163.1).

Table 5. Sequence overlap extension PCRs to obtain full-length coding sequence of the Sox5-BLM cDNA Reaction # Forward primer Reverse primer Template 1 Template 2 Product size (bp) SOE PCR 1 mSox5-195F mSox5-968R PCR product of 195F+536R PCR product of 512F+968R 669

SOE PCR 2 mSox5-195F mSox5-1535R SOE PCR 1 product PCR product of 899F+1535R 1236

SOE PCR 3 mSox5-195F mSox5-2520R SOE PCR 2 product PCR product of 1426F+2520R 2221

Table 6. Results of PCR cloning of Sox5-V3 and S-Sox5 transcripts from TRAF3-/- mouse B lymphomas Predicted size Primer pair PCR products generated from each template cDNA (bp) (bp) Sox5- Forward Reverse Sox5-V3 S-Sox5 6983-2 spl 115-6 spl 33-3 spl 7060-8 asc Testist BLM mSox5-1030F mSox5-1412R 383 236 - 383a 383a 383a 383a 236 mSox5-1030F mSox5-1463R 434 287 - 434b 434b 434b 434b 287 mSox5-1181F mSox5-1412R 232 85 85 232c 232c 232c 232c 85 mSox5-1181F mSox5-1463R 283 136 136 283d 283d 283d 283d 136 mS-Sox5-12F mSox5-1412R - - 214 No No No No 214 mS-Sox5-12F mSox5-1463R - - 265 No No No No 265 mS-Sox5-42F mSox5-1412R - - 184 No No No No 184 mS-Sox5-42F mSox5-1463R - - 235 No No No No 235 a: the PCR products of (1030F+1412R) contain a dominant band of 383 bp, but also a faint band of 236 bp. b: the PCR products of (1030F+1463R) contain a dominant band of 434 bp, but also a faint band of 287 bp. c: the PCR products of (1181F+1412R) contain a dominant band of 232 bp, but also a faint band of 85 bp. d: the PCR products of (1181F+1463R) contain a dominant band of 283 bp, but also a faint band of 136 bp. t: Mouse testis cDNA was used as positive control for the detection of the short isoform of Sox5 (S-Sox5).

181

182

APPENDIX 2: LIST OF ABBREVIATIONS USED

AD 198: N-Benzyladriamycin-14-valerate

B-TRAF3-/-: B cell-specific TRAF3-deficient

BAFF: B-cell activating factor

CLL: B cell chronic lymphocytic leukemia

CRC: colorectal cancer

NHL: Non-Hodgkin lymphoma

MCC: mutated in colorectal cancer, also known as colorectal cancer mutant protein

MCL: mantle cell lymphoma

MM: multiple myeloma

PEP005: ingenol-3-angelate

SMZL: splenic marginal zone lymphoma

TRAF3: tumor necrosis factor receptor (TNF-R)-associated factor 3

WM: Waldenström’s Macroglobulinemia

183

ACKNOWLEDGEMENT OF PREVIOUS PUBLICATIONS

Edwards SK, Moore CR, Liu Y, Grewal S, Covey LR, and Xie P. N- Benzyladriamycin-14-valerate (AD 198) exhibits potent anti-tumor activity on TRAF3-deficient mouse B lymphoma and human multiple myeloma. BMC Cancer, 13: 481, pages 1-20, 2013.

Edwards SKE, Desai A, Moore CR, Liu Y, Hart R.P, and Xie P. Expression and function of a novel isoform of Sox5 in malignant B cells. Leukemia Research, 38: 393-401, 2014.

Edwards SKE, Baron, J, Moore CR, Liu Y, Perlman, D, Hart R.P, and Xie P. Mutated in colorectal cancer (MCC) is a novel oncogene in B lymphocytes. Journal of Hematology & Oncology, 7:56, pages 1-24, 2014.

Moore CR, Edwards SKE, and Xie P. Targeting TRAF3 downstream signaling pathways in B cell neoplasms. Journal of Cancer Science and Therapy, 7.2, pages 067-074, 2015. (Invited review)

Edwards S, Liu Y, Desai A, Baron J, Grewal S, Moore CR, and Xie P. Complex signaling mechanisms of bortezomib in TRAF3-deficient mouse B lymphoma and human multiple myeloma cells. Leukemia Research. In revision

184

BIBLIOGRAPHY

1 Society, A. C. (American Cancer Society, Atlanta, 2014).

2 Ruddon, R. Ch. The Epidemiology of Human Cancer, 62-116 (Oxford University Press, 2007). 3 Morton, L. M. et al. Lymphoma incidence patterns by WHO subtype in the United States, 1992-2001. Blood 107, 265-276, doi:10.1182/blood-2005-06-2508 (2006).

4 Shaffer, A. L., Rosenwald, A. & Staudt, L. M. Lymphoid malignancies: the dark side of B-cell differentiation. Nature reviews. Immunology 2, 920-932, doi:10.1038/nri953 (2002).

5 Rajewsky, K. Clonal selection and learning in the antibody system. Nature 381, 751-758, doi:10.1038/381751a0 (1996).

6 Dudley, D. D., Chaudhuri, J., Bassing, C. H. & Alt, F. W. Mechanism and control of V(D)J recombination versus class switch recombination: similarities and differences. Advances in immunology 86, 43-112, doi:10.1016/S0065- 2776(04)86002-4 (2005).

7 Pasqualucci, L. et al. Hypermutation of multiple proto-oncogenes in B-cell diffuse large-cell lymphomas. Nature 412, 341-346, doi:10.1038/35085588 (2001).

8 Peled, J. U. et al. The biochemistry of somatic hypermutation. Annual review of immunology 26, 481-511, doi:10.1146/annurev.immunol.26.021607.090236 (2008).

9 Kuppers, R. & Dalla-Favera, R. Mechanisms of chromosomal translocations in B cell lymphomas. Oncogene 20, 5580-5594, doi:10.1038/sj.onc.1204640 (2001).

10 Spicuglia, S., Franchini, D. M. & Ferrier, P. Regulation of V(D)J recombination. Current opinion in immunology 18, 158-163, doi:10.1016/j.coi.2006.01.003 (2006).

11 Soulas-Sprauel, P. et al. V(D)J and immunoglobulin class switch recombinations: a paradigm to study the regulation of DNA end-joining. Oncogene 26, 7780-7791, doi:10.1038/sj.onc.1210875 (2007).

12 Seto, M. Genetic and epigenetic factors involved in B-cell lymphomagenesis. Cancer science 95, 704-710 (2004).

13 Pasqualucci, L. et al. AID is required for germinal center-derived lymphomagenesis. Nature genetics 40, 108-112, doi:10.1038/ng.2007.35 (2008).

14 Okazaki, I. M., Kotani, A. & Honjo, T. Role of AID in tumorigenesis. Advances in immunology 94, 245-273, doi:10.1016/S0065-2776(06)94008-5 (2007).

185

15 Moore, C. R. et al. Specific deletion of TRAF3 in B lymphocytes leads to B- lymphoma development in mice. Leukemia 26, 1122-1127, doi:10.1038/leu.2011.309 (2012).

16 Xie, P., Stunz, L. L., Larison, K. D., Yang, B. & Bishop, G. A. Tumor necrosis factor receptor-associated factor 3 is a critical regulator of B cell homeostasis in secondary lymphoid organs. Immunity 27, 253-267, doi:10.1016/j.immuni.2007.07.012 (2007).

17 Bishop, G. A. & Xie, P. Multiple roles of TRAF3 signaling in lymphocyte function. Immunologic research 39, 22-32 (2007).

18 Xie, P., Kraus, Z. J., Stunz, L. L. & Bishop, G. A. Roles of TRAF molecules in B lymphocyte function. Cytokine & growth factor reviews 19, 199-207, doi:10.1016/j.cytogfr.2008.04.002 (2008).

19 Inoue, J. et al. Tumor necrosis factor receptor-associated factor (TRAF) family: adapter proteins that mediate cytokine signaling. Experimental cell research 254, 14-24, doi:10.1006/excr.1999.4733 (2000).

20 Wajant, H., Henkler, F. & Scheurich, P. The TNF-receptor-associated factor family: scaffold molecules for cytokine receptors, kinases and their regulators. Cellular signalling 13, 389-400 (2001).

21 Xu, L. G., Li, L. Y. & Shu, H. B. TRAF7 potentiates MEKK3-induced AP1 and CHOP activation and induces apoptosis. The Journal of biological chemistry 279, 17278-17282, doi:10.1074/jbc.C400063200 (2004).

22 Cerhan, J. R. et al. Genetic variation in 1253 immune and inflammation genes and risk of non-Hodgkin lymphoma. Blood 110, 4455-4463, doi:10.1182/blood-2007- 05-088682 (2007).

23 Keats, J. J. et al. Promiscuous mutations activate the noncanonical NF-kappaB pathway in multiple myeloma. Cancer Cell 12, 131-144 (2007).

24 Annunziata, C. M. et al. Frequent engagement of the classical and alternative NF- kappaB pathways by diverse genetic abnormalities in multiple myeloma. Cancer Cell 12, 115-130 (2007).

25 Demchenko, Y. N. et al. Classical and/or alternative NF-kappaB pathway activation in multiple myeloma. Blood 115, 3541-3552, doi:10.1182/blood-2009-09-243535 (2010).

26 Compagno, M. et al. Mutations of multiple genes cause deregulation of NF-kappaB in diffuse large B-cell lymphoma. Nature 459, 717-721, doi:10.1038/nature07968 (2009).

27 Rahal, R. et al. Pharmacological and genomic profiling identifies NF-kappaB- targeted treatment strategies for mantle cell lymphoma. Nature medicine 20, 87- 92, doi:10.1038/nm.3435 (2014).

186

28 Meissner, B. et al. The E3 ubiquitin ligase UBR5 is recurrently mutated in mantle cell lymphoma. Blood 121, 3161-3164, doi:10.1182/blood-2013-01-478834 (2013).

29 Baldoni, S. et al. NOTCH and NF-kappaB interplay in chronic lymphocytic leukemia is independent of genetic lesion. International journal of hematology 98, 153-157, doi:10.1007/s12185-013-1368-y (2013).

30 Thomas, G. S., Zhang, L., Blackwell, K. & Habelhah, H. Phosphorylation of TRAF2 within its RING domain inhibits stress-induced cell death by promoting IKK and suppressing JNK activation. Cancer research 69, 3665-3672, doi:10.1158/0008- 5472.CAN-08-4867 (2009).

31 Ruiz-Ballesteros, E. et al. Splenic marginal zone lymphoma: proposal of new diagnostic and prognostic markers identified after tissue and cDNA microarray analysis. Blood 106, 1831-1838, doi:10.1182/blood-2004-10-3898 (2005).

32 Hu, H. M., O'Rourke, K., Boguski, M. S. & Dixit, V. M. A novel RING finger protein interacts with the cytoplasmic domain of CD40. The Journal of biological chemistry 269, 30069-30072 (1994).

33 Cheng, G. et al. Involvement of CRAF1, a relative of TRAF, in CD40 signaling. Science 267, 1494-1498 (1995).

34 Mosialos, G. et al. The Epstein-Barr virus transforming protein LMP1 engages signaling proteins for the tumor necrosis factor receptor family. Cell 80, 389-399 (1995).

35 Oganesyan, G. et al. Critical role of TRAF3 in the Toll-like receptor-dependent and -independent antiviral response. Nature 439, 208-211, doi:10.1038/nature04374 (2006).

36 Hacker, H. et al. Specificity in Toll-like receptor signalling through distinct effector functions of TRAF3 and TRAF6. Nature 439, 204-207, doi:10.1038/nature04369 (2006).

37 Saha, S. K. & Cheng, G. TRAF3: a new regulator of type I interferons. Cell cycle 5, 804-807 (2006).

38 Perez de Diego, R. et al. Human TRAF3 adaptor molecule deficiency leads to impaired Toll-like receptor 3 response and susceptibility to herpes simplex encephalitis. Immunity 33, 400-411, doi:10.1016/j.immuni.2010.08.014 (2010).

39 Xu, Y., Cheng, G. & Baltimore, D. Targeted disruption of TRAF3 leads to postnatal lethality and defective T-dependent immune responses. Immunity 5, 407-415 (1996).

40 Gardam, S., Sierro, F., Basten, A., Mackay, F. & Brink, R. TRAF2 and TRAF3 signal adapters act cooperatively to control the maturation and survival signals delivered to B cells by the BAFF receptor. Immunity 28, 391-401, doi:10.1016/j.immuni.2008.01.009 (2008).

187

41 Miller, J. P., Stadanlick, J. E. & Cancro, M. P. Space, selection, and surveillance: setting boundaries with BLyS. Journal of immunology 176, 6405-6410 (2006).

42 Nagel, I. et al. Biallelic inactivation of TRAF3 in a subset of B-cell lymphomas with interstitial del(14)(q24.1q32.33). Leukemia 23, 2153-2155 (2009).

43 Braggio, E. et al. Identification of copy number abnormalities and inactivating mutations in two negative regulators of nuclear factor-kappaB signaling pathways in Waldenstrom's macroglobulinemia. Cancer research 69, 3579-3588 (2009).

44 Miyamoto, A. et al. Increased proliferation of B cells and auto-immunity in mice lacking protein kinase Cdelta. Nature 416, 865-869, doi:10.1038/416865a (2002).

45 Mecklenbrauker, I., Saijo, K., Zheng, N. Y., Leitges, M. & Tarakhovsky, A. Protein kinase Cdelta controls self-antigen-induced B-cell tolerance. Nature 416, 860-865, doi:10.1038/416860a (2002).

46 Martins, J. O. Can PKCdelta be a novel therapeutic target? Journal of leukocyte biology 89, 1-2, doi:10.1189/jlb.0810452 (2011).

47 Ashton-Rickardt, P. G. et al. MCC, a candidate familial polyposis gene in 5q.21, shows frequent allele loss in colorectal and lung cancer. Oncogene 6, 1881-1886 (1991).

48 Morita, R., Hikiji, K., Yamakawa, K. & Nakamura, Y. [Tumor suppressor genes associated with development of human renal cell carcinoma]. Nihon rinsho. Japanese journal of clinical medicine 50, 3100-3105 (1992).

49 Kohonen-Corish, M. R. et al. Promoter methylation of the mutated in colorectal cancer gene is a frequent early event in colorectal cancer. Oncogene 26, 4435- 4441 (2007).

50 Matsumine, A. et al. MCC, a cytoplasmic protein that blocks cell cycle progression from the G0/G1 to S phase. The Journal of biological chemistry 271, 10341-10346 (1996).

51 Pangon, L. et al. The "Mutated in Colorectal Cancer" Protein Is a Novel Target of the UV-Induced DNA Damage Checkpoint. Genes & cancer 1, 917-926, doi:10.1177/1947601910388937 (2010).

52 Wunderle, V. M., Critcher, R., Ashworth, A. & Goodfellow, P. N. Cloning and characterization of SOX5, a new member of the human SOX gene family. Genomics 36, 354-358, doi:10.1006/geno.1996.0474 (1996).

53 Lefebvre, V., Li, P. & de Crombrugghe, B. A new long form of Sox5 (L-Sox5), Sox6 and Sox9 are coexpressed in chondrogenesis and cooperatively activate the type II collagen gene. The EMBO journal 17, 5718-5733, doi:10.1093/emboj/17.19.5718 (1998).

188

54 Lefebvre, V., Behringer, R. R. & de Crombrugghe, B. L-Sox5, Sox6 and Sox9 control essential steps of the chondrocyte differentiation pathway. Osteoarthritis and cartilage / OARS, Osteoarthritis Research Society 9 Suppl A, S69-75 (2001).

55 Ikeda, T. et al. Identification and characterization of the human long form of Sox5 (L-SOX5) gene. Gene 298, 59-68 (2002).

56 Smits, P. & Lefebvre, V. Sox5 and Sox6 are required for notochord extracellular matrix sheath formation, notochord cell survival and development of the nucleus pulposus of intervertebral discs. Development 130, 1135-1148 (2003).

57 Ikeda, T. et al. The combination of SOX5, SOX6, and SOX9 (the SOX trio) provides signals sufficient for induction of permanent cartilage. Arthritis and rheumatism 50, 3561-3573, doi:10.1002/art.20611 (2004).

58 Kinzler, K. W. et al. Identification of a gene located at chromosome 5q21 that is mutated in colorectal cancers. Science 251, 1366-1370 (1991).

59 Nishisho, I. et al. Mutations of chromosome 5q21 genes in FAP and colorectal cancer patients. Science 253, 665-669 (1991).

60 Groden, J. et al. Identification and characterization of the familial adenomatous polyposis coli gene. Cell 66, 589-600 (1991).

61 Fukuyama, R. et al. Mutated in colorectal cancer, a putative tumor suppressor for serrated colorectal cancer, selectively represses beta-catenin-dependent transcription. Oncogene 27, 6044-6055 (2008).

62 Fu, X., Li, L. & Peng, Y. Wnt signalling pathway in the serrated neoplastic pathway of the colorectum: possible roles and epigenetic regulatory mechanisms. Journal of clinical pathology 65, 675-679, doi:10.1136/jclinpath-2011-200602 (2012).

63 Li, L. et al. Wnt signaling pathway is activated in right colon serrated polyps correlating to specific molecular form of beta-catenin. Human pathology 44, 1079- 1088, doi:10.1016/j.humpath.2012.09.013 (2013).

64 Young, T. et al. Mutated in colorectal cancer (Mcc), a candidate tumor suppressor, is dynamically expressed during mouse embryogenesis. Developmental dynamics : an official publication of the American Association of Anatomists 240, 2166-2174, doi:10.1002/dvdy.22712 (2011).

65 Heyer, J., Yang, K., Lipkin, M., Edelmann, W. & Kucherlapati, R. Mouse models for colorectal cancer. Oncogene 18, 5325-5333, doi:10.1038/sj.onc.1203036 (1999).

66 Starr, T. K. et al. A transposon-based genetic screen in mice identifies genes altered in colorectal cancer. Science 323, 1747-1750, doi:10.1126/science.1163040 (2009).

189

67 Sigglekow, N. D. et al. Mutated in colorectal cancer protein modulates the NFkappaB pathway. Anticancer research 32, 73-79 (2012).

68 Arnaud, C. et al. MCC, a new interacting protein for Scrib, is required for cell migration in epithelial cells. FEBS letters 583, 2326-2332 (2009).

69 Pangon, L. et al. The PDZ-binding motif of MCC is phosphorylated at position -1 and controls lamellipodia formation in colon epithelial cells. Biochimica et biophysica acta 1823, 1058-1067, doi:10.1016/j.bbamcr.2012.03.011 (2012).

70 Senda, T., Matsumine, A., Yanai, H. & Akiyama, T. Localization of MCC (mutated in colorectal cancer) in various tissues of mice and its involvement in cell differentiation. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society 47, 1149-1158 (1999).

71 Edwards, S. K. et al. N-benzyladriamycin-14-valerate (AD 198) exhibits potent anti- tumor activity on TRAF3-deficient mouse B lymphoma and human multiple myeloma. BMC cancer 13, 481, doi:10.1186/1471-2407-13-481 (2013).

72 Edwards, S. K., Desai, A., Liu, Y., Moore, C. R. & Xie, P. Expression and function of a novel isoform of Sox5 in malignant B cells. Leukemia research 38, 393-401, doi:10.1016/j.leukres.2013.12.016 (2014).

73 Du, P., Kibbe, W. A. & Lin, S. M. nuID: a universal naming scheme of oligonucleotides for illumina, affymetrix, and other microarrays. Biology direct 2, 16, doi:10.1186/1745-6150-2-16 (2007).

74 Du, P., Kibbe, W. A. & Lin, S. M. lumi: a pipeline for processing Illumina microarray. Bioinformatics 24, 1547-1548, doi:10.1093/bioinformatics/btn224 (2008).

75 Smyth, G. Limma: linear models for microarray data. (Springer, 2005).

76 Xie, P. et al. Enhanced Toll-like receptor (TLR) responses of TNFR-associated factor 3 (TRAF3)-deficient B lymphocytes. Journal of leukocyte biology 90, 1149- 1157, doi:10.1189/jlb.0111044 (2011).

77 Urist, M., Tanaka, T., Poyurovsky, M. V. & Prives, C. p73 induction after DNA damage is regulated by checkpoint kinases Chk1 and Chk2. Genes & development 18, 3041-3054, doi:10.1101/gad.1221004 (2004).

78 Li, J. et al. Nuclear accumulation of HDAC4 in ATM deficiency promotes neurodegeneration in ataxia telangiectasia. Nature medicine 18, 783-790, doi:10.1038/nm.2709 (2012).

79 Zhou, S. et al. MyD88 intrinsically regulates CD4 T-cell responses. Journal of virology 83, 1625-1634, doi:10.1128/JVI.01770-08 (2009).

80 Li, Y., Franklin, S., Zhang, M. J. & Vondriska, T. M. Highly efficient purification of protein complexes from mammalian cells using a novel streptavidin-binding peptide and hexahistidine tandem tag system: application to Bruton's tyrosine

190

kinase. Protein science : a publication of the Protein Society 20, 140-149, doi:10.1002/pro.546 (2011).

81 Porter, J. F., Vavassori, S. & Covey, L. R. A polypyrimidine tract-binding protein- dependent pathway of mRNA stability initiates with CpG activation of primary B cells. Journal of immunology 181, 3336-3345 (2008).

82 Xie, P., Kraus, Z. J., Stunz, L. L., Liu, Y. & Bishop, G. A. TNF receptor-associated factor 3 is required for T cell-mediated immunity and TCR/CD28 signaling. Journal of immunology 186, 143-155, doi:10.4049/jimmunol.1000290 (2011).

83 Wan, Q. et al. Reticulon 3 mediates Bcl-2 accumulation in mitochondria in response to endoplasmic reticulum stress. Apoptosis : an international journal on programmed cell death 12, 319-328, doi:10.1007/s10495-006-0574-y (2007).

84 Simmen, T. et al. PACS-2 controls endoplasmic reticulum-mitochondria communication and Bid-mediated apoptosis. The EMBO journal 24, 717-729, doi:10.1038/sj.emboj.7600559 (2005).

85 Darios, F. et al. Parkin prevents mitochondrial swelling and cytochrome c release in mitochondria-dependent cell death. Human molecular genetics 12, 517-526 (2003).

86 Vento, M. T. et al. Praf2 is a novel Bcl-xL/Bcl-2 interacting protein with the ability to modulate survival of cancer cells. PloS one 5, e15636, doi:10.1371/journal.pone.0015636 (2010).

87 Khan, Z. et al. Accurate proteome-wide protein quantification from high-resolution 15N mass spectra. Genome biology 12, R122, doi:10.1186/gb-2011-12-12-r122 (2011).

88 Ying, W. et al. Highly efficient and selective enrichment of peptide subsets combining fluorous chemistry with reversed-phase chromatography. Rapid communications in mass spectrometry : RCM 23, 4019-4030, doi:10.1002/rcm.4343 (2009).

89 Basso, K. et al. Reverse engineering of regulatory networks in human B cells. Nature genetics 37, 382-390, doi:10.1038/ng1532 (2005).

90 Zhan, F. et al. Gene-expression signature of benign monoclonal gammopathy evident in multiple myeloma is linked to good prognosis. Blood 109, 1692-1700, doi:10.1182/blood-2006-07-037077 (2007).

91 Zhan, F. et al. Global gene expression profiling of multiple myeloma, monoclonal gammopathy of undetermined significance, and normal bone marrow plasma cells. Blood 99, 1745-1757 (2002).

92 Zhan, F. et al. Gene expression profiling of human plasma cell differentiation and classification of multiple myeloma based on similarities to distinct stages of late-

191

stage B-cell development. Blood 101, 1128-1140, doi:10.1182/blood-2002-06- 1737 (2003).

93 Lu, K. T., Dryer, R. L., Song, C. & Covey, L. R. Maintenance of the CD40-related immunodeficient response in hyper-IgM B cells immortalized with a LMP1- regulated mini-EBV. Journal of leukocyte biology 78, 620-629, doi:10.1189/jlb.0305159 (2005).

94 Bouwmeester, T. et al. A physical and functional map of the human TNF-alpha/NF- kappa B signal transduction pathway. Nature cell biology 6, 97-105, doi:10.1038/ncb1086 (2004).

95 Shukla, R. et al. Endogenous retrotransposition activates oncogenic pathways in hepatocellular carcinoma. Cell 153, 101-111, doi:10.1016/j.cell.2013.02.032 (2013).

96 Ewing, R. M. et al. Large-scale mapping of human protein-protein interactions by mass spectrometry. Molecular systems biology 3, 89, doi:10.1038/msb4100134 (2007).

97 Weaver, A. N. & Yang, E. S. Beyond DNA Repair: Additional Functions of PARP- 1 in Cancer. Frontiers in oncology 3, 290, doi:10.3389/fonc.2013.00290 (2013).

98 Thuaud, F., Ribeiro, N., Nebigil, C. G. & Desaubry, L. Prohibitin ligands in cell death and survival: mode of action and therapeutic potential. Chemistry & biology 20, 316-331, doi:10.1016/j.chembiol.2013.02.006 (2013).

99 Cohen-Armon, M. PARP-1 activation in the ERK signaling pathway. Trends in pharmacological sciences 28, 556-560, doi:10.1016/j.tips.2007.08.005 (2007).

100 Artal-Sanz, M. & Tavernarakis, N. Prohibitin and mitochondrial biology. Trends in endocrinology and metabolism: TEM 20, 394-401, doi:10.1016/j.tem.2009.04.004 (2009).

101 Szklarczyk, D. et al. The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored. Nucleic acids research 39, D561-568, doi:10.1093/nar/gkq973 (2011).

102 D'Amico, D., Carbone, D. P., Johnson, B. E., Meltzer, S. J. & Minna, J. D. Polymorphic sites within the MCC and APC loci reveal very frequent loss of heterozygosity in human small cell lung cancer. Cancer research 52, 1996-1999 (1992).

103 Sud, R., Talbot, I. C. & Delhanty, J. D. Infrequent alterations of the APC and MCC genes in gastric cancers from British patients. British journal of cancer 74, 1104- 1108 (1996).

104 Huang, Y. et al. Loss of heterozygosity involves multiple tumor suppressor genes in human esophageal cancers. Cancer research 52, 6525-6530 (1992).

192

105 Guichard, C. et al. Integrated analysis of somatic mutations and focal copy-number changes identifies key genes and pathways in hepatocellular carcinoma. Nature genetics 44, 694-698, doi:10.1038/ng.2256 (2012).

106 Sakamaki, J., Daitoku, H., Yoshimochi, K., Miwa, M. & Fukamizu, A. Regulation of FOXO1-mediated transcription and cell proliferation by PARP-1. Biochemical and biophysical research communications 382, 497-502, doi:10.1016/j.bbrc.2009.03.022 (2009).

107 Harley, M. E., Allan, L. A., Sanderson, H. S. & Clarke, P. R. Phosphorylation of Mcl-1 by CDK1-cyclin B1 initiates its Cdc20-dependent destruction during mitotic arrest. The EMBO journal 29, 2407-2420, doi:10.1038/emboj.2010.112 (2010).

108 Sakurikar, N., Eichhorn, J. M. & Chambers, T. C. Cyclin-dependent kinase-1 (Cdk1)/cyclin B1 dictates cell fate after mitotic arrest via phosphoregulation of antiapoptotic Bcl-2 proteins. The Journal of biological chemistry 287, 39193- 39204, doi:10.1074/jbc.M112.391854 (2012).

109 Zhou, T. B. & Qin, Y. H. Signaling pathways of prohibitin and its role in diseases. Journal of receptor and signal transduction research 33, 28-36, doi:10.3109/10799893.2012.752006 (2013).

110 Kategaya, L. S. et al. Casein kinase 1 proteomics reveal prohibitin 2 function in molecular clock. PloS one 7, e31987, doi:10.1371/journal.pone.0031987 (2012).

111 Marampon, F., Ciccarelli, C. & Zani, B. M. Down-regulation of c-Myc following MEK/ERK inhibition halts the expression of malignant phenotype in rhabdomyosarcoma and in non muscle-derived human tumors. Molecular cancer 5, 31, doi:10.1186/1476-4598-5-31 (2006).

112 Rossi, D. et al. Alteration of BIRC3 and multiple other NF-{kappa}B pathway genes in splenic marginal zone lymphoma. Blood 118, 4930-4934 (2011).

113 Lefebvre, V. The SoxD transcription factors--Sox5, Sox6, and Sox13--are key cell fate modulators. The international journal of biochemistry & cell biology 42, 429- 432, doi:10.1016/j.biocel.2009.07.016 (2010).

114 Kiselak, E. A. et al. Transcriptional regulation of an axonemal central apparatus gene, sperm-associated antigen 6, by a SRY-related high mobility group transcription factor, S-SOX5. The Journal of biological chemistry 285, 30496- 30505, doi:10.1074/jbc.M110.121590 (2010).

115 Denny, P., Swift, S., Connor, F. & Ashworth, A. An SRY-related gene expressed during spermatogenesis in the mouse encodes a sequence-specific DNA-binding protein. The EMBO journal 11, 3705-3712 (1992).

116 Hiraoka, Y., Ogawa, M., Sakai, Y., Kido, S. & Aiso, S. The mouse Sox5 gene encodes a protein containing the leucine zipper and the Q box. Biochimica et biophysica acta 1399, 40-46 (1998).

193

117 Hersh, C. P. et al. SOX5 is a candidate gene for chronic obstructive pulmonary disease susceptibility and is necessary for lung development. American journal of respiratory and critical care medicine 183, 1482-1489, doi:10.1164/rccm.201010- 1751OC (2011).

118 Smits, P. et al. The transcription factors L-Sox5 and Sox6 are essential for cartilage formation. Developmental cell 1, 277-290 (2001).

119 Dy, P. et al. Synovial joint morphogenesis requires the chondrogenic action of Sox5 and Sox6 in growth plate and articular cartilage. Developmental biology 341, 346-359, doi:10.1016/j.ydbio.2010.02.024 (2010).

120 Martinez-Morales, P. L., Quiroga, A. C., Barbas, J. A. & Morales, A. V. SOX5 controls cell cycle progression in neural progenitors by interfering with the WNT- beta-catenin pathway. EMBO reports 11, 466-472, doi:10.1038/embor.2010.61 (2010).

121 Stolt, C. C. et al. SoxD proteins influence multiple stages of oligodendrocyte development and modulate SoxE protein function. Developmental cell 11, 697- 709, doi:10.1016/j.devcel.2006.08.011 (2006).

122 Stolt, C. C., Lommes, P., Hillgartner, S. & Wegner, M. The transcription factor Sox5 modulates Sox10 function during melanocyte development. Nucleic acids research 36, 5427-5440, doi:10.1093/nar/gkn527 (2008).

123 Kwan, K. Y. et al. SOX5 postmitotically regulates migration, postmigratory differentiation, and projections of subplate and deep-layer neocortical neurons. Proceedings of the National Academy of Sciences of the United States of America 105, 16021-16026, doi:10.1073/pnas.0806791105 (2008).

124 Lai, T. et al. SOX5 controls the sequential generation of distinct corticofugal neuron subtypes. Neuron 57, 232-247, doi:10.1016/j.neuron.2007.12.023 (2008).

125 Ehrhardt, G. R. et al. Discriminating gene expression profiles of memory B cell subpopulations. The Journal of experimental medicine 205, 1807-1817, doi:10.1084/jem.20072682 (2008).

126 Charles, E. D. et al. Clonal B cells in patients with hepatitis C virus-associated mixed cryoglobulinemia contain an expanded anergic CD21low B-cell subset. Blood 117, 5425-5437, doi:10.1182/blood-2010-10-312942 (2011).

127 Storlazzi, C. T. et al. Upregulation of the SOX5 by promoter swapping with the P2RY8 gene in primary splenic follicular lymphoma. Leukemia 21, 2221-2225, doi:10.1038/sj.leu.2404784 (2007).

128 Xie, P. & Bishop, G. A. Roles of TNF receptor-associated factor 3 in signaling to B lymphocytes by carboxyl-terminal activating regions 1 and 2 of the EBV-encoded oncoprotein latent membrane protein 1. Journal of immunology 173, 5546-5555 (2004).

194

129 Dull, T. et al. A third-generation lentivirus vector with a conditional packaging system. Journal of virology 72, 8463-8471 (1998).

130 Edwards, S. et al. Mutated in colorectal cancer (MCC) is a novel oncogene in B lymphocytes. Journal of hematology & oncology 7, 56, doi:10.1186/s13045-014- 0056-6 (2014).

131 Tchougounova, E. et al. Sox5 can suppress platelet-derived growth factor B- induced glioma development in Ink4a-deficient mice through induction of acute cellular senescence. Oncogene 28, 1537-1548, doi:10.1038/onc.2009.9 (2009).

132 Lamb, A. N. et al. Haploinsufficiency of SOX5 at 12p12.1 is associated with developmental delays with prominent language delay, behavior problems, and mild dysmorphic features. Human mutation 33, 728-740, doi:10.1002/humu.22037 (2012).

133 Schanze, I. et al. Haploinsufficiency of SOX5, a member of the SOX (SRY-related HMG-box) family of transcription factors is a cause of intellectual disability. European journal of medical genetics 56, 108-113, doi:10.1016/j.ejmg.2012.11.001 (2013).

134 Lee, R. W., Bodurtha, J., Cohen, J., Fatemi, A. & Batista, D. Deletion 12p12 involving SOX5 in two children with developmental delay and dysmorphic features. Pediatric neurology 48, 317-320, doi:10.1016/j.pediatrneurol.2012.12.013 (2013).

135 Ma, S. et al. DNA fingerprinting tags novel altered chromosomal regions and identifies the involvement of SOX5 in the progression of prostate cancer. International journal of cancer. Journal international du cancer 124, 2323-2332, doi:10.1002/ijc.24243 (2009).

136 Zafarana, G. et al. Coamplification of DAD-R, SOX5, and EKI1 in human testicular seminomas, with specific overexpression of DAD-R, correlates with reduced levels of apoptosis and earlier clinical manifestation. Cancer research 62, 1822-1831 (2002).

137 Huang, D. Y. et al. Transcription factor SOX-5 enhances nasopharyngeal carcinoma progression by down-regulating SPARC gene expression. The Journal of pathology 214, 445-455, doi:10.1002/path.2299 (2008).

138 Riker, A. I. et al. The gene expression profiles of primary and metastatic melanoma yields a transition point of tumor progression and metastasis. BMC medical genomics 1, 13, doi:10.1186/1755-8794-1-13 (2008).

139 Ueda, R., Yoshida, K., Kawase, T., Kawakami, Y. & Toda, M. Preferential expression and frequent IgG responses of a tumor antigen, SOX5, in glioma patients. International journal of cancer. Journal international du cancer 120, 1704- 1711, doi:10.1002/ijc.22472 (2007).

195

140 Schlierf, B. et al. Expression of SoxE and SoxD genes in human gliomas. Neuropathology and applied neurobiology 33, 621-630, doi:10.1111/j.1365- 2990.2007.00881.x (2007).

141 Lenburg, M. E., Sinha, A., Faller, D. V. & Denis, G. V. Tumor-specific and proliferation-specific gene expression typifies murine transgenic B cell lymphomagenesis. The Journal of biological chemistry 282, 4803-4811, doi:10.1074/jbc.M605870200 (2007).

142 Carrasco, D. R. et al. The differentiation and stress response factor XBP-1 drives multiple myeloma pathogenesis. Cancer Cell 11, 349-360, doi:10.1016/j.ccr.2007.02.015 (2007).

143 Ramaswamy, S. et al. Multiclass cancer diagnosis using tumor gene expression signatures. Proceedings of the National Academy of Sciences of the United States of America 98, 15149-15154, doi:10.1073/pnas.211566398 (2001).

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

145 Dickens, N. J. et al. Homozygous deletion mapping in myeloma samples identifies genes and an expression signature relevant to pathogenesis and outcome. Clinical cancer research : an official journal of the American Association for Cancer Research 16, 1856-1864, doi:10.1158/1078-0432.CCR-09-2831 (2010).

146 Rinaldi, A. et al. Genome-wide DNA profiling of marginal zone lymphomas identifies subtype-specific lesions with an impact on the clinical outcome. Blood 117, 1595-1604, doi:10.1182/blood-2010-01-264275 (2011).

147 Wang, C., Tai, Y., Lisanti, M. P. & Liao, D. J. c-Myc induction of programmed cell death may contribute to carcinogenesis: a perspective inspired by several concepts of chemical carcinogenesis. Cancer biology & therapy 11, 615-626 (2011).

148 Campaner, S. et al. Cdk2 suppresses cellular senescence induced by the c-myc oncogene. Nature cell biology 12, 54-59; sup pp 51-14, doi:10.1038/ncb2004 (2010).

149 Hoffman, B. & Liebermann, D. A. Apoptotic signaling by c-MYC. Oncogene 27, 6462-6472, doi:10.1038/onc.2008.312 (2008).

150 Hattori, T. et al. Transcriptional regulation of chondrogenesis by coactivator Tip60 via chromatin association with Sox9 and Sox5. Nucleic acids research 36, 3011- 3024, doi:10.1093/nar/gkn150 (2008).

151 Shim, S., Kwan, K. Y., Li, M., Lefebvre, V. & Sestan, N. Cis-regulatory control of corticospinal system development and evolution. Nature 486, 74-79, doi:10.1038/nature11094 (2012).

196

152 Mackay, F. et al. Mice transgenic for BAFF develop lymphocytic disorders along with autoimmune manifestations. The Journal of experimental medicine 190, 1697- 1710 (1999).

153 Zhang, B., Wang, Z., Li, T., Tsitsikov, E. N. & Ding, H. F. NF-kappaB2 mutation targets TRAF1 to induce lymphomagenesis. Blood 110, 743-751, doi:10.1182/blood-2006-11-058446 (2007).

154 Mecklenbrauker, I., Kalled, S. L., Leitges, M., Mackay, F. & Tarakhovsky, A. Regulation of B-cell survival by BAFF-dependent PKCdelta-mediated nuclear signalling. Nature 431, 456-461, doi:10.1038/nature02955 (2004).

155 Lothstein, L., Savranskaya, L., Barrett, C. M., Israel, M. & Sweatman, T. W. N- benzyladriamycin-14-valerate (AD 198) activates protein kinase C-delta holoenzyme to trigger mitochondrial depolarization and cytochrome c release independently of permeability transition pore opening and Ca2+ influx. Anti-cancer drugs 17, 495-502 (2006).

156 Lothstein, L., Savranskaya, L. & Sweatman, T. W. N-Benzyladriamycin-14- valerate (AD 198) cytotoxicty circumvents Bcr-Abl anti-apoptotic signaling in human leukemia cells and also potentiates imatinib cytotoxicity. Leukemia research 31, 1085-1095, doi:10.1016/j.leukres.2006.11.003 (2007).

157 Diaz Bessone, M. I. et al. Involvement of PKC delta (PKCdelta) in the resistance against different doxorubicin analogs. Breast cancer research and treatment 126, 577-587, doi:10.1007/s10549-010-0956-2 (2011).

158 Hampson, P. et al. PEP005, a selective small-molecule activator of protein kinase C, has potent antileukemic activity mediated via the delta isoform of PKC. Blood 106, 1362-1368, doi:10.1182/blood-2004-10-4117 (2005).

159 Serova, M. et al. Effects of protein kinase C modulation by PEP005, a novel ingenol angelate, on mitogen-activated protein kinase and phosphatidylinositol 3-kinase signaling in cancer cells. Molecular cancer therapeutics 7, 915-922, doi:10.1158/1535-7163.MCT-07-2060 (2008).

160 Hampson, P. et al. Kinetics of ERK1/2 activation determine sensitivity of acute myeloid leukaemia cells to the induction of apoptosis by the novel small molecule ingenol 3-angelate (PEP005). Apoptosis : an international journal on programmed cell death 15, 946-955, doi:10.1007/s10495-010-0507-7 (2010).

161 van de Loosdrecht, A. A., Beelen, R. H., Ossenkoppele, G. J., Broekhoven, M. G. & Langenhuijsen, M. M. A tetrazolium-based colorimetric MTT assay to quantitate human monocyte mediated cytotoxicity against leukemic cells from cell lines and patients with acute myeloid leukemia. Journal of immunological methods 174, 311- 320 (1994).

162 Ghoul, A. et al. Epithelial-to-mesenchymal transition and resistance to ingenol 3- angelate, a novel protein kinase C modulator, in colon cancer cells. Cancer research 69, 4260-4269, doi:10.1158/0008-5472.CAN-08-2837 (2009).

197

163 Morse, H. C., 3rd et al. Bethesda proposals for classification of lymphoid neoplasms in mice. Blood 100, 246-258 (2002).

164 Olsnes, A. M. et al. The protein kinase C agonist PEP005 increases NF-kappaB expression, induces differentiation and increases constitutive chemokine release by primary acute myeloid leukaemia cells. British journal of haematology 145, 761- 774, doi:10.1111/j.1365-2141.2009.07691.x (2009).

165 Bilyeu, J. D. et al. Circumvention of nuclear factor kappaB-induced chemoresistance by cytoplasmic-targeted anthracyclines. Molecular pharmacology 65, 1038-1047, doi:10.1124/mol.65.4.1038 (2004).

166 Basu, A. & Pal, D. Two faces of protein kinase Cdelta: the contrasting roles of PKCdelta in cell survival and cell death. TheScientificWorldJournal 10, 2272-2284, doi:10.1100/tsw.2010.214 (2010).

167 Yoshida, K. PKCdelta signaling: mechanisms of DNA damage response and apoptosis. Cellular signalling 19, 892-901, doi:10.1016/j.cellsig.2007.01.027 (2007).

168 Ersvaer, E. et al. The protein kinase C agonist PEP005 (ingenol 3-angelate) in the treatment of human cancer: a balance between efficacy and toxicity. Toxins 2, 174- 194, doi:10.3390/toxins2010174 (2010).

169 Vartanian, R. et al. AP-1 regulates cyclin D1 and c-MYC transcription in an AKT- dependent manner in response to mTOR inhibition: role of AIP4/Itch-mediated JUNB degradation. Molecular cancer research : MCR 9, 115-130, doi:10.1158/1541-7786.MCR-10-0105 (2011).

170 Park, S. S. et al. Insertion of c-Myc into Igh induces B-cell and plasma-cell neoplasms in mice. Cancer research 65, 1306-1315, doi:10.1158/0008- 5472.CAN-04-0268 (2005).

171 Chen, S., Qiong, Y. & Gardner, D. G. A role for p38 mitogen-activated protein kinase and c-myc in endothelin-dependent rat aortic smooth muscle cell proliferation. Hypertension 47, 252-258, doi:10.1161/01.HYP.0000198424.93598.6b (2006).

172 Smith, S. M., Anastasi, J., Cohen, K. S. & Godley, L. A. The impact of MYC expression in lymphoma biology: beyond Burkitt lymphoma. Blood cells, molecules & diseases 45, 317-323, doi:10.1016/j.bcmd.2010.08.002 (2010).

173 Ogbourne, S. M. et al. Antitumor activity of 3-ingenyl angelate: plasma membrane and mitochondrial disruption and necrotic cell death. Cancer research 64, 2833- 2839 (2004).

174 Ogbourne, S. M. et al. Proceedings of the First International Conference on PEP005. Anti-cancer drugs 18, 357-362, doi:10.1097/CAD.0b013e3280149ec5 (2007).

198

175 Challacombe, J. M. et al. Neutrophils are a key component of the antitumor efficacy of topical chemotherapy with ingenol-3-angelate. Journal of immunology 177, 8123-8132 (2006).

176 Hampson, P. et al. The anti-tumor agent, ingenol-3-angelate (PEP005), promotes the recruitment of cytotoxic neutrophils by activation of vascular endothelial cells in a PKC-delta dependent manner. Cancer immunology, immunotherapy : CII 57, 1241-1251, doi:10.1007/s00262-008-0458-9 (2008).

177 Cozzi, S. J., Parsons, P. G., Ogbourne, S. M., Pedley, J. & Boyle, G. M. Induction of senescence in diterpene ester-treated melanoma cells via protein kinase C- dependent hyperactivation of the mitogen-activated protein kinase pathway. Cancer research 66, 10083-10091, doi:10.1158/0008-5472.CAN-06-0348 (2006).

178 Hofmann, P. A. et al. N-Benzyladriamycin-14-valerate (AD 198): a non-cardiotoxic anthracycline that is cardioprotective through PKC-epsilon activation. The Journal of pharmacology and experimental therapeutics 323, 658-664, doi:10.1124/jpet.107.126110 (2007).

179 Benhadji, K. A. et al. Antiproliferative activity of PEP005, a novel ingenol angelate that modulates PKC functions, alone and in combination with cytotoxic agents in human colon cancer cells. British journal of cancer 99, 1808-1815, doi:10.1038/sj.bjc.6604642 (2008).

180 Sweatman, T. W., Seshadri, R. & Israel, M. Pharmacology of N-benzyladriamycin- 14-valerate in the rat. Cancer chemotherapy and pharmacology 43, 419-426, doi:10.1007/s002800050917 (1999).

181 Roaten, J. B. et al. Molecular models of N-benzyladriamycin-14-valerate (AD 198) in complex with the phorbol ester-binding C1b domain of protein kinase C-delta. Journal of medicinal chemistry 44, 1028-1034 (2001).

182 Barrett, C. M. et al. Novel extranuclear-targeted anthracyclines override the antiapoptotic functions of Bcl-2 and target protein kinase C pathways to induce apoptosis. Molecular cancer therapeutics 1, 469-481 (2002).

183 Roaten, J. B. et al. Interaction of the novel anthracycline antitumor agent N- benzyladriamycin-14-valerate with the C1-regulatory domain of protein kinase C: structural requirements, isoform specificity, and correlation with drug cytotoxicity. Molecular cancer therapeutics 1, 483-492 (2002).

184 Ramsay, G., Evan, G. I. & Bishop, J. M. The protein encoded by the human proto- oncogene c-myc. Proceedings of the National Academy of Sciences of the United States of America 81, 7742-7746 (1984).

185 Gregory, M. A. & Hann, S. R. c-Myc proteolysis by the ubiquitin-proteasome pathway: stabilization of c-Myc in Burkitt's lymphoma cells. Molecular and cellular biology 20, 2423-2435 (2000).

199

186 Lameh, J., Chuang, L. F., Israel, M. & Chuang, R. Y. Mechanistic studies on N- benzyladriamycin-14-valerate (AD 198), a highly lipophilic alkyl adriamycin analog. Anticancer research 8, 689-693 (1988).

187 Harstrick, A. et al. Activity of N-benzyl-adriamycin-14-valerate (AD198), a new anthracycline derivate, in multidrug resistant human ovarian and breast carcinoma cell lines. Anti-cancer drugs 6, 681-685 (1995).

188 Ganapathi, R., Grabowski, D., Sweatman, T. W., Seshadri, R. & Israel, M. N- benzyladriamycin-14-valerate versus progressively doxorubicin-resistant murine tumours: cellular pharmacology and characterisation of cross-resistance in vitro and in vivo. British journal of cancer 60, 819-826 (1989).

189 Lothstein, L., Wright, H. M., Sweatman, T. W. & Israel, M. N-benzyladriamycin-14- valerate and drug resistance: correlation of anthracycline structural modification with intracellular accumulation and distribution in multidrug resistant cells. Oncology research 4, 341-347 (1992).

190 Cai, C., Lothstein, L., Morrison, R. R. & Hofmann, P. A. Protection from doxorubicin-induced cardiomyopathy using the modified anthracycline N- benzyladriamycin-14-valerate (AD 198). The Journal of pharmacology and experimental therapeutics 335, 223-230, doi:10.1124/jpet.110.167965 (2010).

191 Fennell, D. A., Chacko, A. & Mutti, L. BCL-2 family regulation by the 20S proteasome inhibitor bortezomib. Oncogene 27, 1189-1197, doi:10.1038/sj.onc.1210744 (2008).

192 Barr, P., Fisher, R. & Friedberg, J. The role of bortezomib in the treatment of lymphoma. Cancer investigation 25, 766-775, doi:10.1080/07357900701579570 (2007).

193 Leonard, J. P., Furman, R. R. & Coleman, M. Proteasome inhibition with bortezomib: a new therapeutic strategy for non-Hodgkin's lymphoma. International journal of cancer. Journal international du cancer 119, 971-979, doi:10.1002/ijc.21805 (2006).

194 Reddy, N. & Czuczman, M. S. Enhancing activity and overcoming chemoresistance in hematologic malignancies with bortezomib: preclinical mechanistic studies. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO 21, 1756-1764, doi:10.1093/annonc/mdq009 (2010).

195 Mujtaba, T. & Dou, Q. P. Advances in the understanding of mechanisms and therapeutic use of bortezomib. Discovery medicine 12, 471-480 (2011).

196 Painuly, U. & Kumar, S. Efficacy of bortezomib as first-line treatment for patients with multiple myeloma. Clinical Medicine Insights. Oncology 7, 53-73, doi:10.4137/CMO.S7764 (2013).

200

197 Driscoll, J. J., Burris, J. & Annunziata, C. M. Targeting the proteasome with bortezomib in multiple myeloma: update on therapeutic benefit as an upfront single agent, induction regimen for stem-cell transplantation and as maintenance therapy. American journal of therapeutics 19, 133-144, doi:10.1097/MJT.0b013e3181ff7a9e (2012).

198 Vallabhapurapu, S. et al. Nonredundant and complementary functions of TRAF2 and TRAF3 in a ubiquitination cascade that activates NIK-dependent alternative NF-kappaB signaling. Nature immunology 9, 1364-1370, doi:10.1038/ni.1678 (2008).

199 Zarnegar, B. J. et al. Noncanonical NF-kappaB activation requires coordinated assembly of a regulatory complex of the adaptors cIAP1, cIAP2, TRAF2 and TRAF3 and the kinase NIK. Nature immunology 9, 1371-1378, doi:10.1038/ni.1676 (2008).

200 Ma, M. H. et al. The proteasome inhibitor PS-341 markedly enhances sensitivity of multiple myeloma tumor cells to chemotherapeutic agents. Clinical cancer research : an official journal of the American Association for Cancer Research 9, 1136-1144 (2003).

201 Hideshima, T. et al. Bortezomib induces canonical nuclear factor-kappaB activation in multiple myeloma cells. Blood 114, 1046-1052, doi:10.1182/blood- 2009-01-199604 (2009).

202 Markovina, S. et al. Bortezomib-resistant nuclear factor-kappaB activity in multiple myeloma cells. Molecular cancer research : MCR 6, 1356-1364, doi:10.1158/1541- 7786.MCR-08-0108 (2008).

203 Matta, H. & Chaudhary, P. M. The proteasome inhibitor bortezomib (PS-341) inhibits growth and induces apoptosis in primary effusion lymphoma cells. Cancer biology & therapy 4, 77-82 (2005).

204 Cavo, M. Proteasome inhibitor bortezomib for the treatment of multiple myeloma. Leukemia 20, 1341-1352, doi:10.1038/sj.leu.2404278 (2006).

205 Skrott, Z. & Cvek, B. Linking the activity of bortezomib in multiple myeloma and autoimmune diseases. Critical reviews in oncology/hematology 92, 61-70, doi:10.1016/j.critrevonc.2014.05.003 (2014).

206 Kapoor, P., Ramakrishnan, V. & Rajkumar, S. V. Bortezomib combination therapy in multiple myeloma. Seminars in hematology 49, 228-242, doi:10.1053/j.seminhematol.2012.04.010 (2012).

207 Pei, X. Y., Dai, Y. & Grant, S. Synergistic induction of oxidative injury and apoptosis in human multiple myeloma cells by the proteasome inhibitor bortezomib and histone deacetylase inhibitors. Clinical cancer research : an official journal of the American Association for Cancer Research 10, 3839-3852, doi:10.1158/1078- 0432.CCR-03-0561 (2004).

201

208 Gomez-Bougie, P. et al. Noxa up-regulation and Mcl-1 cleavage are associated to apoptosis induction by bortezomib in multiple myeloma. Cancer research 67, 5418- 5424, doi:10.1158/0008-5472.CAN-06-4322 (2007).

209 Podar, K. et al. A pivotal role for Mcl-1 in Bortezomib-induced apoptosis. Oncogene 27, 721-731, doi:10.1038/sj.onc.1210679 (2008).

210 Perez-Galan, P., Roue, G., Villamor, N., Campo, E. & Colomer, D. The BH3- mimetic GX15-070 synergizes with bortezomib in mantle cell lymphoma by enhancing Noxa-mediated activation of Bak. Blood 109, 4441-4449, doi:10.1182/blood-2006-07-034173 (2007).

211 Perez-Galan, P. et al. The proteasome inhibitor bortezomib induces apoptosis in mantle-cell lymphoma through generation of ROS and Noxa activation independent of p53 status. Blood 107, 257-264, doi:10.1182/blood-2005-05-2091 (2006).

212 Maynadier, M. et al. Roles of estrogen receptor and p21(Waf1) in bortezomib- induced growth inhibition in human breast cancer cells. Molecular cancer research : MCR 10, 1473-1481, doi:10.1158/1541-7786.MCR-12-0133 (2012).

213 Shen, Y. et al. Bortezomib induces apoptosis of endometrial cancer cells through microRNA-17-5p by targeting p21. Cell biology international 37, 1114-1121, doi:10.1002/cbin.10139 (2013).

214 Canfield, S. E., Zhu, K., Williams, S. A. & McConkey, D. J. Bortezomib inhibits docetaxel-induced apoptosis via a p21-dependent mechanism in human prostate cancer cells. Molecular cancer therapeutics 5, 2043-2050, doi:10.1158/1535- 7163.MCT-05-0437 (2006).

215 Fabre, C. et al. Dual inhibition of canonical and noncanonical NF-kappaB pathways demonstrates significant antitumor activities in multiple myeloma. Clinical cancer research : an official journal of the American Association for Cancer Research 18, 4669-4681, doi:10.1158/1078-0432.CCR-12-0779 (2012).

216 Boynton, R. F. et al. Loss of heterozygosity involving the APC and MCC genetic loci occurs in the majority of human esophageal cancers. Proceedings of the National Academy of Sciences of the United States of America 89, 3385-3388 (1992).

217 Petrelli, A. & Giordano, S. From single- to multi-target drugs in cancer therapy: when aspecificity becomes an advantage. Current medicinal chemistry 15, 422- 432 (2008).

218 Bozic, I. et al. Evolutionary dynamics of cancer in response to targeted combination therapy. eLife 2, e00747, doi:10.7554/eLife.00747 (2013).