Angiopoietin-2 Inhibition by the Investigational Ang-2 Antibody Medi3617 in Renal Cell Carcinoma Induced Angiogenesis

MISSING IMAGE

Material Information

Title:
Angiopoietin-2 Inhibition by the Investigational Ang-2 Antibody Medi3617 in Renal Cell Carcinoma Induced Angiogenesis Therapeutic Implications
Physical Description:
1 online resource (158 p.)
Language:
english
Creator:
Molnar, Nikolett
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Medical Sciences, Physiology and Pharmacology (IDP)
Committee Chair:
Siemann, Dietmar W
Committee Members:
Oh, Suk
Rowe, Thomas C
Law, Brian Keith
Muir, Dave F

Subjects

Subjects / Keywords:
angiopoietin2
Physiology and Pharmacology (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:
Normal 0 false false false EN-US X-NONE X-NONE Angiogenesis, the formation of new blood vessels frompre-existing ones, is a hallmark of cancer. Angiogenesis can be separated intotwo key steps, (a) vascular detabilization mediated by the Angiopoietin/Tie2axis and (b) endothelial cell activation heavily influenced by the VascularEndothelial Growth Factor (VEGF) and its receptor VEGFR-2. The use of VEGF axistargeted anti-angiogenic agents has become a standard of care in many tumorsettings in the last decade. However, their clinical efficacy has been limited.Recently, a new class of anti-angiogenic agents targeting Angiopoetin-2 (Ang-2)has emerged in hopes to circumvent lack of response or resistance seen withVEGF targeted agents. There are still many unanswered questions about thebiology of the Ang/Tie2 axis and vigorous evaluation of agents targeting thisaxis is necessary. The presented research focuses on the preclinical evaluationof an investigational fully human monoclonal antibody, MEDI3617, targetingAng-2 in the renal cell carcinoma setting. We explored the basic principle ofAng-2 dependent vascular destabilization, the loss of endothelial andperi-endothelial cell contacts, and the effect of Ang-2 inhibition. The tumormicroenvironment heavily influences tumor angiogenesis and its effects on theAng-2/Tie2 axis remain controversial. Key microenvironmental factors and theireffect on the Ang-2/Tie2 axis were determined along with the anti-angiogeniceffect of Ang-2 inhibition especially on vascular structure. Real time imagingof tumor microvascular response to both VEGF and Ang-2 targetinganti-angiogenic agents was pursued in the murine dorsal skinfold window chambermodel that revealed limitations to this model. Additionally the anti-tumoreffects of Ang-2 inhibition were explored as well as combinatorial approacheswith VEGF targeting agents both in the angiogenesis and metastatic setting.This work shed light on several important things to consider when using Ang-2targeted agents; future considerations for the use of these agents arediscussed.   /* Style Definitions */ table.MsoNormalTable{mso-style-name:"Table Normal";mso-tstyle-rowband-size:0;mso-tstyle-colband-size:0;mso-style-noshow:yes;mso-style-priority:99;mso-style-parent:"";mso-padding-alt:0in 5.4pt 0in 5.4pt;mso-para-margin:0in;mso-para-margin-bottom:.0001pt;mso-pagination:widow-orphan;font-size:12.0pt;font-family:"Times New Roman","serif";}
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 Nikolett Molnar.
Thesis:
Thesis (Ph.D.)--University of Florida, 2013.
Local:
Adviser: Siemann, Dietmar W.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2013
System ID:
UFE0045740:00001


This item is only available as the following downloads:


Full Text

PAGE 1

ANGIOPOIETIN-2 INHIBITION BY THE INVESTIGATIONAL ANG-2 ANTIBODY MEDI3617 IN RENAL CELL CARCIN OMA INDUCED ANGIOGENESIS: THERAPEUTIC IMPLICATIONS By NIKOLETT MOLNAR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORID A IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013 1

PAGE 2

2013 Nikolett Molnar 2

PAGE 3

To my Mom and Dad who brought me into th is world and supported me with all my endeavors 3

PAGE 4

ACKNOWLEDGMENTS I take this opportunity and express my gratitude to my me ntor Dr. Dietmar Siemann for guiding me throughout the years and sculpting me into the young scientist that I am today. I would like to thank my co mmittee Dr. Thomas Rowe, Dr. Brian Law, Dr. Paul Oh and Dr. Dave Muir for their input throughout my training and diversifying my thinking. I would like to express my many thanks to Stev e McClellan and Craig Monneypenny who have guided me in the earl y phases of live cell imaging and Marda Jorgensen from whom I have learned everyt hing I know about tissue sectioning and immunohistochemistry. A very special thanks to our collaborators Dr. Brian Sorg, Dr. Mamta Wankhede, Dr. Se-woon Choe, and Jennifer Lee I especially thank you Jen for taking the time to teach me the ins and outs of window chamber sur geries, imaging and always being there to help me whenever I needed it; you have been an amazing friend and colleague, cheers to our accomplishments. I would not be here without the love and guidance of my colleagues Divya Sudhan, Sharon Lepler, Veronica Hughes, Chri s Pampo, Dr. Yao Dai and Dr. Lori Rice you all have touched my life in various wa ys and I thank you all for your efforts in shaping the presented research, I am foreve r grateful. Divya, you have been my rock in the lab and I can only hope I have been as helpful to you as you have been to me over the years; keep up t he great work its your turn now! Last but not least I would like to thank my family and friends for their continued support and encouraging words throughout the years. Even though most of you still have no idea what I work on or understand a word when I try to explain, I still love you all for keeping my spirits up when I needed it over the years. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDG MENTS..................................................................................................4LIST OF TABLES............................................................................................................9LIST OF FI GURES ........................................................................................................10ABSTRACT ...................................................................................................................12 CHAPTER 1 INTRODUC TION....................................................................................................14Conventional Canc er Ther apies .............................................................................14Anti-Angiogeni c Ther apy.........................................................................................17 Tumor Vasculature An d Angiogen esis ....................................................................15Angiopoietin/T ie2 Ax is............................................................................................19Angiopoietin-2 In The Tumor Microenvironment .....................................................23Angiopoietin/Tie2 Ta rgeting A gents ........................................................................24Human Monoclonal Angiopoiet in-2 Antibody MEDI3617........................................25Renal Cell Ca rcinoma.............................................................................................25 Motivation And Goal Of Res earch..........................................................................262 INHIBITION OF ENDOTHELIAL/SM OOTH MUSCLE CELL CONTACT LOSS BY THE ANGIOPOIETIN-2 AN TIBODY ME DI3617................................................32Backgroun d.............................................................................................................32Reagent s..........................................................................................................35 Materials And Met hods...........................................................................................35Drug Preparation..............................................................................................35Cell Cult ure.......................................................................................................35Endothelial-Smooth Muscle Cell Co-Culture Sphere Forma tion.......................36Confocal Imaging Of Endot helial-Smooth Muscle Cell Co-Culture Spheres....36Imaging Co-Culture Sphere Form ation And Dest abilizatio n.............................37Human Angiopoieti n-2 ELIS A...........................................................................37Stimulation Of Endothe lial-Smooth Muscle Cell Co -Culture Spheres...............38Result s....................................................................................................................38 Endothelial And Smooth Muscle Cell s Form Co-Cultu re Spheres....................38Ang-2 Inhibitor Impairs Ex ogenous Ang-2 Dependent Sphere Destabilization...............................................................................................38Ang-2 Inhibitor Hinders Endogenous Ang-2 Dependent Sphere Destabilization...............................................................................................39Discussio n..............................................................................................................39 5

PAGE 6

3 IMPLICATIONS OF THE EFFECTS OF TUMOR MICROENVIRONEMNTAL FACTORS ON THE ANG-2 AXIS AND SPHERE DEST ABILIZATION...................48Backgroun d.............................................................................................................48Reagent s..........................................................................................................50 Materials And Met hods...........................................................................................50Cell Cult ure.......................................................................................................51Drug Preparation..............................................................................................51Endothelial-Smooth Muscle Cell Co-Culture Sphere Forma tion.......................52Stimulation Of Endothe lial-Smooth Muscle Cell Co -Culture Spheres...............52Imaging Co-Culture Sphere Form ation And Dest abilizatio n.............................53Human Angiopoieti n-2 ELIS A...........................................................................53Hypoxia............................................................................................................54 Result s....................................................................................................................54Tumor Cells Secrete Minimal Levels Of Ang-2 But Their Conditioned Media Stimulate Ang-2 Secret ion From Endothelial Ce lls And Ang-2 Dependent Destabilization, Ang-2 Inhibitor Impairs Th is Lo ss.........................................54Hypoxia Does Not Affect Tumor Cell Secretion of Ang-2.................................55Tumor Cell Media From Hypoxic Conditions Do Not Affect Endothelial Cell Secretion Of Ang-2 Compared to Tumo r Cell Media From Normal Oxygen Levels ............................................................................................................55VEGF Does Not Significantly Affect A ng-2 Release From Endothelial Cells....56 Endothelial Cell Secretion Of Ang-2 Was Not Altered D ue to Hypoxia.............56 Discussio n..............................................................................................................564 ANGIOPOIETIN-2 INHIBITION EN HANCES THE NORMAL VASCULAR PHENOTYP E..........................................................................................................70Backgroun d.............................................................................................................70Reagent s..........................................................................................................71 Materials And Met hods...........................................................................................71Cell Cult ure.......................................................................................................72Drug Preparation..............................................................................................72Intradermal Angiog enesis Assay......................................................................72Immunohistochem istry......................................................................................73 Result s....................................................................................................................74Ang-2 Inhibitor Impairs Angiogenesi s In Both VHL Normal And Mutated Renal Cell Carc inoma Models .......................................................................74Ang-2 Inhibitor Hinders Both Tumor Grow th And Angiogenesis In Renal Cell Carcinoma Xenogr aft M odel.......................................................................... 74Ang-2 Inhibition Increases Peri-Endothel ial Cell Covered Vessels In Both The Tumor Peripher y And Core....................................................................75Discussio n..............................................................................................................75 6

PAGE 7

5 DORSAL SKINFOLD WINDOW CHAMB ER MODEL TO EVALUATE EARLY TUMOR CELL INDUCED ANGIOGENESIS AND ANTI-ANGIOGENIC THERAP Y...............................................................................................................85Backgroun d.............................................................................................................85Reagent s..........................................................................................................86 Materials And Met hods...........................................................................................86Cell Cult ure.......................................................................................................87Drug Preparation..............................................................................................87Window Chamber Surgery A nd Tumor Init iation ..............................................88Treatment Of Window Chamber Tu mors..........................................................88Hyperspectral Imagi ng......................................................................................89Immunohistochem istry......................................................................................89Serum Angiopoieti n-2 Levels............................................................................90Intradermal Angiog enesis Assay......................................................................90Result s....................................................................................................................91 Human Renal Cell Carcinoma, Caki-2 Growth In The Window Chamber........91VEGF Inhibition Impedes Tumor And Vessel Growth In Caki-2 Window Chamber Tu mors..........................................................................................92Ang-2 Inhibition Does Not Affect Ca ki-2 Tumor Growth In The Window Chamber .......................................................................................................92VEGF And Ang-2 Inhibition In The Intradermal Angio genesis Model...............92Surgery Associated With The Window Chamber Model Leads to An Increase In Serum Ang-2 Lev els...................................................................93Discussio n..............................................................................................................936 ANTI-TUMOR EFFECT OF ANGIOP OIETIN-2 INHBITION AS A MONOAND COMBINATION THERAPY WITH VEGF TARGETED AGENTS..........................107Part A: Anti-Tumor Effect Of Ang-2 M ono-Therapy ..............................................107Materials And Methods ...................................................................................108 Background....................................................................................................107Reagents..................................................................................................108Cell cult ure...............................................................................................108Drug preparation......................................................................................108Xenograft dissoci ation..............................................................................108Growth delay tumor init iation and treat ment.............................................109Lung metastasis tumor init iation and tr eatment........................................109 Results ...........................................................................................................110Ang-2 inhibition does not significa ntly impact renal cell carcinoma xenograft gr owth...................................................................................110Discussio n......................................................................................................111 Ang-2 inhibition does not impact renal cell carcinoma lung metastases..110Part B: Combination Therapy With VEGF Target ed Agents .................................119Background....................................................................................................119Materials And Methods ...................................................................................120 7 Reagents..................................................................................................120

PAGE 8

Cell cult ure...............................................................................................120Intradermal a ssay....................................................................................121 Drug preparation......................................................................................120Immunohistochem istry.............................................................................122 Results ...........................................................................................................122MEDI3617 and Sunitinib combination leads to greater antiangiogenic effect .....................................................................................................122MEDI3617 and Cediranib combination l ead to greater anti-angiogenic effect .....................................................................................................123Discussio n......................................................................................................1257 CONCLUSIONS AND FUTU RE DIRECT IONS....................................................132Summary ..............................................................................................................132Combination With Radiation T herapy................................................................... 136 Current Status Of Angiopoiet in-2/Tie2 Inhibitors...................................................133Combination With Vascula r Disrupting Agents .....................................................138Other Therapeutic Co nsideratio ns........................................................................139Therapeutic Limitations Of Ang-2 Targeted Agents..............................................140 Overall Conc lusions ..............................................................................................141LIST OF REFE RENCES.............................................................................................143BIOGRAPHICAL SKETCH ..........................................................................................158 8

PAGE 9

LIST OF TABLES Table page 1-1 Ang-2/Tie2 axis targeti ng agents and clinical status ...........................................31 9

PAGE 10

LIST OF FIGURES Figure page 1-1 Tumor angi ogenesis. ..........................................................................................28 1-2 Angiogenesis: a tw o step proc ess......................................................................291-3 Therapeutic benefit of Ang-2 target ing...............................................................302-1 Endothelial and smooth muscle cells form co-culture spheres...........................442-2 Exogenous Ang-2 causes sphere destabiliz ation, Ang-2 antibody inhibits this respons e.............................................................................................................452-3 PMA and thrombin are stimulators of Ang-2 secretion from endothelial cells.....462-4 PMA induced endogenous Ang-2 causes sphere destabilization, Ang-2 inhibitor impairs this res ponse. ...........................................................................473-1 Tumor cell involvement in angiogenes is through the Ang2-Tie2 axis................623-2 Tumor cell conditioned media stimulate Ang-2 secret ion from endothelial cells and Ang-2 dependent destabilization, Ang-2 inhibitor hinders this loss......633-3 Tumor cell conditioned media leads to sphere destabiliza tion............................643-4 Tumor cells secrete minimal le vels of Ang-2 under both normoxic and hypoxic condi tions..............................................................................................653-5 Tumor cell conditioned media from hypoxia do not enhance Ang-2 secretion from endothelial cells..........................................................................................663-6 Hypoxia does not enhance tumor cell conditioned media mediated sphere destabilization.....................................................................................................673-7 Hypoxia alone has no significant effe ct on Ang-2 secretion from endothelial cells. ...................................................................................................................683-8 VEGF does not significantly stimulate endothelial cell secretion of Ang-2..........694-1 Intradermal angioge nesis assay.........................................................................784-2 Ang-2 inhibition reduces tumor ce ll induced blood vessel formation..................794-3 Ang-2 inhibition impedes both tumor growth and angiogenesis ..........................804-4 Ang-2 inhibition reduces t he number of tumo r vessels.......................................81 10

PAGE 11

4-5 Normal vasculature is composed of endothelial and peri-endothelial cells.........824-6 Ang-2 inhibition leads to increased normal vascular phenotype in the tumor peripher y............................................................................................................834-7 Ang-2 inhibition leads to increas e normal vascular phenotype within the tumor..................................................................................................................845-1 Human renal cell carcinoma, Caki-2 tumor growth in the window chamber model. .................................................................................................................995-2 VEGF inhibition impedes tumor and ve ssel growth in Caki -2 tumors...............1005-3 VEGF inhibition impedes vessel gr owth in Caki-2 tumors................................1015-4 Ang-2 inhibition does not a ffect Caki-2 tumors.................................................1025-5 Ang-2 inhibition led to r educed tumor vascu lature............................................1035-6 Ang-2 inhibition leads to increased per i-endothelial cell coverage in the tumor periphery. .........................................................................................................1045-7 Ang-2 and VEGF inhibition in the intraderma l assay. .......................................1055-8 Window chamber surgery leads to increase serum Ang2 levels .....................1066-1 Ang-2 inhibition does not affect renal cell carcinoma tu mor growth..................1166-2 Early treatment with Ang-2 inhibitor does not affect renal cell carcinoma tumor growth.....................................................................................................1176-3 Ang-2 inhibition does not affect renal ce ll carcinoma lung metastasis growth..1186-4 Targeting two key st eps in angio genesis. .........................................................1276-5 Ang-2 and VEGF combination therapy with Sunitinib reduce tumor volume and vessel num ber...........................................................................................1286-6 Ang-2 and VEGF combination ther apy with Sunitinib impair tumor angiogenesis....................................................................................................1296-7 Ang-2 and VEGF combination therapy with Cediranib reduce tumor volume and vessel num ber...........................................................................................1306-8 Ang-2 and VEGF combination ther apy with Cediranib impair tumor angiogenesis. ...................................................................................................1317-1 Radiation and VDA comb ination modalit ies.. ....................................................142 11

PAGE 12

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy ANGIOPOIETIN-2 INHIBITION BY THE INVESTIGATIONAL ANG-2 ANTIBODY MEDI3617 IN RENAL CELL CARCIN OMA INDUCED ANGIOGENESIS: THERAPEUTIC IMPLICATIONS By Nikolett Molnar August 2013 Chair: Dietmar W. Siemann Major: Medical Sciences Physiology and Pharmacology Angiogenesis, the formation of new blood vessels from pre-existing ones, is a hallmark of cancer. Angiogenes is can be separated into two key steps, (a) vascular destabilization mediated by the Angiopoietin/T ie2 axis and (b) endothelial cell activation heavily influenced by the Vascular Endothelial Gr owth Factor (VEGF) and its receptor VEGFR-2. In the last decade the use of VEG F axis targeted anti-angiogenic agents has become a common therapy in several oncologic se ttings. However, their clinical efficacy has been limited. Recently, a new class of anti-angiogenic agents targeting Angiopoietin-2 (Ang-2) has emerged in hopes to circumvent lack of response or resistance seen with VEGF targeted agents. There are still many unanswered questions about the biology of the Ang/Tie2 axis and vi gorous evaluation of agents targeting this axis is necessary. The presented research fo cuses on the preclinical evaluation of an investigational fully hum an monoclonal antibody, MEDI3617 targeting Ang-2 in the renal cell carcinoma setting. We explored the basic principle of Ang-2 dependent vascular destabilization, the loss of endot helial and peri-endothelial cell contacts, and the effect of Ang-2 inhibition. The tu mor microenvironment heavily influences tumor 12

PAGE 13

13 angiogenesis and its effects on the Ang-2/Tie2 axis remain controversial. Key microenvironmental factors and their effect on the Ang-2/Tie2 axis were determined along with the anti-angiogenic effect of Ang-2 inhibition especially on vascular structure. Real time imaging of tumor microvascula r response to both VEGF and Ang-2 targeting anti-angiogenic agents was pursued in the murine dorsal skinfold window chamber model that revealed limitations to this model Additionally, the anti-tumor effects of Ang2 inhibition were explored both in the prim ary and metastatic tumor setting. Finally, combinatorial approaches with VEGF targeting agents were evaluated as antiangiogenic modalities. The presented work shed light on several important considerations when using A ng-2 targeted agents and future directions for the use of these agents are discussed.

PAGE 14

CHAPTER 1 INTRODUCTION Cancer remains the second leading cause of death in the United States affecting millions of lives each year. While advances in diagnostics and treatment have led to a 3.4-fold increased patient survival over the last 40 years, many disease settings remain without any progress, high rate of recurrence or fatality fr om metastatic disease (de Moor et al., 2012). Increased knowledge of cancer biology has led to the identification of important factors involved in the development of cancer and the discovery of a plethora of potential therapeutic targets. However, with increased knowledge the growing complexity of this disease is unfolded; as one door closes many more open in the fight against cancer. Nevertheless our efforts bri ng us closer to better management of this disease one step at a time. Conventional Cancer Therapies Conventional cancer therapies such as chemotherapy and radiation therapy target rapidly proliferating cells by inducing DNA damage. These therapies are used in several settings; they can be used with a cura tive intent or palliative treatment to manage the pain associated wit h advanced and metastatic disease. For curative treatments they can be used ei ther (a) as neoadjuvant, before surgical resection of a tumor to reduce the size or im pair the growth of the tumor or as (b) adjuvant treatment with surgery or chemotherapy and radiation therapy combined. Chemotherapeutic agents target rapidly proliferating cells in a cell cycle specific or independent manner (Siu and M oore, 2005). These therapies can also be highly toxic to normal tissues yielding a narrow therapeu tic window. Tumor resistance to these agents stems both from Gompertzian growth of and impaired del ivery to the tumor cells 14

PAGE 15

due to irregular tumor vasculature (C ole and Tannock, 2005; Minchinton and Kyle, 2011; Minchinton and Tannock, 2006) Radiation therapy provides another means to treat cancers. Ionizing r adiation causes DNA damage resulting in tumor cell death (Bristow and Hill, 2005). This type of therapy also has detrimental effects on normal tissues and the therapeutic outcome can be greatly affected by the tumor microenvironment (Bristow and Hill, 2008; Gia ccia and Hall, 2006a). Hypoxic cells are resistant to radiation and pose a real challen ge in the treatment of cancer (Bristow and Hill, 2008; Giaccia and Hall, 200 6b). Efforts to improve oxy genation in the tumor have shown some improvement in treatment outcome (Overgaard and Horsman, 1996). Targeting the tumor vasculature offers a different approach to combat cancer and combinations with chemot herapy and radiati on therapy have shown that these treatments may be complem entary (Horsman and Siem ann, 2006; Siemann, 2011). Tumor Vasculature And Angiogenesis Angiogenesis, the formation of new blood vessels from pre-existing ones is a crucial process in normal vascular develop ment and physiological conditions such as wound healing, reproduction and the menstrual cycle (Papetti and Herman, 2002). The importance of angiogenesis in pathological conditions has been well established (Folkman, 2007). Angiogenesis in physiological conditions is tightly regulated by a balance of pro-and anti-angi ogenic factors, in disease setti ngs this balance is tipped to favor pro-angiogenic factors (Folkman, 2007; Papetti and Herman, 2002). A growing tumor cannot sustain its growth without the initiation of angiogenesis (Folkman, 1971; Goldmann, 1908) (Figure 1-1). As the tu mor grows, cells existing further than the oxygen diffusion distance (70-100 m) from blood vessels, become 15

PAGE 16

hypoxic (Hanahan and Folkman, 1996). Hypoxic cells respond to low oxygen levels by upregulating genes involved in me tabolism and angiogenesis in order to survive (Betof and Dewhirst, 2011; Dewhirst, 2006). Hypoxia I nducible Factor (HIF) is stabilized under low oxygen conditions and triggers transcripti on of pro-angiogenic factors such as Vascular Endothelial Growth Factor (VEGF), a key activator of angiogenesis (Charlesworth and Harris, 2008). Fibroblast grow th factor (bFGF), Notch and delta-like ligand 4, Interleukin-8 (IL-8), heparanase, and matrix metallo-protease 2 (MMP-2) are also important pro-angiogenic factors in tu mor angiogenesis (Ferrara, 2010; Papetti and Herman, 2002). Tumor cells secrete these grow th factors and stimulate endothelial cells to form new vasculature towards the tumo r (Leung, 1989; Shibuya, 2008). As the tumor grows, angiogenesis is continuously activated to sustain tumor growth and the vasculature expands (Figure 1-1). Tumor vasculature is distinct from normal vasculature. Normal vessels are quiescent wi th minimal endothelial cell proliferation, tight endothelial-endothelial ce ll and endothelial-periendothelial mural cell (e.g. smooth muscle cells and pericytes) contacts to regulate vessel permeability (McDonald, 2008 ). Tumor vessels, on the other hand, have highly pr oliferative endothelial cells with loose to minimal endothelial-endothelial and endothelialperi-endothelial cell c ontacts yielding a leaky vascular phenotype (Minchinton an d Kyle, 2011; Vaupel, 2006). Yet despite being highly active, the leaky tu mor vasculature is unable to provide the tumor with its oxygen and nutritional needs leading to tu mor microenvironments t hat have proven to be problematic in the treatment of cancer (Minchinton and Kyle, 2011; Minchinton and Tannock, 2006). 16

PAGE 17

Anti-Angiogenic Therapy Anti-angiogenic therapy target s the tumor vasculature, indirectly affecting tumor growth. It was proposed by the late Judah Folkman and Julie Denekamp over 40 years ago, who hypothesized that depriving the tumo r of oxygen and nutri ents, by targeting the tumor vasculature could have ther apeutic benefits (Denekamp, 1982; Denekamp, 1984; Folkman, 1971). Identificat ion of growth factors and signaling pathways important in angiogenesis resulted in the production of inhibitors and in the past decade antiangiogenic agents have become part of the standard of care in several disease settings (Folkman, 2007; Meadows and Hu rwitz, 2012; Young and Reed, 2012). To date, there are thirteen agents with specif ic or non-specific anti-angi ogenic activity (antibody and small molecule inhibitors) that are approved by the FDA for trea tment of advanced and metastatic renal cell carcinoma (Bevacizumab, Axitinib, Pazopanib,Sorafenib, Sunitinib, Temsirolimus, Everolimus, Interferon ) metastatic colorectal cancer (Bevacizumab, Regorafenib), gastrointestinal cancer (S unitinib), non small cell lung cancer (Bevacizumab), hepatocellular carcinoma (Sor afenib), thyroid cancer (Cabozantinib, Vandetanib), pancreatic neuroendocrine tumors (Sunitinib, Everolimus), glioblastoma (Bevacizumab), multiple myeloma (Tha lidomide, Lenalidomide), astrocytoma (Everolimus); leukemia (Interferon ), melanoma (Interferon ) and AIDS-related Kaposi sarcoma (Interferon ). Many other agents are in clinic al trials for a wide variety of cancers. Bevacizumab (Avastin), a monocl onal antibody targeting the VEGF ligand, has proven to be most successful thus far and is approved for the treatment of glioblastoma, metastatic colorectal, non small cell lung ca ncer, renal cell carcinoma and until recently metastatic breast cancer (Hurwitz et al., 2004; Montero et al., 2012; Pollack, 2011). 17

PAGE 18

Although these agents, in general, alone show minimal anti-tumor effects they have been shown to complement conventiona l cancer therapies in reducing tumor growth (Horsman and Siemann, 2006; Kerbel, 2006; Siem ann, 2011). Combination of such treatments resulted in a slightly wider therapeutic window for both chemotherapy and radiation therapy enhancing tumor cell k ill. Its been proposed that anti-angiogenic treatment yields a window of vascula r normalization that leads to better chemotherapeutic drug delivery to the tumo r core and overall r educed hypoxia within the tumor resulting in better cell kill by radi ation therapy (Bottsford-Miller et al., 2012; Jain, 2005a; Siemann, 2011). Anti-angiogenic therapy does have side effects, most commonly hypertension and occasional ble eding along with impaired physiological angiogenesis such as wound he aling, reproduction and fe tal development (Elice and Rodeghiero, 2012; Hayman et al., 2012; Pralhad et al., 2003). These side effects limit the population of patients that can be treated with such therapy as well as affect the scheduling of treatment for appr oved patients. However, removal of these agents from the patients therapeutic regim ent quickly restores physiological angiogenesis, making it easy to remove or resume treatment before or after major surgeries. Unfortunately, tumor resistance to thes e agents can develop and pose difficulties in successful treatment, lack of overall pat ient survival led to the removal of Bevacizumab for the treatment of metastatic breast cancer (Ebos et al., 2009; Loges et al., 2010; Montero et al., 2012; Pollack, 2011). Tumors can dev elop acquired resistance to anti-angiogenic therapies by upregulation of pathways in angiogenesis that are not inhibited by the therapy (Bergers and Hanahan, 2008). Ev en though VEGF is a major driver of angiogenesis there are other growth factors su ch as basic fibroblast growth 18

PAGE 19

factor (bFGF), interleukin-8 (IL-8), placent al growth factor (PlGF), platelet derived growth factor (PDGF), and insulin like growth factor (IGF) that c an compensate in the absence of VEGF pathway activation (B ergers and Hanahan, 2008). Due to such resistance and compensatory mechanisms ther e is a need for superior anti-angiogenic modalities to fulfill the potential of this va scular targeted therapeutic strategy (Loges et al., 2010). Angiopoietin/Tie2 Axis Tyrosine Kinase with Ig and EGF homology domains-2 (Tie2) and its ligands Angiopoietin-1 (Ang-1), Angiopo ietin-2 (Ang-2), Angiopoieti n-3 and Angiopoietin-4 (Ang4) were identified in the last fifteen y ears (Davis et al., 1996; Partanen et al., 1992; Ribatti et al., 2000; Schnurch and Risau, 1993; Valenzuela et al., 1999). Ang-1 and Ang-2 are the best characteriz ed of the four liga nds and their interactions with Tie2 receptor comprise an import ant endothelial cell-specific receptor tyrosine kinase signaling system in angiogenesis (Augusti n et al., 2009; Loughna and Sato, 2001; Thomas and Augustin, 2009). Tie2 is mainly expressed by vascular and lymphatic endothelial cells and is a transmembrane tyro sine kinase receptor with its tyrosine kinase domain in the cytoplasm (Partanen et al., 1992; Schnurch and Risau, 1993). Tie1 is a receptor that is currently considered an orphan re ceptor, neither angiopoietin ligands are known to directly bind; howev er its interaction with angiopoietins and Tie2 has been shown to be important in lymphatic vascular development (D'Amico et al., 2010; Song et al., 2012; Yuan et al., 2007). Structurally, Ang-1 and Ang-2 ligands have two domains, the N-terminal coiled-coil domain responsible for homo-oligomerization of the ligands into trimers, tetramers and pentam ers and a C-terminal fibrinogen like domain responsible for receptor binding (Davis et al., 2003; Procopio et al., 1999). 19

PAGE 20

Oligomerization of the ligands has been shown to be important for receptor activation but not binding (Procopio et al., 1999). A ng-1 has 498 amino acids and Ang-2 has 496 amino acids, the two ligands share 60% amino acid sequence homology (Davis et al., 1996; Maisonpierre et al., 1997; Procopio et al., 1999). Ang-1 and Ang-2 are secreted proteins t hat interact with Tie2 either in a paracrine (Ang-1) or autocrine (Ang-2) m anner; Ang-1 is expressed and secreted by peri-endothelial mural cells (smooth muscle ce lls, pericytes) while Ang-2 is expressed and secreted by endothelial cells (Maisonpierre et al., 1997). Both angiopoietins bind Tie2 with similar affinities and on the same si te on the receptor at the IgG-like and EGFlike domains (Barton et al., 2005; Fiedler et al., 2003). These ligands, however, have opposing functions. Ang1-Tie2 signaling contro ls vessel quiescence, while Ang2-Tie2 association allows for vessel plasticity (Saharinen and Alitalo, 2011) (Figure 1-2A). Ang1-Tie2 presence and interaction are essential for vessel maturation in embryonic vasculogenesis while Ang-2 expression is expendable for normal embryonic development as shown by loss-of-function st udies (Gale et al., 2002; Maisonpierre et al., 1997). The absence of bot h Ang-1 and Tie2 lead to embryonic lethality due to aberrant vessel remodeling and maturation whil e the over expression of Ang-2 yields similar phenotypes; these genetic studies s upport the agonistic, non-redundant role of Ang-1 and antagonistic role of Ang-2 with Tie2 re ceptor interactions (Maisonpierre et al., 1997; Sato et al., 1995; Suri et al., 1996; Thomas and Augustin, 2009). In the adult, Ang-1 is found in many types of tissues and is constitutively secreted in low levels throughout the body (Maisonpierre et al., 1997). Ang-2 is found in tissues that undergo vessel remodeling and is secreted by endothelial cells at sites of active 20

PAGE 21

angiogenesis (Maisonpierre et al., 1997). An g1-Tie2 maintains a stable vasculature through mediating the tight interaction of endothelial cells with mural cells while Ang2Tie2 functions in their destabilization (Augustin et al., 2009; Maisonpie rre et al., 1997). It is believed that Ang-1/Tie2 interaction l eads to downstream signaling through Rho and mDia inhibiting VEGF/VEGFR2 mediated vasc ular permeability through src while Ang2/Tie2 interaction leads to loss of recept or activation and intracellular signaling (Augustin et al., 2009; Gavard et al., 2008; Thomas and A ugustin, 2009). However, it has been shown that Ang-2 can act as a Tie2 agonist in a context dependent manner mainly in the lymphatics and with Tie-2 expressing tumor associated macrophages (Gale et al., 2002; Mazzieri et al., 2011; Venneri et al., 2007). It appears that the context dependent role of Ang-2 as an agonist or antagonist is mediated by the interaction of Tie2 with Tie1 receptor. Ev en though none of the angiopoietin ligands bind to Tie1 the receptor can heterodimerize with Tie2 mainly affecting the role of Ang-2. Ang-1 alwa ys interacts with Tie2 homodimers however Ang2 can interact with Tie2 homodimers and Tie1/ Tie2 heterodimers. Ang-2 binding to Tie2 homodimers leads to agonistic interactions similar to Ang-1/Tie2 interactions; this is mainly seen in lymphatic vasculature where Tie1 receptor is sparsely expressed (Gale et al., 2002; Song et al., 2012). However in blood endothelial cells Tie1 is commonly expressed leading to an increased Tie1/Tie2 heterodimer interaction with Ang-2 resulting in an antagonistic role of Ang-2 and lack of intracellular signaling leading to vessel destabilization (Kim et al., 2006; S eegar et al., 2010). Recently it has been shown that the ability of Ang-2 to intera ct with both Tie2 homodimers and Tie1/Tie2 heterodimers is a difference of three amino ac ids (proline, glutamine, arginine) in the 21

PAGE 22

protein sequence outside the re ceptor binding site compared to Ang-1 (threonine, alanine, glycine) (Yu et al., 2013). Based on current knowledge of the angiop oietin/Tie2 axis, angiogenesis can be separated into two key steps (a) destabiliza tion of normal vasculature mediated by the Ang-2/Tie2 axis followed by (b) endothelial cell activation by pro-angiogenic factors such as VEGF (Figure 1-2). In the abs ence of pro-angiogenic factors destabilized vasculature has been shown to regress due to endothelial cell apoptosis (Holash et al., 1999; Korff et al., 2001). The angiopoietin ligands are also consider ed important factors in inflammation, Ang-1 as anti-inflammatory and Ang-2 pro-in flammatory (Fiedler and Augustin, 2006; Fiedler et al., 2006). Ang-1 signaling through Tie2 leads to repression of nuclear factorB (NFB ) while Ang-2 interaction with Tie2 acti vates it (Augustin et al., 2009). Ang-1 dependent phosphorylation of Tie2 decreases the synthesis of Ang-2 in endothelial cells by activating the PI3K/Akt pathway and i nactivating the FOXO1 transcription factor (Thomas and Augustin, 2009; Tsigkos et al., 2006) However, Ang-2 is stored in WeibelPalade Bodies and can be quickl y secreted upon microenvironmental changes (Dixit et al., 2008; Fiedler et al., 2004). There are various factors that can lead to Ang-2 secretion from endothelial cells such as hist amine and thrombin that play a role in vasodilation and recruitment of inflammatory cells and cl otting respectively in wound healing (Fiedler et al., 2004; Huang et al., 2002). Hypoxia is another potent upregulator of Ang-2 levels in endothelial cells and ma y lead to Ang-2 release mediated through the COX2-PGE2 pathway ra ther than the HIF-1 dependent route; the inflammatory mediator COX2 is regularly seen in canc erous tissue (Leahy et al., 2000). Finally, 22

PAGE 23

sphingosine-1 phosphate (SP1), secreted by immune cells and platelets, can also lead to Ang-2 release from endothelial cells (J ang et al., 2009). Elevated Ang-2 levels in tumor vasculature have been observed in many cancers and seem to correlate with poor disease prognosis (Ahmad et al., 2001; Ba ch et al., 2007; Bhaskar et al., 2013; Currie et al., 2002; Detjen et al., 2010; Diaz-Sanchez et al ., 2013; Engin et al., 2012; Goede et al., 2010; Helfrich et al., 2009; Imanishi et al., 2007; Lind et al., 2005; Mitsuhashi et al., 2003; Moon et al., 2006; Stratmann et al., 1998; Sugimachi et al., 2003; Tait and Jones, 2004; Tanaka et al., 2002; Tsutsui et al., 2006; Xia et al., 2012; Yamakawa et al., 2004; Zagzag et al., 1999). Angiopoietin-2 In The Tumor Microenvironment Elevated Ang-2 levels have been associat ed with advanced dise ase, progression and poor prognosis in several disease settings such as astrocytoma, glioblastoma, colon, gastric, colorectal, breast, prostate kidney, hepatocellular cancers as well as multiple myeloma, melanoma, and neuroend ocrine tumors (Ahmad et al., 2001; Bhaskar et al., 2013; Currie et al., 2002; Detj en et al., 2010; Engin et al., 2012; Goede et al., 2010; Helfrich et al., 2009; Lind et al ., 2005; Moon et al., 2006; Stratmann et al., 1998; Sugimachi et al., 2003; Tsutsui et al., 2006; Xia et al., 2012; Yamakawa et al., 2004; Zagzag et al., 1999). Ang-2 serum levels significantly increase in patients compared to healthy individuals and pat ients with more advanced disease show significantly higher levels of Ang-2 com pared to patients with earlier stage disease (Detjen et al., 2010; Helfrich et al., 2009; Lind et al., 2005; Pa rk et al., 2007; Sie et al., 2009). The expression of Ang-2 in chronic ly mphocytic leukemia has been shown as a pro-survival phenotype (Xia et al., 2012). Seve ral reports also implicate an integrin Ang-2 interaction in metastatic breast and gl ioma (Carlson et al., 2001; Hu et al., 2006; 23

PAGE 24

Imanishi et al., 2007). Furthermore, in t he presence of Ang-2 tumor associated macrophages (TAMs) express Tie2 and show pro-angiogenic activity (De Palma and Naldini, ; De Palma and Nald ini, 2011). Ang-2 is not onl y a pro-angiogenic but also a pro-inflammatory factor and elevated Ang-2 serum levels correlate with severity of sepsis (systemic inflammatory response to in fection) that often leads to multiple organ dysfunction syndrome, an issue commonl y found in emergency and trauma patients (David et al., 2012; Gallagher et al., 2008; van der Heijden et al., 2009). The presence and importance of Ang-2 in the tumor micr oenvironment and its role in angiogenesis has made this ligand an attractive target as an anti-angiogenic mo dality in cancer therapy. Angiopoietin/Tie2 Targeting Agents In the last few years there has been tr emendous interest in targeting the Ang2/Tie2 axis in angiogenesis in hopes to ci rcumvent the lack of response or adaptive resistance that has been seen in the clinic to VEGF targeted agents. Currently there are no small molecule inhibitors that are selective to Tie2 in clinical development, however several agents in clinical trials do have some activity against the Tie2 axis (Gerald et al., 2013). There are two pepti de-based agents, CVX-060 and Trebananib (AMG386) that are in phase II and III clinical trials respec tively. Trebananib is the furthest in clinical development and targets both Ang-1 and A ng-2, which has raised some questions about the proper way to target the Ang/Tie axis since Ang-1 is considered to be the vascular stabilizer (Coxon et al., 2010; Doi et al ., ; Mita et al., 2010 ; Neal and Wakelee, 2010; Robson and Ghatage, 2011). In gener al Trebananib has not show any progression free survival (PFS) benefits in renal cell carcinoma or breast cancer, however in ovarian cancer ther e seem to be positive anti-tumor effects (Rini et al., ; 24

PAGE 25

Robson and Ghatage, 2011). Several antibodies selectively targeting Ang-2 are being evaluated in phase I clinical trials. For more det ailed information refer to Table 1-1. Human Monoclonal Angiopoietin-2 Antibody MEDI3617 MEDI3617 is a fully human IgG1 kappa monoclonal antibody produced via XenoMouse Technology by MedImmune, LLC. MEDI3617 binds Ang-2 at the brinogen like domain with high selectivity (KD value = 42 pM) and potently blocks Ang-2 binding to the Tie2 receptor (EC = 0.510 nM) (Figure 1-3). Initial preclinical studies have shown antitumor effects in human colon, ovarian, renal and hepatocel lular xenografts (Leow et al., 2012). In preclinical murine models Ang-2 is provided by the mouse endothelium, however mouse and human Ang2 are 85% homologous in amino acid sequence and the binding site of MEDI3617 to Ang-2 is cons erved in the two species therefore can be used both in murine and human treatments (Leow et al., 2012). MEDI3617 is a more potent Ang-2 binding antibody than its precursor 3.19.3 (KD value = 86 pM), which previously had shown vascular an d antitumor effects in several tumor models (Brown et al., 2010). Currently this agent is being evaluated in phase I clinical trials in patients with advanced solid malignancies. Throughout the presented work MEDI3617 will be referred to as the Ang-2 inhibitor. Renal Cell Carcinoma Renal cell carcinoma accounts for about 3% of adult malignancies and 90% of kidney cancers and is a highly vascularized disease driven by VEGF (Curti et al., 2013; Vanchieri and Williams, 2012) Von-Hippel Lindau (VHL) dis ease accounts for 40% of renal cell carcinoma in patients, however the VHL gene is commonly mutated, greater than 90%, in nonhereditary tumors as well (Cur ti et al., 2013; Vanchieri and Williams, 2012). Mutated VHL causes a constitutive ac tivation of the hypoxia inducible factor 25

PAGE 26

(HIF-1 ) leading to high VEGF secretion fr om tumor cells and promotion of angiogenesis even in the absence of hypoxia in the tumor. Patients with renal cell carcinoma show low response to chemotherapeutic agents due to multidrug resistance mediated by p -glycoprotein that is highly express ed by cells (Curti et al., 2013). Furthermore these tumors are rarely sensitive to radi ation therapy. Anti-angiogenic agents have shown beneficial patient response and are now commonly used as both first and second line treatment in renal cell carcinoma. Current ly, eight out of the eleven FDA approved anti-angiogenic ag ents (Bevacizumab, Sorafenib, Sunitinib, Axitinib, Pazopanib, Temsirolimus, Everolimus and Interferon ) are used in renal cell carcinoma making this disease setting by far the most common for anti-angiogenic agents. Motivation And Goal Of Research Anti-angiogenic therapy has become a standard of care in several solid tumors. Even though this therapeutic modality holds promise in pat ient care, there have been several limitations regarding la ck of patient response or occu rrence of patient resistance to currently approved agents in the clinic Our current knowledge of the complex process of angiogenesis and the tumor microenvironment has led to alternative therapeutic strategies. Angiopoietin/Tie2 axis targeted agents mark a new class of antiangiogenics in hopes to eradicate the observed patient resistance with the currently used VEGF targeted agents as well as co mplement conventional therapies. The current research particularly focuses on the monoclonal antibody that is highly specific to Ang-2 (MEDI 3617). At the time this dissert ation was started there were much fewer Ang-2 targeting agents in dev elopment. Mainly Trebananib (AMG386), the Ang-1/Ang-2 peptibody was in phase I clinical trials. The high Ang-2 specificity of 26

PAGE 27

MEDI3617 deemed this antibody superior to Tr ebananib; based on the basic biology of the Ang/Tie2 axis there were many questions concerning the dual inhibition of Ang-1 and Ang-2. Many, including myself, believe that solely Ang-2 should be targeted in this axis. The general concept is that t he tumor microenvironment has high Ang-2 expressions which out-compete Ang-1 for Tie2 binding leading to vascular destabilization. Using an Ang2 neutralizing antibody we hoped to eliminate Ang-2 from the tumor microenvironment and allow for increased Ang-1/ Tie2 binding. This would promote a normal, quiescent vascular phen otype and thus reduce tumor angiogenesis (Figure 1-3). We aimed to ex plore the therapeutic implicat ions of Ang-2 inhibition in renal cell carcinoma, a highly angiogenic disease where anti-angiogenic agents are used both as first and second line treatments. Over the course of our studies we have found benefits and shortcomings to this anti-angiogenic modality that is discussed throughout this work. 27

PAGE 28

Figure 1-1. Tumor angiogenesis. (A) As a tu mor develops it will reach a certain size where cells inside the tumor receive lower levels of oxygen (hypoxia). (B) Hypoxia stabilizes the transcription factor Hypoxia Inducible Factor (HIF-1 ) that in turn upregulates genes involved in angiogenesis. (C) In the tumor microenvironment, endothelial cells fr om nearby vessels secrete Angiopoietin-2 (Ang-2) and tumor cells secrete pro-angiogenic factors to stiumulate the growth of new vessels from nearby normal vasculature. (D) As the tumor develops and grows so does the vasculature with it with the constant stimulation of endothel ial cells to proliferate. 28

PAGE 29

Figure 1-2. Angiogenesis: a two step process. (A) Normal vessels are composed of an endothelial cell layer surrounded by a periendothelial cell layer (e.g. smooth muscle cells or pericytes). Quiescent vessels are non-proliferative and cellcell contacts are maintained by para crine Angiopoietin-1 (Ang-1) and Tie2 receptor interactions. During angiogenesis Angiopoietin-2 (Ang-2) is secreted from endothelial cells and the autocrine in teraction of Ang-2 and Tie2 receptor lead to disruption of cell-cell intera ctions (vessel destabilization). (B) Endothelial cells are exposed to pro-angi ogenic factors (e.g. VEGF, bFGF) in the tumor microenvironment and are activat ed to proliferate, migrate and from tubes of new vessels. 29

PAGE 30

Figure 1-3. Therapeutic benefit of Ang2 targeting. Angiopoietin ligands and Tie2 receptor are responsible for either ma intaining a quiescent vasculature (no endothelial cell prolifer ation) via Ang-1-Tie2 interact ion (A) or destabilizing the vasculature (disrupting cell-cell contacts) via Ang-2-Tie2 interaction (B). Ang2 is released from endothelial cells during an angiogenic stimulus, Ang-2 inhibitor, MEDI3617, binds to the li gand and inhibits its binding to Tie2 receptor (C). By removing Ang-2 fr om the microenviron ment the active angiogenic phenotype of tumors should be reduced (D). 30

PAGE 31

Table 1-1. Ang-2/Tie2 axis targeting agents and clinical status. There is considerable in terest to target the Ang-2/Tie2 axis in tumor angiogenesis. Agent Type Target Stage of Development Company Trebananib (AMG386) Peptibody Ang1, Ang2 Phase III Amgen Regorafenib (BAY 73-4506) Small molecule Tie2, VEGFR1-,2-,3-, Kit, Ret, Raf, PDGFR Phase III Bayer CVX-060 Antibody Ang-2 Phase II Pfizer MGCD265 Small molecule Tie2, c-met, VEGFR1-,2,3Phase I/II MethylGene Inc. MEDI3617 Antibody Ang-2 Phase I MedImmune LLC. CEP-11981 Small molecule Tie2, VEGFR1-, 2-, 3Phase I Cephalon REGN910 Antibody Ang-2 Phase I Regeneron Pharmaceuticals, Inc AMG780 Antibody Ang-2 Phase I Amgen ARRY-614 Small molecule Tie2, VEGFR-2, p38, Abl Phase I ArrayBioPharma DX-2240 Antibody Tie1 Preclin ical Sanofi-Aventis 31

PAGE 32

CHAPTER 2 INHIBITION OF ENDOTHELIAL/SMOOTH MUSCLE CELL CONTACT LOSS BY THE ANGIOPOIETIN-2 ANTIBODY MEDI3617 Chapter 2 explores the use of an endothelial and smooth muscle cell co-culture model to evaluate Ang-2 dependent cell-cell contact loss (destabilization) real time and to evaluate the effect of Ang-2 inhibition using the Ang-2 inhibitor. Studies show the formation and orientation of co-cultu re spheres, the Ang-2 dependent sphere destabilization in the presence of bot h exogenous and endogenous Ang-2 and the subsequent inhibition of this cell-cell cont act loss in the presence of the inhibitor. Background Normal vasculature consists of a single layer of endothelial cells surrounded by peri-endothelial mural cells (smooth muscle cells and pericytes), which form stable vessel networks via tight associations th rough adhesion molecules such as Nand VECadherin (Armulik et al., 2005; Dejana, 2004) Vascular mural cells are important regulators of vascular development, stabiliz ation, maturation and remodeling; abnormal interactions of endothelial and peri-endothelial cells account s for several pathological conditions such as tumor angiogenesis, di abetic microangiopathy, ectopic tissue calcification and stroke and dementia syndr ome (Armulik et al., 2005). Peri-endothelial cells are a very diverse class of cells that may have different phenotypes in the various vascular beds (e.g. artery, arteriole, vein, venule, or capillary); based on the demands of nearby organs, the extent of peri-endothelial cell coverage of the endothelium changes (Armulik et al., 2005). The dynamic nature of endothelial cells al lows for rapid vascular responses to environmental changes that occur under no rmal and pathological conditions (Cleaver and Melton, 2003; Rondaij et al., 2006). Angi ogenesis requires two distinct events 32

PAGE 33

consisting of the disruption of the vascular structure followed by activation of the endothelium. First, the tight association between endothelial-endothelial and endothelial-peri-endothelial cells is disrupted in order to a llow for vessel permeability and exposure of endothelial cells to proangiogenic cytokines. Second, pro-angiogenic factors activate the endothelium to proliferat e, migrate and establis h tubular networks to form new blood vessels. The angiopoietin/Tie2 system is central to the initial phase of angiogenesis, the disruption of endothelial and peri-endothelial cell interactions (Maisonpierre et al., 1997; Scharpfenecker et al., 2005). A ngiopoietin-1 (Ang-1) and Angiopoietin-2 (Ang-2) are secreted proteins that interact with the Tie2 receptor either in a paracrine (Ang-1) or autocrine (Ang-2) manner; Ang-1 is expres sed and secreted by peri-endothelial mural cells while Ang-2 is expre ssed and secreted by endothelial cells (Augustin et al., 2009). Although both angiopoietin s bind Tie2 with similar affiniti es on the same receptor site, they have opposing functions (Davis et al., 2003; Fiedler et al., 2003); Ang-1 stabilizes and Ang-2 destabilizes the vasculature (Augustin et al., 2009; Davis et al., 1996). In the adult, Ang-1 is found in tissues throughout the body and continuously secreted at low levels (Davis et al., 1996; Saharinen and Alitalo, 2011). Ang-1-Tie2 association leads to tyro sine kinase phosphorylation of the Tie2 receptor and downstream signaling maintaining tight adhesion molecule interactions in the vascular cellular networks, between endothelial and peri-endothelial cells (Fukuhara et al., 2008; Gavard et al., 2008; Sahari nen and Alitalo, 2011). Ang-2 is produced by endothelial cells and stored in Weibel-Palade Bodies (WPB) of endothelial cells (Fiedler et al., 2004). WPBs are rod-shaped org anelles that contain several proteins involved in 33

PAGE 34

vascular homeostasis, infla mmation, vasodilation, vasoconstriction, and immune cell adhesion and migration (Lowenstein et al., 2005; Rondaij et al., 2006). The most commonly known WPB protein is von-Willebr and Factor (vWF) involved in hemostasis, others include for example p-selectin, in terleukin-8 endothelin, CD63; Ang-2 colocalizes with vWF in WPBs however not with any other know n WPB proteins (Fiedler et al., 2004; Lowenstein et al., 2005; Rondaij et al., 2006). WPBs are triggered to release their contents via elevated Ca2+ (e.g. thrombin, histami ne, superoxide anion, VEGF, sphingosine-1 phosphate, cerami de) or cAMP (e.g. serotonin, epinephrine, vasopressin) dependent manner and fuse with the endothelial cell plasma membrane using vesicle SNARE (v-SNARE) on the WPB and target -SNARE (t-SNARE) on the plasma membrane (Lowenstein et al., 2005). Upon stim ulation such as inflammation, hypoxia, shear stress at sites of active angiogenesis the WPBs are exocytosed and Ang-2 released from endothelial cells (Lowenstein et al., 2005). Secreted Ang-2 destabilizes the vasculature by autocrine interaction wit h Tie2, antagonizing Ang1 signaling, leading to disruption of adhesion molecule interactions between cells and turning the angiogenic switch towards a pro-angiogenic phenotype (Augustin et al., 2009; Scharpfenecker et al., 2005) The goals of the present in vestigation were to mimi c the interaction between endothelial and peri-en dothelial cells using an endothelialsmooth muscle cell co-culture sphere model and directly assess the impact of the Ang-2 inhibitor on the vascular destabilizing effects of Ang-2 usin g real time imaging of the spheres. 34

PAGE 35

Materials And Methods Reagents Phorbol 12-Myristate 13-Acet ate (PMA) (Sigma), thro mbin (MP Biomedicals, LLC), and human recombinant Angiopoietin-2 (R&D Systems) were dissolved in DMSO, water and PBS + 0.1% BSA respective ly. Human Ang-2 Quantikine ELISA Kits were purchased from R&D Systems and stor ed at 4 C until use. Methylcellulose (viscosity 4000 cP) was obtained from Sigma; stock solutions were dissolved in basal medium (endothelial cell or sm ooth muscle cell) and ultra centrifuged to a clear solution (3500 rpm, 4 C, 3 hr) and the top 45% of supernatants stored at 4 C up to 3 months. CellTracker Orange CMRA, CellTrace Oregon Green 488 Carboxylic Acid Diacetate Succinimidyl Ester, and CellTra cker Violet BMQC (Molecular Probes Invitrogen) were dissolved in DMSO and stored at -20 C. Drug Preparation Angiopoietin-2 inhibitor, MEDI3617, wa s kindly provided by MedImmune, LLC. For in vitro investigations 5 mg/ml stock solutions were diluted to 10 M working solutions in sodium citrate buffer before subsequent dilutions in sterile saline. Stock solutions were kept at -80 C and working concentrations at 4 C. Cell Culture Human umbilical vein endothelial cells (HUVEC) and human umbilical artery smooth muscle cells (HUASMC) were purchased from Clonetics. HUVEC were cultured in Nutrient mixture F-12 Ham, Kaighns supplement ed with 0.03 mg/ml Endothelial Cell Growth Suppl ement (ECGS) and 0.1 mg/m l heparin (Sigma) with 10% fetal bovine serum (FBS). HUASMC were cultured in SmGM-2 (Clonetics). Cells were kept at 37 C, 5% CO2. HUVEC and HUASMC were used between passages 2 and 4. 35

PAGE 36

Endothelial-Smooth Muscle Cell Co -Culture Sphere Formation Co-culture spheres of HUVEC and HU ASMC were generated as previously reported (Korff et al., 2001; Scharpfenecker et al., 2005). Confluent monolayer cultures of HUVEC were traced with 1 g/ml CellTracker Orange CMRA in serum free media for 15 min. Confluent monolayer culture of HUASMC were traced with either 1 g/ml CellTrace Oregon Green 488 Carboxylic Acid Diacetate Succinimidyl Ester (imaging with Nikon eclipse TS100 microscope) or 1 g/ml CellTracker Violet BMQC (imaging with Leica SP5 confocal microscope) in serum free media for 15 min. A 1:1.5 ratio of HUVEC to HUASMC (1000 cells per spheroi d) were suspended in a 1:1 ratio of endothelial and smooth muscle cell complete medium in the presence of 4.2% carboxymethylcellulose into a non-adherent round-bottom 96 well plate (Sigma), 48 hr later the spheres were selected for experiments. Confocal Imaging Of Endothelial-Smooth Muscle Cell Co-Culture Spheres Confocal microscopy was used to evaluate co-culture sphere orientation. HUVEC were tagged with CellTracker Orange CMRA and HUASMC were tagged with CellTracker Violet BMQC prior to sphere formation and imaged with Leica SP5 confocal microscope with an environmental chamber for live imaging at 20x magnification. Spheres were imaged on FluoroDish, FD35-100 (World Precision Instruments, Inc.). 405 Diode laser was used to image HUASMC with CellTracker Violet BMQC and HeNe 543 laser was used to visualize HUVEC with CellTracker Orange CMRA. LAS AF software was used to obtain confocal images of spheroids. Coronal view of spheroids was obtained from a series of 10 z-stack images at 20.42 m per step. 3 dimensiona l images were compiled from a se ries of 30 z-stack images, at 6.19 m per step, using LAS AF software and were processed as 3D projections. 36

PAGE 37

Imaging Co-Culture Sphere Formation And Destabilization Co-culture sphere formation and destabilization, endothelial cell loss, was imaged with a Nikon eclipse TS100 inverted microscope. HUVEC were tagged with CellTracker Orange CMRA and HUASM C were tagged with CellTrace Oregon Green 488 Carboxylic Acid Diacetate Succini midyl Ester prior to sphere formation. HUVEC were fluorescently visualized using TRIC filter and HUASMC were viewed using FITC filter with a Nikon Intensilight C-HGFI lamp. S pheroids were imaged at t=0 and t=4 hr at 10x magnificati on and NIS Elements D 3.2 software was used to obtain individual and merged images. Human Angiopoietin-2 ELISA HUVEC were cultured in 60 mm dishes and grown to confluence. Following media removal and washing 2 ml of media c ontaining either 50 ng/ ml PMA or 1 unit/ml thrombin were added. At various times ther eafter the media was collected, centrifuged (1000 rpm, 4 C, 10 min) and 1.5 ml of super natant collected and stor ed at -20 C until analysis. Cells were trypsinized, harvest ed and counted for each sample. A sandwich ELISA of human Angiopoietin-2 (Quantikine ELISA Kit R&D Systems DY623) was used to analyze Angiopoietin-2 secretion. The manufacturers protocol was followed; briefly, 96-well plates were coated with the capture antibody and allowed to sit overnight at room temperature. Standards and samples were added to the plate followed by the detection antibody, streptavidin HRP and substr ate solution. The concentration of Ang-2 in samples was determined using the optical density at 450 nm. Duplicates of each sample were run and secreted Ang-2 (pg/ml) was normalized to 105 cell number and the baseline was set to 0 based on basal level readings. 37

PAGE 38

Stimulation Of Endothelial-Smooth Muscle Cell Co-Culture Spheres Endothelial and smooth muscle cells were al lowed to form spheres for 48 hr and were then transferred to a non-adhesive round bottom 96 well plate (1 sphere per well) and were live imaged (Nikon eclipse TS100 mi croscope) at t=0 hr prior to any stimulation. Stimulant s, human recombinant Ang-2, thro mbin or PMA, were then added into the wells and spheres were live imaged at t=4 hr. The number of fluorescent tagged HUVEC that detached from t he smooth muscle cell cores wa s counted in the field of view. Statistical signific ance between stimulated groups and control (unstimulated) spheres was determined using the Mann-Wh itney U test at p<0.05. In each experimental group 12 spheres (1 sphere per well) were analyzed. Results Endothelial And Smooth Muscle Cells Form Co-Culture Spheres When combined, endothelial (HUVEC) and smooth muscle cells (HUASMC) form spheroids, which mimic the vascular associati on of these cell types (Korff et al., 2001; Scharpfenecker et al., 2005) Live imaging of HUVEC and HUASMC with red and green cell trackers respectively (Figure 2-1A), was used to demonstrate the time course of endothelial/muscle cell co-culture spheroid fo rmation (Figure 2-1B). Confocal imaging of these spheroids and its 3 dimensional recons truction shows an insideout orientation of spheres with smooth muscle cells found at the core and endothelial cells on the periphery of the sphere (Figure 2-1C). Ang-2 Inhibitor Impairs Exogenous Ang2 Dependent Sphere Destabilization Ang-2 interacts with Tie2 receptor on t he endothelial cells and leads to loss of endothelial cell contact with su rrounding endothelial and per i-endothelial cells (Augustin et al., 2009; Scharpfenecker et al., 2005). The addition of human recombinant Ang-2 38

PAGE 39

(0.5 ng/ml) to co-culture spheres for 4 hr (Figure 2-2A) led to a 5-fold increase in endothelial cell loss from the spheres (p<0.00 01). Administration of the Ang-2 inhibitor (0.5 nM) significantly reduced (~ 3.5-fold) the endothelial cell loss as compared to Ang-2 stimulated spheres (p<0.0001). Figure 2-2B illustrates the live imaging of co-culture spheroids and demonstrates t he endothelial cell loss upon exo genous Ang-2 stimulation and its inhibition in the presence of the Ang-2 inhibitor. Ang-2 Inhibitor Hinders E ndogenous Ang-2 Dependent Sphere Destabilization Stimulation of HUVEC with PMA or th rombin led to endogenous Ang-2 secretion as measured by ELISA (Figure 2-3). PMA expos ure led to detectable Ang-2 secretion at 1 hr and increased to 2.8 and 7.8-fold above c ontrol at 6 and 24 hr respectively (Figure 2-3). Ang-2 secretion was not detectable after 1 hr of thrombin stim ulation and even at 6 and 24 hr exposure the Ang-2 levels were significantly lower than those observed following PMA treatment (Figure 2-3). Treatm ent of spheres for 4 hr with 0.5, 5, 50 ng/ml PMA (Figure 2-4A) resulted in 3.4, 6.2, and 6.8-fold increase in endothelial cell loss from spheres respectively (p<0.0001). Ad ministration of the Ang-2 inhibitor to PMA (50 ng/ml) treated spheres signi ficantly decreased the number of endothelial cells lost from the spheres (Figure 24B). Live imaging of co-culture spheres (Figure 2-4C) illustrates the disruption of endothelial/muscle cell interaction upon PMA stimulation and its inhibition in the presence of the Ang-2 inhibitor. Discussion The importance of angiogenesis in the growth of solid tumors has led to the active pursuit of anti-angiogenic anticanc er therapies (Folkman, 1971). Although such agents have demonstrated some efficacy in the c linic, concerns about lack of treatment response and the potential for resistance to therapy have been raised (Bergers and 39

PAGE 40

Hanahan, 2008; Jain et al., 2006). Resistanc e to anti-angiogenic therapy may occur when inhibition of a particula r signaling pathway is overcome by the activation of another to sustain tumor blood vessel growth (Bergers and Hanahan, 2008; Jain et al., 2006). One possible approach to overcoming such concerns is to target the angiopoietin/Tie2 axis since it is consider ed to be the sole pathway responsible for maintaining blood vessel integrity (Augusti n et al., 2009; Hu and Cheng, 2009). While Ang-1 binding to Tie2 recept or stabilizes the vasculatu re, Ang-2 binding primes the vasculature for angiogenesis by freeing endothe lial cells from peri-endothelial cells, thus making them sensitive to stimulation by pr o-angiogenic factors (Augu stin et al., 2009). Minimizing the dissociation of endothelial an d peri-endothelial cells by targeting Ang-2 therefore would hinder tumor-induced angiog enesis by blocking the access of proangiogenic factors to endothelia l cells, thus making this ligand a desirable target for vascular directed anticancer therapeutic approaches. The present study examined the destabilizing effects of Ang-2 on endothelial/smooth muscle cell contacts in the presence or absence of the Ang-2 inhibitor using a co-culture sphere model and real time imaging (Figure 2-1). The high specificity and neutralizing propert ies of this inhibitor have been previously confirmed by Leow and colleagues (Leow et al., 2012). The results showed that Ang-2 initiates endothelial-peri-endothelial cell contact loss (F igure 2-2); a finding consistent with the previous report that direct contact with smooth muscle cells r egulates endothelial cell quiescence and abrogates the response to VEG F (Korff et al., 2001). Although the model is an inside-out ori entation of the natural appearan ce of the vasculature, it nonetheless mimics the interaction between the two cell types essential in the make-up 40

PAGE 41

of blood vessels. Even though smooth muscle ce lls are usually found on larger arteries while pericytes coat endothelial cells of sma ller arterioles, venules, or capillaries, these two cell types share a common lineage and interact with endothelial cells via NCadherin, a major adhesion molecule betw een endothelial and periendothelial cells that is broken upon Ang-2-Tie2 association during angiogenesis (Armulik et al., 2005; Dejana, 2004). Consistent with previous reports on the endothelial ce ll/peri-endothelial cell interactions (Scharpfenecker et al., 2005) exogenously applied Ang-2 was found to result in a loss of endothelial cells from co -culture spheroids. The current study employed real time imaging of spheres befor e and after stimulati on with Ang-2 (Figure 2-2) and demonstrated that ev en 400-fold lower concentrati ons of Ang-2 (0.5 ng/ml) than previously reported (Scharpfenecker et al., 2005), led to a significant increase in the number of endothelial cells dissociating from the unaffected smooth muscle cell core (Figure 2-2). Because the in situ dest abilization of endothelia l/peri-endothelial cell interactions is modulated by the endogenous re lease of Ang-2 from endothelial cells, experiments also were undertaken to assess whether such Ang-2 release would affect the co-culture spheres. Ang-2 is stored in Weibel-Palade Bodies in endothelial cells and is rapidly secreted during both physiol ogical and pathological angiogenic stimulus (Fiedler et al., 2004; Huang et al., 2002; Krikun et al., 2000; Lowenstein et al., 2005; Rondaij et al., 2006). Chemical stimulation of endothelial ce lls with PMA or thrombin leads to Weibel-Palade Bodies exocytosis and results in Ang2 secretion (Fiedler et al., 2004); a finding confirmed in t he present investigation by ELISA analysis of endogenous Ang-2 secretion from endothelial cells followi ng PMA stimulation (Fi gure 2-3). When co41

PAGE 42

culture spheroids were treated with PMA (50 ng/ml), the number of endothelial cells released was found to be increased ~ 6-fold wit hout an effect on the smooth muscle cell core (Figure 2-4). Similar results were noted when spheroids were treated with the clotting factor thrombin (data not shown). The current studies focused on real ti me imaging of Ang-2 initiated endothelial cell dissociation from smooth muscle cell contacts and the ability of the Ang-2 inhibitor to impair that dissociation. Endothelial/smooth muscle co-culture spheres were exposed to either exogenous Ang-2 (Figure 2-2) or agents to initiate endogenous Ang-2 expression (Figure 2-4) in t he presence or absence of the inhibitor. Loss of endothelial cells from the co-culture spheres was monito red using cell trackers. The results showed that treatment with low dose Ang-2 inhibitor (0.5 nM) c ounteracted the effects of exogenous Ang-2 and inhibited the dissociati on of endothelial from smooth muscle cells (Figure 2-2). Similarly, endothelial cell loss from the co-culture spheroids exposed to PMA could be significantly impaired (~ 3-fold ) in the presence of the Ang-2 inhibitor (Figure 2-4). In a recent in vivo study Mazzieri and colleagues (Mazzieri et al., 2011) reported that treatment with the Ang-2 inhibitor 3.19.3 reduced the number of tumor blood vessels but the remaining tumor vascul ature was heavily coat ed with pericytes. It should be noted, however, that the observa tion was made at the end of an extended treatment period; i.e. at t he end of prolonged Ang-2 depletion which ultimately favors an Ang-1 based stabilization of the endothelial cell/pericyte interaction. The present investigation demonstrates t hat the Ang-2 inhibitor bloc ks the early phases of the angiogenesis process by impairing the init ial destabilization of the endothelial/smooth muscle cell contacts. Therefor e, this study demonstrates t hat Ang-2 targeting with the 42

PAGE 43

inhibitor may have beneficial anti-angiogenic effects by blo cking the initial phase of angiogenesis, the dissociation of endothelial and smooth muscle cells in a stable blood vasculature. The presence of Ang-2 in the tumor microenvironment and circulation has been associated with disease progre ssion (Anargyrou et al., 2008; Helfrich et al., 2009; Lind et al., 2005; Sie et al., 2009; Tait and Jones, 2004; Tsutsui et al., 2006) and enhanced metastatic potential (Carlson et al., 2001; Hu et al., 2006; Im anishi et al., 2007). In the angiogenic process, the angiopoi etin/Tie2 axis is current ly considered to be a nonredundant pathway involved in maintaining bl ood vessel integrity. Consequently, Ang-2 has been considered as a novel anticancer target. The current study supports the rationale for Ang-2 targeting as a therapeutic intervention in the angiogenic process. Experiments using fluorescent imaging of co -culture endothelial/smooth muscle cell spheres clearly demonstrate a significant impact of the Ang-2 inhibitor, a human monoclonal Ang-2 antibody, on Ang-2 dependent destabilization of the endothelial/muscle cell interaction in vitro. Th e present results support the concept that this antibody has inhibitory effects on new bl ood vessel formation via inhibition of Ang-2 induced endothelial/peri-endothelial cell contac t loss. Since minimizing the dissociation of endothelial and peri-endothelial cells by targeting Ang-2 could hinder tumor-induced angiogenesis by blocking the endothelial cells from the influence of pro-angiogenic factors, the Ang-2 inhibitor warrants furt her investigation as a vascular directed anticancer therapy. 43

PAGE 44

Figure 2-1. Endothelial and smooth muscle cells form co-culture spheres. (A) Monolayer of endothelial and smooth muscle cells are added to a 96 wellround bottom plate with complete media and methylcellulose. (B) Formation of an endothelial-smooth muscle cell spher e over a 48 hr time period. Smooth muscle cells (HUASMC) are shown in green (CellTrace Oregon Green) and endothelial cells (HUVEC) are shown in red (CellTracker Orange CMRA). Merged images are also shown. Images were taken with Nikon eclipse TS100. Scale bar represents 100 m (C) Coronal and 3-dimensional view of co-culture sphere with smooth muscle cells located in the core and endothelial cells at the per iphery. Live images were taken with a Leica SP5 confocal microscope. Scale bar represents 100 m. Adapted with permission from ELSEVIER: Microvascular Research 2012; 83(3):290-7. 44

PAGE 45

Figure 2-2. Exogenous Ang-2 causes sphere destabilization, Ang-2 antibody inhibits this response. (A) Stimulation of HUVEC with exogenous human recombinant Ang-2 (0.5 ng/ml) for 4 hr led to the loss of endothelial cells from the smooth muscle cell core. The presence of ME DI3617 (0.5 nM) impairs this loss. Column, mean; bars, SD (n=12 spheres per experimental group). ***, p<0.0001; Mann-Whitney U test. (B) Im ages demonstrating control, Ang-2 (0.5 ng/ml) induced destabilization of co-culture spheres and its inhibition in the presence of MEDI3617 (0.5 nM) at 4 hr Images as in Figure 2-1. Adapted with permission from ELSEVIER: Microvascular Research 2012; 83(3):290-7. 45

PAGE 46

Figure 2-3. PMA and thrombin are stimulator s of Ang-2 secretion from endothelial cells. (A) HUVEC were exposed to PMA (50 ng/ml ) or thrombin (1 unit/ml) for 1,6, or 24 hours. Stimulation of HUVEC with PMA or thrombin led to Ang-2 secretion over time. (B) Ang-2 secret ion was measured by ELISA and data were normalized to 105 cells, baseline readings from basal levels were set at 0. Column, mean; bars, SD of thre e independent experiments. Used with permission from ELSEVIER: Microvascular Research 2012; 83(3):290-7. 46

PAGE 47

Figure 2-4. PMA induced endogenous Ang-2 causes sphere destabilization, Ang-2 inhibitor impairs this response. (A) St imulation of HUVEC with PMA for 4 hr resulted in loss of endothelial cells from the smooth muscle cell core in a dose dependent manner. Column, mean; bars, SD (n=12 spheres per experimental group). ***, P<0.0001; Mann-Whitney U test. (B) MEDI3617 inhibited co-culture sphere destabili zation when stimulated with PMA (50 ng/ml) for 4 hr. Column, mean; bars, SD (n=12 spheres per experimental group). *, p<0.05: **, p<0.01: ***, p< 0.0001; Mann-Whitney U test. (C) Images demonstrating control, PMA (50 ng/ml) induced destabilization of co-culture spheres and its inhibition in the pr esence of MEDI3617 (0.5 nM) at 4 hr. Images as in Figure 2-1.Adapt ed with permission from ELSEVIER: Microvascular Research 2012; 83(3):290-7. 47

PAGE 48

CHAPTER 3 IMPLICATIONS OF THE EFFECTS OF TUMOR MICROENVIRONEMNTAL FACTORS ON THE ANG-2 AXIS AND SPHERE DESTABILIZATION Chapter 2 demonstrated the use of an endothelial-smooth muscle cell sphere model to evaluate Ang-2 dependent sphere dest abilization and its inhibition in the presence of the Ang-2 inhibitor. Chapter 3 explored an area that remains controversial, the involvement of the tumor microenvironment in the Ang-2 axis. The following studies explored the involvement of tumor cells, hypoxia and VEGF in the Ang-2 axis. Results demonstrate that renal cell carcinoma tu mor cells indirectly influence vessel destabilization by secretion of factors that stimulate endothelial cell s to secrete Ang-2. Studies also evaluated factors commonly found in the tumor microenvironment such as hypoxia and VEGF. As demonstrated in Chapte r 2, the presence of the Ang-2 inhibitor impairs Ang-2 dependent s phere destabilization. Background The tumor microenvironment highly influe nces tumor angiogenesis (Betof and Dewhirst, 2011; Weis and Cheresh, 2011). A growing tumor that exceeds the oxygen diffusion limit from normal va sculature experiences hypoxia and nutrient deprivation (Hanahan and Folkman, 1996). This triggers a pro-angiogenic phenotype, tumor cells respond to hypoxia by secret ing pro-angiogenic factors such as VEGF that activates sprouting of new vessels toward the tumor (Betof and Dewhirst, 2011; Charlesworth and Harris, 2008; Dewhirst, 2006). Th e tumor microenvironment is also abundant in tumor associated macrophages (TAM s) and fibroblasts that have been shown to promote tumor growth and angiogenesis (Biswas et al., 2013; Cirri and Chiarugi, 2011; De Palma and Naldini, ; Xing et al., 2010). 48

PAGE 49

The Angiopoietin/Tie2 axis is a ke y component of angiogenesis and vascular remodeling both in physiology and pathology (A ugustin et al., 2009; Davis et al., 1996; Maisonpierre et al., 1997; Partanen et al., 1992; Schnurch and Risau, 1993). Angiopoietin-1 (Ang-1) interaction with Tie2 receptor promotes a quiescent vascular phenotype, maintaining endothelial and periendothelial cell interactions found in normal, non-proliferating vasculature (A ugustin et al., 2009). On the other hand, Angiopoietin-2 (Ang-2) ligand is a Tie2 antagonist leading to vascular destabilization, the loss of peri-endothe lial cell coverage off the endothe lial layer of vasculature (Augustin et al., 2009; Scharpfenecker et al., 2005). Ang-2 is stored in endothelial cells in Weibel-Palade Bodies (WPB) co localized with von Willebrand Factor (vWF), but not other WPB proteins, and is r apidly secreted in response to microenvironmental changes (Fiedler et al., 2004). Ang-2 expression is el evated in various tumor settings, both within tumor tissue and in serum; increased leve ls have been correlated to poorer disease prognosis (Tait and Jones, 2004). Due to the role of Ang-2 in tumor angiogenesis it has been increasingly pursued as an anti-angiogenic modality in cancer therapy. Current ly there are several agents being evaluated in all phases of clinical dev elopment (Gerald et al., 2013). However, our understanding of this axis and its interp lay in the tumor microenvironment remains limited. As a tumor develops and grows it su rpasses the available oxygen and nutrients provided by nearby normal vasculature. Upon low oxygen conditions such as hypoxia tumor cells express the hypox ia inducible factor (HIF-1 ) and upregulate several genes involved in a variety of processes such as angiogenesis and metabolism (Charlesworth and Harris, 2008; Dewhirst, 2006; Maxwel l et al., 1997). Tumor cells promote 49

PAGE 50

angiogenesis by directly secreting pro-angio genic factors such as Vascular Endothelial Growth Factor (VEGF) and Fibroblast Growth Factor (FGF) especially under hypoxic conditions (Maxwell et al., 1997). Whether tumor cells directly affect the Ang-2/Tie2 axis in angiogenesis is still unclear. Several reports have shown that tumor cells do express Ang-2, however there is limited evidence fo r tumor cell secretion of Ang-2 into the microenvironment (Detjen et al., 2010; Engin et al., 2012; Li et al., 2013; Schulz et al., 2011). Furthermore, the understanding of the e ffects of hypoxia and VEGF on the Ang-2 axis is limited (Fiedler et al., 2004; Matsushita et al., 2005; Pichiule et al., 2004; Tsigkos et al., 2006). Chapter 2 demonstrated the use of endothel ial-smooth muscle cell sphere model to mimic normal vasculature and evaluat e Ang-2 dependent cell-cell contact loss (destabilization) and its inhibition in the presence of the Ang-2 inhibitor (Molnar and Siemann, 2012). The current study evaluat ed the tumor microenv ironmental effects such as the interplay between renal cell ca rcinoma, endothelial ce ll, hypoxia and VEGF on the Ang-2 dependent destabiliza tion. The Ang-2 inhibitor was evaluated in the coculture sphere models in vitro Materials And Methods Reagents Methylcellulose (viscosity 4000 cP) was obt ained from Sigma; it was prepared as previously described in Chapter 2. Hum an Ang-2 Quantikine ELISA Kits and human recombinant Vascular Endothelial Growth Fa ctor (VEGF) were purchased from R&D Systems. CellTracker Orange CMRA and CellTrace Oregon Green 488 Carboxylic Acid Diacetate Succinimidyl Ester (Molecular Probes Invitrogen) were dissolved in DMSO and stored at -20 C. 50

PAGE 51

Cell Culture Human umbilical vein endothelial ce lls (HUVEC), human microvascular endothelial cells from the lungs (HMVEC-L), and human umb ilical artery smooth muscle cells (HUASMC) were purchased from Clonet ics. HUVEC were cultured in Nutrient mixture F-12 Ham, Kaighns supplemented with 0.03 mg/ml Endothelial Cell Growth Supplement (ECGS) and 0.1 mg/ml heparin (Sigma) with 10% fetal bovine serum (FBS). HMVEC-L were grown in EGM-2 MV and HUASMC were cultured in SmGM-2 (Clonetics). The human clear cell renal ce ll carcinoma Caki-1 and Caki-2 cell lines were received as a gift from Dr. Susan Knox (Stanford Univer sity); A498 and 786-0 human renal cell carcinoma cell lines were obtained from Dr. Kyung-Mi Bae (University of Florida). Caki-1 and Caki-2 were grown in Dulbeccos modified minimum essential medium (D-MEM, Invitrogen), A498 were grown in Eagles minimum essential medium (MEM, Cellgro), 786-0 were grown in RPMI1640 medium (Sigma). All tumor cell media were supplemented with 10% FBS (Invitrogen) 1% penicillin-streptomycin (Invitrogen), and 1% 200-mmol/L L-glutamine (Invitrogen). Cells were kept at 37 C, 5% CO2. HUVEC, HMVEC-L and HUASMC were used between passages 2 and 4. Caki-1, Caki2, A498 and 786-0 cells were used between passages 2 and 10. Drug Preparation Angiopoietin-2 inhibitor, MEDI3617, wa s kindly provided by MedImmune, LLC. 5 mg/ml stock solutions were diluted to 10 M working solutions in sodium citrate buffer before subsequent dilutions in sterile saline. Stock solutions were kept at -80 C and working concentrations at 4 C. 51

PAGE 52

Endothelial-Smooth Muscle Cell Co -Culture Sphere Formation Co-culture spheres of HUVEC and HU ASMC were generated as previously reported in Chapter 2. Conf luent monolayer cultures of HUVEC were traced with 1 g/ml CellTracker Orange CMRA in seru m free media for 15 min. Confluent monolayer culture of HUASMC were traced with either 1 g/ml CellTrace Oregon Green 488 Carboxylic Acid Diacetate Succinimidyl Ester (imaging with Nikon eclipse TS100 microscope) or 1 g/ml CellTra cker Violet BMQC (imaging with Leica SP5 confocal microscope) in serum free medi a for 15 min. A 1:1.5 ratio of HUVEC to HUASMC (1000 cells per spheroid) were suspended in a 1:1 ratio of endothelial and smooth muscle cell complete medium in the presence of 4.2% carboxymethylcellulose into a non-adherent round-bottom 96 well plate (S igma). After 48 hr, spheres were used for subsequent experiments. Stimulation Of Endothelial-Smooth Muscle Cell Co-Culture Spheres Endothelial and smooth muscle cells were init ially allowed to form spheres for 48 hr at which point they were transferred to a non-adhesive round bottom 96 well plate (1 sphere per well) and were live imaged (Nikon eclipse TS100 micr oscope) at t=0 hr prior to any stimulation. Stimulants, tumor ce ll conditioned media, were then added into the wells (1:1 ratio of tumor cell conditioned me dia and endothelial cell media) and spheres were live imaged at t=4 hr. The number of fluorescent tagged HUVEC that detached from the smooth muscle cell cores was c ounted in the field of view. Statistical significance between stimulated groups and control (unstimulated) spheres was determined using the Mann-Whit ney U test at p<0.05. 52

PAGE 53

Imaging Co-Culture Sphere Formation And Destabilization Co-culture sphere formation and destabilization, endothelial cell loss, was imaged with a Nikon eclipse TS100 inverted microscope. HUVEC were tagged with CellTracker Orange CMRA and HUASM C were tagged with CellTrace Oregon Green 488 Carboxylic Acid Diacetate Succini midyl Ester prior to sphere formation. HUVEC were fluorescently visualized us ing TRIC filter and HUASMC were viewed using FITC filter with a Nikon Intensilight C-HGFI lamp. S pheroids were imaged at t=0 and t=4 hr at 10x magnificati on and NIS Elements D 3.2 so ftware was used to obtain individual and merged images. Human Angiopoietin-2 ELISA Renal cell carcinoma cells (Caki-1, Ca ki-2, 786-0, and A498) were grown in either 100 mm or 60 mm dishes and grown to 70-80% confluence. Following media removal and washing, either 10 ml (100 mm dish) or 2 ml (60 mm dish) media was added to the plates. Plates were either kept at normal oxygen levels (37 C incubator) or placed under hypoxia for various time poi nts. HUVEC and HMVEC-L were cultured in 60 mm dishes and grown to confluence (70-80%). Following media removal and washing, 2 ml of media containing eit her media alone (basal readings), media and tumor cell conditioned media (1:1), or hum an recombinant VEGF (5 25, 50 ng/ml) was added to the plates for various times. The me dia was collected, centrifuged (1000 rpm, 4 C, 10 min) and 1.5 ml of supernatant colle cted and stored at -20 C until analysis. Cells were trypsinized, harvested and counted for each sample. A sandwich ELISA of human Angiopoietin-2 (Quantikine ELISA Kit R&D Systems DY623) was used to analyze Angiopoietin-2 secretion. The manufacturers protocol was followed; briefly, 96well plates were coated with the capture ant ibody and allowed to si t overnight at room 53

PAGE 54

temperature. Standards and samples were added to the plate followed by the detection antibody, streptavidin HRP and substrate so lution. The concentration of Ang-2 in samples was determined using the optical density at 450 nm. Duplicates of each sample were run and secreted Ang-2 (pg/ml) was normalized to 105 cell number. Statistical significance between groups was determined using the student t-test at p<0.05. Hypoxia Special aluminum chambers designed by Dr. Cameron Koch (University of Pennsylvania) were used as hypoxia chambers Tumor (Caki-1, Caki-2, 786-0, A498) or endothelial cells (HUVEC, HMVEC-L) were plated on glass di shes in their respective full serum media. For hypoxia 1% oxygen, 5% CO2 conditions were established. Upon flushing normal oxygen in exch ange to hypoxic conditions the chambers were placed in 37 C incubators for various time points. Results Tumor Cells Secrete Minimal Levels Of Ang-2 But Their Conditioned Media Stimulate Ang-2 Secreti on From Endothelial Cells And Ang-2 Dependent Destabilization, Ang-2 Inhibitor Impairs This Loss There is still controversy whether tumor cells can secrete Ang-2 thus directly affecting angiogenesis (as they do with VEGF) or whether Ang-2 is solely provided by the endothelial cells and found in the tumor stroma. Results show that human renal cell carcinoma cells, Caki-1, Caki-2, 786-0, and A498 secreted minimal levels of Ang-2; 13.4, 8, 8.6 and 24.8 pg/ml of Ang-2 respectively (Figur e 3-2A-D). When conditioned media from tumor cells were added onto endothelial cells the endothelial cells secreted 2.2-2.4-fold higher levels of Ang-2 compar ed to basal conditions; 34.5-, 58-, 50-, and 15-fold higher levels compared to levels that Caki-1, Caki-2, 786-0, and A498 secreted 54

PAGE 55

respectively (p<0.01 C,D; p<0.0001 A,B) (F igure 3-2A-D). When co-culture spheres were stimulated with Caki-1 conditioned m edia there was a significant loss of endothelial cell loss from the sphere core (2 .7-fold); in the presence of the Ang-2 inhibitor Caki-1 conditioned media induced endothelial cell lo ss is significantly reduced by 2.2-fold (p<0.01) (Figur e 3-2E). Figure 3-2F illustrates the endothelial cell loss due to Caki-1 conditioned media and its inhibition in the presence of the Ang-2 antibody. Hypoxia Does Not Affect Tumor Cell Secretion of Ang-2 The tumor microenvironment is fairly hy poxic, which triggers tumor cells to secrete pro-angiogenic factors such as VEGF (Maxwell et al., 1997). The possibility of tumor cells secreting Ang-2 under hypoxic co nditions was evaluated; both short term (acute) and long term (chronic) were assess ed. The results showed that there was no significant difference in Ang-2 secretion from human renal cell carcinoma cells Caki-1 (Figure 3-3C), Caki-2 (Figure 3-3D), 786-0 (Figure 3-3E) or A498 (Figure 3-3F) under hypoxic conditions compared to control; tumor cells secreted minimal levels of Ang-2 regardless of the oxygen st atus in the environment. Tumor Cell Media From Hypoxic Conditi ons Do Not Affect Endothelial Cell Secretion Of Ang-2 Compared to Tum or Cell Media From Normal Oxygen Levels The addition of tumor cell conditioned media significantly increased Ang-2 secretion from endothelial cells under both normal (F igure 3-2A-D) and hypoxic conditions (Figure 3-4A-D). Endothelial cell se creted Ang-2 by an increase of 2.4-, 2.5-, 2.4-, and 2.6-fold in the presence of Caki -1, Caki-2, 786-0, and A498, respectively, conditioned media from hypoxic conditions compared to basal levels (p<0.05) (Figure 34A-D). However, the exposure of tumor cells to hypoxia did not lead to a difference in Ang-2 secretion from endothelial cells compared to normal oxygen conditions. Figure 355

PAGE 56

4E shows that co-culture spheres in the presence of conditioned media from Caki-1 cells, both from normal and hypoxic conditions, led to a significant loss of endothelial cells from the sphere 3.3(p<0.05) and 4.1-fold (p<0.0001) compared to control respectively. Figure 3-2E,F shows that t he Ang-2 inhibitor significantly reduced the number of endothelial cell loss from spheres in the pres ence of Caki-1 conditioned media from normal oxygen conditions; Figur e 3-4E demonstrates that the same occurred in the presence of Caki-1 media from hypoxic conditions with a 1.9-fold (p<0.05) decrease in endothelial cell loss. Endothelial Cell Secretion Of Ang2 Was Not Altered Due to Hypoxia Endothelial cells, HUVEC (Figure 3-5A) and HMVEC-L (Figure 3-5B), under acute or chronic hypoxic conditions did not secrete heightened levels of Ang-2 as compared to normal oxygen levels. VEGF Does Not Significantly Affect A ng-2 Release From Endothelial Cells The possibility of VEGF, found in abunda nce in the tumor microenvironment, being the factor tumor cells secrete and in turn stimulate Ang-2 release from endothelial cells was assessed. VEGF did not significant ly affect HUVEC (Figure 3-8A) or HMVECL (Figure 3-8B) secretion of Ang-2 when eval uated at various time points (Figure 3-8). Discussion The tumor microenvironment plays an im portant role in tumor angiogenesis (Betof and Dewhirst, 2011; We is and Cheresh, 2011). Several factors such as hypoxia, immune cells and stromal cells can have an impact on this process (Weis and Cheresh, 2011). The current study focused on three important aspects of the microenvironment; tumor cells, hypoxia, and pro-angiogenic growth factors. Increased expression of Ang-2 within the tumor microenvironment have been shown to correlate to advanced disease 56

PAGE 57

(Detjen et al., 2010; Engin et al., 2012; Li et al., 2013; Schulz et al., 2011). However, it remains controversial how the microenvironment influences the Ang-2 axis and the origin of Ang-2 in the microenvironment. During the growth of a tumor, hypoxia in the microenvironment leads to tumor cell secretion of pro-angiogenic factors (e.g. VEGF) and the direct influence on new vessel growth (Hanahan and Folkman, 1996). W hether tumor cells contribute to Ang-2 levels in the microenvironment and directly stimulate angiogenesis, as with VEGF (Figure 3-1) is controversial. Reports have demonstrated mRNA le vels of Ang-2 within tumor cells, however, there is lack of evidence for tumor cell secretion of Ang-2 into the microenvironment (Currie et al., 2002; Tait and Jones, 2004; Yamakawa et al., 2004). Chapter 3 evaluated the involv ement of renal cell carci noma, a highly angiogenic disease, in the Ang-2/Tie2 axis during angiogenesis initiation focusing on the interaction of endothelial and peri-endothelial ce lls found in normal vasculature; the effect of the Ang-2 inhibitor on cell-cell contacts was determined. Results indicate that renal cell ca rcinoma cells do express Ang-2 as been previously reported (Currie et al., 2002; Tait and Jones, 2004; Yamakawa et al., 2004), however, secrete minimal levels into the microenvironment in the range of 8-24 pg/ml (Figure 3-2). Nevertheless, the media collect ed off of tumor cells si gnificantly stimulate endothelial cells to release their Ang-2 cont ents into the microenvironment (Figure 3-2). Basal levels of Ang-2 secretion ranged from 170-210 pg/ml but in t he presence of tumor cell conditioned media this range was elev ated to 370-470 pg/ml. The elevated Ang-2 secretion from endothelial cells in the presence of tumor cell conditioned media led to significant destabilization of endothelial and peri-endothelial cell contacts in the sphere 57

PAGE 58

model (Figure 3-3). These studies were conduc ted during a short peri od of time, 4 hrs, therefore the levels of Ang-2 that were detected unlikely a ttribute to the synthesis of Ang-2 but rather the releas e of Ang-2 that were stor ed in Weibel-Palade Bodies (WPB).These results indicate that the tumor cells secrete ce rtain factors that in turn stimulate WPB exocytosis/Ang-2 secretion fr om endothelial cells, therefore suggesting that tumor cells indirectly affect Ang-2 dependent vascular des tabilization. The indirect effects of tumor cells on Ang-2 dependent ve ssel destabilization can be hindered in the presence of the Ang-2 inhibitor (Figure 3-3). Hypoxia is a major factor in the tumor microenvironment that alters tumor cell behavior to promote angiogenesis (Bergers and Benjamin, 2003). A well-known example in tumor angiogenesis is the upregul ation of VEGF by tumor cells as a response to hypoxia in the microenvironm ent and thus the promotion of angiogenesis. Hypoxia in the tumor microenv ironment can occur due to diffusion or perfusion affects both of which tumor cells experience in the course of the tumors growth (Dewhirst et al., 1996; Vaupel et al., 1998). Diffusion limited hypoxia is referred to as chronic or long term hypoxia and occurs as t he tumor grows and expands in size; the oxygen diffusion limit from normal vasculature is 70-100 m and the rapidly growing tumor quickly surpasses this range (Dewhirst et al., 1996; Vaupel et al., 1998). Acute or short term hypoxia occurs due to perfusion limitations; the tumor vasculature is irregular in shape, leaky and has highly proliferative endothelium. These vessels can suddenly shut down and cause hypoxia for minutes to hours to tumor cells that otherwise had access to oxygen and nutrients (Dewhirst et al., 1996). In the present st udy the response of tumor cells in regards to Ang-2 secretion was ev aluated both in the acute and chronic hypoxic 58

PAGE 59

environments. Results show that neither acut e nor chronic hypoxia led to an increase in Ang-2 secretion from tumor cells; in all cases the range of Ang2 secretion remained low in the range of 5-25 pg/ml (Figure 3-4). Hypoxia alters the behavior of tumor ce lls towards a pro-angiogenic phenotype. The effect of tumor cell cond itioned media, collected under hypoxia, on endothelial cell secretion of Ang-2 was evaluated. Results s how that Ang-2 secretion from endothelial cells in the presence of conditioned media co llected from hypoxic conditions was in the range of 410-450 pg/ml (Figure 3-5). These levels were not significantly different from the Ang-2 levels secreted by endothelial cells in the presenc e of tumor cell conditioned media collected under normal oxygen levels 420-470 pg/ml (Figure 3-5). Furthermore, tumor cell conditioned media collected under hypoxic conditions did not significantly increase endothelial-smooth musc le cell contact loss in co-culture spheres compared to results obtained in the presence of tumor cell conditioned media from normal oxygen levels (Figure 3-6). Collectively, these re sults suggest that the tumor cell secreted factors that influence Ang-2 secretion from endothelia l cells are not altered by hypoxia. Finally, these effects can be impa ired in the presence of the A ng-2 inhibitor (Figure 3-6). In the tumor microenvironment tumor cells ar e not the only cells that are exposed to hypoxic conditions, the endothelial cells in the microenvironment also experience the same conditions that tumor cells do. Theref ore, next the effect of hypoxia on the endothelial cells and their behavio r in regards to Ang-2 secretion was evaluated. Endothelial cell secretion of Ang-2 did not change from normoxic (180-290 pg/ml) to hypoxic conditions (140-230 pg/ml ) (Figure 3-7) despite prev ious reports demonstrating hypoxia as a potent trigger for WPB exocytos is in endothelial cells (Lowenstein et al., 59

PAGE 60

2005; Rondaij et al., 2006). These observations could be attributed to the plasticity and dynamics of specific WPB protein release and even though hypoxia has been shown to enhance Ang-2 mRNA levels it may not be a potent stimulator of Ang-2 release from endothelial cells (Rondaij et al., 2006). As mentioned previously, VEGF is highly secreted by renal cell carcinoma cells especially by VHL null or mutant cells and is, therefore, an abunda nt growth factor in the tumor microenvironment even in the absence of hypoxia (Bergers and Benjamin, 2003). The possibility of VEGF as the tumor cell secreted factor that stimulates Ang-2 release from endothelial cells, as observed in this study, was evaluated. The role of VEGF in stimulating Ang-2 release from endothelial cells is controversial. Matsushita and colleagues (Matsushita et al., 2005) showed that VEGF triggers exocytosis of WPB leading to Von-Willebrand Factor (vWF) re lease from endothelial cells. Tsigkos and colleagues (Tsigkos et al., 2006) demonstrated that VEGF leads to Ang-2 release from bovine lung microvascular endot helial cells; however Fiedler and colleagues (Fiedler et al., 2004) reported that VEGF does not cause Ang-2 release from human umbilical vein endothelial cells. In the present study human recombinant VEGF did not significantly affect Ang-2 release from endothelial cells (Figure 3-8). Even though an upward trend was observed the levels of Ang-2 secret ed were much lower compared to those secreted in the presence of tumor cell condi tioned media, therefore VEGF is most likely not the tumor cell secreted factor signifi cantly stimulating the endothelial cells. Sphingosine-1 phosphate could be such factor ; it is a potent st imulator of WPB exocytosis, is secreted from tumor cells and currently pursued as a target in various disease settings (Jang et al., 2009; Lowenstein et al., 2005; Zu He ringdorf et al., 2013). 60

PAGE 61

In conclusion, this study indicates that the renal cell carcinoma tumor microenvironment has a potent effect on Ang-2 release in to the microenvironment by endothelial cells. Renal cell carc inoma cells do not have a direct but rather an indirect effect on the Ang-2/Tie2 axis in tumor angiogenesis; the minimal levels of Ang-2 secretion from tumor cells most likely does not significantly contribute to Ang-2 dependent vessel destabilization. However, tumo r cells secrete factors that potently stimulate Ang-2 release from endot helial cells. Further investi gation into such factors is needed that may lead to better understanding of the tumor microenvironment and potentially new therapeutic tar gets. Nevertheless, Ang-2 inhibition shows promise as an anti-angiogenic modality as its inhibition led to a decrease in endothelial-peri-endothelial cell contact loss in vitro The reported results support the pur suit of the Ang-2 inhibitor, MEDI3617, in the clinic for cancer therapy. 61

PAGE 62

Figure 3-1. Tumor cell involvement in angiogenesis through the Ang-2-Tie2 axis. Tumor cells may directly influence vascular destabilization by secreting Ang-2 (A), or they may indirectly influence this proce ss by secreting factors that in turn stimulate Ang-2 secretion from the endothelium of vessels (B). 62

PAGE 63

Figure 3-2. Tumor cell conditioned media st imulate Ang-2 secret ion from endothelial cells and Ang-2 dependent destabilization, Ang-2 inhibitor hinders this loss. Renal cell carcinoma cell conditioned media (CM) was collected at 6 hr post media change. HUVEC were exposed to tumor cell CM for 4 hr. Ang-2 secretion was measured by ELISA, data were normalized to 105 cells. Tumor cells, Caki-1 (A), Caki-2 (B), 786-0 (C ), and A498 (D) secreted low levels of Ang-2; tumor cell CM however elev ated basal Ang-2 secretion from endothelial cells. Column, mean; bars, SEM of three independent experiments. *, p<0.05; **, p<0.01; Student t-test. 63

PAGE 64

Figure 3-3. Tumor cell conditioned medi a leads to sphere de stabilization. (A) Stimulation of spheres with Caki-1 CM for 4 hr led to significant endothelial cell loss from the sphere core compared to control. The presence of MEDI3617 (0.5 nM) impaired this loss. Column, mean; bars, SEM. Control and Caki-1 CM (n=12), Caki-1 CM plus MEDI3617 (n=7) spheres/group). **, p<0.01; Mann-Whitney U test. (B) Images demonstrating control, Caki-1 CM induced destabilization of co-culture spheres and its inhibition in the presence of MEDI3617 at 4 hr. Endot helial cells (HUVEC) shown in red (CellTracker Orange CMRA) and smooth muscle cells (HUASMC) shown in green (CellTrace Oregon Green). Images tak en with Nikon eclipse TS100; scale bar, 100 m. 64

PAGE 65

Figure 3-4. Tumor cells secrete minima l levels of Ang-2 under both normoxic and hypoxic conditions. (A) The oxyg en diffusion limit is about 70-100 m, tumor cells past this distance are deprived of oxygen and experience chronic or long term hypoxia. (B) Tumor vasculature is ve ry unstable, vessels can shut down and reopen for minutes to hours. Tumor cells around such vessels experience acute or short term hypoxia. (C-D)Tumor cell media was collected at normal oxygen conditions 6 hr post media change, 6 (acute) and 48 (chronic) hr post hypoxia exposure. Ang-2 secretion was measured by ELISA, data were normalized to 105 cells. Normal oxygen levels, short term (acute) or long term (chronic) hypoxic conditions did not affect Caki-1 (C), Caki-2 (D), 786-0 (E), and A498 (F) secretion of Ang-2. Co lumn, mean; bars, SEM of three independent experiments. 65

PAGE 66

Figure 3-5. Tumor cell conditioned media from hypoxia do not enhance Ang-2 secretion from endothelial cells. HUVEC were ex posed to tumor cell CM (normal or acute hypoxia) for 4 hr. Ang-2 secreti on was measured by ELISA, data were normalized to 105 cells. Tumor cell CM from Caki-1 (A), Caki-2 (B), 786-0 (C), and A498 (D) significantly elevated Ang-2 secretion from endothelial cells in both conditions (normal or hypoxic c onditions) compared to basal levels, however no difference in Ang-2 secr etion from endothelial cells was shown between tumor cell CM collected under normal oxygen conditions or hypoxic conditions. Column, mean; bars, SEM of three independent experiments. *, p<0.05; Student t-test. 66

PAGE 67

Figure 3-6. Hypoxia does not enhance tu mor cell conditioned media mediated sphere destabilization. The presence of tumor ce ll CM from either normal or hypoxic conditions led to endothelial cell loss fr om the sphere core. No significant difference in endothelial cells loss was seen between tumor cell CM from normal or hypoxic conditions. The presence of MEDI3617 (0.5 nM) impaired endothelial cell loss due to tumor cell CM from both normal (Figure 3-3) or hypoxic conditions. Column, mean; bars, SEM. Control (n=13), Caki-1 CM normal (n=5); Caki-1 CM hypoxia (n =13); Caki-1 CM plus MEDI3617 (n=8) spheres/group). *, p< 0.05; *** p<0.0001; Mann-Whitney U test. 67

PAGE 68

Figure 3-7. Hypoxia alone has no significant effect on Ang-2 secretion from endothelial cells. HUVEC (A) or HMVEC-L (B) was ex posed to either acute (6 hr) or chronic (24 hr) hypoxia. Neither condition seemed to significantly affect Ang-2 secretion compared to normal oxygen le vels. Column, mean; bars, SEM of three independent experiments. 68

PAGE 69

Figure 3-8. VEGF does not significantly stim ulate endothelial cell se cretion of Ang-2. HUVEC (A) and HMVEC-L (B) were exposed to VEGF (5, 25, or 50 ng/ml) for 1, 4, 8, or 24 hr; overall no significant difference in endothelial cell secretion of Ang-2 was observed. Ang-2 secretion wa s measured by ELISA, data were normalized to 105 cells (A) and 104 cells (B). Column, mean; bars, SEM of three independent experiments. **, p<0.01; Student t-test. 69

PAGE 70

CHAPTER 4 ANGIOPOIETIN-2 INHIBITION ENHANCES THE NORMAL VASCULAR PHENOTYPE Chapters 2 and 3 focused on the in vitro evaluation of Ang-2 dependent vessel destabilization and alluded to microenvironmental e ffects on this axis. In all cases, the presence of the Ang-2 inhibitor led to reduc tion of sphere destabilization suggesting that the inhibitor does neutralize the function of Ang-2. Chapter 4 evaluated the inhibitor as an anti-angiogenic agent; the inhibitors effect on the tu mor vasculature was assessed. Background Anti-angiogenic agents targeting the formati on of new vasculature have become a standard of care in various cancer settings in the last decade (Ebos et al., 2009; Loges et al., 2010). The FDA approved agents current ly used in the clinic mainly target the VEGF pathway, however some also inhi bit the PDGF, cMET and mTOR pathways. In general, these drugs do not show sufficient anti-tumor effects alone, however, complement other conventiona l therapies such as chemotherapy, radiation therapy and surgery. Even though there are various success stories with these agents the full potential of anti-angiogenic ther apies has not been met (Ebos et al., 2009; Loges et al., 2010). Patients are often resistant or becom e resistant over time with prolonged treatment with anti-an giogenics; such resistance st ems from the redundancy of endothelial cell activation and t herefore various other pathways may compensate when one is blocked, allowing the tumor to resume growth (Ebos et al., 2009; Loges et al., 2010). With the recent discovery of the Angiopoi etin and Tie2 axis there is potential to overcome this resistance. This pathway is involved in a different process in angiogenesis than all the currently used ant i-angiogenic agents; inhibiting vascular 70

PAGE 71

destabilization rather than endothelial cell activation. There are several agents now being evaluated both preclinically and clinically that inhibit this axis (Gerald et al., 2013). Normal vasculature is comprised of endothelial and peri-endothelial cells and the Ang-2/Tie2 axis destabilizes the cont acts between the two cells types during angiogenesis. This exposes the normally prot ected endothelial cells to pro-angiogenic factors found in the microenvironment that in turn stim ulate the endothelial cells to proliferate, migrate and fo rm tubes (Armulik et al., 2005; Augustin et al., 2009; Carmeliet, 2005). Therefore, the current anti-angiogenic agents inhibit the second step in angiogenesis or the activati on of the endothelium. The Ang2/Tie2 axis is prevalent in the first step in angiogenesis or the destabilization of normal vasculature. In the present study the effect of the Ang2 inhibitor was evaluated in renal cell carcinoma induced angiogenesis. The following studi es focused on the early initiation of angiogenesis and development of tumor nodules. The structure of the tumor and surrounding vasculature was evaluated upon Ang-2 inhibition. Materials And Methods Reagents MECA-32 was purchased from BioLegend (San Diego, CA), NG2 was obtained from Millipore (Temecula, CA ). AlexaFluor 488 and 594 were purchased from Invitrogen (Grand Island, NY). VectaS hield mounting medium with DAPI was purchased from Vector Labs Inc. (Burlingame, CA). Ti ssue-Tek OCT Compound was purchased from Sakura Finetek (Torrance, CA). 2-methylbutane was obta ined from Thermo Fisher Scientific (Waltham, MA). 71

PAGE 72

Cell Culture Human clear cell renal cell carcinoma Caki -1 and Caki-2 cell lines were received as a gift from Dr. Susan Knox (Stanford University). They were grown in Dulbeccos modified minimum essential medium (D-ME M, Invitrogen) supplemented with 10% FBS (Invitrogen), 1% penicillin-streptomycin (I nvitrogen), and 1% 200-mmol/L L-glutamine (Invitrogen). Cells were kept at 37 C, 5% CO2. Cells were used between passages 2 and 10. Drug Preparation Angiopoietin-2 inhibitor, MEDI3617 wa s kindly provided by MedImmune, LLC. The stock solution (5 mg/ml) was diluted to working concentrations in sodium citrate buffer solution. Stock solutions were kept at -80 C and working concentrations at 4 C. Intradermal Angiogenesis Assay All in vivo procedures were conducted in agr eement with a protocol approved by the University of Florida Institutional An imal Care and Use Committee. Female athymic nu/nu mice were injected intradermally with 105 Caki-2 cells in 10 l volume at four sites on the ventral surface. Beginni ng the day prior to tumor cell injection, mice were treated with IP injection of MEDI3617 (2, 10 mg/kg) every 3 days (3 doses total) up to six days post tumor cell inoculation. Mice were t hen euthanized, tumors measured via calipers and tumor volume (mm3) calculated, assuming the tumor volume to be an ellipsoid, using the following equation: tumor volume = /6 x d1 x d2 x height. Skin flaps were then removed and vessels growing into tumor no dules were counted using a Leica MZ16F dissecting microscope with Leica KL 1500 LCD fiber optic illuminator (Leica Microsystems Inc., Buffalo Grove, IL) at 2.5x original magnification (1,2,3). Images were captured with a Retiga EXi Fast1394 digital CCD camera (QImaging, British Columbia, 72

PAGE 73

Canada) and OpenLab5 software (PerkinElmer Inc., Waltham, MA). Statistical significance between control and treated groups was determined using the MannWhitney U-Test at p<0.05. Immunohistochemistry Intradermal tumors were fresh frozen in OCT and methylbutane and cryosectioned at 5 m thickness using Leica CM 3050S cryostat (Leica Microsystems Inc., Buffalo Grove, IL); se ctions were placed on superfrost plus gold slides (Thermo Fisher Scientific Inc., Wa ltham, MA) and kept at -80 C until immunohistochemical staining. Tissue sections were acetone fi xed for 10 minutes, blocked in 2% normal horse serum in 1x TBS, and incubated overnight at 4 C with MECA-32 and NG2 primary antibodies. Secondar y antibodies AlexaFluor 488 and 594 were added onto slides for 1 hr. Tissue sections were imaged with a Zeiss Axioplan 2 imaging microscope (Carl Zeiss, Inc., Thornwood, NY) with EXFO X-Cite 120 light source (Lumen Dynamics Group Inc., Ontario, Canada). Images were taken with a Retiga EXi Fast digital CCD camera (Qimaging, British Columbia, Canada) and processed in OpenLab5 software (PerkinElmer Inc., Waltham, MA); Rhodamine for MECA32/AlexaFluor594, FITC for NG-2/AlexaFluo r488 and DAPI filters were used. Vessel counts were obtained by taking up to ten random fields/tumor with 20x objective, counting the number of vessels in each random field. The number of peri-endothelial cell covered vessels was counted in the tu mor periphery for each tumor. Statistical significance between control and treated groups was determined using the MannWhitney U-Test at p<0.05. 73

PAGE 74

Results Ang-2 Inhibitor Impairs Angiogenesis In Both VHL No rmal And Mutated Renal Cell Carcinoma Models The effect of Ang-2 inhibition on tumo r cell induced angiogenesis was evaluated in both a VHL normal and mutated renal cell carcinoma model. Athymic nude mice were injected with 1x105 Caki-1 (VHL normal) or Caki-2 (VHL mutated) human renal cell carcinoma cells (Figure 4-1). The mice rece ived a single dose of the Ang-2 inhibitor and tumor nodules were evaluated 3 days post tumo r cell injection (Fig ure 4-2A). Results show that Ang-2 inhibition led to a 1.2-, 2.2(p<0.01), and 3-fold (p<0.001) reduction in the number of Caki-1 tumor nodule associated vessels in a dose dependent manner at 2, 10, and 20 mg/kg respectively (Figure 42B). Similar results were obtained with treatment of Caki-2 nodules, t he Ang-2 antibody led to a 1.4(p<0.01), 1.8(p<0.001), and 2-fold (p<0.001) reduction in tumor nodule associated vasculature in a dose dependent manner (Figure 4-2C,D). Ang-2 Inhibitor Hinders Both Tumor Gr owth And Angiogenesis In Renal Cell Carcinoma Xenograft Model To evaluate the effects of Ang-2 target ing on initiation of angiogenesis and tumor growth, nude mice were inoculated intrade rmally with Caki-2 (mutated VHL) human renal cell carcinoma cells (1x105) and treated with the Ang-2 inhibitor (Figure 4-3A). Results showed that the Ang-2 inhibitor reduced the tumor volume by 2.1 and 3.9-fold (p<0.01) (Figure 4-3B), and vessel number by 1.4 and 2.1-fold (p<0.01, p<0.0001) (Figure 4-3C) at 2 and 10 mg/kg re spectively compared to control. Immunohistochemical analysis revealed that the Ang-2 inhibitor significantly reduced (1.6 and 1.7-fold for 2 and 10 mg/kg respecti vely) the number of vessels in the tumor core compared to control (p<0.01) (Figure 4-4). 74

PAGE 75

Ang-2 Inhibition Increases Peri-Endothelial Cell Covered Vessels In Both The Tumor Periphery And Core Normal vasculature is composed of an endothelial cell layer surrounded by a peri-endothelial cell layer. Vascular structur e of tumors from the intradermal assay (Figure 4-3, 4-4) was determined using i mmunohistochemistry. Figure 4-5 demonstrates normal vasculature; nuclei for both the endot helial (red, MECA-32) and peri-endothelial cell layer (green, NG2) are visible with the peri-endothelial cell layer surrounding the endothelium. Treatment of tumors with the Ang-2 inhibito r led to a 1.5and 3.4-fold increase in the 2 and 10 mg/kg groups, respecti vely, compared to control in the tumor periphery (p<0.05, p<0.0001 respectively) (Fi gure 4-6). A significant difference of 2.2fold was noted in the 10 mg/kg group compar ed to the 2 mg/kg group (p<0.05) (Figure 4-6). Similar trends were seen within the tumor, treatment with the inhibitor led to 2and 4.2-fold increase in the number of peri-endothelial cell covered vessels in the 2 and 10 mg/kg groups, respectively, compared to cont rol (p<0.05, P<0.01 respectively) (Figure 4-7). A significant difference was seen bet ween the 2 and 10 mg/kg groups with a 2.1fold increase in the number of peri-endothelia l cell covered vessels in the 10 mg/kg group (p<0.05) (Figure 4-7). Discussion Renal cell carcinoma is a highly vascula rized disease and anti-angiogenic agents are commonly used as both first and second line treatments in patients. Although in general these agents are beneficial especially in the combination of conventional cancer therapies there are patients who fail to res pond or become resistant to such treatments (Ebos et al., 2009). The intradermal angiogenesis assay (Figure 4-1) allows for the rapid evaluation of anti-angiogenic agents by the assessment of vasculature penetrating the 75

PAGE 76

tumor nodule. The current studies evaluated a new class of anti-angiogenic agents that inhibit the Angiopoietin/Tie-2 axis, the Ang-2 inhibitor, MEDI 3617, and its effect on tumor cell induced angiogenesis was determined. Results show that the Ang-2 inhibitor with a single dose effe ctively impaired the early development of human r enal cell carcinoma induced angiogenesis (Figure 4-2). Since renal cell carcinomas commonly hav e VHL mutations rendering the disease highly aggressive and vascularized, both a VHL normal (Caki-1) and mutated (Caki-2) cell lines were evaluated. In either case t he Ang-2 inhibitor significantly impaired the number of tumor nodule associated vessels in a dose dependent manner by reducing the number of vessels 1.2-3 fold in Caki-1 model (Figure 4-2B) and 1.4-2 fold in Caki-2 model (Figure 4-2C,D). The difference in angi ogenic potential of th e two cell lines were clearly visible, the VHL muta ted cell line, Caki-2, had a 2.3fold higher number of tumor nodule associated vessels than the VHL normal Caki-1 cells did. However the Ang-2 inhibitor was able to impair angiogenesis signi ficantly although it is evident that the range of inhibition was lower in the VHL mutated tumor (Figure 4-2). This observation most likely account for the fact that the VHL mutated Caki-2 cells constitutively expresses VEGF and may escape the Ang-2/Tie2 axis early on. Even though there was a slightly lower inhi bitory response to the Ang-2 inhibitor in VHL mutated Caki-2 cells, when the tumor cells were allowed to grow and develop for twice as long (Figure 4-3A) receiving three doses of the Ang-2 antibody, the results showed a reduction of both the tumor volume (Figure 4-3B) and vessel number (Figure 4-3C). Impairment of tumor vasculature was further conf irmed by immunohistochemical analysis of tumor nodules (Figure 4-4). T hese results are consistent with studies 76

PAGE 77

involving the Ang-2 inhibitor in other tumo r settings as well as various other Ang-2 targeting agents (Brown et al., 2010; Daly et al., 2013; Falcon et al., 2009; Leow et al., 2012). The Ang-2/Tie2 axis is involved in the destabilization of normal vasculature by disrupting the endothelial and peri-endothelial cell contacts t herefore the structure of the tumor and surrounding vasculature were det ermined after treatment with the Ang-2 inhibitor. Results show that Ang-2 inhi bition led to a dose dependent increase in the number of vessels that had peri-endothelial cell coverage both in the tumor periphery (Figure 4-6) and within the tumo r core (Figure 4-7). The in crease in peri-endothelial cell coverage upon treatment with t he Ang-2 inhibitor was also observed with the Ang-1/2 peptibody Trebananib (AMG386) Therefore this phenomenon is not exclusive to MEDI3617 but supports the role of Ang-2 in vascular destabilization (Falcon et al., 2009). Since the sections obtaine d for these tumors only show a snapshot in time it can not be determined whether the pe ri-endothelial cell coverage of vessels was maintained as the tumor nodule developed or whether upon Ang-2 inhibition the tumor associated vessels were able to recruit peri-endothelial cells to the tumor vessels via the Ang1/Tie2 interactions. In either case treatment with the Ang-2 i nhibitor led to an increase in normal vascular phenotype and reduced the num ber of tumor vessels. These data support the advancement of this agent as an anti-angiogenic drug for the treatment of cancer. 77

PAGE 78

Figure 4-1. Intradermal angiogenesis a ssay. Tumor cells were injected at 105 cells underneath the skin on the ventral side at four sites. Tumor cells were allowed to grow for either 3 or 6 da ys post injection at which point tumor nodules were measured (6 day) and ski n flaps removed and the number of blood vessels growing into t he tumor counted (3,6 day) 78

PAGE 79

Figure 4-2. Ang-2 inhibition reduces tumo r cell induced blood vessel formation. Mice were injected intradermally with tumor cells (Figure 4-1) and treated with MEDI3617 (2 ,10, and 20 mg/kg) beginn ing the day prior to tumor cell inoculation (A). The number of blood vessels initiated by (B) human renal cell carcinoma Caki-1 and (C) Caki-2 tumor ce lls were determined at the end of a 3 day period. Lines, mean; bar, SD (n= 16). **, p<0.01: ***, p<0.0001; MannWhitney U test. (D) Representative images of blood vessels induced by Caki2 tumor cells using a dissecting microscope at 2x magnification. Large arrow head, tumor nodule; small arrows, tumo r induced blood vessels. Adapted with permission from ELSEVIER: Microvascular Research 2012; 83(3):290-7. 79

PAGE 80

Figure 4-3. Ang-2 inhibition impedes both tumor growth and angiogenesis. Mice were injected intradermally with Caki-2 renal cell carcinoma cells (Figure 4-1) and treated with MEDI3617 ( 2 or 10 mg/kg) beg inning the day prior to tumor cell inoculation (A). Tumor volume (B) and the number of tumor cell induced blood vessels (C) were determined at the end of a 7 day period. Bar, mean with SEM (n=16); line, median (n=12). **, p<0.01; ***, p<0.0001; Mann-Whitney UTest. 80

PAGE 81

Figure 4-4. Ang-2 inhibition reduces the number of tumor vessels. Mice were treated with MEDI3617 (Figure 4-1). The number of blood vessels within the tumor were evaluated using immunohistochem istry of fresh frozen tissue. Vasculature was stained with MECA-32. Line, median. 0, 2 mg/kg (n=8), 10 mg/kg (n=7) **, p<0.01; Mann-Whitney U-Test. (B) Representative images of the median of each group. Images ta ken with Zeiss Axioplan Imaging2 microscope with 20x objective; scale bar = 140 m. 81

PAGE 82

Figure 4-5. Normal vasculature is co mposed of endothelial and peri-endothelial cells. Arrows, nuclei associated with endothelia l and peri-endothelia l cells. Red, MECA-32 (endothelium); gr een, NG2 (periendothelial cells), blue, DAPI. Images taken with Zeiss Axioplan Imagi ng2 microscope with 63x objective; scale bar = 46 m. 82

PAGE 83

Figure 4-6. Ang-2 inhibition leads to in creased normal vascular phenotype in the tumor periphery. The number of peri-endothelial cell covered vessels increased in a dose dependent manner in the tumor peri phery. Line, median. 0 (n=8), 2 mg/kg (n=7), 10 mg/kg (n=6) *, p<0. 05; ***p<0.0001; Mann-Whitney U-Test. White arrows show peri-endothelial cell covered vessels. Red, MECA-32 (endothelium); green, NG2 ( peri-endothelial cells); blue, DAPI. Images taken with Zeiss Axioplan Imaging2 microscope with 20x objective; scale bar = 140 m 83

PAGE 84

Figure 4-7. Ang-2 inhibition leads to increase normal vascular phenotype within the tumor. The number of per i-endothelial cell covered vessels increased in a dose dependent manner within the tumor core. Line, median. 0 (n=8), 2 mg/kg (n=7), 10 mg/kg (n =6) *, p<0.05; **p<0.001; Mann-Whitney U-Test. White arrows show peri-endothelial cell covered vessels. Red, MECA-32 (endothelium); green, NG2 ( peri-endothelial cells); blue, DAPI. Images taken with Zeiss Axioplan Imaging2 microscope with 20x objective; scale bar = 140 m. 84

PAGE 85

CHAPTER 5 DORSAL SKINFOLD WINDOW CHAMBER MODEL TO EVALUATE EARLY TUMOR CELL INDUCED ANGIOGENESIS AND ANTI-ANGIOGENIC THERAPY Chapter 4 evaluated the ant i-angiogenic effect of Ang-2 inhibition using an intradermal angiogenesis assay. The assay is an endpoint evaluation of impacts on the tumor vasculature. Chapter 5 used an in vivo murine dorsal skinfold window chamber model to evaluate anti-angioge nic therapy real time. This model allows for the assessment of tumor and vascular growth on a daily basis by imaging mice surgically implanted with titanium chamber s bearing tumor. The following work was conducted in collaboration with Jenn ifer Lee and Dr. Brian Sorg in the Biomedical Engineering Department at the University of Florida. Background The dorsal skinfold window chamber is commonly used to evaluate the microvasculature in various settings in vivo The model involves surgical implantation of a titanium window onto exposed microvasculature on the dorsal skin of mice (Moy et al., 2011; Palmer et al., 2011). This method has been used to evaluate for example angiogenesis in endometriosis and tumor developm ent; it was also used to evaluate the early phases of preclinical development of Bevacizumab in t he 1990s (Borgstrom et al., 1998; Borgstrom et al., 1999; Borgstrom et al., 1996; Laschke and Menger, 2007; Li et al., 2000; Yuan et al., 1996). Hyperspectral imaging of hemoglobin saturation has allowed for the evaluation of oxygenation and hypoxia, oxygen transport dynamics and characterization of the abnormal vascular ph ysiology and acute oxygen fluctuations within the tumor microvasculature (Dedeugd et al., 2009; Hardee et al., 2009; Sorg et al., 2005; Wankhede et al., 2010a). Tumor microv ascular response to various therapies 85

PAGE 86

such as radiotherapy, vascular disrupting agents and sickled erythrocytes has also been evaluated using this imaging system (Dewhi rst et al., 2007; Terman et al., 2013; Wankhede et al., 2010b). Settings other than t he tumor microvasculature have also been explored for example the characteriza tion of arteriovenous malformation in hereditary hemorrhagic talengiectasia, t he formation of spontaneous and induced microvascular thrombosis and occlusions and immune cell localization and tissue damage in particle based vaccines (Choe et al., 2010; Park et al., 2009; Wankhede et al., 2010a). Anti-angiogenic agents are commonly evaluated in vivo using either matrigel plug or intradermal assays as well as histologic al assessments of the number and function of tumor associated vasculature (Auerbach, 2008). To date, the murine dorsal skinfold window chamber model has not been widely utiliz ed to evaluate tumor response to antiangiogenic agents using hyperspec tral imaging to assess not only the vascular density of the tumor but also its oxygenation status. This model allows for an in vivo real time assessment of tumor vasculature at microvessel resolution and has tremendous potential to answer several important questions regarding aspects of vascular response to anti-angiogenics such as oxygenation status of the vascu lature. The current study evaluated both a VEGF and Ang2 targeted approach on the induction of early human renal cell carcinoma cell induc ed angiogenesis. Limitations of this model in this particular setting quickly became apparent and are discussed here. Materials And Methods Reagents Mouse Ang-2 ELISA kit was purchased from MyBioSource (San Diego, CA). MECA-32 was purchased from BioLegend (San Diego, CA), NG2 was obtained from 86

PAGE 87

Millipore (Temecula, CA). AlexaFluor 488 and 594 were purchased from Invitrogen (Grand Island, NY). VectaS hield mounting medium with DAPI was purchased from Vector Labs Inc. (Burlingame, CA). Ti ssue-Tek OCT Compound was purchased from Sakura Finetek (Torrance, CA). 2-methylbutane was obta ined from Thermo Fisher Scientific (Waltham, MA). Cell Culture The human clear cell renal cell carcinoma, Caki-2, cell line was received as a gift from Dr. Susan Knox (Stanford University). Caki-2 was grown in Dulbeccos modified minimum essential medium (D-MEM, Invi trogen, Grand Island, NY) supplemented with 10% FBS (Invitrogen, Grand Island, NY), 1% penicillin-streptomycin (Invitrogen, Grand Island, NY), and 1% 200-mmol/L L-glutamine (Invitrogen, Grand Island, NY). Cells were kept at 37 C, 5% CO2. Drug Preparation Ketamine and xylazine were purchased fr om Webster Veterinary (Devens, MA) and prepared in sterile saline. Angiopoietin-2 inhibitor, MEDI3617, was kindly provided by MedImmune, LLC. The stock solution (5 mg/ml) was diluted to the working concentration (10 mg/kg) in sodium citrate buffer solution. Stock solutions were kept at 80 C and working concentrations at 4 C. Sunitinib was obtained from LC Laboratories (Woburn, MA) and stored at -20 C. Worki ng concentration of Sunitinib was prepared fresh daily by making stock and diluent buffe rs of citric acid monohydrate and sodium citrate dihydrate at pH 6.8 and 3.2 respectively. A 1:7 stock to diluent solution was made (~pH 3.3) and acidified to pH 1.0, S unitinib was dissolved, and the solution was adjusted to pH 3.5. Working concentration of Sunitinib was kept at room temperature. 87

PAGE 88

Window Chamber Surgery And Tumor Initiation All in vivo procedures were conducted in agr eement with a protocol approved by the University of Florida Institutional An imal Care and Use Committee. Dorsal skinflap window chamber surgeries were carried out as previously described by Moy and colleagues (Moy et al., 2011). Brie fly, female athymic nu/nu mice (Harlan Laboratories, Indianapolis, IN) were surgically implanted with a titanium window chamber on the dorsal skinflap. During the surgical procedure mice were anesthetized via intraperitoneal (IP) injecti on of ketamine (100 mg/kg) and xylazine (10 mg/kg). Human renal cell carcinoma, Caki-2, tumor was init iated in the window chamber during surgery by injecting cells (2x104 in 10 l volume) subqutaneously in the dorsal skinflap prior to placing a 12 mm diameter number 2 round glass cover slip (Erie Scientific, Portsmouth, NH) over the exposed skin. Post surgic al procedure, animals were housed in an environmental chamber maintained at 33 C and 50% humidity with standard 12 hr light/dark cycles for the re mainder of the study. Treatment Of Window Chamber Tumors Mice were treated with Sunitinib (100 mg/kg) daily via oral gavage or with MEDI3617 (10 mg/kg) every 3 days via IP inje ction, starting the day of window chamber surgery/tumor initiation up to day 11 postsurgery when mice were euthanized. During the study, tumors were measured daily using calipers and tumor volume (mm3) was calculated, assuming that the tumor volume wa s half of an ellipsoid, using the following equation: tumor volume = [ /6 x d1 x d2 x height]. Statistical significance was determined using Mann-Whitney U-Test. 88

PAGE 89

Hyperspectral Imaging The spectral imaging system, image ac quisition, and image processing methods have previously been described (Moy et al., 2011; Sorg et al., 2005). Briefly, window chamber tumors were imaged daily using a Ze iss AxioImager microscope (Carl Zeiss, Inc., Thornwood, NY) with 100-W tungsten halogen lamp, CCD camera thermoelectrically cooled to -20 C (DVC Co., Austin, TX; Model no.1412AM-T2-FW) and C-mounted liquid crystal tunable filter (LCTF) (CRI Inc., Woburn, MA). Tuning of the LCTF and image acquisition with the CCD came ra was automatically controlled with LabVIEW8 software (National Instruments Corp., Austin, TX). Vascular hemoglobin saturation measurements and images were cr eated from the spectral image data, image processing was performed using Matlab software (The Mathwo rks Inc., Natick, MA). During imaging mice were placed on a heated platform and anesthetized with isofluorane (Webster Veterinary, Devens, MA). Immunohistochemistry Eleven days post surgery/tumor cell inoc ulation the mice were euthanized with euthasol (0.01 ml/g) (Webster Veterinary, Devens, MA), titanium window chambers were removed and tumors were fresh froz en in OCT and methylbutane. Tumors were sectioned at 5 m thickness using Leica CM 3050S cr yostat (Leica Microsystems Inc., Buffalo Grove, IL); sections were placed on superfrost plus gold slides (Thermo Fisher Scientific Inc., Waltham MA) and kept at -80 C until immunohistochem ical staining. Tissue sections were acetone fixed for 10 mi nutes, blocked in 2% normal horse serum, and incubated overnight at 4 C with MECA-32 and NG 2 primary antibodies, at room temperature with secondary antib odies AlexaFluor 488 and 594 for 1 hr. Tissue sections 89

PAGE 90

were imaged with a Zeiss Axioplan 2 imaging microscope (Carl Zeiss, Inc., Thornwood, NY) with EXFO X-Cite 120 light source (Lum en Dynamics Group Inc., Ontario, Canada). Images were taken with a Retiga EXi Fast digital CCD camera (QImaging, British Columbia, Canada) and processed in OpenLab5 software (PerkinElmer Inc., Waltham, MA); Rhodamine for MECA-32/AlexaFluor594, FITC for NG-2/AlexaFluor488 and DAPI filters were used. Vessel c ounts were obtained by taking up to ten random fields/tumor at 20x objective, counting the number of vessels in each field. The number of periendothelial cell covered vessels was counted in the tumor periphery for each tumor. Statistical significance bet ween control and treated groups was determined using the Mann-Whitney U-Test at p<0.05. Serum Angiopoietin-2 Levels Blood from the tail vein wa s drawn from mice (2 mice /group) that (1) did not receive surgery (baseline), (2) that receiv ed surgery and ones (3) that received surgery with tumor cells implanted at da ys 3 and 5 post surgery. About 100 l of blood per mouse was collected and placed on ice for 2 hrs to let the blood clot. Blood was then centrifuged for 15 minutes, 4 C at 1000 g. Serum was colle cted and stored at -80 C until analysis. Serum Ang-2 levels were calculated using a m ouse Ang-2 ELISA kit. Based on company recommendation, manufactu rers protocol was altered to load 30 l of either standards or serum and 50 l of 3.3 fold diluted conjugate in 0.9% NaCl per well and incubate for 2 hrs at 37 C; manufacturers protocol was otherwise followed. Intradermal Angiogenesis Assay All in vivo procedures were conducted in agreem ent with a protocol approved by the University of Florida Institutional An imal Care and Use Committee. Female athymic 90

PAGE 91

nu/nu mice were injected intradermally with 105 Caki-2 cells in 10 l volume at four sites on the ventral surface. Beginni ng the day prior to tumor cell injection, mice were treated with either daily oral gavage of Sunitinib ( 100 mg/kg) or IP injection of MEDI3617 (10 mg/kg) every 3 days up to six days post tumor cell inoculation. Mice were then euthanized, tumors measured via calipers and tumor volume (mm3) calculated, assuming the tumor volume to be an ellips oid, using the following equation: tumor volume = /6 x d1 x d2 x height. Skin flaps were then re moved and vessels growing into tumor nodules were counted using a Leica MZ16F dissecting microscope with Leica KL 1500 LCD fiber optic illuminator (Leica Micro systems Inc., Buffalo Grove, IL) at 2.5x original magnification. Images were captured with a Retiga EXi Fast1394 digital CCD camera (QImaging, British Columbia, C anada) and OpenLab5 software (PerkinElmer Inc., Waltham, MA). Statistical signifi cance between control and treated groups was determined using the Mann-Wh itney U-Test at p<0.05. Results Human Renal Cell Carcinoma, Caki-2 Growth In The Window Chamber Renal cell carcinoma is a highly vascula rized disease and anti-angiogenic agents are currently used as both first and second line treatments in patients. Caki-2 cells, VHL mutant and highly vascular and aggressively gr owing cell line, were implanted into the window chamber (Figure 5-1A). Induction of tumor cell induced angiogenesis was seen at day 5-6 post tumor cell im plantation at which point the tumors quickly expanded and became heavily vascularized with hemogl obin saturation 50-80% in the microvasculature at day 8 (Figure 5-1B). 91

PAGE 92

VEGF Inhibition Impedes Tum or And Vessel Growth In Caki-2 Window Chamber Tumors Caki-2 tumors grew rapidly and expanded their microvasculature greatly over time with fairly good oxygenati on. Treatment of mice with the VEGF inhibitor led to a significant 5.2-fold reduction in tumor volu me (p<0.01) (Figure 5-2A) as well as a dramatic impairment of tumo r vasculature and oxygenati on, a reduction from 60-80% hemoglobin saturation to 0-40% in the treated groups (Figure 5-2B). Immunohistochemical analysis of tumor vasc ulature at study endpoint revealed a 2.4fold reduction in the treated groups compar ed to control (p<0.0001) (Figure 5-3). Ang-2 Inhibition Does Not Affect Caki-2 Tumor Growth In The Window Chamber Treatment of mice with t he Ang-2 inhibitor did not show impairment of tumor growth or vascular development (Figure 54). However, immunohistochemical analysis of tumor vasculature at study endpoint reve aled a 1.1-fold reduction in the treated groups compared to control (p<0.05) (Figur e 5-5). Furthermore, analysis of the vascular structure revealed a 4-fold increase in t he number of vessels t hat maintained periendothelial cell coverage in the treated groups compared to c ontrol (p<0.05) (Figure 56). VEGF And Ang-2 Inhibition In The Intradermal Angiogenesis Model To evaluate the inhibition of both the VEGF/VEGFR and Ang2/Tie2 pathways in the absence of the surgery involved with the dorsal skinflold window chamber model an intradermal assay was used. VEGF/VEGFR pathway was inhibited once again with Sunitinib (100 mg/kg) and led to the reducti on of both tumor and vessel growth by 43 (p<0.001) (Figure 5-7A) and 2.5-fold (p<0.0001) (Figure 5-7B) respectively compared to control. Ang-2/Tie2 pathway was inhibited by MEDI3617 (10 mg/kg) and also led to a 92

PAGE 93

reduction of both tumor and vessel growth by 3.1 (p<0.001) (Figure 5-7C) and 1.6-fold (p<0.0001) (Figure 5-7D) respectively compared to control. Surgery Associated With The Window Ch amber Model Leads to An Increase In Serum Ang-2 Levels The circulating Ang-2 levels were determi ned in mice that received surgery in order to explain the discrepancy between results in the window chamber model versus the intradermal assay regarding Ang-2 target ing in angiogenesis. The release of Ang-2 from endothelial cells as a wound healin g response has been previously noted (Kampfer et al., 2001; Kong et al., 2010). Re sults show that mice that underwent surgery have an increased level of Ang-2 in their circulation compared to control mice that did not receive surgery (Figure 5-8). Furthermore, the presence of tumor cells in mice that received surgery further increas ed the Ang-2 levels in the circulation compared to mice that only received surgery. Discussion Vascular-targeted agents that inhibit the formation of new blood vessels from pre-existing ones have become a standard of ca re in several cancer settings over the last decade (Ebos et al., 2009; Loges et al ., 2010). Our understanding of the complexity of the angiogenic process and the tumor microenv ironment led to t he realization that anti-angiogenic therapies have yet to reach their full potential (Ebos et al., 2009; Ferrara, 2010; Loges et al., 2010) With the current FDA approved agents there remains a cohort of patients who do not respond or stop responding to treatment; with this realization and discovery of new important pathways in tumor angiogenesis, there are many new agents currently being evaluated in pr eclinical and clinical trials (Gerald et al., 2013; Loges et al., 2010). 93

PAGE 94

The murine dorsal skinfold window chamber model has been used in the last two decades to study and understand the microvasculature of not only tumors but other diseases as well (Borgstrom et al., 1998; Bo rgstrom et al., 1999; Borgstrom et al., 1996; Choe et al., 2010; Dedeugd et al., 2009; Dewhirst et al., 2007; Hardee et al., 2009; Laschke and Menger, 2007; Li et al., 2000; Moy et al., 2011; Palmer et al., 2011; Park et al., 2009; Sorg et al., 2005; Terman et al ., 2013; Yuan et al., 1996). Recently, the ability to utilize hyperspectral imaging and ev aluate vascular oxygenation status through hemoglobin saturation the window chamber model opened opportunities to study, among others, vascular-targeted agents and their effect on the microvasculature (Wankhede et al., 2010b). The dorsal skinf old window chamber model with high resolution hyperspectral imaging, to our knowledge, has not been widely used to evaluate anti-angiogenc therapy on the early initiation of angiogenesis and development of tumors. Ferrara and colleagues have used the window chamber model with basic imaging techniques to demonstrate the ant i-angiogenic effects of Bevacizumab in preclinical development (Borgstrom et al., 1998; Borgstrom et al., 1999; Borgstrom et al., 1996; Yuan et al., 1996). To dat e, there are no data in the literature using this model to evaluate Ang-2 targeted agents. In the present study, two different classes of anti-angiogenic agents were evaluated. Sunitinib is a small molecule tyrosine kinase inhibitor that targets the VEGFR1-3 and PDGFR and has been FDA approved in 2006 as first line treatment in metastatic kidney cancer (Goodman et al., 2007). The Ang-2 inhibitor, MEDI3617, was used as the Ang-2 targeted agen t. Treatment of the tumor bearing mice began the day of tumor cell injection so that the effects of the anti-angi ogenic agents could be 94

PAGE 95

evaluated on the initiation of tumor cell induced angiogenesis. The current study used the VHL mutated Caki-2 human renal cell carc inoma cells at a small concentration in order to allow a few days before the initiati on of angiogenesis in the window chamber. Results show that Caki-2 cells injected at 2x104 cells initiate angiogenesis 5-6 days post tumor cell injection (Figure 5-1). When Caki-2 tumor bearing mice are treated with 100 mg/kg Sunitinib during an 11 day period t he growth of the tumor is significantly inhibited, 5.2-fold compared to untreated tumors (Figure 5-2A). The development of tumor microvasculature of treated mice is si gnificantly impaired with sparse and poorly oxygenated vessels (Figure 5-2B ). Immunohistochemical analysis of tumors at endpoint further demonstrates the significantly impaire d vasculature of treat ed mice with a 2.4fold reduction in the number of vessels com pared to control tumors (Figure 5-3). The results clearly show that the window cham ber model and hyperspectral imaging can be a useful tool to evaluate agents targeti ng the VEGF pathway and the response of the tumor and microvasculature to such treatment. Resu lts nicely demonstrate not only the inhibition of vascular development but also the poor oxygenation of the microvasculature that t he tumor does possess. On the other hand, the results show a different tumor response to the Ang-2 inhibitor. Mice treated with the inhibitor di d not show any impairment of tumor growth (Figure 5-4A) compared to untreated tumors nor did it seem to have an effect on the vascular density or oxygenation of the vessels (Figure 5-4B). These results were rather puzzling at first but immunohistochemical an alysis revealed a slight but significant reduction in the vessel number (Figure 5-5) as well as a highly significant 4-fold increase in peri-endothelial ce ll covered vessels when tumo rs were treated with the 95

PAGE 96

Ang-2 inhibitor (Figure 5-6) Furthermore, previous results presented in Chapter 4 clearly showed a significant ant i-angiogenic effect of the Ang-2 inhibitor, which is the opposite response that is seen in the window chamber model Therefore, the intradermal assay was repeated with both Suni tinib and MEDI3617 to closely resemble the experiment conducted in the window chamber model. Results from the intradermal assay show that Sunitinib led to a significant 43(Figure 5-7A) and 2.5-fold (Figure 5-7B) reduction in tumor volume and vessel number respectively compared to control correlating to the results seen in the window chamber model. On the contrary to results seen with the Ang-2 inhibitor in the window chamber, treatment with the Ang-2 inhibitor led to a si gnificant 3.1(Figure 5-7C) and 1.6-fold (Figure 5-7D) reduction in tumor volume and vessel number compared to control. Similar results with the VEGF targeted agent but opposing results with the Ang-2 targeted agent led to questions about the difference between the two models. In general, the sole difference between the in tradermal assay and window chamber model is the initial surgery that is involved wit h the window chamber model. Based on the basic biology of the Ang-2 and VEGF axis in ph ysiological response to injury led to the hypothesis that perhaps the su rgery involved with the window chamber model leads to the rapid release of Ang-2 from damaged endothelial cells when the skinflap is cut to expose the vasculature on the skin that is spared. The angiopoietin axis is not only involved with the rapid response to vascular injury as a wound healing response but is also a pro-inflammatory factor (Fiedler and Augustin, 2006; Fiedler et al., 2006; Kampfer et al., 2001; McDonald, 2008 ; Roviezzo et al., 2005; Staton et al.). One can imagine 96

PAGE 97

that the surgery to implant the window chamber would not only elicit a wound healing but also a pro-inflammatory response. To support this hypothesis serum sample s from mice that have not received surgery, mice that received surgery and ones with surgery and tumor cell injection were evaluated at various time point s after surgery. Results demonstrate mice that received surgery had 1.4and 1.5-fold higher Ang-2 levels in the seru m compared to basal levels in mice that did not receive surgery at da ys 3 and 5 post surgery respectively (Figure 58). Furthermore, mice that received surgery and were injected with tumor cells had a slightly higher increase of 1.6-fold com pared to baseline at days 3 and 5 post surgery (Figure 5-8). Due to small animal numbers a nd restrictions on the amount of blood that could be obtained from each mouse the stat istical significance between groups could not be calculated. Nevertheless it is clear that the window chamber surgery did elevate the circulating Ang-2 levels in mice. Evaluating the data betw een the window chamber, intradermal angiogenesis assay and the Ang-2 levels in the serum it ap pears that the surgery led to a release of Ang-2 into the microenvironment of t he window chamber and even though Ang-2 inhibition with the antibody was initiated the day of surgery, perhaps the amount of the inhibitor to Ang-2 ligand was low yieldi ng minimal anti-angiogenic response only noticeable when the tumors were evaluated using immunohistochemical analysis. It is evident that treatment with the inhibitor led to changes in the vasculature both in number (Figure 5-5) and structure (Figure 5-6) suggesting that the inhibitor was attempting to block vascular des tabilization, however, the pr esence of the inhibitor at that particular dose did not elicit a br oader anti-tumor effect. Perhaps an increased 97

PAGE 98

concentration of the inhibitor could lead to significant anti-tumor and anti-angiogenic effects. In conclusion, the murine dorsal skinfold window chamber model is a valuable model to evaluate tumor microvasculature and vascular oxygenation using hyperspectral imaging. Caution with this model should be taken when assessing the vascular response to therapeutic strategies. In this study two differ ent classes of antiangiogenic agents were evaluated. The Ang-2/ Tie2 axis is important in vascular destabilization while the VEGF/VEGFR axis plays a role in endothelial cell activation to proliferate, migrate and form new vessels. While both axes are essential in physiological angiogenesis such as wound hea ling, they nevert heless have very different roles. Endothelial cells store Ang-2 in Weibel-Palade Bodies to be able to quickly respond to environmental changes such as vascular injury (Fiedler et al., 2004; Lowenstein et al., 2005; Rondaij et al., 2006) while VEGF is essential in the formation of new vasculature, a later response in wound healing (Carmeliet, 2005) It is clear that the surgery involved with the window chamber model upsets the normal balance of Ang-2 in the microenvironment leading to skewed results to Ang-2 inhibition. The window chamber model, however, is a great tool to evaluate the inhibition of endothelial cell activation and could be utilized to rapidly and effect ively evaluate anti-angiogenic agents in preclinical studies. 98

PAGE 99

Figure 5-1. Human renal cell carcinoma, Caki-2, tumor growth in the window chamber model. Titanium chambers were surgic ally implanted in nude mice (A) and tumor cells injected (2x104) subcutaneously into the window. (B) Signs of vascular change were visible at about 56 days post tumor cell injection at which point tumors rapidly became more densely vascularized and oxygenated with tumor volume increasing daily. 99

PAGE 100

Figure 5-2. VEGF inhibition impedes tumor and vessel growth in Caki-2 tumors. Mice bearing window chambers with Caki-2 tu mors were treated with Sunitinib. Treatment led to significant reduction in tumor volume (A) and visible inhibition of and poorly oxygenated tumo r vasculature (B). Median + 90/10 percentile; control (n=8), Sunitinib (n=7) (combination of two independent experiments). **, p<0.01, Mann-Whitney U-Test. 100

PAGE 101

Figure 5-3. VEGF inhibition im pedes vessel growth in Caki-2 tumors. Tumors (Figure 52) were frozen and evaluated using immunohistochemistry. Sunitinib treatment led to significant reduction of vessel number within the tumor. Line, median; *, p<0.05 MannWhitney U-Test. Representative images of the median of each group. Red, MECA-32. Images taken with Zeiss Axioplan Imaging2 microscope with 20x objective; scale bar = 140 m. 101

PAGE 102

Figure 5-4. Ang-2 inhibiti on does not affect Caki-2 tumors. Mice bearing window chambers with Caki-2 tumors were tr eated with MEDI3617. Tr eatment did not affect tumor volume (A) or vasculature (B). Median + 90/10 percentile; control (n=8), MEDI3617 (n=7) (combination of two independent experiments). 102

PAGE 103

Figure 5-5. Ang-2 inhibition led to reduc ed tumor vasculature. Tumors (Figure 5-4) were frozen and evaluated using imm unohistochemistry. MEDI3617 treatment led to significant reduction of vessel num ber within the tumor. Line, median; *, p<0.05 Mann-Whitney U-Test. Representat ive images of the median of each group. Red, MECA-32. Images tak en with Zeiss Axioplan Imaging2 microscope with 20x objective; scale bar = 140 m. 103

PAGE 104

Figure 5-6. Ang-2 inhibition leads to in creased peri-endothelial ce ll coverage in the tumor periphery. Window chamber tu mors treated with MEDI3617 were evaluated immunohistochemically for vascular structure. (A) Normal vasculature with peri-endothelial cell coverage. (B) Shows tumor vasculature without peri-endothelial cell coverage. Scale, 46 m. (C) Ang-2 inhibition led to increased number of vessels that had peri-endothelial cell coverage. Line, median; *, p<0.05; MannWhitney U-Test. Representative images of each group. White arrows show peri-endothelial cell covered vessels; yellow arrow shows vessels without peri-endothelial cell coverage. Red, MECA-32 (endothelium); green, NG2 ( peri-endothelial cells); blue, DAPI. Images taken with Zeiss Axioplan Imaging2 microscope with 20x objective; scale bar = 140 m. 104

PAGE 105

Figure 5-7. Ang-2 and VEGF inhibition in th e intradermal assay. Mi ce were injected intradermally with Caki-2 renal cell ca rcinoma cells and treated with either MEDI3617 (10 mg/kg) or Sunitinib (100 mg/kg) beginning the day prior to tumor cell inoculation. Tumor volume (A,C) and the number of tumor cell induced blood vessels (B,D) were dete rmined at the end of a 7 day period. Bar, mean with SEM (n=12) ; line, median (n=12); **, p<0.01; ***, p<0.0001; Mann-Whitney U-Test. 105

PAGE 106

Figure 5-8. Window chamber surgery leads to increase serum Ang-2 levels. The surgery involved with the window chamber morel led to increase Ang-2 levels in the circulation compared to mice that did not undergo surgery. The presence of Caki-2 tumor cells further led to an increase in Ang-2 levels compared to surgery alone. 106

PAGE 107

CHAPTER 6 ANTI-TUMOR EFFECT OF ANGIOPOIE TIN-2 INHBITION AS A MONOAND COMBINATION THERAPY WITH VEGF TARGETED AGENTS Part A: Anti-Tumor Eff ect Of Ang-2 Mono-Therapy Chapters 4 and 5 focused on the in vivo evaluation of the anti-angiogenic effects of Ang-2 inhibition. Chapter 6 explored the anti-tumor effect of Ang-2 inhibition both in the primary tumor and metastatic tumor setting (Part A) as well as combination therapy with VEGF targeted agents (Part B). Background Tumors cannot grow beyond a certain size without the initiation of angiogenesis (Folkman, 1971; Folkman, 2007). Over t he last four decades there has been tremendous interest and progress in angiog enesis targeted therapies (Folkman, 2007). Anti-angiogenic agents have been shown to slow tumor growth and significant tumor growth delay in preclinical studies has led to the approval of several agents for treatment of various solid tumors (Borgstr om et al., 1998; Borgst rom et al., 1999; Borgstrom et al., 1996; Folkman, 2007; G oodman et al., 2007). A new class of antiangiogenic agents the Ang-2/Tie2 target agents has been significantly pursued in the treatment of cancer. Several reports show significant anti-tumor effects with these agents (Brown et al., 2010; Coxo n et al., 2010; Daly et al., 2013; Leow et al., 2012).The current study evaluated the anti-tumor effect of the Ang-2 inhibitor in both VHL normal and mutated human renal cell carcinoma model s as well as a metastatic setting. 107

PAGE 108

Materials And Methods Reagents Protease (P-5147), DNase (DN-25) and collagenase (CO-130) and Bouins solution were obtained from Sigma. Cell culture The human clear cell renal cell carci noma Caki-1 and Caki-2 cell lines were received as a gift from Dr. Susan Knox (Stanford University ); the cells were grown in Dulbeccos modified minimum essential medium (D-MEM, Invitrogen) supplemented with 10% FBS (Invitrogen), 1% penicillin-s treptomycin (Invitrogen), and 1% 200-mmol/L L-glutamine (Invitrogen). Cells were kept at 37 C, 5% CO2. The cells were used between passages 2 and 10. Drug preparation Angiopoietin-2 inhibitor, MEDI3617, wa s kindly provided by MedImmune, LLC. Stock solution (5 mg/ml) was diluted to working concentrations in sodium citrate buffer solution. Stock solutions were kept at -80 C and working concentrations at 4 C. Xenograft dissociation All in vivo procedures were conducted in agr eement with a protocol approved by the University of Florida Institutional An imal Care and Use Committee. Female athymic nu/nu mice were initially injected intramuscula rly (IM) with Caki-1 or Caki-2 cells at about 2x106 in 20 l volume. Tumors were allowed to grow to about 1000 mm3 at which point tumors were harvested, and tumor cell s dissociated from t he stroma. Briefly, tumors were harvested from mice and minc ed with a scissor then placed in 35 ml/tumor enzyme cocktail made of 0.25 mg/ml collagenase, 0.5 mg/m l protease, and 0.4 mg/ml DNase in PBS. Tumor chunks were then inc ubated for 1 hr at 37 C on a rotator. 108

PAGE 109

Dissociated cells were filtered, transferr ed to a new conical tube and centrifuged at 1000 rpm, 10 min at 4 C. Cells were counted and prepared for growth delay injections (as described below). Growth delay tumor initiation and treatment All in vivo procedures were conducted in agr eement with a protocol approved by the University of Florida Institutional An imal Care and Use Committee. Female athymic nu/nu mice were injected IM at 106 cells in 20 l volume with Caki-1 or Caki-2 cells dissociated from previous tumors (as descr ibed above). Anti-tumor effects of the Ang-2 inhibitor were evaluated by (1) allowing tumors to reach a volume of 200 mm3 before treatment began (Caki-1, Caki-2) or (2) begi nning treatment the day of tumor cell injection (Caki-2). Mice were treated with the Ang-2 inhibitor at various doses every three days (twice/week). Tumors were measured with calipers daily and once they reached a volume of 1000 mm3 the mice were euthanized. Lung metastasis tumor initiation and treatment All in vivo procedures were conducted in agr eement with a protocol approved by the University of Florida Institutional An imal Care and Use Committee. Female athymic nu/nu mice were injected with human renal cell carcinoma, Caki-2, cells at 104 in 200 l volume via tail vein. Mice were treated wit h the Ang-2 inhibitor (2,10 mg/kg) every 3 days (twice/week, 8 doses total) starting the day of tumor cell injection. Mice were euthanized 24 days post tumor cell injection, the lungs removed and fixed overnight in Bouins solution. The number of metastases as well as size of each metastasis was counted. Statistical significance between control and treated groups was determined using the Mann-Whitney U-Test at p<0.05. 109

PAGE 110

Results Ang-2 inhibition does not significantly im pact renal cell carcinoma xenograft growth To evaluate the anti-tumor effect of t he Ang-2 inhibitor mice were injected intramuscularly with either Caki-1 or Caki-2 cells. Tumors were grown to a volume of 200 mm3 before treatment was initiat ed. Tumor growth delay to 1000 mm3 was evaluated. Results show that in Caki-1 tumo rs treatment with the A ng-2 inhibitor did not have a significant effect on tumor growth delay yielding only a 1-3 day delay compared to control (Figure 6-1A,B). Even smaller effects of 0-1 day delay were seen in Caki-2 tumors treated with the A ng-2 inhibitor even though the dose was increased compared to doses the Caki-1 bearing mice received (Figure 6-1 C,D). Since significant results were seen with the Ang-2 inhibitor in the early tumor growth and development in the intradermal assay in Chapters 4 and 5, a growth delay experiment was conducted where treatment of mice was in itiated the day of tumor cell injection. However, results from this study were also negative with no growth delay effects in treated groups compared to control (Figure 6-2). Ang-2 inhibition does not impact re nal cell carcinoma lung metastases Angiogenesis is an important process not only for the growth of the primary tumor but also for the develop ment and growth of metastatic lesions. The effect of Ang2 inhibition on lung colonization after tail vein injection of Caki-2 tumor cells was evaluated. Treatment of mice began the day of tumor cell in jection and resumed for a 3week period (Figure 6-3A). Results showed no significant effect on the number of lung metastases between control and treated mice, the median number of lung metastases were 33, 28 and 29 for control, 2 and 10 mg/kg re spectively (Figure 6-3B). Overall there 110

PAGE 111

was no difference between the size of metast atic lesions between control and treated groups (Figure 6-3C) nor the distribution of t he percentage of lesions in a particular size range (Figure 6-3D). Discussion Anti-angiogenic agents have been pursued in the laboratory and the clinic since Dr. Judah Folkman first proposed the importanc e of angiogenesis for tumor growth and the potential for a therapeutic intervention to cut off the blood supp ly to the tumor and impair its growth (Folkman, 1971). The past decade VEGF and other targeted agents have been approved for various solid tumors and are now used as part of the standard of care. The complexity of tumor angiogenesis and the microenvironment that influences the growth of the tumor has been sought after and the more knowledge we gain added complication to this process ar e presented. New targets to maximize the inhibition of tumor angiogenesis are pursued ( Loges et al., 2010). Various groups in a variety of tumor settings are currently investigating the Angiopoietin/Tie2 axis targeting agents (Gerald et al., 2013). Chapters 4 and 5 focused on the anti-angiogenic effect of the Ang-2 inhibitor on early angiogenesis and tumor growth. The present study evaluated the anti-tumor effect of the i nhibitor on both the VHL normal and mutated renal cell carcinoma cell lines. Treatment of tumors began once they have established a substantial size and vasculature to mimi c the status of tumors upon a patients diagnosis with cancer. The results show that the Ang-2 inhibito r has a minor effect on the growth of the Caki-1 tumors, a 1-3 day growth delay comp ared to control (Figure 6-1A,B) however no effect was seen in the growth of the Caki-2 tumors compared to control (Figure 6-1C,D). These results were puzzling especially because Leow and colleagues (2012) have 111

PAGE 112

conducted several experiments using this antibody and have seen successful growth delay effects. However when closely examining the xenograft models they have used and their growth characteristics it is evident that the tumor models used in this study grew much faster allowing for less dosi ng of the inhibitor before the study was terminated. For example, colon cancer LoVo xe nografts grew 2.9-fo ld slower and were dosed 2-3 times more than the Caki-1 tumors used in this study. Renal cell carcinoma 786-0 xenografts grew 2.7-fold slower, colon Colo205 xenografts 1.8-fold slower each allowing for 2-3 more dosings of the inhibito r compared to Caki-1 xenografts. The most aggressive model, HeyA8 ovarian cell line grew 1.3-fold slower allowing for 1-2 more doses of the inhibito r (Leow et al., 2012). The comparison between Caki-2 xenografts used in this model compared to the ones used by Leow and colleagues (2012) were even more drastic. Caki-2 xenografts grew 5.3-, 5-, 2.5-, and 4-fo ld faster than LoVo, 786-0, HeyA8 and Colo205 xenografts respectively reducing the number of doses by 2-4 throughout treatment (Leow et al., 2012). Furthermore, the Caki-2 xenografts we re dosed with 2-5 times higher doses of the inhibitor than the ones reported by Leow and colleagues (2012). These results suggest that both the Caki-1 and especially the Caki-2 tumors grow more aggressively than either xenograft treated with this inhi bitor before. Reports using other Ang-2 targeted agents also used comparable xeno graft models that Leow and colleagues (2012) have used and there are currently no reports with Ang2 targeted agents in either the Caki-1 or Caki-2 xenograft models (Brown et al., 2010; Daly et al., 2013; Leow et al., 2012; Neal and Wakelee, 2010). Due to the aggressive and hi ghly angiogenic profile of both these tumor models we believe that once the tumor vasculature is established in 112

PAGE 113

these tumors there is limited need for ac tivation of the Ang-2 axis. Once a tumor establishes its vasculature it is highly irregular with minima l peri-endothelial cell coverage and high prolif eration rate of the endothelial ce lls (Carmeliet and Jain, 2000; Siemann, 2011). The tumor most likely relie s on pro-angiogenic factors such as VEGF and other pathways to stimul ate endothelial cell prolifer ation and tube formation (Carmeliet, 2005; Carmeliet and Jain, 2000). Most likel y there is activity at the border of the tumor where normal pre-existing vascu lature is destabilized and stimulated, however it is unlikely to be the primary s ource of new blood vesse ls for the tumor. To evaluate whether Ang-2 inhibition is more important in the early phase of tumor development another growth delay expe riment was conducted. Treatment of mice began the day of tumor cell injection and thus the tumor was not given the opportunity to establish its vasculature. Results however were disappointing and there was no significant difference between the control and treated groups (Figur e 6-2). There are several possible explanations such as the use of lower doses of the inhibitor 2 and 10 mg/kg rather than the 20 and 50 mg/kg used in Figure 6-1C,D. Also the tumor reached the endpoint size of 1000 mm3 within 11-12 days allowing only 4 dosings. Perhaps treatment with higher dos es or injection of 10-fold less tumor cells allowing longer time before study endpoint would allow more dosings throughout the treatment period. It is possible that the Ang-2 inhibitor has early phase effects on the tu mor growth however once the tumor escapes the Ang-2 axis it aggressively takes adv antage of the VEGF axis leading to rapid vascularization and thus growth of the tumo r. Perhaps a higher dose of the inhibitor should also be evaluated, however once the tumor escapes the Ang-2 axis there is probably minimal effect s it may have on the dev eloping vasculature 113

PAGE 114

especially in the inner core of the tumor since peri-endothelial ce ll coverage is sparse and loose. Finally, Ang-2 inhibition was evaluated in a metastatic setting. Metastasizing tumor cells need to develop the same way that primary tumors do a nd therefore rely on angiogenesis to expand in size (Weidner et al., 1991). Since our results from the intradermal assay in Chapter 4 showed signif icant impairment of Ca ki-2 tumor growth and vascular development, here t he impairment of lung colony formation was evaluated. Mice were injected with Caki-2 cells via tail vein and treated with the Ang-2 inhibitor during a 3 week period (Figure 6-3A). Results did not show a significant difference in the number of metastatic lesions (Figure 6-3B), size (Figure 6-3C) no r the range of size (Figure 6-3D) between treated and untreated mice. Although this was surprising at first there is a possibility that these tumors co lonize the lung and begin growth at an early time point since by the 21 days most lesions ar e fairly large. There is a possibility that the Ang-2 inhibitor may have an effect very early on but once the tumor escapes the Ang-2 axis it aggressively takes over the VEGF pathway especially in the Caki-2 model since VHL is mutated and the cells produce hi gh levels of VEGF even in the absence of hypoxia. An earlier time point may be eval uated to test this hypothesis. Another possibility may be that once Caki-2 cells in filtrate the lungs they do not rely on angiogenesis for obtaining vasculature but rather co-opt vessels in the heavily vascularized lungs (Aird, 2007). In this case neither Ang-2 nor VEGF pathway inhibiting agents would have an effect on lung colo nization of Caki-2 tumor cells. In conclusion, these studies showed mini mal to no anti-tumor effect when treated with the Ang-2 inhibitor. A possible explanati on for this may be the aggressive nature of 114

PAGE 115

the tumor models chosen to evaluate since others have seen significant tumor growth delay in other less aggressive xenograft mode ls. Another explanatio n may be that the Ang-2 axis and its inhibition may only be relevant very early on during tumor development when the tumor cells stimulate the nearby normal vasculature. Periendothelial cells cover normal vessels and therefor e the Ang-2 axis is necessary to first destabilize the vasculature bef ore the tumor cells can activa te the endothelium to form new vessels (Augustin et al., 2009). Therefor e once the tumor cells have established vasculature from normal pre-existing vessels they can stimulate the formation of new vessels from the highly proliferative and abn ormal tumor vasculature leaving the Ang-2 axis less necessary. 115

PAGE 116

Figure 6-1. Ang-2 inhibition does not affect renal cell carcinom a tumor growth. Mice were injected with 106 human renal cell carcinoma Caki-1 (A,B) or Caki-2 (C,D) cells intramuscularly. Tumors were treated with MEDI3617 beginning when tumor volume reached 200 mm3 and mice were euthanized once tumor reached 1000 mm3. Ang-2 inhibition had no effe ct on tumor growth delay. (A,C) Median responders. (C,D) Median. (A,B) 0, 10, 20 mg/k g (n=6), 2 mg/kg (n=6). (C,D) 0, 20, 50 mg/kg (n=10). *, p<0.05; Mann-Whitney U-Test. 116

PAGE 117

Figure 6-2. Early treatment with Ang-2 inhibitor does not a ffect renal cell carcinoma tumor growth. Mice were injected with 106 human renal cell carcinoma Caki-2 cells intramuscularly. Treatment of tumors began the day of tumor cell injection and continued until tumor volume reached 1000 mm3. Ang-2 inhibition had no effect on tumor grow th delay. (A) Medi an responders. (B) Median (n=10). 117

PAGE 118

Figure 6-3. Ang-2 inhibition does not affe ct renal cell carcinoma lung metastasis growth. (A) Mice were injected with human renal cell carcinoma, Caki-2, cells at 104 via tail vein. Treatment with Ang-2 antibody began the day of tumor cell injection. (B) Ang-2 inhibition did not affect the number of lung metastases (C) the size of individ ual metastases (D) or the frequency of metastases size. (B,C) Line, median (D) mean with SEM 0, 10 mg/kg (n=7), 2 mg/kg (n=5). 118

PAGE 119

Part B: Combination Therapy With VEGF Targeted Agents Background Angiogenesis is a two-step process (a) the normal vasculature is destabilized loosening the endothelia l and peri-endothelial cell contacts in the vasculature at which point (b) pro-angiogenic factors such as VEGF activate the endothelium to proliferate and form new vessels. Currently the majority of FDA approved anti-angiogenic agents target the VEGF pathway and so me have other targets as well such as PDGF, cMET and mTOR. In general anti-angiogenic agents have been complementary to conventional cancer therapies and are used in a variety of tumor settings, however, there are some patients who do not re spond or stop responding after prolonged treatment (Bergers and Hanahan, 2008; Ebos et al., 2009; Loges et al., 2010). To circumvent this issue a new class of ant i-angiogenic agents the A ng-2 targeted agents have been pursued both in the pr eclinical and clinical setting. There is some evidence that the combination of A ng-2 and VEGF targeted agents yiel d superior anti-angiogenic response than either agent alone (Brown et al., 2010; Dandu et al., 2008; Hashizume et al., 2010; Koh et al., 2010). Currently, the combination of the VEGF antibody Bevacizumab and the Ang-1/2 peptibody Trebananib and small molecule VEGFR inhibitor Axitinib with the Ang-2 specific CVX-060 are being evaluated in the clinic (Gerald et al., 2013). There is also an antibody in preclini cal development that targets both the Ang-2 and VEGF ligands (Gerald et al., 2013). The current study evaluated the combinat ion of the Ang-2 inhibitor, MEDI3617, with two small molecule tyrosine kinase inhibitors Sunitinib and Cediranib. Sunitinib is a multikinase inhibitor that is FDA approved as first line treatment for kidney cancer 119

PAGE 120

(Goodman et al., 2007). Cediranib is a VEGFR specific small molecule inhibitor that is currently in clinical development (Zhu et al., 2013). The effects on early tumor development and angiogenesis initiation were evaluated. Materials And Methods Reagents MECA-32 was purchased from BioLegend (S an Diego, CA), AlexaFluor 594 was purchased from Invitrogen (G rand Island, NY). VectaShi eld mounting medium with DAPI was purchased from Ve ctor Labs Inc. (Burlingam e, CA). Tissue-Tek OCT Compound was purchased from Sakura Finete k (Torrance, CA). 2-methylbutane was obtained from Thermo Fisher Scientific (Waltham, MA). Cell culture The human clear cell renal cell carcinoma Caki-2 cell line was received as a gift from Dr. Susan Knox (Stanford University); t he cells were grown in Dulbeccos modified minimum essential medium (D-MEM, In vitrogen) supplemented with 10% FBS (Invitrogen), 1% penicillin-streptom ycin (Invitrogen), and 1% 200-mmol/L L-glutamine (Invitrogen). Cells were kept at 37 C, 5% CO2. The cells were used between passages 2 and 10. Drug preparation Angiopoietin-2 inhibitor, MEDI3617, wa s kindly provided by MedImmune, LLC. Stock solution (5 mg/ml) was diluted to working concentrations in sodium citrate buffer solution. Stock solutions were kept at -80 C and working concentrations at 4 C. Cediranib was kindly provided by AstraZeneca (Wilmington, DE) and stored at 4 C. Working concentration of Cediranib was pr epared fresh daily in 10% volume Tween 80 and 1M HEPES and stored at 4 C. Sunitini b was obtained from LC Laboratories 120

PAGE 121

(Woburn, MA) and stored at -20 C. Worki ng concentration of Sunitinib was prepared fresh daily in stock and diluent buffers of citric acid monohydrate and sodium citrate dihydrate at pH 6.8 and 3.2 respectively at 1:7 stock to diluent solution (~pH 3.3) and acidified to pH 1.0, Sunitinib was disso lved, and the solution adjusted to pH 3.5. Working concentration of Sunitinib was kept at room temperature. Intradermal assay All in vivo procedures were conducted in agr eement with a protocol approved by the University of Florida Institutional An imal Care and Use Committee. Female athymic nu/nu mice were injected intradermally with 105 Caki-2 cells in 10 l volume at four sites on the ventral surface. Beginni ng the day prior to tumor cell injection, mice were treated (1) with either daily oral gavage of Sunitinib ( 10 mg/kg) (7 doses total), IP injection of MEDI3617 (2 mg/kg) every 3 days (3 doses total) or both; or (2) wit h either daily oral gavage of Cediranib (2 mg/kg) (7 doses total), IP injection of MEDI3617 (2 mg/kg) every 3 days (3 doses total) or both up to six days post tumor cell inoculation. Mice were then euthanized, tumors measured via calipers and tumor volume (mm3) calculated, assuming the tumor volume to be an ellips oid, using the following equation: tumor volume = /6 x d1 x d2 x height. Skin flaps were then re moved and vessels growing into tumor nodules were counted using a Leica MZ16F dissecting microscope with Leica KL 1500 LCD fiber optic illuminator (Leica Micro systems Inc., Buffalo Grove, IL) at 2.5x original magnification (1,2,3 ). Images were captured with a Retiga EXi Fast1394 digital CCD camera (QImaging, British Columbia, Canada) and OpenLab5 software (PerkinElmer Inc., Waltham, MA). Statistical significance between control and treated groups was determined using the M ann-Whitney U-Test at p<0.05. 121

PAGE 122

Immunohistochemistry Intradermal tumors were fresh frozen in OCT and methylbutane and cryosectioned at 5 m thickness using Leica CM 3050S cryostat (Leica Microsystems Inc., Buffalo Grove, IL); sect ions were placed on superfros t plus gold slides (Thermo Fisher Scientific Inc., Wa ltham, MA) and kept at -80 C until immunohistochemical staining. Tissue sections were acetone fi xed for 10 minutes, blocked in 2% normal horse serum in 1x TBS, and incubated overnight at 4 C with MECA-32. Secondary antibody AlexaFluor 594 was added onto slides for 1 hr. Ti ssue sections were imaged with a Zeiss Axioplan 2 imaging microscope (Carl Zeiss, Inc., Thornwood, NY) with EXFO X-Cite 120 light source (Lumen Dynamics Group Inc., Ontari o, Canada). Images were taken with a Retiga EXi Fast digital CCD camera (Qimaging, British Columbia, Canada) and processed in OpenLab5 software (PerkinElmer Inc., Waltham, MA); Rhodamine for MECA-32/AlexaFluor594 and D API filters were used. Vessel counts were obtained by taking up to ten random fiel ds/tumor with 20x obje ctive, counting the number of vessels in each field. Statisti cal significance between control and treated groups was determined using the M ann-Whitney U-Test at p<0.05. Results MEDI3617 and Sunitinib combination leads to greater antiangiogenic effect Sunitinib is a multikinase small molecule in hibitor that is currently used as a first line treatment in renal cell ca rcinoma. Caki-2 renal cell ca rcinoma cells were injected intradermally and mice were dosed with the Ang2 inhibitor alone, Sunitinib alone or the combination of the two agents (Figure 6-5A). Results show t hat both agents alone led to reduction of tumor volume a 1.5fold (p=0.06) reducti on with MEDI3617 and 4.6-fold 122

PAGE 123

(p<0.0001) reduction with Sunitinib treatment compared to control, however, the combination of the two agents led to a much more significant reduction of 15.8 fold (p<0.0001) compared to control. The combination of the two agents was also significantly lower than either agent alone with a 10.7(p<0.0001) and 3.4-fold (p<0.0001) decrease compared to MEDI3617 and Sunitinib treatment alone respectively (Figure 6-5B). Evaluating the effect s on the tumor cell induced vasculature results show a 1.7(p<0.0001) and 1.8-fold (p<0.0001) decrease in the number of vessel in MEDI3617 and Sunitinib treated groups compared to control. The combination of the two agents led to a much more significant reduction in vessel number by 3.5-fold (p <0.0001) compared to control. The combination treatment was also significantly better at reducing tumor cell induced blood vessels by 2(p<0.0001) and 1.9-fold (p<0.0001) compared to MEDI3617 and Sunitinib alone respectively (Figure 6-5C). Figure 6.6 demonstrates these results using immunohistochemistry, ev aluating the tumor core rather than the vessels visible around the tumor peripher y as shown in Figure 6-5C. MEDI3617 treatment led to a 2.1(p<0.0001) and Suniti nib a 1.7-fold (p<0.0001) decrease in tumor vasculature compared to contro l. The combination treatment led to a 4.1-fold (p<0.0001) reduction in vessel number compared to c ontrol and a 1.9(p<0.01) and 2.3-fold (p<0.0001) reduction compared to ME DI3617 and Sunitinib treatments alone respectively. MEDI3617 and Cediranib combination lead to greater anti-angiogenic effect Cediranib is a small molecule inhibitor t hat is specific to VEGFR 1-3 unlike the multikinase small molecule inhibitor Sunitinib. Cediranib is currently in Phase II clinical trials in various disease settings. Caki-2 renal cell carcinoma cells were injected 123

PAGE 124

intradermally and mice were dosed with the A ng-2 inhibitor alone, Cediranib alone or the combination of the two agents (Figure 67A). Results show that both agents alone led to significant reduction of tumor volume 3.5fold (p<0.05) reduction with MEDI3617 treatment and a 3.8-fold (p<0 .05) reduction with Cediranib compared to control, however, the combination of the two agents led to a much more significant reduction of 54.6-fold (p<0.0001) compared to control. The combination of the two agents was also significantly lower than either agent alone with a 15.6(p<0.0001) and 14-fold (p<0.01) decrease compared to MEDI3617 and Cediranib treatments alone respectively (Figure 6-7B). Evaluating the effect s on the tumor cell induced vasculature results show a 1.6(p<0.01) and 1.5-fold (p<0.01) decreas e in the number of vessel in MEDI3617 and Cediranib treated groups compared to control. The combination of the two agents led to a much more significant reduction in vessel number by 2.7-fold (p <0.0001) compared to control. The combination treatment was also significantly better at reducing tumor cell induced blood vessels by 1.6(p<0.01) and 1.8-fold (p<0.0001) compared to MEDI3617 and Cediranib alone respectively (Figure 67C). Figure 6.8 demonstr ates these results using immunohistochemistry, evaluating the tu mor core rather than the vessels visible around the tumor periphery as shown in Figur e 6-7C. MEDI3617 treatment led to a 2.5(p<0.01) and Cediranib a 1.4-fold (p<0.01) decrease in tumor vasculature compared to control. The combination treatment led to a 3.5-fold (p<0.01) reduction in vessel number compared to control and a 1.3and 2.4-fold (p<0.05) reduction comp ared to MEDI3617 and Cediranib treatments alone respectively. 124

PAGE 125

Discussion Angiogenesis can be separated into two main events (a) the destabilization of normal vasculature, or the loosening of endothelial and peri-endothelial cell contacts, and (b) the activation of the endothelium to proliferate and form new vessels. The Angiopoietin/Tie2 axis is responsible for the first step in angio genesis or the vessel destabilization while pro-angiogen ic factors such as VEGF activate the endothelial cells. Currently FDA approved antiangiogenic agents target the VEGF pathway and have been shown to complement conventional t herapies such as chemotherapy (2013; Kerbel, 2006), however the issues of lack of patient response and tumor rebound due to acquired patient resistance has been a problem in the clinic (Bergers and Hanahan, 2008; Ebos et al., 2009; Loges et al., 2010). In this study the combination of VEGF targeted agents with the Ang-2 inhibitor, MEDI3617, were evaluated. Targeting two key pathways in angiogenesis should show complementary anti-angiogeni c effects. Two VEGF target ed agents were evaluated, the small molecule multi tyrosine kinase inhi bitor Sunitinib that is FDA approved for treatment of metastatic kidney cancer and t he VEGFR specific small molecule inhibitor Cediranib that is in clinical development fo r a variety of solid tumors (Figure 6-4). Results show that both in the Sunitinib (Fig ure 6-5) and Cediranib combination (Figure 6-6) treatments the combinati on of VEGF and Ang-2 targeting was significantly superior to either modality alone. Tr eatment of tumors with the Ang-2 inhibitor and Sunitinib led to a 1.5and 4.6-fold reduction in the tumor volume (Figure 6-5B) and 1.7and 1.8-fold (Figure 6-5C) reduction in tumor vasculature respectively. However, the combination of the two agents led to a 15.8-fold reduction in tumor volume (Figure 6-5B) and 3.5-fold reduction in tumor vasculature (Figure 6-5C). Similar results were seen with 125

PAGE 126

combination studies with Cediranib. Both the Ang-2 inhibitor and Cediranib led to a 3.5and 3.8-fold reduction in tumor volume (Figur e 6-6B) and 1.6and 1.5-fold reduction in tumor vasculature (Figure 6-6C) respective ly, however, the combination of the two agents led to a much greater anti-tumor and anti-angiogenic effect with a 54.6and 2.7fold reduction respectively (Figure 6-6B,C). In conclusion, the combined targeti ng of the Ang-2 and VEGF pathways showed much better anti-angiogenic effects than either therapeutic modality al one. Two different VEGF targeted agents were evaluated and show ed comparable effects in combination with the Ang-2 inhibitor supporting the pursuit of agents that tar get both pathways in angiogenesis. These results support the combination of the Ang-2 inhibitor, MEDI3617, especially with Sunitini b that is already used in the clinic. 126

PAGE 127

Figure 6-4. Targeting two key steps in angiogenesis. Vessel destabilization depends on the Ang-2/Tie2 axis which can be inhibi ted by the Ang-2 inhibitor, MEDI3617. Endothelial cell activation mediated significantly by the VEGF/VEGFR pathway can be inhibited by the multikinase inhibito r Sunitinib or the VEGFR2 selective small molecule inhibitor Cediranib. 127

PAGE 128

Figure 6-5. Ang-2 and VEGF combination t herapy with Sunitinib reduce tumor volume and vessel number. Mice were injected in tradermally with Caki-2 renal cell carcinoma cells and treated with MEDI3617 (2 mg/kg), Sunitini b (10 mg/kg) or both beginning the day prior to tumor ce ll inoculation. Tumor volume (A) and the number of tumor cell induced blood vessels (B) were determined at the end of a 7 day period. Bar, mean with SEM, control (0) and combination (M+S) (n=20), MEDI3617 (M) and Sunitinib (S) (n=16) ; line, median, control and combination (n=16), MEDI3617 and Sunitinib (n=12); ***, p<0.0001; Mann-Whitney U-Test. 128

PAGE 129

Figure 6-6. Ang-2 and VEGF combinati on therapy with Sunitinib impair tumor angiogenesis. The number of blood vesse ls within the tumor were evaluated using immunohistochemistry of fresh frozen tissue. Vasculature was stained with MECA-32. Line, medi an. Control (n=20), ME DI3617 (n=22), Sunitinib (n=16), combination (n=9). *, p< 0.05; **, p<0.01; ***, p<0.0001; MannWhitney U-Test. Representative images of the median of each group. Red, MECA-32. Images taken with Zeiss Axioplan Imaging2. 129

PAGE 130

Figure 6-7. Ang-2 and VEGF combination t herapy with Cediranib reduce tumor volume and vessel number. Mice were injected in tradermally with Caki-2 renal cell carcinoma cells and treated with MEDI3617 (2 mg/kg), Cedirani b (2 mg/kg) or both beginning the day prior to tumor ce ll inoculation. Tumor volume (A) and the number of tumor cell induced blood vessels (B) were determined at the end of a 7 day period. Bar, mean with SEM (n=16); line, median, (n=16); **, p<0.01; ***, p<0.0001; Mann-Whitney U-Test. 130

PAGE 131

Figure 6-8. Ang-2 and VEGF combinati on therapy with Cediranib impair tumor angiogenesis. The number of blood vesse ls within the tumor were evaluated using immunohistochemistry of fresh frozen tissue. Vasculature was stained with MECA-32. Line, median. (n=6). *, p<0.05; **, p<0.01; Mann-Whitney UTest. Representative images of the m edian of each group. Red, MECA-32. Images taken with Zeiss Axioplan Imagi ng2 microscope with 20x objective; scale bar = 140 m. 131

PAGE 132

CHAPTER 7 CONCLUSIONS AND FUTURE DIRECTIONS Summary The research presented here explored the therapeutic potent ial of an Ang-2 targeted antibody MEDI3617. At the time these studies began this antibody was in preclinical development. It has now moved into phase I clinical trials to be evaluated for patient tolerability. The re search presented here determined the effect of the Ang-2 inhibitor on the endothelial and smooth muscle ce ll contact destabilization, the first step in the angiogenic process. Studi es showed that the presence of Ang-2 leads to cell-cell contact loss, which can be inhibited by the Ang-2 inhibitor. Furthermore, the involvement of the tumor microenvironment in the Ang-2 axis was explored. This area remains controversial and results here sugges t that tumor cells do not significantly contribute Ang-2 to the tumor microenvir onment; however, tumor cells do secrete factors that influence the behavior of endothelial cells. Ther efore tumor cells have an indirect role in the initiation of angioge nesis through the Ang-2 pathway. The presence of the Ang-2 inhibitor impaired the tumor cell induced cell-cell de stabilization both in normal and hypoxic conditions. Hypoxia did not seem to be a major factor in Ang-2 release from endothelial cells neither did it affect the tumor cell secreted factors in regards to stimulation of Ang-2 secretion. The possibilit y of VEGF as the tumor cell secreted factor to influence endothelial cell s to release their Ang-2 contents was evaluated and results did not support such hypothesis. In vivo the Ang-2 inhibitor significantly impai red tumor volume and vessel number and promoted a normal vascular phenotype with peri-endothelial cell coverage of vessels in an early tumor development model. Ho wever, the Ang-2 inhi bitor did not elicit 132

PAGE 133

significant effects when tumors were allowe d to grow large and establish vasculature prior to treatment. No effects were seen in a metastatic model either. When evaluated in the murine dorsal skinfold wi ndow chamber model it became apparent that the model has limitations and that t he natural wound healing and pr o-inflammatory response through the Ang-2 axis skewed results w hen tumors were treated with the Ang-2 inhibitor. No such issues were seen w hen tumors were treated with a VEGF pathway inhibitor. Finally, results support the combin ation of Ang-2 with VEGF inhibition as been shown with two different VEGF pathway small molecule inhibitors. Current Status Of Angiopoietin-2/Tie2 Inhibitors In the last few years there has been tr emendous interest in targeting the Ang2/Tie2 axis in angiogenesis in hopes to ci rcumvent the lack of response or adaptive resistance that has been seen in the c linic to VEGF targeted agents (Bergers and Hanahan, 2008; Ebos et al., 2009; Loges et al., 2010). Currently there are two small molecule inhibitors that could be consider ed Tie2 inhibitors in clinical development. CEP-11981 (Cephalon, Inc) is a pan-VEGFR and Tie2 inhibitor, although it has an IC50 5-fold lower for VEGFR than Tie2 (Hudkin s et al., 2012); Phase I clinical trial (NCT00875264) is complete and t here are no other active trials at this time ARRY-614 (Array BioPharma, Inc) is another small molecu le inhibitor that ta rgets Tie2 and p38; Phase I trial in myelodysplastic syndromes is currently ongoing. ACTB-1003 (ACT Biotech/Bayer AG) is an FGFR/PI3K inhibi tor with comparable activity against Tie2, however there are no clinical tr ials with this agent at this moment (Burd et al., 2010). Several small molecule inhibitors are sometimes referred to as Tie2 inhibitors; however have very low activity toward this rec eptor. For example, R egorafenib (BAY 73-4506; Bayer AG) has 74-, 24-, 6.7and 14-fold higher biochemical activity to VEGFR2, 1, 3 133

PAGE 134

and PDGFR respectively than to Tie2 and it hardly affects the phosphorylation status of Tie2 while significantly reduces those of VEGFR2 and PDGFR (Wilhelm et al., 2011). Carbozantinib (XL-184/BMS-907351; Exelixis/B MS) is another such small molecule inhibitor with 408and 7.9-fold greater acti vity to VEGFR2 and MET respectively than Tie2 (Zhang et al., 2010). Although MG CD-265 (MetylGene, Inc) and Foretinib (Exelixis/GSK) have smaller gaps in activi ty between VEGFR, MET and Tie2, these two compounds are being developed as VEGFR/MET inhibitors (Beaulieu et al., 2011). There are two peptide-based agents, CVX-060 (Pfizer) and Trebananib (AMG386; Amgen) that are in phase II and III c linical trials respectively. Trebananib is the furthest in clinical develop ment with over a dozen ongoing clinical trials in all phases and various disease settings in combinatio n with either chemotherapeutic and/or antiVEGF agents. Trebananib is an A ng-1/2 specific peptibody that has raised questions about this agent. Based on current knowledge about the basic biology of this axis, Ang1/Tie2 interactions support and maintain a normal vascular phenotype while Ang-2/Tie2 interactions lead to vascular destabilizati on and angiogenesis in the presence of proangiogenic factors. Genetic data support the notion that Ang1 is an agonist while Ang-2 is an antagonist of Tie2. The lack of Ang-1 or Tie2 correlates with the same vascular phenotype as over expression of Ang-2 (Ma isonpierre et al., 1997). Therefore, one would assume that in this axis Ang-2 should be the sole target. However, Trebananib has yet to show any adverse effects other than the normal side effects seen with antiangiogenic agents. Although there was an attempt to expl ain the benefit of targeting both Ang-1 and Ang-2 by Coxon and colleagues (Coxon et al ., 2010) the article fell short of the real question which is the possi ble adverse effects of this agent on the 134

PAGE 135

normal vasculature. Recently, Thomas and colleagues (Thomas et al., 2013) demonstrated that Ang-1 inhibition can have adverse effects on normal vascular phenotype. A provocative study by Daly and colleagues (Daly et al., 2013) complicates this issue further. In their study with the Ang-2 antibody (REGN910) they have demonstrated that the addition of Ang-1 to tumor bearing mice in the absence and presence of the Ang-2 inhibitor led to tumo r growth comparable to control. Daly and colleagues interpreted this data suggesting t hat Ang-2 is a Tie2 agonist compensating for lack of Ang-1 in the microenvironmen t. However, they fail to discuss the physiological relevance of Ang-1 overexpr ession and a study by Suri and colleagues (Suri et al., 1998) showed that overexpression of Ang-1 leads to hypervascularization in mice attributing to initiation of vascular branching in the presenc e of VEGF, a process that is mediated by Ang-1/Tie2 in vascular development. Perhaps there is a fine balance of Ang-1 levels that are needed to be maintai ned in order to promote quiescent vascular phenotype and upsetting this balance could in itiate vascular branching ultimately leading to increased vasculature especia lly in the presence of VEGF, an abundant factor in the tumor microenv ironment. These results ther efore could have been skewed by a non-physiological condition that would normally be seen. Clearly, the angiopoietin/Tie axis is co mplex and not fully understood. There are a few agents that are Ang-2 specific; the mo dified antibody mentioned earlier CVX-060 that is currently in Phase I/II trial in combination with Sunitinib. A Phase II trial in combination with Axitinib wa s terminated due to adverse toxicities; new strategies for dosing and scheduling are being evaluated. Fully human monoclonal antibodies specific to Ang-2, ME DI3617 (MedImmune, LLC/AstraZeneca) and 135

PAGE 136

REGN910 (Sanofi/Regeneron) are both in phase I clinical trials (Table 1) (Daly et al., 2013; Leow et al., 2012). Currently all t he Ang/Tie2 pathway inhibitors are being evaluated in combination with chemother apeutic agents and/or anti-VEGF targeted agents, such as Bevacizumab, Sorafenib, S unitinib, that have ov er the years become part of the standard of ca re While there is merit to th ese combinations there are also additional possibilities that have yet to be expl ored preclinically or c linically with these agents. Such considerations are the focu s of the rest of this discussion. Combination With Radiation Therapy Conventional cancer therapies such as chemotherapy and radiation therapy target rapidly proliferating cells by i nducing DNA damage. Radiation therapy uses ionizing radiation that creates reactive oxygen species and causes DNA damage resulting in tumor cell death (Bristow and Hill, 2005; Giaccia and Hall, 2006a; Giaccia and Hall, 2006b). Curative radiotherapy is used either as a single modality, before or after surgical resection of tumors. It can have detrimental effects on normal tissues and the therapeutic outcome can be greatly a ffected by the tumor microenvironment (Bristow and Hill, 2008). Hypoxic cells are resistant to radiation and pose a real problem in the treatment of cancer (Figure 7-1A ) (Bristow and Hill, 2008; Giaccia and Hall, 2006b). Efforts to improve oxygenation in the tumor have s hown some improvement in treatment outcome (Chapman and Whitmore 1984; Coleman, 1985; Giaccia and Hall, 2006b; Overgaard and Horsman, 1996). Targeting the tumor vasculature offers a different approach and combinations with r adiation therapy have shown that these treatments may be complement ary (Figure 7-1A) (Dings et al., 2007; Horsman and Siemann, 2006; Jain et al., 2006; Siemann, 2006; Siemann, 2011) (Figure 7-1A). 136

PAGE 137

There is some evidence that radiation of tumors leads to more aggressive phenotypes of cells that remain. One such reason for aggressive phenotype is the upregulation of pro-angiogenic factors in tumor cells thus t he combination of radiation with anti-angiogenic therapy seems logical (Kargiotis et al., 2010). In general, the combination of radiation and anti-angiogenic therapies result ed in superior anti-tumor effects than either agent al one (Horsman and Siemann, 2006). However, it is now evident that timing of such combinations for maximal effect is important (Senan and Smit, 2007). Studies have suggested that tr eatment of tumors with anti-angiogenic agents may lead to a window of vascular norma lization where the oxygenation of the tumor is improved (Jain, 2005b). Some spec ulate that the normalization window is responsible for the benefits seen in comb ination with chemother apeutics and radiation treatments (Horsman and Siem ann, 2006). Dings and colle agues (Dings et al., 2007) demonstrated that tumors experienced im proved oxygenation post anti-angiogenic treatment and that it was crucial to irradiat e tumors within this window for effective cell kill. Geng and colleagues (Geng et al., 2001) showed that the combination of radiation and anti-angiogenic agent at minimally effe ctive low-dose resulted in the same damage to the tumor as either agent alone at higher doses, however the combination significantly reduced the normal tissue toxiciti es. Benefit of using anti-angiogenic agents as a maintenance therapy post radiation has also been shown (Schueneman et al., 2003). Currently there are 41 open/ ongoing trials combining anti-angiogenic agents with radiation alone or radiation and chemotherapy in several tumor settings such as brain metastasis, gliomas, glioblastomas, head and neck cancer, child glioma, colorectal 137

PAGE 138

cancer, nasopharyngeal carcinoma, child CNS, rectal, non small cell lung cancer, colorectal liver metastasis soft tissue sarcoma, hepatocellular carcinoma, prostate cancer, pelvic cancer. There are no reported preclinical or c linical data combining Ang-2 targeting agents with radiation therapy even though there is convincing evidence that the combination of a vascular targeted agen t and radiation therapy complement each other. In the future, Ang-2 targeted agent s should be pursued in this setting opening a new window of opportunities for this class of agents. Combination With Vascular Disrupting Agents Vascular disrupting agents (VDAs) are a class of vascular targeted agents that work distinctively different from antiangiogenic agents. While anti-angiogenic agents prevent the formation of new vasculature VD As destroy the existing tumor vasculature. VDAs exploit the different characteristics of normal and tumor endothelial cells and specifically target those a ssociated with the tumor (Fi gure 7-1B). Normal endothelial cells are dormant unless physiological angio genesis is required while tumor endothelial cells are continuously dividing, trying to keep up with the vascular demands of the growing tumor. VDAs can be separated into two classes (a) the flavanoids that induce hemorrhagic necrosis and (b) the tubulin depoly merizing/binding agents that disrupt the cytoskeleton of proliferating endothelial cells ultimately leading to vessel occlusion and necrosis (Siemann et al., 2005). Preclinical evaluations show convincing data that the combination of VDAs and anti-angiogenic agents have complementary effects on tumor growth (Chen et al., 2012; Shaked et al., 2010; Siemann and Shi, 2004; Siemann and Shi, 2008). The different mechanisms of action of VDAs and anti-angiogenic agents allows for the destruction of existing tumor vasculature by the VDA and inhi bition of vascular recovery normally seen 138

PAGE 139

after VDA treatment by the anti-angiog enic agent (Figure 7-1B) (Siemann, 2006; Siemann et al., 2004). Even though VDAs ar e effective at destroying the tumor vasculature and have been shown to complem ent conventional therapies such as chemotherapy and radiation therapy they ar e not widely sought after in the clinic. Several clinical trials have been terminated due to toxicity issues. Recently two trials (Phase I/II) were conducted with Combretastatin and Bevacizumab with or without chemotherapy (Patterson et al., 2012). No other trials with VDAs are currently ongoing. To date there are no reported studies combining VDAs and Ang-2 inhibition, however one study alludes to the possible benefits of this comb ination. Welford and colleagues (Welford et al., 2011) demonstrated that Tie2 expressing macrophages can hinder Combretastatin effects and depletion of these macrophages by CXCR4 inhibitor AMD3100 led to improved effects. Targeti ng Ang-2 has also been shown to decrease the amount of Tie2 expressing macrophages in the tumor microenvironment by hindering their recruitment thro ugh Ang-2 (Mazzieri et al., 2011). Other Therapeutic Considerations Ang-2 targeting as an anti-angiogenic ag ent could be beneficial in limiting the development and growth of distant metastases. Furthermore, Ang-2 targeting could be beneficial in targeting tumor associated macrophages (TAMs) especially the ones expressing Tie2 (TEMs). TEMs are recruited to the tu mor microenvironment in the presence of Ang-2 and have been shown to aid tumor angiogenesis (De Palma et al., 2007; De Palma and Naldini, 2011). TAMs in general are known to aid tumor cells as well (De Palma et al., 2007; Solinas et al., 2009). Therefore this therapeutic modality could be useful not only to tr eat the microvasculature of the primary tumor but also eliminate tumor associated macrophages and impair the development of metastases. 139

PAGE 140

Therapeutic Limitations Of Ang-2 Targeted Agents There are possible limitations of Ang-2 targeted agents. For ex ample, there are still some questions regarding the Ang-/Tie axis that remains to be answered. There is limited knowledge on the role of Tie1, however some studies speculate that a Tie1/2 heterodimeric interaction leads to the destabiliz ing effect of Ang-2 ligand (Seegar et al., 2010). Furthermore, the importance of Ang1 in pathological conditions is still controversial. Based on the basic biology one would think that the presence of Ang-1 is important to establish a normal vascular ph enotype and inhibition of this ligand would lead to impairment in the normal vasculature However in the case of the Ang-1/2 peptibody, Trebananib, no overt toxicities have been noticed (Doi et al., 2013; Mita et al., 2010; Neal and Wakelee, 2010; Oliner et al., 2012; Rini et al ., ; Rini et al., 2012; Robson and Ghatage, 2011). Also a study by Jeansson and colleagues (Jeansson et al., 2011) showed that the presence of Ang-1 is not crucial in maintenance of normal quiescent vasculature, however is essentia l for vascular repair upon injury. It has also been shown that Ang-2 can interact with int egrins and aid in endothelial cell migration (Carlson et al., 2001; Hu et al ., 2006; Imanishi et al., 2007); however currently we do not know whether Ang-2 binds integrins on the same site as it does Tie2 receptor and whether the currently available Ang-2 axis inhibitors would also impair the interaction of Ang-2 and integrins and how this would either positively or negativ ely affect the antiangiogenic and anti-tumor effect s of Ang-2 inhibitors. In general, Ang-2 targeting will be beneficia l in inhibition of angiogenesis that occurs from the normal surrounding vascula ture. Since tumor microvasculature is disorganized and lack sufficient peri-endothe lial cell coverage the inhibition of an endothelial-peri-endothelial cell destabilization agent would be out of context. However, 140

PAGE 141

the combination of a VEGF and Ang-2 targeted agent would capitalize on two major pathways in this process and could be more superior. Overall Conclusions Vascular targeted therapies seldom have c linical efficacy alone. However, these agents can enhance benefits from conventional therapies. Combination therapies exploit targeting tumors in various ways and usually require lower dose of each agent limiting normal tissue toxicities, especially when toxicities do not overlap. There is merit to anti-angiogenic combination therapies with radiation ther apy or vascular disrupting agents. However, a killer combination wo uld be the combination of a conventional therapy e.g. chemotherapeutics or radiation that would tar get the rapidly proliferating tumor cells, a VDA that would destroy t he existing tumor vasculature and lead to necrosis of endothelial and tumor cells and an anti-angiogenic agent that would prevent the recovery of the tumor vasculature. Bas ed on current data the co mbination of Ang-2 and VEGF targeted modalities would be more effective anti-angiogenic agent than either alone. Furthermore, Ang-2 tar geting would add an extr a layer of tumor microenvironment effects by inhibiting the recruitment of tumor associated macrophages that have been shown to aid tumor and vascular growth. A combination modality of this magnitude would require rigorous preclinical and clinical evaluation for optimal dosing and timing of each agent to reach maximal tumor cell kill. This therapeutic modality would target major pathwa ys involved in all solid tumor growth and development including metastasis and thus should be considered in the future for cancer treatment. 141

PAGE 142

142 Figure 7-1. Radiation and VDA combinatio n modalities. (A) Anti-angiogenic and radiation therapies combi ned. i) Areas of hypoxia within the tumor are common and cells at such areas are re sistant to radiat ion and therefore survive while cells under normal oxygen conditions are killed. Surviving cells can lead to tumor relapse. ii) The treat ment of tumors with anti-angiogenic agents could lead to a window of vascu lar normalization thus reducing the areas of hypoxia within the tumor. De creased levels of hypoxia within the tumor lead to increased cell ki ll upon irradiation. (B) Anti-angiogenic and vascular disrupting agents combined. Treatment of tumors with vascular disrupting agents leads to the destruction of tumor vasculature and necrosis. i) A viable rim of tumor cells and vessels usually remain after treatment and angiogenesis is induced leadi ng to regrowth of the tumor and its vasculature. ii) The treatment of tumors with anti -angiogenic agents after treatment with vascular disrupting agents inhibits the regrowth of the tumor and its vasculature.

PAGE 143

LIST OF REFERENCES Ahmad, S. A., et al., 2001. The effects of angiopoietin-1 and -2 on tumor growth and angiogenesis in human colon cancer. Cancer Res. 61 1255-1259. Aird, W. C., 2007. Phenotypic heterogeneity of the endothelium: II. Representative vascular beds. Circ Res. 100 174-190. Anargyrou, K., et al., 2008. Normalization of the serum angiopoietin-1 to angiopoietin-2 ratio reflects response in refractory/resistant multiple myelom a patients treated with bortezomib. Haematologica. 93 451-454. Armulik, A., et al., 2005. Endothelial/ pericyte interactions. Circ Res. 97 512-523. Auerbach, R., Models of Angiogenesis. Angiogenesis: an integrative approach from science to medicine. In: J. Folkman, W. D. Figg, Eds. Angiogenesis: an integrative approach from science to m edicine. Springer Science + Business Media, LLC New York, NY, 2008. Augustin, H. G., et al., 2009. Control of vascular morphogenesis and homeostasis through the angiopoietin-Tie system Nat Rev Mol Cell Biol. 10 165-177. Bach, F., et al., 2007. Angiopoietins in malignancy. Eur J Surg Oncol. 33 7-15. Barton, W. A., et al., 2005. Structure of the angiopoietin-2 receptor binding domain and identification of surfaces involved in Tie2 recognition. Structure. 13 825-832. Beaulieu, N., et al., 2011. MGCD265, an Orally Active Me t/VEGFR Multitargeted Kinase Inhibitor in Phase II Clinical Development, Potently Inhibits Clinically Relevant Met Mutants. Bergers, G., Benjamin, L. E., 2003. Tumo rigenesis and the angiogenic switch. Nat Rev Cancer. 3 401-410. Bergers, G., Hanahan, D., 2008. Modes of resistance to anti-angiogenic therapy. Nat Rev Cancer. 8 592-603. Betof, A. S., Dewhirst, M. W., Establishi ng the tumor microenvironment. In: D. W. Siemann, Ed. Tumor Microenvironment. Wi ley-Blackwell Ltd, West Sussex, United Kingdom, 2011. Bhaskar, A., et al., 2013. Synergistic effe ct of vascular endothelial growth factor and angiopoietin-2 on progression free survival in multiple myeloma. Leuk Res. 37 410-415. Biswas, S. K., et al., 2013. Tumor-associated macrophages: functional diversity, clinical significance, and open questions. Semi n Immunopathol DOI. 10.1007/s00281013-0367-7. 143

PAGE 144

Borgstrom, P., et al., 1996. Complete inhibition of angiogenesis and growth of microtumors by anti-vascular endothelia l growth factor neutralizing antibody: novel concepts of angiostatic therapy fr om intravital videomicroscopy. Cancer Res. 56 4032-4039. Borgstrom, P., et al., 1998. Neutralizing anti-vascular endothelial growth factor antibody completely inhibits angiogenesis and grow th of human prostate carcinoma micro tumors in vivo. Prostate. 35 1-10. Borgstrom, P., et al., 1999. Importance of VEG F for breast cancer angiogenesis in vivo: implications from intravital microsc opy of combination treatments with an antiVEGF neutralizing monoclonal antibody and doxorubicin. Anticancer Res. 19 4203-4214. Bottsford-Miller, J. N., et al., 2012. Re sistance and escape from antiangiogenesis therapy: clinical implications and fu ture strategies. J Clin Oncol. 30 4026-4034. Bristow, R. G., Hill, R. P., Molecular and Ce llular Basis of Radiotherapy. In: H. R. Tannock IF, Bristow RG, and Harrington L, Ed. The Basic Science of Oncology. McGraw-Hill Medical Publishing Division, 2005. Bristow, R. G., Hill, R. P., 2008. Hypoxia and metabolism. Hypoxia, DNA repair and genetic instability. Nat Rev Cancer. 8 180-192. Brown, J. L., et al., 2010. A human monoclonal anti-ANG2 antibody leads to broad antitumor activity in combination wit h VEGF inhibitors and chemotherapy agents in preclinical models. Mol Cancer Ther. 9 145-156. Burd, A., et al., 2010. ACTB-1 003 a unique oral pan FGFR and PI3K pathway inhibitor with divergent modes of acti on. Molecular targeted ther apies preclinical Poster 141. Carlson, T. R., et al., 2001. Direct cell adhesion to the angiop oietins mediated by integrins. J Biol Chem. 276 26516-26525. Carmeliet, P., Jain, R. K., 2000. Angiogenesis in cancer and other diseases. Nature. 407 249-257. Carmeliet, P., 2005. Angiogenesis in life, disease and medicine. Nature. 438 932-936. Chapman, J. D., Whitmore, G. F., 1984. Chemic al Modifiers of Canc er Treatment. Intl J Radiat Oncol Biol Phys 10 1161-1813. Charlesworth, P. J. S., Harris, A. L., Hypoxic Regulation of Angiogenesis by HIF-1. In: J. Folkman, W. D. Figg, Eds. Angiogenesis : An Integrative Approach from Science to Medicine. Springer Science + Busi ness Media, LLC, New York, NY, 2008. 144

PAGE 145

Chen, F., et al., 2012. Enhanced antitumor efficacy of a vascular disrupting agent combined with an antiangiogenic in a rat liver tumor model evaluated by multiparametric MRI. PLoS One. 7 e41140. Choe, S. W., et al., 2010. Intravital micr oscopy imaging of macrophage localization to immunogenic particles and co-localized ti ssue oxygen saturation. Acta Biomater. 6 3491-3498. Cirri, P., Chiarugi, P., 2011. Cancer associat ed fibroblasts: the dark side of the coin. Am J Cancer Res. 1 482-497. Cleaver, O., Melton, D. A., 2003. Endothelial signaling duri ng development. Nat Med. 9 661-668. Cole, S. P. C., Tannock, I. F., Drug Resistance. In: I. F. Ta nnock, et al., Eds.The Basic Science of Oncology. McGraw-Hill M edical Publishing Division, 2005. Coleman, C. N., 1985. Hypoxic Cell Radios ensitizers: Expectations and Progress in Drug Development. Intl J Radiat Oncol Biol Phys. 11 323-329. Coxon, A., et al., 2010. C ontext-dependent role of angiopoi etin-1 inhibition in the suppression of angiogenesis and tumor gr owth: implications for AMG 386, an angiopoietin-1/2-neutralizing pept ibody. Mol Cancer Ther. 9 2641-2651. Currie, M. J., et al., 2002. Expression of the angiopoietins and their receptor Tie2 in human renal clear cell carcinomas; regul ation by the von Hi ppel-Lindau gene and hypoxia. J Pathol. 198, 502-510. Curti, B., et al., Renal Cell Carcinoma. emedicine.medscape.com [acessed November 5, 2010]. D'Amico, G., et al., 2010. Loss of endothelia l Tie1 receptor impairs lymphatic vessel development-brief report. Arte rioscler Thromb Vasc Biol. 30 207-209. Daly, C., et al., 2013. Angiopoietin-2 functions as a Tie2 agonist in tumor models, where it limits the effects of VEGF inhibition. Cancer Res. 73 108-118. Dandu, R., et al., 2008. Design and synthesis of dihydroindazolo[5,4-a]pyrrolo[3,4c]carbazole oximes as potent dual inhibi tors of TIE-2 and VEGF-R2 receptor tyrosine kinases. Bioorganic & Medicinal Chemistry Letters. 18, 1916. Davis, S., et al., 1996. Isolation of angiopoi etin-1, a ligand for the TIE2 receptor, by secretion-trap expression cloning. Cell. 87, 1161-1169. Davis, S., et al., 2003. Angiopoietins have distinct modular domains essential for receptor binding, dimerization and superclustering. Nat Struct Biol. 10 38-44. 145

PAGE 146

David, S., et al., 2012. Angiopoietin-2 may c ontribute to multiple organ dysfunction and death in sepsis. Crit Care Med. 40 3034-3041. de Moor, J. S., et al., 2012. Cancer surviv ors in the United States: prevalence across the survivorship trajectory and implic ations for care. Cancer Epidemiol Biomarkers Prev. 22 561-570. De Palma, M., et al., 2007. Tie2-ex pressing monocytes: r egulation of tumor angiogenesis and therapeutic implications. Trends Immunol. 28 519-524. De Palma, M., Naldini, L., 2011. Angiopoietin-2 TI Es up macrophages in tumor angiogenesis. Clin Cancer Res. 17 5226-5232. Dedeugd, C., et al., 2009. Multimodal optical imaging of microvessel network convective oxygen transport dynamics. Appl Opt. 48 D187-D197. Dejana, E., 2004. Endothelial ce ll-cell junctions: happy together. Nat Rev Mol Cell Biol. 5 261-270. Denekamp, J., 1982. Endothelial cell proliferation as a novel approach to targeting tumour therapy. Br J Cancer. 45 136-139. Denekamp, J., 1984. Vascular endothelium as t he vulnerable element in tumours. Acta Radiol Oncol. 23 217-225. Detjen, K. M., et al., 2 010. Angiopoietin-2 promotes disease progression of neuroendocrine tumors. Clin Cancer Res. 16 420-429. Dewhirst, M. W., et al., 1996. Microvascular studies on the origins of perfusion-limited hypoxia. Br J C ancer Suppl. 27 S247-S251. Dewhirst, M. W., The cycle between angiogenesis, perfusion, and hypoxia in tumors. In: B. Teicher, Ed. Cancer Drug Disco very and Development: Cancer Drug Resistance. Humana Press Inc, Totowa, NJ, 2006. Dewhirst, M. W., et al., 2007. Exploring the role of HI F-1 in early angiogenesis and response to radiotherapy. Radiother Oncol. 83 249-255. Diaz-Sanchez, A., et al., 2013. Serum angiopoi etin-2 level as a predictor of tumor invasiveness in patients with hepatocellular carcinoma. Scand J Gastroenterol. 48, 334-343. Dings, R. P., et al., 2007. Scheduling of radi ation with angiogenesis inhibitors anginex and Avastin improves therapeutic outcome via vessel normalization. Clin Cancer Res. 13 3395-3402. 146

PAGE 147

Dixit, M., et al., 2008. Shear stress-induced activation of the AMP-activated protein kinase regulates FoxO1a and angiopoietin-2 in endothelial cells. Cardiovasc Res. 77, 160-168. Doi, T., et al., 2013. Phase 1 study of trebananib (AMG 386), an angiogenesis targeting angiopoietin-1/2 antagonist, in Japanese patients with advanced solid tumors. Cancer Chemother Pharmacol. 71 227-235. Ebos, J. M., et al., 2009. Tumor and host-me diated pathways of resistance and disease progression in response to antiangi ogenic therapy. Clin Cancer Res. 15 50205025. Elice, F., Rodeghiero, F., 2012. Side effect s of anti-angiogenic drug s. Thromb Res. 129 Suppl 1 S50-S53. Engin, H., et al., 2012. Plasma concentrati ons of angiopoietin-1, angiopoietin-2 and Tie2 in colon cancer. Eur Cytokine Netw. 23 68-71. Falcon, B. L., et al., 2009. Contrasting acti ons of selective inhibitors of angiopoietin-1 and angiopoietin-2 on the no rmalization of tumor blood vessels. Am J Pathol. 175 2159-2170. Ferrara, N., 2010. Pathways mediating VEG F-independent tumor angiogenesis Cyt Growth Factor Rev. 21 21-26. Fiedler, U., et al., 2003. Angiopoietin-1 and angiopoietin-2 share the same binding domains in the Tie-2 receptor involvi ng the first Ig-like loop and the epidermal growth factor-like repeats. J Biol Chem. 278, 1721-1727. Fiedler, U., et al., 2004. The Ti e-2 ligand angiopoietin-2 is st ored in and rapidly released upon stimulation from endot helial cell Weibel-Pal ade bodies. Blood. 103 41504156. Fiedler, U., et al., 2006. Angiopoietin-2 s ensitizes endothelial cells to TNF-alpha and has a crucial role in the induction of inflammation. Nat Med. 12 235-239. Fiedler, U., Augustin, H. G., 2006. Angi opoietins: a link between angiogenesis and inflammation. Trends Immunol. 27 552-558. Folkman, J., 1971. Tumor angiogenesis: ther apeutic implications. N Engl J Med. 285 1182-1186. Folkman, J., 2007. Angiogenesis: an organizing principle for drug discovery? Nat Rev Drug Discov. 6 273-286. Fukuhara, S., et al., 2008. Differential f unction of Tie2 at cellcell contacts and cellsubstratum contacts regulated by angiopoietin-1. Nat Cell Biol. 10 513-526. 147

PAGE 148

Gale, N. W., et al., 2002. Angiopoietin-2 is required for postnatal angiogenesis and lymphatic patterning, and only the latter ro le is rescued by Angiopoietin-1. Dev Cell. 3 411-423. Gallagher, D. C., et al., 2008. Circulating angiopoietin 2 correlates with mortality in a surgical population with acute lung injury /adult respiratory distress syndrome. Shock. 29 656-661. Gavard, J., et al., 2008. Angiopoietin-1 prevents VEGF-induced e ndothelial permeability by sequestering Src through mDia. Dev Cell. 14 25-36. Geng, L., et al., 2001. Inhibition of vascula r endothelial growth factor receptor signaling leads to reversal of tumor resistan ce to radiotherapy. Cancer Res. 61 24132419. Gerald, D., et al., 2013. Angiopoie tin-2: an attractive target for improved antiangiogenic tumor therapy. Cancer Res. 73 1649-1657. Giaccia, A. J., Hall, E. J., Cell Survival Cu rves. In: A. J. Giaccia, E. J. Hall, Eds. Radiobiology for the Radiologist. Lippincott Williams and Wilkins, Philadelphia PA, 2006a. Giaccia, A. J., Hall, E. J., Ox ygen Effect and Reoxygenation. In: A. J. Giaccia, E. J. Hall, Eds. Radiobiology for the Radiolog ist. Lippincott Williams and Wilkins, Philadelphia, PA, 2006b. Goede, V., et al., 2010. Identific ation of serum angiopoietin-2 as a biomarker for clinical outcome of colorectal cancer patients treated with bevacizumab-containing therapy. Br J Cancer. 103 1407-1414. Goldmann, E., 1908. The Growth of Malignant Disease in Man and the Lower Animals, with special reference to the Vascu lar System. Proc R Soc Med. 1 1-13. Goodman, V. L., et al., 2007. Approval summary: sunitinib for the treatment of imatinib refractory or intolerant gastrointesti nal stromal tumors and advanced renal cell carcinoma. Clin Cancer Res. 13 1367-1373. Hanahan, D., Folkman, J., 1996. Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell. 86 353-364. Hardee, M. E., et al., 2009. Novel imaging pr ovides new insights into mechanisms of oxygen transport in tumors. Curr Mol Med. 9 435-441. Hashizume, H., et al., 2010. Complementary actions of inhibitors of angiopoietin-2 and VEGF on tumor angiogenesis and growth. Cancer Res. 70 2213-2223. Hayman, S. R., et al., 2012. VEGF inhibition, hypertension, and renal toxicity. Curr Oncol Rep. 14, 285-294. 148

PAGE 149

Helfrich, I., et al., 2009. Angi opoietin-2 levels are associat ed with disease progression in metastatic malignant mel anoma. Clin Cancer Res. 15 1384-1392. Holash, J., et al., 1999. Vessel cooption, regr ession, and growth in tumors mediated by angiopoietins and VEGF. Science. 284, 1994-1998. Horsman, M. R., Siemann, D. W., 2006. Pathophy siological Effects of VascularTargeting Agents and the Implications for Combination with Conventional Therapies. Cancer Res 66 11520-11539. Hu, B., et al., 2006. Angiopoietin 2 induces g lioma cell invasion by stimulating matrix metalloprotease 2 expression through the alphavbeta1 integrin and focal adhesion kinase signaling pat hway. Cancer Res. 66, 775-783. Hu, B., Cheng, S. Y., 2009. Angiopoietin-2: development of inhibitors for cancer therapy. Curr Oncol Rep. 11 111-116. Huang, Y. Q., et al., 2002. Thrombin in duces increased expression and secretion of angiopoietin-2 from human umbilica l vein endothelial cells. Blood. 99 1646-1650. Hudkins, R. L., et al., 2012. Synthesis and biological profile of the pan-vascular endothelial growth factor receptor/t yrosine kinase with immunoglobulin and epidermal growth factor-like homology domai ns 2 (VEGF-R/TIE-2) inhibitor 11-(2methylpropyl)-12,13-dihydro-2-methyl-8-(pyrimidin-2-ylamino)-4H-indazolo[5, 4a]pyrrolo[3,4-c]carbazol-4-one (CEP-11981) : a novel oncology therapeutic agent. J Med Chem. 55, 903-913. Hurwitz, H., et al., 2004. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal c ancer. N Engl J Med. 350 2335-2342. Imanishi, Y., et al., 2007. Angiopoietin-2 stimulates breast cancer metastasis through the alpha(5)beta(1) integrin-mediated pathway. Cancer Res. 67 4254-4263. Jain, R. K., 2005a. Antiangiogenic therapy fo r cancer: current and emerging concepts. Oncology. 19 7-16. Jain, R. K., 2005b. Normalization of tumo r vasculature: an emerging concept in antiangiogenic ther apy. Science. 307 58-62. Jain, R. K., et al., 2006. Lessons from phase III clinical trials on anti-VEGF therapy for cancer. Nat Clin Pract Oncol. 3 24-40. Jang, C., et al., 2009. Angiopoietin-2 exo cytosis is stimulated by sphingosine-1phosphate in human blood and lymphatic endothel ial cells. Arterioscler Thromb Vasc Biol. 29 401-407. Jeansson, M., et al., 2011. Angiopoietin-1 is essential in mouse vasculature during development and in response to injury. J Clin Invest. 121 2278-2289. 149

PAGE 150

Kampfer, H., et al., 2001. Expressional regul ation of angiopoietin-1 and -2 and the tie-1 and -2 receptor tyrosine kinases during cutaneous wound healing: a comparative study of normal and impaired repair. Lab Invest. 81 361-373. Kargiotis, O., et al., 2010. Effects of irradiation on tumor cell survival, invasion and angiogenesis. J Neurooncol. 100 323-338. Kerbel, R. S., 2006. Antiangiogenic therapy: a universal chemosensiti zation strategy for cancer? Science. 312, 1171-1175. Kim, K. L., et al., 2006. Inte raction between Tie receptors modulates angiogenic activity of angiopoietin2 in endothelial pr ogenitor cells. Cardiovasc Res. 72 394-402. Koh, Y. J., et al., 2010. Double antiangio genic protein, DAAP, targeting VEGF-A and angiopoietins in tumor angiogenesis, me tastasis, and vascular leakage. Cancer Cell. 18 171-184. Kong, B., et al., 2010. Surgical procedure as an inducer of tumor angiogenesis. Exp Oncol. 32 186-189. Korff, T., et al., 2001. Blood vessel maturati on in a 3-dimensional spheroidal coculture model: direct contact with smooth muscl e cells regulates endothelial cell quiescence and abrogates VEGF responsiveness. Faseb J. 15 447-457. Krikun, G., et al., 2000. Expre ssion of angiopoietin-2 by human endometrial endothelial cells: regulation by hypoxia and inflammation. Biochem Biophys Res Commun. 275 159-163. Laschke, M. W., Menger, M. D., 2007. In vitro and in vivo approaches to study angiogenesis in the pathophysiology and therapy of endometriosis. Human Reprod Update 13 331-342. Leahy, K. M., et al., 2000. Role of cycl ooxygenases in angiogenesis. Curr Med Chem. 7 1163-1170. Leow, C. C., et al., 2012. MEDI3617, a human anti-angiopoietin 2 monoclonal antibody, inhibits angiogenesis and tumor growth in human tumor xenograft models. Int J Oncol. 40 1321-1330. Leung, D. W., 1989. Vascular e ndothelial growth factor is a secreted angiogenic mitogen. Science 246 1306-1309. Li, C. Y., et al., 2000. Initia l stages of tumor cell-induced angiogenesis: evaluation via skin window chambers in rodent models. J Natl Cancer Inst. 92 143-147. Li, C., et al., 2013. Angiopoietin-2 ex pression is correlated with angiogenesis and overall survival in oral squamous cell carcinoma. Med Oncol. 30 571-581. 150

PAGE 151

Lind, A. J., et al., 2005. Angi opoietin 2 expression is relat ed to histological grade, vascular density, metastases, and outcome in prostate cancer. Prostate. 62 394399. Loges, S., et al., 2010. Mechanisms of re sistance to anti-angiogenic therapy and development of third-generat ion anti-angiogenic drug candidates. Genes Cancer. 1 12-25. Loughna, S., Sato, T. N., 2001. Angiopoietin and Tie sig naling pathways in vascular development. Matrix Biol. 20 319-325. Lowenstein, C. J., et al., 2005. Regulation of Weibel-Pal ade body exocytosis. Trends Cardiovasc Med. 15 302-308. Maisonpierre, P. C., et al., 1997. Angiopoiet in-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science. 277 55-60. Matsushita, K., et al., 2005. Vascular endothe lial growth factor regulation of WeibelPalade-body exocytosis. Blood. 105 207-214. Maxwell, P. H., et al., 1997. Hypoxia-inducible factor-1 modulates gene expression in solid tumors and influences both angiogene sis and tumor growth. Proc Natl Acad Sci U S A. 94 8104-8109. Mazzieri, R., et al., 2011. Targeting the ANG2/TIE2 axis inhibits tumor growth and metastasis by impairing angiogenesis and disabling rebounds of proangiogenic myeloid cells. Cancer Cell. 19 512-526. McDonald, D. M., Angiogenesis and Vascular Remodeling in Inflammation and Cancer: Biology and Architecture of the Vasculatu re. In: J. Folkman, W. D. Figg, Eds. Angiogenesis: An Integrative Approach fr om Science to Medicine. Springer New York, NY, 2008. Meadows, K. L., Hurwitz, H. I., Anti-VEGF Therapies in the Clinic. In: M. Klagsburn, P. D'Amore, Eds. Perspectives in Medicine Cold Spring Harbor Laboratory Press, New York, 2012. Minchinton, A. I., Tannock, I. F., 2006. Drug penetration in solid tumours. Nat Rev Cancer. 6 583-592. Minchinton, A. I., Kyle, A. H., Drug Penetration and Therapeutic Resistance. In: D. W. Siemann, (Ed.), Tumor Microenvironment. Wiley-Blackwell Lt d, West Sussex, United Kingdom, 2011. Mita, A. C., et al., 2010. Phase 1 study of AMG 386, a selectiv e angiopoietin 1/2neutralizing peptibody, in combination with chemotherapy in adults with advanced solid tumors. Clin Cancer Res. 16 3044-3056. 151

PAGE 152

Mitsuhashi, N., et al., 2003. Angiopoietin s and Tie-2 expression in angiogenesis and proliferation of human hepatocellu lar carcinoma. Hepatology. 37 1105-1113. Molnar, N., Siemann, D. W., 2012. Inhibition of endothelial/smooth muscle cell contact loss by the investigational angiopoieti n-2 antibody MEDI3617. Microvasc Res. 83, 290-297. Montero, A. J., et al., 2012. Bevacizumab in the treatment of metastatic breast cancer: friend or foe? Curr Oncol Rep. 14 1-11. Moon, W. S., et al., 2006. Expression of angiopoietin 1, 2 and their common receptor Tie2 in human gastric carcinoma: implication for angiogenesis. J Korean Med Sci. 21 272-278. Moy, A. J., et al., 2011. Wide-field func tional imaging of blood flow and hemoglobin oxygen saturation in the rodent dorsal window chamber. Microvasc Res. 82 199209. Neal, J., Wakelee, H., 2010. AMG-386, a selective angiopoietin-1/-2-neutralizing peptibody for the potential treatment of cancer. Curr Opin Mol Ther. 12 487-495. Oliner, J. D., et al., 2012. AMG 386, a sele ctive angiopoietin 1/2-ne utralizing peptibody, inhibits angiogenesis in models of ocular neovascular diseases. Invest Ophthalmol Vis Sci. 53 2170-2180. Overgaard, J., Horsman, M. R., 1996. Modification of Hypoxia Induced Radioresistance in Tumors by the Use of Oxygen and Sensitizers. Semin Radiat Oncol. 6 10-21. Palmer, G. M., et al., 2011. In vivo optical molecular imag ing and analysis in mice using dorsal window chamber models applied to hypoxia, vasculature and fluorescent reporters. Nat Prot. 6 1355-1366. Papetti, M., Herman, I. M., 2002. Mec hanisms of normal and tumor-derived angiogenesis. Am J Physiol Cell Physiol. 282 C947-C970. Park, J. H., et al., 2007. Serum angiopoietin -2 as a clinical marker for lung cancer. Chest. 132 200-206. Park, S. O., et al., 2009. Real-time imaging of de novo arteriovenous malformation in a mouse model of hereditary hemorrhagic telangiectasia. J Clin Invest. 119 34873496. Partanen, J., et al., 1992. A novel endothelial cell surface receptor tyrosine kinase with extracellular epidermal growth factor homology domains. Mol Cell Biol. 12 16981707. 152

PAGE 153

Patterson, D. M., et al., 2012. Phase I c linical and pharmacokinetic evaluation of the vascular-disrupting agent OXi4503 in patients with advanced solid tumors. Clin Cancer Res. 18 1415-1425. Pichiule, P., et al., 2004. Hypoxic regulation of angiopoietin-2 expression in endothelial cells. J Biol Chem. 279 12171-12180. Pollack, A., F.D.A. Revokes Approval of Avastin for Use as Breast Cancer Drug. The New York Times New York 2011, pp. B1. Pralhad, T., et al., 2003. Concept, mechani sms and therapeutics of angiogenesis in cancer and other diseases. J Pharm Pharmacol. 55 1045-1053. Procopio, W. N., et al., 1999. Angiopoietin-1 and -2 coiled coil domain s mediate distinct homo-oligomerization patterns, but fi brinogen-like domains mediate ligand activity. J Biol Chem. 274 30196-30201. Ribatti, D., et al., 2000. The di scovery of angiogenic factors: a historical review. Gen Pharmacol. 35 227-231. Rini, B., et al., 2012. AMG 386 in combination with sorafenib in patients with metastatic clear cell carcinoma of the kidney : a randomized, double-blind, placebocontrolled, phase 2 study. Cancer. 118 6152-6161. Robson, E. J., Ghatage, P., 2011. AMG 386: profile of a nov el angiopoietin antagonist in patients with ovarian cancer. Expert Opin Investig Drugs. 20 297-304. Rondaij, M. G., et al., 2006. Dynamics and plasticity of Weibel-Palade bodies in endothelial cells. Arterioscler Thromb Vasc Biol. 26 1002-1007. Roviezzo, F., et al., 2005. A ngiopoietin-2 causes inflammation in vivo by promoting vascular leakage. J Pharmacol Exp Ther. 314 738-744. Saharinen, P., Alitalo, K., 2011. The yin, t he yang, and the angiopoietin-1. J Clin Invest. 121 2157-2159. Sato, T. N., et al., 1995. Distinct roles of t he receptor tyrosine kinases Tie-1 and Tie-2 in blood vessel formation. Nature. 376 70-74. Scharpfenecker, M., et al., 2005. The Tie-2 ligand angiopoietin -2 destabilizes quiescent endothelium through an inter nal autocrine loop mechanism. J Cell Sci. 118, 771780. Schnurch, H., Risau, W., 1993. Ex pression of tie-2, a mem ber of a novel family of receptor tyrosine kinases, in the endothelial cell lineage. Development. 119 957968. 153

PAGE 154

Schueneman, A. J., et al ., 2003. SU11248 maintenance therapy prevents tumor regrowth after fractionated irradiation of murine tumor models. Cancer Res. 63 4009-4016. Schulz, P., et al., 2011. Angi opoietin-2 drives lymphatic metastasis of pancreatic cancer. Faseb J. 25 3325-3335. Seegar, T. C., et al., 2010. Tie1-Tie2 interactions medi ate functional differences between angiopoietin ligands. Mol Cell. 37 643-655. Senan, S., Smit, E. F., 2007. Design of clin ical trials of radiation combined with antiangiogenic therapy. Oncologist. 12 465-477. Shaked, Y., et al., Combin ing Antiangiogenic Drugs with Vascular Disrupting Agents Rationale and Mechanisms of Action. In: T. Meyer, Ed. Vascular Disruptive Agents for the Treatment of Cancer. Springer Science+Business Media LLC, New York, NY, 2010. Shibuya, M., Vascular Permeabili ty/Vascular Endothelial Growth Factor. In: J. Folkman, W. D. Figg, Eds. Angiogenesis: An In tegrative Approach from Science to Medicine. Springer Science + Busine ss Media, LLC, New York, NY, 2008. Sie, M., et al., 2009. The angiopoietin 1/angiopoietin 2 balance as a prognostic marker in primary glioblastoma mu ltiforme. J Neurosurg. 110 147-155. Siemann, D. W., et al., 2004. Vascular-targeti ng therapies for treatment of malignant disease. Cancer. 100 2491-2499. Siemann, D. W., Shi, W., 2004. Efficacy of combined ant iangiogenic and vascular disrupting agents in treatment of solid tumors. Int J Ra diat Oncol Biol Phys. 60 1233-1240. Siemann, D. W., et al., 2005. Differentia tion and definition of vascular-targeted therapies. Clin Cancer Res. 11 416-420. Siemann, D. W., Tumor Vasculature: a Target for Anticancer Therapies In: D. W. Siemann, Ed. Vascular Targeted Therapies in Oncology. Wiley Ltd., West Sussex, United Kingdom, 2006. Siemann, D. W., Shi, W., 2008. Dual Targeting of Tumor Vasculature: Combining Avastin and Vascular Disrupting Agents (C a4P or Oxi4503). Anticancer Res. 28 2027-2032. Siemann, D. W., 2011. The unique characterist ics of tumor vasculature and preclinical evidence for its selective disruption by Tumor-Vascular Disrupting Agents. Cancer Treat Rev. 37 63-74. 154

PAGE 155

Siu, L. L., Moore, M. J., Pharmacology of Anticancer Drugs. In: I. F. Tannock, et al., Eds. The Basic Science of Oncology. Mc Graw-Hill Medical P ublishing Division, 2005. Solinas, G., et al., 2009. Tumo r-associated macrophages (TAM) as major players of the cancer-related inflammation. J Leukoc Biol. 86 1065-1073. Song, S. H., et al., 2012. Tie1 regulates the Tie2 agonistic role of angiopoietin-2 in human lymphatic endothelial cells. Biochem Biophys Res Commun. 419 281286. Sorg, B. S., et al., 2005. Hyperspectral imaging of hemoglobin saturation in tumor microvasculature and tumor hypoxia development. J Biomed Opt 10 044004-111. Staton, C. A., et al., 2010. Angiopoietin-1, angiopoietin-2 and Tie-2 receptor expression in human dermal wound repair and scarring. Br J Dermatol. 163 920-927. Stratmann, A., et al., 1998. Cell type-s pecific expression of angiopoietin-1 and angiopoietin-2 suggests a role in gliobl astoma angiogenesis. Am J Pathol. 153 1459-1466. Sugimachi, K., et al., 2003. Angiopoiet in switching regulates angiogenesis and progression of human hepatocellular carcinoma. J Clin Pathol. 56 854-860. Suri, C., et al., 1996. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell. 87 1171-1180. Suri, C., et al., 1998. Increas ed vascularization in mice overexpressing angiopoietin-1. Science. 282 468-471. Tait, C. R., Jones, P. F., 2004. Angiopoietins in tumours: the angiogenic switch. J Pathol. 204 1-10. Tanaka, F., et al., 2002. Expr ession of angiopoietins and its clinical significance in nonsmall cell lung cancer. Cancer Res. 62 7124-7129. Terman, D. S., et al., 2013. Sickle erythro cytes target cytotoxics to hypoxic tumor microvessels and potentiate a tumo ricidal response. PLoS One. 8 e52543. Thomas, M., Augustin, H. G., 2009. The role of the Angiopoietins in vascular morphogenesis. Angiogenesis. 12 125-137. Thomas, M., et al., 2013. A novel angiopoie tin-2 selective fully human antibody with potent anti-tumoral and anti-angiogenic efficacy and superior side effect profile compared to Pan-Angiopoietin-1/2 inhibitors. PLoS One. 8 e54923. 155

PAGE 156

Tsigkos, S., et al., 2006. Regulation of Ang2 release by PTEN/PI3-kinase/Akt in lung microvascular endothelial cells. J Cell Physiol. 207 506-511. Tsutsui, S., et al., 2006. Angiopoietin 2 expr ession in invasive ductal carcinoma of the breast: its relationship to the VEGF expr ession and microvessel density. Breast Cancer Res Treat. 98 261-266. Valenzuela, D. M., et al., 1999. Angiopoietins 3 and 4: diverging gene counterparts in mice and humans. Proc Natl Acad Sci U S A. 96 1904-1909. van der Heijden, M., et al., 2009. Circulating angiopoietin-2 levels in the course of septic shock: relation with fluid balance, pulmonar y dysfunction and mortality. Intensive Care Med. 35 1567-1574. Vanchieri, C., Williams, S. C. P., Managing the nations research portfolio: an annual plan and budget proposal for fiscal year 2013. The national cancer program, National Cancer Institute. NIH Publication, 2012. Vaupel, P., et al., 1998. Modulat ion of tumor oxygenation. In t J Radiat Oncol Biol Phys. 42, 843-848. Vaupel, P., Abnormal Microvasculature and Defective Microcirculatory Function in Solid Tumors. In: D. W. Siemann, Ed. Vascular Targeted Therapies in Oncology. Wiley Ltd, West Sussex, United Kingdom, 2006. Venneri, M. A., et al., 2007. Identificati on of proangiogenic TIE2 -expressing monocytes (TEMs) in human peripheral blood and cancer. Blood. 109 5276-5285. Wankhede, M., et al., 2010a. Spectral imagi ng reveals microvessel physiology and function from anastomoses to thromboses. J Biomed Opt. 15 011111. Wankhede, M., et al., 2010b. In vivo functional differences in microvascular response of 4T1 and Caki-1 tumors after treat ment with OXi4503. Oncol Rep. 23 685-692. Weidner, N., et al., 1991. Tumor angiogenesi s and metastasis--correlation in invasive breast carcinoma. N Engl J Med. 324 1-8. Weis, S. M., Cheresh, D. A., 2011. Tu mor angiogenesis: molecular pathways and therapeutic targets Nat Med. 17 1359-1367. Welford, A. F., et al., 2011. TIE2-expre ssing macrophages limit the therapeutic efficacy of the vascular-disrupting agent combre tastatin A4 phosphate in mice. J Clin Invest. 121 1969-1973. Wilhelm, S. M., et al., 2011. Regorafeni b (BAY 73-4506): a new oral multikinase inhibitor of angiogenic, stromal and onc ogenic receptor tyrosine kinases with potent preclinical antitumor activity. Int J Cancer. 129 245-255. 156

PAGE 157

157 Xia, Y., et al., 2012. Angiogenic factors in chronic lymphocytic leukemia. Leuk Res. 36 1211-1217. Xing, F., et al., 2010. Cancer associated fi broblasts (CAFs) in tumor microenvironment. Front Biosci. 15 166-179. Yamakawa, M., et al., 2004. Ex pression of angiopoietins in r enal epithelial and clear cell carcinoma cells: regulation by hypoxia and participation in angiogenesis. Am J Physiol Renal Physiol. 287 F649-F657. Young, R. J., Reed, M. W., 2012. Antiangiogenic therapy: concept to clinic. Microcirculation. 19, 115-125. Yu, X., et al., 2013. Structural basis for angi opoietin-1-mediated signal ing initiation. Proc Natl Acad Sci U S A. 110 7205-7210. Yuan, F., et al., 1996. Time-dependent vascu lar regression and permeability changes in established human tumor xenografts i nduced by an anti-vascular endothelial growth factor/vascular permeability factor antibody. Proc Natl Acad Sci U S A. 93, 14765-14770. Yuan, H. T., et al., 2007. Activation of t he orphan endothelial receptor Tie1 modifies Tie2-mediated intracellular signali ng and cell survival. Faseb J. 21 3171-3183. Zagzag, D., et al., 1999. In si tu expression of angiopoietins in astrocytomas identifies angiopoietin-2 as an early marker of tumor angiogenesis. Exp Neurol. 159 391400. Zhang, Y., et al., 2010. XL-184, a MET, VEGFR-2 and RET kinase inhibitor for the treatment of thyroid cancer, glioblas toma multiforme and NSCLC. IDrugs. 13 112-121. Zhu, A. X., et al., 2013. Efficacy, safety, pharmacokinetics, and biom arkers of cediranib monotherapy in advanced hepatocellular carcinoma: a phase II study. Clin Cancer Res. 19 1557-1566. Zu Heringdorf, D. M., et al., 2013. Pharma cology of the sphi ngosine-1-phosphate signalling system. Handb Exp Pharmacol. 215 239-253.

PAGE 158

BIOGRAPHICAL SKETCH Nikolett Molnar obtained her Bachelor of Science degree in biology from Gettysburg College, Pennsyl vania (2003-2007) and joined the Interdisciplinary Biomedical Sciences Program at the University of Florida Co llege of Medicine in 2008. Under the mentorship of Dr. Dietmar Siemann, Miss Molnar has presented her research in several international conferences such as the American Association for Cancer Research (AACR), Molecular Targets and Cancer Therapeutics, The International Workshop on the Tumor Microenvironment, and t he International Symposium on AntiAngiogenic Therapy as well as participated in the Shands Cancer Center Research Day and the College of Medicine Research Day. Throughout her training she has received honors and awards such as the NIH T32 Trai ning Grant in Cancer Biology, best predoctoral poster award at the Shands Cancer Center Research Day and junior investigator award at the International Workshop on the Tumor Microenvironment. She has also successfully collabor ated with a laboratory at the Department of Biomedical Sciences that led to co-authorship on poste r presentations and a publication. She has shared her passion for cancer biology re search by teaching a course entitled Cancer Biology and Therapeutics to honors high school student s during the summer for three years as part of the University of Flor ida Student Science Tr aining Program. Upon completion of her PhD program in Medical Sciences in August 2013, she aspires to remain in the cancer research field. 158