I Calcium/Calmodulin-Dependent Protein Kinase Kinase

Calcium/Calmodulin-Dependent Protein Kinase Kinase 2 (CaMKK2) Regulates

Dendritic Cells and Myeloid Derived Suppressor Cells Development

in the Lymphoma Microenvironment

Wei Huang

Department of Pathology Duke University

Date: ______Approved:

______Nelson Chao, Co-Supervisor

______Soman Abraham, Co-Supervisor

______Benny Chen

______Luigi Racioppi

______Herman Staats

Dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Pathology in the Graduate School of Duke University 2016 i

ABSTRACT

Calcium/Calmodulin-Dependent Protein Kinase Kinase 2 (CaMKK2) Regulates

Dendritic Cells and Myeloid Derived Suppressor Cells Development

in the Lymphoma Microenvironment

Wei Huang

Department of Pathology Duke University

Date______Approved:

______Nelson Chao, Co-Supervisor

______Soman Abraham, Co-Supervisor

______Benny Chen

______Luigi Racioppi

______Herman Staats

An abstract of a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Pathology in the Graduate School of Duke University

2016 i

Abstract

Calcium (Ca2+) is a known important second messenger. Calcium/Calmodulin

(CaM) dependent protein kinase kinase 2 (CaMKK2) is a crucial kinase in the calcium signaling cascade. Activated by Ca2+/CaM, CaMKK2 can phosphorylate other CaM kinases and AMP-activated protein kinase (AMPK) to regulate cell differentiation, energy balance, metabolism and inflammation. Outside of the brain, CaMKK2 can only be detected in hematopoietic stem cells and progenitors, and in the subsets of mature myeloid cells. CaMKK2 has been noted to facilitate tumor cell proliferation in prostate cancer, breast cancer, and hepatic cancer. However, whethter CaMKK2 impacts the tumor microenvironment especially in hematopoietic malignancies remains unknown.

Due to the relevance of myeloid cells in tumor growth, we hypothesized that CaMKK2 has a critical role in the tumor microenvironment, and tested this hyopothesis in murine models of hematological and solid cancer malignancies.

We found that CaMKK2 ablation in the host suppressed the growth of E.G7 murine lymphoma, Vk*Myc myeloma and E0771 mammary cancer. The selective ablation of CaMKK2 in myeloid cells was sufficient to restrain tumor growth, of which could be reversed by CD8 cell depletion. In the lymphoma microenvironment, ablating

CaMKK2 generated less myeloid-derived suppressor cells (MDSCs) in vitro and in vivo.

Mechanistically, CaMKK2 deficient dendritic cells showed higher Major

Histocompatibility Class II (MHC II) and costimulatory factor expression, higher chemokine and IL-12 secretion when stimulated by LPS, and have higher potent in stimulating T-cell activation. AMPK, an anti-inflammatory kinase, was found as the relevant downstream target of CaMKK2 in dendritic cells. Treatment with CaMKK2 selective inhibitor STO-609 efficiently suppressed E.G7 and E0771 tumor growth, and reshaped the tumor microenvironment by attracting more immunogenic myeloid cells and infiltrated T cells.

In conclusion, we demonstrate that CaMKK2 expressed in myeloid cells is an important checkpoint in tumor microenvironment. Ablating CaMKK2 suppresses lymphoma growth by promoting myeloid cells development thereby decreasing MDSCs while enhancing the anti-tumor immune response. CaMKK2 inhibition is an innovative strategy for cancer therapy through reprogramming the tumor microenvironment.

Dedication

This work is in memory of my beloved grandparents, Dr. Huan-Qiu Huang and

Ms. Jie Zhang - for the endless love from earth to heaven.

Contents

Abstract ...... iv

List of Tables ...... xii

List of Figures ...... xiii

List of Abbreviation ...... xvi

Acknowledgements ...... xxi

1. Introduction ...... 1

1.1 Myeloid Cells in Cancer ...... 1

1.1.1 Myeloid Derived Suppressor Cells ...... 2

1.1.2 Dendritic Cells ...... 8

1.2 Calcium/Calmodulin-Dependent Protein Kinase Kinase 2 (CaMKK2) ...... 12

1.2.1 Calcium/Calmodulin-Dependent Protein Kinase Cascade ...... 12

1.2.2 Structure and Distribution of CaMKK2 ...... 14

1.2.3 Downstream Targets of CaMKK2 ...... 19

1.2.3.1 CaMKs ...... 19

1.2.3.2 AMPK ...... 21

1.2.4 CaMKK2 in Cell Development ...... 24

1.2.5 CaMKK2 in Cancer ...... 27

1.3 STO-609 ...... 31

1.4 Lymphoma ...... 32

vii

1.5 Research Objective ...... 33

2. Experimental Procedures ...... 36

2.1 Methods for Establishing in vivo Lymphoma Model in CaMKK2 Mice ...... 36

2.1.1 E.G7 Cell Culture ...... 36

2.1.2 Mice ...... 36

2.1.3 Intracranial Tumor Model ...... 37

2.1.4 Flank Tumor Model ...... 37

2.1.5 ELISpot Analysis ...... 38

2.1.6. Tumor Component Analysis ...... 38

2.1.6.1 Tumor Pathology Staining ...... 38

2.1.6.2 Tumor Digestion ...... 39

2.1.6.3 Flow Cytometry Analysis ...... 39

2.1.7 In Vivo Bioluminescence Imaging ...... 39

2.1.8 T-Cell Depletion Treatment ...... 40

2.1.9 STO-609 Treatment ...... 40

2.2 Methods for Establishing Other in vivo Cancer Models in CaMKK2 Mice ...... 41

2.2.1 E0771 Breast Cancer Cell Line and Breast Cancer Model ...... 41

2.2.2. Vk*Myc Myeloma Cells and Myeloma Model ...... 42

2.3 Methods for Establishment of in vitro Lymphoma-Exposed Cell Model with Cells

Derived from CaMKK2 Mice ...... 42

2.3.1 Generating Bone Marrow Derived Dendritic Cells ...... 42

2.3.2 Generating Myeloid Derived Suppressor Cells ...... 43

viii

2.3.3 Coculture of BMDC and T cells ...... 44

2.3.3.1 Primary Cell Separation ...... 44

2.3.3.2. T-cell CFSE Labeling ...... 45

2.3.3.3 Seeding of Cells ...... 45

2.3.4 Cytokine Secretion Analysis ...... 46

2.3.5 Flow Cytometry ...... 46

2.3.5.1 Myeloid Cells Staining ...... 46

2.3.5.2 Costimulatory Factor Staining ...... 46

2.3.5.3 T Cells Staining ...... 47

2.3.5.4 Cell Analysis ...... 47

2.3.6 Cytospin and Staining ...... 47

2.4 Methods for Molecular Biology ...... 48

2.4.1 Western Blot ...... 48

2.4.2 Real-Time Quantitative RT-PCR Assay ...... 49

2.4.3 RT2 Profiler PCR Arrays ...... 50

2.5 Statistical Analysis ...... 51

3. The Role of CaMKK2 in Regulating Lymphoma Growth ...... 52

3.1 Introduction ...... 52

3.2 Result ...... 53

3.2.1 CaMKK2 ablation suppresses lymphoma growth in vivo ...... 53

3.2.2. CaMKK2 ablation in myeloid cells is sufficient for lymphoma suppression ...... 58

3.2.3. CD8 T-cell ablation reverses lymphoma suppression in CaMKK2-/- mice...... 60

3.3 Conclusions ...... 63

4. The Role of CaMKK2 in Regulating Myeloid-Cell Development in Lymphoma

Environment ...... 66

4.1 Introduction ...... 66

4.2 Results ...... 68

4.2.1. CaMKK2 ablation facilitates myeloid cells maturation ...... 68

4.2.1.1 CaMKK2 ablation limits MDSC generation in vitro ...... 68

4.2.1.2. CaMKK2 ablation limits MDSC generation in tumor bearing mice ...... 74

4.2.2 CaMKK2 is expressed in dendritic cells ...... 75

4.2.3 CaMKK2-/- DCs lead to enhanced T-cell stimulation ...... 80

4.2.3.1 CaMKK2-/- dendritic cells present higher CD11c+I-A+ percentage ...... 80

4.2.3.2 CaMKK2-/- dendritic cells have higher co-stimulatory factors expression ...... 81

4.2.4 AMPK pathway is involved in CaMKK2-mediating myeloid cell development .... 95

4.3 Conclusions ...... 97

5. The Role of CaMKK2 in Other Tumor Models ...... 100

5.1 Introduction ...... 100

5.2 Results ...... 101

5.2.1 CaMKK2 regulates myeloma progression ...... 101

5.2.2 CaMKK2 regulates breast cancer progression ...... 102

5.3 Conclusions ...... 104

6. STO-609 as a Potential Immune-Regulating Treatment for Cancers ...... 105

6.1 Introduction ...... 105

6.2 Results ...... 105

6.2.1 STO-609 suppresses lymphoma and breast cancer progression in vivo ...... 105

6.2.2 STO-609 remodels the immune microenvironment of breast cancer ...... 107

6.3 Conclusion ...... 109

7. Discussion and Future Perspectives ...... 110

7.1 Summary of findings ...... 110

7.2 Potential applications and future perspectives...... 114

7.2.1 Inhibiting CaMKK2 can inhibit AMPK specifically in myeloid cells ...... 114

7.2.2 Inhibition of CaMKK2 decreases MDSC generation and function in lymphoma 117

7.2.3 Inhibition of CaMKK2 provides a potential way to enhance DC vaccine efficacy in

lymphoma by promoting DCs maturation and function ...... 121

7.2.3 STO-609 is a potential immune therapy for cancers ...... 125

7.3 Conclusion ...... 127

References ...... 128

Biography ...... 154

List of Tables

Table 1: Primer sequences for real-time PCR assay ...... 50

xii

List of Figures

Figure 1: Tumor growth of mice with E.G7 subcutaneous injection ...... 54

Figure 2: Bioluminescent imaging of mice with E.G7 subcutaneous injection ...... 55

Figure 3: CaMKK2 Mice with E.G7 intracranial injection ...... 56

Figure 4: Bioluminescent imaging of mice with E.G7 intracranial injection ...... 58

Figure 5: E.G7 tumor growth in mice with myeloid lineage CaMKK2 ablation ...... 59

Figure 6: in vivo CD8 depletion strategy ...... 61

Figure 7: CD8 depletion confirmed by flow cytometry ...... 62

Figure 8: CD8 depletion reverses tumor growth suppression in CaMKK2-/- mice ...... 63

Figure 9: CaMKK2 promoter is active in in vitro MDSC induced by E.G7 supernatant. .. 69

Figure 10: CaMKK2-/- marrow generates less MDSCs in vitro ...... 70

Figure 11: CaMKK2-/- marrow generates less CD11b+Ly6Ghi MDSCs ...... 71

Figure 12: in vitro generated CaMKK2-/- myeloid cells have higher percentage of CD11b+I-A+ cells ...... 72

Figure 13: Myeloid cells generated from CaMKK2-/- HSC have higher CD11b+ I-A+ percentage ...... 73

Figure 14: CaMKK2 ablation limits MDSC generation in vivo ...... 74

Figure 15: The CaMKK2 promoter is active in splenic CD11c+ DCs ...... 75

Figure 16: CaMKK2 promoter activity is developmentally regulated in DCs...... 77

Figure 17: CaMKK2 promoter is active in bone marrow derived dendritic cells ...... 78

Figure 18: CaMKK2 is expressed in BMDCs ...... 79

xiii

Figure 19: The percentage of CD11c+I-A+ is higher in BMDCs generated from CaMKK2-/- marrow...... 80

Figure 20: BMDC from CaMKK2-/- marrow have higher co-stimulatory factors expression after interacting with T cells...... 82

Figure 21: Cytokine secretion analysis in supernatant from coculture of BMDC with T cells...... 83

Figure 22: Cytokine secretion analysis in supernatant from BMDCs stimulated with LPS...... 84

Figure 23: CaMKK2 ablation BMDC better stimulates T-cell activation ...... 85

Figure 24: T cells cocultured with CaMKK2-/- BMDCs better express T-cell activation marker CD69, which can be blocked by anti-IL-12 mAb ...... 86

Figure 25: The proliferated CD25+ cells are FoxP3 negative ...... 87

Figure 26: CaMKK2-/- BMDC better stimulates OT1 T-cell activation ...... 89

Figure 27: in vivo DCs have higher co-stimulatory factors expression in normal CaMKK2- /- mice ...... 90

Figure 28: Surface marker expression indicates DC compartment from CaMKK2-/- have better T-cell stimulation function ...... 91

Figure 29: DC subsets in CaMKK2 WT and CaMKK2-/- tumor-bearing mice ...... 92

Figure 30: Splenocytes from CaMKK2-/- EG7 tumor-bearing mice have higher IFN-γ secretion comparing to WT ...... 94

Figure 31: Immunoblots shows AMPK pathway changes in BMDC ...... 96

Figure 32: Quantification of AMPK pathway immunoblot ...... 97

Figure 33: WT and CaMKK2-/- mice with myeloma cell injection ...... 101

Figure 34: E0771 tumor growth in CaMKK2 mice ...... 103

Figure 35: E0771 tumor growth in mice with myeloid lineage CaMKK2 ablation ...... 103

xiv

Figure 36: Efficacy of STO-609 on E.G7 lymphoma ...... 106

Figure 37: Efficacy of STO609 on E0771 breast cancer ...... 107

Figure 38: STO609 treatment reprograms the tumor microenvironment...... 108

List of Abbreviation

°C degree Celsius

Ab antibody

ACC Acetyl-CoA Carboxylase

AMPK AMP-activated protein kinase

APC allophycocyanin

APC antigen presenting cells

ATCC American Type Culture Collection

BM bone marrow

BMDC bone-marrow derived dendritic cells

C-RPMI complete RPMI medium

C-terminal carboxy-terminal

Ca2+ calcium

CaM calmodulin

CaMK Ca2+/CaM dependent protein kinase

CaMKK Ca2+/CaM dependent protein kinase kinase

cDC conventional dendritic cells

CFSE carboxyfluorescein succinimidyl ester

cm3 cubic centimeter

xvi

CMP common myeloid progenitor

CO2 carbone dioxide

Cy cyanine

DC dendritic cell

DCp dendritic cell progenitor

DMEM Dulbecco's Modified Eagle Medium

DMSO Dimethyl sulfoxide

DNA deoxyribonucleic acid

DNase I Deoxyribonuclease I

DTT dithiothreitol

EDTA Ethylenediaminetetraacetic acid

EG7 E.G7-OVA [derivative of EL4] cell line

FACS fluorescence activated cell sorting

FBS fetal bovine serum

FITC fluorescein isothiocyanate g gram

GFP green fluorescent protein

GM-CSF granulocyte-macrophage colony-stimulating factor

xvii

H&E hematoxylin and eosin

HBSS Hank's Balanced Salt Solution

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HSC hematopoietic stem cell i.p. intraperitoneal i.v. intravenous

Ig immunoglobin

IL interleukin

KLS c-Kit+Lin-/loScal1+

KO knock out

Lin lineage mAb monoclonal antibody

MDSC myeloid derived suppressor cell

MFI median fluorescence intensity mg milligram mM millimolar

N-terminal amino-terminal

OVA ovalbumin p- phosphor-

xviii

PB pacific blue

PBS phosphate-buffered saline

PCR polymerase chain reaction

PDC plasmacytoid dendritic cells

PE phycoerythrin

PerCp peridinin chlorophyll pg picogram rm recombinant mouse

RNA ribonucleic acid

RPMI Roswell Park Memorial Institute

RT reverse transcription s.c. subcutaneous

SDS sodium dodecyl sulfate

SEM Standard error of the mean

Ser Serine shRNA short hairpin RNA

TCR T-cell receptor

Thr Threonine

TLR Toll-like receptor

xix

Treg regulatory T cells w/ with wo without

WT wild type

µg microgram

µl microliter

Acknowledgements

I am deeply grateful to many individuals and entities that gave me great support and positive impacts in my long and memorable journey towards my degree at Duke

University. I would first and foremost thank my mentor and advisor, Dr. Nelson Chao, for his support, guidance, patience, and generosity during all these years. I thank him for giving me great freedom to explore different projects and research interest, and the full support and wise advice in my career development. I would love to thank Dr. Luigi

Racioppi, for guiding me step by step towards the completion of the degree, challenging me critically in science, and encouraging me positively in life. I want to express my sincere appreciation to all my thesis committee members including Dr. Benny Chen, Dr.

Soman Abraham, Dr. Herman Staats, and formerly Dr. Xiao-Fan Wang, for their invaluable and unreserved support both scientifically and emotionally. Special thanks are given to Dr. Shijie Sheng, my former mentor at Wayne State University, for the unceasing caring and encouraging all along the years. I am honored to have so many great mentors from whom I have learned a great deal in their attitude and philosophy towards career, family, life, humanity, and society. I also want to thank both present and past members of Chao’s Lab for all the stimulating discussions, support, encouragement, and love all along these years. I especially thank Yaping Liu, Sadhna Piryani, and formerly William Lento, for the fruitful collaborations that have directly contributed to

xxi

this dissertation. I am also indebted to the staffs at Duke Cancer Institute’s Flow Core

Facility, Duke animal facilities, and Duke Human Vaccine Institute for their technical support, without which this research would not have been possible.

I am incredibly thankful to God for providing the churches, fellowships, and friends in Detroit and in North Carolina, whose prayers, support and love have helped raise me up in all my endeavors. Thank Ned Chung for revising this dissertation. Thank

Gladys Ang, Xuhui Bao, Leina Lu, Ting Zhou, Chunbai Zhang, Vicky Yeh, Qin Wang,

Teresa Lee, Chen Yang, Jessica Jin, Xiaoting Wang, Joanne Lin, for all the company, comfort, friendship, help and support during all these ups and downs. Many thanks go to Grace Hsu and James Wang, David Chang and Wen-Liang Chung, Wei-Te Chiang and Huei-Li Chu, Pai-Lien Chen and Hsiao-Chuan Tian, for the caring and love and prayers all along. Thank Benny Chen’s family – Benny and Liling, Michelle and Oliver, for making me feel at home here. I am also deeply grateful and indebted to my family – my parents and grandparents, my brother Chinson Leung, for understanding and forgiving my years of absence, giving me unconditional love and endless support, and having the faith in me when I don’t have any. I love you more than words can say.

I am forever grateful for the memorable journey at Duke University, in which I have experienced a different world, a different life, a different me. I thank everyone whom I have encountered during these years, for adding different colors to my life.

xxii

1. Introduction

Cancer is a collection of related disease in which abnormal cells undergo uncontrolled cell growth with potential to spread and invade to other parts of the body.

It has tremendously impacted societies worldwide. Growing evidences has shown that during the cancer progression, cancer cells could induce an inflammatory and immune suppressive environment to impair the normal maturation and function of immune cells. (Bronte, 2009; Gabrilovich & Nagaraj, 2009; Schmid & Varner, 2010) (Coussens,

Zitvogel, & Palucka, 2013) Calcium/Calmodulin dependent protein kinase kinase 2

(CaMKK2) is a crucial kinase in the calcium signaling cascade. CaMKK2 has been reported to be involved in mediating cell differentiation, inflammation, and cancer progression in different studies. (Racioppi & Means, 2012) These characteristics make

CaMKK2 a potential crucial regulator in the cancer microenvironment and a potential treatment target for cancer therapy. In this chapter, I am going to discuss the key concepts of CaMKK2 and cancer immune regulation, the recent updates of the field, and the rationale and research objectives of this dissertation.

1.1 Myeloid Cells in Cancer

Myeloid cells, including macrophages, monocytes, and dendritic cells, are a collection of hematopoietic cells differentiated from the common myeloid progenitors.

(Gabrilovich et al., 2007) In effective antitumor immunity, antigen-presenting cells

(APCs), especially dendritic cells (DCs), are required to process and present tumor antigen to T cells, and thereby facilitating T-cell activation and eventually eliminating tumor cells. (Ali et al., 1993; Banchereau & Steinman, 1998) Thus, the immunostimulatory characteristics of mature APCs are important to bridge innate immunity and adaptive immunity together. However, under the cancerous condition, the balance of myeloid-cell differentiation and maturation is interrupted, leading to an accumulation of immature myeloid cells, increased number of monocytes and macrophages, and decreased number and function of matured dendritic cells in both mice models and human patients. (Almand et al., 2001; Gabrilovich, Ostrand-Rosenberg,

& Bronte, 2012) (Kusmartsev & Gabrilovich, 2002) Instead of presenting tumor antigen and activating T-cell response, the immature myeloid cells have impaired immune function and therefore inhibit the T-cell response, ultimately creating an immune suppressive state to protect tumors.

1.1.1 Myeloid Derived Suppressor Cells

The suppressive function of some myeloid cells in infection was reported as early as 1970s and 1980s, but the term of myeloid-derived-suppressor cells (MDSCs) was not formally introduced until 2007. (Gabrilovich et al., 2007) In the past decade, a large amount of information and evidence have been accumulated concerning the biology, clinical relevance, and potential impact in various pathological conditions. (Gabrilovich

et al., 2012; Kusmartsev & Gabrilovich, 2002; Youn, Nagaraj, Collazo, & Gabrilovich,

2008)

Broadly defined, MDSCs are immature myeloid cells that are not commonly detected in physiologic conditions. However, MDSC have been found to accumulate in pathological conditions including severe or chronic inflammation, autoimmune diseases, transplantation, and cancers. (Gabrilovich et al., 2007) Distinct from the terminally differentiated cells including macrophages, DCs, monocytes or neutrophils, MDSCs include some myeloid progenitors and immature mononuclear cells and can suppress immune response. Based on morphology and phenotypes, MDSCs are categorized in monocytic MDSC (M-MDSCs) and polymorphonuclear MDSCs (PMN-MDSCs). (Youn et al., 2008) Although M-MDSC is considered more immune suppressive, PMN-MDSCs are more prevalent in cancer and severe infection conditions. In human, M-MDSCs are

CD14+HLA-DR-/lo cells and PMN-MDSCs are CD14-CD11b+CD33+CD15+ cells. There is another population of MDSC in human defined as Lin-HLA-DR-CD33+, which is enriched by myeloid progenitors. Although these three catergories can be identified in cancer patients, there is no distinct marker to distinguish MDSCs specifically. (Damuzzo et al., 2015; Talmadge & Gabrilovich, 2013)

MDSCs in mice are broadly identified as CD11b+Gr1+ cells and further categorized according to Gr-1 subgroup markers as monocytic MDSC (CD11b+Gr-1lo, or

CD11b+Ly6ChighLy6G-) and granulocytic MDSC (CD11b+Gr-1high, or

CD11b+Ly6CloLy6G+). While both subsets of MDSCs can suppress immune responses,

M-MDSCs are stronger at single-cell level, and can suppress both antigen-specific and nonspecific T-cell response using nitric oxide (NO) generation and suppressive cytokines. PMN-MDSCs can suppress only antigen-specific T-cell response by generating reactive oxygen species (ROS) to interfere the binding of peptides to MHC molecules, thus blocking T-cell function. In addition to suppressing T-cell function,

MDSCs can also suppress the immune system by expressing arginase, inducible NOS

(iNOS), COX2, and TH2 cytokines including TGF-β and IL-10. MDSCs can also induce

Treg and suppress the function of NK cells and dendritic cells in cancer environment.

(Ribechini, Greifenberg, Sandwick, & Lutz, 2010; Youn et al., 2008)

MDSCs facilitating tumor growth has been confirmed in multiple cancer models in both human and mice. In most solid tumors including breast cancer, mesothelioma, colon cancer, hepatic cancer, renal cancer, and lung cancer, PMN-MDSCs are significantly expanded and associated with a poor prognosis; while M-MDSC is more prevalent in melanoma, multiple myeloma, lymphoma, and prostate cancer. (Diaz-

Montero et al., 2009; Filipazzi et al., 2007; Hoechst et al., 2008; Kusmartsev et al., 2008;

Montero et al., 2012; Movahedi et al., 2008; Serafini, Mgebroff, Noonan, & Borrello, 2008;

Veltman et al., 2010; Vuk-Pavlovic et al., 2010; Zhang et al., 2013; Solito et al., 2011) In the

cancer microenvironment, cancer cells and the stroma secrete various growth factors including granulocyte-macrophage CSF (GM-CSF), macrophage CSF (M-CSF), granulocyte CSF (G-CSF) to support the expansion of immature myeloid cells, while the abundant cytokines including IL-6, IL-1, IL-10, IFN-γ, TNF-α can further activate the suppressive function of MDSCs. (Damuzzo et al., 2015; Dolcetti et al., 2010) During the regulation of MDSC expansion and activity, the STAT transcription factor family especially STAT3, is considered as a critical factor by regulating arginase production through C/EBP and IFN regulatory factor pathway. STAT3 also prevents the differentiation of MDSCs into mature dendritic cells. (Nefedova et al., 2004) HIF-1α can regulate MDSC activity by upregulating iNOS and arginase. (Doedens et al., 2010) HIF-

1α can also up regulate the expression of programmed cell death ligand-1 (PD-L1) on myeloid cells, and faciliate their differentiation into tumor-associated macrophages

(TAMs). (G. Liu et al., 2014; Shirato et al., 2009) Recent research also indicates that inflammatory mediators including high mobility group box 1 (HMGB1) and PPARγ also control MDSC function. (L. Wu, Wang, Qu, Yan, & Du, 2011)

In return for the favorable environment generated by cancer cells, MDSCs support tumor growth and metastasis by directly interacting with tumor cells to induce stemness and increase survival, and also remodel the tumor microenvironment by suppressing immune killing and establishing pre-metastatic niches. MDSCs can secrete

VEGF and its analogues, and matrix metaloproteinases (MMPs) to induce angiogenesis, support tumor growth and facilitate metastasis. (Gabrilovich et al., 1996) MDSC-derived

IL-6 has been found to induce tumor stemness in a breast cance mice model. The accumulation of MDSCs also recruits IL-17-producing T cells. (H. Zhang et al., 2015) The changes and potential mechanisms of MDSC-tumor interaction in different tumor models seem independent and may not by syngeneic to each other, adding to the complexity of controling MDSCs in tumors.

Currently, the strategies to therapeutically target MDSCs include 1) eliminating

MDSCs by chemotherapy; 2) deactivating MDSCs by targeting NO and ROS generation, and 3) skewing the myelopoeisis from MDSC accumulation to differentiation towards mature DCs and macrophages. (Marvel & Gabrilovich, 2015; Najjar & Finke, 2013) To eliminate MDSCs by chemotherapy, low dose of gemcitabine, cisplatin, 5-fluorouracil have been reported in murine models to delete MDSCs while preserving the T-cell functions. Immune therapies aiming to augment the vaccine-induced immune responses are underway in mice by using engineered peptides selectively targeting TRAIL receptors on MDSCs. Since ROS and NO are essential components in MDSC-mediated immune suppression, it is crucial to upregulate the expression of antioxidant enzymes or downregulate the expression and activity of arginase, iNOS and COX2. Tadalafil, the phosphodiesterase-5 (PDE-5) inhibitor, has been recently reported in a clinical trial of

head/neack cancer patients to efficiently reduced the circulating MDSCs, lowered iNOS and arginase expression, and increased the number of tumor-specific T cells. (Serafini et al., 2006) In the mouse cancer models, MDSCs have been reported to have a shorter life span compared to normal immature monocytes and neutrophils (Condamine et al.,

2014), and can be converted to dendritic cells and macrophages according to the environmental stimulation . All-trans-retinoic acid (ATRA) that can help MDSCs differentiate into immunogenic DCs has been tested on renal cancer carcinoma patients and lung cancer patients.(Kusmartsev et al., 2008; Nefedova et al., 2007) Sunitinib inhibits STAT3 and other signaling pathways including VEGF and c-kit has been demonstrated efficiently in reducing MDSC level in renal cell carcinoma patients. (Ko et al., 2009) Skewing and re-educating MDSCs to immunogenic antigen-presenting cells may be a promising way to improve antitumor immunity in patients.

Recent data depleting PMN-MDSCs in cancer patients question if PMN-MDSC eradication truly impacts tumor progression, and the strong ability of M-MDSC to expand and compensate the loss of PMN-MDSCs also makes the problem complex.

(Damuzzo et al., 2015) It is more likely that the MDSCs induction is a dynamic process according to the pathological environment, and MDSCs suppress the immune system using multiple mechanisms throughout the progression of the diseases. As a result, using multiple strategies to target MDSCs may be necessary in cancer therapy.

1.1.2 Dendritic Cells

Dendritic cells (DCs) are derived from bone marrow progenitors and migrate to all tissues especially the sites of potential antigen entry, to further differentiate and mature. The maturation of DCs is a complex process and can be activated by Toll-like receptors (TLRs), interferons (IFNs), and tumor necrosis factor receptor (TNF-R) family.

The mature DCs can capture, process the antigens including invading microbes and abnormal tumor antigens to T cells, and further stimulate T-cell activation and clonal proliferation. (Banchereau et al., 2000; Steinman, 2001) This efficient antigen-presenting capability makes DCs the molecular sensor of the environment and links the innate and adaptive immune responses together. Besides T cells, mature DCs can also interact with

NK cells and B cells for further activation and differentiation. However, antigen presentation mediated by immature DCs can induce T cell tolerance due to the lack of costimulatory signals. (Steinman, 2001) Thus, the maturation of DCs with proper costimulatory factors expression is crucial for effective T-cell activation and differentiation.

DCs are categorized into different subsets according to their different differentiated stages, cell surface markers, and functions. Immature DCs (imDCs) with

FLT-3 ligand expression are produced from CD34+ hematopoietic stem cells from bone marrow. (McKenna et al., 2000) In human, while the CD11c+CD1a+ imDCs developed

from common lymphoid progenitors (CLP) later differentiate to Langerhans cells in skin epidermis, CD11c+CD1a- imDCs developed from common myeloid progenitors (CMP) later differentiate to interstitial imDCs in skin dermis and other tissues. (Ito et al., 1999;

Kadowaki, Antonenko, & Liu, 2001; Strunk, Egger, Leitner, Hanau, & Stingl, 1997) These two subsets have different phenotypes and functions. While Langerhans cells are strong at CD8 T-cell priming, interstitial DCs are more active in large-molecule antigen processing through macropinocytosis, IL-10 production, and B cell activation. The production and diversification of imDCs from CD34+ stem cells are considered as pathogen-independent and in a steady state. Besides imDCS, the DC precursors (pre-

DCs) including monocytes (pre-DC1) and plasmacytoid cells (pre-DC2) are also differentiated from hematopoietic stem cells. (Y. J. Liu, 2001) While imDCs have the recognizable morphology and moderate ability to stimulate T-cell activation and high mobility in culture, pre-DCs are more prone to regulate innate immunity against microbe infections. (M. Y. Liu, Xu, & Song, 2001; Y. J. Liu, 2001) Pre-DC1 and Pre-DC2, again, have distinct phenotypes and destinations. Pre-DC1 expresses the myeloid antigens including CD11b, CD11c, CD14 and CD33, and can differentiate into immature myeloid DCs (im-DC1) in culture with GM-CSF and IL-4, while Pre-DC2 are more lymphocyte-specific in mRNA level and can differentiate into im-DC2 while cultured with IL-3. When activated by CD40 ligand, DC1s can secrete large amount of IL-12 and

strongly mediate T-cell differentiation towards T helper type 1 (TH1) and cytotoxic T cells, while DC2 are more prone to induce TH2 responses and suppress CD8+ T cells by generating IL-10. (Y. J. Liu, Kanzler, Soumelis, & Gilliet, 2001) In mice, splenic DCs are separated into CD8a+CD11b- and CD8a-CD11b+. While the CD8+ DCs are tend to secrete IL-12, mediate TH1 response, cross-prime CD8+ T cells, CD8- DCs prefer to induce TH2 response. CD4 marker is later also found expressed on DCs. While the

CD8+CD40- DCs prone to mediate TH1 response, CD8-CD4+ DCs and CD8-CD4- DCs are so far considered similar in function. Even though pre-DC1 and pre-DC2 have the potential to develop into TH1 or TH2 DCs, the differentiation is also affected by antigen encountering and cytokines environment including IFNs and TNF-a. Although CD4 and

CD8 are lymphoid markers, they do not indicate the origin of peripheral DCs, but more functional difference between DC subsets. (Brasel, De Smedt, Smith, & Maliszewski,

2000)

DC maturation is a complex process. DC maturation is defined by the capability of DCs converting from antigen-capturing to antigen-presenting. The matured DCs express high level of antigen-presenting molecules (MHC I, MHC II) and costimulatory molecules (CD40, CD80, CD86, CD54), and release of large amount of IL-12 that can induce a Th1 immune response through STAT4 pathway and promotes the IFNγ secretion. (Banchereau et al., 2000; Banchereau & Steinman, 1998; Cella, Sallusto, &

Lanzavecchia, 1997; Fukao et al., 2001; Lanzavecchia & Sallusto, 2000) On the contrary,

IL-10 is the cytokine secreted by immature DCs. IL-10 blocks the DC maturation and induces a Th2 immune response. Blocking IL-10 using functional antibody can facilitate

DC maturation and increase IL-12 secretion. Once matured, DCs are insensitive to IL-10, by an unknown mechanism. (Corinti, Albanesi, la Sala, Pastore, & Girolomoni, 2001;

Fukao et al., 2001; la Sala et al., 2001) Toll-like receptors (TLRs) are known as the crucial bridge between innate and adaptive immune responses. TLRs mediate DC activation and maturation through the NF-κB and MAP kinase pathway, up-regulating the expression of MHC and costimulatory molecules, and increasing the production of cytokines. All these factors are necessary to sustain the clonal expansion and differentiation of the newly activated T cells. The newly activated T cells can in turns drive DC maturation independently of TLRs or even specific antigens. (Ardavin,

Amigorena, & Reis e Sousa, 2004)

In the tumor environment, malignant cells can interact with different cells in the environment including macrophages, granulocytes and monocytes, fibroblasts, mast cells, T cells, epithelial and endothelial cells. All these cells can secrete diverse growth factors, chemokines, interleukins, and many other enzymes and small molecules to reshape the microenvironment into a chronic inflammatory status. The chronic inflammatory environment facilitates the tumor cell survival and proliferation, while

blunting the DCs’ function as the environmental sensor and monitor. (Coussens et al.,

2013) The exposure to cancer-derived factors such as thymic stromal lymphopoietin

(TSLP) can also skew the DCs maturation in tumor microenvironment towards TH2-type inflammation, inducing more TH2 T cells and M2 macrophages and eventually a tumor- friendly microenvironment. Therefore, facilitating DCs maturation and reprograming the DCs differentiation from TH2 to TH1 with increased T-cell costimulatory function would be a potential strategy for cancer therapy.

1.2 Calcium/Calmodulin-Dependent Protein Kinase Kinase 2 (CaMKK2)

1.2.1 Calcium/Calmodulin-Dependent Protein Kinase Cascade

Calcium (Ca2+) plays a pivotal role in cellular processes. In signal transduction,

Ca2+ is a ubiquitous and essential second messenger that regulates many important pathways involving cell survival and apoptosis, proliferation, motility, secretion, and development. (Hook & Means, 2001) In mammalian cells, calcium level is strictly regulated. While the intercellular Ca2+ is released from bones and tightly controlled by the parathyroid hormones, the intracellular Ca2+ stores in mitochondria and the endoplasmic reticulum, and can be released during certain cellular events. The elevated intracellular Ca2+ can directly act on target proteins, or bind to various intracellular Ca2+- binding proteins to further trigger the signaling cascades.

Calmodulin (CaM) is one of the major calcium signaling transducers. (Berridge,

Bootman, & Roderick, 2003; Hook & Means, 2001) CaM is a highly conservative protein that acts as a primary Ca2+ receptor. It consists of four high-affinity Ca2+-binding motifs.

Once bound to Ca2+, the conformational changes in CaM enables it to further interact with other Ca2+-directed targets including Ca2+/CaM-dependent protein kinases (CaMK).

CaMKs are Ca2+/CaM-dependent Ser-Thr kinases with different ability to phosphorylate distinct substrates. Unlike the other family members, CaMKIII is the phosphorylase kinase and dedicated to the restricted repertoire as myosin light chain kinase (MLCK).

Except CaMKIII, all the other members including CaMKI subfamily (α, β, γ, δ), CaMKII subfamily (α, β, γ, δ), CaMKIV, and CaMK kinase (CaMKK) family (α and β, also known as CaMKK1 and CaMKK2), are all multifunctional kinases with broad substrates. (Braun

& Schulman, 1995; Hudmon & Schulman, 2002) The multifunction kinases are similar in homology and domain structures including a conserved N-terminal kinase domain, a central regulatory domain responding for auto-inhibition, and an overlapping CaM- binding domain. Once CaMK binds to Ca2+/CaM complex, a conformational change occurs in the autoregulatory domain and releases the auto-inhibition.

Despite the similarities in general structures and common regulation and activation, the CaMKs are functionally distinct from each other according to their specific distribution, subcellular localization, specific substrates, and regulation

mechanisms. While CaMKII can be fully activated only by binding to Ca2+/CaM complex, to fully activate CaMKI and CaMKIV still requires the phosphorylation of the activation loop threonine by the upstream Ca2+/CaM-dependent kinase CaMKK1 or

CaMKK2. This forms the CaM-depndent kinase cascade. (Soderling, 1999; Tokumitsu,

Enslen, & Soderling, 1995) As the goal of this dissertation research is to explore the role of CaMKK2 in regulating tumor microenvironment, the focus of the discussion will continue on the current known about the CaMKK2, its function on tumor and the environment, and the downstream signaling pathways in regulating cancer progression.

1.2.2 Structure and Distribution of CaMKK2

The Ca2+/CaM-dependent kinase kinase (CaMKK) family is known as the upstream kinases to activate CaMKI and CaMKIV. CaMKKs family includes two isoforms CaMKK1 and CaMKK2 encoded by Camkk1 and Camkk2 respectively. CaMKK1 and CaMKK2 are 50% identical at amino acid level and both predominantly expressed in cytoplasm. (Fujiwara et al., 2015) Both these two enzymes are highest expressed in brain but in different subregions, moderate expressed in the testis, thymus, spleen, and low in some other cells and tissues. (Anderson et al., 1998) Other known downstream kinases of CaMKKs besides CaMKI and CaMKIV include AMP-dependent kinase

(AMPK) and protein kinase B (PKB). (Hawley et al., 2005; Hook & Means, 2001; Hurley et al., 2005) In particular, AMPK is known as a ubiquitous energy sensor in all tissues,

and also a direct and physiological substrate of CaMKK2 that has the highest affinity to

Ca2+/CaM.

As reviewd by Racioppi and Means (Racioppi & Means, 2012), CaMKK2 is a highly conserved 66-68-kDa kinase that exists in rat, mouse, and human. Like other

CaMK family members, CaMKK2 also has the unique N- and C-terminal domains, a regulatory domain with overlapping autoinhibitory and CaM-binding regions, and a central Ser/Thr-directed kinase domain. Although CaMKK2 shares the similarity of the kinase homology domain to other CaM kinases at the amino acid sequences level in cloned cDNAs, CaMKK2 is distinct with a unique 22-residue Pro/Arg/Gly-rich insert between the protein substrate motifs and the ATP-binding motifs. (Tokumitsu et al.,

1999) In human, CaMKK2 is on chromosome 12q24.2, over 40kb pairs. It has 18 exons with 17 introns. CaMKK2 major transcripts are generated by the last two exons that have the polyadenylation sites, or by alternative splicing of internal exons 14 and/or 16 regulated by PKA and CaMKIV. (L. S. Hsu et al., 2001; Ishikawa et al., 2003) Human

CaMKK2 gene is identified for the consensus DNA binding sequence for several transcription factors that are typically restrictedly expressed in stem cell progenitors.

These transcription factors include Ikaros, GATA-binding factor 1 (GATA1), and Runt- related transcription factor 1 (RUNX1). (Inoue, Shiga, & Ito, 2008; Tijssen et al., 2011;

Zagami, Zusso, & Stifani, 2009) The prevalence of these transcription factors may

explain the restricted expression of CaMKK2 to some specific cell types, and its role in regulating neural development and hematopoiesis. (Kokubo et al., 2009; Teng, Racioppi,

& Means, 2011)

CaMKK2 is highly expressed in many areas of the brain including the olfactory bulb, hippocampus, dentate gyrus, amygdala, hypothalamus, and cerebellum, and at lower levels in testis, spleen and lung. (Anderson et al., 1998; Edelman et al., 1996;

Tokumitsu et al., 1995) In some isolated murine cells including preadipocytes, embryonic fibroblasts, and hepatocytes, CaMKK2 is detectable. (Lin et al., 2015; Lin,

Ribar, & Means, 2011) In adults outside the brain, CaMKK2 is found nearly exclusively in the hematopoietic stem cells, progenitors, and some myeloid cells. (Racioppi,

Noeldner, Lin, Arvai, & Means, 2012; Teng et al., 2011) The expression of CaMKK2 is high in peritoneal macrophages and bone marrow derived macrophages.

CaMKK2 was first discovered in brain and it plays critical roles in the development of neurons and brain physiology. CaMKK2 regulates hippocampal memory, especially the long-term memory formation, by activating the cAMP response- element binding protein (CREB) and regulating the learning-induced neuronal cytoskeleton remodeling via the CaMKK/CaMI cascade. (Peters et al., 2003)

CaMKK2/CaMIV/CREB signaling pathway is essential for granule cell precursors

(GCPs) proliferation and migration, thus controls the cerebellar granule cell

development. (Kokubo et al., 2009; M. Sato, Suzuki, & Nakanishi, 2006; Sotelo, 2004; V.

Y. Wang & Zoghbi, 2001) CaMKK2 interferes with the Ghrelin, an important hormone produced by the intestine that exerts a potent central orexigenic effect by acting on hypothalamic neurons to regulate various hormonal and nutrient signals, by interacting with its downstream signaling component AMPK. (Anderson et al., 2008; Morton,

Cummings, Baskin, Barsh, & Schwartz, 2006) Other effects of CaMKK2 on brain functions includes regulating serotonin secretion by CaMKK2/CaMKIV/CREB pathway, and CaMKK2 is considered a schizophrenia susceptibility gene due to several highly disease-associated single-nucleotide polymorphisms (SNPs). (Atakhorrami et al., 2015)

CaMKK2 has been reported to promote adipogenesis and hepatic glucose metabolism in response to overnutrition in obesity and diabetes. (Anderson et al., 2012)

Comparing to WT mice with matched food consumption, CaMKK2-/- mice have more fat accumulation in physiological condition, yet are protected from overnutrient- induced obesity, glucose intolerance and insulin resistance. CaMKK2 expression decreases as preadipocytes mature into adipocytes, indicating that CaMKK2 regulates preadipocyte differentiation. This will be further discussed in the later section about cell differentiation (Section 1.2.4). In hepatocytes, CaMKK2 participates in glucose metabolism by promoting gluconeogenesis and suppressing lipogenesis through the

CaMK/CREB/HDAC5 pathway.

The role of CaMKK2 in regulating hematopoietic and immune cells is a recent discovery. (Teng et al., 2011) In hematopoietic cells, CaMKK2 is highly expressed in hematopoietic stem cells and the progenitors, and the expression gradually decreases as cells differentiate. This suggests that CaMKK2 may regulate hematopoietic development. The recent studies confirm that CaMKK2 and its substrate CaMKIV play an important role in HSC survival and maintenance, and granulopoiesis. This will be further discussed in the later section about CaMKK2-mediated cell differentiation

(Section 1.2.4). In the differentiated immune cells, CaMKK2 is selectively expressed in macrophages, while CaMIV is found to regulate the activation and survival also in dendritic cells. (Illario et al., 2008) Besides, CaMKI is implicated as the regulator of TLR4 singling pathway, which is crucial for dendritic cell activation. In addition to CaMKI and CaMKIV, AMPK is another downstream target of CaMKK2 and known as a regulator in the inflammatory response and macrophage activation. (Racioppi et al.,

2012; Sag, Carling, Stout, & Suttles, 2008) These evidence indicate that CaMKK2 may be a crucial regulator in the inflammatory responses. Following on the interest of this dissertation, the role of CaMKK2 in regulating immune cell development and inflammation will be further investigated in the following chapters.

1.2.3 Downstream Targets of CaMKK2

CaMKI, CaMKIV, and AMPK are the well-characterized substrates of CaMKK2.

CaMKK2 phosphorylates CaMKI and CaMKIV on their activation loop Thr residues to fully activate these kinases. (Soderling, 1999) The phosphorylation of CaMKIV also enables the translocation of the kinase into the nucleus and regulates gene transcription.

AMPK is an additional direct and physiological substrate of CaMKK2. (Green,

Anderson, & Means, 2011; Hurley et al., 2005) Unlike with CaMKIV, only the active conformation of CaMKK2 can bind to AMPK. Thus, even though CaMKK2 can have autonomous activity against CaMKI and CaMKIV by auto-phosphorylation, to activate

AMPK it still requires Ca2+/CaM binding.

1.2.3.1 CaMKs

CaMKI and CaMKIV are both CaMKK2 downstream targets, which require

CaMKK2 binding for full activation. (Soderling, 1999) Among all four isoforms, CaMKIα was the first to be identified and is also the most-well studied. (Picciotto, Czernik, &

Nairn, 1993) CaMKI is broadly expressed in all tissues, but most of its biological functions are studied in neural system. CaMKK/CaMKI pathway is reported to regulate neuronal motility, neuronal branching, spine formation, and synaptogenesis. (Saneyoshi et al., 2008; Wayman et al., 2004; Wayman, Lee, Tokumitsu, Silva, & Soderling, 2008)

Recent studies revealed the role of CaMKI in regulating cell cycle by activating

cyclinD/CDK4 during G1 to S transition. (Kahl & Means, 2004) CaMKI has been reported as a component in TLR4 signaling pathway where it regulates immune response in sepsis; however, more confirmatory studies are still needed. (X. Zhang et al., 2011)

CaMKIV has been considered as an important transcriptional regulator. It has two isoforms, CaMKIVα and CaMKIVβ. (Sun, Means, LeMagueresse, & Means, 1995)

CaMKIV is expressed only in specific tissues including some distinct regions of the brain, the thymus, testis, spleen, ovary, and certain immune cells. (Means, Ribar, Kane,

Hook, & Anderson, 1997; S. L. Wang, Ribar, & Means, 2001) In brain, the CaMKK2-

CaMKIV cascade plays an important role in memory formation and cerebellar granule cell development. (Anderson et al., 1998; Kokubo et al., 2009; Ribar et al., 2000) In

CaMKIV knock-out mice, HSCs are reported with significant defects in survival and proliferation, leading to the hematopoietic abnormal hyper-proliferation and exhaustion, which is resulted from the decreased levels of CREB phosphorylation and the related proteins. (Kitsos et al., 2005) These evidence show that CaMKK2/CaMKIV has an important role in HSC regulation. CaMKIV has been reported to regulate T-cell function through CREB-dependent pathway, and has been associated with the survival and activation of dendritic cells in inflammatory response. (Anderson, Ribar, Illario, &

Means, 1997; Illario et al., 2008)

In general, CaMKK2/CaMKs plays important role in regulating neural development and function, however, its role in mediating immune responses warrants further investigation.

1.2.3.2 AMPK

AMPK is a direct and physiological downstream target of CaMKK2. It is known as a conserved central metabolic sensor that allows cross-talk between the cell metabolism and signaling networks. (Hardie, Ross, & Hawley, 2012; Mihaylova & Shaw,

2011) AMPK is a heterotrimeric Ser/Thr kinase complex that has a catalytic α subunit and two regulatory β and γ subunits to make up to 12 different heterotrimers. These subunits exist in different isoforms to show tissue specificity, and contribute to different cell functions including cell growth, proliferation, differentiation in normal cells and tumor cells. (Fox, Phoenix, Kopsiaftis, & Claffey, 2013; Hawley et al., 2010; Xiao et al.,

2011)

AMPK serves as an energy sensor at the cell level to restore energy homeostasis.

During the metabolic stress condition such as hypoxia or glucose deprivation, ATP level is reduced while ADP and AMP amount increase. The direct binding of ADP and/or

AMP to the AMPK γ subunit can lead to conformational changes that promote phosphorylation of Thr172 within the α subunit, leading to AMPK activation. (Hardie et al., 2012) The upstream Ser/Thr protein kinases of AMPK include tumor suppressor liver

kinase beta 1 (LKB1) and CaMKK2 (Hawley et al., 2005) While LKB1 activates AMPK during energy stress, CaMKK2 activates AMPK responding to the increased intracellular

Ca2+ level regardless of the energy status and compensates the absent or mutated function of LKB1. (Hawley et al., 2005) The increased reactive oxygen species (ROS) is also reported to activate AMPK in an LKB1-independent way. Once activated, the

AMPK can switch the energy balance from anabolic pathways to catabolic pathways to maintain energy homeostasis, or phosphorylate the transcription factors and coactivators to mediate cell growth and survival. (Vander Heiden, Cantley, &

Thompson, 2009) AMPK can also initiate autophagy by activating serine/threonine- protein kinase (ULK1), and increase mitochondrial biogenesis by activating peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α)/SIRT1 pathway.

(Mihaylova & Shaw, 2011) In most quiescent cells including DCs, macrophages, and unstimulated immune cells, oxidative metabolism is mainly used to generate ATP.

(Pearce & Everts, 2015) However, in the rapidly proliferating mammalian cells including cancer cells and activated immune cells, AMPK is activated by glucose deprivation, and control the elevated aerobic glycolysis (Warburg effect) by switching the Warburg effect to the oxidative metabolism. (Faubert et al., 2013)

While naive immune cells tend to use oxidative metabolism, activated immune cells in inflammation favors aerobic glycolysis. The activation of AMPK regulates

inflammation mainly by down-regulating glycolysis and promoting oxidative metabolism, thus steering the immune cells from pro-inflammatory to anti- inflammatory phenotype. Exposing dendritic cells to LPS can results in a reduced

AMPK phosphorylation besides DC maturation and increased glucose consumption, while knocking down AMPK or suppressing AMPK using 5-Aminoimidazole-4- carboxamide ribonucleotide (AICAR) can enhance this phenomenon. (Krawczyk et al.,

2010) Similar data are seen in macrophages. When macrophages are exposed to LPS or fatty acid, AMPK is inactivated through Thr172 dephosphorylation.(Sag et al., 2008)

Down regulating AMPK in macrophages when exposed to LPS can increase proinflammatory cytokines secretion, while an AMPK mutant in active form leads to more anti-inflammatory cytokines.(Yang, Kahn, Shi, & Xue, 2010; Yang et al., 2012) The AMPK activating drug can suppress macrophages pro-inflammatory responses but not in macrophages with a mutated AMPK.(Jeong et al., 2009) Bone marrow derived macrophages from AMPK-β1 knockout mice with a decreased of AMPK activity are detected with reduced mitochondrial proteins and decreased fatty acid oxidation, while the pro-inflammatory markers are increased after stimulation.(Galic et al., 2011) These data demonstrate that by mediating changes in metabolism, AMPK is more likely to induce macrophages to polarize towards anti-inflammatory M2 phenotype rather than pro-inflammatory M1 phenotype. And this effect may also apply to dendritic cells

because of the similarity of metabolism and function between dendritic cells and macrophages. The ability for activated AMPK to promote myeloid cells polarization towards anti-inflammatory phenotype makes it a potential target cancer immune therapy.

1.2.4 CaMKK2 in Cell Development

The role of CaMKK2 in mediating cell development was initially investigated in neural system. In the hippocampus, CaMKK2/CaMKI cascade is required for forming the normal axonal growth cone morphology and outgrowth, regulating the dendritic arborization, and developing the spine and synapse. (Cao et al., 2011; Saneyoshi et al.,

2008) CaMKK2 deficiency could lead to the abnormal neuronal cytoskeleton modeling, and result in impaired memory formation. (Peters et al., 2003) In cerebellar development, CaKK2/CaMKIV pathway is essential for the granule cell precursors to cease proliferation in the external granule layer and subsequently migrate to the internal granule layer to form synaptic connections with Purkinje cells. (Kokubo et al., 2009; M.

Sato et al., 2006; Sotelo, 2004)

In bone remodeling, CaMKK2 inhibits the osteoblast (OBs) differentiation from mesenchymal stem cells (MSCs). (Cary et al., 2013; Pritchard et al., 2015) MSCs from

Camkk2-/ mice have higher alkaline phosphatase positive OBs, and are more prone to differentiate into OBs even without the BM microenvironment. The intrinsic effect of

CaMKK2 ablation on MSCs result in the higher bone mass and more active OBs in

Camkk2-/ mice. CaMKK2-CaMKIV-pCREB pathway and the inhibition effect of CaMKK2 on PKA activation have been assumed to involve in the process. While promoting OBs activity, CaMKK2 ablation also impairs osteoclasts formation through the CaMKK2-

CREB-NFATc1 signaling cascade. These results indicate the intrinsic role of CaMKK2 in regulating of MSC differentiation into OBs and OCs, thus maintaining the balance of bone remodeling.

CaMKK2 also regulates adipogenesis. (Y. Y. Chen, Lee, Hsu, Wei, & Tsai, 2012;

Lin et al., 2011) Lost of CaMKK2 results in the excessive accumulation of white adipose tissue and enlarged adipocytes. The expression of CaMKK2 is detectable in preadipocytes, but not in mature adipocytes. During adiopogenesis, CaMKK2 expression is markedly decreased, while the mRNAs encoding the early adipogenic genes CCAAT/enhancer binding protein (C/EBP) is increased. By activating AMPK to maintain preadipocyte factor 1 (Pref-1)-ERK-Sox9 signaling, the expression of CaMKK2 in preadipocytes prevents the differentiation towards white adipocyte tissues.

A recent report shows the role for CaMKK2 in restricting the commitment and differentiation of myeloid progenitors into the granulocytic lineage.(Teng et al., 2011)

CaMKK2-/- mice have more common myeloid progenitor (CMP) cells with an accelerated granulopoietic phenomenon in the marrow. When transplanted to irradiated mice, the

donor cells from CaMKK2-/- mice increased the production of mature granulocytes in the marrow as well as in peripheral blood compared to donor cells from WT mice. In in vitro culture, CaMKK2-/- myeloid progenitors are more prone to form CFU-G colonies, and differentiate into Gr1+CD11b+ cells. Similar to adipogenesis, although CaMKK2 is detectable in HSC and CMP, it is down regulated as cell differentiate and no longer detectable in mature granulocytes. The fact that CaMKK2-/- myeloid progenitors expressed higher level of C/EBPα and PU.1 proteins may explain why CaMKK2 leads to the myeloid lineage commitment in HSCs and myeloid progenitors differentiation.

Adding back functional CaMKK2 can reverse this phenomenon while none of its downstream targets, including CaMKI, CaMKIV, and AMPK, are responsible for this function. Thus, CaMKK2 is concluded to play a cell-intrinsic role in restricting granulocyte differentiation in myeloid progenitors independently of CaMK or AMPK.

CaMKIV is another downstream target of CaMKK2 that has been studied in the hematopoietic process. (Kitsos et al., 2005) The CaMKIV pathway includes the phosphorylation of CREB and interaction with the CREB-binding protein and Bcl-2. It is also involved in regulating the survival and function of HSC, and the lifespan, differentiation, maturation and function of several mature hematopoietic cells including

DCs, thymocytes, and T cells. These evidence indicate a broader role of CaMKIV in hematopoiesis. However, since most of the experiments are done with CaMKIV knock-

out mice which have a severe hematopoiesis defect, it may not be a good model to illustrate the downstream effect of CaMKK2.

1.2.5 CaMKK2 in Cancer

The relation between CaMKK2 and cancer was initially studied in prostate cancer (PC). (Frigo et al., 2011; Massie et al., 2011) PC is one of the sex hormone-affected malignancies and is heavily regulated by the androgen receptor (AR) signaling pathway.

The binding of the sex hormones including testosterone and dihydrotestosterone to the

AR can drive this transcription factor to translocate into the nuclei and regulates the expression of genes that involve cell survival, cell growth, and metastasis. (Isaacs &

Isaacs, 2004) Androgen-deprivation therapy (ADT) is an effective therapy for advanced

PC patients. It can efficiently control the symptoms of the disease and decrease the cancer biomarker level of prostate-specific antigen (PSA). Similar to other chemotherapies, long-term exposure to ADT can induce refractory PC cells, resulting in castration-resistant prostate cancer (CRPC). (C. D. Chen et al., 2004) However, although the CRPCs are insensitive to ADT, the AR-mediated signaling is still found to be active within the cancer cells to maintain the cell survival and progression.

While CaMKK2 is hardly detectable in normal prostate tissue, the increased expression of CaMKK2 is confirmed in PC samples in multiple independent clinical trials, and is further elevated along with disease progression. (Karacosta, Foster,

Azabdaftari, Feliciano, & Edelman, 2012; Karacosta et al., 2016; Massie et al., 2011; Shima et al., 2012) The increased expression of CaMKK2 is closely correlated with the advance stages of the malignancy and poor prognosis. By analyzing the AR-regulated genes in the androgen-sensitive LNCaP adenocarcinoma cell lines, CaMKK2 has been revealed to be a primary gene regulated by AR. CaMKK2 also has the androgen-responsive element

(ARE) in its promoter, therefore its expression can be up-regulated by androgen stimulation, and controls the proliferation and metabolism of PC cells according to AR signals. (Nelson et al., 2002) Various means such as blocking CaMKK2 with small interfering RNA (siRNA) or its pharmacological inhibition STO609, or blocking its downstream target protein AMPK, can prevent the androgen-driven migration and invasion of the PC cells. (Frigo et al., 2011; Karacosta et al., 2012; Massie et al., 2011)

During the analysis the genomic sequencing of AR-regulated genes in two different PC cell lines, LNCap and VCap, AR was found to regulate the aerobic glycolysis and anabolic synthesis in prostate cancer cells by stimulating the expression of key anaerobic enzymes. This was meadiated by the CaMKK2-AMPK-mTOR pathway, but not the

CaMKs-CREB pathway. (Frigo et al., 2011) These data point to CaMKK2 as the key player in regulating the metabolic changes in both hormone-sensitive PCs and CRPCs.

In a later study, AR-dependent stimulation not only increase CaMKK2 expression in

LNCap cells, but also promote CaMKK2 translocation from cytoplasm into the

perinuclear or nucleus to further increase the level of AR-regulated transcriptional activity and protein level to control cell cycle and proliferation. (Karacosta et al., 2012)

Thus, AR and CaMKK2 together form a feedback cycle to promote PC cell growth and progression.

Another important mechanism for CaMKK2-mediating PC progression is the activation of macrophages. By comparing the normal and the cancer-containing area of the prostate specimens from the same PC patients, macrophages infiltration has been found in cancer lesions in almost all the cancer samples, and the interaction between macrophages with the stroma cells has also been noted. (Luo et al., 2007; Nonomura et al., 2011; Zhu et al., 2006) Many publications have reported that CaMKK2 could trigger the activation of macrophages by regulating its downstream targets, CaMKI and AMPK in TLR signal cascade. (Racioppi et al., 2012) The inflammatory response benefits tumor growth by generating a unique ecosystem of hypoxia and accumulated cytokines that favors malignant cell growth. These inflammatory cytokines can also attract the myeloid progenitors and induce tumor-associated macrophages (TAMs), blunt the normal maturation and function of antigen presentation cells, as well as suppress T cell function. (Racioppi, 2013) Blocking CaMKK2 using its pharmacological inhibitor STO609 can reduce the phosphorylation of CaMKI and AMPK, limit the macrophage activation in response to pathogen-associated or damage-associated molecule patterns (PAMPs

and DAMPs), thus attenuating the accumulation of inflammatory cytokines secreted by activated macrophages. (Karacosta et al., 2012) Recent research has also shown that

CaMKK2 can be regulated by miRNA-224 in PC cell lines, indicating that CaMKK2 may be a crucial target regulated by multiple mechanisms to control the PC progression. (Fu et al., 2015)

Besides PCs, the role of CaMKK2 in regulating other cancers has also been recently reported. In breast cancer, another sex hormone-regulated malignancy,

CaMKK2/CaMKI has been found to facilitate the growth of the MCF7 human cell line by activating the ERK pathway, and this effect can be blocked by STO-609. (Fox et al.,

2013; Rodriguez-Mora, LaHair, McCubrey, & Franklin, 2005; Schmitt, Abell, Wagner, &

Davare, 2010) CaMKK2/CaMKs can also control MCF7 cell growth through interacting with the cell cycle progression, and acting as a downstream kinase of Purkinje cell protein 4/peptide 19 (PCP4/PEP19) to prevent MCF cell apoptosis. (Hamada et al., 2014)

AMPK-mTOR pathway has been demonstrated as the key regulator in breast cancer progression. Although LKB1 is closely associated with AMPK activation in breast cancer,

CaMKK2 is still considered an important regulator given its high affinity to AMPK, especially in the LKB1-deficient mutants. A very recent report also revealed the role of

CaMKK2 in gastric adenocarcinoma. (Subbannayya et al., 2015) CaMKK2 expression was found 7-fold higher in human gastric tumor tissue in the quantitative proteomics

analysis, and confirmed in 94% of the gastric cancer cases by immunochemistry (IHC) labeling. Silencing CaMKK2 by using siRNA decreases AMPK phosphorylation and results in the decreased gastric cancer cell proliferation. Similar result has been observed in hepatocellular carcinoma (HHC). (Lin et al., 2015) In human HHC samples, upregulation of CaMKK2 is confirmed by microarray and IHC, and the expression level of

CaMKK2 is inversely correlated with patient survival. A higher expression of CaMKK2 has been found in all eight hepatic cancer cell lines tested in the study. Loss of CaMKK2 function through gene ablation or pharmacological inhibition can inhibit hepatic cancer growth. And this effect is mediated by CaMIV/mTOR/S6K pathway rather than AMPK.

All these data show that CaMKK2 is likely to be a central metabolic regulator controlling metabolism, cancer cell growth and progression in various cancers.

1.3 STO-609

STO-609 is a selective and cell-permeable inhibitor of CaMKK through ATP competition. STO-609 can selectively inhibit the activities of both CaMKK1 and

CaMKK2, but no significant effect on down stream kinases including CaMKI and

CaMKIV. (Tokumitsu et al., 2002) The sensitivity of CaMKK2 to the compound is ~5 folds higher than that of CaMKK1. This difference is caused by a single amino acid substitute in these two isoforms – Val269 in CaMKK2, and Leu263 in CaMKK1.

(Tokumitsu, Inuzuka, Ishikawa, & Kobayashi, 2003) Because of the specificity of STO-

609 to CaMKK2 and the established CaMKK2-AMPK pathway, STO-609 has been often used to identify the role of AMPK in different cancer models. STO-609 has been tested to suppress tumor growth in cell lines and mice models of prostate cancer, colon cancer, breast cancer, and hepatic cancer. The use of STO-609 in mice also increases bone mass by regulating the differentiation of osteoblast and suppressing osteoclast as describe before.(Cary et al., 2013; Pritchard et al., 2015)

1.4 Lymphoma

Lymphoma is a malignancy of lymphocytes. Because of the wide distribution of lymph tissue all through the body, lymphoma can theoretically begin anywhere.

Hodgkin lymphoma and non-Hodgkin lymphoma (NHL) are two main types of lymphoma and can occur in any age. While Hodgkin lymphoma is more lymphocytic origin and more uniform, NHL has many different types that form from different types of white blood cells. The treatment and the prognosis of lymphoma depend on the type and stage of the diseases. The conventional treatment includes chemotherapy, radiotherapy, immunotherapy, and even surgery. (Arulogun, Hertzberg, & Gandhi,

2016; Multani, White, & Grillo-Lopez, 2001; Xu & Fedoriw, 2016) Although primary central nervous system lymphoma (PCNSL) is rare, lymphoma can sometimes migrate to the central nervous system (CNS) and forms secondary CNS lymphoma (SCNSL).

(James L. Rubenstein, 2013; B. L. Liu, Cheng, Zhang, Zhang, & Cheng, 2009; Schmitz &

Wu, 2015) The CNS lymphomas in general are hard to treat and prone to relapse. In some types of lymphomas, dendritic cells (DCs) are found to accumulate in the lesion, and the immature DCs are linked to the progression of diseases. (C. L. Berger et al., 2002;

Martin-Martin et al., 2015; Romano et al., 2015; Y. Sato et al., 2015; Schlapbach et al.,

2010; Tadmor, Fell, Polliack, & Attias, 2013) Several clinical trials infused dendritic cells pulsed with tumor lysates in vitro into patients to stimulate T-cell response. These DC vaccines with increased DC function have been shown to efficiently control lymphoma progression. (Maier et al., 2003; Suehiro et al., 2015; Timmerman et al., 2002) Immune checkpoint molecules including program death-1 (PD-1) and its ligand PD-L1, and cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) are also found increased in various lymphoma patients and animal models. (Armand, 2015a, 2015b; F. J. Hsu &

Komarovskaya, 2002) In the recent clinical trials, PD-1 blockades have been shown as a substantial therapeutic activity in refractory Hodgkin’s lymphoma, diffuse large B-cell lymphoma, and other hematopoitic malignancies. (Ansell et al., 2015; Armand et al.,

2013; R. Berger et al., 2008; Villasboas & Ansell, 2016; Westin et al., 2014)

1.5 Research Objective

Based on the evidence above, we can conclude that CaMKK2 and its downstream targets are most likely involved in regulating the cellular metabolism in different cancer cells, mediating inflammation, and inhibiting several cell types differentiation. However,

the cross-talk between these elements in cancer microenvrionment has not been investigated. For example, the role of CaMKK2 in cancers is only studied within cancer cells but not the microenvironment. The studies of CaMKK2/AMPK mediated inflammation are based on LPS-induced models or infection models, while whether the phenomenon also exists in cancer is unknown. The inhibitory effect of CaMKK2 on cell differentiation is studied in knock-out mice in physiological condition, yet no information available if CaMKK2 is also involved in the distorted immune cells maturation in a malignancy. From the point of view of immunology, cancer can be considered as a chronic disease with the following characteristics: 1) cells exhibiting abnormal cellular metabolic status, 2) the presence of an inflammatory environment, and

3) suppressed immune cell differentiation and activation. CaMKK2 has been implicated to mediate every part of these three elements. Thus, the role of CaMKK2 in cancer progression may have been under-appreciated. Furthermore, the current evidences of

CaMKK2 in cancers are only derived from solid tumor models. The role of CaMKK2 in hematopoietic malignancies has not yet been investigated.

To advance the knowledge of CaMKK2 in tumor microenvironment, especially in lymphoma, we use E.G7 cell line, a commonly used murine lymphoma cell line, as our cancer model, to study the microenvironment changes especially myeloid cell differentiation and function under cancer circumstance. We hypothesize that CaMKK2

has a major role in shaping a tumor-protective microenvironment by suppressing myeloid cell development and function. In Chapter 3, we use two different in vivo lymphoma models in CaMKK2-/- mice to show that CaMKK2 regulates lymphoma growth through the myeloid compartment. In Chapter 4, we further investigate the mechanism how CaMKK2 impacts myeloid cell development and function in lymphoma. To prove that the impact of CaMKK2 on tumor growth and microenvironment is not lymphoma-specific, we validate the data in other two tumor models including breast cancer and myeloma in Chapter 5, and using the CaMKK2 inhibitor STO-609 on lymphoma and breast cancer to suppress cancer growth in Chapter

6. In the last chapter, we conclude that CaMKK2 regulates lymphoma progression by reshaping the cancer environment, and discuss the future study direction for CaMKK2 as a potential target in cancer immune therapy.

2. Experimental Procedures

2.1 Methods for Establishing in vivo Lymphoma Model in CaMKK2 Mice

2.1.1 E.G7 Cell Culture

E.G7 cells were purchased from American Type Tissue Culture Collection

(Rockville, MD, USA). E.G7 cells are derived from EL4 C57BL/6 (H-2 b) mouse lymphoma cell line EL4 and stably transfected with OVA expressing plasmid. E.G7 cells were further stably transfected with GFP-Luciferase expressing plasmid by Dr. Yiping

Yang’s lab. Cells were cultured in RPMI 1640 (Gibco, MA, USA) supplied with 4.5g/L glucose (Sigma-Aldrich, MO, USA), 2mM L-glutamine, 1.5g/L sodium bicarbonate,

10mM HEPES, 1.0mM sodium pyruvate (all from Gibco, MA, USA), 0.05mM 2- mercaptoethanol, 0.4mg/ml G418 (Sigma-Aldrich, MO, USA), and 10% fetal bovine serum (Hyclone, MA, USA). Cells were kept in humidified 37°C CO2 incubator.

2.1.2 Mice

C57BL/6 mice were purchased from the Jackson Laboratory (CA, USA). Wild- type (WT) and CaMKK2 global knock-out (Camkk2-/-) mice were generated as described

(Anderson, Ribar, 2008) and back-crossed to C57BL/6 for more than 7 generations.

Tg(Camkk2-EGFP)DF129Gsat mice were originally provided by the Mutant Mouse

Regional Resource Centers (MMRRCC) and back-crossed to C57BL/6 for more than 7 generations. (Gong et al., 2003) In general, this transgenic strain has the coding sequence for enhanced green fluorescent protein (EGFP) inserted into the mouse genomic 36

chromosome at the ATG transcription initiation codon of the CaMKK2 gene. Thus, the expression of the EGFP reporter is driven by the CaMKK2 gene. LysMCre+;CaMKK2fl/fl and its genotype control LysMCre+;CaMKK2+/+ were generated by crossing B6.129P2-

Lyz2tm1(cre)Ifo/J mice from the Jackson Laboratory with CaMKK2lox/lox mice. All the mice were used between 8-16 weeks, with gender and age matched in experimental groups and control groups. The animals were housed in Duke University animal facilities under a 12 hour light/12 hour dark cycle with food and water ad libitum provided. All animal care and experimental procedures were approved by National Institute of Health and

Duke University Institutional Animal Care and Use Committee, and in compliance with the guidelines.

2.1.3 Intracranial Tumor Model

The intracranial tumor model was generated with the help of Dr. David Synder.

In brief, 1×104 E.G7 cells were suspended in 25µl 10% methylcellulose and stereotactically implanted into the right caudate nucleus of the brain 8-12 weeks

CaMKK2 WT and KO mice as previously described. (Sampson et al., 1999) Mice were monitored by physical examination every other day and bioluminescence imaging each week. The moribund mice with severe hunching and decreased activity were sacrificed.

2.1.4 Flank Tumor Model

1×105 E.G7 cells were suspended in 100 µl PBS and implanted subcutaneously to the right side flank of the mice. Tumors were monitored using bioluminescence imaging

each week. Tumor sizes were measured with calibrator every 3-4 days and calculated as:

Length × Width × Width /2. In some experiments 5×105 cells were used as indicated.

Mice were sacrificed when tumor reached 2 cm3.

2.1.5 ELISpot Analysis

CaMKK2 WT and KO mice were injected with E.G7 cells subcutaneously to generate tumor models. On day 21, mice were sacrificed and the spleens were removed using sterilized scissors. Single-cell splenocyte suspension was generated as previously described. 5×105 splenocytes were suspended in 200 µl complete DMEM media with 10%

FBS as described before and seeded into mouse IFN-gamma ELISpot kit (R&D, MN,

USA). OVA protein was added at 1mg/ml as stimulator. Cells were cultured in a humidified incubator with 5% CO2 for 24 h, and reactions were processed as instructed.

The plate was read using a CTL-ImmunoSpot® S6 FluoroSpot Line (Cellular Technology

Ltd, OH, USA) and the data was analyzed with ImmunoSpot software (Cellular

Technology Ltd, OH, USA).

2.1.6. Tumor Component Analysis

2.1.6.1 Tumor Pathology Staining

Tumor-bearing mice were sacrificed in CO2 chamber. The tumors were removed using sterilized sectioning scissors and forceps. The tissues were fixed in formaldehyde and further embedded in paraffin. The slides were cut at 7µm and stained with H&E and examined under microscope.

2.1.6.2 Tumor Digestion

The removed tumors were further minced with a surgical scalpel. The tissue were added to 5ml HBSS w/ Ca2+ and Mg2+ (ThermoFisher Scientific, MA, USA) supplied with collagenase 2 mg/ml and DNase 0.1 mg/ml (Roche, Basel, Switzerland) in gentleMACS C tubes and dissecting with gentleMACS dissociator (Miltenyi, Bergisch

Gladbach, Germany). The samples were incubated in 37°C for 15min and passed through the 70 µm cell strainers (Coring, NY, USA). Cells were washed twice in PBS w/

2% FBS before staining.

2.1.6.3 Flow Cytometry Analysis

For myeloid cells staining, 5×106 cells were incubated with FITC anti-mouse

CD11c, PE anti-mouse/human CD11b, PE-Cy7 anti-mouse Ly6G, APC anti-mouse F4/80

(all from BioLegend, CA, USA), PerCpCy5.5 anti-mouse Ly6C, Fixable Viability Dye eFluor® 450 (eBioscience, CA, USA) at 4°C for 15min in dark and washed with PBS w/

2% FBS. The stained cells were analyzed using a BD FACSCanto flow cytometer (BD, NJ,

USA). Data is analyzed using FlowJo (TreeStar, OR, USA).

2.1.7 In Vivo Bioluminescence Imaging

A fresh stock solution of D-Luciferin (Perkin Elmer, MA, USA) was diluted in

DPBS at 30mg/ml and filtered sterilize through 0.2µm filter. Each mouse was injected with 100µl stock solution intra-peritoneally 10-15min before imaging. Mice were anesthetized with isoflurane (Butler-Schein, OH, USA) and shaved with a shaving

trimmer (Wahl, IL, USA) and imaged with IVIS Lumina XR (Perkin Elmer, MA, USA) for 1min. Image was analyzed with IVIS Lumina LivingImage Software (Perkin Elmer,

MA, USA).

2.1.8 T-Cell Depletion Treatment

CaMKK2 KO mice were injected peritoneally with anti-mouse CD8 monoclonal antibody or its IgG isotype (BioXCell, NH, USA) each at 200µl at the concentration of

1mg/ml at 4 days and 1 days before tumor inoculation, and every 3 days after tumor cell injection. The effect of CD8 depletion was monitored by flow cytometry using 50µl peripheral blood staining with APC anti-mouse CD8, APC-Cy7 anti-mouse CD45.2

(Biolegend, CA, USA) at the same time. The mice were injected with 1×105 E.G7 cells to the right flank subcutaneously on day 0. The tumor size was monitored with a calibration every 3-4 days. Tumor volume was calculated as indicated before. The experiment was terminated when the tumor reached 2 cm3.

2.1.9 STO-609 Treatment

STO-609 was purchased from TORIS Bioscience (Bristol, UK). The C57BL/6 mice were purchased from the Jackson Laboratory (CA, USA). The mice were inoculated with

5×105 E.G7 cells to the right flank subcutaneously on day 0. On day 7, the tumor growth was measured using bioluminescent imaging. All the mice with visible tumor under luciferin detection were randomized into STO-609 treatment group or DMSO control group. The mice were injected peritoneally with 200µl STO-609 at concentration of

20mM or DMSO as control every 3-4 days for two weeks. The tumor size was monitored using a calibrator every 3-4 days. The experiment was terminated when the tumor reached 2 cm3.

2.2 Methods for Establishing Other in vivo Cancer Models in CaMKK2 Mice

2.2.1 E0771 Breast Cancer Cell Line and Breast Cancer Model

E0771 mouse breast cancer cell line was purchased from ATCC. Cells were cultured in RPMI-1640 (Gibco, MA, USA) supplied with 2mM L-glutamine, 1.5g/L sodium bicarbonate, 1.0mM sodium pyruvate (all from Gibco, MA, USA), and 8% fetal bovine serum (Hyclone, MA, USA). Cells were kept in humidified 37°C CO2 incubator.

Cells were passed at 70% confluence.

For tumor inoculation, cells were detached from the flasks using 0.25% trypsin-

0.53mM EDTA solution. Cells were washed with PBS and resuspended in plain PBS at the concentration of 2×106 cells/ml. For each female mouse, 2×105 cells in 100µl cell suspension were injected subcutaneously into the first breast fat pat on the right. Tumor growth was measured using a calibrator every 3-4 days. Mice were sacrificed when the tumor reached 2cm3.

For STO-609 treatment, the same strategy and procedure was used as in mice inoculated with E.G7 lymphoma cells. For tumor component analysis, the mice with tumor were sacrificed in CO2. Tumors were dissected and digested and stained for flow cytometry analysis as described before. 41

2.2.2. Vk*Myc Myeloma Cells and Myeloma Model

Vk*Myc myeloma cells were precious gift from Dr. Marta Chesi. The generation and characteristics of the cells were described in publication. (Chesi et al., 2008) Tumor cells were purified using Ficoll (Cedarland, Ontario, Canada) from the spleens of

C57BL/6 mice that were inoculated with myeloma cells and showed detectable M-spike.

For tumor inoculation, 5×105 cells were injected through tail vein into each mouse.

Peripheral blood was collected from maxillary sinus every week since Week 4. Serum was collected from the blood by high speed centrifuging, and gamma immunoglobulin was detected using protein electrophoresis and staining (Helena Laboratories, Texas,

USA). Mice were monitored every day and sacrificed when become moribund.

2.3 Methods for Establishment of in vitro Lymphoma-Exposed Cell Model with Cells Derived from CaMKK2 Mice

2.3.1 Generating Bone Marrow Derived Dendritic Cells

The CaMKK2 WT and KO mice were sacrificed in a CO2 chamber and sterilized with 70% alcohol. The femur and tibia bones were removed using the sterilized sectioning scissors and forceps and further crunched in 5ml PBS with 2% FBS and 2mM

EDTA using a ceramic molar. The liquid was further passed through a 70µm cell strainers (Coring, NY, USA) to form the single cell suspension. After spinning down the cells, the erythrocytes were lysed using the red blood cell lysis buffer (Biolegend, CA,

USA) followed by two washes in PBS with 2% FBS and 2mM EDTA. The cells were seeded to 6-well plates (Coring, NY, USA) at 5×106 per well in 5ml DMEM (Gibco, MA,

USA) supplied with 10% FBS, 2mM L-glutamine, 1.5g/L sodium bicarbonate, 10mM

HEPES, 1.0mM sodium pyruvate (all from Gibco, MA, USA), 10% fetal bovine serum

(Hyclone, MA, USA), 10nM recombinant mouse GM-CSF and 10nM recombinant mouse

IL-4 (both from Biolegend, CA, USA), and with or without 50% E.G7 supernatant. The media was half-changed on day 3. The floating cells were collected on day 5 for flow cytometry analysis and further enriched by anti-mouse CD11c Microbeads (Miltenyi,

Bergisch Gladbach, Germany). The cells with the purity > 85% were used. In some experiments as indicated, the cells were stained with FITC anti-mouse CD11c, PE anti- mouse/human CD11b (Biolegend, CA, USA). Double positive cells were sorted with BD

FACS Aria II cell sorter (BD, NJ, USA).

2.3.2 Generating Myeloid Derived Suppressor Cells

Bone marrow cells were prepared as described before. Cells after RBC lysis were seeded to 6-well plates (Coring, NY, USA) at 1×106 per well in 5ml DMEM (Gibco, MA,

USA) supplied with 10% FBS, 2mM L-glutamine, 1.5g/L sodium bicarbonate, 10mM

HEPES, 1.0mM sodium pyruvate (all from Gibco, MA, USA), 10% fetal bovine serum

(Hyclone, MA, USA), 40nM recombinant mouse GM-CSF and 40nM recombinant mouse

IL-6 (Biolegend, CA, USA), and with or without 50% E.G7 supernatant. The supernatant was half-changed on Day 3. The floating cells were collected on Day 5 and enriched by anti-mouse Gr-1 Microbeads (Miltenyi, Bergisch Gladbach, Germany). The cells with the purity > 90% were used. In some experiments as indicated, the cells were stained with

APC anti-mouse Gr-1, PE anti-mouse/human CD11b (Biolegend, CA, USA). Double positive cells were sorted with BD FACS Aria II cell sorter (BD, NJ, USA).

2.3.3 Coculture of BMDC and T cells

2.3.3.1 Primary Cell Separation

BMDCs were prepared as described. T cells were separated from spleens of

C57BL/6 mice. The spleens were removed using sterilized sectioning scissors and smashed with the end of the plugs from the 1.5ml syringes on a 70µm cell strainers

(Coring, NY, USA). Cells were flushed into 50ml centrifuge tubes (BD falcon, MA, USA) using PBS with 2% FBS and 2mM EDTA. After spinning down the cells, the erythrocytes were removed using the red blood cell lysis buffer (Biolegend, CA, USA) followed by two washes. The T cells were further enriched using the Pan T Cell Isolation Kit II

(Miltenyi, Bergisch Gladbach, Germany). A fraction of BMDCs and T cells were stained with FITC anti-mouse CD11c or FITC anti-mouse CD3 (Biolegend, CA, USA) respectively, and detected by flow cytometry for enrichment purity. BMDCs with

CD11c+ purity > 90% and T cells with CD3+ purity > 95% were used.

For antigen-specific T cells, C57BL/6-Tg(TcraTcrb)1100Mjb/J (OT1) or B6.Cg-

Tg(TcraTcrb)425Cbn/J (OT2) mice (the Jackson Laboratory, ME, USA) at 8-16 weeks were used as T-cell donors. T cells were enriched as described above and further stained with PE Mouse Vα2 TCR rat anti-mouse mAb (Invitrogen, CA, USA), and APC anti-

mouse CD4 or CD8 mAb (BioLegend, CA, USA) accordingly. The double-positive cells were sorted with the BD FACS Aria II cell sorter (BD, NJ, USA).

2.3.3.2. T-cell CFSE Labeling

After two washings in PBS with 2% FBS and 2mM EDTA, 1×107 T cells were resuspended in 1 ml plain PBS with 2µM CFSE (MitoSciences, OR, USA) and incubated in 37°C for 20 min. 35ml RPMI with 10% FBS were subsequently added to the cells and incubated for another 10min. After the incubation, the cells were washed twice in 2%

FBS and 2mM EDTA. Cell number was determined.

2.3.3.3 Seeding of Cells

1×105 CFSE-labeled T cells and 2×104 BMDCs were seeded into 96-well plates

(Corning, NY, USA) with 200µl complete DMEM media with 10% FBS as described before. Purified NA/LE hamster anti-mouse CD3e (BD Biosciences, CA, USA) was added to the culture media at 0.1 ug/ml.

For antigen-specific T cells, OVA 245-261 or OVA 323-339 peptide (Invivogen,

CA, USA) was added accordingly to the culture media at 0.01ug/ml.

For IL-12 neutralizing experiment, purified NA/LE rat anti-mouse IL-12

(p40/p70) (BD Pharmingen, CA, USA) was added to the culture media at 0.2 ng/ml.

Cells were cultured in humidified incubator with 5% CO2 for 24 h, 48 h, or 72 h as indicated in different experiments and further analyzed using flow cytometry.

2.3.4 Cytokine Secretion Analysis

After 24 h of coculturing, 100 µl supernatant was carefully collected to a new 96- well plate. The plate was sealed with an adhesive seal (BioRad, CA, USA) and stored in

-20 °C. The supernatant was analyzed with MILLIPLEX® MAP Mouse

Cytokine/Chemokine Magnetic Bead Panel (EMD Millipore, Darmstadt, Germany) using

Luminex® (Luminex, TX, USA).

2.3.5 Flow Cytometry

2.3.5.1 Myeloid Cells Staining

For MDSC and BMDC, only floating cells were collected. 1×106 cells were spun down in BD Falcon round-bottom polysterene tubes (BD, NY, USA). The cell pellet was incubated with 50µl PBS with 2% FBS with FITC anti-mouse CD11c, PE anti- mouse/human CD11b, PE-Cy7 anti-mouse Ly6G, APC anti-mouse F4/80, APC-Cy7 anti- mouse I-A (all from BioLegend, CA, USA), PerCP Cy5.5 anti-mouse Ly6C, and Fixable

Viability Dye eFluor® 450 (eBioscience, CA, USA) at 4°C for 15min. Cells were washed once with 2ml PBS with 2% FBS and fixed with 200µl 1.5% PFA in 4°C till analysis.

2.3.5.2 Costimulatory Factor Staining

Cells were treated as indicated above and stained with FITC anti-mouse CD80,

PE anti-mouse CD274, PE-Cy5 anti-mouse CD40 (all from eBioscience, CA, USA), PE-

Cy7 anti-mouse/human CD11b, APC anti-mouse CD86, and APC-Cy7 anti-mouse

CD11c (all from BioLegend, CA, USA). Cells were fixed in 200µl 1.5% PFA in 4°C till analysis.

2.3.5.3 T Cells Staining

All the cells from 96-well plates were collected to flow tubes and spun down. The cell pellets were stained with PerCp-Cy5.5 anti-mouse CD69, APC anti-mouse CD8a,

APC-Cy7 anti-mouse CD4 (Biolegend, CA, USA), PE-Cy7 anti-mouse CD25, and Fixable

Viability Dye eFluor® 450 (eBioscience, CA, USA) at 4°C for 15min. Cells were further penetrated using the FoxP3 staining buffer set (eBioscience, CA, USA) as instructed and stained with PE anti-mouse/human FoxP3 (Biolegend, CA, USA). Cells were fixed in

200µl 1.5% PFA in 4°C until analysis.

2.3.5.4 Cell Analysis

The stained cells were analyzed using a BD FACSCanto flow cytometer (BD, NJ,

USA). Flow-Count Fluorospheres (Beckman Coulter, CA, USA) were added to the sample for absolute count measurement. Data was analyzed using FlowJo (TreeStar, OR,

USA).

2.3.6 Cytospin and Staining

MDSCs were generated as described. On Day 5, the floating cells were carefully collected and enriched by anti-mouse Gr-1 Microbeads (Miltenyi, Bergisch Gladbach,

Germany). The enriched MDSCs were washed twice in plain PBS, resuspended at the concentration of 2×104 cells/200µl in complete RPMI supplied with 50% FBS, and

carefully added to the assemblied cuvettes with mounted slides. The cells were spun down at 500rpm for 5min. The slides were carefully detached from the cuvettes, dried in air for 30min and followed by 4% paraformaldehyde (PFA) fixation for another 30min at room temperature. The slides were stored in plain PBS at 4°C until the Giemsa staining

(Sigma-Aldrich, MO, USA) was performed as instructed.

2.4 Methods for Molecular Biology

2.4.1 Western Blot

To generate dry cell pellets, cells were spun down in Eppendorf tubes at 2000 rpm for 10min with supernatant carefully removed. The cell pellets were stored at -80°C.

To generate cell lysate, Halt Protease Inhibitor Cocktail and M-PER Mammalian Protein

Extraction Reagent (both from ThermoFisher Scientific, MA, USA) were added to the dry pellets as instructed. The lysate was sonicated (QSonica, CT, USA) and spun down at 14,000 rpm for 10 min. The supernatant was carefully taken and aliquoted to clean

Eppendorf tubes. SDS 10× loading buffer was added to the lysate and boiled at 85°C for

5min. 50µl sample was added to each well of the Mini-PROTEAN® TGX™ Precast Gels

(Bio-Rad, CA, USA) and run in SDS Tris/Glycine buffer (ThermoFisher Scientific, MA,

USA) at 120V. The protein was transferred to 0.2µm nitrocellulose membrane using

Trans-Blot® Turbo transfer system (Bio-Rad, CA, USA). The membrane was blocked in

Odyssey TBS blocking buffer (LI-COR Biosciences, NE, USA) at room temperature for

30min and incubated in the first antibody at 4°C for overnight. All the first antibodies

were monoclonal, and used at 1:1000 dilution in the blocking buffer except anti-β actin at 1:5000. The first antibodies used are listed: purified mouse anti-CaM kinase kinase

(BD Biosciences, CA, USA), anti-S6 ribosomal protein (54D2) mouse mAb, anti-acetyl-

CoA carboxylase (C83B10) rabbit mAb, anti-phosphor-ACC (Ser79 D7D11) rabbit mAb, anti-phosphor-p38 MAPK (T180/Y182 D3F9) rabbit mAb, anti-phosphor-AMPK alpha

(T172 40H9) rabbit mAb, anti-AMPK alpha (F6) mouse mAb, anti-β actin mouse mAb

(all from Cell Signaling, CA, USA). After the incubation, the membranes were washed 3 times in TBST (Cell Signaling, CA, USA) for 5 min, and further incubated in secondary antibodies anti-mouse IgG Alexa Fluor 680 (Invitrogen, CA, USA) or anti-rabbit IgG

IRDye800 conjugated (Rockland Immunochemicals, PA, USA) accordingly at 1:5000 dilution at room temperature for 30 min, followed by 3 washes with TBST for 5min. The image was taken using Odyssey CLx (LI-COR, NE, USA) and analyzed using

ImageStudio (LI-COR, NE, USA).

2.4.2 Real-Time Quantitative RT-PCR Assay

Cell pellets were generated as described before. RNA was extracted using

RNeasy Mini Kit (Qiagen, Hilden, Germany) and checked for quality using a NanoDrop

2000c (Thermo Scientific, DE, USA). To synthesize first-strand cDNA, RNA sample was mixed with 10× DNaseI buffer and DNaseI at room temperature for 30min and then with 25mM EDTA at 65°C for another 10min; Oligo(dT) primer and dNTP mix were added and incubated at 72°C for 10min, and eventually with 5× first strand buffer, DTT,

RNaseOut ribonuclease inhibitor and MLV reverse transcriptase (all reagents from

Invitrogen, CA, USA) at 37°C for 50min and subsequently 70°C for 15min. Quantitative real-time PCR was performed using iQ SYBR Green Supermix (Bio-Rad, CA, USA) with respective primers and cDNA and run by the CFX96 Real-Time System (Bio-Rad, CA,

USA).

Table 1: Primer sequences for real-time PCR assay

Camkk2-Fwd 5’-CATGAATGGACGCTGC-3’

Camkk2-Rev 5’-TGACAACGCCATAGGAGCC-3’

Mouse GAPDH-Fwd 5’-CGACTTCAACAGCAACTCCCACTCTTCC -3’

Mouse GAPDH-Rev 5’-TGGGTGGTCCAGGGTTTCT TACTCCTT-3’

LysMCre-Fwd 5’-CCCAAGAAGAAGAGGAAGGTGTCC-3’

LysMCre-Rev 5’-CCCAGAAATGCCAGATTACG-3’

PCG-1α-Fwd 5’-TCACCCTCTGGCCTGACAAATCTT -3’

PCG-1α-Rev 5’-TTTGATGGGCTACCCACAGTGCT -3’

2.4.3 RT2 Profiler PCR Arrays

CaMKK2 WT and KO BMDCs were generated from individual mice (n=5) and pooled during the generation of cell pellets. RNA was extracted using RNeasy Mini Kit followed by DNase clean-up using a RNase-Free DNase Set (Qiagen, Hilden, Germany) and checked for quality using a NanoDrop 2000c (Thermo Scientific, DE, USA) before

proceeding to reverse transcription. cDNA was generated using RT2 First Strand Kit

(Qiagen, Hilden, Germany) as instructed, and further mixed with RT² SYBR Green qPCR

Mastermix (Qiagen, Hilden, Germany) and evenly distributed to the 96-well plate of RT²

Profiler™ PCR Array Mouse AMPK Signaling (Qiagen, Hilden, Germany). The reaction was performed using the CFX96 Real-Time System (Bio-Rad, CA, USA). The data was uploaded to the web resource the GeneGlobe Data Analysis Center (Qiagen, Hilden,

Germany) for analysis.

2.5 Statistical Analysis

Statistical analysis was performed using Prism GraphPad (GraphPad Software,

Inc. CA, USA) or Excel (Microsoft, WA, USA). For survival studies, Mantel-Cox test was used. For tumor size and bioluminescent measurement, two-way ANOVA test was used. For flow cytometry, real-time quantitative RT-PCR assay, and cytokine detection analysis, student’s t test was used to compare between two groups. Level of significance was set at P < 0.05. Bar graphs represent mean ± SEM.

3. The Role of CaMKK2 in Regulating Lymphoma Growth

3.1 Introduction

CaMKK2, an important kinase that regulates the calcium/camodulin kinase cascade, has been reported to regulate cancer progression in different cancer models including prostate cancer (PC), breast cancer, colon cancer, and hepatic cancer. In these cancer cells, CaMKK2 expression was elevated. It serves as a key player to regulate the cancer cell metabolic status and thus sustain the cancer cell survival and proliferation.

Another important role of CaMKK2 in cancer is mediating macrophage activation to reshape tumor microenvironment. In human PC samples, macrophages have been noted to infiltrate in the tumor and interact with stromal cells. CaMKK2 can suppress the activation of macrophages thus inhibiting the inflammatory response.

Therefore, CaMKK2 generates a tumor-protective microenvironment with enriched anti- inflammatory cytokines to sustain tumor growth and induces tumor-associated macrophages.

Lymphoma is one of the most common hematopoietic malignancies. It is the cancer of lymph tissue that may occur in any part of the body and involves patients of all age. Although most types of the lymphoma occur outside of the central nerve system

(CNS), lymphoma can occur primarily in the CNS or metastasis to the CNS. CNS lymphoma is relatively hard to treat because of the special location. Clinical evidences and animal models have shown the important role of dendritic cell function in

modulating lymphoma progression, indicating that lymphoma may be highly dependent on the microenvironment. Since the focus of CaMKK2 related research in cancer has hitherto been limited to solid tumor cells, here we use E.G7 lymphoma cell, a commonly used EL4-derived murine lymphoma cell line, as our model. By injecting

E.G7 cells subcutaneously or intracranially into CaMKK2 ablation mice, we mimic the different lymphoma models and study how CaMKK2 changes the tumor microenvironment to impact lymphoma progression.

3.2 Result

3.2.1 CaMKK2 ablation suppresses lymphoma growth in vivo

Subcutaneously injecting E.G7 cells into mice is a conventional murine lymphoma model. In our experiments, E.G7 1×105 were injected subcutaneously into the right flank of WT and CaMKK2-/ -(KO) mice. The tumor growth was monitored by measuring the size with calibrators, and bioluminescent imaging every week. In the WT mice, the tumors were palpable since Day 7 and gradually increased in size (Figure 1,

Panel A) and in luciferin photon count (Figure 1, Panel B), while in CaMKK2-/ - mice the tumors maintained statistically smaller (P=0.001) with lower photon counts (P=0.002).

The mice were sacrificed when the tumors reached 2 cm3 as required in animal protocol.

By Day 32, bioluminescent image was performed on the remaining mice. The image showed that the tumors were growing vigorously in WT mice, but only one still

detectable in CaMKK2-/- mice (Figure 1, Panel C). This result indicates that CAMKK2 ablation efficiently suppresses lymphoma in subcutaneous injection model.

A. B. Flank Injected EG7: Tumor volume Flank Injected EG7: Luciferin Count 2000 1×1010 WT WT KO 8×109 KO

3 1500

6×109 1000 P=0.001 P=0.002 4×109 n=10 Volume/mm 500 n=10 Photon Counts 2×109

0 0 0 7 14 17 21 0 7 14 17 21 Days Days

C. Control Luciferase CaMKK2 -/- Luciferase

Figure 1: Tumor growth of mice with E.G7 subcutaneous injection

WT and CaMKK2-/- mice (n=10) were injected with E.G7 1×105 s.c. in the flank. A). The tumors were measured every 3-4 days. P=0.001. B). Bioluminescent imaging was monitored every week. Photon count was calculated by software. P=0.001. C). Bioluminescent image of mice on Day 32.

The lack of tumor in CaMKK2-/- mice could be caused by tumor engraftment failure, or rejection after successful engraftment. To discriminate these two possibilities, tumor tracing for early time points is deemed necessary. In subcutaneous flank model, bioluminescent imaging was performed every 3-4 days after tumor injection (Figure 2).

It appears that in the first two weeks, tumors in CaMKK2-/- mice were detectable and growing, although slower than those in the WT group. After two weeks, while continuous tumor growth was observed in WT mice, the tumor growth in CaMKK2-/- mice gradually suspended and even began to shrink. Bioluminescent imaging not only demonstrated the different growth patterns of E.G7 cells in WT and CaMKK2-/- mice, but also confirmed the successful engraftment of tumor in CaMKK2-/- mice. The gradual tumor shrinkage in CaMKK2-/- mice after the initial two weeks suggests the potential tumor rejection mediated by immune system.

Figure 2: Bioluminescent imaging of mice with E.G7 subcutaneous injection

WT (n=4) and CaMKK2-/- mice (n=5) were injected with E.G7 1×105 s.c. and followed up by bioluminescent imaging every 3-4 days. The tumors in WT group grew larger gradually, while the tumors in CaMKK2-/- group appeared around Day 7-14 and gradually shrinked and disappeared afterwards.

As previously stated, lymphoma can occur or migrate to CNS to form CNS lymphoma, which is usually refractory to conventional lymphoma treatment, resulting in high mortality. Unlike the subcutaneous flank model in which CaMKK2-expressing adjacent parenchymal cells are absent, CaMKK2 is highly expressed in neurons in CNS system. Thus, this intracranial model of lymphoma will be a helpful means to investigate whether CaMKK2-enriched parenchymal cells have an impact on lymphoma growth.

Figure 3: CaMKK2 Mice with E.G7 intracranial injection

WT and CaMKK2-/- mice (n=10) with 1×104 E.G7 cells intracranial injection were monitored every other day. A) Curve to show mice with posture changes. Most WT mice show abnormal posture from day 10-20 after injection. B) Survival curve. Mice were sacrificed when showing moribund. Data was pooled from two individual experiments. Log-rank (Mantel-Cox) Test was used for analysis. P<0.0001.

To answer this question, we used the intracranial lymphoma model. 1×104 E.G7 cells suspended in 25µl methylcellulose were injected into the right caudate nuclei of WT and CaMKK2-/- mice. The mice were monitored by physical examination every other day.

Similar to the subcutaneous injection model, in WT group, mice start to show abnormal gesture and movement including hunching, circling, paralysis, decreased appetite and body weight loss from 10 days after injection (Figure 3, Panel A). Most WT mice were sacrificed before 25 days after injection when they were moribund (Figure 3, Panel B).

Comparing to WT mice, only 1 in 10 CaMKK2-/- mice with E.G7 intracranial injection was sick and needed to be sacrificed before the end of the experiment (Figure 3). The difference of tumor growth in WT and KO is statistically significant (P<0.0001).

To better monitor the tumor growth, weekly bioluminescent imaging was performed on all mice. With the help of imaging, intracranial tumors were clearly shown in WT group on Day 21, yet not in CaMKK2-/- mice (Figure 4, Panel B). On Day 21, the photon count of luciferin signal was significantly different (P<0.001) between the two groups (Figure 4, Panel A). These findings corroborate the conclusion in the intracranial tumor model, indicating that CaMKK2 expression in parenchymal cells does not impact lymphoma progression. The significant difference of tumor growth in the WT and

CaMKK2-/- mice is more likely due to the systemic effect mediated by CaMKK2.

Considering that CaMKK2 is also expressed in macrophages and some hematopoietic progenitors, it is likely that the CaMKK2-ablation impacts the immune responses to suppress the growth of lymphoma.

A Intracranal EG7 Injection: photon count 1010 *** 109 WT KO 108

107

106 Photon Counts 105

104 7 14 21 28 35 Days B Control Luciferase CaMKK2 -/- Luciferase

Figure 4: Bioluminescent imaging of mice with E.G7 intracranial injection

WT and CaMKK2-/- mice (n=10) were injected with 1×104 E.G7 cells i.c. and monitored by bioluminescent imaging every week. A). The photon count was calculated by imaging software. The difference in photon count between two groups was significant on Day 21. P<0.001. B. Bioluminescent image of mice on Day 21.

3.2.2. CaMKK2 ablation in myeloid cells is sufficient for lymphoma suppression

CaMKK2 has been reported to express in macrophages and regulates granulocyte development. To further verify if CaMKK2 regulates lymphoma development through myeloid cells, LyMCre mice were crossed with CaMKK2fl/fl mice to generate mice with conditional CaMKK2 ablation in myeloid lineage,

LyMCre+;CaMKK2fl/fl.

EG7 tumor volume - s.c. 1E5 600 Cre+;+/+ n=8 ) 3 Cre+;fl/fl n=10 400

200 P=0.0004 Tumor volume (mm volume Tumor 0 0 5 10 15 20 25 Days

Figure 5: E.G7 tumor growth in mice with myeloid lineage CaMKK2 ablation

E.G7 1×105 was injected subcutaneously into LysMCre+;CaMKK2fl/fl mice (n=10) and the genotype littermates control LysMCre+;CaMKK2+/+ mice (n=8). Tumor size was monitored every 3-4 days. Similar to E.G7 in CaMKK2-/- mice, tumor growth LysMCre+;CaMKK2fl/fl mice were significantly suppressed comparing to that in WT mice. P=0.0004.

According to literature, LysMCre mice are specific and highly efficient murine model for Cre-mediated deletion of loxP-flanked target genes in myeloid cells. It has been reported to have 83-98% deletion efficiency in mature macrophages, almost 100% in granulocytes, and 16% partial deletion in CD11c+ splenic dendritic cells in double mutant mice. (Clausen, Burkhardt, Reith, Renkawitz, & Forster, 1999) The

LyMCre+;CaMKK2fl/fl mice and the genotype control LyMCre+;CaMKK2wt/wt were challenged with E.G7 cells 1×105 subcutaneously, and monitored every 3-4 days. Tumor sizes were significantly smaller in LyMCre+;CaMKK2fl/fl mice compared to the genotype

control litermates. (Figure 5) The tumor growth curve in LyMCre+;CaMKK2fl/fl mice and the genotype control were similar to the ones in WT and CaMKK2-/- mice. Given that

CaMKK2 ablation in myeloid cells is sufficient for lymphoma suppression, our finding indicates that CaMKK2 regulates lymphoma growth through the myeloid compartment.

3.2.3. CD8 T-cell ablation reverses lymphoma suppression in CaMKK2-/- mice.

Myeloid cells, especially dendritic cells and MDSCs, can impact anti-tumor immunity through regulating T-cell function. T cells, especially CD8 T cells, can be activated by antigen presenting cells (APCs) such as mature DCs to proliferate, expand, and differentiate into memory T cells or effector cells including cytotoxicity T lymphocytes (CTLs) thus control tumor progression. CTLs are generally CD8+ cells, and

CD8 CTLs have been proven to be essential for anti-tumor immunity. (Gajewski,

Schreiber, & Fu, 2013; Janeway CA Jr, 2001; Toes, Ossendorp, Offringa, & Melief, 1999)

To investigate whether T cells participate in CaMKK2-mediated lymphoma growth, we depleted CD8 cells in tumor-bearing mice. Anti-mouse CD8 monoclonal antibody (mAb) was injected into CaMKK2-/- mice 4 days and 1 day before tumor inoculation, and every

3-4 days after the tumor injection, as indicated in Figure 6. The comparable amount of

IgG isotype was used in control group at the same time.

Figure 6: in vivo CD8 depletion strategy

E.G7 1×105 was injected subcutaneously into CaMKK2-/- mice on Day 0. The arrows show the days to inject anti-CD8 mAb or its IgG isotype control i.p.. Generally, the injection for CD8 depletion was performed 4 days and 1 day before, and every 3 to 4 days after the tumor inoculation.

To determine the efficiency of CD8 cell depletion, CD8% in peripheral blood was monitored by flow cytometry at different time points (Figure 7). Before mAb injection,

CD8% in CD45+ cells in the peripheral blood was comparable in the mAb treatment group and control group. After 1 does of injection, CD8% in the control group still maintained at ~10%, while the percentage in the treatment group dropped to less than

1%. The difference was statistically significant (P<0.0001). As the injections continue,

CD8% remained low in treatment group after tumor inoculation, whereas the percentage was comparable to day -5 in control group (P<0.01). CD8 cells in antibody treatment group were successfully depleted.

CD8% in CD45+ cells in blood 15 ** ns **** Anti-CD8 mAb 10 Isotype

5 Percentage (%) Percentage

Day -5 Day -1 Day 4

Figure 7: CD8 depletion confirmed by flow cytometry

Peripheral blood from mice with anti-CD8 depletion and controls was taken on day -5 (before depletion antibodies injection), day -1 (after first-dose antibodies depletion, before E.G7 injection), day 4 (after E.G7 injection). CD8% was determined by flow cytometery. Student t test was used for statistical analysis. Before the CD8 depletion, the CD8% was comparable between the treatment group and control group (p value shows no significance). After antibody injection, the CD8 cells were successfully depleted on day -1 (p<0.0001) and day 4 (p<0.001). n=5.

After two doses of mAb or isotype control injection, E.G7 1×105 were injected s.c. into CaMKK2-/- mice on Day 0. Tumor size was measured every 3-4 days. During the experiment, 1/5 of the mice in control group died of accident and was censored.

Consistent with the previous experiments with identical tumor dose in CaMKK2-/- mice, the tumors in untreated control group and isotype IgG control group were not detectable or remain at very small sizes. However, the tumors in the group with anti-

CD8 mAb treatment develop vigorously (Figure 8) and led to significant differences in the tumor growth pattern (P<0.0001).

CAMKK2 KO with EG7 tumor volume 2500 ) 3 Anti-CD8 n=5

m 2000

m Isotype IgG n=4 (

1500 No treatment n=4 me olu

v 1000

r o

m 500

Tu P<0.0001 0 0 5 0 5 0 1 1 2 Days

! Figure 8: CD8 depletion reverses tumor growth suppression in CaMKK2-/- mice

CaMKK2-/- mice injected with E.G7 1×105 were monitored every 3-4 days. The mice with no antibody treatment or IgG isotype injection had no or few tumor growth. The tumors in the mice with anti-CD8 depletion were significantly larger than the controls (p<0.0001).

3.3 Conclusions

By injecting E.G7 cells intracranial or subcutaneously into WT and CaMKK2-/- mice, we successfully observed the effect of CaMKK2 ablation on lymphoma growth in two different animal models using the same lymphoma cell line. In both intracranial model and the flank injection model, lymphoma tumor growth was sufficiently suppressed in CaMKK2-/- mice. The primary difference between the flank injection model

and the intracranial injection model is the expression of CaMKK2 in parenchymal cells – i.e. high CaMKK2 expressed in neurons, while absent in the flank. The similarity of tumor growth curves in two models indicates that rather than the CaMKK2 expressing parenchymal cells, CaMKK2 impacts tumor growth by other CaMKK2-expressing cells, and regulating these cells may provide a new way to suppress CNS lymphoma.

CaMKK2 has been reported to express in macrophages and other myeloid cells.

By injecting E.G7 cells into LysMCre+;CaMKK2fl/fl mice in which CaMKK2 was specifically ablated in myeloid lineage, we observed the similar tumor growth curves as in CaMKK2 global WT and KO mice. This result suggests that CaMKK2 ablation regulates tumor growth through myeloid cells. To further illustrate if CaMKK2 ablation in myeloid cells impacts anti-tumor immunity through T cells, we used anti-CD8 monoclonal antibody to deplete CD8 cells in CaMKK2-/- mice and challenged the mice with E.G7 cells. While there was very limited tumor growth in the CaMKK2-/- mice with no treatment or IgG isotype control, the tumor suppression in CaMKK2-/- mice was reversed after CD8 cells depletion. This finding further confirms that CaMKK2 regulates lymphoma growth by interfering the interaction of myeloid cells and T cells to suppress

T-cell anti-tumor immunity.

In this chapter, we uses two different lymphoma injection models in two different mice strains to confirm that CaMKK2 ablation suppress lymphoma progression. CaMKK2 expression in myeloid cells plays a crucial role in lymphoma

growth by interfering T-cell immunity. In the next chapter, we will further investigate the underlying mechanism on how CaMKK2 impacts myeloid cell and T-cell interaction in lymphoma microenvironment.

4. The Role of CaMKK2 in Regulating Myeloid-Cell Development in Lymphoma Environment

4.1 Introduction

In the last chapter we demonstrate that CaMKK2 expression in myeloid cells facilitate lymphoma progression by interfering T-cell anti-tumor immunity. In cancer environment, myeloid cells were crucial in mediating anti-tumor immunity. On one hand, antigen-presenting cells (APCs) especially mature dendritic cells (DCs), are known important to process and present tumor antigen to T cells, stimulate T-cell activation and clonal expansion, and augment T-cell anti-tumor effect. On the other hand, tumor-associated macrophages (TAMs) can trigger inflammatory responses to sustain the tumor survival and proliferation, as well as blunt the T-cell immunity.

Moreover, the accumulation of immature myeloid cells secret suppressive and anti- inflammatory cytokines to impair T-cell function and induce regulatory T cells (Treg).

These cells were later termed as myeloid-derived suppressor cells and there is increasing evidence for their T-cell suppressing function.

The inhibitory effect of CaMKK2 in mediating cell differentiation and maturation has been confirmed in bone remodeling, adipogenesise, and granulopoiesis. Upon the depletion of CaMKK2 in hematopoietic stem cells and myeloid progenitors, the cells show an intrinsic tendency to commit and differentiate into mature granulocytes. We wished to interrogate whether CaMKK2 exerts a brake-effect in cancer environment to suppress myeloid cells from maturation.

CaMKK2 is reported as an essential regulator in macrophage-induced inflammation. In macrophages, the loss of CaMKK2 leads to impaired cytokine secretion, morphological changes, and phagocytosis when activated by LPS. These effects are caused by the uncoupling of TLR4 cascade, resulting in the failure to recruit and phosphorylate PYK2, ERK, Jun, AKT, and NFκB. Moreover, AMPK, a direct downstream target of CaMKK2, also mediates inflammation upon phosphorylation, regulates the cellular metabolism, and inhibits the maturation of dendritic cells and macrophages towards a TH1 phenotype.

Based on all these evidence, we hypothesize the CaMKK2 can suppress myeloid cell development in lymphoma microenvironment; thereby inducing MDSCs and impairing the dendritic cell function to stimulate T cells. During this process, AMPK is likely to be involved as the downstream target. To test this hypothesis, we generated bone marrow derived dendritic cells (BMDCs) in vitro from WT and CaMKK2-/- mice exposed to E.G7 supernatant, to observe the cell maturation under tumor exposing environment. Coculture of BMDCs with T cells was used to monitor T-cell stimulation function. AMPK pathway was interrogated in BMDCs to explore the potential mechanism. Eventually, we analyzed the myeloid cell compartment in WT and CaMKK2-

/- E.G7 tumor bearing mice. From these experiments, we sought to reveal how CaMKK2 mediates myeloid cell development and function in the lymphoma microenvironment.

4.2 Results

4.2.1. CaMKK2 ablation facilitates myeloid cells maturation

4.2.1.1 CaMKK2 ablation limits MDSC generation in vitro

Since the E.G7 tumor growth is suppressed in CaMKK2-/- mice, it is difficult to determine the causal effect of the changes of myeloid cells in CaMKK2-/- mice in terms of tumor suppression. To generate a comparable tumor-exposing environment, we used in vitro generated MDSCs from mouse bone marrow to study how CaMKK2 impacts myeloid cells development.

To induce in vitro MDSCs, bone marrow cells from CaMKK2 EGFP reporter mice were cultured in the complete DMEM media with 10% FBS (D-10) supplied with GM-

CSF and IL-4, with or without 50% E.G7 supernatant added. The media was half- changed carefully at Day 3. The floating cells were collected and analyzed by flow cytometry on Day 5. As shown in Figure 9, E.G7 supernatant successfully induced more

CD11b+GR1+ MDSCs compared to that of the control with D-10 media only. (Figure 9,

Panel A and B) These MDSCs were EGFP+, indicating that CaMKK2 promoter was active. The same method was applied to generate MDSCs from WT and CaMKK2-/- mice with the present of E.G7 supernatant, and the cells were analyzed using the same gating strategy. (Figure 9, Panel C) The percentage of MDSCs generated from CaMKK2-/- mice was lower compared to that of WT (P<0.05), suggesting that CaMKK2 ablation limits

MDSC generation in vitro in the present of E.G7 supernatant.

Figure 9: CaMKK2 promoter is active in in vitro MDSC induced by E.G7 supernatant.

Cells from CaMKK2-EGFP reporter mice marrow were cultured in vitro with cytokines with or without 50% E.G7 supernatant for 5 days. Cells were collected and analyzed by flow cytometry on Day 5. A). Flow cytometry layout shows E.G7 supernatant induced more CD11b+GR1+ MDSCs compared to D-10 media alone. The CD11b+GR1+ cells were EGFP+. B) Histogram for CD11b+GR1+ cells induced by D-10 and E.G7 supernatant. n=2. The experiment was repeated once. C). CD11b+Gr1+ MDSCs generated from WT and CaMKK2-/- mice marrow with 50% E.G7 supernatant. CaMKK2-/- marrow generated less MDSCs compared to WT. (P<0.05) n=3. The experiment was repeated twice.

G/4 G/4 + E.G7 sup M-MDSC 3000000

2500000 *** * M-MDSC 2000000

1500000

Cell count 1000000 Control G-MDSC 500000 0

G/4 WT G/4 KO

G/4+EG7 supG/4+EG7 WT sup KO

G-MDSC 3500000 CaMKK2-/- 3000000 * 2500000 2000000 * 1500000

Cell count 1000000 500000 0

G/4 WT G/4 KO

G/4+EG7 supG/4+EG7 WT sup KO

Figure 10: CaMKK2-/- marrow generates less MDSCs in vitro

Cells from WT and CaMKK2-/- mice marrow were cultured in vitro with cytokines GM- CSF and IL-4 (G/4), with or without 50% E.G7 supernatant, for 5 days. Cell medium was half-changed on Day 3. Cells were collected and analyzed by flow cytometry on Day 5. The flow panel showed the gating strategy for M-MDSC (Ly6ChighLy6Gdim) and G-MDSC (Ly6CdimLy6Ghigh). Cell numbers were normalized by flueroscent microbeads. In both conditions with or without E.G7 supernatant, the absolute cell counts of both M-MDSC and G-MDSC were lower in the cells generated from CaMKK2-/- marrow compared to WT. (P<0.05)

We further analyzed the different components in CD11b+GR1+ MDSCs by staining the cells with Ly6C, Ly6G and I-A. Similar to the result of CD11b+GR1+

MDSCs, the absolute number of both Ly6ChighLy6Gdim (M-MDSC) and Ly6CdimLy6Ghigh

(G-MDSC) in the cells generated from CaMKK2-/- mice were lower compared to that of

WT (Figure 10. P<0.05) in both conditions with or without E.G7 supernatant, suggesting that less MDSCs were generated from CaMKK2-/- marrow. The cells generated from

CaMKK2-/- mice also had lower percentage of CD11b+Ly6G+ total events compared to that of WT (Figure 11. P<0.05). Further analysis of these cells showed that the percentage of CD11b+I-A+ was significantly higher in CaMKK2-/- cells compared to the WT ones

(Figure 12, P<0.01), indicating that the CaMKK2-/- MDSCs were more matured and differentiated.

Figure 11: CaMKK2-/- marrow generates less CD11b+Ly6Ghi MDSCs

Cells from WT and CaMKK2-/- mice marrow were cultured in vitro with 50% E.G7 supernatant and cytokines in the media for 5 days. Cell medium was half-changed on Day 3. Cells were collected and analyzed by flow cytometry on Day 5. The percentage of CD11b+Ly6Ghi was lower in cells generated from CaMKK2-/- marrow compared to WT. (P<0.05)

Considering the presence of progenitors at different hematopoietic stages in the bone marrow, we further selected sorted KSL cells for culture. The cells were sorted as

previously described, and seeded to 96-well plates with 50% E.G7 supernatant and cytokines, cultured for 5 days, and analyzed by flow cytometry. Similar to the cells generated from whole bone marrow, cells generated from CaMKK2-/- KSL also had higher CD11b+I-A+ expression. (Figure 13) This result again confirms that stem cells with CaMKK2 ablation have the intrinsic inclination towards myeloid cells maturation.

Control CaMKK2 -/- CD11b+ I-A+ 80 **

40 CD11b

20 % of parent event

I-A Control CaMKK2-/-

Figure 12: in vitro generated CaMKK2-/- myeloid cells have higher percentage of CD11b+I-A+ cells

Cells from WT and CaMKK2-/- mice marrow were cultured in vitro with 50% E.G7 supernatant and cytokines in the media for 5 days. Cell medium was half-changed on Day 3. Cells were collected and analyzed by flow cytometry on Day 5. Myeloid cells generated from CaMKK2-/- mice have higher percentage of CD11b+I-A+ double positive cells (P<0.01).

WT CaMKK2-/-

A - I

10 19.7

CD11b

Figure 13: Myeloid cells generated from CaMKK2-/- HSC have higher CD11b+ I-A+ percentage

HSCs sorted from WT and CaMKK2-/- marrow were placed in 96-well plates and cultured with 50% E.G7 supernatant and cytokines. Myeloid cells generated from CaMKK2-/- HSC have higher CD11b+I-A+ percentage. For cell number limitation, only two wells were seeded for each genotype. The experiment was repeated once.

4.2.1.2. CaMKK2 ablation limits MDSC generation in tumor bearing mice

To further examine the role of CaMKK2 in myeloid development, we investigated MDSC generation in tumor-bearing mice. WT and CaMKK2-/- mice were challenged with E.G7 1×105 subcutaneously. The mice were sacrificed after 14 days.

Splenocytes were stained for myeloid cell markers after RBC lysis as previously described. As detected in CD11b+ myeloid compartment, tumor challenged WT mice had about 60% Ly6GhighLy6Cdim cells, while CaMKK2-/- mice only had 10-20% (Figure 14).

Significantly less Ly6GhighLy6Cdim cells were accumulated in the spleen pf CaMKK2-/- tumor-bearing mice (P<0.001). In light of the fact that Ly6GhighLy6Cdim cell (ganulocytic

MDSC) is the most proliferated MDSC compartment, this result indicates that CaMKK2 ablation suppresses MDSC proliferation and accumulation in lymphoma environment.

G-MDSC

% 70 ***

low 60 50 Ly6C

hi 40 30 20 10

CD11b+Ly6G 0

Control CaMKK2-/-

Figure 14: CaMKK2 ablation limits MDSC generation in vivo

WT and CaMKK2-/- mice challenged with E.G7 1×105 were sacrificed on day 14. Splenocytes from these mice were stained for flow cytometry. WT tumor-bearing mice have significant higher Ly6GhiLy6Clow MDSC cells accumulation in the spleen comparing to CaMKK2-/- mice (p<0.001). n=5. Repeated 3 times. 74

4.2.2 CaMKK2 is expressed in dendritic cells

To further investigate how CaMKK2 impacts myeloid cells in the lymphoma microenvironment, we first verified the expression of CaMKK2 in myeloid cells.

CaMKK2-EGFP reporter mice were used to trace the CaMKK2 promoter activity in different cell compartment. Dendritic cells have been known as important APCs to stimulate T cells in immunity. We first tested DCs in spleen. The splenocytes from

CaMKK2-EGFP reporter mice were separated and stained for flow cytometry analysis.

The mature dendritic cells, which were gated as CD11b+CD11c+I-A+, were EGFP+

(Figure 15), indicating that CaMKK2 reporter is active in splenic dendritic cells.

CD11b

CD11c I-A EGFP

Figure 15: The CaMKK2 promoter is active in splenic CD11c+ DCs

Splenocytes of CAMKK2-EGFP reporter mice were stained for myeloid cells. Most of the mature dendritic cells (CD11b+ CD11c+ I-Ahigh) were EGFP+, indicating that CaMKK2 promoter is active in splenic dendritic cells.

To identify the expression of CaMKK2 in different DC subsets, we further gated the splenic DCs to three different subsets according to their differentiation stages:

CD11b+CD11cint as dendritic cell progenitor (DCp), CD11b+CD11c+ as conventional dendritic cells (cDC), CD11b-CD11c+ as plasmatoid dendritic cells (PDC). (Figure 16,

Panel A) EGFP expression was found to be highest in DCp, while in cDC and PDC showed relatively low EGFP expression (Figure 16, Panel B, C). EGFP signal significantly lowered as the dendritic cells matured from DCp to I-A+ DCp (P<0.01), and further from cDC (P<0.001) and PDC (Figure 16, Panel D). This finding indicates that the

CaMKK2 promoter activity is developmentally regulated in DC subsets as DCs differentiate. In DCp population, CaMKK2 promoter was highly activated. As the DCp matures, I-A was expressed, while CaMKK2 promoter activity was significantly decreased. In more differentiated dendritic cells such as cDC and PDC, CaMKK2 promoter activity was found relatively low. This result suggests that CaMKK2 plays a more important role in dendritic cell development and differentiation, rather than regulation of mature dendritic cell function.

Since in vitro DC generation provides a more efficient way to generate tumor- exposed DCs, we decided to use in vitro DC for further mechanism studies. We first selected CaMKK2-EGFP reporter mice to confirm CaMKK2 promoter activity in in vitro

DCs. Bone marrow cells from CaMKK2-EGFP reporter mice were collected and cultured in 6-well plates with 50% E.G7 supernatant and cytokines as decribed in previous

section. Cells were cultured for 5 days and detected with flow cytomery. As shown in

Figure 17, CD11b+CD11c+ bone marrow derived dendritic cells (BMDCs) comprised roughly 75-90% of the live cells. Comparing to the cells generated from C57BL/6 mice

(negative controls, shown in blue), the BMDCs generated from EGFP mice had positive

EGFP expression, suggesting that CaMKK2 reporter was active in BMDCs.

Figure 16: CaMKK2 promoter activity is developmentally regulated in DCs.

Splenic dendritic cells from CaMKK2-EGFP mice were further gated to different subsets of DCs according to the expression level of CD11b and CD11c. A) Gating of DC subsets. Dendritic cell progenitors (DCp) are CD11b+CD11cint. Conventional dendritic cells (cDC) are CD11b+CD11c+. Plasmacytoid dendritic cells are CD11b-CD11c+. B) DCp cells gated as CD11b+CD11cintCD4-CD8- have high EGFP expression. Most of these cells were EGFP bright. C) Comparing EGFP expression in different subset of

DCs. DCp has the brightest EGFP expression, while cDC and PDC were relatively dim. D) EGFP level in different DC subsets. EGFP has highest expression in DCp, and gradually lower down in I-A+ DCp (P<0.01) and cDC (P<0.001). PDC has limited EGFP expression.

Figure 17: CaMKK2 promoter is active in bone marrow derived dendritic cells

This figure demonstrates the gating of BMDCs. BMDCs was induced from CaMKK2- EGFP reporter mice. The last plot shows that the CD11b+CD11c+ cells from the reporter mice express high EGFP (the red histogram). Cells from C57BL/6 were used as negative control (the blue histogram).

A. B. WT CaMKK2-/- CaMKK2 RNA

**** 1.0 CAMKK2

0.5 Actin Fold Change

0.0

WT KO

Figure 18: CaMKK2 is expressed in BMDCs

BMDC from CaMKK2 WT mice express CAMKK2. A) Immunoblot of BMDC. WT BMDC have clear CaMKK2 bands, while CaMKK2-/- BMDC do not. B) Quantitative RT-PCR of BMDC. CaMKK2 mRNA level in WT BMDC was significantly higher than BMDC generated from CaMKK2-/- mice (P<0.0001).

To further validate the expression of CaMKK2 in WT and CaMKK2-/- BMDCs, we applied the same method to generate BMDC from CaMKK2 WT and CaMKK2-/- mice.

Nonadherent cells were carefully collected on Day 5 to generate cell pellets. Western blot with anti-mouse CaMKK2 antibody successfully detected CaMKK2 expression in WT

BMDC but not in CaMKK2-/- BMDC (Figure 18, Panel A). Quantitative realtime RT-PCR showed a significantly higher level of CaMKK2 mRNA in WT BMDC than that of

CaMKK2-/- BMDC (Figure 18, Panel B. P<0.0001). Both western blot and real-time PCR confirmed CaMKK2 expression in BMDCs. 79

4.2.3 CaMKK2-/- DCs lead to enhanced T-cell stimulation

4.2.3.1 CaMKK2-/- dendritic cells present higher CD11c+I-A+ percentage

Figure 19: The percentage of CD11c+I-A+ is higher in BMDCs generated from CaMKK2-/- marrow.

Dendritic cells were induced from bone marrow of WT and CaMKK2-/- mice (n=4) in vitro with E.G7 supernatant and cytokines. Cells from CaMKK2-/- marrow have higher CD11c+I-A+ percentage (P<0.001).

During dendritic cell maturation, the expression of MHCs molecules and T-cell costimulatory factors (e.g. CD40, CD80, CD86) is up-regulated. To further study how

CaMKK2 ablation impacts DC maturation and function in the lymphoma environment, we generated BMDC from CaMKK2 WT and CaMKK2-/- mice in vitro with 50% E.G7 supernatant and cytokines. Cells were analyzed with flow cytometry on Day 5. Similar

to myeloid cells generated in vitro, BMDCs from CaMKK2-/- mice have higher percentage of CD11c+I-A+ in CD11b+ compartment (Figure 19. P<0.001), indicating enhanced dendritic cell maturation.

4.2.3.2 CaMKK2-/- dendritic cells have higher co-stimulatory factors expression

The expression of costimulatory factors on in vitro BMDCs was also assayed.

Unlike the difference in splenic DCs between WT and CaMKK2-/-mice, BMDCs from WT and CaMKK2-/- mice have no difference in costimulatory molecule expression (Figure 20,

Panel A). However, when 1×104 BMDCs were co-cultured with 1×105 T cells for 24 hours in the present of anti-CD3 antibody, CaMKK2-/- BMDCs have significantly higher expression of CD40 (P<0.01) and CD86 (P<0.001) compared to WT BMDCs (Figure 20,

Panel B). The difference of costimulatory factors expression before and after interacting with T cells suggests that CaMKK2 ablation in BMDC increases the response ability of

DC to T cells and may stimulate better T cells proliferation and function.

To verify this hypothesis, 50µl supernatant from the coculture wells was carefully collected at 24h for cytokine detection by MILLIPLEX® MAP Mouse

Cytokine/Chemokine Magnetic Bead Panel using Luminex. Among many other detected cytokines, IFN-γ and IL-12 are two major cytokines that relate to dendritic cells and T- cell activation. As expected, the concentration of IFN-γ and IL-12 were higher in the coculture of CaMKK2-/- BMDC with T cells (Figure 21). IL-12 can be secreted by dendritic cells and is a powerful signal to stimulate T-cell activation. IFN-γ on the contrary, is

mainly secreted by T cells, and considered as a T-cell activation marker. IFN-γ can also further enhance the DC function. The increased level of IL-12 and IFN-γ in CaMKK2-/-

BMDC with T-cells coculture strongly suggests that BMDC with CaMKK2 ablation better stimulates T-cell activation, which in turn enhances BMDC function.

Figure 20: BMDC from CaMKK2-/- marrow have higher co-stimulatory factors expression after interacting with T cells.

Flow cytometry on BMDC from WT and CaMKK2-/- before and after interacting with T cells. A). BMDC from CaMKK2 WT and CaMKK2-/- mice (n=4) at day 5. CD40 and CD86 expression level have no difference between WT and CaMKK2-/-. B). BMDC 1×104 coculture with T cells 1×105 for 24 hours. BMDCs were mixed from 4 mice as indicated. The culture was triplicated. CaMKK2-/- BMDC had significantly higher expression of CD40 (P<0.01) and CD86 (P<0.001) comparing to WT BMDC.

IFN-y IL-12(p70) 1200 60 * 1000 *

800 40

600

400 20 Conc (pg/mL) Conc (pg/mL) 200

0 0

KO DC+T cells+anti-CD3 WT DC+ T cells +anti-CD3 KO DC+T cells+anti-CD3 WT DC+ T cells +anti-CD3

Figure 21: Cytokine secretion analysis in supernatant from coculture of BMDC with T cells.

BMDC 1×104 coculture with T cells 1×105 with anti-CD3 mAb in the media. At 24 hours, 100µl supernatant was carefully removed from each well and detected for cytokine secretion by Luminex. The wells were triplicated. From the wells T cells cocultured with CaMKK2-/- BMDCs, IFN-γ and IL-12 concentration were higher comparing to WT wells (P<0.05).

At the same time, we also stimulated the BMDCs with LPS to test if CaMKK2-/-

BMDCs have better function. BMDCs 1×105 from WT or CaMKK2-/- were seeded into 96- wells and stimulated with LPS for 24h. Supernatant was carefully collected and the cytokine secretion was analyzed by the Luminex platform. Most of the cytokines (G-

CSF, TNF, IL-12p40, IL-6, IL-1, IL-15, IL-18) and chemokines (CCL2, CCL3, CCL4, CCL5,

CXCL1, CXCL2, CXCL10) were higher in CaMKK2-/- wells comparing to those in WT.

(Figure 22) This result suggests a better maturation of DCs with stronger response to

LPS stimulation.

G-CSF TNF IL12(p40) IL6 250 * 400 80 1500 200 *** 300 60 ** 1000 * 150 200 40 pg/ml pg/ml pg/ml 100 pg/ml 500 100 20 50

0 0 0 0

WT KO WT KO WT KO WT KO

IL1a IL1b IL15 IL18 100 50 20 60

80 40 ** ** ** * 15 40 60 30 10 pg/ml 40 pg/ml 20 pg/ml pg/ml 20 5 20 10

0 0 0 0

WT KO WT KO WT KO WT KO

CCL2 CCL3 CCL4 CCL5 80 ** 1500 1500 5000 4000 * 60 1000 ** 1000 *** 3000 40 pg/ml pg/ml pg/ml pg/ml 2000 500 500 20 1000

0 0 0 0

WT KO WT KO WT KO WT KO

CXCL1 CXCL2 LIX 4000 CXCL10 8000 400 300

3000 ** 6000 * 300 ** 200 2000 4000 200 pg/ml pg/ml pg/ml pg/ml 100 1000 2000 100

0 0 0 0 WT KO WT KO WT KO WT KO

Figure 22: Cytokine secretion analysis in supernatant from BMDCs stimulated with LPS.

BMDCs 1×105 generated from WT and CaMKK2-/- mice (n=3) were exposed to LPS in the media in triplicate. At 24 hours, 100µl supernatant was carefully removed from each well and detected for cytokine secretion by Luminex. After LPS stimulation, most of the cytokines and chemokines secreted by CaMKK2-/- BMDCs were higher than WT. Error bar indicates standard error of the mean. * P<0.05. ** P<0.01. *** P<0.001.

A. CD25

CFSE WT KO No DCs

B. CD25

CFSE WT KO No DCs C. CD4 CD25+ proliferated CD8 CD25+ proliferated

80 ** 100 ** 80 60 60 40 40 20 20 CD25+ Proliferated % CD25+ Proliferated % 0 0

WT KO WT KO

Figure 23: CaMKK2 ablation BMDC better stimulates T-cell activation

BMDC 1×104 coculture with T cells 1×105 with anti-CD3 mAb for 72 hours. T cells were pre-stained with CFSE. In Panel A CD4 cells were seeded and in Panel B CD8 cells were seeded. Both CD4 and CD8 T cells with BMDC from CaMKK2-/- mice have higher percentage of CD25+CFSEproliferated cells, which means that BMDC from CaMKK2-/- better stimulates T-cell activation. T cells alone were used as negative control.

CD4 CD25+ CD69 MFI - 72h CD8 CD25+ CD69 MFI - 72h *** 150 300 ** *** **

100 200 **** *** CD69 MFI CD69 MFI 50 100

0 0

WT KO WT KO

WT anti-IL12 KO anti-IL12 WT anti-IL12 KO anti-IL12

Figure 24: T cells cocultured with CaMKK2-/- BMDCs better express T-cell activation marker CD69, which can be blocked by anti-IL-12 mAb

BMDC 1×104 coculture with T cells 1×105 for 72 hours. With BMDC from CaMKK2-/-mice, both CD4 and CD8 CD25+CFSE proliferated cells had higher CD69 MFI compared to cells cocultued with WT BMDC. (P<0.001 for CD4, P<0.01 for CD8, respectively) Adding anti-IL-12 antibody in the culture decreased the CD69 expression on T cells interacted with CaMKK2-/- BMDCs. (P<0.001 for CD4, P<0.01 for CD8, respectively)

To test if CaMKK2 ablation affects the ability of BMDC in T-cell activation,

BMDC were cocultured with T cells that had been pre-stained with CFSE. BMDC 1×104 and T cells 1×105 were cocultured with anti-CD3 antibody for 72 hours. T cell proliferation was monitored by flow cytometry tracing CFSE dilution. CD25 was used as a T-cell activation marker. After 72 hours of culture, most of the CD4 and CD8 cells proliferated, as shown by CFSE dilution (Figure 23, Panel A and B). However, in these proliferating cells, the ones interacting with CaMKK2-/- BMDC have higher CD25+ expression (Figure 23, Panel C. P<0.01). The proliferating CD25+ cells were further

analyzed with CD69, another T-cell activation marker. Again, both CD4 and CD8 cells cocultured with CaMKK2-/- BMDCs have higher CD69 expression (Figure 24. P<0.001 for

CD4 cells, P<0.01 for CD8 cells).

Figure 25: The proliferated CD25+ cells are FoxP3 negative

BMDC 1×104 coculture with T cells 1×105 for 72 hours with anti-CD3 mAb in the media. The CD25+CFSEproliferated cells were stained for FoxP3 to identify Treg cells. Cells from coculture wells with both CAMKK2 WT and CaMKK2-/- have low FoxP3 expression, indicating these cells were activated T cells but not Treg cells. T cells cocultured with BMDC without anti-CD3 mAb were used as positive control for FoxP3 staining.

Considering that IL-12 is crucial for T-cell activation, and CaMKK2-/- BMDCs secrete higher level of IL-12 when interacting with T cells, we added anti-IL-12 neutralizing antibody into the coculture system, hoping to reverse the T-cell activation.

As shown in Figure 24, although in the presence of anti-IL-12 mAb, the CD69 expression on T cells interacted with CaMKK2-/- BMDCs was still higher compared to that with WT

ones (P<0.001 for CD4, P<0.0001 for CD8, respectively), it was significantly decreased comparing to the ones activated by CaMKK2-/- BMDCs without anti-IL-12 mAb (P<0.001 for CD4, P<0.01 for CD8, respectively). This evidence indicates that CaMKK2-/- BMDCs better activate T cells because of the elevated secretion of IL-12, and this advantage can be blocked by neutralizing IL-12 using antibody.

Since CD25 is also a marker for regulatory T cells, the proliferated CD4+CD25+ cells were further stained for FOXP3 intracellular staining. Comparing to the quiescent T cells from coculture without anti-CD3 mAb stimulation, the BMDC cocultred T cells with anti-CD3 mAb stimulation were FoxP3 negative (Figure 25), indicating that the proliferating CD25+ cells were not Treg cells.

To investigate whether CaMKK2 -/- BMDCs impact the antigen-dependent T-cell activation, OVA peptide (SIINFEKL) was chosen to stimulate BMDCs cocultured with

OT-1 T cells. Splenocytes from OT-1 mice were enriched by R&D anti-mouse CD3 T-cell enrichment columns, and followed by MACS magnetic CD4 depletion. The CD8 cells were pre-stained with CFSE and cocultured with BMDCs with SIINFEKL peptide with the final concentration of 10 ng/ml in media. At 72 hours, the percentage of

CD25+CFSEdim and CD69+CFSEdim were detected higher in T cells cocultured with

CaMKK2-/- BMDCs compared to the ones with WT BMDCs (Figure 26). This result further solidates the fact that CaMKK2-/- BMDCs are more potent to stimulate T-cell antigen-dependent activation.

Figure 26: CaMKK2-/- BMDC better stimulates OT1 T-cell activation

BMDC from WT and CaMKK2-/- mice 2×104 cells were cocultured with 1×105 OT-1 T cells with the presence of titrated SIINFEKL peptide at 10ng/ml for 72h. All wells were triplicated. OT-1 T cells were analyzed for proliferation by CFSE and activation by CD25 and CD69 flow cytometry. The percentage of CD25+CFSEdim and CD69+CFSEdim were higher in T cells cocultured with CaMKK2-/- BMDC, indicating CaMKK2-/- BMDC can better stimulates OT-1 antigen-dependent activation.

To compare the basic level of costimulatory factors on dendritic cells, splenocytes from normal WT and CaMKK2-/- mice were stained for costimulatory factors and analyzed by flow cytometry. As illustrated in Figure 27, in the CD11b+CD11c+ compartment, CaMKK2-/- splenic DCs had higher CD80 (P<0.0001) and CD40 (P<0.05) expression than those of WT splenic DCs. We also analyzed I-A expression and PD-L1 expression in the splenic DCs. Not surprisingly, CaMKK2-/- splenic DCs have higher I-A+ percentage. This result corroborates the finding of BMDCs from WT and CaMKK2-/- mice.

The CaMKK2 WT splenic DCs also have higher PD-L1 expression (Figure 28). The collective flow cytometry results suggest that CaMKK2-/- dendritic cells have better ability for antigen presentation and T-cell stimulation.

CD80 CD40 100 **** 300 *

80 200 60 MFI 40 MFI 100 20

0 0

WT Spleen KO Spleen WT Spleen KO Spleen

Figure 27: in vivo DCs have higher co-stimulatory factors expression in normal CaMKK2-/- mice

Flow cytometry of splenic dendritic cells from normal CaMKK2 WT (n=4) and CaMKK2-/- mice (n=4). CaMKK2-/- DCs have significantly higher CD80 (P<0.0001) and CD40 (P<0.05) expression.

PD-L1%

80 *

Percentage (%) Percentage 20

WT KO

Figure 28: Surface marker expression indicates DC compartment from CaMKK2-/- have better T-cell stimulation function

Flow cytometry of spleen dendritic cells from normal CaMKK2 WT (n=4) and CaMKK2-/- mice (n=4). A higher percentage of I-A+ DCs and lower percentage of PD-L1+ DCs were found in CaMKK2-/- mice. (P<0.05)

Considering the various abilities of different DC subsets to stimulate T cells, we are interested to investigate the splenic DC subsets of CaMKK2 WT and CaMKK2-/- mice at baseline. The splenocytes from CaMKK2 WT and CaMKK2-/- mice were stained for anti-mouse CD11b, anti-mouse CD11c, anti-mouse I-A, anti-mouse CD4 and CD8. The conventional DC (cDC) was gated as CD11b+CD11c+. The cDCs were further gated as

CD4+ cDC, CD8+ cDC, and double-negative (DN) cDC (Figure 29). Similar to the myeloid cells cultured in vitro, the splenic cDC from CaMKK2-/- mice had higher I-A expression (P<0.05). When the cDC were further gated into CD4 and CD8 subsets, CD4+ cDCs and DN cDCs in CaMKK2-/- mice also had higher I-A expression (P<0.05).

Moreover, the DC subsets in CaMKK2 WT and CaMKK2-/- mice were also slightly different. While CaMKK2-/- mice had more CD4+ cDCs (P<0.01), the WT mice had more

DN cDCs (P<0.001). This result may indicate that the differentiation and maturation of dendritic cells are different in CaMKK2 ablated mice, leading to changes in immune functions.

DC#subsets#in#EG7.bearing#WT#and#CaMKK2./.#mice$

cDC$

DC$CD4+$

80 20000 20000 * * * 60 *** 15000 15000 **

10000 10000 40 % of cDC I-A (MFI) I-A I-A (MFI) I-A 5000 5000 20

0 0 0 WT KO WT KO WT KO WT KO WT KO WT KO WT KO CD8$ CD4$ DN$ CD8$ CD4$ DN$

Figure 29: DC subsets in CaMKK2 WT and CaMKK2-/- tumor-bearing mice

Splenic DCs from CaMKK2 WT and CaMKK2-/- mice were analyzed by flow cytometry. In CD11b+CD11c+ conventional DC population and CD11b+CD11c+CD4+ DC population, the I-A expression was higher in CaMKK2-/- cells compared to WT cells (P<0.05). There were more CD4+ cDCs in CaMKK2-/- mice (P<0.01), while CD8-CD4- cDCs were more dominant in WT mice (P<0.001). n=7.

Finally, to confirm that the CaMKK2-ablated environment enhances anti-tumor activity, WT and CaMKK2-/- mice were challenged with E.G7 cells subcutaneously and sacrificed at Day 21. After cell separation, 5×106 splenocytes from the tumor-bearing mice were seeded into ELISpot plate precoated with anti-mouse IFN-γ antibody and cultured with OVA protein at 1mg/ml for 24h. ELISpot assay was performed as per instructions from the manufacturer. Even though there were a low number of spots, the wells with cells from CaMKK2-/- mice had doubled the number of detected spots comparing to WT wells (Figure 30) and the difference is statistically significant (P<0.05).

Given that E.G7 cells express OVA protein, using ELISpot to detect antigen-specific T cells against OVA protein can demonstrate the T-cell immunity towards E.G7 cells in tumor-bearing mice. However, since OVA is not a major antigen that drives tumor rejection, the amount of OVA antigen-specific MHC complexes on DCs (one hundred or less per cells) and the OVA antigen-specific T-cell clones frequency (~1/100,000 or less) were very limited.(Banchereau et al., 2000) Here we input 5×105 cells/well with OVA peptide and culture for 24h. The low frequency of antigen-specific cells in general may explain the sparse IFN- γ positive cells in both genotypes of tumor bearing mice. And the difference in result of ELISpot indicates that CaMKK2 ablation mice have better anti- tumor T-cell immunity.

ELISpot IFN-gamma 15 *

Spot count 5

WT KO

Figure 30: Splenocytes from CaMKK2-/- EG7 tumor-bearing mice have higher IFN-γ secretion comparing to WT

CaMKK2 WT and KO mice (n=4) were challenged with EG7 tumor cells subcutaneously. The mice were then sacrificed at day 21. The splenocytes were mixed to 1 sample each genotype, and seeded to the ELISpot plate precoated with IFN-γ antibody at the concentration of 5×105 cells/well and cultured for 24h with 1mg/ml OVA protein. The wells were triplicated. Wells with cells from CaMKK2-/- mice have more spots detected comparing to WT wells (P<0.05).

4.2.4 AMPK pathway is involved in CaMKK2-mediating myeloid cell development

CaMKK2 has been known to phosphorylate 3 downstream kinases: CaMKI,

CaMKIV, and AMPK. Among these three kinases, AMPK is widely expressed in different tissues and considered to have the highest affinity to CaMKK2. To investigate the molecular mechanism of CaMKK2 in regulating immune cells, we first use RT2 PCR array to analyze AMPK pathway in BMDCs from CaMKK2 WT and CaMKK2-/- mice.

Compared to WT BMDCs, there was 4.11-fold decrease for Camkk2, 2.82 fold decrease for Camkk1, and 2.36 fold decrease for Ppargc1a, and 2.47 fold increase for Acacb in

CaMKK2-/- BMDCs (data not shown). This result was further confirmed by quantitative real-time PCR. However the transcription level of Ppargc1a and Acacb were very low.

Western blot was later used to detect the expression of some crucial proteins in

AMPK pathway. The cell number was determined and normalized before generating cell pellets. Proteins from 2×105 cells were denatured and added to each lane. The membranes were duplicated, and blotted with anti-mouse total ACC, anti-mouse phosphor-ACC, anti-mouse total AMPK, anti-mouse phosphor-AMPK, anti-mouse S6, anti-mouse phospho-S6, and beta-actin was used as loading control (Figure 31). All the signals were normalized to actin. As calculated, the protein level of total AMPK, total S6 and total ACC were comparable in WT and CaMKK2-/- BMDC, however, p-AMPK

(P<0.001) and p-ACC (P<0.05) were significantly lower in CaMKK2-/- BMDC, while p-S6

(P<0.05) was higher (Figure 32). ACC is a direct downstream target of AMPK and an

indicator of AMPK activity, while S6 is a remote target in AMPK-mTOR pathway. The changes of the phosphorylation of these two downstream emzymes in BMDCs illustrate that AMPK is involved in CaMKK2-mediated tumor immunity.

WT | CaMKK2-/-

Total ACC

p- ACC

Total AMPK

p- AMPK

p- S6

Actin

Figure 31: Immunoblots shows AMPK pathway changes in BMDC

BMDCs were generated from individual CaMKK2 WT and CaMKK2-/- mice (n=3). Cell number was normalized before generating cell pellets. Western blog was performed and incubated with antibodies of key proteins in AMPK pathway. Β-Actin was used as loading control.

Figure 32: Quantification of AMPK pathway immunoblot

Histograms of BMDC immunoblot quantification. The total proteins of AMPK, S6, and ACC were all comparable between CaMKK2 WT and CaMKK2-/-. But CaMKK2-/- BMDCs have significantly lower expression of phosphor-AMPK (P<0.001) and phosphor-ACC (P<0.05), and higher expression of phosphor-S6 (P<0.05). n=3. All proteins normalized to β-Actin.

4.3 Conclusions

In this chapter, we used both in vitro and in vivo models to illustrate that

CaMKK2 ablation facilitates myeloid cells maturation and limits MDSC accumulation.

By using CaMKK2-EGFP reporter mice, we first show that CaMKK2 reporter was active 97

in splenic dendritic cells, yet the promoter activity was decreased as DCs differentiated.

In BMDCs, CaMKK2 was expressed confirmed by reporter activity, western blot, and quantitative real-time RT-PCR. CaMKK2-/- BMDCs had higher I-A expression, better responsed to LPS stimulation, and expressed higher level of T-cell costimulatory molecules after being cocultured with T cells. In the presence of T cells, CaMKK2-/-

BMDCs also secreted higher level of IL-12, better induced T-cell IFN-γ secretion, and better stimulated antigen dependent and non-dependent T-cell activation. At baseline level, CaMKK2-/- splenic DCs have higher I-A percentage and lower PD-L1 percentage, and higher expression level of CD80 and CD40. In tumor bearing mice, the DC subsets in WT and CaMKK2-/- mice were different, yet I-A expression was still higher in CaMKK2-

/- DCs. By using ELISpot, it was confirmed that splenocytes from CaMKK2-/- tumor bearing mice had higher IFN-γ secretion towards OVA protein. These collective evidence indicate that CaMKK2 ablation regulates DC differentiation in lymphoma environment, facilitates DC maturation and enhances DC stimulatory function, and therefore better stimulates T-cell anti-tumor activity.

As a downstream target of CaMKK2, AMPK regulates fatty acid oxidation, mitochondrial biogenesis and oxidative metabolism. Western blot shows that AMPK phosphorylation was inhibited in CaMKK2-/- BMDCs, and phosphorylation of the downstream targets of AMPK including ACC and S6 were also altered. Based on this information we could conclude that AMPK is involved in CaMKK2-mediated BMDC

differentiation. Further functional assays are needed to confirm the mechanism how

AMPK impacts DC functions in CaMKK2 abaltion microenvironment.

5. The Role of CaMKK2 in Other Tumor Models

5.1 Introduction

In the previous two chapters, we have illustrated that CaMKK2 regulates lymphoma progression by suppressing myeloid cell maturation, resulting in MDSC accumulation and impaired DC function. Although different strains of mice were used,

E.G7 was the only cell line that we tested for the lymphoma model. To exclude the possibility that the CaMKK2-environment effect is cell-line specific, we used different cancer models to repeat the tumor growth experiment.

Multiple myeloma is a hematopoietic malignancy caused by abnormal plasma cells accumulating in the bone marrow. The excessive plasma cells secrete the monoclonal protein that can be detected by serum protein electrophoresis. In mice, primary multiple myeloma model is established in Vk*Myc mutation mice. The Vk*Myc myeloma cells are then purified from the mice with detectable monoclonal gammopathy, and propagated in healthy mice through tail vein injection.

E0771 is a commonly used murine breast cancer cell line. Although CaMKK2 has been reported to regulate the tumor growth in some breast cancer cell lines, it is unknown whether CaMKK2 impacts the breast cancer microenvironment. Hence, we here used E0771 as the solid tumor model to investigate how CaMKK2 regulated the tumor microenvironment.

100

5.2 Results CaMKK2&/&(mice(are(5.2.1 CaMKK2 regulates myelomaresistant progression(to( Vk*MYC(myeloma(

M#Spike!

B. CaMKK2#/#!

p = 0.015! WT!

Fig.Fig. 3. 3. Vk12598 myeloma myeloma does does not not engraft engraft in in CaMKK2-/-mice. 10^5 Vk12598 myeloma cells were FigureCaMKK2-/-mice. 33: WT and CaMKK2 10^5-/- mice Vk12598 with myeloma myeloma cell-/- cells injection were inoculatedinoculated i.v. in in wild-type wild-type (WT) (WT) and and CaMKK2 CaMKK2 mice.-/- mice. WT and CaMKK2-/- mice(A)(A) wereSPEPSPEP injected was performedperformed with Vk*Myc on on mouse myelomamouse bled bled cells at 5at 5 weeks. ×510 weeks.6 i.v.. Mice were monitored for survivalThe Theevery positionposition other ofofday. thethe Serum albuminalbumin was is is indicated,collected indicated, andto andmonitor brackets brackets M -spike every week. A) Image of ashow gel to the show different M-spike. globulin M - componentsspike at 6 weeks of the serum. after injection was detected from CaMKK2show WT the mice different serum globulin using protein components electrophoresis, of the serum. but invisible in CaMKK2-/- mice. B) SurvivalArrowheadArrowhead curve emphasizes of WT and CaMKK2 M-spike M-spike -/- in mic in WT e WT (n=10) mouse. mouse. after (B) myeloma (B) cell injection. Most of the SurvivalWTSurvival mice curve. die between pp == 0.015. 0.015. Day 45-56, while all the mice in CaMKK2-/- group survive (P=0.015).

101 !!

Vk*Myc myeloma cells were purified from spleens of Vk*Myc mice with detectable monoclonal gammopathy.(Chesi et al., 2008) 5×106 Vk*Myc cells were injected into WT and CaMKK2-/- mice through tail vein injection. The mice were monitored every other day and bled to test for monoclonal gammopathy every week. The monoclonal band in the serum, also called M-spike, appeared in WT group around 42 days after injection, however not in CaMKK2-/- mice (Figure 33, Panel A). After 45 days, most of the

WT mice with myeloma injection died because of myeloma, whereas 100% of the

CaMKK2-/- mice survived up to 100 days. (Figure 33, Panel B) The difference is significant

(P=0.015). Similar to E.G7 lymphoma, CaMKK2 ablation environment also suppresses myeloma development.

5.2.2 CaMKK2 regulates breast cancer progression

Considering myeloma and lymphoma were both hematopoietic malignancies, we further selected a solid tumor model to test the role of CaMKK2 ablation in tumor growth. 2×105 E0771 murine breast cancer cells were injected into the first fat pad of WT and CaMKK2-/- females. The tumor size was measured every 2-3 days. Similar to the previous observations, the tumors developed gradually in the WT mice, while the tumors grew at a much slower pace in CaMKK2-/- mice. (Figure 34. P=0.04) The same result was noted in LysMCre+;CaMKK2+/+ and LysMCre+;CaMKK2fl/fl mice (Figure 35.

P=0.009), indicating that CaMKK2 ablation in myeloid lineage is sufficient to suppress breast cancer growth.

102

E0771 tumor volume 1500 WT

3 KO 1000

P=0.04 500 tumor volume/mm

0 0 7 9 14 16 Days

Figure 34: E0771 tumor growth in CaMKK2 mice

CaMKK2 WT and CaMKK2-/- female mice (n=10) were injected 2×105 E0771 cells s.c. into the fat pad of the first nipple on the right side. Tumor size was monitored every 2-3 days. The tumor growth was significantly faster in WT comparing to CaMKK2-/- mice (P=0.04).

LysM-Cre Mice w/ E0771: Tumor Size 2000 Cre+;+/+

3 Cre+;fl/fl 1500

P=0.009 1000

500 tumor volume/mm

0 0 7 10 14 17 Days

Figure 35: E0771 tumor growth in mice with myeloid lineage CaMKK2 ablation

2×105 E0771 cells were injected s.c. into the breast fat pads of female LysMCre+;CaMKK2fl/fl mice and the genotype control littermates LysMCre+;CaMKK2+/+ mice (n=8). Tumor size was monitored every 2-3 days. The tumor growth in LysMCre+;CaMKK2+/+ mice was significantly faster than in LysMCre+;CaMKK2fl/fl mice (P=0.009). 103

5.3 Conclusions

In this chapter, we used two different cancer models- multiple myeloma cells and breast cancer cells, inoculated into WT and CaMKK2-/- mice, and obtained similar tumor suppression pattern to E.G7 cells in CaMKK2-/- mice. The tumor suppression pattern of breast cancer progression in LysMCre+;CaMKK2fl/fl mice again confirms that

CaMKK2 mediates tumor growth by regulating myeloid cells, as with lymphoma model.

We now could conclude that the impact of CaMKK2 on cancer microenvironment is exerted through myeloid cells, and this phenomenon is not cell-line specific, but universal for lymphoma, multiple myeloma, and breast cancer. CaMKK2 is a crucial immune mediator and potential target in the general cancer immunity. In the next chapter, we are going to discuss the use of pharmaceutical inhibitor of CaMKK2 as a cancer therapy.

104

6. STO-609 as a Potential Immune-Regulating Treatment for Cancers

6.1 Introduction

STO-609 is a selective inhibitor for CaMKKs. The efficiency of STO-609 towards

CaMKK2 is 5-fold higher than that to CaMKK1. In most studies involving CaMKK2 or

AMPK, STO-609 is often used in vivo or in vitro to suppress CaMKK2-AMPK pathway, thus validating the participation of CaMKK2 or AMPK. In cancer studies, STO-609 has been tested in prostate cancer, breast cancer, colon cancer, gastric cancer, and hepatic cancer for its capability in regulating cancer cell survival and proliferation. Based on the reported effect of STO-609 on other cancers and the role of CaMKK2 in regulating tumor microenvironment, we here seletected STO-609 as a potential treatment for lymphoma, in the hope that it could suppress CaMKK2 thus reshaping the tumor microenvironment, and leading to possible tumor suppression.

6.2 Results

6.2.1 STO-609 suppresses lymphoma and breast cancer progression in vivo

STO-609 was tested on E.G7 lymphoma model to determine if CaMKK2 pharmaceutical inhibitor could suppress tumor growth. E.G7 5×105 were injected to

C57BL/6 mice subcutaneously 10 days before the start of treatment. On Day 0, the mice were screened with bioluminescent imaging to confirm tumor implantation, and randomized into the STO-609 treatment group and DMSO vehicle control group

105

according to their tumor signals. The mice with tumors were then received peritoneal injection of 200µl 20mM STO-609 or DMSO as vehicle control every 2-3 days. Tumor size was measured at the same time. As shown in Figure 36, lymphoma growth was significantly suppressed after STO-609 (P<0.0001), compared to the ones with vehicle control. (Figure 36) The same phenomenon was also observed in C57BL/6 female mice injected with E0771 breast cancer cells (Figure 37). In both lymphoma and breast cancer models, STOEfficacy-609 effectively of suppress CaMKK2ed tumor growth. Inhibitor These results STO-609 collaborate the previous findingsIn in mice Mammary with CaMKK2 ablat Tumored. Treatment

2500 Vehicle 2000 STO-609

1500

1000 *

500 Tumor size (mm3) size Tumor

0 0 5 10 15 20

Time (days)

Figure 36: Efficacy of STO-609 on E.G7 lymphoma

C57BL/6 mice injected with E.G7 5×105 were inoculated 10 days before treatment and tumor implantation was confirmed by bioluminescent imaging on day 0 of treatment. Mice were randomized into DMSO control group (n=9) and STO-609 treatment group (n=6). 20mM STO609 or DMSO control were injected i.p. every 2-3 days. Tumor size 106

Efficacy of CaMKK2 Inhibitor STO-609 was monitored. The tumor volumes were significantly different between STO-609 treatment group and InDMSO Mammary control group (P<0.0001). Tumor Treatment

2500 Vehicle 2000 STO-609

1500

1000 *

500 Tumor size (mm3) size Tumor

0 0 5 10 15 20

Time (days)

Figure 37: Efficacy of STO609 on E0771 breast cancer

E0771 2×105 were injected into the first fat pat on the right side in C57B/6 female mice 7 days before STO-609 treatment. 20mM STO-609 or DMSO vehicle control were injected i.p. every 2-3 days. Tumor size was monitored. The tumor volumes were significantly different between STO609 treatment group and DMSO control group (P<0.05).

6.2.2 STO-609 remodels the immune microenvironment of breast cancer

To better analyze the microenvironment changes after the STO-609 treatment, the breast tumor masses from both STO-609 treatment group and DMSO vehicle control group were dissected and digested to single cells and stained for flow cytometry. Flow cytometry data showed that the percentage of Ly6C+I-A+ cells and CD4+ T cells were increased in the STO-609 tumor treated (P<0.05), and the percentage of CD3 T cells, CD8

107

T cells were also significantly higher (P<0.01) in STO-609 treated group when compared to the control group (Figure 38). The increased percentage of PD-1+ cells in T cells after

STO-609 treatment (P<0.01) also indicates that the accumulated T cells were activated

(Figure 38). These changesSTO-609 suggest that STOTreatment-609 changes the Reprogram tumor microenvironment by facilitating the myeloidThe-cell maturation Tumor and Microenvironment increasing the T-cell infiltration .

Monos-Ly6C+ I-A+ NK T-cells

15 20 50 ** * 40 15 10 30 10 20 5 5 10 CD3+ (% of CD45.2) Ly6C I-A+ (% of Ly6C of (% I-A+ Ly6C

0 CD3- NK1.1+ (% of CD45.2) 0 0 Vehicle STO Vehicle STO Vehicle STO Vehicle STO Vehicle STO Vehicle STO

CD8+ T-cells CD4+ T-cells PD1+ T-cells 20 15 20 15 ** ** 15 * 10 10 10

5 5 5 CD3+ PD1+ (% of CD45.2)

CD3+ CD8+ (% of CD45.2) 0 CD3+ CD4+ (% of CD45.2) 0 0 Vehicle STO Vehicle STO Vehicle STO Vehicle STO Vehicle STO Vehicle STO

Figure 38: STO609 treatment reprograms the tumor microenvironment.

C57B/6 mice with E0771 tumors were treated with STO609 or vehicle every 3-4 days. The tumors were digested and the cell component was analyzed by flow cytometry when the mice were sacrificed. After STO609 treatment, there was increased percentage of CD3+ T cells (P<0.01), CD8+ T cells (P<0.01) and CD4+ T cells (P<0.05) in the STO609 treated tumors. The percentage of PD1+ T cells (P<0.01) and Ly6C+I-A+ cells (P<0.05) were also increased.

108

6.3 Conclusion

By injecting STO-609, the CaMKK2-selective inhibitor, into the C57BL/6 mice with lymphoma or breast cancer, we successfully suppressed the tumor growth as hypothesized. The similarity of tumor growing curves between the ones generated from

CaMKK2 WT and KO mice and the ones from STO-609 treated C57BL/6 mice strongly suggests that STO-609 suppressing tumor growths by inhibiting CaMKK2. The analysis of STO-609 treated breast tumors component also showed the changes of tumor microenvironment with increased T-cell infiltration and better myeloid maturation.

Unlike the previous published studies that had mainly focused on the role of

CaMKK2 pathway inside the tumor cells or the direct effect of STO-609 on tumor cells, our study reveled the CaMKK2-mediated regulation of tumor microenvironment.

CaMKK2 inhibition in tumor microenvironment increases anti-tumor immunity by facilitating myeloid cells maturation and T-cell accumulation. The impact of STO-609 on different tumors also reveals its trememdous potential as an immune regulator in future cancer treatment.

109

7. Discussion and Future Perspectives

7.1 Summary of findings

CaMKK2, as a crucial regulator in calcium-signaling, energy metabolic sensor, and inflammation mediator, has been discussed in many publications and briefly reviewed in the first chapter of this dissertation. All the previous CaMKK2-related studies focus on solid tumors and the role of CaMKK2 as a signaling kinase in malignant cells. Considering that CaMKK2 also mediates other cell activities including differentiation and inflammation (Racioppi, 2013; Racioppi et al., 2012; Teng et al., 2011), we believed that the role of CaMKK2 in cancer progression is underappreciated. In this dissertation, we used E.G7 lymphoma cell line in different in vivo and in vitro models with CaMKK2 ablation or inhibition to study how CaMKK2 impacts lymphoma microenvironment. For the first time we revealed that CaMKK2 facilitates lymphoma progression by inhibiting myeloid cells maturation, promoting MDSC generation, and suppressing dendritic cell functions, thus reshaping the lymphoma microenvironment.

There are several crucial findings in our research that advanced the knowledge in this field.

First, instead of mediating tumor cells growth as reported before, CaMKK2 reshapes the tumor microenvironment. In Chapter 3, we injected E.G7 subcutaneously in the flank or intracranially in the brain of the WT and CaMKK2-/- mice. The tumor cells were successfully engrafted in both genotypes (Figure 2); however, the progression in

110

the later phase was suppressed in CaMKK2-/- mice. (Figure1, Figure 3) There was no difference between the intracranial model and the flank model, indicating that CaMKK2 does not regulate tumor cells through the expression in parenchymal cells, but through a more systemic mechanism, for example, impacting the immune system. This hypothesis was confirmed by the lymphoma suppression in LyMCre+;CaMKK2fl/fl mice, a strain that

CaMKK2 is specifically ablated in the myeloid lineage. (Figure 5) Using anti-CD8 cell depletion could reverse tumor suppression in CaMKK2-/- tumor-bearing mice. (Figure 8)

These findings correlate the previous report of CaMKK2 expression in several immune cells including macrophages, and can regulate macrophages activation and induce

TAMs. (Racioppi et al., 2012) Our results further confirmed that CaMKK2 regulates myeloid cell development. CaMKK2 ablation limits MDSC generation in vivo and in vitro in the lymphoma microenvironment (Figure 9-14). We also demonstrated that CaMKK2 is expressed in splenic DCs and in vitro generated BMDCs (Figure 15-18). Considering that DCs are the most important APCs in the myeloid compartment to stimulate T-cell function, this information may provide a new aspect to understanding how CaMKK2- mediated tumor progression.

Second, in addition to adiopogenesis and granulopoiesis, we proved that

CaMKK2 also regulates dendritic cell differentiation and maturation through AMPK pathway. Compared to WT BMDCs, BMDCs generated from CaMKK2-/- bone marrow in the presence of E.G7 supernatant were more maturated by showing higher percentage of

111

I-A+, costimulatory factor expression, and IL-12 secretion when stimulated by T cells or

LPS. By analyzing the splenic DC subsets from CaMKK2-EGFP mice, we found that

CaMKK2 promoter was active in the splenic CD11c+ DCs, and the promoter activity was developmentally decreased as the DCs matured - while CaMKK2 expression was high in

DC progenitors, it was dramatically decreased once DCs matured with upregulated I-A expression; CaMKK2 reporter activity was very low in conventional DCs and plasmacytoid DCs. This information suggests that CaMKK2 is important in regulating

DC differentiation, but less in regulating the mature DC functions.

The same phenomenon has also been observed in the previous studies in adipogenesis and granulopoiesis. During adipogenesis, CaMKK2 is expressed in preadipocytes but not in mature adipocytes. CaMKK2 ablation or inhibition in preadipocytes enhances th differentiation of preadipocytes into mature adipocytes by reducing preadipocyte factor 1 and Sox9, thus increases the expression of C/EBPβ and

C/EBPδ, and resulting in accelerated adipogenesis. AMPK is the downstream signaling pathway in this regulation. (Lin et al., 2011) Similar to adiposity regulation, CaMKK2 expression is high in hematopoietic stem cells and myeloid progenitors, but gradually decreases and eventually becomes undetectable as granulocytes differentiate and mature. CaMKK2 ablation results in increased CMP number and accelerated granulopoietic phenotype in the marrow. In bone marrow transplantation and in vitro induction, CaMKK2-/- progenitors have the intrinsic tendency for granulocytic fate

112

commitment and differentiation. Although C/EBPα and PU.1 expressions are elevated in

CaMKK2-/- myeloid progenitors, the inhibitory effect of CaMKK2 in granulocytic differentiation is independent of the conventional targets including AMPK, CaMKI, and

CaMKIV. (Teng et al., 2011) Our finding of CaMKK2 expression in splenic DCs correlates the CaMKK2 expression with the changes in both adipogensis and granulopoiesis. This piece of data further supports that CaMKK2 mediates cell lineage commitment, differentiation, and maturation. And this phenomenon is likely to be universal instead of being limited to specific cell types.

Furthermore, we found that the AMPK phosphorylation was significantly lower in CaMKK2-/- BMDCs. Along with the decreased in phosphor-AMPK, the phosphorylation of AMPK’s downstream targets such as ACC and S6 were also changed. Unlike ACC that is directly phosphorylated by AMPK, S6 phosphorylation is a marker for mTORC1 activity and therefore is an indirect downstream target of AMPK.

These changes implicates that CaMKK2 ablation inhibits AMPK phosphorylation and enhances mTOR activity, which may change the metabolism status inside the BMDCs.

Third, CaMKK2-mediated tumor suppression may be universal. To eliminate the possibility of the E.G7 cells in CaMKK2-/- mice were eradicated as a cell-line specific incidence, we used Vk*Myc myeloma cells and E0771 breast cancer cells as additional cancer models. Similar to the E.G7 cells, both myeloma cells and breast cancer cells were suppressed in the CaMKK2-/- mice, indicating that the CaMKK2-mediated tumor

113

suppression is universal in different cancers. STO-609, the pharmacological inhibitor of

CaMKK2, also efficiently suppressed E.G7 and E0771 tumor progression. In addition,

STO-609 altered the microenvironment of E0771 tumor by attracting more T cells into the tumor and increasing the percentages of I-A+ cells in the Ly6C+ compartment. These findings provide evidence for STO-609 to become a potential immune regulator in cancer treatment.

7.2 Potential applications and future perspectives.

7.2.1 Inhibiting CaMKK2 can inhibit AMPK specifically in myeloid cells

AMPK activation has been reported to inhibit DC maturation. The metabolic transition from oxidative phosphorylation (OXPHOS) to aerobic glycolysis is essential for TLR-mediated DC activation and maturation. PI3K/Akt signaling mediates the metabolic switch, while AMPK and IL-10 can inhibit it. (Krawczyk et al., 2010) AMPK phosphorylation diminishes rapidly as DCs exposed to LPS. Using shRNA to reduce

AMPK level promotes LPS-induced DC activation and increased the expression of IL-

12p40 and CD86, while activating AMPK with AICAR inhibites the expression of LPS- mediated IL-12p40. IL-10 can interfere with the glycolysis and suppress the DC activation. Carroll et al. later used AMPKα1-deficient mice to show that AMPKα1- deficient macrophages and DCs are more potent in stimulating an inflammatory response and enhance the capacity for antigen presentation to promote Th1 and Th17 responses. (Carroll, Viollet, & Suttles, 2013) In response to TLR and CD40 stimulation,

114

AMPKα1-deficient macrophges and DCs have higher levels of proinflammatory cytokines and decreased production of IL-10. These results indicate that AMPK activation inhibits TLR-mediated DC activation and maturation. In CaMKK2-/- BMDCs,

AMPK activation is inhibited due to the reduced phosphorylation, which may directly contribute to TLR-dependent DC activation and maturation.

The downstream targets of AMPK are also involved in DCs maturation. The mammalian target of rapamycin (mTOR) is usually activated by RAS/PI3K/Akt pathway to regulate cell survival and proliferation. AMPK activation can inhibit mTOR complex 1

(mTORC1) by inhibiting its upstream kinase, the tuberous sclerosis complex (TSC), thus reduces the phosphorylation of mTOR downstream targets including S6K1 and 4E-BP1. mTOR has been reported as a necessity for maturation of GM-CSF/IL-4 induced BMDCs.

Rapamycin, a mTOR inhibitor, can interurupt this maturation process. The upregulation of MHC and costimulatory molecules are suppressed in BMDCs when cultured in the presence of rapamycin. (Hackstein et al., 2003. Taner, 2005 #1211) Furthermore, rapamycin-matured DCs fail to promote T-cell activation; instead, they induce T-cell tolerance and promote the Foxp3+ Treg generation. (Turnquist et al., 2007) In our experiment, although the direct effect of mTOR using rapamycin inhibition is not tested, the increased S6 phosphorylation suggests an increased activity of mTOR in CaMKK2-/-

BMDCs. Other findings in our study in DC maturation and functions in CaMKK2-/-

BMDCs are also supported by the aforementioned publications.

115

Acetyl-CoA Ccarboxylase-1 (ACC1) is a direct downstream target of AMPK that controls lipogenesis. There is increasing evidence to show that lipid production and consumption are critical in DC biology. Inhibiting lipid-body assembly or preventing triacylglycerol (TAG) accumulation abrogates the MHCI cross-presenting capability of the DCs. (Bougneres et al., 2009) Rehman et al. later showed that blockading fatty-acid synthesis increases DC capacity to secrete selective proinflammatory cytokines including

IL-12 and MCP-1, and better induced CTL responses in T cells and NK cells. (Rehman et al., 2013) By analyzing the DCs from mouse models as well as cancer patients with lymphoma, colon, and breast cancers, Herber et al. found that the lipid level, especially

TAG level, was elevated in DCs. (Herber et al., 2010) Using ACC1 inhibitor to normalize its expression in DCs can restore the functional activity of lipid-laden DCs, and enhance their potency when used as a cancer vaccine. ACC1 phosphorylation is decreased in

CaMKK2-/- BMDCs. This result implicates an inhibited state of lipid synthesis, although further supportive data will be needed.

The inhibitory role of AMPK in regulating immune cell maturation and function provides a new target for potential immune therapy. However, the role of AMPK in regulating cancers is controversial. As an important downstream target of LKB1, a crucial tumor suppressor, AMPK activation can inhibit tumor cell progression by limiting glycolysis and reversing the Warburg effect. (Evans, Donnelly, Emslie-Smith,

Alessi, & Morris, 2005; Faubert et al., 2013; Faubert, Vincent, Poffenberger, & Jones, 2015;

116

Hardie, 2015; Liang & Mills, 2013) Moreover, AMPK is a ubiquitous and crucial enzyme in the physiological states for most of the cells, especially in several vital organs such as liver, hearts, and muscles. It is impossible to directly target AMPK body-wide and reverses its function specifically in the immune cells. Although further studies are needed to verify the role of AMPK in CaMKK2 deficient cells, our study for the first time reveals a possibility to target AMPK specifically in myeloid cells to augment the immune response, but not affecting its normal function in other tissues.

7.2.2 Inhibition of CaMKK2 decreases MDSC generation and function in lymphoma

MDSCs are a collection of immature cells originataed from the myeloid compartment that accumulate in the body under certain pathological conditions. These immature myeloid cells have decreased expression of MHC-II and costimulatory factors, thus lack of the capability to activate T cells but induce Treg cells. In mice, MDSCs are generally described as CD11b+Gr-1+ cells. Based on the Gr-1 isoforms, MDSCs are further categorized into Gr-1int Mo-MDSCs and Gr-1high PMN-MDSCs. While PMN-

MDSCs are highly dependent on arginase 1 (ARG1) myeloperoxidase to produce reactive oxygen species (ROS), Mo-MDSC can also up-regulate inducible nitric oxide synthase (iNOS) to generate NO. Both ARG1 and iNOS can produce peroxynitrites. The excessive ROS and peroxynitrite produced by MDSC can lead to T-cell unresponsiveness, causing inflammation, and inducing TAMs generation. (Gabrilovich

& Nagaraj, 2009; Talmadge & Gabrilovich, 2013) 117

In lymphoma, the accumulation of MDSCs has been noted. Among different cancer cell lines, the supernatant from EL4 murine lymphoma cells is the strongest inducer for MDSC in vitro. (Youn et al., 2008) Inoculating mice with EL4 cells also induces significant MDSC accumulation. (Youn et al., 2008) In the A20 B-cell lymphoma mice model, MDSCs serves as the tolerogenic APCs and induces Treg. (Serafini et al.,

2008) Romano et al. reported that in Hodgkin Lymphoma patients, all circulating MDSC subsets are increased. (Romano et al., 2015) Sato et al. observed similar result in non-

Hodgkin lymphoma, in which the increased MDSCs were inversely correlated with the number of NK cells. In diffuse B-cell lymphoma (DLBCL) patients, increased level of M-

MDSCs has also been noted and has been associated with poor prognosis. (Tadmor et al., 2013) Wu et al. later support this result based on their study with an independent

DLBCL patient cohort (C. Wu et al., 2015), and Zhang et al. also observed the same trend in extranodal NK/T cell lymphoma patients (H. Zhang et al., 2015). Recently, Zhang et al. reported that the CD14+HLA-DRlow/neg monocytes in lymphoma patients cross-talked with lymphoma cells through Hsp27, and subsequently induced chemotherapy resistance. (Z. J. Zhang et al., 2015) All these findings point to the crucial role of MDSCs in lymphoma, and targeting MDSCs could be a potential treatment option for the disease.

Therapeutic targeting of MDSCs includes at least four aspects: 1) promoting myeloid-cell differentiation using all-trans retinoic acid (ATRA); 2) inhibiting MDSC

118

expansion by neutralizing tumor-derived factors; 3) inhibiting MDSC function by blocking COX2 to intervene the signaling pathways for suppressive factors production;

4) eliminating MDSCs using chemotherapy such as gemcitabine. (Gabrilovich &

Nagaraj, 2009)

In our previous discussion, we already illustrated that CaMKK2 is an intrinsic inhibitor for granulocytosis and myeloid cell differentiation in cancer. CaMKK2-/-mice inoculated with E.G7 cells had fewer MDSCs accumulation compared to WT mice. In in vitro culture with E.G7 supernatant, CaMKK2-/- marrow cells and KSL cells also showed higher percentage of I-A+. These evidences indicate that inhibiting CaMKK2 can reduce

MDSC generation and accumulation in cancer environment.

AMPK phosphorylation was significantly decreased in CaMKK2-/- BMDCs, indicating that AMPK activity could be regulated by CaMKK2 in BMDCs, and is more likely in MDSC as well. Similar to dendritic cells, the activation of MDSCs requires high central carbon metabolism activity including glycolysis, glutaminolysis and TCA cycle activity. AMPK is the key enzyme for this metabolic transition. The immunosuppressive activity of MDSCs mainly depends on inducible nitric oxide synthase (iNOS) and arginase 1 (ARG1). These two enzymes can convert L-arginine (L-Arg) to nitric oxide

(NO) or L-ornithine, resulting in the accumulation of NO derivatives to inhibit T-cell immunity. Hammani et al (Hammami et al., 2012) reported that AMPK phosphorylation was increased during MDSC maturation along with the increased L-Arg metabolic

119

enzymes activity, and inhibiting AMPK by its inhibitor Compound C resulted in the inhibition of L-Arg metabolizing enzymes activity, thus abolishing MDSCs immunosuppression. This for the first time provided the evidence for the directly regulation of MDSC immunosuppressive activity by AMPK. Besides glycolysis, AMPK also enhances fatty acid uptake and oxidation. (O'Neill, Holloway, & Steinberg, 2013)

Hossain et al. reported that fatty acid uptake and activated fatty acid oxidation were increased in tumor-infiltrated MDSCs. Recently, Wu et al. reported that blocking mTOR, a downstream target inhibited by AMPK, interrupts the differentiation and immunosuppressive function of M-MDSC in both allogeneic skin grafting model and cancer model. Inhibiting mTOR with rapamycin dramatically reduces iNOs and Arg-1 expression in M-MDSCs, and also significantly blocks the glycolytic pathway during M-

MDSC induction. Using Metformin, an AMPK activator, on the contrary, can rescue the inhibitory effect of rapamycin on M-MDSC differentiation, and significantly improved iNOS expression and immunosuppressive function. (T. Wu et al., 2016) All these data support that AMPK-mTOR pathway is an intrinsic factor essential for M-MDSC differentiation and function.

Furthermore, tumor-derived lactic acid, a byproduct of glycolysis, has been reported to induce M2 tumor-associated macrophages and MDSC. (Colegio et al., 2014;

Husain, Huang, Seth, & Sukhatme, 2013) It is well accepted that AMPK promotes macrophage polarization to an anti-inflammatory functional phenotype and increases

120

the secretion of IL-6, IL-10, TNF-α. (Sag et al., 2008) These anti-inflammatory cytokines were crucial in inducing MDSC generation and activation. Thus, inhibiting AMPK in macrophages can reverse the anti-inflammatory state, depriving the MDSCs from the necessary cytokines, and thus reducing MDSC generation. Based on all the evidence stated above, we are convinced that AMPK is a crucial element in MDSC generation and function. In CaMKK2-/- mice, the inhibition of AMPK in myeloid cells may be strongly relevant to the decreased generation of MDSC.

In conclusion, targeting CaMKK2 is a new potential strategy to control MDSC generation and function in tumor microenvironment.

7.2.3 Inhibition of CaMKK2 provides a potential way to enhance DC vaccine efficacy in lymphoma by promoting DCs maturation and function

In lymphoma, the role of dendritic cells in regulating tumor progression has been noted for a long time. In cutaneous T-cell lymphoma (CTCL), immature DCs have been reported to be essential in sustaining the CTCL survival, proliferation, and malignancy in vitro.(C. L. Berger et al., 2002(Edelson, 2001 #1166) Maturing these DCs in the culture system through the activation of TCR or CD40 on DCs enhanced mixed lymphocyte culture stimulation, and eventually leaded to apoptotic tumor cell death. To sustain the survival and immature status of the DCs, CTCL secretes IL-10 to prevent DC maturation. At the same time, the immature DCs can phagocytose the apoptotic cancer cells and present in the class II complexes as an antigen, to drive the mutated CD4

121

malignant colony to proliferation. This balance will be destroyed once the DCs are matured. In a recent study, Schlapbach et al. found a significant increase of all DC subsets from different DC origins in the lesion of human patient CTCL samples.

(Schlapbach et al., 2010) While various types of DCs are found to be located in close proximity to tumor cells, the interstitial DCs with immature markers have been shown to be in contact with FoxP3+CD25+ Tregs. In light of the important role of DCs in lymphoma, different methods have been used to boost DC functions, leading to the development of DC vaccine therapy in lymphoma. Hsu et al. pulsed autologous DCs ex vivo with tumor-specific idiotype protein, and transfused them into 4 patients with follicular B-cell lymphomas as a vaccine therapy. (F. J. Hsu et al., 1996) Anti-tumor cellular immune responses were observed in all patients, and the clinical outcomes of the patients were encouraging. In 2002, the same group from Stanford University reported another clinical trial with 35 patients treated with the same approach, and concluded that Id-pulsed DC vaccination can induce T-cell responses and durable tumor regression. (Timmerman et al., 2002) In a recent pilot study of three previously treated patients with adult T cell leukemia/lymphoma, Suehiro et al. also reported that Tax peptide-pulsed DC vaccine can increase tumor-specific cytotoxic T lymphocyte response without severe adverse events, and confirmed that Tax peptide-pulse DC vaccine is a safe and promising immunotherapy for ATL. (Suehiro et al., 2015) Instead of peptide- pulsed DC vaccines, Gatza et al. used DC pulsed with tumor cell lysate from C6VL, a

122

murine T-cell lymphoma line to immune mice and then challenged the mice with C6VL lymphoma. (Gatza & Okada, 2002) They found that the C6VL lysate-pulsed DCs stimulate tumor-specific cellular response, provide a more efficient vaccination against tumor compared to DC vaccine pulsed with tumor idiotype protein, and are therapeutic in tumor-bearing mice. One year later, intranodal injection of autologous tumor-lysate- pulsed dendritic cells was reported in 8 CTCL patients. (Maier et al., 2003) All the patients developed tumor-specific delayed-type hypersensitivity reaction, and most reached objective clinical response. Intranodal injection of autologous tumor-lysate- pulsed DCs is considered well-tolerated and immunologically effective in CTCL patients. In a recent report, Hira et al. found a significantly increased vaccine efficacy in lymphoma-bearing mice by using the combination therapy of recombinant IL-15 (rIL-15) and DCs pre-activated with recombinant IL-15 (rIL-15), and then pulsed with whole tumor cell lysates. As of March 2016, the search of “cancer vaccines” at http://clinicaltrials.gov yields 1,743 results, in which 378 studies are about “dendritic cell vaccine”, including 102 studies related to hematopoietic malignancies. It is well accepted that facilitating DC maturation by vaccine adjuvants or DC activators such as CD40 agonists is the key of success to establishing immunity rather than tolerance. (Palucka &

Banchereau, 2013)

Among the cytokines secreted by mature DCs, IL-12 is an absolute requirement for TLR signaling. IL-12 secreted by DCs or macrophages is essential to induce CD4+

123

naïve cells towards the T helper 1 (TH1) phenotype for the secretion of IFN-γ. (Hsieh et al., 1993; Macatonia et al., 1995) Newly activated CD4+ T cells can amplify but not initiate this effect. (Ardavin et al., 2004) IL-12 secretion in DCs is triggered by the independent ligation of either CD40 or MHC II molecules, and inhibited by IL-4 and IL-

10. (Koch et al., 1996; Manetti et al., 1993) When studying the role of T-bet in mediating

TH1 cell commitment, Mullen et al. discovered that IL-12/STAT4 not only serves as a growth signal to induce TH1 survival and proliferation, but is as a trans-activator to prolong IFN-γ synthesis through interacting with CREB-binding protein, an IFN-γ gene coactivator.(Mullen et al., 2001) As early as in 1999, Melero et al. reported that injecting

DCs with IL-12 producing adenoviral transfection into mouse tumor nodules induced by colon cancer cell line could successfully eradicate 50-100% well-established tumor malignancies. (Melero et al., 1999) This finding was later confirmed in other animal models and clinical trials (Carreno et al., 2013; Gollob et al., 2000; Lasek, Zagozdzon, &

Jakobisiak, 2014; Mortarini et al., 2000; Portielje et al., 1999; Tugues et al., 2015; S. N.

Zhang et al., 2011) More clinical trials with IL-12 therapy in various solid tumors are on going. Promoting DC maturation to increase IL-12 secretion may provide a new way to augment the efficacy of DC vaccination.

In our study, CaMKK2-/- dendritic cells, including in BMDCs and splenic DCs, showed better maturation and T-cell stimulation functions. When stimulating BMDCs with LPS to activate TLR signaling, we observed a significant increase in the secretion of

124

most chemokines and cytokines in CaMKK2-/- cells. IL-12 and IL-15, the two important cytokines that stimulate T-cell activation and expansion, were significantly elevated; while IL-4 and IL-10, the two inhibitory factors for TH1 differentiation and IL-12 secretion, remained stable. Considering CaMKK2-/- BMDCs have higher MHCII and

CD40 expression, it is not surprising that when cocultured with T cells, CaMKK2-/-

BMDCs yielded higher IL-12 secretion, better induced IFN-γ from T cells, and better stimulated T-cell activation and proliferation.

Our results suggest that inhibiting CaMKK2 in lymphoma environment can be a new strategy to induce DC maturation and enhance DC functions, thus augmenting the

DC vaccine efficacy.

7.2.3 STO-609 is a potential immune therapy for cancers

STO-609, the pharmacological inhibitor of CaMKK2 has been used in different cancer models including prostate cancer, liver cancer, gastric cancer, and breast cancer.

STO-609 inhibits cancer cell proliferation and progression by blocking CaMKK2 thus interfering the phosphorylation of the downstream signaling including CaMKI and

AMPK. In all these models, STO-609 has been proven to successfully control tumor growth in vitro and in vivo.

In our research, we further demonstrate that STO-609 is an efficient therapy to suppress E.G7 lymphoma and E0771 breast cancer by changing the tumor environment.

In the previous sections, we already discussed the rationale for inhibiting CaMKK2 as a

125

cancer therapy to decrease MDSC accumulation and function, increase DC maturation and function to improve T-cell anti-tumor function. According to our previous publications and the data from www.immgen.org, outside of the brain, CaMKK2 expression is highly restricted to certain immune cells, hematopoietic progenitors. This fact provides a possibility to specifically target AMPK in the myeloid cells to enhance

APC functions, while not interfering the physiological function of AMPK in other tissue and organs.

Intriguingly, when analyzing the breast cancer component, we found more T cells, especially CD8 T cells, accumulated in the tumor tissues in the STO-609 treated mice. However, in these T cells, the percentage of Program Death 1 (PD-1) positive was also elevated. PD-1 was first discovered in activated T and B lymphocytes and considered as lymphocyte activation marker. It was later recognized to inhibit T-cell function and induce T-cell exhaustion. (Agata et al., 1996; Goldberg et al., 2007; Ishida,

Agata, Shibahara, & Honjo, 1992) Monoclonal antibodies targeting PD-1 and its ligand

PD-L1 have been proven efficientive in controlling cancer growth in multiple clinical trials. The presence of tumor-infiltrating lymphocytes (TIL) in breast cancer specimens has been shown to be an important marker for better prognosis in breast cancer. (Adams et al., 2014; Liedtke et al., 2008; Loi et al., 2014) TIL recruitment is suppressed by the Ras-

MAPK pathway. Recently, Loi et al. reported that it was more efficient to combine anti-

PD-1/PD-L1 immunotherapy with Ras-MAPK pathway inhibitor in the triple-negative

126

breast cancer (TNBC) mouse models. In our model, STO-609 successfully changed the tumor microenvironment and recruited more lymphocytes to the lesion. The increased

T-cell infiltration with higher percentage of PD-1 indicates the possibility for a new combined immunotherapy using STO-609 with the immune checkpoint blockade such as anti-CTLA-4 antibody or anti-PD-1/PD-L1 antibody.

In conclusion, using STO-609 to inhibit CaMKK2 can provide a new therapeutic alternative that covers multiple mechanisms to impact tumor microenvironment and inhibit tumor progression.

7.3 Conclusion

Using E.G7 as a lymphoma model, our study successfully reveals that CaMKK2 regulates lymphoma progression by inhibiting dendritic cell maturation and inducing

MDSCs. Inhibition of CaMKK2 can facilitate DC differentiation and maturation, reduce

MDSC accumulation, thus reshape the lymphoma microenvironment. The use of STO-

609 could inhibit CaMKK2 and therefore provides a new potential strategy for lymphoma immune therapy.

127

References

Adams, S., Gray, R. J., Demaria, S., Goldstein, L., Perez, E. A., Shulman, L. N., . . . Badve, S. S. (2014). Prognostic value of tumor-infiltrating lymphocytes in triple-negative breast cancers from two phase III randomized adjuvant breast cancer trials: ECOG 2197 and ECOG 1199. J Clin Oncol, 32(27), 2959-2966. doi: 10.1200/JCO.2013.55.0491

Agata, Y., Kawasaki, A., Nishimura, H., Ishida, Y., Tsubata, T., Yagita, H., & Honjo, T. (1996). Expression of the PD-1 antigen on the surface of stimulated mouse T and B lymphocytes. Int Immunol, 8(5), 765-772.

Ali, A., Patterson, S., Cruickshank, K., Rudge, P., Dalgleish, A. G., & Knight, S. C. (1993). Dendritic cells infected in vitro with human T cell leukaemia/lymphoma virus type-1 (HTLV-1); enhanced lymphocytic proliferation and tropical spastic paraparesis. Clin Exp Immunol, 94(1), 32-37.

Almand, B., Clark, J. I., Nikitina, E., van Beynen, J., English, N. R., Knight, S. C., . . . Gabrilovich, D. I. (2001). Increased production of immature myeloid cells in cancer patients: a mechanism of immunosuppression in cancer. J Immunol, 166(1), 678-689.

Anderson, K. A., Lin, F., Ribar, T. J., Stevens, R. D., Muehlbauer, M. J., Newgard, C. B., & Means, A. R. (2012). Deletion of CaMKK2 from the liver lowers blood glucose and improves whole-body glucose tolerance in the mouse. Mol Endocrinol, 26(2), 281-291. doi: 10.1210/me.2011-1299

Anderson, K. A., Means, R. L., Huang, Q. H., Kemp, B. E., Goldstein, E. G., Selbert, M. A., . . . Means, A. R. (1998). Components of a calmodulin-dependent protein kinase cascade. Molecular cloning, functional characterization and cellular localization of Ca2+/calmodulin-dependent protein kinase kinase beta. J Biol Chem, 273(48), 31880-31889.

Anderson, K. A., Ribar, T. J., Illario, M., & Means, A. R. (1997). Defective survival and activation of thymocytes in transgenic mice expressing a catalytically inactive

128

form of Ca2+/calmodulin-dependent protein kinase IV. Mol Endocrinol, 11(6), 725- 737. doi: 10.1210/mend.11.6.0011

Anderson, K. A., Ribar, T. J., Lin, F., Noeldner, P. K., Green, M. F., Muehlbauer, M. J., . . . Means, A. R. (2008). Hypothalamic CaMKK2 contributes to the regulation of energy balance. Cell Metab, 7(5), 377-388. doi: 10.1016/j.cmet.2008.02.011

Ansell, S. M., Lesokhin, A. M., Borrello, I., Halwani, A., Scott, E. C., Gutierrez, M., . . . Armand, P. (2015). PD-1 blockade with nivolumab in relapsed or refractory Hodgkin's lymphoma. N Engl J Med, 372(4), 311-319. doi: 10.1056/NEJMoa1411087

Ardavin, C., Amigorena, S., & Reis e Sousa, C. (2004). Dendritic cells: immunobiology and cancer immunotherapy. Immunity, 20(1), 17-23.

Armand, P. (2015a). Checkpoint blockade in lymphoma. Hematology Am Soc Hematol Educ Program, 2015(1), 69-73. doi: 10.1182/asheducation-2015.1.69

Armand, P. (2015b). Immune checkpoint blockade in hematologic malignancies. Blood, 125(22), 3393-3400. doi: 10.1182/blood-2015-02-567453

Armand, P., Nagler, A., Weller, E. A., Devine, S. M., Avigan, D. E., Chen, Y. B., . . . Gordon, L. I. (2013). Disabling immune tolerance by programmed death-1 blockade with pidilizumab after autologous hematopoietic stem-cell transplantation for diffuse large B-cell lymphoma: results of an international phase II trial. J Clin Oncol, 31(33), 4199-4206. doi: 10.1200/JCO.2012.48.3685

Arulogun, S., Hertzberg, M., & Gandhi, M. K. (2016). Recent treatment advances in Hodgkin Lymphoma: a concise review. Intern Med J. doi: 10.1111/imj.13051

Atakhorrami, M., Rahimi-Aliabadi, S., Jamshidi, J., Moslemi, E., Movafagh, A., Ohadi, M., . . . Darvish, H. (2015). A genetic variant in CAMKK2 gene is possibly associated with increased risk of bipolar disorder. J Neural Transm (Vienna). doi: 10.1007/s00702-015-1456-7

129

Banchereau, J., Briere, F., Caux, C., Davoust, J., Lebecque, S., Liu, Y. J., . . . Palucka, K. (2000). Immunobiology of dendritic cells. Annu Rev Immunol, 18, 767-811. doi: 10.1146/annurev.immunol.18.1.767

Banchereau, J., & Steinman, R. M. (1998). Dendritic cells and the control of immunity. Nature, 392(6673), 245-252. doi: 10.1038/32588

Berger, C. L., Hanlon, D., Kanada, D., Dhodapkar, M., Lombillo, V., Wang, N., . . . Edelson, R. (2002). The growth of cutaneous T-cell lymphoma is stimulated by immature dendritic cells. Blood, 99(8), 2929-2939.

Berger, R., Rotem-Yehudar, R., Slama, G., Landes, S., Kneller, A., Leiba, M., . . . Nagler, A. (2008). Phase I safety and pharmacokinetic study of CT-011, a humanized antibody interacting with PD-1, in patients with advanced hematologic malignancies. Clin Cancer Res, 14(10), 3044-3051. doi: 10.1158/1078-0432.CCR-07- 4079

Berridge, M. J., Bootman, M. D., & Roderick, H. L. (2003). Calcium signalling: dynamics, homeostasis and remodelling. Nat Rev Mol Cell Biol, 4(7), 517-529. doi: 10.1038/nrm1155

Bougneres, L., Helft, J., Tiwari, S., Vargas, P., Chang, B. H., Chan, L., . . . Guermonprez, P. (2009). A role for lipid bodies in the cross-presentation of phagocytosed antigens by MHC class I in dendritic cells. Immunity, 31(2), 232-244. doi: 10.1016/j.immuni.2009.06.022

Brasel, K., De Smedt, T., Smith, J. L., & Maliszewski, C. R. (2000). Generation of murine dendritic cells from flt3-ligand-supplemented bone marrow cultures. Blood, 96(9), 3029-3039.

Braun, A. P., & Schulman, H. (1995). The multifunctional calcium/calmodulin-dependent protein kinase: from form to function. Annu Rev Physiol, 57, 417-445. doi: 10.1146/annurev.ph.57.030195.002221

130

Bronte, V. (2009). Myeloid-derived suppressor cells in inflammation: uncovering cell subsets with enhanced immunosuppressive functions. Eur J Immunol, 39(10), 2670-2672. doi: 10.1002/eji.200939892

Cao, W., Sohail, M., Liu, G., Koumbadinga, G. A., Lobo, V. G., & Xie, J. (2011). Differential effects of PKA-controlled CaMKK2 variants on neuronal differentiation. RNA Biol, 8(6), 1061-1072. doi: 10.4161/rna.8.6.16691

Carreno, B. M., Becker-Hapak, M., Huang, A., Chan, M., Alyasiry, A., Lie, W. R., . . . Linette, G. P. (2013). IL-12p70-producing patient DC vaccine elicits Tc1-polarized immunity. J Clin Invest, 123(8), 3383-3394. doi: 10.1172/JCI68395

Carroll, K. C., Viollet, B., & Suttles, J. (2013). AMPKalpha1 deficiency amplifies proinflammatory myeloid APC activity and CD40 signaling. J Leukoc Biol, 94(6), 1113-1121. doi: 10.1189/jlb.0313157

Cary, R. L., Waddell, S., Racioppi, L., Long, F., Novack, D. V., Voor, M. J., & Sankar, U. (2013). Inhibition of Ca(2)(+)/calmodulin-dependent protein kinase kinase 2 stimulates osteoblast formation and inhibits osteoclast differentiation. J Bone Miner Res, 28(7), 1599-1610. doi: 10.1002/jbmr.1890

Cella, M., Sallusto, F., & Lanzavecchia, A. (1997). Origin, maturation and antigen presenting function of dendritic cells. Curr Opin Immunol, 9(1), 10-16.

Chen, C. D., Welsbie, D. S., Tran, C., Baek, S. H., Chen, R., Vessella, R., . . . Sawyers, C. L. (2004). Molecular determinants of resistance to antiandrogen therapy. Nat Med, 10(1), 33-39. doi: 10.1038/nm972

Chen, Y. Y., Lee, M. H., Hsu, C. C., Wei, C. L., & Tsai, Y. C. (2012). Methyl cinnamate inhibits adipocyte differentiation via activation of the CaMKK2-AMPK pathway in 3T3-L1 preadipocytes. J Agric Food Chem, 60(4), 955-963. doi: 10.1021/jf203981x

Chesi, M., Robbiani, D. F., Sebag, M., Chng, W. J., Affer, M., Tiedemann, R., . . . Bergsagel, P. L. (2008). AID-dependent activation of a MYC transgene induces

131

multiple myeloma in a conditional mouse model of post-germinal center malignancies. Cancer Cell, 13(2), 167-180. doi: 10.1016/j.ccr.2008.01.007

Clausen, B. E., Burkhardt, C., Reith, W., Renkawitz, R., & Forster, I. (1999). Conditional gene targeting in macrophages and granulocytes using LysMcre mice. Transgenic Res, 8(4), 265-277.

Colegio, O. R., Chu, N. Q., Szabo, A. L., Chu, T., Rhebergen, A. M., Jairam, V., . . . Medzhitov, R. (2014). Functional polarization of tumour-associated macrophages by tumour-derived lactic acid. Nature, 513(7519), 559-563. doi: 10.1038/nature13490

Condamine, T., Kumar, V., Ramachandran, I. R., Youn, J. I., Celis, E., Finnberg, N., . . . Gabrilovich, D. I. (2014). ER stress regulates myeloid-derived suppressor cell fate through TRAIL-R-mediated apoptosis. J Clin Invest, 124(6), 2626-2639. doi: 10.1172/JCI74056

Corinti, S., Albanesi, C., la Sala, A., Pastore, S., & Girolomoni, G. (2001). Regulatory activity of autocrine IL-10 on dendritic cell functions. J Immunol, 166(7), 4312- 4318.

Coussens, L. M., Zitvogel, L., & Palucka, A. K. (2013). Neutralizing tumor-promoting chronic inflammation: a magic bullet? Science, 339(6117), 286-291. doi: 10.1126/science.1232227

Damuzzo, V., Pinton, L., Desantis, G., Solito, S., Marigo, I., Bronte, V., & Mandruzzato, S. (2015). Complexity and challenges in defining myeloid-derived suppressor cells. Cytometry B Clin Cytom, 88(2), 77-91. doi: 10.1002/cyto.b.21206

Diaz-Montero, C. M., Salem, M. L., Nishimura, M. I., Garrett-Mayer, E., Cole, D. J., & Montero, A. J. (2009). Increased circulating myeloid-derived suppressor cells correlate with clinical cancer stage, metastatic tumor burden, and doxorubicin- cyclophosphamide chemotherapy. Cancer Immunol Immunother, 58(1), 49-59. doi: 10.1007/s00262-008-0523-4

132

Doedens, A. L., Stockmann, C., Rubinstein, M. P., Liao, D., Zhang, N., DeNardo, D. G., . . . Johnson, R. S. (2010). Macrophage expression of hypoxia-inducible factor-1 alpha suppresses T-cell function and promotes tumor progression. Cancer Res, 70(19), 7465-7475. doi: 10.1158/0008-5472.CAN-10-1439

Dolcetti, L., Peranzoni, E., Ugel, S., Marigo, I., Fernandez Gomez, A., Mesa, C., . . . Bronte, V. (2010). Hierarchy of immunosuppressive strength among myeloid- derived suppressor cell subsets is determined by GM-CSF. Eur J Immunol, 40(1), 22-35. doi: 10.1002/eji.200939903

Edelman, A. M., Mitchelhill, K. I., Selbert, M. A., Anderson, K. A., Hook, S. S., Stapleton, D., . . . Kemp, B. E. (1996). Multiple Ca(2+)-calmodulin-dependent protein kinase kinases from rat brain. Purification, regulation by Ca(2+)-calmodulin, and partial amino acid sequence. J Biol Chem, 271(18), 10806-10810.

Evans, J. M., Donnelly, L. A., Emslie-Smith, A. M., Alessi, D. R., & Morris, A. D. (2005). Metformin and reduced risk of cancer in diabetic patients. BMJ, 330(7503), 1304- 1305. doi: 10.1136/bmj.38415.708634.F7

Faubert, B., Boily, G., Izreig, S., Griss, T., Samborska, B., Dong, Z., . . . Jones, R. G. (2013). AMPK is a negative regulator of the Warburg effect and suppresses tumor growth in vivo. Cell Metab, 17(1), 113-124. doi: 10.1016/j.cmet.2012.12.001

Faubert, B., Vincent, E. E., Poffenberger, M. C., & Jones, R. G. (2015). The AMP-activated protein kinase (AMPK) and cancer: many faces of a metabolic regulator. Cancer Lett, 356(2 Pt A), 165-170. doi: 10.1016/j.canlet.2014.01.018

Filipazzi, P., Valenti, R., Huber, V., Pilla, L., Canese, P., Iero, M., . . . Rivoltini, L. (2007). Identification of a new subset of myeloid suppressor cells in peripheral blood of melanoma patients with modulation by a granulocyte-macrophage colony- stimulation factor-based antitumor vaccine. J Clin Oncol, 25(18), 2546-2553. doi: 10.1200/JCO.2006.08.5829

Fox, M. M., Phoenix, K. N., Kopsiaftis, S. G., & Claffey, K. P. (2013). AMP-Activated Protein Kinase alpha 2 Isoform Suppression in Primary Breast Cancer Alters

133

AMPK Growth Control and Apoptotic Signaling. Genes Cancer, 4(1-2), 3-14. doi: 10.1177/1947601913486346

Frigo, D. E., Howe, M. K., Wittmann, B. M., Brunner, A. M., Cushman, I., Wang, Q., . . . McDonnell, D. P. (2011). CaM kinase kinase beta-mediated activation of the growth regulatory kinase AMPK is required for androgen-dependent migration of prostate cancer cells. Cancer Res, 71(2), 528-537. doi: 10.1158/0008-5472.CAN- 10-2581

Fu, H., He, H. C., Han, Z. D., Wan, Y. P., Luo, H. W., Huang, Y. Q., . . . Zhong, W. D. (2015). MicroRNA-224 and its target CAMKK2 synergistically influence tumor progression and patient prognosis in prostate cancer. Tumour Biol, 36(3), 1983- 1991. doi: 10.1007/s13277-014-2805-0

Fujiwara, Y., Hiraoka, Y., Fujimoto, T., Kanayama, N., Magari, M., & Tokumitsu, H. (2015). Analysis of Distinct Roles of CaMKK Isoforms Using STO-609-Resistant Mutants in Living Cells. Biochemistry, 54(25), 3969-3977. doi: 10.1021/acs.biochem.5b00149

Fukao, T., Frucht, D. M., Yap, G., Gadina, M., O'Shea, J. J., & Koyasu, S. (2001). Inducible expression of Stat4 in dendritic cells and macrophages and its critical role in innate and adaptive immune responses. J Immunol, 166(7), 4446-4455.

Gabrilovich, D. I., Bronte, V., Chen, S. H., Colombo, M. P., Ochoa, A., Ostrand- Rosenberg, S., & Schreiber, H. (2007). The terminology issue for myeloid-derived suppressor cells. Cancer Res, 67(1), 425; author reply 426. doi: 10.1158/0008- 5472.CAN-06-3037

Gabrilovich, D. I., Chen, H. L., Girgis, K. R., Cunningham, H. T., Meny, G. M., Nadaf, S., . . . Carbone, D. P. (1996). Production of vascular endothelial growth factor by human tumors inhibits the functional maturation of dendritic cells. Nat Med, 2(10), 1096-1103.

Gabrilovich, D. I., & Nagaraj, S. (2009). Myeloid-derived suppressor cells as regulators of the immune system. Nat Rev Immunol, 9(3), 162-174. doi: 10.1038/nri2506

134

Gabrilovich, D. I., Ostrand-Rosenberg, S., & Bronte, V. (2012). Coordinated regulation of myeloid cells by tumours. Nat Rev Immunol, 12(4), 253-268. doi: 10.1038/nri3175

Gajewski, T. F., Schreiber, H., & Fu, Y. X. (2013). Innate and adaptive immune cells in the tumor microenvironment. Nat Immunol, 14(10), 1014-1022. doi: 10.1038/ni.2703

Galic, S., Fullerton, M. D., Schertzer, J. D., Sikkema, S., Marcinko, K., Walkley, C. R., . . . Steinberg, G. R. (2011). Hematopoietic AMPK beta1 reduces mouse adipose tissue macrophage inflammation and insulin resistance in obesity. J Clin Invest, 121(12), 4903-4915. doi: 10.1172/JCI58577

Gatza, E., & Okada, C. Y. (2002). Tumor cell lysate-pulsed dendritic cells are more effective than TCR Id protein vaccines for active immunotherapy of T cell lymphoma. J Immunol, 169(9), 5227-5235.

Goldberg, M. V., Maris, C. H., Hipkiss, E. L., Flies, A. S., Zhen, L., Tuder, R. M., . . . Drake, C. G. (2007). Role of PD-1 and its ligand, B7-H1, in early fate decisions of CD8 T cells. Blood, 110(1), 186-192. doi: 10.1182/blood-2006-12-062422

Gollob, J. A., Mier, J. W., Veenstra, K., McDermott, D. F., Clancy, D., Clancy, M., & Atkins, M. B. (2000). Phase I trial of twice-weekly intravenous interleukin 12 in patients with metastatic renal cell cancer or malignant melanoma: ability to maintain IFN-gamma induction is associated with clinical response. Clin Cancer Res, 6(5), 1678-1692.

Gong, S., Zheng, C., Doughty, M. L., Losos, K., Didkovsky, N., Schambra, U. B., . . . Heintz, N. (2003). A gene expression atlas of the central nervous system based on bacterial artificial chromosomes. Nature, 425(6961), 917-925. doi: 10.1038/nature02033

Green, M. F., Anderson, K. A., & Means, A. R. (2011). Characterization of the CaMKKbeta-AMPK signaling complex. Cell Signal, 23(12), 2005-2012. doi: 10.1016/j.cellsig.2011.07.014

135

Hackstein, H., Taner, T., Zahorchak, A. F., Morelli, A. E., Logar, A. J., Gessner, A., & Thomson, A. W. (2003). Rapamycin inhibits IL-4--induced dendritic cell maturation in vitro and dendritic cell mobilization and function in vivo. Blood, 101(11), 4457-4463. doi: 10.1182/blood-2002-11-3370

Hamada, T., Souda, M., Yoshimura, T., Sasaguri, S., Hatanaka, K., Tasaki, T., . . . Tanimoto, A. (2014). Anti-apoptotic effects of PCP4/PEP19 in human breast cancer cell lines: a novel oncotarget. Oncotarget, 5(15), 6076-6086. doi: 10.18632/oncotarget.2161

Hammami, I., Chen, J., Murschel, F., Bronte, V., De Crescenzo, G., & Jolicoeur, M. (2012). Immunosuppressive activity enhances central carbon metabolism and bioenergetics in myeloid-derived suppressor cells in vitro models. BMC Cell Biol, 13, 18. doi: 10.1186/1471-2121-13-18

Hardie, D. G. (2015). Molecular Pathways: Is AMPK a Friend or a Foe in Cancer? Clin Cancer Res, 21(17), 3836-3840. doi: 10.1158/1078-0432.CCR-14-3300

Hardie, D. G., Ross, F. A., & Hawley, S. A. (2012). AMPK: a nutrient and energy sensor that maintains energy homeostasis. Nat Rev Mol Cell Biol, 13(4), 251-262. doi: 10.1038/nrm3311

Hawley, S. A., Pan, D. A., Mustard, K. J., Ross, L., Bain, J., Edelman, A. M., . . . Hardie, D. G. (2005). Calmodulin-dependent protein kinase kinase-beta is an alternative upstream kinase for AMP-activated protein kinase. Cell Metab, 2(1), 9-19. doi: 10.1016/j.cmet.2005.05.009

Hawley, S. A., Ross, F. A., Chevtzoff, C., Green, K. A., Evans, A., Fogarty, S., . . . Hardie, D. G. (2010). Use of cells expressing gamma subunit variants to identify diverse mechanisms of AMPK activation. Cell Metab, 11(6), 554-565. doi: 10.1016/j.cmet.2010.04.001

Herber, D. L., Cao, W., Nefedova, Y., Novitskiy, S. V., Nagaraj, S., Tyurin, V. A., . . . Gabrilovich, D. I. (2010). Lipid accumulation and dendritic cell dysfunction in cancer. Nat Med, 16(8), 880-886. doi: 10.1038/nm.2172

136

Hoechst, B., Ormandy, L. A., Ballmaier, M., Lehner, F., Kruger, C., Manns, M. P., . . . Korangy, F. (2008). A new population of myeloid-derived suppressor cells in hepatocellular carcinoma patients induces CD4(+)CD25(+)Foxp3(+) T cells. Gastroenterology, 135(1), 234-243. doi: 10.1053/j.gastro.2008.03.020

Hook, S. S., & Means, A. R. (2001). Ca(2+)/CaM-dependent kinases: from activation to function. Annu Rev Pharmacol Toxicol, 41, 471-505. doi: 10.1146/annurev.pharmtox.41.1.471

Hsieh, C. S., Macatonia, S. E., Tripp, C. S., Wolf, S. F., O'Garra, A., & Murphy, K. M. (1993). Development of TH1 CD4+ T cells through IL-12 produced by Listeria- induced macrophages. Science, 260(5107), 547-549.

Hsu, F. J., Benike, C., Fagnoni, F., Liles, T. M., Czerwinski, D., Taidi, B., . . . Levy, R. (1996). Vaccination of patients with B-cell lymphoma using autologous antigen- pulsed dendritic cells. Nat Med, 2(1), 52-58.

Hsu, F. J., & Komarovskaya, M. (2002). CTLA4 blockade maximizes antitumor T-cell activation by dendritic cells presenting idiotype protein or opsonized anti-CD20 antibody-coated lymphoma cells. J Immunother, 25(6), 455-468.

Hsu, L. S., Chen, G. D., Lee, L. S., Chi, C. W., Cheng, J. F., & Chen, J. Y. (2001). Human Ca2+/calmodulin-dependent protein kinase kinase beta gene encodes multiple isoforms that display distinct kinase activity. J Biol Chem, 276(33), 31113-31123. doi: 10.1074/jbc.M011720200

Hudmon, A., & Schulman, H. (2002). Structure-function of the multifunctional Ca2+/calmodulin-dependent protein kinase II. Biochem J, 364(Pt 3), 593-611. doi: 10.1042/BJ20020228

Hurley, R. L., Anderson, K. A., Franzone, J. M., Kemp, B. E., Means, A. R., & Witters, L. A. (2005). The Ca2+/calmodulin-dependent protein kinase kinases are AMP- activated protein kinase kinases. J Biol Chem, 280(32), 29060-29066. doi: 10.1074/jbc.M503824200

137

Husain, Z., Huang, Y., Seth, P., & Sukhatme, V. P. (2013). Tumor-derived lactate modifies antitumor immune response: effect on myeloid-derived suppressor cells and NK cells. J Immunol, 191(3), 1486-1495. doi: 10.4049/jimmunol.1202702

Illario, M., Giardino-Torchia, M. L., Sankar, U., Ribar, T. J., Galgani, M., Vitiello, L., . . . Racioppi, L. (2008). Calmodulin-dependent kinase IV links Toll-like receptor 4 signaling with survival pathway of activated dendritic cells. Blood, 111(2), 723- 731. doi: 10.1182/blood-2007-05-091173

Inoue, K., Shiga, T., & Ito, Y. (2008). Runx transcription factors in neuronal development. Neural Dev, 3, 20. doi: 10.1186/1749-8104-3-20

Isaacs, J. T., & Isaacs, W. B. (2004). Androgen receptor outwits prostate cancer drugs. Nat Med, 10(1), 26-27. doi: 10.1038/nm0104-26

Ishida, Y., Agata, Y., Shibahara, K., & Honjo, T. (1992). Induced expression of PD-1, a novel member of the immunoglobulin gene superfamily, upon programmed cell death. EMBO J, 11(11), 3887-3895.

Ishikawa, Y., Tokumitsu, H., Inuzuka, H., Murata-Hori, M., Hosoya, H., & Kobayashi, R. (2003). Identification and characterization of novel components of a Ca2+/calmodulin-dependent protein kinase cascade in HeLa cells. FEBS Lett, 550(1-3), 57-63.

Ito, T., Inaba, M., Inaba, K., Toki, J., Sogo, S., Iguchi, T., . . . Ikehara, S. (1999). A CD1a+/CD11c+ subset of human blood dendritic cells is a direct precursor of Langerhans cells. J Immunol, 163(3), 1409-1419.

James L. Rubenstein, N. K. G., Gabriel N. Mannis, Amanda K. LaMarre, Patrick Treseler. (2013). How I treat CNS lymphomas. Blood, 122(14), 2318-2330. doi: 10.1182/blood-2013-06-453084

Janeway CA Jr, T. P., Walport M, et al. . (2001). T cell-mediated cytotoxicity. Immunobiology: The Immune System in Health and Disease. (5th edition ed.). New York: Garland Science.

138

Jeong, H. W., Hsu, K. C., Lee, J. W., Ham, M., Huh, J. Y., Shin, H. J., . . . Kim, J. B. (2009). Berberine suppresses proinflammatory responses through AMPK activation in macrophages. Am J Physiol Endocrinol Metab, 296(4), E955-964. doi: 10.1152/ajpendo.90599.2008

Kadowaki, N., Antonenko, S., & Liu, Y. J. (2001). Distinct CpG DNA and polyinosinic- polycytidylic acid double-stranded RNA, respectively, stimulate CD11c- type 2 dendritic cell precursors and CD11c+ dendritic cells to produce type I IFN. J Immunol, 166(4), 2291-2295.

Kahl, C. R., & Means, A. R. (2004). Regulation of cyclin D1/Cdk4 complexes by calcium/calmodulin-dependent protein kinase I. J Biol Chem, 279(15), 15411- 15419. doi: 10.1074/jbc.M312543200

Karacosta, L. G., Foster, B. A., Azabdaftari, G., Feliciano, D. M., & Edelman, A. M. (2012). A regulatory feedback loop between Ca2+/calmodulin-dependent protein kinase kinase 2 (CaMKK2) and the androgen receptor in prostate cancer progression. J Biol Chem, 287(29), 24832-24843. doi: 10.1074/jbc.M112.370783

Karacosta, L. G., Kuroski, L. A., Hofmann, W. A., Azabdaftari, G., Mastri, M., Gocher, A. M., . . . Edelman, A. M. (2016). Nucleoporin 62 and Ca(2+) /calmodulin dependent kinase kinase 2 regulate androgen receptor activity in castrate resistant prostate cancer cells. Prostate, 76(3), 294-306. doi: 10.1002/pros.23121

Kitsos, C. M., Sankar, U., Illario, M., Colomer-Font, J. M., Duncan, A. W., Ribar, T. J., . . . Means, A. R. (2005). Calmodulin-dependent protein kinase IV regulates hematopoietic stem cell maintenance. J Biol Chem, 280(39), 33101-33108. doi: 10.1074/jbc.M505208200

Ko, J. S., Zea, A. H., Rini, B. I., Ireland, J. L., Elson, P., Cohen, P., . . . Finke, J. H. (2009). Sunitinib mediates reversal of myeloid-derived suppressor cell accumulation in renal cell carcinoma patients. Clin Cancer Res, 15(6), 2148-2157. doi: 10.1158/1078- 0432.CCR-08-1332

Koch, F., Stanzl, U., Jennewein, P., Janke, K., Heufler, C., Kampgen, E., . . . Schuler, G. (1996). High level IL-12 production by murine dendritic cells: upregulation via 139

MHC class II and CD40 molecules and downregulation by IL-4 and IL-10. J Exp Med, 184(2), 741-746.

Kokubo, M., Nishio, M., Ribar, T. J., Anderson, K. A., West, A. E., & Means, A. R. (2009). BDNF-mediated cerebellar granule cell development is impaired in mice null for CaMKK2 or CaMKIV. J Neurosci, 29(28), 8901-8913. doi: 10.1523/JNEUROSCI.0040-09.2009

Krawczyk, C. M., Holowka, T., Sun, J., Blagih, J., Amiel, E., DeBerardinis, R. J., . . . Pearce, E. J. (2010). Toll-like receptor-induced changes in glycolytic metabolism regulate dendritic cell activation. Blood, 115(23), 4742-4749. doi: 10.1182/blood- 2009-10-249540

Kusmartsev, S., & Gabrilovich, D. I. (2002). Immature myeloid cells and cancer- associated immune suppression. Cancer Immunol Immunother, 51(6), 293-298. doi: 10.1007/s00262-002-0280-8

Kusmartsev, S., Su, Z., Heiser, A., Dannull, J., Eruslanov, E., Kubler, H., . . . Vieweg, J. (2008). Reversal of myeloid cell-mediated immunosuppression in patients with metastatic renal cell carcinoma. Clin Cancer Res, 14(24), 8270-8278. doi: 10.1158/1078-0432.CCR-08-0165 la Sala, A., Ferrari, D., Corinti, S., Cavani, A., Di Virgilio, F., & Girolomoni, G. (2001). Extracellular ATP induces a distorted maturation of dendritic cells and inhibits their capacity to initiate Th1 responses. J Immunol, 166(3), 1611-1617.

Lanzavecchia, A., & Sallusto, F. (2000). Dynamics of T lymphocyte responses: intermediates, effectors, and memory cells. Science, 290(5489), 92-97.

Lasek, W., Zagozdzon, R., & Jakobisiak, M. (2014). Interleukin 12: still a promising candidate for tumor immunotherapy? Cancer Immunol Immunother, 63(5), 419-435. doi: 10.1007/s00262-014-1523-1

Liang, J., & Mills, G. B. (2013). AMPK: a contextual oncogene or tumor suppressor? Cancer Res, 73(10), 2929-2935. doi: 10.1158/0008-5472.CAN-12-3876

140

Liedtke, C., Mazouni, C., Hess, K. R., Andre, F., Tordai, A., Mejia, J. A., . . . Pusztai, L. (2008). Response to neoadjuvant therapy and long-term survival in patients with triple-negative breast cancer. J Clin Oncol, 26(8), 1275-1281. doi: 10.1200/JCO.2007.14.4147

Lin, F., Marcelo, K. L., Rajapakshe, K., Coarfa, C., Dean, A., Wilganowski, N., . . . York, B. (2015). The camKK2/camKIV relay is an essential regulator of hepatic cancer. Hepatology, 62(2), 505-520. doi: 10.1002/hep.27832

Lin, F., Ribar, T. J., & Means, A. R. (2011). The Ca2+/calmodulin-dependent protein kinase kinase, CaMKK2, inhibits preadipocyte differentiation. Endocrinology, 152(10), 3668-3679. doi: 10.1210/en.2011-1107

Liu, B. L., Cheng, J. X., Zhang, X., Zhang, W., & Cheng, H. (2009). Limited role of surgery in the management of primary central nervous system lymphoma (Review). Oncol Rep, 22(3), 439-449.

Liu, G., Bi, Y., Shen, B., Yang, H., Zhang, Y., Wang, X., . . . Chu, Y. (2014). SIRT1 limits the function and fate of myeloid-derived suppressor cells in tumors by orchestrating HIF-1alpha-dependent glycolysis. Cancer Res, 74(3), 727-737. doi: 10.1158/0008-5472.CAN-13-2584

Liu, M. Y., Xu, S. M., & Song, G. H. (2001). [Dendritic cells and protozoan infection]. Zhongguo Ji Sheng Chong Xue Yu Ji Sheng Chong Bing Za Zhi, 19(6), 367-370.

Liu, Y. J. (2001). Dendritic cell subsets and lineages, and their functions in innate and adaptive immunity. Cell, 106(3), 259-262.

Liu, Y. J., Kanzler, H., Soumelis, V., & Gilliet, M. (2001). Dendritic cell lineage, plasticity and cross-regulation. Nat Immunol, 2(7), 585-589. doi: 10.1038/89726

Loi, S., Michiels, S., Salgado, R., Sirtaine, N., Jose, V., Fumagalli, D., . . . Sotiriou, C. (2014). Tumor infiltrating lymphocytes are prognostic in triple negative breast cancer and predictive for trastuzumab benefit in early breast cancer: results from the FinHER trial. Ann Oncol, 25(8), 1544-1550. doi: 10.1093/annonc/mdu112

141

Luo, J. L., Tan, W., Ricono, J. M., Korchynskyi, O., Zhang, M., Gonias, S. L., . . . Karin, M. (2007). Nuclear cytokine-activated IKKalpha controls prostate cancer metastasis by repressing Maspin. Nature, 446(7136), 690-694. doi: 10.1038/nature05656

Macatonia, S. E., Hosken, N. A., Litton, M., Vieira, P., Hsieh, C. S., Culpepper, J. A., . . . O'Garra, A. (1995). Dendritic cells produce IL-12 and direct the development of Th1 cells from naive CD4+ T cells. J Immunol, 154(10), 5071-5079.

Maier, T., Tun-Kyi, A., Tassis, A., Jungius, K. P., Burg, G., Dummer, R., & Nestle, F. O. (2003). Vaccination of patients with cutaneous T-cell lymphoma using intranodal injection of autologous tumor-lysate-pulsed dendritic cells. Blood, 102(7), 2338- 2344. doi: 10.1182/blood-2002-08-2455

Manetti, R., Parronchi, P., Giudizi, M. G., Piccinni, M. P., Maggi, E., Trinchieri, G., & Romagnani, S. (1993). Natural killer cell stimulatory factor (interleukin 12 [IL-12]) induces T helper type 1 (Th1)-specific immune responses and inhibits the development of IL-4-producing Th cells. J Exp Med, 177(4), 1199-1204.

Martin-Martin, L., Lopez, A., Vidriales, B., Caballero, M. D., Rodrigues, A. S., Ferreira, S. I., . . . Almeida, J. (2015). Classification and clinical behavior of blastic plasmacytoid dendritic cell neoplasms according to their maturation-associated immunophenotypic profile. Oncotarget, 6(22), 19204-19216. doi: 10.18632/oncotarget.4146

Marvel, D., & Gabrilovich, D. I. (2015). Myeloid-derived suppressor cells in the tumor microenvironment: expect the unexpected. J Clin Invest, 125(9), 3356-3364. doi: 10.1172/JCI80005

Massie, C. E., Lynch, A., Ramos-Montoya, A., Boren, J., Stark, R., Fazli, L., . . . Mills, I. G. (2011). The androgen receptor fuels prostate cancer by regulating central metabolism and biosynthesis. EMBO J, 30(13), 2719-2733. doi: 10.1038/emboj.2011.158

McKenna, H. J., Stocking, K. L., Miller, R. E., Brasel, K., De Smedt, T., Maraskovsky, E., . . . Peschon, J. J. (2000). Mice lacking flt3 ligand have deficient hematopoiesis

142

affecting hematopoietic progenitor cells, dendritic cells, and natural killer cells. Blood, 95(11), 3489-3497.

Means, A. R., Ribar, T. J., Kane, C. D., Hook, S. S., & Anderson, K. A. (1997). Regulation and properties of the rat Ca2+/calmodulin-dependent protein kinase IV gene and its protein products. Recent Prog Horm Res, 52, 389-406; discussion 406-387.

Melero, I., Duarte, M., Ruiz, J., Sangro, B., Galofre, J., Mazzolini, G., . . . Prieto, J. (1999). Intratumoral injection of bone-marrow derived dendritic cells engineered to produce interleukin-12 induces complete regression of established murine transplantable colon adenocarcinomas. Gene Ther, 6(10), 1779-1784. doi: 10.1038/sj.gt.3301010

Mihaylova, M. M., & Shaw, R. J. (2011). The AMPK signalling pathway coordinates cell growth, autophagy and metabolism. Nat Cell Biol, 13(9), 1016-1023. doi: 10.1038/ncb2329

Montero, A. J., Diaz-Montero, C. M., Deutsch, Y. E., Hurley, J., Koniaris, L. G., Rumboldt, T., . . . Gluck, S. (2012). Phase 2 study of neoadjuvant treatment with NOV-002 in combination with doxorubicin and cyclophosphamide followed by docetaxel in patients with HER-2 negative clinical stage II-IIIc breast cancer. Breast Cancer Res Treat, 132(1), 215-223. doi: 10.1007/s10549-011-1889-0

Mortarini, R., Borri, A., Tragni, G., Bersani, I., Vegetti, C., Bajetta, E., . . . Anichini, A. (2000). Peripheral burst of tumor-specific cytotoxic T lymphocytes and infiltration of metastatic lesions by memory CD8+ T cells in melanoma patients receiving interleukin 12. Cancer Res, 60(13), 3559-3568.

Morton, G. J., Cummings, D. E., Baskin, D. G., Barsh, G. S., & Schwartz, M. W. (2006). Central nervous system control of food intake and body weight. Nature, 443(7109), 289-295. doi: 10.1038/nature05026

Movahedi, K., Guilliams, M., Van den Bossche, J., Van den Bergh, R., Gysemans, C., Beschin, A., . . . Van Ginderachter, J. A. (2008). Identification of discrete tumor- induced myeloid-derived suppressor cell subpopulations with distinct T cell- suppressive activity. Blood, 111(8), 4233-4244. doi: 10.1182/blood-2007-07-099226 143

Mullen, A. C., High, F. A., Hutchins, A. S., Lee, H. W., Villarino, A. V., Livingston, D. M., . . . Reiner, S. L. (2001). Role of T-bet in commitment of TH1 cells before IL-12- dependent selection. Science, 292(5523), 1907-1910. doi: 10.1126/science.1059835

Multani, P., White, C. A., & Grillo-Lopez, A. (2001). Non-Hodgkin's lymphoma: review of conventional treatments. Curr Pharm Biotechnol, 2(4), 279-291.

Najjar, Y. G., & Finke, J. H. (2013). Clinical perspectives on targeting of myeloid derived suppressor cells in the treatment of cancer. Front Oncol, 3, 49. doi: 10.3389/fonc.2013.00049

Nefedova, Y., Fishman, M., Sherman, S., Wang, X., Beg, A. A., & Gabrilovich, D. I. (2007). Mechanism of all-trans retinoic acid effect on tumor-associated myeloid- derived suppressor cells. Cancer Res, 67(22), 11021-11028. doi: 10.1158/0008- 5472.CAN-07-2593

Nefedova, Y., Huang, M., Kusmartsev, S., Bhattacharya, R., Cheng, P., Salup, R., . . . Gabrilovich, D. (2004). Hyperactivation of STAT3 is involved in abnormal differentiation of dendritic cells in cancer. J Immunol, 172(1), 464-474.

Nelson, P. S., Clegg, N., Arnold, H., Ferguson, C., Bonham, M., White, J., . . . Lin, B. (2002). The program of androgen-responsive genes in neoplastic prostate epithelium. Proc Natl Acad Sci U S A, 99(18), 11890-11895. doi: 10.1073/pnas.182376299

Nonomura, N., Takayama, H., Nakayama, M., Nakai, Y., Kawashima, A., Mukai, M., . . . Tsujimura, A. (2011). Infiltration of tumour-associated macrophages in prostate biopsy specimens is predictive of disease progression after hormonal therapy for prostate cancer. BJU Int, 107(12), 1918-1922. doi: 10.1111/j.1464-410X.2010.09804.x

O'Neill, H. M., Holloway, G. P., & Steinberg, G. R. (2013). AMPK regulation of fatty acid metabolism and mitochondrial biogenesis: implications for obesity. Mol Cell Endocrinol, 366(2), 135-151. doi: 10.1016/j.mce.2012.06.019

144

Palucka, K., & Banchereau, J. (2013). Dendritic-cell-based therapeutic cancer vaccines. Immunity, 39(1), 38-48. doi: 10.1016/j.immuni.2013.07.004

Pearce, E. J., & Everts, B. (2015). Dendritic cell metabolism. Nat Rev Immunol, 15(1), 18-29. doi: 10.1038/nri3771

Peters, M., Mizuno, K., Ris, L., Angelo, M., Godaux, E., & Giese, K. P. (2003). Loss of Ca2+/calmodulin kinase kinase beta affects the formation of some, but not all, types of hippocampus-dependent long-term memory. J Neurosci, 23(30), 9752- 9760.

Picciotto, M. R., Czernik, A. J., & Nairn, A. C. (1993). Calcium/calmodulin-dependent protein kinase I. cDNA cloning and identification of autophosphorylation site. J Biol Chem, 268(35), 26512-26521.

Portielje, J. E., Kruit, W. H., Schuler, M., Beck, J., Lamers, C. H., Stoter, G., . . . Aulitzky, W. E. (1999). Phase I study of subcutaneously administered recombinant human interleukin 12 in patients with advanced renal cell cancer. Clin Cancer Res, 5(12), 3983-3989.

Pritchard, Z. J., Cary, R. L., Yang, C., Novack, D. V., Voor, M. J., & Sankar, U. (2015). Inhibition of CaMKK2 reverses age-associated decline in bone mass. Bone, 75, 120-127. doi: 10.1016/j.bone.2015.01.021

Racioppi, L. (2013). CaMKK2: a novel target for shaping the androgen-regulated tumor ecosystem. Trends Mol Med, 19(2), 83-88. doi: 10.1016/j.molmed.2012.12.004

Racioppi, L., & Means, A. R. (2012). Calcium/calmodulin-dependent protein kinase kinase 2: roles in signaling and pathophysiology. J Biol Chem, 287(38), 31658- 31665. doi: 10.1074/jbc.R112.356485

Racioppi, L., Noeldner, P. K., Lin, F., Arvai, S., & Means, A. R. (2012). Calcium/calmodulin-dependent protein kinase kinase 2 regulates macrophage- mediated inflammatory responses. J Biol Chem, 287(14), 11579-11591. doi: 10.1074/jbc.M111.336032

145

Rehman, A., Hemmert, K. C., Ochi, A., Jamal, M., Henning, J. R., Barilla, R., . . . Miller, G. (2013). Role of fatty-acid synthesis in dendritic cell generation and function. J Immunol, 190(9), 4640-4649. doi: 10.4049/jimmunol.1202312

Ribar, T. J., Rodriguiz, R. M., Khiroug, L., Wetsel, W. C., Augustine, G. J., & Means, A. R. (2000). Cerebellar defects in Ca2+/calmodulin kinase IV-deficient mice. J Neurosci, 20(22), RC107.

Ribechini, E., Greifenberg, V., Sandwick, S., & Lutz, M. B. (2010). Subsets, expansion and activation of myeloid-derived suppressor cells. Med Microbiol Immunol, 199(3), 273-281. doi: 10.1007/s00430-010-0151-4

Rodriguez-Mora, O. G., LaHair, M. M., McCubrey, J. A., & Franklin, R. A. (2005). Calcium/calmodulin-dependent kinase I and calcium/calmodulin-dependent kinase kinase participate in the control of cell cycle progression in MCF-7 human breast cancer cells. Cancer Res, 65(12), 5408-5416. doi: 10.1158/0008-5472.CAN-05- 0271

Romano, A., Parrinello, N. L., Vetro, C., Forte, S., Chiarenza, A., Figuera, A., . . . Di Raimondo, F. (2015). Circulating myeloid-derived suppressor cells correlate with clinical outcome in Hodgkin Lymphoma patients treated up-front with a risk- adapted strategy. Br J Haematol, 168(5), 689-700. doi: 10.1111/bjh.13198

Sag, D., Carling, D., Stout, R. D., & Suttles, J. (2008). Adenosine 5'-monophosphate- activated protein kinase promotes macrophage polarization to an anti- inflammatory functional phenotype. J Immunol, 181(12), 8633-8641.

Sampson, J. H., Archer, G. E., Villavicencio, A. T., McLendon, R. E., Friedman, A. H., Bishop, W. R., . . . Friedman, H. S. (1999). Treatment of neoplastic meningitis with intrathecal temozolomide. Clin Cancer Res, 5(5), 1183-1188.

Saneyoshi, T., Wayman, G., Fortin, D., Davare, M., Hoshi, N., Nozaki, N., . . . Soderling, T. R. (2008). Activity-dependent synaptogenesis: regulation by a CaM-kinase kinase/CaM-kinase I/betaPIX signaling complex. Neuron, 57(1), 94-107. doi: 10.1016/j.neuron.2007.11.016

146

Sato, M., Suzuki, K., & Nakanishi, S. (2006). Expression profile of BDNF-responsive genes during cerebellar granule cell development. Biochem Biophys Res Commun, 341(2), 304-309. doi: 10.1016/j.bbrc.2005.12.184

Sato, Y., Shimizu, K., Shinga, J., Hidaka, M., Kawano, F., Kakimi, K., . . . Fujii, S. I. (2015). Characterization of the myeloid-derived suppressor cell subset regulated by NK cells in malignant lymphoma. Oncoimmunology, 4(3), e995541. doi: 10.1080/2162402X.2014.995541

Schlapbach, C., Ochsenbein, A., Kaelin, U., Hassan, A. S., Hunger, R. E., & Yawalkar, N. (2010). High numbers of DC-SIGN+ dendritic cells in lesional skin of cutaneous T-cell lymphoma. J Am Acad Dermatol, 62(6), 995-1004. doi: 10.1016/j.jaad.2009.06.082

Schmid, M. C., & Varner, J. A. (2010). Myeloid cells in the tumor microenvironment: modulation of tumor angiogenesis and tumor inflammation. J Oncol, 2010, 201026. doi: 10.1155/2010/201026

Schmitt, J. M., Abell, E., Wagner, A., & Davare, M. A. (2010). ERK activation and cell growth require CaM kinases in MCF-7 breast cancer cells. Mol Cell Biochem, 335(1-2), 155-171. doi: 10.1007/s11010-009-0252-9

Schmitz, N., & Wu, H. S. (2015). Advances in the Treatment of Secondary CNS Lymphoma. J Clin Oncol, 33(33), 3851-3853. doi: 10.1200/JCO.2015.63.1143

Serafini, P., Meckel, K., Kelso, M., Noonan, K., Califano, J., Koch, W., . . . Borrello, I. (2006). Phosphodiesterase-5 inhibition augments endogenous antitumor immunity by reducing myeloid-derived suppressor cell function. J Exp Med, 203(12), 2691-2702. doi: 10.1084/jem.20061104

Serafini, P., Mgebroff, S., Noonan, K., & Borrello, I. (2008). Myeloid-derived suppressor cells promote cross-tolerance in B-cell lymphoma by expanding regulatory T cells. Cancer Res, 68(13), 5439-5449. doi: 10.1158/0008-5472.CAN-07-6621

147

Shima, T., Mizokami, A., Miyagi, T., Kawai, K., Izumi, K., Kumaki, M., . . . Namiki, M. (2012). Down-regulation of calcium/calmodulin-dependent protein kinase kinase 2 by androgen deprivation induces castration-resistant prostate cancer. Prostate, 72(16), 1789-1801. doi: 10.1002/pros.22533

Shirato, K., Kizaki, T., Sakurai, T., Ogasawara, J. E., Ishibashi, Y., Iijima, T., . . . Ohno, H. (2009). Hypoxia-inducible factor-1alpha suppresses the expression of macrophage scavenger receptor 1. Pflugers Arch, 459(1), 93-103. doi: 10.1007/s00424-009-0702-y

Soderling, T. R. (1999). The Ca-calmodulin-dependent protein kinase cascade. Trends Biochem Sci, 24(6), 232-236.

Sotelo, C. (2004). Cellular and genetic regulation of the development of the cerebellar system. Prog Neurobiol, 72(5), 295-339. doi: 10.1016/j.pneurobio.2004.03.004

Steinman, R. M. (2001). Dendritic cells and the control of immunity: enhancing the efficiency of antigen presentation. Mt Sinai J Med, 68(3), 160-166.

Strunk, D., Egger, C., Leitner, G., Hanau, D., & Stingl, G. (1997). A skin homing molecule defines the langerhans cell progenitor in human peripheral blood. J Exp Med, 185(6), 1131-1136.

Subbannayya, Y., Syed, N., Barbhuiya, M. A., Raja, R., Marimuthu, A., Sahasrabuddhe, N., . . . Chatterjee, A. (2015). Calcium calmodulin dependent kinase kinase 2 - a novel therapeutic target for gastric adenocarcinoma. Cancer Biol Ther, 16(2), 336- 345. doi: 10.4161/15384047.2014.972264

Suehiro, Y., Hasegawa, A., Iino, T., Sasada, A., Watanabe, N., Matsuoka, M., . . . Okamura, J. (2015). Clinical outcomes of a novel therapeutic vaccine with Tax peptide-pulsed dendritic cells for adult T cell leukaemia/lymphoma in a pilot study. Br J Haematol, 169(3), 356-367. doi: 10.1111/bjh.13302

148

Sun, Z., Means, R. L., LeMagueresse, B., & Means, A. R. (1995). Organization and analysis of the complete rat calmodulin-dependent protein kinase IV gene. J Biol Chem, 270(49), 29507-29514.

Tadmor, T., Fell, R., Polliack, A., & Attias, D. (2013). Absolute monocytosis at diagnosis correlates with survival in diffuse large B-cell lymphoma-possible link with monocytic myeloid-derived suppressor cells. Hematol Oncol, 31(2), 65-71. doi: 10.1002/hon.2019

Talmadge, J. E., & Gabrilovich, D. I. (2013). History of myeloid-derived suppressor cells. Nat Rev Cancer, 13(10), 739-752. doi: 10.1038/nrc3581

Teng, E. C., Racioppi, L., & Means, A. R. (2011). A cell-intrinsic role for CaMKK2 in granulocyte lineage commitment and differentiation. J Leukoc Biol, 90(5), 897-909. doi: 10.1189/jlb.0311152

Tijssen, M. R., Cvejic, A., Joshi, A., Hannah, R. L., Ferreira, R., Forrai, A., . . . Gottgens, B. (2011). Genome-wide analysis of simultaneous GATA1/2, RUNX1, FLI1, and SCL binding in megakaryocytes identifies hematopoietic regulators. Dev Cell, 20(5), 597-609. doi: 10.1016/j.devcel.2011.04.008

Timmerman, J. M., Czerwinski, D. K., Davis, T. A., Hsu, F. J., Benike, C., Hao, Z. M., . . . Levy, R. (2002). Idiotype-pulsed dendritic cell vaccination for B-cell lymphoma: clinical and immune responses in 35 patients. Blood, 99(5), 1517-1526.

Toes, R. E., Ossendorp, F., Offringa, R., & Melief, C. J. (1999). CD4 T cells and their role in antitumor immune responses. J Exp Med, 189(5), 753-756.

Tokumitsu, H., Enslen, H., & Soderling, T. R. (1995). Characterization of a Ca2+/calmodulin-dependent protein kinase cascade. Molecular cloning and expression of calcium/calmodulin-dependent protein kinase kinase. J Biol Chem, 270(33), 19320-19324.

149

Tokumitsu, H., Inuzuka, H., Ishikawa, Y., Ikeda, M., Saji, I., & Kobayashi, R. (2002). STO- 609, a specific inhibitor of the Ca(2+)/calmodulin-dependent protein kinase kinase. J Biol Chem, 277(18), 15813-15818. doi: 10.1074/jbc.M201075200

Tokumitsu, H., Inuzuka, H., Ishikawa, Y., & Kobayashi, R. (2003). A single amino acid difference between alpha and beta Ca2+/calmodulin-dependent protein kinase kinase dictates sensitivity to the specific inhibitor, STO-609. J Biol Chem, 278(13), 10908-10913. doi: 10.1074/jbc.M213183200

Tokumitsu, H., Takahashi, N., Eto, K., Yano, S., Soderling, T. R., & Muramatsu, M. (1999). Substrate recognition by Ca2+/Calmodulin-dependent protein kinase kinase. Role of the arg-pro-rich insert domain. J Biol Chem, 274(22), 15803-15810.

Tugues, S., Burkhard, S. H., Ohs, I., Vrohlings, M., Nussbaum, K., Vom Berg, J., . . . Becher, B. (2015). New insights into IL-12-mediated tumor suppression. Cell Death Differ, 22(2), 237-246. doi: 10.1038/cdd.2014.134

Turnquist, H. R., Raimondi, G., Zahorchak, A. F., Fischer, R. T., Wang, Z., & Thomson, A. W. (2007). Rapamycin-conditioned dendritic cells are poor stimulators of allogeneic CD4+ T cells, but enrich for antigen-specific Foxp3+ T regulatory cells and promote organ transplant tolerance. J Immunol, 178(11), 7018-7031.

Vander Heiden, M. G., Cantley, L. C., & Thompson, C. B. (2009). Understanding the Warburg effect: the metabolic requirements of cell proliferation. Science, 324(5930), 1029-1033. doi: 10.1126/science.1160809

Veltman, J. D., Lambers, M. E., van Nimwegen, M., Hendriks, R. W., Hoogsteden, H. C., Aerts, J. G., & Hegmans, J. P. (2010). COX-2 inhibition improves immunotherapy and is associated with decreased numbers of myeloid-derived suppressor cells in mesothelioma. Celecoxib influences MDSC function. BMC Cancer, 10, 464. doi: 10.1186/1471-2407-10-464

Villasboas, J. C., & Ansell, S. (2016). Checkpoint Inhibition: Programmed Cell Death 1 and Programmed Cell Death 1 Ligand Inhibitors in Hodgkin Lymphoma. Cancer J, 22(1), 17-22. doi: 10.1097/PPO.0000000000000164

150

Wang, S. L., Ribar, T. J., & Means, A. R. (2001). Expression of Ca(2+)/calmodulin- dependent protein kinase IV (caMKIV) messenger RNA during murine embryogenesis. Cell Growth Differ, 12(7), 351-361.

Wang, V. Y., & Zoghbi, H. Y. (2001). Genetic regulation of cerebellar development. Nat Rev Neurosci, 2(7), 484-491. doi: 10.1038/35081558

Wayman, G. A., Kaech, S., Grant, W. F., Davare, M., Impey, S., Tokumitsu, H., . . . Soderling, T. R. (2004). Regulation of axonal extension and growth cone motility by calmodulin-dependent protein kinase I. J Neurosci, 24(15), 3786-3794. doi: 10.1523/JNEUROSCI.3294-03.2004

Wayman, G. A., Lee, Y. S., Tokumitsu, H., Silva, A. J., & Soderling, T. R. (2008). Calmodulin-kinases: modulators of neuronal development and plasticity. Neuron, 59(6), 914-931. doi: 10.1016/j.neuron.2008.08.021

Westin, J. R., Chu, F., Zhang, M., Fayad, L. E., Kwak, L. W., Fowler, N., . . . Neelapu, S. S. (2014). Safety and activity of PD1 blockade by pidilizumab in combination with rituximab in patients with relapsed follicular lymphoma: a single group, open- label, phase 2 trial. Lancet Oncol, 15(1), 69-77. doi: 10.1016/S1470-2045(13)70551-5

Wu, C., Wu, X., Zhang, X., Chai, Y., Guo, Q., Li, L., . . . Zhang, L. (2015). Prognostic significance of peripheral monocytic myeloid-derived suppressor cells and monocytes in patients newly diagnosed with diffuse large b-cell lymphoma. Int J Clin Exp Med, 8(9), 15173-15181.

Wu, L., Wang, G., Qu, P., Yan, C., & Du, H. (2011). Overexpression of dominant negative peroxisome proliferator-activated receptor-gamma (PPARgamma) in alveolar type II epithelial cells causes inflammation and T-cell suppression in the lung. Am J Pathol, 178(5), 2191-2204. doi: 10.1016/j.ajpath.2011.01.046

Wu, T., Zhao, Y., Wang, H., Li, Y., Shao, L., Wang, R., . . . Zhao, Y. (2016). mTOR masters monocytic myeloid-derived suppressor cells in mice with allografts or tumors. Sci Rep, 6, 20250. doi: 10.1038/srep20250

151

Xiao, B., Sanders, M. J., Underwood, E., Heath, R., Mayer, F. V., Carmena, D., . . . Gamblin, S. J. (2011). Structure of mammalian AMPK and its regulation by ADP. Nature, 472(7342), 230-233. doi: 10.1038/nature09932

Xu, M. L., & Fedoriw, Y. (2016). Lymphoma Microenvironment and Immunotherapy. Surg Pathol Clin, 9(1), 93-100. doi: 10.1016/j.path.2015.10.001

Yang, Z., Kahn, B. B., Shi, H., & Xue, B. Z. (2010). Macrophage alpha1 AMP-activated protein kinase (alpha1AMPK) antagonizes fatty acid-induced inflammation through SIRT1. J Biol Chem, 285(25), 19051-19059. doi: 10.1074/jbc.M110.123620

Yang, Z., Wang, X., He, Y., Qi, L., Yu, L., Xue, B., & Shi, H. (2012). The full capacity of AICAR to reduce obesity-induced inflammation and insulin resistance requires myeloid SIRT1. PLoS One, 7(11), e49935. doi: 10.1371/journal.pone.0049935

Youn, J. I., Nagaraj, S., Collazo, M., & Gabrilovich, D. I. (2008). Subsets of myeloid- derived suppressor cells in tumor-bearing mice. J Immunol, 181(8), 5791-5802.

Zagami, C. J., Zusso, M., & Stifani, S. (2009). Runx transcription factors: lineage-specific regulators of neuronal precursor cell proliferation and post-mitotic neuron subtype development. J Cell Biochem, 107(6), 1063-1072. doi: 10.1002/jcb.22221

Zhang, B., Wang, Z., Wu, L., Zhang, M., Li, W., Ding, J., . . . Zhao, K. (2013). Circulating and tumor-infiltrating myeloid-derived suppressor cells in patients with colorectal carcinoma. PLoS One, 8(2), e57114. doi: 10.1371/journal.pone.0057114

Zhang, H., Li, Z. L., Ye, S. B., Ouyang, L. Y., Chen, Y. S., He, J., . . . Li, J. (2015). Myeloid- derived suppressor cells inhibit T cell proliferation in human extranodal NK/T cell lymphoma: a novel prognostic indicator. Cancer Immunol Immunother, 64(12), 1587-1599. doi: 10.1007/s00262-015-1765-6

Zhang, S. N., Choi, I. K., Huang, J. H., Yoo, J. Y., Choi, K. J., & Yun, C. O. (2011). Optimizing DC vaccination by combination with oncolytic adenovirus coexpressing IL-12 and GM-CSF. Mol Ther, 19(8), 1558-1568. doi: 10.1038/mt.2011.29

152

Zhang, X., Guo, L., Collage, R. D., Stripay, J. L., Tsung, A., Lee, J. S., & Rosengart, M. R. (2011). Calcium/calmodulin-dependent protein kinase (CaMK) Ialpha mediates the macrophage inflammatory response to sepsis. J Leukoc Biol, 90(2), 249-261. doi: 10.1189/jlb.0510286

Zhang, Z. J., Bulur, P. A., Dogan, A., Gastineau, D. A., Dietz, A. B., & Lin, Y. (2015). Immune independent crosstalk between lymphoma and myeloid suppressor CD14HLA-DR monocytes mediates chemotherapy resistance. Oncoimmunology, 4(4), e996470. doi: 10.1080/2162402X.2014.996470

Zhu, P., Baek, S. H., Bourk, E. M., Ohgi, K. A., Garcia-Bassets, I., Sanjo, H., . . . Rose, D. W. (2006). Macrophage/cancer cell interactions mediate hormone resistance by a nuclear receptor derepression pathway. Cell, 124(3), 615-629. doi: 10.1016/j.cell.2005.12.032

153

Biography

Wei Huang was born in Guangzhou, Guangdong, P. R. China on November 10th,

1985 to Xiao-Yi Huang and Wei-Hong Song. She grew up with her grandfather Huan-

Qiu Huang who was the president of Sun Yat-Sen University, and her grandmother Jie

Zhang. In 2004, Wei graduated from Guangzhou No.6 High School. She then attended

Peking University, majoring in Clinical Medicine. During this time, she received

Exceptional Medical Student Scholarship (2005, 2006, 2007), Award for Outstanding

Academic Performance (2008, 2009), and graduated as the Top-1 student of her class with the National Scholarship for Outstanding Graduate. After receiving her medical degree in 2009, Wei went on to study prostate cancer in Dr. Shijie Sheng’s Lab at Wayne

State University in Detroit, MI. In 2010, she was recruited to the doctoral training program in Pathology at Duke University in Durham, NC, where she pursued her doctoral degree under the mentorship of Dr. Nelson Chao and focus on cancer immunology and T-cell immunology. Her doctoral thesis is entitled

“Calcium/Calmodulin-Dependent Protein Kinase Kinase 2 (CaMKK2) Regulates

Dendritic Cells and Myeloid Derived Suppressor Cells Development in the Lymphoma

Microenvironment”. Her work contributed to another two papers in preparation.

Publication:

Lento, W., T. Ito, C. Zhao, J. R. Harris, W. Huang, C. Jiang, K. Owzar, S. Piryani, L. Racioppi, N. Chao, T. Reya . 2014. Loss of β-catenin triggers oxidative stress and impairs hematopoietic regeneration. Genes Dev. 28: 995–1004.

154