<%BANNER%>

Acquisition of Anoikis Resistance by Attenuation of Mitochondrial Respiration and Reactive Oxygen Species

Permanent Link: http://ufdc.ufl.edu/UFE0043595/00001

Material Information

Title: Acquisition of Anoikis Resistance by Attenuation of Mitochondrial Respiration and Reactive Oxygen Species
Physical Description: 1 online resource (132 p.)
Language: english
Creator: Kamarajugadda, Sushamadevi
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: anoikis -- metabolism -- metastasis
Biochemistry and Molecular Biology (IDP) -- Dissertations, Academic -- UF
Genre: Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Cancer cells commonly exhibit aberrant glucose metabolism characterized by a preference for aerobic glycolysis rather than mitochondrial oxidation. However, the significance of this phenomenon, known as the Warburg effect, remains incompletely understood. To metastasize, cancer cells must overcome matrix detachment-induced apoptosis, or anoikis. It is unclear whether tumor metabolism contributes to anoikis resistance and metastasis. Here we show when detached from matrix, untransformed mammary epithelial cells (MCF10A) undergo metabolic reprogramming by markedly upregulating pyruvate dehydrogenase (PDH) kinase 4 (PDK4) through estrogen-related receptor gamma (ERR?), thereby attenuating the flux of pyruvate into mitochondrial oxidation. Depletion of PDK4 or activation of PDH enhances mitochondrial respiration and oxidative stress in suspended MCF10A cells and sensitizes them to anoikis. Therefore, decreased glucose oxidation confers resistance to anoikis in untransformed mammary epithelial cells. Consistent with this finding, matrix-detached MCF10A cells also elevate the antioxidant manganese superoxide dismutase (MnSOD) to alleviate oxidative stress and prolong survival in suspension. Cancer cells, due to reduced glucose oxidation, inherently possess survival advantage in suspension. Normalization of glucose metabolism by activating PDH restores anoikis in metastatic breast cancer cells MDA-MB-231. Depletion of PDK1 curtails MDA-MB-231 cells metastatic potential by reducing the lung tumor incidence in vivo. Taken together, our study suggests that intervention of tumor metabolism by developing therapy against PDK may open new avenues for anti-metastasis treatment .
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sushamadevi Kamarajugadda.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Lu, Jianrong.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043595:00001

Permanent Link: http://ufdc.ufl.edu/UFE0043595/00001

Material Information

Title: Acquisition of Anoikis Resistance by Attenuation of Mitochondrial Respiration and Reactive Oxygen Species
Physical Description: 1 online resource (132 p.)
Language: english
Creator: Kamarajugadda, Sushamadevi
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: anoikis -- metabolism -- metastasis
Biochemistry and Molecular Biology (IDP) -- Dissertations, Academic -- UF
Genre: Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Cancer cells commonly exhibit aberrant glucose metabolism characterized by a preference for aerobic glycolysis rather than mitochondrial oxidation. However, the significance of this phenomenon, known as the Warburg effect, remains incompletely understood. To metastasize, cancer cells must overcome matrix detachment-induced apoptosis, or anoikis. It is unclear whether tumor metabolism contributes to anoikis resistance and metastasis. Here we show when detached from matrix, untransformed mammary epithelial cells (MCF10A) undergo metabolic reprogramming by markedly upregulating pyruvate dehydrogenase (PDH) kinase 4 (PDK4) through estrogen-related receptor gamma (ERR?), thereby attenuating the flux of pyruvate into mitochondrial oxidation. Depletion of PDK4 or activation of PDH enhances mitochondrial respiration and oxidative stress in suspended MCF10A cells and sensitizes them to anoikis. Therefore, decreased glucose oxidation confers resistance to anoikis in untransformed mammary epithelial cells. Consistent with this finding, matrix-detached MCF10A cells also elevate the antioxidant manganese superoxide dismutase (MnSOD) to alleviate oxidative stress and prolong survival in suspension. Cancer cells, due to reduced glucose oxidation, inherently possess survival advantage in suspension. Normalization of glucose metabolism by activating PDH restores anoikis in metastatic breast cancer cells MDA-MB-231. Depletion of PDK1 curtails MDA-MB-231 cells metastatic potential by reducing the lung tumor incidence in vivo. Taken together, our study suggests that intervention of tumor metabolism by developing therapy against PDK may open new avenues for anti-metastasis treatment .
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sushamadevi Kamarajugadda.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Lu, Jianrong.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043595:00001


This item has the following downloads:


Full Text

PAGE 1

1 ACQUISITION OF ANOIKIS RESISTANCE BY ATTENUATION OF MITOCHONDRIAL RESPIRATION AND REACTIVE OXYGEN SPECIES By SUSHAMA KAMARAJUGADDA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FU LFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Sushama Kamarajugadda

PAGE 3

3 To my parents and my husband for their unconditional love and support

PAGE 4

4 ACKNOWLEDGMENTS Firstly, I would like to t hank my advisor Dr. Jianrong Lu for giving me an opportunity to work in his lab. He is a great scientist and has exceptional intellectual ability. His guidance and technical expertise have helped me achieve my research goals. His enthusiasm and ardent zeal for science keeps all of our lab members motivated. I sincerely thank him for his support and encouragement during my Ph.D. program. Next, I would like to thank my committee members; Dr. Kevin Brown, Dr. Jorg Bungert, Dr. Susan Frost, and Dr. Lizi Wu. Th ey have been very supportive of my research project. Their insight has always helped my project go in the right direction. I thank them for their valuable comments, guidance and critical evaluation of my work. I would like to Dr. Brown for his advice and guidance all through my Ph.D. He has always been my second mentor in giving both personal and professional advices. I am very thankful to him for all his help at every step of my Ph.D. program. I would like to thank Dr. Susan Frost for her valuable advice during my Medical guild research competition. She is an amazing teacher and has always made sure to correct my mistakes. I thank Dr. Frost for all her guidance. I would like to thank Dr. Jorg Bungert for providing me with his input for my qualifying propo sal. I thank him for all his help and support. Next, I would like to thank Dr. Nicholas Simpson for teaching me NMR technique. He is a great teacher. I admire his enthusiasm, patience and perseverance. I had a great time working with him and thank him ver y much for all his help. I would like to thank Dr. Anna Maria from Christiaan Leeuwenburgh lab for her advice and technical assistance.

PAGE 5

5 Next, I would like to thank all current and past lab members for all their help and advice. I would like to thank my undergraduate Ms. Lauren Stemboroski for helping me with my experiments. She is very diligent and hard working undergraduate. I thank my friends: Dr. Vasumathi Kode, Dr. Mansi Parekh, Dr. Su Nayak, Dr. Lisa Dyer and Dr. Carolina Pardo for their help, advi ce and support. They are all wonderful people. and encouragement. I thank my parents, older brother and my loving husband for their unconditional love and support. I specially th ank my husband for putting up with me all through my Ph.D. program. He has always been my confidence booster and a great critic of my work.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 10 LI ST OF ABBREVIATIONS ................................ ................................ ........................... 12 ABSTRACT ................................ ................................ ................................ ................... 15 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 17 Ove rview ................................ ................................ ................................ ................. 17 Anoikis in Normal Cells ................................ ................................ ..................... 17 Definition and significance ................................ ................................ ......... 17 Mechanisms underlying anoikis ................................ ................................ 17 Anoikis Resistance in Tumor Cells ................................ ................................ ... 19 Signficance of anoikis resistance ................................ ............................... 19 Mechanisms underlying anoikis resistance ................................ ................ 19 Glucose Metabolism ................................ ................................ ......................... 21 Importance and s ignificance ................................ ................................ ...... 21 Lactate dehydrogenase (LDH) ................................ ................................ ... 21 Pyruvate dehydrogenase complex (PDC) ................................ .................. 22 Regulation of PDC enzyme complex ................................ ......................... 23 Reactive oxygen species (ROS) ................................ ................................ 25 Tumor Metabolism ................................ ................................ ............................ 27 Warburg effect and its significance ................................ ............................ 27 Advantages of Warburg effect ................................ ................................ .... 27 Oncogenes inducing the Warburg effect ................................ .................... 29 Mutations in mitochondrial enzymes that favor the Warburg effect ............ 30 Human Breast Cancer ................................ ................................ ...................... 31 Types of Breast cancer ................................ ................................ .............. 31 2 MATERIALS AND METHODS ................................ ................................ ................ 37 Cell Culture ................................ ................................ ................................ ............. 37 Reagents ................................ ................................ ................................ ................ 37 Plasmids ................................ ................................ ................................ ................. 38 Poly Hema Coated Plates for Suspension Cu lture Cells ................................ ........ 38 RNA Isolation, and Northern blotting ................................ ................................ ....... 38 Reverse Transcription, and Real Time PCR ................................ ........................... 39 Short hairpin RNA (shRNA) Vector Construction ................................ .................... 40

PAGE 7

7 Plasmid DNA MiniPrep ................................ ................................ ........................... 42 Retroviral Short Hai rpin RNA Production and Transduction of Target Cells ........... 42 Protein Isolation and Immunoblotting ................................ ................................ ...... 44 Site Directed Mutagenesis ................................ ................................ ...................... 45 Lentiviral Over Expression of Constitutively Active FLAG ........... 46 Measurement of PDH Activity ................................ ................................ ................. 47 Trypan Blue Exclusion Assay ................................ ................................ ................. 48 Caspase 3/7 Activity Assay ................................ ................................ ..................... 48 Annexin V/7 AAD Analysis ................................ ................................ ...................... 49 Measurement of Oxygen Consumption Rate ................................ .......................... 49 Measurement of Intracellular ROS ................................ ................................ .......... 50 Measurement of Intracellular Lactate ................................ ................................ ...... 50 Experimental Metastasis Assay in Mice ................................ ................................ .. 51 Tissue Sectioning and H&E Stainin g ................................ ................................ ...... 51 Statistics ................................ ................................ ................................ ................. 52 3 INFLUENCE OF GLUCOSE METABOLISM ON ANCHORAGE INDEPENDENT SURIVIVAL IN MAMMARY EPITHELIAL CELLS ................................ ................... 55 Background ................................ ................................ ................................ ............. 55 Results ................................ ................................ ................................ .................... 56 Induction of PDK upon Detachment from ECM ................................ ................ 56 Upregulation of PDK4 Antagonizes Anoikis ................................ ...................... 57 Activation of PDH Sensitizes Cells to Anoikis ................................ ................... 59 Depletion of PDK4 Increases Mitochondrial Oxidation ................................ ..... 60 Treatment with Antioxidant Rescues PDK4 Depleted Cells from Anoikis ......... 61 Estrogen Related Receptor Activates PDK4 in Response to Cell Detachment ................................ ................................ ................................ ... 62 Summary ................................ ................................ ................................ ................ 64 4 PDK PROMOTES ANO IKIS RESISTANCE AND TUMOR METASTASIS IN BREAST CANCER CELLS ................................ ................................ ..................... 80 Background ................................ ................................ ................................ ............. 80 Results ................................ ................................ ................................ .................... 81 Induction of PDK4 Promotes Anoikis Resistance in RAS Transformed Mammary Epithelial Cells ................................ ................................ .............. 81 Induction of PDK4 in Ras transformed MCF10A cells (10ACA1.1) ............ 81 PDK4 resists anoikis in Ras transformed MCF10A cells (MCF10ACA1.1) ................................ ................................ ..................... 82 Warburg Effect Promotes Anoikis Resistance and Metastasis in Breast T umor Cells ................................ ................................ ................................ ... 82 Matrix detachment favors Warburg effect in MDA MB 231 cells ................ 82 Forced activation of mitochondrial oxidation induce s anoikis in MDA MB 231 ................................ ................................ ................................ ... 83 PDK1 enhances Warburg effect in matrix detached MDA MB 231 cells .... 84

PAGE 8

8 Depletion of PDK1 activates mitoc hondrial oxidation and abrogates anoikis resistance in MDA MB 231 cells ................................ ................. 85 PDK1 enhances breast tumor metastasis in vivo ................................ ....... 86 Summ ary ................................ ................................ ................................ ................ 87 5 ANTIOXIDANT PROTECTION FROM DETACHMENT INDUCED OXIDATIVE STRESS IN MAMMARY EPITHELIAL CELLS ................................ ........................ 98 Background ................................ ................................ ................................ ............. 98 Endogenous Antioxidants Defense System ................................ ...................... 98 Manganese Superoxide Dismutase (MnSOD) ................................ .................. 98 Good ROS and Bad ROS ................................ ................................ ................. 99 Results ................................ ................................ ................................ .................. 100 Induction of MnSOD in Matrix Detached Mammary Epithelial Cells ............... 100 Depletion of MnSOD Sensitizes Mammary Epithelial Cells to Anoikis ........... 101 Epistatic Relationship Between PDK4 and MnSOD upon Matrix Detachment 102 Summary ................................ ................................ ................................ .............. 102 6 CONCLUSIONS AND FUTURE DIRECTIONS ................................ .................... 109 Conclusions ................................ ................................ ................................ .......... 109 Future Directions ................................ ................................ ................................ .. 116 LIST OF REFERENCES ................................ ................................ ............................. 120 BIOGRAPHICAL SK ETCH ................................ ................................ .......................... 132

PAGE 9

9 LIST OF TABLES Table page 2 1 Primers used for real time RT PCR ................................ ................................ ... 54 2 2 sh RNA Oligos designed for retrovirus mediated knockdown ............................. 54 2 3 Antibodies used for Immunoblotting (IB) ................................ ............................ 54

PAGE 10

10 LIST OF FIGURES Figure page 1 1 Mechanisms regulating anoikis and anoikis resistance.. ................................ ... 34 1 2 Pyruvate dehydrogenase complex (PDH) and its reaction. ............................... 35 1 3 Regulation of PDH complex.. ................................ ................................ ............. 36 2 1 Cloning vector information for microRNA adapted retroviral vector.. ................. 53 3 1 Matrix detachment induces PDK4 expression in mammary epithelial cells. ...... 65 3 2 Induction of PDK4 at RNA and protein level in MCF10A suspension cells. ....... 66 3 3 Expression of LDHA in attached and suspended MCF10A cells. ...................... 67 3 4 Depletion of PDK4 using retroviral short hairpin R NA in MCF10A cells. ........... 68 3 5 Knockdown of PDK4 sensitizes MCF10A cells to anoikis ................................ 69 3 6 Detachment from matrix attenuates PD H activity in MCF10A cells. .................. 71 3 7 ............ 72 3 8 Activation of PDH sensitizes MCF10A cells to anoikis.. ................................ .... 73 3 9 Depletion of PDK4 increases mitochondrial oxidation in MCF10A cells. ........... 74 3 10 Antioxidant treatment rescues PDK4 depleted MCF10A cells from anoikis.. ..... 75 3 11 .......................... 76 3 12 .... 77 3 13 ................. 78 3 14 Summary model. ................................ ................................ ............................... 79 4 1 Detachmen t from matrix upregulates PDK4 in MCF10ACA1.1 .. ........................ 88 4 2 Depletion of PDK4 sensitizes MCF10ACA1.1 cells to anoikis. .......................... 89 4 3 Mat rix detachment attenuates PDH activity in MDA MB 231cells. .................... 90 4 4 MB 231 cells., ...... 91 4 5 Forced activation of PDH induces anoikis in MDA MB 231 cells. ...................... 92

PAGE 11

11 4 6 Induction of PDK1 at protein level in MDA MB231 cells ................................ .... 93 4 7 Depletion of PDK1 switches glucose metabolism in MDA MB 231 cells.. ......... 95 4 8 Depletion of PDK1 abolishes anoikis resistance in MDA MB 231 cells.. ........... 96 4 9 Depletion of PDK1 abrogates tumor metastasis in vivo. ................................ ... 97 5 1 Induction of MnSOD in HMEC and MCF10A suspended cells.. ...................... 103 5 2 Knockdown of MnSOD in MCF10A cells. ................................ ........................ 104 5 3 Depletion of MnSOD leads to increased anoikis in MCF10A.. ......................... 105 5 4 Epistatic relationship b etween PDK4 and MnSOD in MCF10A cells.. ............. 107 5 5 Summary model. ................................ ................................ ............................. 108 6 1 Expression of PDK4 in different tumor cells lines. ................................ ........... 119

PAGE 12

12 LIST OF ABBREVIATION S Anoikis Detachment induced cell death 7 AAD 7 aminoactinomycin D APAF 1 Apoptotic protease activating factor 1 ASK1 Apoptosis signal regulating kinase ATP Adenosine trinucleotide phosphate BCL2 B cell lymphoma 2 BAD Bcl2 associated death promoter protein BAX B cl2 associated X protein BID BH3 interacting domain death agonist BMF Bcl2 modifying factor CA Carbonic anhydrase CCAAT/enhancer DCA Dichloroacetate DCIS Ductal c arsinoma in situ ECM Extracellular matrix EGFR Epithelial growth fac tor receptor EMT Epithelial to mesenchymal transition ER Estrogen receptor ERR Estrogen related receptors ErbB2 Erythroblastosis oncogene B ERK Extracellular signal regulated Kinase ETC Electron transport chain EV Empty vector FADD Fas associated via de ath domain

PAGE 13

13 FADH 2 Flavin adenosine dinucleotide FAK Focal adhesion kinase FH Fumarate hydratase FLIP FLICE (FADD like IL converting enzyme) inhibitory protein FOXO3A Forkhead family of transcription factor 3a G6P Glucose 6 phosphate GPx Glutathione peroxidase GSH L Glutathione (reduced form) HER2 Human epidermal growth factor receptor 2 H&E Hematoxylin and Eosin HK Hexokinase IDC Invasive ductal c arcinoma IDH2 Isocitrate dehydrogenase 2 ILC Invasive lobular c arcinoma ILK Integrin linked kinase LA Lipoic acid LCIS Lobular carcinoma in situ LDH Lactate dehydrogenase MCL1 Myeloid cell leukemia 1 protein MEK Mitogen act ivated protein kinase kinase MnSOD Manganese superoxide dismutase NADH Nicotinamide adenine dinucleotide NADPH Nicotinamide adenine dinucleotide phosphate Nuclear factor kappa light chain enhancer of activated B cells OCR Oxygen consumption rate

PAGE 14

14 PE Phycoerythrin Poly Hema Poly(2 hydroxyethyl methacrylate) PPP Pentose phosphate pathway PTEN Phosphatase and tensin homolog PI3K Phosphoinositide 3 Kinase PDH/C Pyruvate dehydrogenase complex PDK Pyruvate dehydrogenase kinase PFK 1 Phosphofructokinase 1 Peroxisome proliferator 1 alpha PKM2 Phospho fructokinase isoform M2 P eroxisome proliferator activated recep tor PR Progesterone receptor p53 Tumor protein 53 ROS Reactive oxygen species SDH Succinate dehydrogenase (SDH) shRNA short hairpin RNA SOD Superoxide dismutase TCA Tricarboxylic acid cycle VEGF Vascular endothelial growth factor VHL von Hippel Lindau v Src Viral sarcoma XIAP X linked inhibitor of apoptosis protein

PAGE 15

15 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ACQUISITION OF ANOIKIS RESISTANCE BY ATTENUATION OF MITOCHONDRIAL RESPIRATION AND REACTIVE OXYGEN SPECIES By Sushama Kamarajugadda December 2011 Chair: Jianrong Lu Major: Medical Sciences Biochemistry and Molecular Biology Cancer cells commonly exhibit aberrant glucose metabolism characterized by a preference for aerobic glycolysis rather than mitochondrial oxidation. However, the significance of this phenomenon, known as the Warburg effect, remains incompletely understood. To metastasize, cancer cells must overcome matrix detachment induced apoptosis, or anoikis. It is unclear whether tumor metabolism contributes to anoikis resistance and metastasis. Here we show when detached from matrix, untransformed mammary epithelial ce lls (MCF10A) undergo metabolic reprogramming by markedly upregulating pyruvate dehydrogenase (PDH) kinase 4 (PDK4) through estrogen related receptor gamma (ERR ), thereby attenuating the flux of pyruvate into mitochondrial oxidation Depletion of PDK4 or a ctivation of PDH enhances mitochondrial respiration and oxidative stress in suspended MCF10A cells and sensitizes them to anoikis. Therefore, decreased glucose oxidation confers resistance to anoikis in untransformed mammary epithelial cells Consistent wi th this finding, matrix detached MCF10A cells also elevate the antioxidant manganese superoxide dismutase (MnSOD) to alleviate oxidative stress and prolong survival in suspension

PAGE 16

16 Cancer cells, due to reduced glucose oxidation, inherently possess survival advantage in suspension. Normalization of glucose metabolism by activating PDH restores anoikis in metastatic breast cancer cells MDA MB 231. D epletion of PDK 1 curtails MDA MB 231 cells metastatic potential by reducing the lung tumor incidence in vivo Ta ken together our study suggests that intervention of tumor metabolism by developing therapy against PDK may open new avenues for anti metastasis treatment.

PAGE 17

17 CHAPTER 1 INTRODUCTION Overview Anoikis in Normal Cells Definition and significance Anoikis i triggered by loss of cell adhesion or inappropriate cell anchorage 1 It prevents a detached cell from reattaching to an inadequate or inappropriate m atrix 2 The physiological relevance of anoikis is to maintain tissue homeostasis, and proper development 3 Binding of integrin proteins to the extracellular matrix (ECM) provides survival signals, which helps in establishing appropriate cell matrix interactions. Any association of detached cells with inappropriate matrix wou ld lead to incorrect integrin engagement followed by activation of proapoptotic genes, and consequently anoikis 4 Mechanisms underlying anoikis Under normal physiological conditions, cells are protected from anoikis in three different scenarios: (i) During reattachment of the cell to an appro priate matrix, (ii) detachment of a cell to move towards a chemo attractant, and (iii) through cell cell contacts 2 It is well established that the integrins associated with the ECM are critical for cell survival 4 They are activated by various signa ling molecules such as focal adhesion kinase (FAK), integrin linked kinase (ILK), phosphoinositide 3 kinase (PI3K)/Akt or protein kinase B (PKB), and extracellular signal related kinase (ERK) 5,6 Upon detachment of cells from ECM, anoikis is activated either by an extrinsic /death receptor pathway 7,8 or an intrinsic/mitochondrial pathway 9 (Figur e 1 1). In the intrinsic pathway, cells undergo apoptosis due to activation of pro apoptotic B cell

PAGE 18

18 lymphoma 2 (BCL2) family proteins. This is followed by release of cytochrome C from mitochondria and activation of downstream caspase events. Several report s have shown that pro apoptotic BH3 only proteins of BCL2 family like Bim 5 Bcl2 modifying factor (BMF) 9 11 and Bcl2 associated death promoter protein (BAD) 12 are activated during cell detachment and trigger anoikis. These pro apoptotic genes are activated by different signaling pathways such as fork head family of transcription factor 3a (FOXO3a) signaling 13 Matrix detachment inactivates survival signaling pathways such as epidermal growth factor receptor mitogen activated protein kinase kinase extracellular signal regulated kinase ( ERK1/2) 14 and epithelial growth factor receptor (EGFR) involving PI3K pathw ays 5 which may result in activation of certain pro apoptotic genes. Therefore, oncogenes such as erythroblastosis oncogene B2 (ErbB2) and viral sarcoma (v Src) suppress anoikis through ERK mitogen protein kinase pathway in cultured mammary epithelial ac ini 15 It has been shown that hypoxia suppresses Bim and BMF expression via ERK signaling and inhibits luminal clearing during morphogenesis in an in vitro three dimensional mammary epithelial culture 16 On t he other hand, t he extrinsic pathway involving death receptors like Fas and tumor necrosis ) may trigger apoptosis in certain cells upon detachment 17 This pathway activates caspase 8, which may trigger apoptosis with or without the involvement of intrinsic pathway. The pro apoptotic BH3 interacting domain death agonist (BID) protein is cleaved by caspase 8 and the truncated BID activates the intrinsic pathway during anoikis 18 To date, the involvement of the death receptor pathway in anoikis is ambiguous. However, the involvement of intrinsic pathway is well established.

PAGE 19

19 Another mechanism driving anoikis during detachment is protein 53 (p53). p53 is a crucial molecule that regulates cell cycle, maintains genomic stability, inhibits angiogenesis and is henc e known as a tumor suppressor protein. It has been shown that inhibition of p53 increases anoikis resistance in fibroblasts 19 and thyroid epithelial cells 20 v 3 integrin activates p53 leading to anoikis in endotheli al cells 21 Th e tumor suppressor gene phosphatase and tensin homolog (PTEN), which is transactivated by p53, restores anoikis in breast cancer cells 22 These studies confirm the role of p53 in inducing anoikis upon detachment from ECM in different cell lines. Anoikis Resistance in Tumor Cells Si gnficance of anoikis resistance Unlike normal cells, tumor cells are resistant to anoikis. Anoikis resistance is a hallmark of malignant tumors for an anchorage independent growth and survival 23 The underlying purpose of evad ing anoikis by a tumor cell is to survive in an environment outside its own niche. Anoikis resistance has been implicated in tumor invasion and metastasis. Mechanisms underlying anoikis resistance Tumor cells adopt different mechanisms to evade anoikis. On e of them is the constitutive activation of survival signals like PI3K, mitogen activated protein kinase kinase (MEK)/ERK etc, for proliferation, survival, and migration 24 Another common strategy is the activation of epithelial to mesenchymal transition (EMT) by activating transcription factors such as snail, twist, slug, etc., EMT changes the tumor cell morphology by suppressing epithelial markers like E Cadherin and up regulating mesenchymal markers like vimentin and fibronectin 25 allowing the tumor cells to

PAGE 20

20 migrate and invade the distant organs ev en in the absence of matrix. In addition, a change in the expression pattern of integrin proteins plays a critical role in anoikis resistance. It has been shown that MCF10A cells overexpressing ErbB2 upregulate tyrosine phosphorylation 26 Another plausible protection fr om anoikis implicates reactive oxygen species (ROS) in tumor cells. A recent report suggested that detachment from extracellular matrix leads to accumulation of ROS in prostrate cancer cells, which results in the activation of survival signals such as EGFR via src kinase 27 However, the implication of ROS in anoikis is still controversial. Tumor cells upregulate apoptosis inhibiting proteins such as FLICE (FADD like IL converting enzyme) inhibitory protein (FLIP), an inhibitor of caspase 8, and X linked inhibitor of apoptosis protein (XIAP), an inhibitor of caspase 3/7, upon loss of anchorage to evade apoptosis 23 (Figure 1 1). Over expression of oncogenes such as ras has been reported in some tumor cells, which would activate survival signaling pathways 28 The hypoxic environment prevailing within a tumor also helps in promoting anoikis resistance by upregulating different survival and proliferation genes 29 Hypox ia regulated genes will be discussed in detail in tumor metabolism section. Apart from the signaling pathways and the inhibitors of pro apoptotic genes, the role of metabolism in regulating anoikis resistance is quite intriguing. A genome wide micro array study performed by Schmelzle T et al ., showed an upregulation of pyruvate dehydrogenase kinase isozyme 4 (PDK4), an important enzyme that regulates glucose metabolism, in MCF10A suspension culture 10 The microarray data indicate a plausible

PAGE 21

21 involvement of glucose metabolism in regulating anoikis. A detailed description of glucose metabolism is in the fol lowing section. Glucose Metabolism Importance and significance Glucose metabolism is one of the major fuels for energy production. Glucose is transported across the plasma membrane via glucose receptors (Glut). The first rate determing step in glucose meta bolism is conversion of glucose to glucose 6 phosphate (G6P), which is carried out by Hexokinase (HK) enzyme. G6P enters the glycolytic pathway to generate Nicotinamide adenine dinucleotide (NADH), Adenosine triphosphate (ATP), and pyruvate and/or the pent ose phosphate pathway (PPP) to generate Nicotinamide adenine dinucleotide phosphate (NADPH) and pentose sugars for nucleotide biosynthesis. In the presence of oxygen, pyruvate will enter mitochondria and be processed in the Krebs cycle or tricarboxylic aci d (TCA) cycle to produce energy in the form of ATP required to drive various cellular processes. However in the absence of oxygen (anaerobic condition), pyruvate is converted into lactate. Lactate dehydrogenase (LDH) LDH is an enzyme that cat alyzes the interconversion of pyruvate to lacta te under low oxygen conditions LDH has five isoenzymes produced from two polypeptide chains M and H in different combinations 30 The two genes LDHA and LDHB encode the polypeptide chains M and H respectively. The five isoenzymes are; LDH1 B4, LDH2 B3A1, LDH3 B2A2, LDH4 BA3, LDH5 A4 31 Of all the isoforms, LDH5 is considered to be efficient in catalyzing pyruvate to lactate con versi on. LDH1 is efficient in pyruvate formation, which is fed into the Krebs cycle 32

PAGE 22

22 Pyruvate dehydrogenase complex (PDC) Function and Importance: PDC is a multienzyme complex present in mitochondria. PDC carries out the irreversible conversion of p yruvate to acetyl CoA and releases CO 2 and NADH. Thus, PDC activates the Krebs cycle and produces adenosine trinucleotide phosphate (ATP) via electron transport chain (ETC). When carbohydrate content is high in the body, PDC exists in an active dephosphor ylated form to oxidize glucose and produce energy 33 Under low carbohydrate content, PDC exist s in an inactive phosphorylated form to limit the glucose consumption by non neuronal tissue 33 The key players to regulate phosphorylated and dephosporylated PDC forms in a tissues specific manner are pyruvate dehyrogenase kinase (PDK) 34 and pyruvate dehydrogenase phospha tase (PDP) 35 respectively. Structure and mechanism of PDC : The enzymatic reactions of PDC are catalyzed by E1 pyruvate dehydrogenase, E2 dihydrolipoyl transacetylase, and E3 dihydrolipoamide dehydrogenase (E3) components in a sequential manner (Figure 1 2) 36 The core structure of PDC consists of a multidomain E2 and E3 binding protein (E3BP). The two N terminal lipoyl domains of E2 and one lipoyl domain of E3BP are designated as L1, L2, and L3, respectively 37 The complex consists of 60 subunits of E2, which carries out the transacetylati on reaction. E2 and E3BP provide E1 binding and E3 binding domain via a 20 30 amino acid linker region 38 There are five sequential steps involved in the conversion of pyruvate to acetyl CoA by the PDC enzyme complex 37 ; E1 catalyzes decarboxylation of pyruvate in the presence of the cofactor thiamine diphosphate.

PAGE 23

23 E1 catalyzes reductive acetylation of lipoyl lysine prosthetic groups present on the lipoyl domains of E2 and E3BP. It is considered to be the rate limiting step. E2 catalyzes the transfer of the acetyl group from dihydrolipoyl prosthetic groups to CoA (Cofactor). E3 accepts the reducing equivalents and regenerates lipoyl prosthetic groups and meanwhile, reduces the thiol FAD system. E3 furt her transfers the reducing equivalents from FAD to NAD + thus producing NADH+H + Regulation of PDC enzyme complex As mentioned above, the PDC complex is highly regulated by PDK and PDP enzymes to maintain the glucose metabolic homeostasis within the tissu es depending on carbohydrate availability (Figure 1 3) 34 Below they are discussed in detail; (i) Pyruvate dehydrogenase kinase (PDK): PDK is a serine kinase family enzyme, which regulates glucose metabolism by its inhibitory effects on PDC enzyme. PDK inactivates PDC e nzyme by phosphoryl subunit. It exists as four different isoenzymes designated as PDK1, PDK2, PDK3, and PDK4 35 The expre ssion pattern for four isoenzymes is tissue specific, which helps in regulating PDC activity depending on their metabolic requirements 39 Of all, PDK2 is widely distributed with high expression in liver, heart and kidney. PDK4 is expressed in liver, heart, kidney, and oxidative skeletal muscle whereas expression of PDK1 is limited to heart, and PDK3 is abundantly expressed in testis 39 PDK is a dimer consisting of a C terminal catalytic domain (cat) and a regulatory N terminal (R domain) 38 The ATP/ADP binding site is located on the cat domain and binding of regulatory ligands occurs at the R domain. PDK binds to the PDC complex

PAGE 24

24 via the lipoyl domain of E2 and thus obtains access to the E1 substrate 38 The binding affinity of the four isoforms for the lipoyl domain of E2 subunit is different, which in turn regulates their enzymatic activity. Greatest binding affinity is observed for PDK3 followed by PDK1=PDK2 and PDK4, respectively 36 PDC complex is phosphorylated at subunit by PDK enzym subunit at three differ ent serine residues; ser 293 ( phosphorylation site 1), ser 300 (phos phorylation site 2) and, ser 232 (phosphorylaiton site 3) 40 The activity of PDK is enhanced by PDH products i.e., acetyl CoA and NADH+H + On the other hand, ADP and increased pyruvate levels inhibit PDK activity to reactivate PDC 39 Of all four i soenzymes PDK2 is very sensitive to its regulatory effectors like NADH and acetyl CoA (positive regulators) and, ADP and pyruva te (negative regulators). On the other hand PDK3 is insensitive to the inhibitory effects of pyruvate and ADP when compared with PDK1, 2, and 4 36,39 PDK1 41,42 and PDK3 43,44 genes are activated by hypoxia inducible transcription tumor growth. PDK2 and PDK4 genes are elevated during starvation 45 and diabetes 46 in various tissues to reduce the oxidation of pyruvate to acetyl CoA The different acti vators of PDK2 and PDK4 are: peroxisome proliferator coactivator 1 alpha (PGC 47 estrogen related receptors (ERRalpha and ERRgamma ) 48 peroxisome proliferator activated receptor (PPAR ) 42,49 thyroid hormone (T(3)) 50 and CCAAT/enhancer 51 Inhibition of PDK to activate PDC is an important therapeutic strategy in the treatment of diabetes 52 heart ischemia, and in cancer 53 PDK4 activity is inhibited in diabetic obese zuker mice by AZ7545 inhibitor 52 AZ7545 binds to lipoyl domain of E2

PAGE 25

25 component and prevents binding of PDK. Dichloroacetate (DCA) is another inhibitor of PDK, which is widely used to treat lact ic acidosis in children deficient of PDC 54 and in cancer cells where activation of PDC lead s to tumo r regression 53 DCA prevents PDK activity by binding to L2 binding pocket at N terminal region of PDK 55 (ii) Pyruvate dehyrogenase phosphatase (PDP) : PDP belong s to 2c class of protein phosphatases. It activates PDH enzyme by d PDP has two isoforms, PDP1 and PDP2 consisting of ~56kda catalytic subunit (PDPc) and a large ~95.6kDa regulatory subunit (PDPr) 35 Both isoforms require Mg +2 for their activation and their activity is highly regulated by this metal. Insulin upregulates PDP activity and activates the PDC complex to metabolize the hi gh glucose content present within the tissue 37 Reactive oxygen species (ROS) Reactive oxygen speci es (ROS) are the free radicals derived from the molecular oxygen. They are highly unstable, reactive and cause damage to the DNA, protein, and lipids. The major source of ROS production is the mitochondrial electron transport chain (ETC) 56 The TCA cycle not only oxidizes acetyl CoA to CO 2 but also produces reducing equivalents such as NADH + and flavin adenine dinucleotide (FADH 2 ). These reducing water molecule via ETC During this process, some electrons are leaked out of the respiratory complexes and react with oxygen molecule to form reactive oxygen species 57 The first ROS species produced is the highly unstable superoxide anion (O 2 ), which is readily dismutated by superoxide dismutases to generate hydroperoxyl (HO 2 ). The two molecules of HO 2 with each other react to form hydrogen peroxide (H 2 O 2 ) and water (H 2 O). Unlike superoxide anion, H 2 O 2 is not a strong oxidant. In the

PAGE 26

26 prese nce of transition metal ions such as Fe 2+ H 2 O 2 is converted into a powerful oxidizing agent that damages the DNA and leads to lipid oxidation 58 Another initial ROS species generated is nitric oxide (NO ). The sources of NO are vascular endothelium, nerve terminals and mitochondrial NO synthase (NO S) 59 The interaction between NO and O 2 produces ONOO which is quickly converted into a very cytotoxic peroxynitrous acid (ONOOH) rad ical. Lower levels of ROS helps in cellular proliferation and gene transcription by activating signaling pathways 60 Higher levels of ROS damage the cellular DNA, lipid or proteins. Therefore, maintaining redox balance within the cell is crucial for cell survival and maintenance. Increased ROS levels are eliminated by upregulation of RO S scavengers such as superoxide dismutase (SOD1, 2, and 3), glutathione peroxidase, peroxiredoxins, glutaredoxin, thioredoxin and catalase. A more detailed description about endogenous antioxidants is provided in Chapter five. Apart from mitochondria, ROS are generated during detoxification of cytochrome P450 61 and b 5 families of enxymes 62 peroxisomes 63 and plasma membrane bound oxidases such as phagocytic NADPH oxidase 64 etc. The role of reactive oxygen species in regulating anoikis is controversial. Several reports suggest that the impact of ROS on anoikis is bot h positive 65 and negative 66 Although, the growing body of evidence indicates accumulation of ROS upon detachment from ECM leads to cell death in endothelial cells 67 mammary epithelial cells 68 keratinocytes 69 etc. Glucose metabolism is one of the major fuels to produce ATP and in the process generates ROS in normal cells. However, cancer cells have altered glucose

PAGE 27

27 metabolism, which helps them survive and proliferate in varying microenvironments. A detailed description is in the following section. Tumor Metabolism Warburg effect and its significance Aberrant energy me tabolism is one of the hallmarks of cancer cells 70 Tumor cells have an altered glucose metabolism when compared with normal cells. Unlike normal cells, tumor cells utilize glycolysis (conversion of pyruvate to lactate) for their energy production even in the presen 71 Otto Warburg first observed this phenomenon in 1920 and it is therefore known as the It is now well accepted that enhanced glycolysis facilitates cell proliferation 72 which requires not only ATP but als o synthesis of nucleotides, lipids, and proteins. Advantages of Warburg effect Mitochondrial respiration yields higher energy (38 ATP) when compared with glycolysis (2ATP). Despite this advantage, most tumor cells favor glycolysis over mitochondrial respir ation. Initially, Otto Warburg assumed that cancer cells favor aerobic glycolysis due to damaged mitochondria 71 However, subsequent reports disproved this assumption by showing functional and intact mitochondria in different tumor cells 72 74 making it difficult to understand the significance of less energy efficient aerobic glycolysis over mitochondrial respiration in cancer cells. In general, cancer cells consume much more glucose through increased glucose uptake and glycolysis. There are several theories to explain the advantages of the War burg effect or the altered glucose metabolism in cancer cells. The first and the most popular theory attributed to hypoxic environment in tumors. Hypoxia stabilizes and

PAGE 28

28 s everal glycolytic genes such as Glut1 receptors, Hexokinase II, LDHA 75 PDK1 41,42 and PDK3 43 etc., So, mitochondrial respiration may be very effi cient in producing 18 times the ATP per mole of glucose but the rate of anaerobic glycolysis is 100 times higher in tumor cells 76 In addition, increased glycolysis within the tumor cells result in increased release of H + ions into the surrounding environment leading to acidosis 75 The acidified environment of tumor cells increases the expression of H + transporters and Na +2 H + exchangers t o attenuate the surrounding cellular toxicity. The second theory to explain the selective advantage of Warburg effect in tumor cells is due to the stimulation of pentose phosphate pathway (PPP) The PPP pathway has both oxidative and non oxidative branch es. The oxidative branch utilizes glucose to yield ribose, which is used for the production of building blocks such as RNA and DNA. It has been shown that tumor cells upregulate transketolase, an important enzyme of PPP pathway, for their proliferation 77 The oxidative branch also generates NADPH. NADPH is an important cofactor of glutathione reductase, which is an endogenous antioxidant. Thus, the production of NADPH induces antio xidant defense mechanism within the cells to prevent oxidative damage 78 Moreover, the highly proliferating cancer cells require increasing amounts of biomass (amino acids, lipids and DNA) for their growth and survival. To meet these demands, t he int ermediate metabolites produced during glycolysis may be shunted as anabolic precursors for the synthesis of amino acids and nucleosides 72 The third theory to explain the preferential use of aerobic glycolysis in tumor cells is to escape ROS from mitochondria. Since mitochondria are the major source of ROS

PAGE 29

29 production, the actively dividing tumor cells may evade mitochondrial respiration to prevent excessive release of ROS. There have been some reports that support this reduced ROS production, rescued the cells from apoptotic induction 42 and reduced the tumor size in mice xenograft models 79 (2) pharmacological inhibition of PDK2 with DCA in duced apoptosis in Hela cells by decreasing mitochondrial membrane potential, and increasing H 2 O 2 production and Kv channels 53 and (3) Depletion of LDHA has reduced tumor growth and maintenance due to increased ROS levels 73 Therefore, the Warburg effect in cancer cells not only provides macromolecular biosynthetic elements required for its growth and survival but also maintains cellular redox homeostasis However, it is unclear how glucose metabolism regulates tumor cells during metastasis. Oncogenes inducing the Warburg effect While the Warburg effect is favored by most of the cancer cells i t is still unclear whether it is the cause or the effect of malignant phenotype Cancer is a genetic disease resulting from mutatio ns in cellular pathways that trigger abnormal cell growth and proliferation Several oncogenes and tumor suppressors are closely connected to metabolic pathways and any alteration in their activity can promote aerobic glycolysis 80 The activated Ras or Src oncogenes have been shown to increase glucose uptake and activate a number of glycolytic enzymes 81 Another oncogenic transcription factor Myc has been shown to directly upregulate LDHA at RNA level 82 and several other glycolytic genes as well 83 Further, Myc induces pho sphofructose kinase isoform M2 (PKM2), which converts phosphoenolpyruvate to pyruvate 84 Thu s, establishing the critical role of Myc in inducing aerobic glycolysis similar to HIF 1. The oncogene AKT,

PAGE 30

30 which is activated by PI3K signaling, enhances aerobic glycolysis by activating hexokinase 2, phosphofructokinase 1 (PFK 1), and several other glyco lytic genes 85 Apart from oncogenes, the loss of tumor suppressor gene functions such as PTEN 86 and p53 87 favor aerobic glycolysis in some tumor cells. Mutations in mitochondrial enzymes that favor the Warburg effect The mutations in mitochondrial metabolic enzymes have reinforced the role of metabolism in tumorigenesis. The three key enzymes of TCA cycle with mutations are: Fumarate hydratase (FH), succinate dehydrogenase (SDH), and isocitrate dehyd rogenase 2 (IDH2). The mutations in FH and SDH lead to accumulation of fumarate and succinate intermediates, which inhibits prolyl hydroxylases that are responsible for the d 8 8 As a result, even in the presence of normal aerobic glycolysis to trigger tumorigenesis in some tumors 89 IDH2 mutations have been found in low grade gliomas 90,91 Similar to FH and SDH mutations, IDH2 mutat ion The altered glucose metabolism and its role in tumor growth and survival have been well studied. Another important feature of cancer cells is to metastasize and invade distant organs. Resistance to anoikis is a critical step fo r the tumor cells to undergo metastasis. Therefore, it is very important to explore the key regulators that promote anoikis resistance. Since the Warburg effect has been implicated in solid tumor growth, it is quite intriguing to understand its role in ano ikis resistance and tumor metastasis in vivo To elucidate the functional role of glucose metabolism in anoikis resistance and tumor metastasis, we have chosen human breast cancer as our model system.

PAGE 31

31 Human Breast Cancer Breast cancer is the cancer that begins in different areas of breast; the ducts, and the lobules. It is the second most common cancer in the United States for women and its incidence rate is 122.8 per 100,000 women each year 92 Recent advances of early detection and diagnosis of breast cancer has reduced the risk of death. But, if tumor metastasizes to dist ant organs, the mortality rate is still high. Breast cancer can invade different organs such as lung, liver, brain and bones via the lymphatic system. Breast cancer is highly heterogeneous in nature. Types of Breast cancer The different types of breast ca ncers include non invasive, invasive, recurrent and metastatic types. These classifications are based on their invasiveness, point of origin, and hormone receptor status. Ductal Carc inoma in situ (DCIS) It is non invasive type of breast cancer occurring inside the milk ducts and has not spread beyond its point of origin. This cancer has the best prognosis because they express hormone receptor positive cells (ACS). It is not life threatening but the recurrence rate is under 30% Invasive Ductal Carcinoma (IDC) It is sometimes referred to as infiltrating ductal carcinoma and is the most common type of breast cancer. It is invasive type of ductal carcinoma, which invades the fatty tissues of breast and the lymph nodes (ACS). It can affect women of any age but most common in older women (ACS). About 80% of all invasive breast cancers are IDS. Lobular Carcinoma In Situ (LCIS) It is the non invasive type of breast cancer occurring in the milk producing lobules. Like DCIS, they are responsive to hormone ther apy as they contain hormone receptor cells.

PAGE 32

32 Invasive Lobular Carcinoma (ILC) it is the second most common type of breast cancer after IDS. About 10% of all invasive breast cancers are ILC. The cancer begins in the milk producing lobules of the breast. O ver the time, ILC can spread to lymph nodes and the other areas of the body. In our entire study, we use untransformed immortalized mammary epithelial cell lines MCF10A, primary immortalized human mammary epithelial cell lines (HMEC), Ras transformed mamma ry epithelial cell line (10ACA1.1), and triple negative (Estrogen receptor negative (ER ), progesterone receptor negative (PR ), and human epidermal growth factor receptor 2 (HER2 ) negative), breast cancer cell line (MBDA MB 231) to study the role of gluc ose metabolism in promoting anoikis resistance and tumor metastasis. MCF10A cells are derived from fibrotic tissue and they are ER negative, EGF receptor negative, HER2 negative but E Cadherin positive. They are immortalized but do not form tumors in vivo HMEC cells are derived from normal human reduction mammoplasty tissue, immortalized and ER positive cells. They do not form tumors in vivo Therefore, these two cell lines are used as control or normal cells. MDA MB 231 is a triple negative and derived fr om an adenocarcinoma tissue. It is a highly aggressive and metastatic tumor cell line. The goal of our study is to elucidate the role of glucose metabolism in regulating anoikis in both normal and metastatic breast cell lines. We discovered that upon detac hment from matrix, the normal mammary epithelial cells and metastatic breast cancer cells upregulate PDK to reduce mitochondrial oxidation to evade ROS and resist anoikis. Further, we demonstrated that either forced activation of PDH or depletion of PDK4 s ensitized these cells to anoikis in suspension culture condition. We further

PAGE 33

33 demonstrated that metastatic breast tumor cells take advantage of the Warburg effect upon detachment from matrix to resist anoikis and metastasize to distant organs. Our study imp licates PDKs as potential therapeutic targets for the breast tumor metastasis.

PAGE 34

34 Figure 1 1. Mechanisms regulating anoikis and anoikis resistance. Upon matrix detachment, the intrinsic and extrinsic apoptotic signaling pathways are activated causing cell death/anoikis. Cancer cells activate FLIP or XIAP to inhibit these pathways and resist anoikis. See text for detail description. FLIP= FLICE (FADD like IL converting enzyme) inhibitory protein, XIAP= X linked inhibitor of apoptosis protein, FADD= Fas as sociated via death domain, Bcl2= B cell lymphoma 2, Bmf= Bcl2 modifying factor, APAF 1= Apoptotic protease activating factor 1.

PAGE 35

35 Figure 1 2. Pyruvate dehydrogenase complex (PDH) and its reaction. Schematic representation of PDH structure showing its sub units E1, E2, and E3. The chemical equation of the reaction carried out by PDH enzyme is represented below. For more detailed description of this complex, see the text. E1 pyruvate dehydrogenase, E2 dihydrolipoyl transacetylase, and E3 dihydro lipoamide deh ydrogenase.

PAGE 36

36 Figure 1 3. Regulation of PDH complex. The diagram represents the negative and positive regulators of PDH complex. A detailed description is provided in the text. PDH= Pyurvate dehydrogenase complex, PDK= Pyruvate dehydrogenase kinase, PDP= Pyruvate dehydrogenase phosphatase.

PAGE 37

37 CHAPTER 2 MATERIALS AND METHODS Cell Culture The immortalized human breast epithelial cell line MCF10A and primary human mammary epithelial cells HMEC were purchased from ATCC. The Ras transformed MCF10CA1.1 cells were obtained from Barbara Ann Karmanos Cancer Institute. All these three cell lines were cultured in 50/50 mix ( DMEM/F12 ) medium (Cellgro #10 090 CV ) supplemented with 5% hor se serum (GIBCO #16050122 ), 20 ng/mL epide rmal growth factor (EGF, Sigma #E 9644 ), 10 g/mL insulin (Sigma #I 1882 ), 0.5 g/mL hydrocortisone (Sigma #H 0135), 100 g/mL streptomycin and 100 units/mL penicillin (Cell gro #30 002 CI) The transformed Human embryonic kidney (HEK) 293 cells Pheonix A and HEK293 FT cells were purchased from ATC C. The metastatic breast cancer cell line MDA MB 231 was a kind gift from Dr. Kevin Brown (University of Florida). All these cell 101 CV) supplemented with 10% bovine calf serum (BCS, HyClo g/mL streptomycin and 100 units/mL penicillin All assays in MCF10A and HMEC were performed 24 hours after incubation and in MCF10ACA1.1 and MDA MB 231 were performed 48 hours after incubation under attached and suspended conditions unless otherwise noted. Reagents Poly(2 hydroxyethyl methacrylate) (Poly Hema) was obtained from Sigma,Cat #P3932. The pharmacological ROS scavengers or antioxidants the reduced form of L

PAGE 38

38 Lipoic acid (Cat no #T5625 ) were purchased from Sigma Aldrich. Plasmids Full length human cDNAs of PDH (Cat no. # MHS1011 58712) PDK1 (Cat no. # MHS1010 7429799) 2 (Cat no. # MHS1011 61425) 3 (Cat no. # MHS1010 73978) and 4 (Cat no. # MHS1010 7429396) were obtained from OpenB iosystems. Poly Hema Coated Plates for Suspension Culture Cells The adherent cells were cultured under suspension culture conditions by coating the plates with poly mg/mL of Poly Hema (10x Stock ) was made by dissolving in 95% ethanol and rotating overnight at 65C incubator for complete solubility. The plates were coated with 1:10 dilution of 10X stock (final conc of 12 mg/mL) dissolved in 95% ethanol and air dried. For 35 mm plates, 0.5 mL of 12 mg/mL poly hema respectively was used to completely coat the plates. Once the plates were completely dry, the cells were added and grown with specific culture media for a specific time period depending on the assay performed. RNA Isolation, and Northern b lotting RNA was extracted from both attached and suspended MCF10A cells following hema coated suspended culture conditions in 35 mm plates for 24 hours incubation. For the ad herent cells, the media was aspirated and 1 mL Trizol reagent (Invitrogen #15596 026) was added to homogenize the cells. The suspended cells were collected by centrifugation at 900 rpm for 3 minutes, followed by homogenization of cells with 1 mL Trizol re agent To separate RNA from DNA and protein contents, 0.1 mL of chloroform was added to each homogenized samples. Followed by, centrifugation a t

PAGE 39

39 12,000 x g for 15 minutes at 4C to separate aqueous and phenol chloroform phases. was extracted from each sample to a new tube. The RNA was precipitated with 75% isopropanol (v/v) and centrifuged at 12,000 g for 10 minutes at 4C Later, t he RNA pellets were washed with 1 mL of 75% ethanol. After washing, the RNA pellets were air dried for 5 minutes before dissolving in sterile filtered TE (10mM Tris pH 8.0, 1mM EDTA) and stored at 80C until used The concentration of extracted RNA was measured at 260/280 nm wavelength using UV spectrophotometer Approximately 10 onto 0.8% agarose gel to separate the RNA using gel electrophoresis. After separation, the RNA from the agarose gel was transferred to nylon membrane in 10X SSC buffer (made from 20X SSC: 3 M NaCl, 0.3 M Na 3 Citrate.H 2 O, adjust the pH to 7.0 with 1N HCl) overnight at room temperature. After transfer, the membrane was hybridized with a specific probe radioactively labeled with 32 P in a hybridization buffer (Millipore, #S4031) at 42C overnight in a ro tating incubator. The probes were ~2.3 kb, ~1.5 kb, ~0.8 kb, and ~0.4 kb targeted to PDK1, PDK2, PDK3, and PDK4 respectively. They were generated by excising a part of cDNA from each gene with a suitable restriction enzyme digestion. After 24 hours of incu bation with radiolabeled probe, the membrane was washed with 100 mL of 0.2X SSC with 0.1% SDS at 55C for 5 15 minutes until the background noise was reduced. Later, the membrane was dried, wrapped in a saran wrap to perform autoradiography. Reverse Trans cription, and Real Time PCR RNA was extracted from the target cells following the same procedure as described above. g of total RNA from attached and suspended cells was added

PAGE 40

40 to a reaction mixture containing DEPC treated ddH 2 M dNTP, and 5 nM random primers to a total volume of entire mixture was incubated at 70C for 5 minutes, and qu L of 10X M MuLV RT buffer (NEB), 1 L RN L M MuLV Reverse Transcriptase (NEB) were added to the re were incubated at 42C for 1 hour Followed by heat inactivation at 65C for 20 minutes, dilution of the sample with ddH 2 O Each real cDNA generated from M primer mix L ddH 2 O, L of 2X SYBR Green PCR Master Mix (Applied Biosystems). The primers used for four PDKs were from the published report 93 All the samples were run in t riplicates for each reaction and results were expressed after normalization with endogenous beta actin expression as relative quantities Reactions with no template were also included on real time PCR plate for each set of primers as negative control. More than two fold difference in g ene expression was considered as significant. The thermal cycling parameters were : 95C for 10 minutes, 40 cycles of 95C for 15 se conds for denaturing step and 60C for 60 second s for product extension At the end of each run, melting curve analysis was performed StepOne (48 well), or StepOnePlus (96 well) real time PCR machines (Applied Biosystems) were used for data collecti on. Prime rs used were listed in Table 2 1 S hort hairpin RNA (shRNA) Vector Construction Oligos for shRNA construction were designed using shRNA psm2 designer at RNAi Central (http://cancan.cshl.edu/RNAi_central/RNAi.cgi?type=shRNA). The accession number s for h were entered to

PAGE 41

41 design specific shRNAs The target sequences for all the genes were listed in Table 2 2. To generate a high fidelity oligo, it was broken into two fragments when ord ering from Invitrogen (Table 2 2 ). The bre aks in the two fragments were designed with overlapping, complimentary loop regions so they anneal and extend during PCR into the full length oligo. The desiccated oligos were dissolved in TE buffer and combined to a final M m i r30 PCR primers were used for amplification using Phusion High Fidelity DNA Polymerase (Finnzymes). Mir30 pri mer sequences were as follows: F orward 5' AAGCCCTTTGTACACCCTAAGCCT 3' and reverse 5' ACCTGGTGAAACTCACCCAGGGATT 3'. The PCR was performed by mixing of 10 mM dNTP mix, and distilled water fol lowed by 35 cycles of 1) denatu ratio QIAquick Gel Extraction Kit (Qiag en). The purified fragment was digested with XhoI and also digested in parallel. The pLMP vector sequence and inform ation can be found in Figure 2 1 (Openbiosystems). Th C for 30 minutes. The digested fragment and vector were mixed at a 7:1 ratio, respectively and ligated with 1 L of T4 DNA ligase and 1X ligase buffer (NEB) in 10 L final volume for 1 hour overnight. The ligated DNA by heat

PAGE 42

42 mL ampicillin LB agar plates and incubat clones. Clones were manua mL Plasmid DNA MiniPrep 1 mL of the overnight culture was transferre d to a microcentrifuge tube. The bacterial cells were centrifuged at 8000 rpm for 1 minute. The cell pellets were resuspended with 200 L P1 Buffer (50 mM Tris Cl pH 8.0, 10 mM EDTA) each. The cells were then lysed by adding 200 L P2 Lysis Buffer (200mM N aOH, 1% SDS w/v) to each tube and mixed by inversion. Next, 200 L P3 Neutralization Buffer (3 M potassium acetate) were added to each tube and mixed by inversion. The samples were centrifuged at 14000 rpm for 5 minutes to pellet precipitated proteins. 0. 5 mL of the supernatant was transferred to a new tube and reserved. The pr otein pellet was discarded. 1 mL of 100% ethanol was added to the reserved supernatant and mixed to precipitate DNA. The precipitated DNA was pelleted by centrifugation at 14000 rpm f or 10 minutes. The supernatant was discarded. The DNA pellets were washed once in 1 mL 70% ethanol and the supernatants were removed. The DNA pellets were air dried for 5 to10 minutes. The DNA was dissolved in 30 to 50 L TE (1 M Tris Cl, 0.5 M EDTA pH 8) containing 20 g/mL RNase A. Retroviral Short Hairpin RNA Production and Transduction of Target Cells Purified DNA was sequenced to confirm shRNA oligo insertion prior to retrovirus production. To generate retroviral p articl 1 ) containing the specific knockdown sequence was transiently transfected into

PAGE 43

43 transformed HEK293 cells called Phoenix. The Phoenix cells were co transfected with Gag Pol a nd Env to facilitate packaging of shRNA into retroviral particles and increase their production. Cells were seeded at ~60% confluency 24 hours prior to transfection. They were transfected using Turbofect Transfection Reagent (Fermentos, #R0531) by 2 L of the transfection reagent in 2 00 L PBS. The two diluted reagents were inc ubated at room temperature for 2 0 minutes to form a cationic lipid mediated transfection complex. The complex was added directly to cells dropwise. The cells were then incubated from 12 hours to overnight before switching to fresh media. The cells were incubated for an additional 48 hours. The virus containing media was collected and passed through a 45 m filter to exclude cell debris. The viral media was aliquoted and used immediate ly to infect target cells. Excess aliquots were stored at 80 C, or disposed after bleaching. Target cells were trypsinized and plated 24 hours prior to retroviral infection. Adherent cells were infected by replacing culture media with the infection cockt ail, which 46 consisted of 1:1 viral media: culture media and 4 g/mL polybrene. The cells were incubated for 24 hours, and then the infection cocktail was replaced with fresh media. The cells were incubated for an additional 24 hours. After the 24 hours of recovery in fresh media, the cells were treated with 2 g/mL puromycin dihydrochloride (Cellgro) to begin selection of transformed cell. Uninfected target cells were treated in parallel to estimate selection completion, which is typically complete 48 ho urs after addition of puromycin. The selection media was replaced with fresh culture media after selection and transformed cells were allowed to expand to desired density. The cells were then trypsinized and dissociated into a uniform

PAGE 44

44 suspension and aliquo ted. The transformed cell stocks were stored in freezing media (bovine serum albumin containing 10% v/v DMSO) at 80 culture, or further selected for monoclonal culture. Protein I solation and Immunoblotting Cells grown under attached and suspended culture conditions were first washed in cold PBS for two times. Then, the cells were lysed with 50 L of lysis buffer (50 mM Tris pH7.5, 1 mM EDTA, 1% (v/v) SDS, 1% 2 mercaptoethanol, 20 mM dithiothreitol). The sam ples were boiled for 10 minutes to completely lyse the cells Later, 6X sample lading buffer (4X Tris SDS pH 6.8, 30% glycerol. 10% SDS, 0.6M dithiothreitol, 0.012% bromophenol blue) was added to all the samples. The samples were either stored at 20C until use or loaed onto the SDS PAGE gel for analysis. The samples were resolved by first adding the appropriate amount of 6X sample buffer and boiled for 5 minutes. The samples were loaded onto 10% Tris HCl polyacrylamide separating gels /4% stacking gel at 1mm thickness for electroporation in 1X running buffer (25 mM Tris, 190 mM glycine, 0.2% SDS). The gel was then electrotransferred onto polyvinylidene fluoride (PVDF) membrane using Trans Blot Semi Dry Electrophoretic Transfer Cell (BioRad) in 1X transfer buffer (20 mM Tris, 192 mM glycine, 10% methanol). Membranes were stained with Fast Green (0.1% Fast Green FCF, 50% methanol, 10% acetic acid) for 5 minutes at room temperature to ensure transfer and equal loading. Stained membranes were washed 2 3 times in TBST (30 mM Tris pH 7.5, 200 mM NaCl, 0.1% (v/v) Tween 20), and incubated in 3% (w/v) non fat dry milk in TBST blocking solution for one hour at room temperature on a shaker. Blocked membranes were rinsed in TBST, followed by, probing with diluted

PAGE 45

45 membranes were then washed 3 times at room temperatu re in TBST for 5 minutes each. Next, they were incubated in diluted peroxidase conjugated secondary antibodies in TBST for 30 minutes at room temperature. The membranes were then washed 3 times at room temperature in TBST for 5 minutes each. The membrane w as probed with Pierce ECL substrate solution (Thermo Scientific) to detect the bound antibodies. Followed by, autoradiography to expose the membrane to X ray film. Specific primary antibodies, secondary antibodies and their respective dilutions listed in T able 2 3. Site D irected Mutagenesis The constitutively active PDH was generated by PCR using full length PDH cDNA S300 were mutated to alanine by site directed mutagenesis using P CR with the following set of PCR primers: S232A: Forward TGGAATGGGAACA GCT GTTGAGAG 300A: Forward TACCACGGACAC GCC ATGAGTGACCCGGGAGTC GCT T RP CGTGTACGG TAAGCGACTCCCGGGTC DNA sequence coding for the mutation FLAG following set of primers: Forward GCTAGC reverse C TCGAG TTATTTATC GTCATCGTCTTTGTAGTC ACTGACTGACTTA DNA sequence are underlined and highlighted in bold and underlined respectively. The PCR reac tion was set up for a total

PAGE 46

46 2 0. The PCR cycling conditio ns were: denat uring step lowed by 35 cycles of 1) denaturing Followed by, final extension step minutes. The amplicon was clone d into a lentiviral expression vector pCSCGW2 using restriction enzyme digestion of NheI and XhoI. Lentiviral Over Expression of Constitutively Active FLAG To produce lentiviral particles, the pCSCGW2 plasmid containing consti as empty vector (EV) were transiently transfected into the HEK293FT a transformed HEK293 cell line TurboFect as transfection reagent. Along with the above each of t wo lentivir al helper plasmids ; MD2G (envelope plasmid) and PAX (packing plasmid) were co transfected to facilitate virus production and packaging. After 48 hours post transfection, the media carrying viral particles was collected and filtered through a filter t o eliminate cell debris. The viral media was harvested to either infect the target cells or stored at 80 C. MB 231 cells were plated 24 hours prior to lentiviral infection in a 60 mm petripla tes After 24 hours, the media from the cultured cells (MCF10A and MDA MB 231) was replaced with viral g/mL of polybrene was added to the viral cocktai l to increase viral transduction efficiency After 24 hours of incubation, the viral media was replaced with fresh media for the recovery of the cells. After 48 hours post infection, the cells are transferred from 60 mm to 100 mm plates for amplication. La ter, the cells were sorted

PAGE 47

47 for GFP positive cells using FACSORT instrument at ICBR Flow Cytometry Core facility, University of Florida. After sorting, one million GFP positive cells were collected from each group; EV and FPDH E1 recovery of GFP sorted cells, immunoblotting (IB) was performed using anti FLAG antibody to confirm the expression of FLAG MCF10A and MDA MB 231 cells. Measurement of PDH A ctivit y PDH activity was measured using Dipstick assay kit from MitoSciences ( # MSP30) following their protocol Cells were grown under attached and suspended culture conditions in 35 mm plates. Then the cells were trypsinized, washed twice in cold PB S and collected by centrifugation at 900 rpm for 3 minutes. After washing, the cells were lysed by adding 5 volumes of 10X sample buffer and 1/10 volume of detergent provided by the kit. The samples were incubated for 10 minutes on ice followed by centrifu gation at 3000 rpm for 10 minutes. During this step, the mitochondrial extract was isolated as supernatant from the cell debris. The extract was immediately used for protein c oncentration using BCA kit. One mg of extract from each sample was loaded onto 9 6 well plate. The equal volume of blocking solution provided by the kit was added to the wells with the sample. The dipsticks provided by the kit were gently added to the sample mix in the microplate well. The samples were allowed to wick up onto the dips ticks towards wicking pads dipstick to wash. After 5 by kit per one dipstick) was added to an empty microwell for each dipstick. Now, the

PAGE 48

48 wicking pad was removed from each dipstick and placed in a well with activity buffer. The signal appeared 5 7mm from the bottom of the dip stick within 20 minutes. After one hour, the signal was completely developed. So, the dipsticks were transferred to dipsticks were dried and the signal was measured by usin g Canon flatbed scanner. The PDH activity was represented in arbitrary units and the experiment was repeated at three different times. Trypan Blue Exclusion Assay Cells were grown under attached and suspended culture conditions in 35 mm plates. The suspe nded cells were collected by centrifugation at 900 rpm for 3 minutes. Then both the attached and suspended cells were trypsinized, washed in PBS once, f the sample mixture was added to the hemocytometer slide and the number of live (unstained) vs dead (stained blue) cells were counted. The percentage of cell viability was determined by dividing total number of unstained cells to total number of stained+u nstained cells X 100. The experiment was repeated at least for three different times in duplicates. Caspase 3/7 Activity Assay Caspase 3/7 assay w as performed using caspase 3/7 g lo assay kit (# G8090) from Promega. Cells were grown under attached and suspe nded culture conditions in 35 mm plates. The suspended cells were collected by centrifugation at 900 rpm for 3 minutes. Then both the attached and suspended cells were trypsinized, washed in PBS twice and resuspended in 3 volumes of hypertonic buffer (HTB) Cells were lysed in hypertonic buff er (HTB: 10 mM HEPES, pH7.9 at 4 C, 1.5 mM MgCl 2 10 mM KCl, add

PAGE 49

49 0.2 mM PMSF, 0.5 M DTT freshly added before use). The protein concentration of samples was measured using the BCA kit. 75 100 g of protein samples were l oaded in each of the 96 well microplates using HTB. Equal volume of caspase 3/7 substrate provided by the kit was added to all the samples in the wells. The samples were incubated for 30 minutes at room temp erature on a shaker. Later, caspase 3/7 activity was measured in relative luminiscence unit (RLU) per second using luminometer instruction. The samples were assayed in duplicates at three independent experiments. Annexin V/7 A AD Analysis Cells were grown under attached and suspended culture conditions in 35 mm plates. The suspended cells were collected by centrifugation at 900 rpm for 3 minutes. The attached were trypsinized. Both attached and suspended cells were washed twice with PBS, followed by 1x10 5 cells staining with Annexin V (Phycoerythrin) and 7 AAD (aminoactinomycin D) from BD Pharmingen for 15 min utes Facsort. Cellquest so ftware was used to analyze the data as in [FL2H, FL3H] log scale two dimensional diagram. The results were from three independent samples. Measureme nt of Oxygen Consumption R ate Oxygen consumption was measured using 96 well Oxygen biosensor plates from BD Biosciences. Cells were grown under attached and suspended culture conditions in 35 mm plates. Adherent cells were trypsinized and counted. O ne million cells from attached and suspension cultures were loaded to 96 well BD biosensor plate to measure oxyge n consumption rate for two hours at 10 min utes interval. Fluorescence

PAGE 50

50 was measured using excitation/emission ( Ex/Em ) 590/630. The assay was performed in duplicates from three independent samples. Measurement of I ntracellular ROS Intracellular ROS was measu red using the Amplex red hydrogen peroxide / peroxidase kit from Invitrogen. Cells were grown under attached and suspension conditions in 35 mm plates. Cells were trypsinized, washed twice in PBS, followed by lysis using hypertonic buff er (HTB: 10 mM HEPE S, pH7.9 at 4C, 1 .5 mM MgCl 2 10 mM KCl, add 0.2 mM PMSF fresh before use ). Equal amount of protein was loaded in 96 well plate and ROS levels were measured by adding equal volume of amplex red hydrogen peroxide/ peroxidase substrate as per protocol Fluorescence measured using Ex/EM of 590/630. The assay was performed in duplicates from three independent samples. Measurement of Intracellular L actate Intracellular lactate was measured using the l actate assay kit from Biovision (# K607 100) Cells were grown under attached and suspended culture conditions in 35 mm plates. The suspended cells were collected by centrifugation at 900 rpm for 3 minutes. Then both the attached and suspended cells were trypsinized, washed in PBS twice and resuspende d in 1 mL PBS. O ne million cells were added to 96 well microplate. The volume was adj enzyme mix provided by kit) was added to each well with sample. The samples were incubated for 30 minutes at room temperature. Later, measured lactate levels for fluorescence at Ex/Em of 590/630. The assay was performed in duplicates from three independent samples.

PAGE 51

51 Experimental Metastasis A ssay in M ice Four to five weeks old female SCID/Beige mice were purchased from Harlan. Empty vector and PDK1 depleted MDA MB 231 cells were cultured under adherent conditions. After trypsinization, two million PBS, and injected via intravenous route in a total number of six mice for each group. Forty days post injection; lungs were harvested from the mice after euthanasia. After harvesting, the lungs were washed twice in 5 mL PBS lung and incubated overnight on a shaker for fixation. The tumor nodules turned white the surface of each lung. The University of Florida IACUC approved all procedures. Tis sue Sectioning and H&E Staining Tissue sectioning, processing, and Hematoxylin and Eosin (H&E) staining were performed by the Cell and Tissue Analysis Core (CTAC) facility at University of Florida. Lung tissues were paraffin the sections were deparaffinized and rehydrated. For deparafinization, the following steps were followed; the slides with lung sections were dipped for 2 minutes in xylene. This step was repea ted three additional times followed by, two times wash in absolute ethanol for 1 minute each. Then, washed the slide twice in 95% ethanol for 30 seconds, 70% ethanol for 45 seconds and finally in water for 1 min). After deparaffinization, the sections were stained with Hematoxylin, a basic dye that stains the nuclei. The sections were washed with water and citric acid, known as clarifier. It removes the excess hemotoxylin dye from the sections. Then, the slides were treated with potassium hydroxide followed by a quick rinse in water. The sections were equilibrated in 95% ethanol. Then, the sections were stained with Eosin for 15 30 seconds. Eosin is an acid

PAGE 52

52 dye used most commonly to counterstain hematoxylin. After Eosin staining, the slides were washed three times in 100% ethanol followed by xylene wash. After staining, the sections were mounted and pictures were taken using upright light microscopy at 5X magnification. Statistics Data were represented as the mean S.D. tailed t test was used to calculate p values and p<0.05 regarded as statistically significant

PAGE 53

53 Figure 2 1 Cloning vector information for microRNA adapted retroviral vector. Cloning vector information for microRNA adapted retroviral vector. A) Vector map and unique restric tion sites of MSCV LMP cloning vector. B) Xho1 EcoR1 cloning site for shRNAmir expression using retroviral 5 PGK promoter (Ppgk) drives expression of selection cassettes. Puror cassette allows for selection of stable integrates. IRES GFP served as marker for stable integration. Abbreviations: LTR is long terminal repeats. MiR is microR NA. Ppgk is phosphoglycerate kinase promoter. Puro is puromycin. IRES is internal ribosome entry site. GFP is green fluorescence protein

PAGE 54

54 Table 2 1. Primers used for real time RT PCR Name Forward Reverse Tm PDK1 CTATGAAAATGCTAGGCGTCTG T AACCACTTGTATTGGCTGT CC 60C PDK2 AGGACACCTACGGCGATGA TGCCGATGTGTTTGGGAT G G 60C PDK3 GCCAAAGCGCCAGACAAAC CAACTGTCGCTCTCATTGA GT 60C PDK4 TTATACATACTCCACTGCACCA ATAGACTCAGAAGACAAAGC C T 60C MnSOD TACGTGAACAACCTGAAC TATCTGGGCTGTAACATCT 60C ERR TCTTGCTAATTCAGACTCCAT GCAGTGTCATCAGCATCTTG 60C Actin AGAAAATCTGGCACCACACC AGAGGCGTACAGGGATAGCA 60C Table 2 2. shRNA Oligos designed for retrovirus mediated knockdown Oligo Name Gene targeted Sequence shPDK1 PDK1 TTCTACATGAGTCGCATTTCAA shPDK4 1 P DK4 ACCAACGCCTGTGATGGATAAT shPDK4 2 PDK4 TATTTATCATCTCCAGAATTAA shMnSOD MnSOD AAGGAACAACAGGCCTTATTCC CAGTGGGAGCTACAGTTCA Table 2 3. Antibodies used for Immunoblotting (IB) Antibody Species Isotype Company Catolog number Dilution FLAG N/A Mouse IgG Sigma F 1804 1:3000 PDK1 Human Rabbit IgG Cell Signaling 3820s 1:1000 PDK2 Human Rabbit IgG Epitomics 2282 1 1:2000 PDK4 Human Rabbit IgG Abgent AP7041b 1:500 MnSOD Human Rabbit IgG Santa Cruz 30080 1:1000 LDHA Human Rabbit IgG Cell Signal ing 2012s 1:1000 ERR Human Rabbit IgG Gene Tex 108166s 1:2000 Tubulin Human Mouse IgG Sigma T 6199 1:5000 HRP Donkey anti Mouse Mouse Donkey IgG Jackson Immunoresearch 715 035150 1:5000 HRP Donkey anti Rabbit Rabbit Donkey IgG Jackson Immunoresearch 715 035152 1:8000

PAGE 55

55 CHAPTER 3 INFLUENCE OF GLUCOSE METABOLISM ON ANCHORAGE INDEPENDENT SURIVIVAL IN MAMMARY EPITHELIAL CELLS Background Apart from providing energy and biofuels for the maintenance of life, glucose metabolism also regulates cell death processes such as necrosi s, autophagy and apoptosis 94 Of these processes, the effect of glucose metabolism on apoptosis is very well studied. Several groups demonstrated that glucose deprivation leads to attenuation of anti apoptotic genes such as myelo id cell leukemia 1 protein (MCL 1), activation of BH3 only proteins such as Noxa and Bim 95 and activation of proapoptotic genes such as Bcl2 associated X protein (BAX) 96,97 Furthermore, it has been reported that NADPH, which is primarily generate d from the pentose phosphate pathway (PPP), protects cells from apoptosis by reducing ROS levels under various conditions 98 The impact of glucose metabolism on anoikis (detachment induced cell death) is still being studied. Schafer et al ., demonstrated when detached from ECM, the untransformed mammary epithelial cells reduce glucose uptak e leading to ATP deficiency kn 68 Furthermore, the attenuation of glucose uptake due to matrix detachment reduces the pentose phosphat e pathway (PPP), hence, reduces antioxidant capacity, and con sequentl y increases ROS production 68 The above study focused on antioxidant rescue from anchorage independent cell death but it is not clear how the metabolic impairment is connected to anoikis/detachment induced cell death. The main focus of our study is to understand how glucose metabolism influences anchorage independent survival in human mammary epithelial cells. Our study provides evidence that decreased glucose oxidation promotes anchorage independen t survival

PAGE 56

56 by attenuating ROS production. Furthermore, we show for the first time the functional role of PDKs in cell survival. PDKs inhibit the entry of pyruvate into TCA cycle and hence reduce glucose oxidation, promoting anoikis resistance in untransfo rmed mammary epithelial cells. Results Induction of PDK upon D etachment from ECM Pyruvate, derived from glycolytic activity, is primarily converted to acetyl CoA in the mitochondria by the pyruvate dehydrogenase (PDH) complex in normal cells. In most cance r cells, pyruvate is mostly converted into lactate by the lactate dehydrogenase (LDH). Thus, the fate of pyruvate represents a pivotal point between normal and tumor cell metabolism. Furthermore, glucose metabolism plays a critical role in regulating cell viability under variety of stressful conditions 98 As normal cells show increased sensitivity to anoikis when detach ed from matrix, we decided to investigate the influence of glucose metabolism on anoikis in normal mammary epithelial cells by examining the regulation of PDH and LDH upon cell detachment from ECM. As described in introduction, PDH is a multi subunit enzym atic complex and its kinases (PDKs) 1 4 99 We examined the expression of all four PDK isoenzymes at RNA level by quantitative RT PCR in two different mammary epithelial cell models; primary human mammary epithelial cell line (HMEC) and untran sformed human mammry epithelial cell line (MCF10A) under attached and suspended conditions. PDK4 transcript in suspended HMEC cells (Figure 3 1A), and PDK2 and PDK4 isoenzymes in suspended MCF10A cells (Figure 3 1B) were significantly upregulated, whereas PDK1

PAGE 57

57 and PDK3 showed only marginal changes (Figure 3 1). Thus, the results demonstrate that upon matrix detachment PDK4 is induced in HMEC and MCF10A cells. The rest of the study is focused on MCF10A (untransformed, immortalized mammary epithelial cell lin e). To validate if PDK4 is the major PDK isoenzyme in suspended MCF10A cells, we measured the absolute quantity of all four PDK RNAs. PDK1 and PDK4 transcripts were found to be the most abundant PDKs in suspended MCF10A cells, whereas PDK2 was negligible ( Figure 3 2A). Consistent with RNA expression, PDK4 isoenzyme was abundant at the protein level in suspended MCF10A cells and not detected under adherent conditions (Figure 3 2B). Although PDK1 was upregulated at the RNA level in suspended MCF10A cells (Fig ure 3 1B), no significant change at the protein level was observed under attached and suspended conditions (Figure 3 2B). PDK2 was not detectable at the protein level regardless of the cell matrix contact status (Figure 3 2B). Another important enzyme tha t regulates the fate of pyruvate is LDH enzyme. It interconverts pyruvate to lactate in the cytoplasm. Therefore, we decided to examine its role in matrix detachment. As mentioned in introduction, LDH5 is the most efficient isoenzyme encoded by the LDHA ge ne. Hence, we investigated the expression of LDHA in MCF10A under attached and suspended conditions. Western blotting showed no significant change in protein levels of LDHA when cells lost adhesion to ECM (Figure 3 3). Thus, the results confirm the inducti on of PDK4 when detached from matrix but no significant change in the LDH enzyme. Upregulation of PDK4 Antagonizes Anoikis To define the effect of matrix detachment induced PDK4 on anoikis, we depleted PDK4 in MCF10A cells using two independent retroviral short hairpin RNAs (shRNAs).

PAGE 58

58 In parallel, we transduced MCF10A cells with retrovirus carrying empty vector (EV) to serve as a control. The efficient knockdown of PDK4 was verified by both Northern blotting analysis (Figure 3 4A) and quantitative RT PCR (Fi gure 3 4B) in attached and suspended MCF10A cells. Thus, the two independent shRNAs targeted against PDK4 significantly diminished 80% of PDK4 expression. We further investigated if the depletion of PDK4 caused increased anoikis in MCF10A cells. Normally MCF10A cells undergo anoikis or detachment induced cell death upon detachment from matrix 15,16 Depletion of PDK4 showed a further increase of cell death in MCF10A suspension cells (based on Trypan blue exclusion assay) (Figure 3 5A). Since PDK4 was not expressed in adherent MCF10A cells, knockdown of PDK4 did not induce any significant cell death (Figure 3 5A). To validate that the increase in cell death was due to increased apoptosis, we first measured the caspas e matrix, no significant induction of caspase activity was observed in either EV or PDK4 depleted MCF10A cells (Figure 3 5B). Upon detachment, PDK4 depleted MCF10A cells displa yed higher caspase activity than the suspended EV cells (Figure 3 5B). To further validate that depletion of PDK4 enhanced apoptosis in MCF10A suspension cells, we performed another apoptosis assay i.e., PE annexin V/7 AAD analysis. Similar to caspase acti vity, no significant difference in PE annexinV/7 AAD staining was detected between the EV and PDK4 depleted MCF10A attached cells (Figure 3 5C and 3 5D). However, detachment from matrix showed an increased PE annexinV/7 AAD staining in both cell groups (Fi gure 3 5C). Compared to EV cells, PDK4 depleted cells exhibited elevated early stage (29% vs. 14%) and late stage (11% vs. 5%) apoptosis when

PAGE 59

59 detached from matrix (Figure 3 5D). Thus, the results suggest that upon detachment from matrix, mammary epithelial cells upregulate PDK4 to resist anoikis and prolong their survival in suspension. Activation of PDH Sensitizes Cells to Anoikis PDKs phosphorylate PDH enzyme and inactivate its activity. Since MCF10A cells showed a significant upregulation of PDK4 in sus pension cells, we measured the PDH activity using Mitosciences DipStick assay kit under attached and suspended culture conditions. The PDH activity was significantly downregulated in matrix detached cells compared with attached cells (Figure 3 6). The resu lts suggest the mammary epithelial cells potently activate PDK enzyme when detached from matrix to reduce glucose oxidation by decreasing PDH activity. To investigate the functional significance of reduced PDH activity in matrix detached cells, we decided to study the effect of PDH activation on anoikis. PDKs inactivate PDH by phosphorylating the 293,and 300 99 We constructed these serines with alanines, inserted a FLAG peptide at the C terminal end, and was confirmed by Western blotting using anti FLAG antibody (Figure 3 7A). To confirm displayed increased PDH activity unde r both attached and suspended conditions (Figure 3 7B). We investigated the effect of increased PDH activation on anoikis. When MCF10A

PAGE 60

60 cell death under suspended conditions (Figure 3 8A). No detectable cell death by PDH 8A). To determine if the increased cell death is due to increased apoptosis, we measured caspase 3/7 activity. As expected, the MCF10A cells transduced with caspase activity in suspension cells (Figure 3 8B). These findings suggest that increased PDH activity sensitizes MCF10A cells to anoikis. Because activation of PDH essentially phenocopied depletion of PDK4, we conclude tha t PDK4 modulates anoikis through PDH. Depletion of PDK4 Increases Mitochondrial Oxidation Pyruvate is converted irreversibly into acetyl CoA by the PDH enzyme to turn on the TCA cycle. As described in the introduction, PDH is tightly regulated by PDK and PDPs. PDKs inhibit PDH activity and attenuate the TCA cycle. Therefore, we expect that depletion of PDK4 in MCF10A would lead to increased PDH activity. An increased PDH activity would increase the TCA cycle to produce ATP or energy for other biosynthetic process. During the process of ATP production by oxidative phosphorylation, the reactive oxygen species (ROS) would be generated as a byproduct. Infact, mitochondrion is the main source for the production of ROS. Therefore, we hypothesize that depletion of PDK4 increases ROS levels in MCF10A suspension cells. To test this model, we measured PDH activity in PDK4 depleted MCF10A cells grown under attached and suspended conditions. As expected, the knockdown of PDK4 increased PDH in suspended cells (Figure 3 9 A). Since PDK4 was not expressed in adherent cells, depletion of PDK4 did not alter the PDH activity in MCF10A attached cells (Figure 3 9A). Increased PDH activity should increase mitochondrial oxidation of

PAGE 61

61 glucose and this in turn increases oxygen consump tion rate. To test this, we measured the oxygen consumption rate (OCR) in PDK4 depleted cells using BD oxygen biosensor plates. There was no significant difference in OCR between EV and PDK4 depleted MCF10A cells under adherent condition (Figure 3 9B). Upo n detachment, MCF10A carrying EV cells significantly reduced the oxygen consumption rate (Figure 3 9B), which was consistent with increased PDK4 expression (Figure 3 1B) and reduced PDH activity (Figure 3 6) during the process. The suspended PDK4 depleted MCF10A cells showed significantly higher OCR than the suspended control cells (Figure 3 9B). As the depletion of PDK increased mitochondrial respiration in MCF10A suspended cells, we investigated if the depletion of PDK4 had any impact on the ROS producti hydrogen peroxide/peroxidase assay kit. Although the MCF10A cells undergo anoikis, but no significant increase in ROS levels was observed in suspended cells compared to adherent one s (Figure 3 9C). Consistent with increased mitochondrial activity, PDK4 depleted MCF10A cells accumulated higher ROS levels compared to the control cells under suspension culture (Figure 3 9C). Thus, our results suggest that depletion of PDK4 augments mito chondrial oxidation and ROS levels specifically in matrix detached MCF10A cells. Treatment with Antioxidant Rescues PDK4 Depleted Cells from Anoikis Based on our results; we infer that depletion of PDK4 or increased PDH activity sensitizes cells to anoiki s. Furthermore, depletion of PDK4 increased ROS levels in suspended MCF10A cells. ROS are known to cause damage to all cellular components due to their oxidative nature. Since Mitochondria are the major sources for ROS production and plays a critical role in the regulation of cell death especially intrinsic

PAGE 62

62 apoptotic pathway (as described in the introduction), we decided to investigate the impact of ROS on anoikis in PDK4 depleted MCF10A cells. To test if ROS contributed to anoikis, we treated cells with tw o different antioxidants: reduced form of glutathione (GSH) and lipoic acid (LA), to scavenge the intracellular ROS. GSH is one of the most prevalent intracellular reducing agents to maintain redox homeostasis. GSH reacts with ROS to form a relatively less stable GSSG (oxidized form), which is quickly converted back to the reduced form (GSH) by endogenous glutathione reductase. Thus, the reduced form of glutathione removes ROS. lipoic acid (LA ) is another antioxidant that is readily converted into the reduced form by intracellular antioxidant enzymes such as glutat hione reductase. The reduced form has potent antioxidant effects, which scavenges ROS. The treatment with antioxidants had little or no effect on apoptosis of adherent cells, but significantly suppressed detachment induced apoptosis i n both EV and PDK4 dep leted MCF10A cells, as measured by caspase 3/7 activity (Figure 3 10 ). Although MCF10A EV cells did not show any increase in ROS levels upon detachment, the antioxidant treatment rescued the se cells from anoikis. The plausible explanation could be upon det achment from matrix, the epithelial cells may become sensitive to ROS. Taken together, our results suggest that when detached from matrix, mammary epithelial cells upregulate PDK4 to attenuate PDH activity and hence the mitochondrial oxidation of pyruvate This metabolic reprogramming reduces oxidative stress, and promotes resistance to anoikis. Estrogen Related Receptor Activates PDK4 in Response to Cell D etachment Our study indicates that PDK4 plays a critical role in matrix detachment induced metabolic shift and promotes anoikis resistance. Therefore, it is important to determine

PAGE 63

63 the key activator of PDK4 following matrix detachment. It was well established that PDK4 was a direct target of the estrogen related receptors (ERRs) 100 103 ERRs are orphan nuclear receptors and play a key role in regulating mitochondrial biogenesis and fatty acid oxidation 104 ERR have three members: ERR ERR and ERR The first two members are expressed in metabolically active tissues, whereas ERR is largely restricted to embryonic cells 104 We f irst examined the expression of ERR and ERR in MCF10A cells under attached and suspended culture conditions. The western blotting of ERR did not disp lay any significant change at protein level upon detachment (Figure 3 11A ). By contrast, ERR was significantly induced at RNA level by cell detachment (Figure 3 11B ). Since ERR was dramatically upregulated in suspended cells, we investigated its role in the induction of PDK4. To test that, we efficient ly depleted ERR with a lentiviral shRNA, as shown by qRT PCR (Figure 3 12A ). Upon depletion of ERR the induction of PDK4 wa s substantially reduced at RNA level in MCF10A suspended cells (Figure 3 12B). Next, we tested if the depletion of ERR had a simi lar effect on anoikis as PDK4 depletion in MCF10A. Consistent with PDK4 depleted MCF10A cells, depletion of ERR significantly induced cell death in suspended cells as shown by trypan blue exclusion assay (Figure 3 13A). Furthermore, this increased cell de ath was due to increased apoptosis (as shown by caspase 3/7 assay) in ERR depleted MCF10A suspended cells (Figure 3 13B). No significant cell death (Figure 3 13A) or apoptosis (Figure 3 13B) was observed in ERR depleted adherent cells. Thus, our results confirm that the induction of PDK4 in MCF10A suspension culture cells is by ERR and depletion of

PAGE 64

64 Summary Glucose metabolism not only produces energy in the form of ATP but also r egulates cell death processes. Anoikis is a form of cell death that occurs when an epithelial cell detaches from ECM. We focused on understanding how glucose metabolism was affected by matrix detachment and how it impacted anoikis or detachment induced apo ptosis. Based on our results, we conclude that upon detachment from matrix, untransformed mammary epithelial cells potently upregulate ERR expression, which in turn activates the transcription of PDK4 and, consequently results in decreased glucose oxidati on and prolonged cell survi val in suspension (Figure 3 14 ).

PAGE 65

65 Figure 3 1. Matrix detachment induces PDK4 expression in mammary epithelial cell s. HMEC and MCF10A cells were grown in a 35 mm plates under attached and poly hema coated suspended culture condi tions for 24 hrs. After 24 hrs of incubation, the cells were collected to extract RNA using Trizol reagent. 2 of RNA was used from each sample to synthesize cDNA by Moloney Murine Leukemia Virus (M run realtime quantitative PCR reaction with SYBR green PCR mix. The relative RNA levels of PDK1, PDK2, PDK3, a nd PDK4 were normalized to Actin in (A) HMEC and (B) MCF10A. These data represent triplicate experiments. Error bars were expressed as standard error of the test. Att = Attached, Susp = Suspended and PDK = Pyruvate dehydrogenase kinase isozyme 1, 2, 3, and 4.

PAGE 66

66 Figure 3 2. Induction of PDK4 at RNA and protein level in MCF10A suspension cells. MCF10A cells were grown in a 35 mm plates under attached and poly hema coate d suspended culture conditions for 24 hr s. (A ) Realtime qRT PCR. After 24 hr s of incubation, the cells were collected to extract RNA using Trizol reagent. 2 Moloney Murine Leukemia Virus (M cDNA was used to run realtime quantitative PCR reaction with SYBR green PCR mix. 100 ng of pure cDNAs of PDK1, 2, 3, and 4 resp ectively were diluted to 1, 10, 100, and1000 times and realtime quantitative PCR was run simultaneously to generate standard curve for each PDK isozyme. The absolute expression of PDK1, 2,3 and 4 under attached and suspended culture conditions in MCF10A ce lls were obtained by comparing the ct values with the respective PDK sta ndard curve. The data represent three independent experiments. Error bars were expressed as standard error of the mean (S.E.M.). Statistical analysis was performed using paired St udent test. (B ) Immunoblotting. After 24 hr s of incubation, the cells under attached and suspended culture conditions were collected and lysed in denature lysis Mercaptoethanol. The samples were analyzed by 10% SDS PAGE, followed by Immu noblotting using antibodies against PDK1, 2, 4, Tubulin as loading control Att =Attached, Susp = Suspended, and PDK = Pyruvate dehydrogenase kinase isozyme 1, 2, 3, and 4.

PAGE 67

67 Figure 3 3 Expression of LDHA in attached and suspended MCF10A cells MCF 10A cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 24 hrs. After 24 hrs of incubation, the cells under attached and suspended culture conditions were collected and lysed in denature lysis buffer with Mercaptoethanol. The samples were analyzed by 12% SDS PAGE, followed by Immunoblotting using polyclonal anti LDHA antibody and mouse anti tubulin antibody as loading control. Att = Attached, Susp = Suspended, and LDHA = Lactate dehyrogenase isoform A.

PAGE 68

68 Figure 3 4. Depletion of PDK4 using retroviral short hairpin RNA in MCF10A cells. MCF10A EV and PDK4 depleted MCF10A cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 24 hrs. After 24 hrs of incu bation, the cells were collected to extract RNA using Trizol reagent. (A was loaded onto 0.8% agarose gel, followed by 32P labeled northern blotting was performed to analyze knockdown efficiency of PDK4 using PDK4 specific probe. 18s RNA was used for loading. (B ) Realtime qRT P was used from each sample to synthesize cDNA by Moloney Murine Leukemia Virus (M run realtime quantitative PCR reaction with SYBR green PCR mix. The relative RNA levels of PDK4 were nor Actin. These data represent triplicate experiments. Error bars were expressed as standard error of the mean (S.E.M.). Statistical analysis was performed using paired test. A or Att = Attached, S or Susp = Suspended, EV = Empty vector, sh = shorthairpin RNA 1 and 2, and PDK4 = Pyruvate dehydrogenase kinase isozyme 4.

PAGE 69

69 Figure 3 5 Knockdown of PDK4 sensitizes MCF10A cells to anoikis MCF10A EV and PDK4 depleted MCF10A cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions. (A) Trypan Blue (TB) exclusion assay After 48 hrs of incubation, the cells were collected from all the samples, trypzined, washed, and resuspended in 1 mL PBS. 1:1 ratio of 0.4% trypan blue dye was added to 50 viable cells was counted by excluding dead cells stained blue using hemocytometer. (B) Caspase 3/7 glo assay. After 24 hrs of incubation, the cells were collected, washed and lyed in hypertonic buffer (HTB) to extract cytop HTB for each sample was loaded onto 96 well microplate. Equal volume of caspase 3/7 glo substrate was added to measure caspase 3/7 activity using Luminometer. (C, D) PE Annexin V/ 7 AAD a nalysis. After 24hr of incubation, the cells were collected washed twice in PBS and 1*10 5 cells were Annexin V and 7 AAD were added to the cells and incubated for 15 minutes at room temperature. Later, the cell s were immediately analyzed by FACScan machine. A total number of 10,000 cells were analyzed for each sample (C) The dot plot representing the percentage of total number of cells that were unstained or live (Lower left

PAGE 70

70 quadrant), stained positive for PE An nexin V only (Lower right quadrant), stained positive for PE Annexin V and 7 AAD (Upper right quadrant) and stained positive for 7 AAD alone (Upper left quadrant). FL2 H x axis represents PE Annexin V positive and FL3 H y axis represent 7 AAD positive cell s. (D ) Statistical representation of total percentage of apoptotic cells (percentage of PE AnnexinV only + percentage of PE AnnexinV and 7 AAD cells). All error bars represent standard deviation (n=3). Statistical analysis was performed using paired Studen test Att = Attached, Susp = S uspended, EV = Empty vector, sh = shorthairpin RNA 1 and 2, PDk4 = P yruva te dehydrogenase kinase isozyme, PE = Phycoerythrin, 7 AAD = 7 Aminoactinomycin D, and FL2H or FL3H = fluorescence emission at its highest peak re presented in logarithmic scale.

PAGE 71

71 Figure 3 6. Detachment from matrix attenuates PDH activity in MCF10A cells. PDH activity was measured using Mitosciences PDH Dipstick assay kit. MCF10A cells were grown in a 35 mm plates under attached and poly hema coate d suspended culture conditions for 24 hrs. After 24 hrs of incubation, the cells were collected, washed in PBS, and lysed by adding 5 volumes of 10X sample buffer + 1/10 volume of detergent provided by the PDH activity measured using Mitoscience dipstick a ssay kit. After measuring protein concentration using BCA kit, 1 mg of protein was loaded onto 96 well microplate to analyze PDH activity by the kit. The signal produced by PDH activity was quantitatively represented as arbitrary units using densitometry. The data represent n=3. Att = Attached, Susp = Suspended, and PDH = Pyruvate dehydrogenase complex.

PAGE 72

72 Figure 3 MCF10A cells were transduced with lentivirus carrying empty vector and FLAG EV and 10A FPDH in a 35 mm plates under attached and pol y hema coated suspended culture conditions for 24 hrs. (A) Immunoblotting. After 24 hrs of incubation, the cells under attached and suspended culture conditions were collected and lysed in Mercaptoethanol. The samples wer e analyzed by 10% SDS PAGE, followed by Immunoblotting using mouse polyclonal anti FLAG antibody (1:2000) and mouse anti tubulin antibody (1:5000) as loading control. (B ) PDH activity assay. After 24 hrs of incubation, the cells were collected, washed in P BS, and lysed by adding 5 volumes of 10X sample buffer + 1/10 volume of detergent provided by the PDH activity measured using Mitoscience dipstick assay kit. After measuring protein concentration using BCA kit, 1 mg of protein was loaded onto 96 well micro plate to analyze PDH activity by the kit. Att = Attached, Susp = of pyruvate dehydrogenase complex.

PAGE 73

73 Figure 3 8. Activation of PDH sensitizes MCF10A cells to anoikis. MCF10A EV and 10A FPDH attached and poly hema coated suspended culture conditions for 24 hrs. A) Trypan Blue (TB) exclusion assay After 24 hrs of incubation, the cells were collected from all the samples, trypzined, washed, and r esuspended in 1 mL PBS. 1:1 ratio of viable cells was counted by excluding dead cells stained blue using hemocytometer. (B) Caspase 3/7 glo assay. After 24 hrs of incubation, the cells we re collected, washed and lyed in hypertonic buffer (HTB) to extract HTB for each sample was loaded onto 96 well microplate. Equal volume of caspase 3/7 glo substrate was added to mea sure caspase 3/7 activity using Luminometer. All error bars represent standard deviation (n=3). Att =

PAGE 74

74 Figure 3 9. Depletion of PDK4 increases mitochondrial oxidation in MCF10A suspended cells. MCF10A EV and PDK4 depleted MCF10A cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 24 hrs. (A) PDH activity assay. After 24hr of incubation the cells were collected, washed in PBS, and lysed by adding 5 volumes of 10X sample buffer + 1/10 volume of detergent provided by the PDH activity measured using Mitoscience dipstick assay kit. After measuring protein concentration using BCA kit, 1 mg o f protein was loaded onto 96 well microplate to analyze PDH activity by the kit. The signal produced by PDH activity was quantitatively represented as arbitrary units using densitometry. (B) Measurement of oxygen consumption rate (OCR). After 24 hrs of inc ubation, one million cells were loaded onto the 96 well BD Oxygen biosensor plates to measure oxygen consumption for 2hrs at 10 min interval. (C) Measurement of ROS by Amplex red hydrogen peroxide/peroxidase assay. After 24 hrs incubation, the cells were c ollected, washed in PBS, and lysed the cells with hypertonic buffer. Equal amount of protein was loaded onto 96 well microplate and Amplex red hydrogen peroxide/peroxidase substrate was added to analyze the ROS levels at fluorescence wavelength of 590/630 nm

PAGE 75

75 (excitation/emission). All error bars represent standard deviation (n=3). Att = Attached, Susp = S uspended, EV = Empty vector, sh = short hairpin RNA, PDK4 = P yruva te dehydrogenase kinase isozyme, PDH= Pyruvate dehydrogenase complex, OCR= Oxygen consump tion rate, ROS = Reactive oxygen species, and FLU= Fluoroscence units. Figure 3 10. Antioxidant treatment rescues PDK4 depleted MCF10A cells from anoikis. Caspase 3/7 glo assay. MCF10A EV and PDK4 depleted MCF10A cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 48 hrs. The cells were treated with (i) untreated represented as (ii) 2.5 mM L glutathione represented as GSH, and (iii) 100 Lipoic acid represented as LA for 48 hrs. After 48 hrs of incubation, the cells were collected, washed and lyed in hypertonic buffer (HTB) to extract HTB for each sample was loa ded onto 96 well microplate. Equal volume of caspase 3/7 glo substrate was added to measure caspase 3/7 activity using Luminometer. All error bars represent standard deviation (n=3). Att = Attached, Susp = Suspended, EV = Empty, sh = short hairpin RNA, PDK 4 = Pyruvate dehydrogenase kinase isozyme 4, ( ) = untreated, GSH = L glutathione, and Lipoic acid.

PAGE 76

76 Figure 3 MCF10A cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 24 hrs. (A) Immunoblotting. After 24 hrs of incubati on, the cells under attached and suspended culture conditions were Mercaptoethanol. The samples were analyzed by 10% SDS PAGE, followed by Immunoblotting using polyclonal anti e anti tubulin antibody as loading control. (B) Realtime qRT PCR. After 24 hrs of incubation, the cells from each sample to synthesize cDNA by Moloney Murine Leukemia Virus (M MulV) r quantitative PCR reaction with SYBR green PCR mix. The relative RNA Actin. All error bars represent standard deviation (n=3) Att = Attached, Susp = Suspended, ERRa = Estrogen related receptor isoform alpha, and ERRg = Estrogen related receptor isoform gamma

PAGE 77

77 Figure 3 12. Depletion of ERR PDK4 induction in MCF10A suspended cells. MCF10A cells were transduced with a retroviral empty vecto r ("EV") or shRNA targeting ERR ("shERRg"). Cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 24 hrs and total RNA was extracted and anal yzed by quantitative RT PCR. (A) Depletion of ERR in MCF10A (B ) Attenuation of PDK4 induction following cell detachment in ERR depleted MCF10A. All error bars represent standard deviation (n=3) Att = Attached, Susp = Suspended, ERRg = Estrogen related receptor isoform gamma, and PDK4 = Pyruvate dehydrogenase kinase sh = shorthairpin RNA.

PAGE 78

78 Figure 3 MCF10A cells were transduced with a retroviral empty vector ("EV") or shRNA targeting ERR ("shERRg"). Cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 24 hrs. (A) Trypan Blue (TB) exclusion assay. After 24hr of incubation, the cells were collected from all the samples, trypzined, washed, and resuspended in 1 mL PBS. 1:1 ratio of 0.4% trypan blue dye was The percentage of viable cells was counted by excluding dead cells stained blue using hemocytometer. (B) Caspase 3/7 glo assay. After 24hr of incubation, the cells were collected, washed and lyed in hypertonic buffer (HTB) to ext well microplate. Equal volume of caspase 3/7 glo substrate was added to measure caspase 3/7 activity using Luminometer. All error bars rep resent standard deviation (n=3). Att = Attached, Susp = Suspended, ERRg = Estrogen related receptor isoform gamma, and sh = shorthairpin RNA.

PAGE 79

79 Figure 3 14. Summary model. Upon detachment from ECM, the MCF10A cells ate mitochondrial oxidation in order to prevent ROS production and thereby resist anoikis.

PAGE 80

80 CHAPTER 4 PDK PROMOTES ANOIKIS RESISTANCE AND TUMOR METASTASIS IN BREAST CANCER CELLS Background Tumor cells are formed due to genetic alterations in signaling path ways leading to abnormal cell growth and survival. The two major signaling pathways that are often found mutated in different tumors are Ras extracellular signal regulated kinase (ERK) and the Phosphatidyllinositol 3 OH kinase (PI(3)K) AKT pathways 29 These aberrant mutations cause inappropriate cell growth and survival. The rapidly dividing and proliferating tumor cells eventually outgrow their vascular supply, resulting in a hypoxic environment (low oxygen levels). The major consequence of hypoxia is to activate different hypoxia inducible genes, which include glycolysis related genes such as Hexokinase II, Glut1 receptors, LDHA, PDK1, PDK3, pH regulated genes such as Carbonic anhydr ase 9 and CA12, and angiogenic genes such as vascular endothelial growth factor (VEGF) and angiopoietin 2 105 All these genes help tumor growth and invasion. Furthermore, the hypoxic environment favors tumor metastasis by activating epithelial to mesenchymal transition (EMT), regulating integrin interactions to facilitate tissue rupture and invasion, and upregulating vascular endothelial growth factor (VEGF) for angiogenesis and vascular formation 106 Cancer cells possess both altered energy metabolism and resistance to anoikis but the connection between them is yet to be elucidated Schafer et al., were the first to demonstrate the potential link between glucose metabolism and anoiki s in untransformed mammary epithelial cells 68 Their report showed that detachment from ECM led to metabolic defect in MCF10A cells. Forced overexpression of oncoprotein ErbB2/HER2 rescued these cel ls from oxidative stress, restored ATP generation, and

PAGE 81

81 supported anchorage independent cell survival However, it is less clear how ATP deficiency is linked to cell death. In the current study we show that breast tumor cells have attenuated oxidative meta bolism of glucose compared to the normal cells. Furthermore, detachment from ECM in cancer cells exhibits a dramatic shift from mitochondrial respiration towards glycolysis by upregulating PDKs In other words, the tumor cells show a further increase in ae robic glycolysis to survive and metastasize to distant organs. Activation of mitochondrial oxidation either by depletion of PDK or forced over expre ssion of PDH cells to anoikis and abrogates their metastatic potential These observations indicate that the Warburg effect in tumor cells promotes anoikis resistance and contributes to the malignant phenotypes. Results Induction of PDK4 Promotes Anoikis Resistance in RAS Transformed Mammary Epithelial Cells Induction of PDK4 in Ras transformed MCF10A cells (10ACA1.1) untransformed mammary epithelial cells (Cha pter 3). Next, we investigated if a similar survival mechanism existed in Ras transformed MCF10A derivative cell line (MCF10A CA1.1). These cells not only harbor the activated oncogene Ras but also certain unindentified spontaneous mutations which led to generation of a highly tumorogenic cell line 107 Similar to MCF10A, MCF10CA1.1 cells upregulated PDK4 at RNA level un der the detached condition as shown by northern blotting (Figure 4 1). No de tectable expression of PDK4 was observed under attached culture condition (Figure 4 1).

PAGE 82

82 PDK4 resists anoikis in Ras transformed MCF10A cells (MCF10ACA1.1) To test the functional significance of PDK4 in MCF10ACA1.1 PDK4 was depleted in MCF10CA1.1 cells wi th two independent retroviral shRNAs and their knockdown efficiency was confirmed by northern blotting. The depletion of PDK4 in MCF10ACA1.1 was approximately 80% by shPDK4 1 and ~90% by shPDK4 2 when compared to the empty vector (EV) under detached condit ion (Figure 4 1). Unlike untransformed MCF10A cells, MCF10CA1.1 transduced with EV did not exhibit significant cell death when placed in suspension; however, depletion of PDK4 sens itized them to cell death (Figure 4 2A ) The significant decrease in cell vi ability in PDK4 depleted MCF10ACA1.1 detached cells was due to increased anoikis as shown by caspase 3/7activity (Figure 4 2B) Collectively, our results indicate that matrix detachment leads to induction of PDK4 to promote anoikis resistance in both untr ansformed and malignant mammary epithelial cells. Warburg Effect Promotes Anoikis Resistance and Metastasis in Breast Tumor Cells Matrix detachment favors Warburg effect in MDA MB 231 cells The hallmark of cancer cells is anoikis resistance. Since most ca ncer cells prefer aerobic glycolysis to mitochondrial respiration, we hypothesized that reduced mitochondrial oxidation provides survival advantage to malignant tumor cells when detached from matrix. To test our hypothesis we investigated the impact of W arburg effect on anoikis resistance in an aggressive, metastatic, and highly glycolytic triple n egative breast cancer cell line MDA MB 231 108 First, we compared the PDH activity in MDA MB 231 cells cultured under attached and suspended conditions. As expected,

PAGE 83

83 the MDA MB 231 cells displayed lower PDH activity under attached conditions (Figure 4 3) compared to MCF10A (Figure 1 7); however, the suspended MDA MB 231 cells demonstrated a significant decrease in PDH activity compared to attached cells (Figure 4 3 ). Thus, our data suggest that upon detachment from matrix, breast tumor cells demonstrate a further reduction in mitochondrial activity. Forced activation of mitochondrial oxidation induces anoikis in MDA MB 231 We investigated the significance of reduced mi tochondrial activity in MDA MB 231 under suspension culture conditions. As performed in MCF10A, we ectopic ally expres sed the constitutively active form of PDH with c terminal FLAG tag in MDA MB 231 using lentiviral transduction method. Meanwhile, we in fected MDA MB 231 with lentivirus carrying empty vector (EV) as control. Overexpression of PDH was confirmed by western blotting (Figure 4 4A). To test if the ectopically expressed PDH was enzymatically active, we measured PDH activity in these ce lls. The MDA MB 231 cells expressing constitutively active form of PDH showed significantly increased PDH activity compared to MDA MB 231 cells expressing EV (Figure 4 4B). Next, we investigated the impact of increased mitochondrial activity on anoik is in MDA MB 231 cells. Since cancer cells are resistant to detachment induced cell death, MDA MB 231 cells transduced with EV did not display significant cell death under detached culture conditions (Figure 4 5A); however, forced increase of PDH activity sensitized MDA MB 231 cells to cell death only under detached culture conditions (Figures 4 5A). The increased anoikis was attributed to increased apoptosis in MDA MB 231 cells expressing constitutively active form of PDH (Figure 4 5B). T hese results suggest that breast tumor cells significantly attenuate mitochondrial oxidation to survive longer after detachment from matrix.

PAGE 84

84 PDK1 enhances Warburg effect in matrix detached MDA MB 231 cells To this point, our results s uggest that breast tumor cells display a significant decrease of mitochondrial oxidation upon matrix detachment to survive longer. To test if the PDK isozymes regulate the mitochondrial activity under detached condition as observed in MCF10A and RAS transf ormed MCF10A cells, we examined the expression of all four PDKs under attached and suspended culture conditions. PDK4 RNA was significantly upregulated in MDA MB 231 cells upon matrix detachment (Figure 4 6A) but its over all abundance was trivial when com pared with other PDKs (Figure 4 6B ). PDK1 was the predominant PDK isoenzyme in MDA MB 231 cells both under attached and detached culture conditions as identified by quantitative RT PCR (Figure 4 6B). Although the RNA level of PDK1 in MDA MB 231 cells did n ot change upon detachment from matrix, its protein level was significantly increased compared to attached cells (Figure 4 6C). These data indicate that MDA MB 231 cells upregulate PDK1 to downregulate mitochondrial oxidation in suspension. Therefore, to a ctivate mitochondrial oxidation in MDA MB 231 cells, we depleted the predominant PDK isoenzyme i.e., PDK1 using two independent retroviral short hairpin RNAs. The knockdown efficiency of PDK1 was confirmed by western blotting (Figure 4 7A). The depletion o f PDK1 led to increased PDH activity in MDA MB 231 cells both under attached and detached culture conditions (Figure 4 7B), indicating a glucose metabolic shift toward mitochondrial oxidation from glycolysis. To test that, we measured intracellular lactate levels in PDK1 depleted MDA MB 231 cells. As expected, MDA MB 231 cells carrying EV produced increased lactate levels in suspended cells compared to attached ones (Figure 4 7C); however, depletion of PDK1 significantly decreased lactate production both un der attached and detached

PAGE 85

85 culture conditions (Figure 4 7C). Since we observed a significant decrease in lactate levels and increase in PDH activity, we measured the oxygen consumption rate (OCR) in PDK1 depleted MDA MB 231 cells to confirm an increase in m itochondrial respiration. Consistent with our observations, MDA MB 231 cells displayed decreased OCR in detached cells compared to attached cells (Figure 4 7D), but knockdown of PDK1 increased O 2 consumption rate in these cells both under attached and susp ended culture conditions (Figure 4 7D ). These results suggest that matrix detachment further enhances the glycolytic phenotype in MDA MB 231 cells and depletion of PDK1 in these cells reverses the Warburg effect shifting glucose metabolism from glycolysi s towards mitochondrial oxidation. Depletion of PDK1 activates mitochondrial oxidation and abrogates anoikis resistance in MDA MB 231 cells Depletion of PDK1 significantly reversed the Warburg effect in MDA MB 231 cells both under attached and detached c ulture conditions; therefore, we investigated the potential effect of this metabolic shift on anoikis. Not surprisingly, MDA MB 231 cells expressing EV did not show any significant cell death (Figure 4 8A) or apoptosis (Figure 4 8B) when grown in suspensio n Although depletion of PDK1 increased PDH activity and reversed the Warburg effect in MDA MB 231 cells under attached culture condition, it did not induce cell death (Figure 4 8A) or apoptosis (Figure 4 8B) Only when MDA MB 231 cells were detached from matrix did depletion of PDK1 induce cell death (Figure 4 8A) and apoptosis (Figure 4 8B) Together, these results suggest that normalization of glucose metabolism by depletion of PDK or activation of PDH is capable of restoring anoikis in breast cancer ce lls.

PAGE 86

86 PDK1 enhances breast tumor metastasis in vivo Resistance to anoikis is one of the critical steps for tumor metastasis. Depletion of PDK1 sensitized the highly malignant MDA MB 231 cells to anoikis by reversing the Warburg effect in vitro therefore, w e examined the physiological relevance of this metabolic manipulation on cancer metastasis. A common experimental model to analyze tumor metastasis in vivo is tail vein injectio n experimental metastasis assay. It measures the ability of cancer cells to sur vive in the blood circulation, extravasate, and form metastatic colonies at distant secondary sites. MDA MB 231 cells carrying EV and PDK1 depleted MDA MB 231 cells were cultured under adherent conditions, then trypsinized, and equal number of cells (1*10 6 cells/200 intravenously injected into immunodeficient mice (Six mice in each group) After 40 days post injection control MDA MB 231 cells gave rise to about 100 tumor nodules per lu ng in all 6 mice analyzed (Figure 4 9A ). Consistent with o ur in vitro observations, the mice injected with PDK1 depleted MDA MB 231 cells produced significantly less number of lu ng tumor nodules (Figure 4 9A ). The immunohistochemical staining of Hemotoxylin and Eosin (H&E) did not show any significant morphologic al differences between control and PDK1 depleted lung tumor tissues (Figure 4 9A). There was a five fold decrease in the lung tumor nodule formation in mice injected with PDK1 depleted MDA MB 231 cells compared to control mice (Figure 4 9B). Therefore, rev ersal of the Warburg effect by depletion of PDK1 not only sensitized metastatic cancer cells to anoikis in vitro but also profoundly decreased their metastatic potential in vivo

PAGE 87

87 Summary Tumor cells exhibit aerobic glycolysis / Warburg effect and resistan ce to anoikis to survive and thrive the harsh environments. It has been well established that Warburg effect helps tumor growth and survival. However, its role in promoting anoikis resistance and tumor metastasis glucose is still not clear. Therefore, we f ocused our study understanding the role of altered glucose metabolism in regulating anoikis resistance to promote metastasis in vivo Based on our results, we conclude that upon detachment from ECM, tumor cells take advantage of their altered glucose metab olism i.e., Warburg effect by upregulating the PDKs to survive and invade the distant organs. Furthermore, depletion of PDK or forced activation of PDH increases mitochondrial respiration and, thus sensitizes these tumor cells to anoikis and abrogates thei r metastatic potential in vivo. Therefore, our study implicates that the PDK enzymes are potential therapeutic drug targets to prevent both tumor growth and metastasis.

PAGE 88

88 Figure 4 1. Detachment from matrix upregulates PDK4 in Ras transformed MCF10A cells (MCF10ACA1.1). Northern blotting. 10ACA1.1 EV and PDK4 depleted 10ACA1.1 cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 48 hrs. After 48 hr s of incubation, the cells were collected to extract RNA us ing Trizol reagent. each sample was loaded onto 0.8% agarose gel, followed by 32 P labeled northern blotting was performed to analyze knockdown efficiency of PDK4 using PDK4 specific probe. 18s RNA was used for loading. A = Attached, S = S uspended, EV= Empty vector, sh = Shorthairpin, and PDK4 = Pyruvate dehydrogenase kinase isozyme 4.

PAGE 89

89 Figure 4 2. Depletion of PDK4 sensitizes MCF10ACA1.1 cells to anoikis. Activation of PDH sensitizes MCF10A cells to anoikis. 10ACA1.1 EV and PDK4 deplete d 10ACA1.1 cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 48 hrs. (A) Trypan Blue (TB) exclusion assay After 48hr of incubation, the cells were collected from all the samples, trypzined, washed, and resuspended in 1 mL PBS. 1:1 ratio of 0.4% was counted by excluding dead cells stained blue using hemocytometer. (B) Caspase 3/7 glo assay. Only 10ACA1.1 shPDK4 2 was used for thi s assay. After 48 hrs of incubation, the cells were collected, washed and lyed in well microplate. Equal vol ume of caspase 3/7 glo substrate was added to measure caspase 3/7 activity using Luminometer. All error bars represent standard deviation (n=3). Att = Attached, Susp = Suspended, EV = Empty vector, sh =Shorthairpin, and PDK4=Pyruvate dehydrogenase kinase i sozyme 4.

PAGE 90

90 Figure 4 3. Matrix detachment attenuates PDH activity in MDA MB 231cells. PDH activity assay. MDA MB 231 cells were grown in a 35 mm plates under attached and poly hema coated suspended culture condition for 48 hrs. After 48 hrs incubation, th e cells were collected, washed in PBS, and lysed by adding 5 volumes of 10X sample buffer + 1/10 volume of detergent provided by the PDH activity measured using Mitoscience dipstick assay kit. After measuring protein concentration using BCA kit, 1 mg of pr otein was loaded onto 96 well microplate to analyze PDH activity by the kit. Att = Attached, Susp = Suspended, and PDH = Pyruvate dehydrogenase complex.

PAGE 91

91 Figure 4 MB 231 cells. MDA MB 231 cells were transduced with lentivirus carrying empty vector and FLAG EV and MDA231 FPDH plates under attached and poly hema coated suspended culture conditions for 48 hrs. After 48 hrs of incubation, the cells under attached and suspended culture conditions were Mercaptoethanol. The s amples were analyzed by 10% SDS PAGE, followed by Immunoblotting using mouse polyclonal anti FLAG antibody (1:2000) and mouse anti tubulin antibody (1:5000) as loading control. (B ) PDH activity assay. MDA231 EV and MDA231 FPDH mm plates under attached culture condition for 48 hrs. After 48 hrs of incubation, the cells were collected, washed in PBS, and lysed by adding 5 volumes of 10X sample buffer + 1/10 volume of detergent provided by the PDH activity measured using Mitoscienc e dipstick assay kit. After measuring protein concentration using BCA kit, 1 mg of protein was loaded onto 96 well microplate to analyze PDH activity by the kit. Att = Attached, Susp = Suspended, EV = Empty pyruvate dehydrogenase complex.

PAGE 92

92 Figure 4 5. Forced activation of PDH induces anoikis in MDA MB 231 cells. MDA231 EV and MDA231 attached and poly hema coated suspended culture conditions for 48 hrs. (A ) Trypan Blue (TB) exclusion assay After 48 hrs of incubation, the cells were collected from all the samples, trypzined, washed, and resuspended in 1 mL percentage of viable cells was c ounted by excluding dead cells stained blue using hemocytometer. (B) Caspase 3/7 glo assay. After 48 hrs of incubation, the cells were collected, washed and lyed in hypertonic buffer (HTB) to by HTB for each sample was loaded onto 96 well microplate. Equal volume of caspase 3/7 glo substrate was added to measure caspase 3/7 activity using Luminometer. All error bars represent standard deviation (n=3). Att = Attached, Susp = Suspe

PAGE 93

93 Figure 4 6. Induction of PDK 1 at protein level in MDA MB231 upon detachment from matrix. MDA MB 231 cells were grown in a 35 mm plates under attached and p oly hema coated suspended culture conditions for 48 hrs. (A, B) After 48 hrs of incubation, the cells were collected to extract RNA using Trizol reagent. 2 Murine Leukemia Virus (M MulV) rev used to run realtime quantitative PCR reaction with SYBR green PCR mix. (A) Relative PDK expression. The relative RNA levels of PDK1, PDK2, PDK3, Actin. (B) Absolute PDK express ion. 100 ng of pure cDNAs of PDK1, 2, 3, and 4 respectively were diluted to 1, 10, 100, and1000 times and realtime quantitative PCR was run simultaneously to generate standard curve for each PDK isozyme. The absolute expression of PDK1, 2,3 and 4 under att ached and suspended culture conditions in MDA MB 231 cells were obtained by comparing the ct values with the respective PDK standard curve. (C ) Immunoblotting. After 48 hr s of incubation, the cells under attached and suspended culture conditions were colle Mercaptoethanol. The samples were analyzed by 10% SDS PAGE, followed by Immu noblotting using rabbit polyclonal anti PDK1 (1:1000) and mouse anti Tubulin antibody (1:5000) as loading control The data r epresent three independent experiments. Error bars were expressed as standard error of the

PAGE 94

94 test. Att = Attached, Susp = Suspended, and PDK = Pyruvate dehydrogenase kinase isozyme 1, 2, 3, and 4.

PAGE 95

95 Figure 4 7. Depletion of PDK1 switches glucose metabolism towards mitochondrial oxidation in MDA MB 231 cells. MDA231 EV and MDA231 shPDK1 cells were grown in 35 mm plates under attached and poly hema coated suspended culture conditions f or 48 hrs. (A) Immunoblotting. After 48 hr s of incubation, the cells under attached and suspended culture conditions were Mercaptoethanol. The samples were analyzed by 10% SDS PAGE, followed by Immu noblotting using rabbit polyclonal anti PDK1 (1:1000) and mouse anti Tubulin antibody (1:5000) as loading control (B) PDH activity assay. After 48 hrs of incubation, the cells were collected, washed in PBS, and lysed by adding 5 volumes of 10X sample bu ffer + 1/10 volume of detergent provided by the PDH activity measured using Mitoscience dipstick assay kit. After measuring protein concentration using BCA kit, 1 mg of protein was loaded onto 96 well microplate to analyze PDH activity by the kit. (C) Lact ate assay. After 48 hrs of incubation, o ne million cells were added to 96 well microplate. Equal volume of reaction mix provided by kit was added to each sample to measure lactate levels at Ex/EM 0f 590/630 nm. (D) Measurement of oxygen consumption rate (O CR). After 48 hrs of incubation, one million cells were loaded onto the 96 well BD Oxygen biosensor plates to measure oxygen consumption for 2hrs at 10 minutes interval. The data represent three independent experiments. Error bars were expressed as standar d error of the test. Att = Attached, Susp = Suspended, EV =Empty vector, PDK1 = Pyruvate dehydrogenase kinase 1.

PAGE 96

96 Figure 4 8. Depletion of PDK1 abolishes anoikis resistance in MDA MB 231 cells. MDA231 EV and MDA231 shPDK1 cells were grown in 35 mm plates under attached and poly hema coated suspended culture conditions for 48 hrs. (A) Trypan Blue (TB) exclusion assay After 48 hrs of incubation, the cells were collected from all the samples, trypzined, washed, and resuspended in 1 mL PBS. 1:1 ratio of 0.4% trypan blue dye was added percentage of viable cells was counted by excluding dead cells stained blue using hemocytometer. (B) Caspase 3/7 glo assay. After 48 hrs of incubation, the cells were collected, washed and lyed in hypertonic buffer (HTB) to extract by HTB for each sample was loaded onto 96 well microplate. Equal volume of caspase 3/7 glo substrate was added to measure caspase 3/7 activity using Luminometer. All error bars represe nt standard deviation (n=3). Att = Attached, Susp = Suspended, EV= Empty vector, and PDK = Pyruvate dehydrogenase kinase.

PAGE 97

97 Figure 4 9. Depletion of PDK1 abrogates tumor metastasis in vivo MDA231 EV and MDA231 shPDK1 cells were grown in 100 mm plates und er attached condition. The cells were trypsinized and washed twice in PBS. Two million 5 weeks old nude mice for each group (231 EV and 231 shPDK1). Forty days post injection, the lungs from the injected mice were harvested. (A) The entire lung picture with tumor nodules and H&E staning. For counting the nodules, was fixed in 4% paraformaldehyde (PFA) solu tion for 24 hrs at room temperature. After 24 hrs, the lung is stored in 70% ethanol until tissue processing. The tissue processing, sectioning, paraffin embedding, and H&E staining procedure was performed by University of Florida Core tissue sectioning fa cility. Detailed description of H&E staining is in the text. Upper panel lung from the mouse injected with MDA231 EV and lower panel lung from the mouse injected with 231 shPDK1. (B) Statistical analysis. The lung tumor nodules for all six mice in each g roup are statistically represented. Statistical analysis was performe test. EV = Empty vector, sh =short hairpin, and PDK1 = Pyruvate dehydrogenase kinase 1.

PAGE 98

98 CHAPTER 5 ANTIOXIDANT PROTECTI ON FROM DETACHMENT I NDUCED OXIDATIVE STR E SS IN MAMMARY EPITHELIAL CELLS Background Endogenous Antioxidants Defense System The endogenous ROS levels are tightly regulated by both non enzymatic and enzymatic antioxidant components. The non enzymatic antioxidants include hydrophilic and lipophilic r tocopherol, ascorbic acid, reduced coenzyme Q10 and glutathione 109 The enzymatic antioxidant defense system includes superoxide dismutases (SOD), which dismutase highly reactive supero xide anion to less reactive hydrogen peroxide. There are three different forms of SOD enzyme localized at different regions; SOD1 or copper superoxide dismutase (CuSOD) in cytosol 110 SOD2 or manganese superoxide dismutase (MnSOD) in mitochondrion 111 and SOD3 in extracellular matrix region 112 Hydrogen peroxide is maintained at low levels by catalase and glutathione peroxidase (Gpx) enzymes, which convert H 2 O 2 to H 2 O and O 2 113 The other enzymes that are involved in the reduction of oxidized forms of antioxidant molecules are; thioredoxin reductase, glutathione reductases and peroxiredoxin 113 Therefore, the endo genous antioxidants balance the ROS levels to prevent damage from oxidative stress and maintain redox homeostasis. Manganese Superoxide Dismutase (MnSOD) MnSOD is one of the major antioxidants that scavenge the superoxide anion to generate hydrogen peroxid e. It is a tetrameric enzyme localized to mitochondrion encoded by a nuclear gene located on chromosome 6q25. MnSOD is an antioxidant that resists apoptosis and protects cells from oxidative stress 114 inflammatory cytokines and ionizing radiation 115 The mice lacking MnSOD enzyme die immediately after birth,

PAGE 99

99 thus, illustrating its importance in maintenance of life 116,117 The mice carrying heterozygous MnSOD gene display reduction in MnSOD activity, increase in mitochondrial oxidative damage, and higher incidence of tumor in older mice 118 Several reports state that MnSOD acts as a tumor suppressor gene 119,120 High levels of MnSOD have been reported to increase tumor metastasis and invasion 121,122 MnSOD is highly expressed in estrogen receptor (ER) negative human breast cancer lines (M DA MB 231, SKBR3) compared to ER dependent human breast cancer cell lines (MCF7, T47D) 123 The role of MnSOD in tumor growth, invasion and metastasis is controversial, but it is an important antioxidant that prevents oxidative stress and inhibit s apoptosis. Good ROS and Bad ROS The physiological role of ROS is very dynamic and depends on its overall production. Mitochondrion generates ROS in a continuous process under aerobic conditions. Low or transient levels of ROS is beneficial for normal ce lls because it stimulates cellular proliferation and survival signaling pathways 124 ROS oxidizes the cysteine residues of kinases and phosphotases and thus, activates the signaling molecules inv olved in proliferation, survival and growth. Some of those molecules may include apoptosis signal regulating kinase (ASK1) 125 ERK 126 PI3K/AKT 127 and HIF1 128 etc. Excessive production of ROS causes damage to all cellular components leading to cell death or senescence. The oxidative nat ure of ROS leads t o alterations in mitochondrial membrane components such as lipid peroxidation, or protein thiol oxidation, which results in mitochondrial membrane permeabilization 129 This is followed by release of apoptosis inducing factors including cytochrome c from

PAGE 100

100 mitochondria and subsequent activation of the caspase cascade 130 Furthermore, increased ROS production oxidizes cytochrome c leading to pro apoptotic activity 130 Under phy siological conditions, oxidants and anti oxidants are in harmony to maintain redox homeostasis. Any perturbation that causes imbalance between these two components leads to oxidative stress. Matrix detachment causes an increase in the level of ROS in e ndo thelial cells 67 and mammary epithe lial cells 68 Our data indicate that upon detachme nt, mammary epithelial cells modulate glucose metabolism by upregulating PDK4 to attenuate mitochondrial respiration to evade ROS production and resist anoikis (Fig ure 1 15). Furthermore, treatment with antioxidants not only rescued PDK4 depleted MCF10A ce lls but also the control MCF10A cells from apoptosis under detached culture conditions (Fig ure 1 11), suggesting matrix detachment renders mammary epithelial cells sensitive to ROS. To validate our observations, we examine d the consequence of increase d oxi dative stress resulting from depletion of an endogenous antioxidant. Since our work mainly focuses on mitochondrial metabolism and ROS generated from it, we studied the impact of mitochondria antioxidant manganese superoxide dismutase (MnSOD or SOD2) on an oikis o r detachment induced apoptosis Our work demonstrates for the first time that when detached from ECM, the mammary epithelial cells also modulate their endogenous antioxidant MnSOD levels to reduce oxidative stress and promote anoikis resistance. Re sults Induction of MnSOD in Matrix Detached Mammary Epithelial C ells MnSOD catalyzes the dismutation of mitochondrial superoxide into hydrogen peroxide and oxygen, and hence is a key anti oxidant. Also, our data suggest that matrix

PAGE 101

101 detachment in mammary epi thelial cells modulate glucose metabolism to evade ROS production from mit ochondria. Therefore, we first examined the induction of MnSOD in HMEC and MCF10A at RNA level under attached and suspended culture conditions. We unexpectedly found that like PDK4, MnSOD RNA levels were significantly increased following detach ment of HMEC (Figure 5 1A) and MCF10A cells (Figure 5 1B). Consistent with RNA level, t he protein level of MnSOD was also marked ly elevated in MCF10A suspended culture cells (Figure 5 2 ). Thus, the data indicate upon detachment from extracellular matrix, mammary epithelial cells induce antioxidant MnSOD to combat oxidative stress to protect from anoikis. Depletion of MnSOD Sensitizes Mammary Epithelial Cells to Anoikis We further investigated the effect of MnSOD depletion on a noikis in MCF10A. MnSOD was depleted using retroviral shRNA and the knockdown efficiency was confirmed by western blotting (Figure 5 2) Although MnSOD was expressed under attached culture conditions, depletion of MnSOD had little effect on cell death (Figure 5 3A ) While in suspension culture conditions, knockdown of MnSOD significantly induced cell death in detached MCF10A cells (Figure 5 3A). To investigate if the increased cell death is due to increased apoptosis, we perf ormed caspase 3/7 activity and PE annexin V/7 AAD staining in MnSOD depleted MCF10A cells in attached and suspended culture cells. As expected, knockdown of MnSOD displayed incre ased caspase 3/7 activity (Figure 5 3B ) and PE a nnexin V/7 AAD staining (Figur e 5 3C ) when cells were in suspension. Compared to control cells, MnSOD depleted cells exhibited elevated early stage (22% vs. 14%) and late stage (10% vs. 5%) apoptosis when detached from matrix (Figure 5 3D). Consistent with TB exclusion assay, no signif icant increase in apoptosis was observed in MnSOD

PAGE 102

102 depleted MCF10A cells under adherent culture condition. These results support that increased oxidative stress enhances anoikis. Epistatic Relationship B etween PDK4 and MnSOD upon Matrix Detachment Upon deta chment, mammary epithelial cells upregulated both PDK4 and MnSOD. Our data provide evidence that depletion of either gene sensitized MCF10A to anoikis. Therefore, we decided to investigate if these two genes have an epistatic relationship. At RNA level, PD K4 depleted MCF10A ce lls expressed significantly higher MnSOD levels than control cells fo llowing cell detachment (Figure 5 4A ). Conversely, MnSOD depleted MCF10A cells induced mo re PDK4 than control cells in suspended culture condition (Figure 5 4B ). Howe ver, the induction of MnSOD in PDK4 depleted MCF10A suspended cells was significantly higher than PDK4 induction in MnSOD depleted MCF10A suspended culture cells. In addition, unlike PDK4, induction of MnSOD was not dependent on ERR (Figure 5 4C) But, li ke PDK4 depleted MCF10A cells, ERR depleted MCF10A cells induced more MnSOD suspended cells (Figure 5 4C). These findings suggest that upon matrix detachment, PDK4 and MnSOD genes m ay compensate each other to evade ROS production and resist anoikis. Summ ary Taken together, we conclude that upon detachment, mammary epithelial cells simultaneously upregulate both ERR PDK4 and MnSOD to reduce ROS levels (Figure 2 5). The loss of one gene may compensate for the other. Thus, it suggests an epistatic relations hip exists between the two genes. These alterations in both metabolism and endogenous antioxidant system confer increased resis tance to anoikis by eliminating oxidative stress and extend survival of cells in suspension (Figure 5 5 ).

PAGE 103

103 Figure 5 1. Induction of MnSOD in HMEC and MCF10A suspended cells. HMEC and MCF10A cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 24 hrs. After 24 hrs of incubation, the cells were collected to extract RNA using Trizol r used from each sample to synthesize cDNA by Moloney Murine Leukemia Virus (M quantitative PCR reaction with SYBR green PCR mix. The relative RNA levels of MnSOD wer Actin in (A) HMEC and (B) MCF10A. These data represent triplicate experiments. Error bars were expressed as standard error of the mean (S.E.M.). Statistical analysis was test. Att = Attached, Susp = Suspended and MnSOD = Manganese superoxide dismutase.

PAGE 104

104 Figure 5 2. Knockdown of MnSOD in MCF10A cells. MCF10A cells were transduced with retrovirus carrying empty vector (10A EV) and short hairpin RNA against MnSOD (10A shMnSOD). The cells were grown under attached and poly hema coated suspended cells for 24 hrs. After 24 hrs of incubation, the cells were collected and lysed in denature lysis buffer Mercaptoethanol. The samples were analyzed by 12 % SDS PAGE, followed by Immu noblotting using rabbit polyclonal anti MnSOD (1:1000) and mouse anti t ubulin antibody (1:5000) as loading control Att = Attached, Susp = Suspended, and MnSOD = Man ganese superoxide dismutase.

PAGE 105

105 Figure 5 3. Depletion of MnSOD leads to increased anoikis in MCF10A. MCF10A EV and MnSOD depleted MCF10A cells were grown in a 35 mm plates under attached and poly hema coated suspended culture conditions for 24 hrs. (A) Try pan Blue (TB) exclusion assay After 24 hrs of incubation, the cells were collected from all the samples, trypzined, washed, and resuspended in 1 mL PBS. 1:1 ratio of 0.4% trypa percentage of viable cells was counted by excluding dead cells stained blue using hemocytometer. (B) Caspase 3/7 glo assay. After 24 hrs of incubation, the cells were collected, washed and lyed in hypertonic buf fer (HTB) to by HTB for each sample was loaded onto 96 well microplate. Equal volume of caspase 3/7 glo substrate was added to measure caspase 3/7 activity using Luminometer. (C D) PE Annexin V/ 7 AAD analysis. After 24 hrs of incubation, the cells were collected washed twice in PBS and 1*10 5 cells were Annexin V and 7 AAD were added to the cells and incubated for 15 minutes at room temperature. Later, the cells were immediately analyzed by FACScan machine. A total number of

PAGE 106

106 10,000 cells were analyzed for each sample (C) The dot plot representing the percentage of total number of cells that were unstained or live (Lower left quadrant) stained positive for PE Annexin V only (Lower right quadrant), stained positive for PE Annexin V and 7 AAD (Upper right quadrant) and stained positive for 7 AAD alone (Upper left quadrant). FL2 H x axis represents PE Annexin V positive and FL3 H y axis r epresent 7 AAD positive cells. (D ) Statistical representation of total percentage of apoptotic cells (percentage of PE AnnexinV only + percentage of PE AnnexinV and 7 AAD cells). All error bars represent standard deviation (n=3). Statistical analysis was p test Att = Attached, Susp = S uspended, EV = Empty vector, MnSOD = Manganese superoxide dismutase, PE = Phycoerythrin, 7 AAD = 7 Aminoactinomycin D, and FL2H or FL3H= fluorescence emission at its highest peak represented in logarithmic scale.

PAGE 107

107 Figure 5 4. Epistatic relationship between PDK4 and MnSOD in MCF10A suspended culture cells. MCF10A EV, 10A shPDK4, and 10A under attached and poly hema coated suspended culture conditions for 24 hrs. After 24 hrs of incubation, the cells were collected to extract RNA using by Moloney Murine Leukemia Virus (M cDNA was used to run realtime quantitative PCR reaction with SYBR green PCR mix. (A) qRT PCR for MnSOD expression in EV, PDK4 depleted 10A and MnSOD depleted 10A cells (B) qRT PCR for PDK4 in 10A EV, PDK4 depleted MCF10A and MnSOD depleted MCF10A cells and (C) MnSOD in 10A EV and 10A s. These data represent triplicate experiments. Att = Attached, Susp = Suspended EV = Empty vector, sh = Short hairpin RNA, PDK4 = Pyruvate dehydrogenase kinase isozyme 4, shPDK4 = PDK4 epleted MCF10A cells.

PAGE 108

108 Figure 5 5. Summary model. Detachment from ECM induces MnSOD in MCF10A in PDK4 to prevent ROS production and resist anoikis. The loss of PDK4 is compromised by MnSOD and vice versa in order to evade oxidative str ess during detachment and prolong cell survival.

PAGE 109

109 CHAPTER 6 CONCLUSIONS AND FUTURE DIRECTIONS Conclusions The present study examines the role of glucose metabolism in the regulation of anoikis or detachment induced cell death and its impact on breast canc er metastasis in vivo Upon detachment from ECM, the normal epithelial cells undergo cell death known as anoikis to prevent irregular growth. Unlike normal cells, tumor cells resist anoikis to survive through blood circulation, invade and metastasize to d istant organs. Compared to normal cells, tumor cells have altered glucose metabolism known as aerobic glycolysis or the Warburg effect. Since tumor metastasis is one of the leading causes of death among cancer patients, it is critical to understand the mol ecular mechanisms that contribute to this process Although the Warburg effect contributes to tumor initiation, growth, and survival, its role in tumor metastasis is yet to be elucidated. To study the role of glucose metabolism in the development of anoiki s resistance and tumor metastasis, we used untransformed mammary epithelial cells (MCF10A and HMEC), Ras transformed mammary epithelial cells (MCF10ACA1.1) and highly metastatic breast cancer cells (MDA MB 231) as our model systems. The first part of this work examines the impact of glucose metabolism on cell survival during detachment in normal mammary epithelial cells. We have demonstrated that in response to cell detachment HMEC and MCF10A cells markedly upregulate PDK4 to inhibit PDH, thereby att enuat ing the conversion of glycolysis derived pyruvate to acetyl CoA in the mitochondria and subsequent flux into the TCA cycle. This metabolic reprogramming diverts glucose derived carbons away from mitochondrial oxidative metabolism, decreases oxygen consumpt ion and mitochondrial ROS

PAGE 110

110 production in MCF10A cells specifically under suspension culture conditions. These data indicate that matrix attachment has a profound effect on glucose oxidation in mammary epithelial cells. While Brugge et al ., concomitantly rep orted the induction of PDK4 leading to attenuation of mitochondrial activity in MCF10A suspension cells 131 the impact of detachment triggered metabolic shift on anoikis proves to be the novel aspect of this study In evaluating the functional significance of PDK4 induction in MCF10A cells sp ecifically under suspension culture conditions, we have demonstrated that depletion of PDK4 accelerates anoikis by activating intrinsic apoptotic pathway. This increased cell death in MCF10A suspension cells by depletion of PDK4 is due to the activation of PDH, which in turn increases mitochondrial respiration and oxidative stress. In other words, MCF10A cells attenuate mitochondrial respiration by upregulating PDK4 to delay apoptosis upon matrix detachment. To prove that attenuation of mitochondrial oxidat ion prolongs cell survival under suspension culture conditions, we have constitutively activated PDH in MCF10A cells. Not surprisingly, activation of PDH sensitizes MCF10A cells to anoikis. These findings reveal how untransformed mammary epithelial cells m anage to survive longer in the absence of matrix by reprogramming glucose metabolism. This study also provides direct evidence that decreased glucose oxidation promotes anoikis resistance in normal mammary epithelial cells. Depletion of PDK4 induces signi ficant cell death in MCF10A suspension cells but demonstrates a slight shift towards glucose oxidation. In other words, elimination of PDK4 may not have completely relieved PDH enzyme activity and hence we observe a partial increase in glucose oxidation. O ne possible explanation for this incomplete

PAGE 111

111 activation of PDH enzyme is the presence of other PDK isozymes. Although PDK4 and PDK2 are induced upon matrix detachment in MCF10A cells, PDK1 and PDK4 are the most abundant of all the four isozymes. Also, the e nzymatic inhibitory effect of PDK1 is higher than PDK4. Therefore, depletion of PDK4 in MCF10A cells partially activates the PDH enzyme due to the existence of PDK1. Similar to PDK4 depletion, it is possible that depletion of PDK1 in MCF10A cells may sensi tize them to anoikis. PDK4 depleted MCF10A cells markedly undergo anoikis in suspension due to increased oxidative stress through mitochondrial respiration. Pharmacological treatment with antioxidants rescued both the control and PDK4 depleted MCF10A cells from apoptosis under suspension culture conditions. Schafer et al ., demonstrated that matrix detachment increases ROS production in MCF10A 68 but our data indicate no significant change in ROS leve ls. A plausible reason for antioxidants to protect control MCF10A cells from anoikis could be that matrix detachment may render cells sensitive to oxidative stress. This observation is further reinforced when MCF10A cells exhibited increased apoptosis due to forced activation of PDH specifically in suspension cells but not in attached ones. In other words, forced activation of PDH increases oxidative stress in MCF10A cells both under attached and suspended culture conditions but only suspension cells are se nsitive to oxidative stress and therefore undergo anoikis. It is intriguing that MCF10A cells activate the ERR PDK4 axis following matrix detachment. Expression of PDK4 is usually stimulated by starvation, which consequently suppresses PDH and curtails g lucose oxidation. This response conserves glucose reserves and allows the switch from the utilization of glucose to fatty acids as an energy source. On the other hand, ERRs are key regulators of fatt y acid oxidation It

PAGE 112

112 was recently observed that matrix de tachment led to reduce d gluc ose uptake in MCF10A cells 68 Therefore, cell detachment may mimic glucose starvation, and trigger the ERR PDK4 program and metabolic shift in MCF10A cells Consistent wi th cell's metabolic shift to reduce glucose oxidation and oxidative stress response to detachment, MCF10A and HMEC suspended cells also stimulate expression of MnSOD, which mitigates the oxidative stress and extends the viability of cells in suspension. Co upling of a decrease in ROS generation and an increase in cellular antioxidant capacity to scavenge ROS represents an integrated strategy for cells to avoid oxidative damage and survive longer under the s tress of matrix detachment. These observations revea l how cells manage to control oxidative stress to resist anoikis. Consistent with our observations it has been reported that expression of MnSOD is elevated in aggressive breast cancer cells and contributes to tumor invasiveness and metastasis 12 3 Our data provide the first direct evidence that attenuation of mitochondrial respiration promotes anoikis resistance in normal mammary epithelial cells. The second part of our work focuses on examining the influence of glucose metabolism on anoikis re sistance and tumor metastasis in breast cancer cells. Since the untransformed mammary epithelial cells induce PDK4 to resist anoikis, we decided to examine the role of PDKs in Ras transformed mammary epithelial cells (MCF10ACA1.1). As expected, MCF10ACA1.1 cells significantly upregulate PDK4 upon matrix detachment to protect cells from anoikis. Although MCF10ACA1.1 are highly metastatic in nature and induced PDK4 upon matrix detachment, they do not form significantly larger tumor nodules in vivo (data not s hown). It is more clinically relevant to study the impact of glucose

PAGE 113

113 metabolism on anoikis resistance in a breast cancer cell line derived from human patients. Therefore, we have focused our attention to a highly metastatic, aggressive, and triple negative breast cancer cell line MDA MB 231. MDA MB 231 cells are highly glycolytic in nature. Because cancer cells exhibit a preference for aerobic glycolysis and a low rate of mitochondrial oxidation of glucose, we hypothesize that the Warburg effect intrinsica lly bestows a survival advantage upon cancer cells when they are detached from the matrix, and facilitates their metastatic spreading. Consistent with our hypothesis, MDA MB 231 cells exhibit significantly lower PDH activity than the normal mammary epithel ial cells under adherent culture conditions. Surprisingly, matrix detachment significantly attenuates PDH activity, increases lactate levels and decreases OCR in MDA MB 231 cells. These observations prove that upon detachment from matrix, the highly aggres sive breast cancer MDA MB 231 cells further enhance the Warburg effect to promote anoikis resistance. Indeed, normalization of glucose metabolism in MDA MB 231 cells by activation of PDH redirects pyruvate towards mitochondrial oxidation and restores their sensitivity to anoikis. Therefore, the attenuation of mitochondrial respiration protects breast cancer cells from detachment induced cell death by evading oxidative stress from mitochondria. Like MCF10A, HMEC and MCF10ACA1.1 cell lines, MDA MB 231 cells i nduce PDK4 upon detachment from matrix. But, the abundance of PDK1 is significantly higher than all other three PDK isozyme. Also, PDK1 protein levels are upregulated when detached from the matrix in MDA MB 231 cells. Therefore, we have decided to examine its role in attenuating mitochondrial oxidation and promoting anoikis resistance and

PAGE 114

114 tumor metastasis in vivo As expected, depletion of PDK1 in MDA MB 231 enhances mitochondrial PDH activity, decreases lactate levels and increases OCR both under attached and detached culture conditions. However, this metabolic shift does not effect the MDA MB 231 cells under adherent conditions but markedly promotes anoikis in detached cells. These data help explain why breast cancer cells enhance the Warburg effect upon d etachment from the matrix. Anoikis resistance is an important step during tumor metastasis. For cancer cells to metastasize to distant organs, they need to detach from the primary tumor site, enter the circulatory system or lymphatic system, develop resis tance to anoikis to survive through circulation and thus invade the distant organs for secondary site. So, resistance to detachment induced cell death (anoikis) is a critical step for cancer cells to survive and metastasize. Our work has demonstrated that PDK1 is a key molecule for promoting anoikis resistance in MDA MB 231 cells. Consistent with our in vitro data, depletion of PDK1 in MDA MB 231 cells markedly reduced lung tumor nodules in an in vivo metastatic experimental assay. Altered glucose metabolis m is believed to support tumor cell proliferation and our results add new insights into the significance of the Warburg effect in cancer metastasis, and establish PDKs as an important regulator of anchorage independent cell survival as well as tumor metast asis. The latter may serve as the basis for anti metastasis therapeutic interventions Reprogramming tumor metabolism, in particular inhibition of PDKs to stimulate glucose oxidation, should sensitize cancer cells to anoikis, and hence may impede the detac hment of cancer cells from the primary site, kill circulating tumor cells, and eradicate disseminated tumor cells at secondary sites before they re establish conducive cell matrix interactions.

PAGE 115

115 Increased glucose uptake and glycolysis in cancer cells provid es substrates for macromolecular biosynthesis required for cell proliferation This convincingly explains at least part of the advantage provided by the Warburg effect. However, it is puzzling as to why after the glycolytic process the end product pyruvat e in cancer cells is primarily dispensed as lactate instead of entering the TCA cycle even under normal oxygen tensions. This phenomenon might be attributed to the detrimental consequence of mitochondrial oxidation. I ncreased glucose consumption through ae robic glycolysis by cancer cells (for the need of biosynthesis) conceivably gives rise to increased pyruvate production. If most pyruvate enters the mitochondrial oxidative pathway as in normal cells, the resultant increased ROS would probably disrupt the cellular redox balance and jeopardize cancer cells' capability to cope with stressful conditions such as loss of matrix attachment. By shunting pyruvate away from mitochondria, the Warburg effect helps cancer cells avoid generation of excessive ROS and res ist anoikis. In this regard, increased aerobic glycolysis also stimulates the reactions of the PPP, which is a principal pathway to generate the reducing equiv alent, NADPH By evading production of extra ROS and increasing antioxidant defense, the Warburg effect maintains redox homeostasis, and thus promotes anoikis resistance and metastasis. It is well molecules to promote solid tumor growth and survival 41,43 DCA is a potential drug available to inhibit PDK enzyme activity by preventing the ph osphorylation of PDH enzyme. DCA is a well studied molecule to prevent lactic acidosis in children 54 Several recent reports have demonstrated that treatment with DCA activates mitochondrial respiration in cancer cells leading to mitochondrial membrane permeabilization, restores

PAGE 116

116 apoptosis, kills cancer cells in vitro and shrinks tumor in rats 53 Further, Michelakis and group performed phase II clinical trials using DCA to treat glioblastoma patients 132 DCA successfully regressed the tumor with minimal side effects in all four out of five patients. Although, DCA has been successful in treating certain cancer models, higher dosage of DCA (>25mg/kg/day) causes peripheral neuropathy, neurotoxicity, and gait disturbances. In addition, unlike PDK2, PDK1 and PDK4 are highly resistant to DCA 133 Therefore inhibition of PDK1 and PDK4 requires higher amounts of DCA, which might presumably cause severe neuropathic symptoms. Based on our work, it is evident that PDK1 promotes anoikis resistance and tumor metastasis in breast cancer cells. Therefore, it is cri tical to design additional drugs that can potentially inhibit PDK1 to prevent both cancer growth and metastatic spreading. Future D irections The present study focuses mainly on understanding the role of glucose metabolism in regulating anoikis resistance in both normal mammary epithelial cells and metastatic breast cancer cells. We have demonstrated that upon detachment, mammary epithelial cells simultaneously upregulate PDK to attenuate mitochondrial oxidation and MnSOD to evade oxidative stress. This is an important strategy for normal mammary epithelial cells to survive longer under detached conditions. Also, we have established that metastatic breast cancer cells further enhance the Warburg effect upon detachment from matrix by upregulating PDK1 to pro mote anoikis resistance and tumor metastasis in vivo Based on our work, it is quite evident that detachment from matrix induces MnSOD in MCF10A cells. Also, MnSOD has been implicated in tumor progression and metastasis. Therefore, it is imperative to iden tify the molecule that activates MnSOD

PAGE 117

117 during detachment. Several studies have reported that the nuclear factor kappa light chain enhancer of MnSOD 134 It is well established th apoptosis in several cell lines 135,136 So, depleting RelA (p65), an important molecule in regulated genes is an important future directi on. After depletion of p65 in MCF10A, we will examine the expression of MnSOD in these cells both under attached and detached culture conditions. According to our hypothesis, the depletion of p65 may repress the MnSOD RNA levels in detached cells but not in adherent cells. Thus, we may be able to provide the detailed information of NFkB MnSOD pathway and how it regulates anoikis resistance in mammary epithelial cells. Our data provide evidence that normal mammary epithelial cells induce PDK4 upon detachment to prevent anoikis and survive longer. Although MDA MB 231 induced PDK4 upon detachment, its overall abundance is trivial. Therefore, it will be quite intriguing to examine the expression of PDK4 in different tumor cells other than breast cancer cells. Based on Standford microarray profiling, A549 (Lung adenocarcinoma) and SKMEL5 (Melanoma cell line) cells show higher PDK4 expression levels. Our preliminary experiments confir m that the expression of PDK4 at RNA level is higher in A549 and SKMEL5 when compared with MCF10A (Figure 6 1). So, future work will include depletion of PDK4 and perform anoikis assay to test its functional significance in these tumor cells. Finally, the key finding of our work is that PDK1 can be utilized as a potential drug target for breast cancer metastasis. Thus, future work will focus on developing effective

PAGE 118

118 drug inhibiting PDK1 activity. Additionally, since DCA is not effective in inhibiting PDK1 en zyme, it is essential to design potential inhibitors for PDK1 to prevent tumor metastasis in vivo

PAGE 119

119 Figure 6 1. Expression of PDK4 in different tumor cells lines. MCF10A, A549 and SKMEL5 cells were grown in 35 mm plates under attached condition for 24 hr s. After 24 hrs of incubat ion, the cells were collected to extract RNA by used from each sample to run quantitative RT PCR. The relative PDK4 RNAL actin. After normalization, the f old induction of PDK4 expression in MCF10A was set to one. The fold induction of PDK4 in A549 and SKMEL5 cells were compared to PDK expression in MCF10A. These data represent n=3. A549= lung adenocarcinoma cells, and SKMEL5= melanoma cells.

PAGE 120

120 LIST OF REF ERENCES 1 Frisch, S. M. & Francis, H. Disruption of epithelial cell matrix interactions induces apoptosis. Journal of Cell Biology 124 619 626, doi:10.1083/jcb.124.4.619 (1994). 2 Chiarugi, P. & Giannoni, E. Anoikis: A necessary death program for anchorage dependent cells. Biochemical Pharmacology 76 1352 1364, doi:10.1016/j.bcp.2008.07.023 (2008). 3 Gilmore, A. P. Anoikis. Cell Death and Differentiation 12 1473 1477, doi:10.1038/sj.cdd.4401723 (2005). 4 Frisch, S. M. & Screaton, R. A Anoikis mechanisms. Current Opinion in Cell Biology 13 555 562, doi:10.1016/s0955 0674(00)00251 9 (2001). 5 Reginato, M. J. et al. Integrins and EGFR coordinately regulate the pro apoptotic protein Bim to prevent anoikis. Nature Cell Biology 5 733 740, doi:10.1038/ncb1026 (2003). 6 Parsons, J. T. Focal adhesion kinase: the first ten years. Journal of Cell Science 116 1409 1416, doi:10.1242/jcs.00373 (2003). 7 Frisch, S. M. Evidence for a function of death receptor related, death domain containing prote ins in anoikis. Current Biology 9 1047 1049, doi:10.1016/s0960 9822(99)80455 2 (1999). 8 Rytomaa, M., Martins, L. M. & Downward, J. Involvement of FADD and caspase 8 signalling in detachment induced apoptosis. Current Biology 9 1043 1046, doi:10.1016/s09 60 9822(99)80454 0 (1999). 9 Puthalakath, H. et al. Bmf: A proapoptotic BH3 only protein regulated by interaction with the myosin V actin motor complex, activated by anoikis. Science 293 1829 1832, doi:10.1126/science.1062257 (2001). 10 Schmelzle, T. et a l. Functional role and oncogene regulated expression of the BH3 only factor Bmf in mammary epithelial anoikis and morphogenesis. Proceedings of the National Academy of Sciences of the United States of America 104 3787 3792, doi:10.1073/pnas.0700115104 (20 07). 11 Hausmann, M. et al. BCL 2 modifying factor (BMF) is a central regulator of anoikis in human intestinal epithelial Cells. Journal of Biological Chemistry 286 26533 26540, doi:10.1074/jbc.M111.265322 (2011). 12 Idogawa, M., Adachi, M., Minami, T., Y asui, H. & Imai, K. Overexpression of bad preferentially augments anoikis. International Journal of Cancer 107 215 223, doi:10.1002/ijc.11399 (2003).

PAGE 121

121 13 Gilley, J., Coffer, P. J. & Ham, J. FOXO transcription factors directly activate bim gene expression a nd promote apoptosis in sympathetic neurons. Journal of Cell Biology 162 613 622, doi:10.1083/jcb.200303026 (2003). 14 Ley, R. et al. Extracellular signal regulated kinases 1/2 are serum stimulated "Bim(EL) kinases" that bind to the BH3 only protein Bim(E L) causing its phosphorylation and turnover. Journal of Biological Chemistry 279 8837 8847, doi:10.1074/jbc.M311578200 (2004). 15 Reginato, M. J. et al. Bim regulation of lumen formation in cultured mammary epithelial acini is targeted by oncogenes. Molec ular and Cellular Biology 25 4591 4601, doi:10.1128/mcb.25.11.4591 4601.2005 (2005). 16 Whelan, K. A. et al. Hypoxia Suppression of Bim and Bmf Blocks Anoikis and Luminal Clearing during Mammary Morphogenesis. Molecular Biology of the Cell 21 3829 3837, doi:10.1091/mbc.E10 04 0353 (2010). 17 Valentijn, A. J., Zouq, N. & Gilmore, A. P. Anoikis. Biochemical Society Transactions 32 421 425, doi:10.1042/bst0320421 (2004). 18 Valentijn, A. J. & Gilmore, A. P. Translocation of full length bid to mitochondria d uring anoikis. Journal of Biological Chemistry 279 32848 32857, doi:10.1074/jbc.M313375200 (2004). 19 Ilic, D. et al. Extracellular matrix survival signals transduced by focal adhesion kinase suppress p53 mediated apoptosis. Journal of Cell Biology 143 5 47 560 (1998). 20 Vitale, M. et al. Apoptosis induced by denied adhesion to extracellular matrix (anoikis) in thyroid epithelial cells is p53 dependent but fails to correlate with modulation of p53 expression. Febs Letters 462 57 60, doi:10.1016/s0014 579 3(99)01512 4 (1999). 21 Stromblad, S., Becker, J. C., Yebra, M., Brooks, P. C. & Cheresh, D. A. Suppression of p53 activity and p21(WAF1/CIP1) expression by vascular cell integrin alpha v beta 3 during angiogenesis. Journal of Clinical Investigation 98 42 6 433, doi:10.1172/jci118808 (1996). 22 Lu, Y. L. et al. The PTEN/MMAC1/TEP tumor suppressor gene decreases cell growth and induces apoptosis and anoikis in breast cancer cells. Oncogene 18 7034 7045, doi:10.1038/sj.onc.1203183 (1999). 23 Simpson, C. D., Anyiwe, K. & Schimmer, A. D. Anoikis resistance and tumor metastasis. Cancer Letters 272 177 185, doi:10.1016/j.canlet.2008.05.029 (2008). 24 Liotta, L. A. & Kohn, E. C. The microenvironment of the tumour host interface. Nature 411 375 379, doi:10.1038/3 5077241 (2001).

PAGE 122

122 25 Lopez, D., Niu, G., Huber, P. & Carter, W. B. Tumor induced upregulation of Twist, Snail, and Slug represses the activity of the human VE cadherin promoter. Archives of Biochemistry and Biophysics 482 77 82, doi:10.1016/j.abb.2008.11.01 6 (2009). 26 Haenssen, K. K. et al. ErbB2 requires integrin alpha 5 for anoikis resistance via Src regulation of receptor activity in human mammary epithelial cells. Journal of Cell Science 123 1373 1382, doi:10.1242/jcs.050906 (2010). 27 Giannoni, E. et al. Redox regulation of anoikis: reactive oxygen species as essential mediators of cell survival. Cell Death and Differentiation 15 867 878, doi:10.1038/cdd.2008.3 (2008). 28 Liu, Z. P. et al. Oncogenic Ras inhibits anoikis of intestinal epithelial cells by preventing the release of a mitochondrial pro apoptotic protein Omi/HtrA2 into the cytoplasm. Journal of Biological Chemistry 281 14738 14747, doi:10.1074/jbc.M508664200 (2006). 29 Pouyssegur, J., Dayan, F. & Mazure, N. M. Hypoxia signalling in cancer and approaches to enforce tumour regression. Nature 441 437 443, doi:10.1038/nature04871 (2006). 30 Markert, C. L., Shaklee, J. B. & Whitt, G. S. Evolution of a gene. Science 189 102 114, doi:10.1126/science.1138367 (1975). 31 Thornburg, J. M. et al. Tar geting aspartate aminotransferase in breast cancer. Breast Cancer Research 10 doi:R8410.1186/bcr2154 (2008). 32 Giatromanolaki, A. et al. Lactate dehydrogenase 5 expression in non Hodgkin B cell lymphomas is associated with hypoxia regulated proteins. Leu kemia & Lymphoma 49 2181 2186, doi:10.1080/10428190802450629 (2008). 33 Sugden, M. C. & Holness, M. J. Mechanisms underlying regulation of the expression and activities of the mammalian pyruvate dehydrogenase kinases. Archives of Physiology and Biochemist ry 112 139 149, doi:10.1080/13813450600935263 (2006). 34 Gudi, R., Bowkerkinley, M. M., Kedishvili, N. Y., Zhao, Y. & Popov, K. M. Diversity of the pyruvate dehydrogenase kinase gene family in humans. Journal of Biological Chemistry 270 28989 28994 (1995 ). 35 Huang, B. L., Wu, P. F., Popov, K. M. & Harris, R. A. Starvation and diabetes reduce the amount of pyruvate dehydrogenase phosphatase in rat heart and kidney. Diabetes 52 1371 1376, doi:10.2337/diabetes.52.6.1371 (2003). 36 Baker, J. C., Yan, X. H., Peng, T., Kasten, S. & Roche, T. E. Marked differences between two isoforms of human pyruvate dehydrogenase kinase. Journal of Biological Chemistry 275 15773 15781, doi:10.1074/jbc.M909488199 (2000).

PAGE 123

1 23 37 Roche, T. E. & Hiromasa, Y. Pyruvate dehydrogenase kinase regulatory mechanisms and inhibition in treating diabetes, heart ischemia, and cancer. Cellular and Molecular Life Sciences 64 830 849, doi:10.1007/s00018 007 6380 z (2007). 38 Hiromasa, Y. & Roche, T. E. Facilitated interaction between the pyruvat e dehydrogenase kinase isoform 2 and the dihydrolipoyl acetyltransferase. Journal of Biological Chemistry 278 33681 33693, doi:10.1074/jbc.M212733200 (2003). 39 Bowker Kinley, M. M., Davis, W. I., Wu, P. F., Harris, R. A. & Popov, K. M. Evidence for exist ence of tissue specific regulation of the mammalian pyruvate dehydrogenase complex. Biochemical Journal 329 191 196 (1998). 40 Korotchkina, L. G. & Patel, M. S. Site specificity of four pyruvate dehydrogenase kinase isoenzymes toward the three phosphoryla tion sites of human pyruvate dehydrogenase. Journal of Biological Chemistry 276 37223 37229, doi:10.1074/jbc.M103069200 (2001). 41 Papandreou, I., Cairns, R. A., Fontana, L., Lim, A. L. & Denko, N. C. HIF 1 mediates adaptation to hypoxia by actively downr egulating mitochondrial oxygen consumption. Cell Metabolism 3 187 197, doi:10.1016/j.cmet.2006.01.012 (2006). 42 Kim, J. W., Tchernyshyov, I., Semenza, G. L. & Dang, C. V. HIF 1 mediated expression of pyruvate dehydrogenase kinase: A metabolic switch requ ired for cellular adaptation to hypoxia. Cell Metabolism 3 177 185, doi:10.1016/j.cmet.2006.02.002 (2006). 43 Lu, C. W., Lin, S. C., Chen, K. F., Lai, Y. Y. & Tsai, S. J. Induction of pyruvate dehydrogenase kinase 3 by hypoxia inducible factor 1 promotes metabolic switch and drug resistance. Journal of Biological Chemistry 283 28106 28114, doi:10.1074/jbc.M803508200 (2008). 44 Lu, C. W. et al. Overexpression of pyruvate dehydrogenase kinase 3 increases drug resistance and early recurrence in colon cancer. The American journal of pathology 179 1405 1414 (2011). 45 Sugden, M. C., Orfali, K. A., Fryer, L. G. D., Holness, M. J. & Priestman, D. A. Molecular mechanisms underlying the long term impact of dietary fat to increase cardiac pyruvate dehydrogenase kin ase: Regulation by insulin, cyclic AMP and pyruvate. Journal of Molecular and Cellular Cardiology 29 1867 1875, doi:10.1006/jmcc.1997.0425 (1997). 46 Harris, R. A., Huang, B. L. & Wu, P. F. Control of pyruvate dehydrogenase kinase gene expression. Advance s in Enzyme Regulation, Vol 41 41 269 288, doi:10.1016/s0065 2571(00)00020 0 (2001).

PAGE 124

124 47 Ma, K., Zhang, Y., Elam, M. B., Cook, G. A. & Park, E. A. Cloning of the rat pyruvate dehydrogenase kinase 4 gene promoter Activation of pyruvate dehydrogenase kinas e 4 by the peroxisome proliferator activated receptor gamma coactivator. Journal of Biological Chemistry 280 29525 29532, doi:10.1074/jbc.M502236200 (2005). 48 Zhang, Y. et al. Estrogen related receptors stimulate pyruvate dehydrogenase kinase isoform 4 g ene expression. Journal of Biological Chemistry 281 39897 39906, doi:10.1074/jbc.M608657200 (2006). 49 Wu, P. F., Peters, J. M. & Harris, R. A. Adaptive increase in pyruvate dehydrogenase kinase 4 during starvation is mediated by peroxisome proliferator a ctivated receptor alpha. Biochemical and Biophysical Research Communications 287 391 396, doi:10.1006/bbrc.2001.5608 (2001). 50 Attia, R. R. et al. Regulation of Pyruvate Dehydrogenase Kinase 4 (PDK4) by Thyroid Hormone role of the peroxisome proliferator activated receptor gamma COACTIVATOR (PGC 1 alpha). Journal of Biological Chemistry 285 2375 2385, doi:10.1074/jbc.M109.039081 (2010). 51 Attia, R. R. et al. Regulation of Pyruvate Dehydrogenase Kinase 4 (PDK4) by CCAAT/Enhancer binding Protein beta (C/E BP beta). Journal of Biological Chemistry 286 23799 23807, doi:10.1074/jbc.M111.246389 (2011). 52 Mayers, R. M. et al. AZD7545, a novel inhibitor of pyruvate dehydrogenase kinase 2 (PDHK2), activates pyruvate dehydrogenase in vivo and improves blood gluco se control in obese (fa/fa) Zucker rats. Biochemical Society Transactions 31 1165 1167 (2003). 53 Bonnet, S. et al. A mitochondria K+ channel axis is suppressed in cancer and its normalization promotes apoptosis and inhibits cancer growth. Cancer Cell 11 37 51, doi:10.1016/j.ccr.2006.10.020 (2007). 54 Stacpoole, P. W. et al. Controlled clinical trial of dichloroacetate for treatment of congenital lactic acidosis in children. Pediatrics 117 1519 1531, doi:10.1542/peds.2005 1226 (2006). 55 Kato, M., Li, J. Chuang, J. L. & Chuang, D. T. Distinct structural mechanisms for inhibition of pyruvate dehydrogenase kinase isoforms by AZD7545, dichloroacetate, and radicicol. Structure 15 992 1004, doi:10.1016/j.str.2007.07.001 (2007). 56 Turrens, J. F. Mitochondria l formation of reactive oxygen species. Journal of Physiology London 552 335 344, doi:10.1111/j.1469 7793.2003.00335.x (2003).

PAGE 125

125 57 Stowe, D. F. & Camara, A. K. S. Mitochondrial Reactive Oxygen Species Production in Excitable Cells: Modulators of Mitochondr ial and Cell Function. Antioxidants & Redox Signaling 11 1373 1414, doi:10.1089/ars.2008.2331 (2009). 58 Rush, J. D., Maskos, Z. & Koppenol, W. H. Distinction between hyroxyl radical and ferryl species. Methods in Enzymology 186 148 156 (1990). 59 Gao, S J. et al. Docking of endothelial nitric oxide synthase (eNOS) to the mitochondrial outer membrane A pentabasic amino acid sequence in the autoinhibitory domain of eNOS targets a proteinase K cleavable peptide on the cytoplasmic face of mitochondria. Jo urnal of Biological Chemistry 279 15968 15974, doi:10.1074/jbc.M308504200 (2004). 60 Droge, W. Free radicals in the physiological control of cell function. Physiological Reviews 82 47 95 (2002). 61 Trachootham, D., Alexandre, J. & Huang, P. Targeting can cer cells by ROS mediated mechanisms: a radical therapeutic approach? Nature Reviews Drug Discovery 8 579 591, doi:10.1038/nrd2803 (2009). 62 Thannickal, V. J. & Fanburg, B. L. Reactive oxygen species in cell signaling. American Journal of Physiology Lung Cellular and Molecular Physiology 279 L1005 L1028 (2000). 63 Tolbert, N. E. & Essner, E. Microbodies peroxisomes and glyoxysomes. Journal of Cell Biology 91 S271 S283, doi:10.1083/jcb.91.3.271s (1981). 64 Babior, B. M. NADPH oxidase: An update. Blood 93 1464 1476 (1999). 65 Giannoni, E., Fiaschi, T., Ramponi, G. & Chiarugi, P. Redox regulation of anoikis resistance of metastatic prostate cancer cells: key role for Src and EGFR mediated pro survival signals. Oncogene 28 2074 2086, doi:10.1038/onc.2009.7 7 (2009). 66 Young, C. D. & Anderson, S. M. Rah, rah, ROS: metabolic changes caused by loss of adhesion induce cell death. Breast Cancer Research 11 doi:30710.1186/bcr2417 (2009). 67 Li, A. E. et al. A role for reactive oxygen species in endothelial cell anoikis. Circulation Research 85 304 310 (1999). 68 Schafer, Z. T. et al. Antioxidant and oncogene rescue of metabolic defects caused by loss of matrix attachment. Nature 461 109 U118, doi:10.1038/nature08268 (2009).

PAGE 126

126 69 Lam, C. R. I. et al. TAK1 regulate s SCF expression to modulate PKB alpha activity that protects keratinocytes from ROS induced apoptosis. Cell Death and Differentiation 18 1120 1129, doi:10.1038/cdd.2010.182 (2011). 70 Hanahan, D. & Weinberg, R. A. Hallmarks of Cancer: The Next Generation Cell 144 646 674, doi:10.1016/j.cell.2011.02.013 (2011). 71 Warburg, O. Origin of cancer cells. Science 123 309 314, doi:10.1126/science.123.3191.309 (1956). 72 Vander Heiden, M. G., Cantley, L. C. & Thompson, C. B. Understanding the Warburg effect: th e metabolic requirements of cell proliferation. Science (New York, N.Y.) 324 1029 1033 (2009). 73 Fantin, V. R., St Pierre, J. & Leder, P. Attenuation of LDH A expression uncovers a link between glycolysis, mitochondrial physiology, and tumor maintenance. Cancer Cell 9 425 434, doi:10.1016/j.ccr.2006.04.023 (2006). 74 Moreno Sanchez, R., Rodriguez Enriquez, S., Marin Hernandez, A. & Saavedra, E. Energy metabolism in tumor cells. Febs Journal 274 1393 1418, doi:10.1111/j.1742 4658.2007.05686.x (2007). 75 Gillies, R. J., Robey, I. & Gatenby, R. A. Causes and consequences of increased glucose metabolism of cancers. Journal of Nuclear Medicine 49 24S 42S, doi:10.2967/jnumed.107.047258 (2008). 76 Bartrons, R. & Caro, J. Hypoxia, glucose metabolism and the War burg's effect. Journal of Bioenergetics and Biomembranes 39 223 229, doi:10.1007/s10863 007 9080 3 (2007). 77 Boros, L. G. et al. Oxythiamine and dehydroepiandrosterone inhibit the nonoxidative synthesis of ribose and tumor cell proliferation. Cancer Rese arch 57 4242 4248 (1997). 78 Cairns, R. A., Harris, I. S. & Mak, T. W. Regulation of cancer cell metabolism. Nature Reviews Cancer 11 85 95, doi:10.1038/nrc2981 (2011). 79 McFate, T. et al. Pyruvate dehydrogenase complex activity controls metabolic and m alignant phenotype in cancer cells. Journal of Biological Chemistry 283 22700 22708, doi:10.1074/jbc.M801765200 (2008). 80 Koppenol, W. H., Bounds, P. L. & Dang, C. V. Otto Warburg's contributions to current concepts of cancer metabolism. Nature Reviews C ancer 11 325 337, doi:10.1038/nrc3038 (2011). 81 Flier, J. S., Mueckler, M. M., Usher, P. & Lodish, H. F. Elevated levels of glucose transport and transporter messenger RNA are induced by ras or src oncogenes. Science 235 1492 1495, doi:10.1126/science.3 103217 (1987).

PAGE 127

127 82 Shim, H. et al. c Myc transactivation of LDH A: Implications for tumor metabolism and growth. Proceedings of the National Academy of Sciences of the United States of America 94 6658 6663, doi:10.1073/pnas.94.13.6658 (1997). 83 Dang, C. V c myc target genes involved in cell growth, apoptosis, and metabolism. Molecular and Cellular Biology 19 1 11 (1999). 84 David, C. J., Chen, M., Assanah, M., Canoll, P. & Manley, J. L. HnRNP proteins controlled by c Myc deregulate pyruvate kinase mRNA s plicing in cancer. Nature 463 364 U114, doi:10.1038/nature08697 (2010). 85 Robey, R. B. & Hay, N. Is Akt the "Warburg kinase"? Akt energy metabolism interactions and oncogenesis. Seminars in Cancer Biology 19 25 31, doi:10.1016/j.semcancer.2008.11.010 (2 009). 86 Zundel, W. et al. Loss of PTEN facilitates HIF 1 mediated gene expression. Genes & Development 14 391 396 (2000). 87 Bensaad, K. et al. TIGAR, a p53 inducible regulator of glycolysis and apoptosis. Cell 126 107 120, doi:10.1016/j.cell.2006.05.03 6 (2006). 88 Pollard, P. J. et al. Accumulation of Krebs cycle intermediates and over expression of HIF1 alpha in tumours which result from germline FH and SDH mutations. Human Molecular Genetics 14 2231 2239, doi:10.1093/hmg/ddi227 (2005). 89 Porcelli, A M. et al. The genetic and metabolic signature of oncocytic transformation implicates HIF1 alpha destabilization. Human Molecular Genetics 19 1019 1032, doi:10.1093/hmg/ddp566 (2010). 90 Reitman, Z. J., Parsons, D. W. & Yan, H. IDH1 and IDH2: not your ty pical oncogenes. Cancer Cell 17 215 216, doi:10.1016/j.ccr.2010.02.024 (2010). 91 Gross, S. et al. Cancer associated metabolite 2 hydroxyglutarate accumulates in acute myelogenous leukemia with isocitrate dehydrogenase 1 and 2 mutations. Journal of Experi mental Medicine 207 339 344, doi:10.1084/jem.20092506 (2010). 92 Altekruse, S. F., McGlynn, K. A. & Reichman, M. E. Hepatocellular Carcinoma Incidence, Mortality, and Survival Trends in the United States From 1975 to 2005. Journal of Clinical Oncology 27 1485 1491, doi:10.1200/jco.2008.20.7753 (2009). 93 Degenhardt, T. et al. Three members of the human pyruvate dehydrogenase kinase gene family are direct targets of the peroxisome proliferator activated receptor beta/delta. Journal of Molecular Biology 372 341 355, doi:10.1016/j.jmb.2007.06.091 (2007).

PAGE 128

128 94 Edinger, A. L. & Thompson, C. B. Death by design: apoptosis, necrosis and autophagy. Current Opinion in Cell Biology 16 663 669, doi:10.1016/j.ceb.2004.09.011 (2004). 95 Alves, N. L. et al. The Noxa/Mcl 1 axis regulates susceptibility to apoptosis under glucose limitation in dividing T cells. Immunity 24 703 716, doi:10.1016/j.immuni.2006.03.018 (2006). 96 Chi, M. M. Y., Pingsterhaus, J., Carayannopoulos, M. & Moley, K. H. Decreased glucose transporter e xpression triggers BAX dependent apoptosis in the murine blastocyst. Journal of Biological Chemistry 275 40252 40257, doi:10.1074/jbc.M005508200 (2000). 97 Vander Heiden, M. G. et al. Growth factors can influence cell growth and survival through effects o n glucose metabolism. Molecular and Cellular Biology 21 5899 5912, doi:10.1128/mcb.21.17.5899 5912.2001 (2001). 98 Zhao, Y., Wieman, H. L., Jacobs, S. R. & Rathmell, J. C. Mechanisms and methods in glucose metabolism and cell death. Programmed Cell Death, General Principles for Studying Cell Death, Pt A 442 439 457, doi:10.1016/s0076 0879(08)01422 5 (2008). 99 Patel, M. S. & Korotchkina, L. G. Regulation of the pyruvate dehydrogenase complex. Biochemical Society Transactions 34 217 222 (2006). 100 Wende, A. R., Huss, J. M., Schaeffer, P. J., Giguere, V. & Kelly, D. P. PGC 1 alpha coactivates PDK4 gene expression via the orphan nuclear receptor ERR alpha: a mechanism for transcriptional control of muscle glucose metabolism. Molecular and Cellular Biology 2 5 10684 10694, doi:10.1128/mcb.25.24.10684 10694.2005 (2005). 101 Araki, M. & Motojima, K. Identification of ERR alpha as a specific partner of PGC 1 alpha for the activation of PDK4 gene expression in muscle. Febs Journal 273 1669 1680, doi:10.1111/j.17 42 4658.2006.05183.x (2006). 102 Dufour, C. R. et al. Genome wide orchestration of cardiac functions by the orphan nuclear receptors ERR alpha and gamma. Cell Metabolism 5 345 356, doi:10.1016/j.cmet.2007.03.007 (2007). 103 Giguere, V. Transcriptional Con trol of Energy Homeostasis by the Estrogen Related Receptors. Endocrine Reviews 29 677 696, doi:10.1210/er.2008 0017 (2008). 104 Eichner, L. J. & Giguere, V. Estrogen related receptors (ERRs): A new dawn in transcriptional control of mitochondrial gene ne tworks. Mitochondrion 11 544 552, doi:10.1016/j.mito.2011.03.121 (2011).

PAGE 129

129 105 Carroll, V. A. & Ashcroft, M. Targeting the molecular basis for tumour hypoxia. Expert Reviews in Molecular Medicine 7 1 16, doi:10.1017/s1462399405009117 (2005). 106 Sullivan, R. & Graham, C. H. Hypoxia driven selection of the metastatic phenotype. Cancer and Metastasis Reviews 26 319 331, doi:10.1007/s10555 007 9062 2 (2007). 107 Santner, S. J. et al. Malignant MCF10CA1 cell lines derived from premalignant human breast epithel ial MCF10AT cells. Breast Cancer Research and Treatment 65 101 110, doi:10.1023/a:1006461422273 (2001). 108 Gatenby, R. A. & Gillies, R. J. Why do cancers have high aerobic glycolysis? Nature Reviews Cancer 4 891 899, doi:10.1038/nrc1478 (2004). 109 Adda bbo, F., Montagnani, M. & Goligorsky, M. S. Mitochondria and Reactive Oxygen Species. Hypertension 53 885 892, doi:10.1161/hypertensionaha.109.130054 (2009). 110 Chen, Y., Azad, M. B. & Gibson, S. B. Superoxide is the major reactive oxygen species regulat ing autophagy. Cell Death and Differentiation 16 1040 1052, doi:10.1038/cdd.2009.49 (2009). 111 Kahl, R., Kampkotter, A., Watjen, W. & Chovolou, Y. Antioxidant enzymes and apoptosis. Drug Metabolism Reviews 36 747 762, doi:10.1081/dmr 200033488 (2004). 1 12 Kinnula, V. L. & Crapo, J. D. Superoxide dismutases in the lung and human lung diseases. American Journal of Respiratory and Critical Care Medicine 167 1600 1619, doi:10.1164/rccm.200212 1479SO (2003). 113 Mates, J. M. Effects of antioxidant enzymes in the molecular control of reactive oxygen species toxicology. Toxicology 153 83 104 (2000). 114 Mohr, A., Bueneker, C., Gough, R. P. & Zwacka, R. M. MnSOD protects colorectal cancer cells from TRAIL induced apoptosis by inhibition of Smac/DIABLO release. Oncogene 27 763 774, doi:10.1038/sj.onc.1210673 (2008). 115 Kiningham, K. K., Oberley, T. D., Lin, S. M., Mattingly, C. A. & St Clair, D. K. Overexpression of manganese superoxide dismutase protects against mitochondrial initiated poly(ADP ribose) polymer ase mediated cell death. Faseb Journal 13 1601 1610 (1999). 116 Li, Y. B. et al. Dilated cardiomyopathy and neonatal lethality in mutant mice lacking manganese superoxide dismutase. Nature Genetics 11 376 381, doi:10.1038/ng1295 376 (1995).

PAGE 130

130 117 Lebovitz, R. M. et al. Neurodegeneration, myocardial injury, and perinatal death in mitochondrial superoxide dismutase deficient mice. Proceedings of the National Academy of Sciences of the United States of America 93 9782 9787, doi:10.1073/pnas.93.18.9782 (1996). 118 Van Remmen, H. et al. Life long reduction in MnSOD activity results in increased DNA damage and higher incidence of cancer but does not accelerate aging. Physiological Genomics 16 29 37, doi:10.1152/physiolgenomics.00122.2003 (2003). 119 Church, S. L et al. Increased manganese superoxide dismutase expression suppresses the malignant phenotype of human melanoma cells. Proceedings of the National Academy of Sciences of the United States of America 90 3113 3117, doi:10.1073/pnas.90.7.3113 (1993). 120 C ullen, J. J. et al. Dicumarol inhibition of NADPH: Quinone oxidoreductase induces growth inhibition of pancreatic cancer via a superoxide mediated mechanism. Cancer Research 63 5513 5520 (2003). 121 Malafa, M., Margenthaler, J., Webb, B., Neitzel, L. & Ch ristophersen, M. MnSOD expression is increased in metastatic gastric cancer. Journal of Surgical Research 88 130 134, doi:10.1006/jsre.1999.5773 (2000). 122 Janssen, A. M. L. et al. Superoxide dismutases in gastric and esophageal cancer and the prognostic impact in gastric cancer. Clinical Cancer Research 6 3183 3192 (2000). 123 Kattan, Z., Minig, V., Leroy, P., Dauca, M. & Becuwe, P. Role of manganese superoxide dismutase on growth and invasive properties of human estrogen independent breast cancer cells Breast Cancer Research and Treatment 108 203 215, doi:10.1007/s10549 007 9597 5 (2008). 124 Weinberg, F. & Chandel, N. S. Reactive oxygen species dependent signaling regulates cancer. Cellular and Molecular Life Sciences 66 3663 3673, doi:10.1007/s0001 8 009 0099 y (2009). 125 Saitoh, M. et al. Mammalian thioredoxin is a direct inhibitor of apoptosis signal regulating kinase (ASK) 1. Embo Journal 17 2596 2606, doi:10.1093/emboj/17.9.2596 (1998). 126 Knebel, A., Rahmsdorf, H. J., Ullrich, A. & Herrlich, P. Dephosphorylation of receptor tyrosine kinases as target of regulation by radiation, oxidants or alkylating agents. Embo Journal 15 5314 5325 (1996). 127 Sachsenmaier, C. et al. Involvement of growth factor receptors in the mammalian UVC response. Cell 78 963 972, doi:10.1016/0092 8674(94)90272 0 (1994).

PAGE 131

131 128 Vercesi, A. E., Kowaltowski, A. J., Grijalba, M. T., Meinicke, A. R. & Castilho, R. F. The role of reactive oxygen species in mitochondrial permeability transition. Bioscience Reports 17 43 52, do i:10.1023/a:1027335217774 (1997). 129 Orrenius, S., Gogvadze, V. & Zhivotovsky, B. Mitochondrial oxidative stress: Implications for cell death. Annual Review of Pharmacology and Toxicology 47 143 183, doi:10.1146/annurev.pharmtox.47.120505.105122 (2007). 130 Vaughn, A. E. & Deshmukh, M. Glucose metabolism inhibits apoptosis in neurons and cancer cells by redox inactivation of cytochrome c. Nature Cell Biology 10 1477 U1228, doi:10.1038/ncb1807 (2008). 131 Grassian, A. R., Metallo, C. M., Coloff, J. L., St ephanopoulos, G. & Brugge, J. S. Erk regulation of pyruvate dehydrogenase flux through PDK4 modulates cell proliferation. Genes & Development 25 1716 1733, doi:10.1101/gad.16771811 (2011). 132 Michelakis, E. D. et al. Metabolic Modulation of Glioblastoma with Dichloroacetate. Science Translational Medicine 2 doi:31ra3410.1126/scitranslmed.3000677 (2010). 133 Whitehou.S, Cooper, R. H. & Randle, P. J. Mechanism of activation of pyruvate dehydrogenase by dichloroacetate and other halogenated carboxylic acid s. Biochemical Journal 141 761 774 (1974). 134 Morgan, M. J. & Liu, Z. g. Crosstalk of reactive oxygen species and NF kappa B signaling. Cell Research 21 103 115, doi:10.1038/cr.2010.178 (2011). 135 Yan, S. R. et al. Activation of NF kappa B following de tachment delays apoptosis in intestinal epithelial cells. Oncogene 24 6482 6491, doi:10.1038/sj.onc.1208810 (2005). 136 Lin, D. C. et al. PLK1 Is Transcriptionally Activated by NF kappa B during Cell Detachment and Enhances Anoikis Resistance through Inhi biting beta Catenin Degradation in Esophageal Squamous Cell Carcinoma. Clinical Cancer Research 17 4285 4295, doi:10.1158/1078 0432.ccr 10 3236 (2011).

PAGE 132

132 BIOGRAPHICAL SKETCH Sushama Kamarajugadda was born and brought up in India. She graduated from Holy Mary Girls High School. Her undergraduate degree was in agricultural sciences at Sushama Kamarajugadda moved to United States to pursue higher education. In the fall of & microbiology at University of Central Florida, Orlando. In the summer of 2004, she t of malaria vaccine in transgenic plants using chloroplast transformation technique. summ er, 2006. In the fall of 2006, Sushama Kamarajugadda was admitted into the interdisciplinary program (IDP) in biomedical sciences at the College of Medicine at the projec t. She worked on breast cancer and her main focus was to understand the role of glucose metabolism in promoting anoikis r esistance and tumor metastasis.