Role of Activin Receptor-Like Kinase 1 (ALK1) in Regulation of Angiogenesis

Material Information

Role of Activin Receptor-Like Kinase 1 (ALK1) in Regulation of Angiogenesis
Choi, Eun-Jung
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (121 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Medical Sciences
Molecular Cell Biology (IDP)
Committee Chair:
Oh, Suk P.
Committee Members:
Liao, Daiqing
Narayan, Satya
Liu, Chen


Subjects / Keywords:
Angiogenesis ( jstor )
Blood vessels ( jstor )
Capillaries ( jstor )
Cardiovascular system ( jstor )
Cells ( jstor )
Endothelial cells ( jstor )
Hereditary hemorrhagic telangiectasia ( jstor )
Ligands ( jstor )
Receptors ( jstor )
Tumors ( jstor )
Molecular Cell Biology (IDP) -- Dissertations, Academic -- UF
alk1, angiogenesis, blood, cells, endothelial, hht, signaling, tgf, tumor, vessels
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Medical Sciences thesis, Ph.D.


Hereditary hemorrhagic telangiectasia (HHT), also known as Rendu-Osler-Weber syndrome, is a genetic vascular disease inherited in an autosomal dominant manner. Its clinical manifestations are recurrent nosebleeds, multiple mucocutaneous hemorrhages and arteriovenous malformations (AVMs) in multiple organs including the brain, lung, liver and gastrointestinal tract. Three distinctive types of HHT have been categorized depending on their different disease-causing genes. The HHT type 1 (HHT1), type 2 (HHT2) and a combined syndrome of HHT and Juvenile Polyposis (JP-HHT) are predisposed by various homozygous mutations in ENDOGLIN (ENG), Activin receptor-Like Kinase 1 (ACVRL1; ALK1) and SMAD4 genes, respectively. Interestingly, all of these genes encode proteins that are implicated in the transforming growth factor (TGF)-? superfamily signaling. Therefore, HHT has been considered as a TGF-? disease. Previously, our in vivo studies have demonstrated that conditional deletion of the Alk1 gene in endothelial cells (ECs) is sufficient for development of AVMs in various vascular beds. In order to investigate the molecular mechanism by which endothelial Alk1-deficiency leads to AVM formation, we have characterized Alk1-null pulmonary endothelial cell lines at the biochemical and cellular levels. We found that Alk1-null pulmonary ECs displayed significantly increased migratory properties in vitro and in vivo in response to an angiogenic factor. In tube forming assays on Matrigel, Alk1-null ECs formed a vascular network with higher density, length, and thickness compared to corresponding control ECs, indicating that ALK1 may modulate responses to an angiogenic stimulus during angiogenesis. Interestingly, while an inhibitory effect of BMP-9 in tube formation was blunted in Alk1-null ECs, BMP-9-mediated downstream SMAD1/5 phosphorylation was unaffected. Taken together, this data suggest that ALK1 is an important cellular modulator for angiogenic stimuli during angiogenesis via a SMAD-independent pathway. Since ALK1-deficient ECs form AVMs in angiogenic environment, we examined whether ALK1-deficiency can impact tumor formation or growth by inducing non-productive blood vessels feeding to tumor. We observed that Alk1-deletion considerably inhibited formation and growth of Lewis lung carcinoma (LLC). We found numerous AVMs in the peripheral tumor microcirculation and disorganized blood vessels in the tumor. This data indicate that ALK1 can be a novel target of tumor angiogenic therapy. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
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.
Thesis (Ph.D.)--University of Florida, 2009.
Adviser: Oh, Suk P.
Electronic Access:
Statement of Responsibility:
by Eun-Jung Choi.

Record Information

Source Institution:
Rights Management:
Copyright Choi, Eun-Jung. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
LD1780 2009 ( lcc )


This item has the following downloads:

Full Text




2 2009 Eun Jung Choi


3 To my loving mom, husband and sister


4 ACKNOWLEDGMENTS There have been so many people who have helped me complete my graduate work. I would like to give my great thanks to all of them from the bottom of my heart. I am sure that I would never follow my queries in science this far without them. I thank my loving family for their unconditional and endless love, encouragement and support. E specially, my mom, InSook Kim, always cheers, inspires and strengthens me in the ups and downs throughout my life. I am so grateful to my wonderful and lovely husband, Koo Yung Jung. He is always my big brother who constantly loves, cares and motivates me. I also thank my sister, SuJung Choi for her lifetime friendship with me sharing good and bad times. I would like to give my special thanks to my mentor Dr. S. Paul Oh for his outstanding support and guidance. I have been inspired and awed by his persistent passion toward science. Dr. Oh helped me gain an understanding of critical thinking and creativity that are absolutely needed in the field of research. He also helped me build my enthusiasm and knowledge to succeed in science. I also thank my dissertation committee members, Dr. Satya Narayan, Dr. Daiqing Liao and Dr. Chen Liu for their constructive comments, assistance and encouragement. I also truly acknowledge professors from my undergraduate study, Drs. Tae Young Jung, ChangWon Kang, Gi l Saeng Jung, HongYang Park and HoonTaek Lee for being continuously mentors and supporters. I am also indebted to all past and current Oh lab members for providing great companionship. I am especially grateful to Naime Fliess for not only her help in res earch but also her kindness and support. Her pleasant personality, generosity and great sense of humor made our lab a friendly and warm place. I would like to acknowledge Dr. YoungJae Lee and Melissa Chen for their advice and assistance in working with ce lls. I truly thank Dr. Chastity Bradford for her helpful discussions Ha Long Nguyen and Jairo Tabora for their generous helps


5 in preparation of presentations and writings. I would also like to thank KwonHo Hong and Chul Han for their great friendship and helps scientifically and personally. Lastly, I would like to express my gratitude to Kimberly Hodges for her kind assistance and concern on handling all administrative works properly and timely.


6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................9 LIST OF FIGURES .......................................................................................................................10 ABSTRACT ...................................................................................................................................11 C H A P T E R 1 INTRODUCTION ..................................................................................................................13 Organization of Vascular Network .........................................................................................13 Anatomy and Function of Blood Vessels ........................................................................13 Vascular Circulatory System ...........................................................................................14 Angiogenesis ...........................................................................................................................14 Types of Angiogenesis ....................................................................................................14 Overview of Angiogenesis ..............................................................................................15 Angiogenic Signaling Pathways .............................................................................................15 Vascular Endothelial Growth Factor (VEGF) Family Signaling ....................................16 Fibroblast Growth Factor (FGF) Family Signaling .........................................................17 Transforming Growth Factor (TGF).......................................17 Notch/Delta Family Signaling .........................................................................................18 Eph/ephrin Family Signaling ...........................................................................................19 Other Family Signaling ...................................................................................................20 Hereditary Hemorrhagic Telangiectasia (HHT) .....................................................................21 Clinical Features of HHT ................................................................................................21 Arteriovenous Malformation (AVM) ..............................................................................22 Clinical Management and Treatment ..............................................................................23 Transgenic Mouse Models for HHT2 ..............................................................................23 Aim and Significance of the Study .........................................................................................24 2 MOLECULAR AND CELLULAR CHARACTERIZATION OF ALK1 DELETED PULMONARY ENDOTHELIAL CELLS (ECS) ..................................................................27 Background .............................................................................................................................27 Overview of Hereditary Hemorrhagic Telangiectasia (HHT) .........................................27 Signal Transduction for TGF ..................................................................27 Vascular Endothelial Specific ALK1 Signaling .............................................................29 Animal Models for HHT .................................................................................................30 Is ALK1 Signaling Independent from TGF ...............................30 HHT Like Vascular Lesions Induced by Alk1 Deletion at the Adult Stages ..................31


7 Results .....................................................................................................................................33 Establishment of Genetically Modified Pulmonary EC Line ..........................................33 Development of Alk12f/1f ( Alk1 Null Heterozygote) Derived Alk11f/1f ( Alk1 Null Homozygote) ECs ........................................................................................................34 Alk1 Null ECs Showed Increased Migratory Index in Response to an Angiogenic Challenge in vitro .........................................................................................................34 Alk1 Deficient ECs Resulted in Excessive, Disorganized and Enlarged Tubular Network Formation upon an Angiogenic Challenge in vitro .......................................35 Antiangiogenic Effect of BMP 9 was Blunted, But That of TGF Alk1 null ECs ...............................................................................................................37 Alk1 Deletion Caused Higher Migratory Indication of ECs and Abnormal Blood Vessel Formation in vivo .............................................................................................40 Biochemical Characterization of Alk1 Null Pulmonary ECs ..........................................41 Discussion ...............................................................................................................................44 3 NOVEL PHYSIOLOGICAL EFFECT(S) OF ARTERIOVENOUS MALFORMATIONS (AVMS) ON TUMOR VASCULATURE ..........................................68 Background .............................................................................................................................68 Tumor Angiogene sis .......................................................................................................68 Antiangiogenic Strategies Targeting Cancers .................................................................69 Lessons from Numerous VEGF/VEGFR Inhibitor Based Clinical Trials ......................69 Emerging Novel Antiangiogenic Targets ........................................................................70 A Novel Anti Tu mor Effect of Alk1 Deletion Induced AVMs on Tumor Vasculature ..................................................................................................................71 Results .....................................................................................................................................73 Initiation of Tumorigenesis Was Suppressed in Alk1 Deleted Adult Mutant Mice ........73 Progression of Tumorigenesis Was Significantly Inhibited in Alk1 Deleted Adult Mutant Mice .................................................................................................................75 AVMs Were Resulted From Alk1 Deficiency in Peripheral TumorFeeding Blood Vessels .........................................................................................................................76 Alk1 De letion during Tumor Angiogenesis Caused Disruption of Tumor Vascular Network ........................................................................................................................77 Discussion ...............................................................................................................................79 4 CONCLUSIONS AND FUTURE STUDIES .........................................................................88 5 MATERIALS AND METHODS ...........................................................................................94 Transgenic Mice .....................................................................................................................94 Overall Cell Culture Conditions .............................................................................................94 Establishment of Alk12f/1f and Alk11f/1f Pulmonary Endothelial Cells (pECs) .........................94 Sorting pECs by Fluorescent Activated Cell Sorting (FACS) ...............................................95 Genomic DNA PCR Analysis ................................................................................................96 RT PCR Analysis ...................................................................................................................96 Western Blotting .....................................................................................................................97 in vitro Endothelial Migration Assay .....................................................................................98


8 in vitro Tube Formation Assay on Matrigel ...........................................................................99 in vivo Matrigel Plug Angiogenesis Assay .............................................................................99 X Gal Staining ......................................................................................................................100 in vivo Subcutaneous Tumor Generation ..............................................................................101 in vivo Intramuscular Tumor Generation ..............................................................................102 Latex Dye Injection ..............................................................................................................102 Histology and Immunohistochemistry ..................................................................................103 Statistics ................................................................................................................................104 LIST OF REFERENCES .............................................................................................................107 BIOGRAPHICAL SKETCH .......................................................................................................121


9 LIST OF TABLES Table page 51 Summary of primers used for genomic PCR analysis. ....................................................105 52 Summary of primers used for RT PCR analysis. .............................................................106


10 LIST OF FIGURES Figu re page 11 TGF ........................................................................26 21 Experimental scheme for cellular and molecular characteriza tion of Alk1 heterozygous( Alk12f/1f) and Alk1 homozygous ( Alk11f/1f) pulmonary ECs. ........................52 22 Establishment of R26+/CreER; Alk12f/1f (equivalent to Alk1 null heterozygote; Alk1+/ -) parental pulmonary endothelial cell line. ...........................................................................53 23 Derivation of R26+/CreER; Alk11f/1f (equivalent to Alk1 null homozygote; Alk1/-) ECs from parental EC line. ........................................................................................................54 24 Alk1 null pulmonary ECs displayed elevated migratory index upon angiogenic factor challenge. ...........................................................................................................................55 25 in vitro Matrigel tube formation assay. ..............................................................................57 26 Alk1 null ECs formed thicker tube like structures that are resistant to regression.. ..........58 27 ALK1 deficiency resulted in an increase in length of tubes and sprouting of ECs.. .........60 28 Inhibitory effect of BMP 9 on angiogenesis was diminished, whereas that of TGF existed in Alk1 null ECs .....................................................................................................62 29 in vivo Alk1 deletion also resulted in higher migratory and proangiogenic properties of ECs. ................................................................................................................................64 210 ALK1 signaling in pulmonary ECs was not specific for SMAD dependent nor t he ERK MAPK pathways. ......................................................................................................65 211 BMP specific SMAD1/5/8 pathway for ALK1 signaling was compensated by other TGF ........................................................................67 31 Initiation of tumor growth was repressed in R26+/CreER; Alk11f/1f mut ant mice. .................83 32 Tumor growth was significantly inhibited in Alk1 null mutant mice. ...............................84 33 Latex day injection revealed Alk1 del etion caused AVMs in tumor feeding blood vasculature.. .......................................................................................................................85 34 Tumor blood vessels were destroyed by Alk1 deficiency. .................................................86 35 Tumo r vascular integrity was disrupted by Alk1 deletion. ................................................87


11 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ROLE OF ACTIVIN RECEPTORLIKE KINASE 1 (ALK1) IN REGULATION OF ANGIOGENESIS By Eun Jung Choi August 2009 Chair: S. Paul Oh Major: Medical Sciences Molecular Cell Biology Hereditary hemorrhagic telangiectasia (HHT), also known as RenduOsler Weber syndrome, is a genetic vascular disease inherited in an autosomal dominant manner. Its clinical manifestations are recurrent nosebleeds, mult iple mucocutaneous hemorrhages and arteriovenous malformations (AVMs) in multiple organs including the brain, lung, liver and gastrointestinal tract. Three distinctive types of HHT have been categorized depending on their di fferent disease causing genes. The HHT type 1 (HHT1), type 2 (HHT2) and a combined syndrome of HHT and Juvenile Polyposis ( JP HHT ) are predisposed by various homozygous mutations in ENDOGLIN ( ENG), Activin receptor Like Kinase 1 ( ACVRL1 ; ALK1 ) and SMAD4 genes, respectively. Interestingly, all of these genes encode proteins that are implicated in the transforming growth factor (TGF) superfamily signaling. Therefore, HHT has been considered as a TGF disease. Previously, our in vivo studies have demonstrated that conditional deletion of the Alk 1 gene in endothelial cells (ECs) is sufficient for development of AVMs in various vascular beds. In order to investigate the molecular mechanism by which endothelial Alk1deficiency leads to AVM formation, we have characterized Alk1 null pulmonary endothelial cell lines at the biochemical and cellular levels. We found that Alk1 null pulmonary EC s displayed significantly


12 increased migratory properties in vitro and in vivo in response to an angiogenic factor. In tube forming assays o n Matrigel, Alk1 null ECs formed a vascular network with higher density, length, and thickness compared to correspond ing control ECs, indicating that ALK1 may modulate responses to an angiogenic stimulus during angiogenesis. Interestingly, while an inhibitory effect of BMP 9 in tube formation was blunted in Alk1 null ECs, BMP 9mediated downstream SMAD1/5 phosphorylation was unaffected. Taken together, this data suggest that ALK1 is an important cellular modulator for angiogenic stimuli during angiogenesis via a SMADindependent pathway. Since ALK1 deficient ECs form AVMs in angiogenic environment, we examined whether ALK1deficiency can impact tumor formation or growth by inducing nonproductive blood vessels feeding to tumor. We observed that Alk1 deletion considerably inhibited formation and growth of Lewis lung carcinoma (LLC) We found numerous AVM s in the peripheral tumor microcirculation and disorganized blood vessels in the tumor. This data indicate that ALK1 can be a novel target of tumor angiogenic therapy.


13 CHAPTER 1 INTRODUCTION Organization of Vascular Network The development and maintenance of blood c irculation are essential processes for the vertebrate life. The blood circulation throughout the body is achieved through blood vessels which roughly consist of arteries, veins and capillaries. The hierarchical vascular system is organized into a functiona l vascular network required and established by a fine tuned orchestration of the cross talk between numerous molecular signaling pathways and several different cell types during the developmental stages (1) It functions mainly to deliver adequate supplies including oxygen and nutrients and eliminate wastes. Anatomy and Function of Blood Vessels Capillaries are the thinnest and most ample blood vessels (1) The structure of their walls consists of a single layer of endothelial cells (ECs; endothel ium) surrounded by a layer of a basement membrane (BM) and pericytes. Such a structure allows the exchange of molecu les between blood and tissues. As compared to capillaries, arterioles an d venules are sheathed with more mural cells (1) acting as a b ridge between arteries and veins with capillary beds. Arteries and veins share the same fundamental structure (1) The walls of these large vessels are comprised of three layers from the inner most layer to the outer most layer: an intima, media and a dventitia. The intima and media are sheets containing endothelial cells and smooth muscle cell s (SMCs), respectively, while the adventitia is a zone comprised of fibroblasts, an extracellular matrix (ECM) and elastic laminae. The vascular diameter, vessel tone and blood flow are modulated by SMCs and elastic laminae.


14 Vascular Circulatory System The circulation is a tightly regulated and closed system that delivers blood throughout the body. It starts from the left ventricle of the heart from which oxygenated blood enters into arteries The arteries are connected to arterioles that lead arterial blood into the capillary bed. The capillary bed is a network of the smallest and the thinnest blood vessels where the blood supplies oxygen, water and nutrients to surrounding tissues as well as removes local carbon dioxide and waste products The deoxygenated blood flows from the capillary bed through venules into the veins and returns to the heart. As capillaries provide actual conduits enabling blood to supply all t issues of the body, its sustained permeability is crucial for ensuring molecular exchanges and maintaining vascular homeostasis. Angiogenesis Nascent blood vessels are formed by two processes: vasculogenesis and angiogenesis. Vasculogenesis occurs mainly during embryonic development in which a primary capillary plexus is established from the differentiation of endothelial progenitor cells (2) On the other hand, angiogenesis refers the formation of new blood vessels by sprouting or splitting fr o m pre existing vessels during both development and in postnatal life (3) In the adult, angiogenesis is usually initiated in response to either physiological or pathological stimuli. Physiological angiogenesis takes place during wound healing, inflammation(4, 5) in the female reproduct ive system (6) and is tightly regulated. However, in pathological conditions, such as proliferative re tinopathy, rheumatoid arthritis and tumorigenesis (4, 5) it is uncontrolled. Types of Angiogenesis Two disti nctive types of angiogenesis have been described (7) One is sprouting angiogenesis which is the true branching of capillaries from pre existing vessels. It begin s with degradation of ECM by proteolytic enzymes followed by migration and proliferation of EC s


15 towards an angiogenic stimulus. After formation of the vascular lumen, the vascular endotheli um is stabi lized and functionally mature (7) On the other hand, nonsprouting angiogenesis, also known as intussusception is initially seen in the embryonic lung (7, 8) It occurs when the vessel lumen widens and pre existing vessels are subsequently split by transcapillary pillars follow ing pr oliferation of vascular ECs (7) I t is also caused by coalescence and division of capillaries. Overview of Angiogenesis Angiogenesis is a multiple step process that is roughly divided into two phases of activation and resolution (9 12) The activation phase is characterized by initiation and progression steps. Upon angiogenic stimuli, permeability of pre existing capillaries or post capillary venules is increased EC migration and s pr outing into ECM follow s basement membrane degradation. ECs then prolife rate and form a capillary lumen. Conversely, during the resolution phase, EC proliferation and migration are inhibited and BM is reconstit uted. ECs then differentiate and recruit perivascular cells to promote maturation and stabilization of newly formed vessels. Angiogenic Signaling Pathways Many signaling pathways underlying angiogenesis have been identified from genetic studies of animal models (7, 11) The most extensivel y investigated ones are the vascular e ndothelial growth factor (VEGF), fibroblast growth factor (FGF) and transforming growth factor (TGF). R ecently more angiogenic pathways have been discovered (13) such as the Notch/Delta (14 16) ,Eph/Ephrin (17, 18) Hedgehog (19) Sprouty (20) and Rounda bout/Slit (19, 21, 22) families These were initially found to play crucial roles in embryonic vascular development and differentiation Thus, finding these involved in postnatal angiogenesis allows us to broaden our understanding of developmental blood vessel establishment as well as ph ysiological and pathological angiogenesis.


16 Vascular Endothelial Growth Factor (VEGF) Family Signaling The VEGF family includes three tyrosin e kinase receptors (VEGFR 1, 2 and 3), seven ligands (VEGF A, B, C, D, E, placenta growth factor (PL GF) 1 and 2) and two coreceptors (neuropilin (NRP) 1 and 2) (13, 23) Several of the VEGF ligands exert various angiogenic effects depending on their different binding specificities for each of three receptors. The s ignal is then tranduced through several identified secondary pathways such as the Akt and MAPK pathways (13) Originally two receptors, VEGFR 1 (24) and VEGFR 2 (25) were discovered on ECs (23) VEGFR1, also known as FLT 1 ( f ms like tyrosine kinase 1 ) binds to VEGF A, VEGF B and PLGF. It functions differently depending on developmental stages, physiologi cal and pathological conditions and cell types (23) I t is known as a potent VEGF antagonist for VEGF activity either by binding to VEGF with high affinity (26) or by inhibiting VEGFR 2regulated signaling (27) VEGFR2, also known as KDR ( kinase insert domain receptor), is a receptor for VEGF A, VEGF C and VEGF E (13, 23) It is known as the main receptor mediating effects of VEGF A on EC proliferation, migration, invasion, survival (28, 29) and permeability of microvessels (30) during angioge nesis. It ha s been reported that activation of VEGFR 2 alone could efficiently induce angiogenesis in vitro and in vivo (23, 31) Lastly, VEGFR3, also known as FLT4 (fms like tyrosine kinae 4) binds to VEGF C, VEGF D (13, 23) and is highly expressed in embryonic developmental vasculature (13) I n the adult, it is present in lymphatic vessels and implicated in lymphangiogenesi s (23, 32, 33) However, its persistent role in the vasculature regulating VEGFR 2 signaling to maintain vascular integrity has been also suggested (23, 34)


17 Fibroblast Growth Factor (FGF) Family Signaling In vertebrates, there are 22 polypeptide growth factor (FGF114 and 1623) and four tyrosine kinase receptor (FGFR 1, 2, 3, and 4) members of the FGF family (35) They have multiple functions on various cell types that are strictly modulated by the activity and receptor specificity of FGF (35) In the embryonic development, their functions are involved i n patterning, li mb formation, brain development as well as mesoderm and neural induction (36) via control of cell proliferation, migration and differentiation (35) I n adult stages, they regulate wound healing and angiogenic processes. During angiogenesis, acidic FGF (FGF1) and bas ic FGF (FGF2) play si gnificant roles in elevating EC proliferation and tube formation (37) They are believed to be more potent angiogenic factors than VEGF or platelet derived growth factor (PDGF) (37) Tran sforming Growth Factor (TGF)The TGFtransmembrane receptors, including type I (RI), type II (RII), type III (RIII), and several ligands ( TGF orphogenetic proteins (BMPs), growth and differentiation factors (GDFs ) and nodal ) (38, 39) They control various biological processes such as embryogenesis, embryonic patterning, growth, cell cycle regulation and immunosuppression (38, 40, 41) through their influence on proliferation, migration, differentiation and survival of several cell types (39, 42, 43) Furthermore, from genetic studies of mice lacking different TGF the importance of TGF has been well established (7, 11, 38) The cellular effects of TGF rfamily are diversified by combinatorial associations between i ts receptors and ligands (38, 39, 44) In general, dimeri c type II receptors are associated with one of ligands and recruit a specif ic type I receptor, forming a heteromeric complex. T he


18 autophosphorylated type II receptor phosphorylates the type I receptor The activated type I receptor then transduces the sig nal by phosphorylating as series of intracellular signaling molecules, the SMADs. Depending on the ligand/receptor activated, there are two TGF SMADdependent pathways distinguished by the type of receptor regulated SMADs (R SMADs) mediating the signal. SMAD 2/3 is involved with the TGF SMAD1/5/8 modulate BMP signa ls (45 48) After their ph osphorylation by the type I receptor, R SMADs make heteromeric complexes with the common SMAD (co SMAD), named SMAD4. Finally, these complexes translocate into the nucleus where they interact with transcription factors and co factors (co activators and co repressors) to control target gene expression (49, 50) Additionally this signal ing transduction can be enhanced by auxiliary co receptors (type III receptors) GLYCAN, at the cell surface (48) or impeded by the binding of inhibitory SMAD6 and SMAD7 (I SMADs) to the heterodimeric receptor complex (39) (Figure 1 1). Notch/Delta Family Signaling Notch signaling was originally identified in the development of Drosophila (15) It consists of four Notch rec eptors (Notch1, 2, 3 and 4) and five ligands (Jagged 1, 2, Delta 1, 3 and 4) (13, 16) and transduces the signal through the association between one cell bound ligand and another nonautonomous cell containing a receptor (13) Upon the activation of Notch, its intracellu lar domain is cleaved, released and subsequently translocated into the nucleus to activate target genes that inhibit differentiation of cells an d cause cells to prolifer ate. All members are widely present in at least one type of blood vessels (e.g. arteries, veins or capillaries ) or vascular structures like VSMCs or pericytes (13) However, only Notch 4 and Delta 4 are expressed in capillaries (51) Moreover, Notch4 is arterial specific (52) while Delta like 4 is specifically expressed in ECs (53, 54)


19 It was shown that Notch signaling plays an important negative role on endothelial fu nctions during angiogenesis (55) Recently, it has been suggested that VEGF signaling regulates Notch receptor and Delta ligand gene expression (56, 57) It was shown that VEGF through V EGFR 1 and VEGFR 2, but not basic FGF (FGF2), induced expression of Notch1 and its ligand Dll4 (Delta 4) in human arterial endothelial cells (HAECs) (56) Another study further demonstrated that in capillary like network formation and in activation of N otch1, Notch4, VEGFR2 was downregulated, resulting in inhibition of EC proliferation in response to VEGF, but not bFGF (57) It is noteworthy to menti on that expression of Dll4 is almost absent in adult tissues; however, it is highly expressed in the areas where angiogenesis is active and in the tumor vasculature (13, 51, 53) And upon the ac tivation of Dll4 by VEGF signaling, it makes a negative feedback loop to block VEGF induced tumor angiogenesis (57, 58) Eph/ephrin Family Signaling Similar to the Notch/Delta family members, Eph receptors and ephrin ligands are both membrane bound proteins (13, 51) Eph receptor s are transmembrane molecules containing a tyrosine kinase activity and divided i nto two groups, EphA and EphB. Depending on how they are inserted into the plasma membrane, ephrin ligands are classified into A and B types and show remarkable indiscriminati on towards receptor binding (13, 51) These family members are presen t in a broad range of tissues during embryonic and adult stages. Nonetheless, several Eph receptors and ephrin ligands including the EphB3 and EphB4 r eceptors and ephrinA1, ephrinB1 and ephrinB2 ligands have been identified to be specifically expressed in vascular endothelium (13, 51) It has been suggested that ephrinA1 is involved in inflammatory angiogenesis activated by tumor necrosis factor (13, 51, 59) while ephrinB1 increases endothelial capillary like network formation and cell attachment in vivo (13, 51, 60)


20 Other Family Signaling There are three human homologues of the Drosophila hedgehog gene; sonic hedgehog ( SHH), desert hedgehog ( DHH) and Indian hedgehog ( IHH) (13, 51) Th ese associate with the Patched1 receptor to activate the t ranscription factors Gil1, Gil2 and Gil3. Hedgehog signaling is important for the for mation of limb, bone, lung, gut and hair follicles (51) The expression patterns o f three hedgehog proteins are differ ent and deficiency of each one results in distinct phenotypic defects (13, 51) A role for the Hedgehog si gnaling in angiogenesis was implied by evidence showing that its dysregulation causes vascular defects. SHH did not affect EC pr oliferation or migration in vitro (19) However, it induced express ion of VEGF, angiopoietins 1 and 2 from interstitial mesen chymal cells, suggesting an indirect role in angiogenesis by exerting its upstream function on angiogenic factors (19) Furthermore, in mice, administration of SHH resulted in unique enlarged vessels in ischemic hind limbs, indicating its potent angiogenic activity in viv o (19) In mammals, four intracellular isoforms of Spry have been identified. S prouty (Spry) was initially discovered as an inhibitor of FGF signaling in the development of Drosophila tracheal system (13, 51) Bec ause of many similarities between mammalian angiogenesis and Drosophila tracheal development, a role of Spry in angiogenesis was expected (13, 51) The first in vivo evidence was reported to show an inhibitory role for Spry during angiogenesis (20) It was found that over expression of the mouse Spry4 (mSpry4) derived by a recombinant ade noviral vector in the endothelium of a developing mouse embryo resulted in reduced branching of small vessels from large vessels (20) Moreover, in human umbilical vein endothelial cells (HUVECs), in vitro mSpry4 transfection led to decreased cellular migrati on and cell cycle arrest at the G1/S phase through receptor tyrosine kinase pathways such as bFGF (FGF2) or VEGF induced MAPK signaling (13, 20, 51)


21 Lastly, there are two Roundabout receptors and three S lits in hu man (51) The expression of these family m embers was believed to be restricted to cells of neuronal lineage and they are impl icated in regulation of repulsive axonal guidance (21, 61) However, a novel R oundabout receptor, magic roundabout (ROBO4), has been recently identified to be exclusively expressed in ECs (62) Under physiological conditions, it i s absent in adult tissues, however, it i s strongly expressed in some tumor vasculature including bladder, brain, and colonmetastatic liver tumors (22) In this respect, ROBO4 is emerging as a promising antiangiogenic therapeutic target (13, 51) Hereditary Hemorrhagic Telangiectasia (HHT) Disrupted TGF a diversity of diseases including cancer autoimmune diseases, fibrosis and vascular disorders (44, 63, 64) HHT, also known as Osler RenduWeber syndrome, is a n autosomal dominant vascular disease characterized by epistaxis (spontaneous and recurrent nosebleeds), mucocutaneous telangiectases and arteriovenous malformations (AVMs) in internal organs such as brain, lung, liver and gastrointestinal tract (38, 44, 65, 66) It affects 1 in 5,000 10,000 individuals (67 69) worldwide Various types of mutation have been mostly identified in three genes. ENDOGLIN ( ENG ) is responsible for HHT type 1 (HHT1) (7073) Activin receptor Like Kinase 1 ( ALK1 ) is accountable for the HHT type 2 (HHT2) (7476) and SAMD4 is associated with a combined syndrome of HHT and Juvenile Polyposis ( JP HHT ) (77) Clinical Features of HHT The clinical symptoms o f HHT are highly variable even within the same HHT family. HHT is diagnosed by clinical and molecular methods (78) In t erms of clinical evaluation, the affected individual is initially diagnosed to have HHT by t he presence of at least three manifestations according to the Curacao criteria: 1) spontaneous recurrent nosebleeds


22 (e pistaxes) 2) multiple mucocutaneous telangiec tasias on the lips, tongue, oral mucosa, or fingertip 3) visceral involvement (AVM in the brain, lung, liver, gastrointestinal tract, or spine) and 4) appropriate family history (65) Confirmation by m olecular testing is by a genetic test for mutations in the ENDOGLIN, ALK1 and SMAD 4 genes (78) Spontaneous recurrent nosebleeds are the earliest and most common clinical indications of HHT (78) On average, telangiectases of the skin and mucosa typica lly appear when a person is in their 30s and worsen with age. Many patients suffer mild symptoms and live a normal life span (78) A majority of patients possess A VMs in at least one of following: pulmonary (~50%), hepatic (~30%), cerebral (~10%), spinal (~1%) circulations but are usual ly asymptomatic (78, 79) C omplications that arise from HHT symptoms involve chronic a nd severe anemia, stroke and brain abscess. R arely more severe complications can lead to deep venous thromboses, symptomatic liver disease, severe pulmonary hypert ension, pregnancyrelated death and spinovascular accidents (78) Arteriovenous Malformation (AVM) AVM is the ha llmark vascular lesion of HHT. It is a focal pathological condition where arteries and veins are int erconnected without intervening capillaries (38) In a histopathological view, these lesions contain dilated and fused vessels with thin walls. Therefore, they tend to be easily ruptured, ble e d and cannot nurture surrounding tissues well Moreover, AVMs in certain internal organs such as the brain and lung, can be life threatening due to direct arteriovenous (AV) shunts (80, 81) An overview of the establishment of these vascular anomalies was offered by a systemic electron microscopic study (38, 82) It started on the dilation of post capillary venules followed by t he enlargement of their lumen. Arterioles became dilated, and then the venules progressively enlarged and expanded, resulting in two to four AVMs that accompanied disappearance of the capillary bed.


23 Clinical Management and Treatment Si nce each HHT patient has variable and different ial clinical manifestations, treatment options should be designed specifically for each patient (83) T herapeutic options for nosebleeds are tranexamic acid treatments estrogen/progesterone preparations, use of vasodilators, laser phot ocoagulation, dermoplasty of the nasal mucosa and embolization of nasal arteries (65, 8386) Telangiectasias of the skin and the mucosae are commonly treated by medical or laser therapy (84) The gastrointestinal hemorrhages is treated with estrogen/progestero ne, danazol or octreotide (83) T hermic cauterization and laser photocoagulation using endoscopy are performed if there is multiple GI hemorrhaging. Severe anemia caused by recurrent nosebleeds and gastrointestinal bleeding is managed b y blood transfusion and/or iron replacement therapy (83) Transcatheter embolotherapy with balloon or stainless steel coils and surgery are used for the treatment of pulmonary AVMs (PAVMs) (8385, 87) Stereotaxic radiotherapy and transcatheter embolo therapy are performed to treat cerebral fistula (84) Severe liver insufficiency requires organ transplantation (83, 88) Transgenic Mouse Models for HHT2 ALK1 is one of seven type I receptors of TGF (89, 90) Biochemical studies suggested that Activin A (90, 91) TGF and TGF ALK1 (92) Recently BMP 9 and BMP 10 have been further identified (93 95) In general, upon its stimulation, ALK1 phosphorylates and activates the BMP induced SMAD dependent intracellular pathway (38) that is mediated by R SMAD 1/5/8 (96 98) To investigate the role of ALK1 in HHT pathogenesis, genetically modified mouse models have been generated and utilized (38, 44, 97, 99) The Alk1 knockout (KO) was embryonic lethal during midgestation (E10.5 E11.5). Alk1 KO embryos displayed vascular defects due to the elevated activation phase of angiogenesis and the failure of differentiation and recruitment of


24 vascular smooth muscle cells (VSMCs) (97) Furthermore, hyper dilated and fused blood vessels as well as AVMs were found. Despite all of these findings that clearly demonstrate the importance of ALK1 regualted TGF the Alk1 KO lethality makes it not suitable for studying the pathogenesis of HHT. On the other hand, he terozygous Alk1 mice developed HHT like phenotype s (38, 100) ; age dependent subcutaneous and mucocutaneous vascular lesions as well as hemorrhages in the lung, intestine, liver, brain and spleen (100) All of these pathological conditions are st rikingly similar to clinical indications observed in human HHT patients. In that respect, they have been a valuable animal model for elucidating the crucial role of ALK1 for the vascular homeostasis maintenance and the HHT manifestation. Aim and Significan ce of the Study The genes that are responsible for a majority of cases of the inherited vascular disorder HHT had been known for over a decade now. However, the exact pathogenic mechanism(s) of the HHT remains obscure. Treatment options for HHT largely depend on management of symptoms, but the dependable and consistent therapy is currently unavailable. Thus, the primary goal of this study was to determine ALK1mediated endothelial signaling pathway(s) that may be responsible for AVM development seen in HHT. To study ALK1 endothelial signaling during angiogenesis, pulmonary endothelial cells were molecularly and cellulary characterized in response to various angiogenic stimuli using in vivo as well as in vitro assays The results from this study will provide the potential therapeutic target(s) that can prevent or cure vascular lesions in HHT patients. So far, there is no study examining whether de novo ALK1 deficiency causes AVMs in the established vasculature of the adult. Recently, our laboratory generated a transgenic mouse line, using the Cre/LoxP system, in which the Alk1 gene was deleted after tamoxifen treatment. These


25 conditional knockout (cKO) mice developed AVMs in adult vessels undergoing angiogenesis. E mploying this mouse line, the next aim of this study was to investigate a novel physiological effect of AVMs in tumor vasculature in vivo The main purpose of tumor induced angiogenesis is to meet the elevated metabolic needs of tumors as they grow. S ince microcirculation of AVMs is nonproductive due to the lack of capillaries and AV shunts their formation in blood vessel surrounding tumors will starve tumor mass, thereby leading to t umor growth inhibition and/or regression.


26 Figure 11. TGF ransduction through ALK1. The TGF than 30 known ligands that are roughly divid ed into TGF There are seven type I (R I), five type II (R II) a nd two type III (CoR) receptor family members Its si gnal transduction starts with the binding of a ligand to specific type II receptors. The activated type II receptor then recruits and phosphorylates a type I receptor. This R II and R I complex propagates t he signal through intracellular signaling molecule s SMAD s. In general, depending on ligands, the receptor (R) SMAD2/3 pathway mediates TGF activated signaling, whereas BMP ligands utilize the R SMAD 1/5/8 pat hway. At the end of the signal transduction, the complex of R Smad and common (co) SMAD4 enters into the nucleus and regulates expression of many target genes which are involved in various cellular events such as cell proliferat ion, migration, differentiation and apoptosis. The signal is positively regulated by coreceptor binding or negatively controlled by inhibitory (I) SMADs 6 and 7 interaction with the R I and R II complex. A LK1 is one of seven TGF receptors (ALK1 7). It has been reported that among numerous TGF ligands, TGF A, BMP 9 and BMP 10 are putative physiological ligands for ALK1. Additionally, in vitro studies indicated that ALK1 transduce the signal through the SMAD1/5/8 regualted BMP pathway.


27 CHAPTER 2 MOLECULAR AND CELLULAR CHARACTERIZATION OF ALK1 DELETED PULMONARY ENDOTHELIAL CELLS (ECS) Background Overview of Hereditary Hemorrhagic Telangiectasia (HHT) HHT (also known as RenduOsler Weber syndrome) is a genetic vascular disease that is inherited in an autosomal dominant manner and affects 1 in 5,000 10,000 people worldwide (67 69) A person is diagnosed with HHT when he/she possesses at least three of the following four criteria: 1) a related family history 2) epistaxes (spontaneous and recurrent nosebleeds) 3) multiple mucocutaneous telangiectases, and 4) AVMs (arteriovenous malformations) in major viscer al organs including the brain, lung and liver (65) Depending on the different genetic loci encoding HHT causing autosomal genes, HHT is categorized into five different types. Several heterozygous mutations in ENDOGLIN ( ENG on chromosome 9) and Activin receptor Like Kinase 1 ( ACVRL1 ; ALK1 on chromosome 12) genes cause the HHT type 1 (HHT1) (7073) and HHT type 2 (HHT2) (7476) respectively. A subset of HHT patients showing mutations in the SMAD4 gene on chromosome 18 develop a combined syndrome of HHT and Juveni le Polyposis ( JP HHT ) (77) Additionally, two other genetic loci were recently identified and mapped to chromosomes 5 (101) and 7 (102) These contribute to HHT types 3 (HHT3) and 4 (HHT4), re spectively. Interestingly, the most commonly mutated genes ( ENG, ALK1 and SMAD4 ) in HHT encode proteins that are components of the transforming growth factor signal transduction pathway. Therefore, HHT has been considered a disease caused by dysr egulation of TGF Signal Transduction for TGFThe TGFconsists of a large number of secreted pleiotropic cytokine ligands and several receptors (38, 39) There are roughly four different subfamilies containing TGF


28 activin/inhibin, myostatin/ growth and differentiation factors (GDF) and bone morphogenetic proteins (BMP) ligands and three distinctive types of recept or (41, 64) In mammals, seven type I (RI; ALK1 7), five type II (RII; TGF II) and two GLYCAN) transmenbrane receptors have been identified (39, 44) Type I and type II receptors are serine/threonine kinases, whereas type III receptors have short intracellular domains lacking a kinase domain (44) As compared to type II receptors, type I receptors contain a unique glycine/serine (GS) domain followed by the kinase domains ( 39) T he TGF by various combinations among these members and redundant in the association between ligands and type II receptors as well as type II and type I receptors Its signaling affects cellular proliferation, differentiation, migration and apoptosis (39, 42, 43) M any ge netic studies in mice have described the importance of the TGF subfamilies for normal vascular development and maintenance (7, 11, 38) This pathway is also implicated in a variety of physiological processes such as development, embryogenesis, embryonic patte rning, immunological regulation and growth (38, 40, 41) In the resting state of TGF pt ors are present as homodimers on the cell surface (39) I ts signal transduction is initiated by the binding of ligand(s) to specific autophosphorylated type II receptors (38, 39, 44) T he activated type II receptor recruits a type I receptor, forming a heteromeric complex and transphophorylates the GS motif of the type I receptor (39) T he signal is propagated via phosphorylation at the C terminal serines of the cytoplasmic mediators, SMAD proteins, by the type I receptors (39) There are three distinctive types of SMADs, receptorregulated (R)SMAD s (SMAD1, 2, 3, 5 and 8), common (co) SMAD (SMAD4) and inhi bitory (I) SMADs (SMAD6 and 7). R SMAD2 and 3 are TGF specific, whereas R SMAD 1, 5 and 8 are BMP specific (45 48) All


29 signals (phosphorylated R SMADs) by the SMAD dependent pa thway s converge with SMAD4, resulting in heteromeric complexes. Upon their entrance into the nucleus, these complexes are associated with transcription factors, co factors and regulate various target genes (49, 50) Additionally, it is believed that the interaction b etween type II receptors with their ligands are modulated by type III receptors (66, 103) A lso the association of heteromeric receptor complexes with R SMADs are hindered by I SMAD6 and 7, resulting in a negative feedback loop (39) There are also SMAD independent pathways that have been identified to mediate TGF including the MAPK and PI3K pathways (39, 104) Vascular Endothelial Specific ALK1 Signaling ALK1 is one of seven transmembrane type I receptors for the TGF serine/threonine kinase activity (89, 90) Because of the uncertainty of its spec ific ligand(s) and downstream target(s), ALK1 had been considered an orphan receptor (90) B iochemical studies have suggested that ALK1 is a putative physiological receptor for some TGF Activin A (90, 91) TGF and TGF (92) based upon its binding ability. More recently, BMP ligands such as BMP 9 and BMP 10 have been reported to bind to ALK1 and elicit the ALK1 signaling as well (9395) Alk1 is primarily expressed in vascular endothelial cells (ECs) (105107) A dynamic spatiotemporal expression of Alk1 had been described by using Alk1 LacZ knockin (KI) mice (106) Developmental expression of Alk1 was predominant in arterial ECs up to postnatal growth stages, while its expression w as reduced in the adult stage. However, when angiogenesis was initiated by either wounding or tumorigenesis in the adult, its expression reappeared in nascent and remodeling arteries Furthermore, anothe r study reported that in the embryos, Alk1 expression was confined to sites where angiogenesis was active in endothelium (108)


30 Animal Models for HHT The clini cal onset and progression of HHT is highly variable implying a significant role for modifying factors (38) To investigate genetic as well as environmental factors for the pathogenesis of the HHT, numerous groups have generated and studied several genetically modified mouse models (38, 44, 108, 109) Homozygous nul l mutations in the Eng (110 112) and Alk1 (97, 99) genes were embryonic lethal between E10.5 11.5 due to severe vascular defects. S ince heterozygous mice for each gene developed HHT like vascular lesions and HHT is a dominantly inherited genetic disease, they have been valuable animal models for studying the HHT. Eng+/ mice exhibited telangiectases, recurrent nosebleeds and AVMs (111, 113, 114) Alk1+/ mice showed age dependent subcutaneous and mucocutaneous vascular lesions and hemorrhages in organs involving lung, intestine, liver, brain, and spleen (100) This data suggest haploinsufficiency of ENG and ALK1 is responsible for HHT1 and HHT2 manifestations, respectively (66) However, these heterozygous mice also had limitations for investigating the molecular pathogenesis of HHT. They were highly variable in their onset of vascular lesions and HHT like phenotypes Additionally, penetrance varied among mouse strain s (100, 111, 113, 114) To overcome embryonic lethality and these limitations, a conditional knockout (cKO) strategy has been developed by taking advantages of the the Cre/LoxP system producing mouse m odels in which recapitulation of HHT vascular lesions can be created in a desired location predictably and consistently (78, 115) Is ALK1 Signaling Independent from TGFHHT has been considered a TGF However, our most recent in vivo studies in mice and zebrafish (115) indicated tha t neither ALK5 nor TGFBR2 is required for the ALK1 signaling, suggesting that HHT might not be a TGF (115) In this study, cKO alleles for Alk1 Alk5 or Tgfbr2 were deleted specifically in the vascular endothelium


31 using a novel Cre deleter line in which Cre recombinase was predominantly expressed in pulmonary ECs from E13.5. S evere vascular defects included excessive, convoluted, dilated blood vessels and thin vessel walls as well as AVMs, w hich are ver y similar to vascular lesions seen in HHT resulted from the Alk1 deletion. These mice were embryonic lethal between E16.518.5. In contrast, mice with Alk5 or Tgfbr2conditional deletion did not exhibit such abnormal vascular phenotypes. Furt hermore, recent biochemical data suggested that BMP9 might be a physiological ligand for ALK1 (9395) In that respect, the functions of ALK1 signaling might be exerted via the BMP subfamily pathway involving BMP ligand(s) a nd type II receptors ( ActRIIA, ActRIIB, or BMPR2). All of these studies strongly suggest that ECs ar e the primary cell type affected by the ALK1 deficiency and EC specific Alk1 deletion is sufficient for the development of the vascular abnormality. HHT Lik e Vascular Lesions Induced by Alk1 Deletion at the Adult Stages In our most recent preliminary study, we investigate d the role of the ALK1 signaling in the adult vasculature To induce the global Alk1 conditional allele deletion, we employed the ROSA26 Cre ER knockin deleter strain in which Cre recombinase is ubiquitously expr essed and can be activated upon tamoxifen (TM) treatment. Between 9 to 21 days after a single TM injection, the adult mice carrying homozygous Alk1 deletion died. A n autopsy revealed he morrhages in the lungs and gastrointestinal tract, while no other organs s howed such features. To explain why only these organs were affected by the Alk1 deletion, we speculated that ongoing angiogenesis might be required for the formation of abnormal blood vessels in response to Alk1 deletion, based upon our previous observations. Firstly, Alk1 i s persistently expressed in the pulmonary vasculature in the adult stage. B oth pulmonary and gastrointestinal blood vessels showed severe vascular defects upon Alk 1deletion at embr yonic and adult stages. Secondly, expression of the Alk1 gene was reinitiated by woundor tumorigenesis induced angiogenesis


32 during adulthood (106) To test this hypothesis, we performed an in vivo woundinduced angiogenesis assay and found that homozygous Alk1 deletion and proangiogenic cues are required and sufficient for the formation of HHT like vascular abnormalities in the adult mice. Intere stingly, we could not observe these pathological features in Alk1 heterozygous mice with wounding, indicating that homozygous deletion of the Alk1 gene in ECs might be required for development of such vascular lesions in the adult angiogenesis. H ere, we ex amined how Alk1 deletion affected molecular and cellular traits of pulmonary E Cs upon angiogenic challenges. Our next hypothesis was that the ALK1 signaling in ECs controls cellular responses to angiogenic factors during angiogenesis. Thus, when there is a deficiency in ALK1, the appearance of dilated, thin walled, excessive, tortuous blood vessels and formation of AVMs (115) may result due to angiogenic dysregulation in ECs. To test this we isolated and establishe d two new pulmonary EC lines that were either Alk1 heterozygous or homozygous We found that Alk1 deleted ECs exhibited increased migratory and sprouting properties in response to proangiogenic cues both in vitro and in vivo Furthermore, Alk1 null ECs fo rmed an irregular disorganized and enlarged tube like structure. Next, we investigated which ligand and intracellular pathway, TGF would be relevant for such cellular behaviors. In the absence of ALK1, the tubular network was significantly regre ssed by TGF treatment. However, this inhibitory effect was not seen with BMP 9 treatment of Alk1 null pulmonary ECs, indicating that the BMP pathway is more specific for ALK1 signaling. C onsistent with this data, our Western blot analysis suggested that the activation of ERK in Alk1 null pulmonary ECs was greater in BMP 9 treatment than TGF However, the induction of SMAD dependent pathways was unchanged wit h or without ALK1, implying that the cytoplasmic pathway(s) mediating the ALK1 signal ing may be SMADindependent.


33 Results Establishment of Genetically Modified Pulmonary EC Line T he overall experimental scheme is summarized in Figure 2 1 Pulmonary ECs were isolated from the whole lung of a R26+/CreER; Alk12f/1f mouse. The 2f means that ex ons 4 to 6, which encode the transmembrane domain of the Alk1 gene are floxed with two LoxP sequences within the Alk1 cKO allele, while 1f refers the Alk1 nulle allele (Figure 2 3A). Thus, R26+/CreER; Alk12f/1f ECs are equivalent to heterozygous Alk1 null ( Alk1+/ -) ECs. Additionally, in this mouse line, cDNA encoding a Cre ER fusion protein is inserted into the ROSA26 locus ( R26+/CreER). And upon treatment with the tamoxifen (TM), a chemical compound that recognizes and binds specifically mutated estrogen recept or (ER), Cre recom binase becomes active. Therefore, when R26+/CreER; Alk12f/1f ECs are treated with TM, activated Cre recombinase recognizes two LoxP sequences and deletes the Alk1 cKO allele, resulting in homozygous Alk1 null ( Alk 11 f/1f; Alk1/-) EC s lacking the transmembrane region of ALK1 (Figure 2 3A) After isolation and immortalization, ECs were fluorescently and endothelialspecifically labeled with Dio Ac LDL and then sorted by the fluorescent activated cell sorting (FACS) system Typically, t he culture is enriched with more than 95% ECs (Figure 2 2A). Subsequently, endothelial specific properties of sorted ECs were further evaluated by their morphology and desired biomarker gene s expression Morphologically, isolated ECs exhibited the EC characterist ic cobble stone shape and cell c ell contact inhibition at confluency (Figure 2 2B) And the expression of various known EC specific marker genes, such as Flk1 Tie2 Endoglin (Eng) Endothelin (Edn) and Alk 1, w as confirmed by RT PCR analyses. To obt ain more homogeneous populations of ECs, several clones were isolated and further analyzed by methods described above. Based on its typical polygonal shape and high expression of Alk1 and Edn


34 (Figure 2 2 C), clone #28 was chosen and used for further analyse s. Finally, the parental EC line carrying the R26+/CreER; Alk12f/1f transgene w as established. Development of Alk12f/1f ( Alk1 Null Heterozygote)Derived Alk11f/1f ( Alk1 Null Homozygote) ECs To obtain the R26+/CreER; Alk11f/1f ( Alk1 null homozygote) EC line the parental R26+/CreER; Alk 12 f/1f ECs were treated with 1 TM (Figure 2 3A). After a 2 days TM treatment, the 2f Alk1 cKO allele was deleted and thereby undetectable, but only the resulting 1f Alk1 null allele was detected by genomic PCR (Fig ure 23B) and Southern blot (Figure 2 3C) analyses. This genetic modification was maintained in normal endothelial cell medium (ECM ) for at least 7 days without additional TM treatment (Fig ure 2 3B). By culturing ECs in normal ECM for a week, we could avoid any possible side effects from the TM treatment. Furthe rmore, RT PCR analysis showed that Alk1 transcripts were observed in TMuntreated parental cells, but were undetectable in TM treated cells (Fig ure 2 3D). The TM treatment did not affect morphology nor cellular and molecular characteristics of ECs. Hereaft er, R26+/CreER; Alk12f/1f and R26+/CreER; Alk11f/1f are denoted as control ( Alk12f/1f) and mutant ( Alk11f/1f) respectively. Alk1 Null ECs Showed Increased Migratory Index in Response to an Angiogenic Challenge in v itro Migratory property is one of the imp ortant features to focus on when studying the cellular phenotypes of ECs. Two well established mig ration assays are widely used: two dimensional (2D) woundinduced migration (116) and three dimensional (3D) modified Boyden chamber (117) assays. In a previous study, in agreement with our preliminary data from the in vivo woundinduced angiogenesis, we found that without angiogenic stimulation, there was no difference in migration between control (Alk12f/1f) and mutant ( Alk11f/1f) ECs. Thus, basic fibroblast growth factor (bFGF, 50 ng/ml), a well known angiogenic factor, was added into the normal ECM containing 20% fetal bovine serum (FBS). In the 2D woundinduced migration


35 assay, we observed that Alk1 null mutant ECs migrated faster than control ECs (Figure 24A). The rate of cha nge (mean SE; 0.079 mm/hr 0.0016; n=9) in migratory distance of mutant cells was significantly higher than that (mean SE; 0.062 mm/hr 0.0016; n=9) of control cells ( p < 0.0001). T he overall change (mean SE; 0.99 mm 0.024; n=9) in migration of m utant cells for 12 hours was significantly higher than that (mean SE; 0.77 mm 0.024; n=9) of control cells (p < 0.0001) (Figure 24B, C). Consistent with the result s in the 3D migration chamber assay, t he average (mean SE; 29.94 0.14; n=6) number of migrating mutant cells for 48 hours was significantly higher than that (mean SE; 17.33 1.57; n=6) of control cells ( p<0.0002) (Figure 2 4F) However, there i s a possibility that unknown factor(s) in ECM containing 20% FBS may be influencing some of the migratory property of the ECs. Consequently, we repeated both migration assays in chemically defined growth factor and serum free ECM containing bFGF (50 ng/ml). W e obtained the same result s in that mutant cells exhibited a significantly higher rate of change (mean SE; 0.077 mm/hr 0.0015; n=9) and overall change (mean SE; 0.95 mm 0.019; n=9) in migration as compared to control cells (mean SE; 0.062 mm/hr 0.0015 and 0.76 mm 0.019; n=9) ( p < 0.0001) (Figure 2 4D, E). Furthermore, in the 3D modified Boyden chamber assay, a significantly more number of mutant cells (mean SE; 17.29 4.96; n=6) migrated than control cells (mean SE; 7.75 2.87; n=6) ( p < 0.0001 ) (Figure 24G). Alk1 Deficient ECs Resulted in Excessive, Disorganized and Enla rged Tubular Network Formation upon an Angiogenic Challenge in v itro To study endothelial outgrowth or sprouting during the activation phase of angiogenesis in vitro several gel systems such as basement membrane, collagen and fibrin matrices have been developed (117) These contain components of extracellular matrix (ECM) that ar e i mportant for the cellular physiology of ECs (118) We employed Matrigel, a bas ement membrane rich ECM


36 (119) to examine branching morphogenesis of Alk12f/1f control and Alk11f/1f mutant pulmonary ECs. bFGF (50 ng/ml) was added to the culture to determine the ECs behavior during an angiogenic stimulation. After the control (Alk12f/1f) or mutant ( Alk11f/1f) cell suspension s in chemically defined growth factor and serum free ECM were seeded onto the Matrigel, the morphological changes in their sprouting were photographed every 3 hours up to 12 hours and terminated at 24 hours. Overall, the appearance of result ing tube like networks was quite different between control and mutant ECs cultures in response to bFGF treatment (Figure 2 5A). A tubular ne twork formed by control ECs consisted of slowly sprouting, thin and few tubes between 3 and 6 hours (Figure 26A). In contrast, Alk1 null mutant ECs developed an excessive and disorganized tubular network due to fast branching and thick tubes during the same period (Figure 2 6B) In addi tion between 9 to 12 hours (Figure 2 6C) cordlike structure s in the control culture began to regress and many of them disappeared after 24 hours (Figure 26E). However, similar structures formed by mutant cells were resistant to regression during this time period (Figure 2 6D and F). To quantify and statistically an alyze these data, all images from the assay were processed by imaging software. By doing so, we could obtain quantitative readouts such as coverage area of tube like structures on the Matrigel (Figure 2 6A F), total length of these structures (Figure 2 7A and B) and number of sprouting tubes from each nodule (Figure 27D and E) to examine differing cellular phenotypes between control and mutant cells during in vitro induced angiogenesis. S tatistical analyses confirmed that Alk1 null mutant cells formed significantly more and thicker tube like structures indicated by an increased area of coverage at all time points in response to bFGF (Figure 26G). On the other hand, the tubular network developed by


37 control cells showed less area of coverage largely due to thin tube like structures (Figure 2 6G). Although tube like structures in both control and mutant cultures showed significant regression between 12 to 24 hours, Alk12f/1f control cells showed significantly rapid regression compared to their counterpart Alk 1deleted mutant cells (Figure 2 6G). This data indicates that ALK1 deficiency prevented endothelial tubes from regression in an angiogenic condition. In terms of total length (Figure 2 7C) and total number (Figure 2 7F) of sprouting tubes, there was no si gnificant difference between control and mutant ECs by 12 hours. However, mutant ECs showed a considerable increase in both readouts after 24 hours (Figure 2 7C and F), further confirming that ALK1deficiency leads to an enhanced resistance to tubular regression. Antiangioge nic Effect of BMP 9 was Blunted, But That of TGF in Alk1 null ECs B ased on in vitro data, it was believed that ALK1 propagates the TGF TGF and Activin A) signal through the BMP specific SMAD1/5/8 intra cellular pathway in a combination with the TG F However, the existence of an unidentified ligand different from TGF implied (92) In three recent studies, two BMP ligands (BMP 9 and 10) have been reported to specifically bind to ALK1 and activate the SMAD 1/5/8 pathway. One group showed that BMP 9 and BMP 10 induced phosphorylation of SMAD 1/5/8 in microvascular ECs. T hey als o activated expression of genes, such as ID1 ID2 SMAD6 SMAD7 ENG and BMPRII which was previously suggested to be derived by caALK1 (94) And these two factors were anti angiogenic by inhibiting growth and migration of ECs. However, this activation was abrogated when Alk1 and Bmpr 2 were silenced. The same group also found that in human serum only the BMP 9 neutralizing antibody could inhibit serum induced SMAD1/5 phosphorylation (95) Furthermore, from two in vivo angiogenesis assays, it was demonstrated that BMP 9 potently pre vented sprouting angiogenesis.


38 Thus, they argue that BMP 9 might be a circulating specific ligand for ALK1 and drive a physiological effect of ALK1. The conclusion from another group further supported these findings (93) In their study, it was observed that BMP 9 bound to ALK1 and BMPR2 with high affinity in ECs. T hey also found that SMAD1/5 and caALK1responsive genes were activated by BMP 9. Addition ally BMP 9 suppressed proliferation and migration of bFGF stimulated bovine aortic endothelial cells (BAECs) as well as vascular endothelial growth factor (VEGF)derived angiogenesis. Thus, we examined whether these previous data showing an antiangiogenic function of the ALK1 signaling by BMP 9 stimulation would be true in our pulmonary ECs Moreover, in comparison with the TGF activate antiangiogenic effects of ALK1 signaling in these ECs in an angiogenic condition. bFGF (50 ng/ml) was added in combination with either TGF (5 ng/ml) or BMP 9 (20 ng/ml) into chemically defined grow th factor and serum free ECM. Depending on its different concentration, TGF 1 exerts a biphasic effect in a variety of biological processes (11) bF GF or VEGF induced sprouting of bovine microvascular endothelial ce lls was prom oted by 200 to 500 pg/ml of TGF 10 ng/ml of TGF in vitro angiogenesis (11) The concentration of circulating BMP 9 in healthy human sera and plasma has been determined recently (95) The physiological concentration of circulating BMP 9 was varied between 2 and 12 ng/ml and its mean level was 6.2 0.6 ng/ml (95) Here, to study a maximum inhibitory effect of BMP 9 during angiogenesis we examined 20 ng/ml of BMP 9 tha t is much higher than its known biologically active concentration. Endothelial sprouting was phot ographed at various time points (3, 6, 9, 12 and 24 hours) post seeding on the Matrigel (Figure 2 5B and C). C ontrol cells less branched and more


39 regressed over the time in response to TGF (Figure 2 5B) or BMP 9 (Figure 25C) as compared to observations with bFGF alone (Figure 2 5A), indicating that their inhibitory effects on endothelial outgrowth were maintained in the presence of the ALK1. By contrast, Alk1 deleted mutant cells did not appear to be affected by either TGF 5B) or BMP 9 (Fig ure 2 5C) treatments. There was an increase in capillary like network density comparable to that of mutants ECs in response to bFGF alone (Figure 25A). Furthermore, such tube like structures were resistant to regression despite the presence of TGF MP 9 (Figure 2 5A C), indicating increased vessel stability due to the absence of ALK1. Next, all of pictures were processed by imaging software to obtain quantitative parameters, and then these readouts were statistically analyzed. The surface area of tub e formation by Alk12f/1f control ECs was significantly less than that of Alk11f/1f mutant ECs at all time points in response to TGF 8A) or BMP 9 treatments (Figure 2 8B). Moreover, tube like structures in the control rapidly retreated over time (Figure 2 8A and B), indicating their inhibitory effect s were present in control ECs. It is important to note that such antiangiogenic effects led to less surface area but more regression as compared to corresponding cultures with bFGF alone (Figure 2 6G), indicating that ALK1 in pulmonary ECs suppressed uncontrolled endothelial outgr owth upon a proangiogenic cue and its inhibitory effect was further improved by the addition of antiangiogenic factors (TGF 9). In contrast to control cells, Alk1 null mutant cells greatly induced tube like structures despite the presence of TGF 8A) or BMP 9 (Figure 28B) by 3 hours. However, these structures signific antly and rapidly regressed between 3 to 24 hours upon TGF 8A), but slowly regressed and much longer sustained in the presence of BMP 9 (Figure 2 8B). This suggests that without ALK1, the inhibitory effect by BMP 9, and not TGF as blunted, indicating that BMP 9


40 might be a more specific or the relevant physiological ligand for the ALK1 signaling in pulmonary ECs. On the other hand, the total length of such tubular networks (Figure 2 8C and D) and the total number of sprouting from nodules (Figure 2 8E and F) in response to TGF BMP 9 stimulation was not much different during early stages up to 12 hours between Alk1 heterozygous and Alk1 null ECs. However, at 24 hours, Alk1 deleted mutant ECs showed a higher index for these readouts (Figure 2 8C F), furthe r demonstrating that ALK1 deficiency resulted in increased vessel stability. Importantly, unlike results from coverage area readout, we did not observe either TGF or BMP 9specific antiangiogenic effects depending on the presence of ALK1 in sprouting a nd elongating properties (Figure 28C F). Therefore, our data from the in vitro tube formation assay were in agreement with previously reported conclusions (9395) that BMP 9 might be a specific ligand for ALK1 and the BMP induced ALK1 signaling exerted an antiangiogenic role during angiogenesis Alk1 Deletion Caused Higher Migratory Indicati on of ECs and Abnormal Blood Vessel Formation in v ivo Importantly, we found that in vivo Alk1 deletion also led to a higher migratory activity among ECs and abnormal nascent blood vessels Two Alk1 cKO alleles in R26+/CreER; Alk12f/2f mice were deleted by intraperitoneal injection of TM, resulting in adult Alk1 null mutant mice As a control, R26+/+; Alk12f/2f mice were used and also injected with TM. Subsequently, Matrigel containing bFGF (250 ng/ml) was subcutaneously injected into the dorsal region of mi ce. Between 7 to 10 days post injection, mice were sacrificed and Matrigel plugs were collected for further histological analysis. To easily visualize migrating ECs and newly formed vessels into the Matrigel plug, we utilized control and mutant mice contai ning the Flk1lacZKI allele. In these mice, expression of the lacZ gene can be easily detected by X gal staining where the EC specific Flk1 gene is expressed.


41 M acroscopic o bservation of plugs from Alk1 deficient mutant mice revealed several vessels extendi ng from the skin and penetrating into the Matrigel plug (n=5) (Figure 2 9B), while none of the plugs from control mice showed this (n=6) (Figure 2 9 A). Histological analysis demonstrated that migrating cells were found only in the peripheral region of the Matrigel plug in the controls (Figure 2 9C, E, G) However, in the mutants, ECs significantly migrated into the center of the Matrigel plug (Figure 29D, F, H) Moreover, Alk1 null ECs formed several blood vessel structures (Figure 2 9D, F, H), as compared to just a few tiny vessels by control ECs at the edge of the Matrigel plug (Figure 29G). These newly formed blood vessels in the Matrigel plug from muta nt mice were dilated, irregular and disorganized (Figure 2 7D, F, H). Thus, such findings further suggested that Alk1 deficient ECs were more migratory and proangiogenic in response to the angiogenic stimuli. Biochemical Characterization of Alk1Null Pulmonary ECs To determine what intracellular downstream pathway is differentially regulated by ALK1 signa ling in pulmonary ECs, Alk1 null ECs were biochemically studied by Western blot analysis. The canonical TGF signaling consists of two different SMAD dependent intracellular pathways, TGF activated SMAD2/3 and BMP ligands induced SMAD1/5/8 pathways (45 48) TGF 3 was originally believed to be the TGF ligand that activate ALK1 signaling (92) However, recent studies have suggested BMP 9 and BMP 10 to be responsible for inducing ALK1 activation (93 95) In general, ligand activated ALK1 propagates the signal through phosphorylation of SMAD 1/5/8. Here, we examined which ligand, TGF 9, is more speific for the ALK1 signaling by measuring phosphorylation levels of either SMAD2/3 or SMAD1/5/8, respectively, in the presence or absence of ALK1. Furthermore, other SMAD independent signal transduction pathways (104) were stu died to see whether the ALK1 signaling may function in a SMAD independent manner in our ECs.


42 For examination of SMAD dependent TGF overnight serum starvation, both parental Alk1 heterozygous ( Alk12f/1f) and TM treated Alk1 null ( Alk11f/1f) ECs were treated with either TGF 9 (20 ng/ml) for 30 minutes. The total protein levels of SMAD 2 were unchanged by ALK1 and each treatment (Figure 2 10A, B). As expected, in both the Alk 12f/1f and Alk11f/1f ECs SMAD2/3 TGF activated by TGF unchanged by BMP 9 treatment. Interestingly, the basal level of phosphorylated SMAD2/3 was moderately decreased in Alk1 null ECs, indi cating slightly reduced TGF 10A, B). On the other hand, SMAD1/5/8 BMP pathway was markedly elevated by BMP 9, whereas a slight increase was observed by TGF ure 2 10C, D) The total protein level s of SMAD 1 were consta nt with o r without ALK1 and treatments. Surprisingly, despite the absence of ALK1, phosphorylation of SMAD1/5/8 was still considerably elevated by the BMP 9 treatment (Fig ure 210C, D) T his result implied that SMAD 1/5/8mediated BMP pathway may not be a m ajor intracellular pathway of the ALK1 signaling in lung ECs. Overall, signaling components for the canonical TGF Alk12f/1f control and Alk11f/1f mutant pulmonary ECs. Next, we performed similar West ern blot analyses to determine whether a SMAD independent signaling pathway may be mediating the ALK1 signal. There have been numerous studies demonstrating SMAD independent intracellular pathways including MAPK and PI3K pathways for the TGF (39, 104) A recent report proposed that constitutively active (ca) ALK1 negatively re gulated the activation of ERK MAPK induced by human dermal microvascular endothelial cells (HMVEC d) wounding, thereby inhibiting EC migration (120) Based on this finding, we expected that the activation of ERK1/2 MAPK pathway would be


43 increased in respo nse to the absence of ALK1. The total amount of ERK1/2 was same regardless of ALK1 and stimulations (Fig ure 2 10E, F). Unexpectedly, basal and induced phosphoERK1/2 levels were slightly decreased in the absence of ALK1, although not statistically signific ant (Figure 2 10 E, F), indicating that the ERK MAPK pathway may be a ALK1 independent pathway in pulmonary ECs. When these ECs were further stimulated by the pro angiogenic factor bFGF (50 ng/ml) in combination with either TGF 9, similar trends were observed (Figure 2 10 B, D, F). However, in response to angiogenic stimulation, the basal level of both phosphoSMAD2/3 and SMAD1/5/8 was decreased (mean of controls: mean of mutants fo r Smad1; from 0.95: 1.02 to 0.30: 0.39 and for Smad2; from 0.59: 0.44 to 0.21: 0.16). T heir changes in induction were also reduced as compared to those without bFGF treatment (mean of controls: mean of mutants for pSmad1/5/8; from 3.24: 3.06 to 1.10: 1.18 and for pSmad2/3; from 2.20: 2.10 to 1.41: 1.50) However, the activation pattern of ERK1/2 was opposite. In the presence of bFGF, its basal level (mean of controls: mean of mutants for ERK1/2; from 0.30: 0.25 to 0.45: 0.37) and change in induction (mean of controls: mean of mutants for pERK1/2; from 0.23: 0.18 to 0.48: 0.31 with TGF 9) were increased. Overall, there was no obvious difference in SMAD dependent pathways between parental control Alk12f/1f and TM t reated mutant Alk11f/1f pulmonary ECs. The activation of TGF specific SMAD2/3 was greatly induced by TGF 9. On the other hand, the phosphoSMAD1/5/8 was significantly up regulated by BMP 9, while it was slightly increased with TGF 1. Interestingly, in Alk12f/1f control ECs, the phosphorylation of ERK1/2 was unchanged or slightly decreased by TGF 9 treatments (Figure 2 10E, F). However, although not statistically significant, the reduction in the amount of phsophoERK1/2 in re sponse


44 to TGF treatment in Alk11f/1f mutant ECs was greater than that in response to BMP 9 (Figure 210E, F), im plying that the negative function of BMP 9 on the activation of ERK1/2 was reduced in the absence of ALK1 Discussion Despite advances made i n the elucidation of a crucial role for vascular endothelial ALK1 in the establishment and maintenance of vascular integrity, many issues remain unclear. B iological effect(s) and molecular target(s) of ALK1 in vascular ECs still need to be determined. R esu lts have been contradictory because of limitations in the experimental approach. First due to the complexity of the TGF in data depending on endothelial cell type s and culture conditions used. Second, because of an absence of a physiological ligand for ALK1, the constitutively active (ca) ALK1 is mainly used to avoid the simultaneous activation of another TGF However, as the ligand independent activation of ALK1 can induce a broad range of downstream signaling pathway, finding the specific ALK1regualted intracellular sign aling cascade which is likely relevant to the pathogenesis of HHT2 is limited. Based on this background and our previous studies, the rationale for experimental approach presented can be summarized in three points. First, the expression of Alk1 is persiste nt in vascular ECs of the lungs at the adult, while its expression is decreased in most systemic vascular ECs after neonatal periods Furthermore, in mice, the pulmonary vasculature including gastroint estinal (GI) vascular system is affected by Alk1 deletion at both embryonic and adult stages. Second, since primary cells could not be cultured in vitro for a long time with many passages, ECs were immortalized by the SV40 T antigen transfection to avoid variability among the distinct EC batches from different isolation. More importantly, R26+/CreER; Alk12f/1f ( equivalent to Alk1+/ -) and TM treated R26+/CreER; Alk11f/1f ( equivalent to Alk1/-) ECs have the


45 same origin. Therefore, we could reduce variation depending on age and genetic background of different mi c e. Third, unlike using siRNA to block the ALK1 signaling, our genetic approach would result in the complete blockage of the ALK1 signaling. The primary goal of this study was to determine an endothelial specific role for ALK1 signaling during angiogenesis. Based on our previous studies, ALK1 deficiency in vascular ECs is the most likely cause of HHT like vascular lesions during embry onic and adult stages of mice. However, there was no evidence that explains w hy only the pulmonary and gastrointestinal vasculatures are consistently affected by the ALK1 deficiency in the adult mice. O ur previous study hinted that diminished ALK1 expression during postnatal development was reactivated by active angiogenesis (106) Thus, we tested the relationship between Alk1 deletion and ongoing angiogenesis for the development of HH T like vascular abnormalities. Our in vivo woundinduced angiogenesis assay showed both factors were necessary and sufficient for the formation of abnormal blood vessels that were disorganized, irregular, ex cessive, tortuous and AV shunted. Such results led us to investigate how ALK1 deficiency affects cellular and biochemical properties of vascular ECs during angiogenesis. Another emerging issue was that ALK1 may not propagate signals in a TGF pathwaydependent manner, therefore, HHT might not be a TGF Three recent in vitro reports demonstrated that ALK1 specifically bound to BMP 9 and B MP 10 in combination with BMPR2 T his association induced activation of the BMP specific SMAD1/5/8 pathway and resulted in inhibition of proliferation and migration of ECs as well as suppression of sprouting angiogenesis (9395) This finding was further strengthened by our in vivo data suggesting that HHT like pathological vascular features were caused by the endothelial Alk1 deficiency, but not by Alk5 or Tgfbr2deletion in vascular ECs (115) This raised the possibility that an other TGF


46 superfamily signal transduction, most likely BMP ligands implicated pathway, may be relevant for the ALK1 signaling, rather than the TGF induced signaling pathway. Although there has been no in vivo evidence showing that deletion of B MP 9 and/or BMP 10 cause HHT like phenotypes, a recent study suggested that RNA interferencemediated silencing of Bmpr2 expression caused vascular abnormality including mucosal bleedings and the lack of mural cells on vessel walls (121) Thus, they conferred a possible link between the perturbation of BMP pathway induced by BMP ligands through BMPR2 and HHT pathological vessel lesions. W e initially examined endothelial specific cellular phenotypes such as proliferation, migration and tube formation of Alk1 null lung ECs without proangiogenic cues. C onsistent with our in vivo data, there was no difference in their phenotypes for migration and tube formation in the absence of an angiogenic factor even in the media containing 20% FBS. However, we observed that during bFGF induced in vitro angiogenesis, Alk1 deficient pulmonary ECs exhibited signifi cantly higher migratory properties than Alk1 heterozygous ECs in both 2D woundinduced and 3D modified B oyden chamber assays. When these ECs were seeded onto the Matrigel, they organized into a tube like structure in response to angiogenic bFGF. Alk1 null ECs rapidly developed an excessive, irregular and thick tube like network. However Alk1 hetero zygous ECs contained less, long and thin cords. Moreover, the cord like structure s in the mutant cell culture was sustained much lo nger than those of control cell culture. Although we could not see any TGF or BM P 9specific difference in migration properties between mutant and control cells, we observed BMP 9mediated antiangiogenic effects were blunted in Alk1 deleted ECs in the tube formation assay, indicating higher specificity of BMP 9 over TGF 1 for the ALK1 signaling. Further quantitative in vivo assay using subcutaneous implantation of Matrigel containing bFGF also demonstrated the elevated migratory and pro-


47 angiogenic character istics of Alk1 deleted ECs. However, a drawback in our cell lines i s that we could not see a significant difference in proliferation rates between Alk1 heterozygous and homozygous ECs since they were immortalized It has been controversy regarding the role of ALK1 on ECs during angiogenesis. Previously, our group found that in Alk1 KO embryos, a number of genes that act in the activation phase of angiogenesis showed an incr ease in their mRNA levels. Thus, we suggested that ALK1 signaling was implicated in the resolution phase of angiogenesis and functioned to inhibit EC proliferation and migration as well as promoting VSMC recruitment (97) A nother group su pported our findings and reported that target genes of caALK1 inhibit proliferation and migration of human microvascular endothelial cells from the dermis (HMVEC d), implying the role of ALK1 in the maturation phase of angiogenesis (122) Furthermore, it was demonstrated that interruption of the acvrl1 gene increased endothelial cell number in cranial vessels of zebrafish (123) However, others suggested that ALK1 signaling enhance s proliferation and migration of E Cs, proposing ALK1 involvement i n the activation phase of angiogenesis (98, 107, 124) Our results from this study further su pport a significa nt role of endothelial specific ALK1 signaling for the resolution phase of angiogenesis. Alk1 deficient ECs resulted in increased migr ation, excessive tube formation and resistance for regression, f eatures of the activation phase of angiogenesis. Therefore in ECs without ALK1, the transition of ECs from activation to resolution is perturbed and uncontrolled angiogenesis lead to the formation of an abnormal vascular network. Taken together, our findings at the cellular level provide a hint for the pathogeni c mechanism underlying HHT like vascular lesions containing dilated, thin walled, excessive and


48 tortuous blood vessels and AVMs. Mutations in the Alk1 gene result in a decrease in ALK1 expression, causing perturbation in the functions of ECs during blood vessel development and remodeling. The Alk1 deficiency in ECs lead s to persistent activation phase of angiogenesis, thereby superfluous blood vessels formed. Consequently, without proliferation termination and maturation, the vessels develop enlarged lumen and thin vessel walls In addition, it was thought that dilated and fused blood vessels surrounding AVMs mainly resulted from hemodynamic forces d uring the pathogenesis of HHT. However, in our in vitro data from the tube formation assay, peculiar thick and fuse d capillary like cords or tube developed solely by Alk1 deletion in ECs. These fragile blood vessels are prone to rupturing, causing recurrent bleedings from telangiectasia (focal dilation of blood vessels) with in the mucocutaneous layers of the sk in including lips, tongue, nose and finger tips of HHT patients. Moreover, due to the lack of a switch to the resolution phase of angiogenesis, the telangiectasias can worsen and the vessels can progress into AVMs AV shunts may develop from AVMs, leading to dilated and tortuous veins as well as fusion of ar terial and venous blood flow. It is noteworthy that perturbation of Notch/Delta signaling, another known important angiogenic signaling pathway, also developed AVMs in mice (125, 126) Thi s pathway has been implicated in arteriovenous specification (14, 16, 127, 128) Next we searched which signaling pathway(s ) may be responsible for these cellular phenotypes. The purpose of examining the mechanism would be to find potential therapeutic target(s) that could be designed to prevent or alleviate clinical symptoms of the HHT. The TGFsignaling pathway is diverse and complex, and proceeds in numerous SMAD dependent and independent pathways (39) Two majo r SMADdependent pathways are stimulated by the ligand binding to the type II and Type I receptor complex (39, 44) (Figur e 2 11A). T he TGF ligand -


49 mediated pathway propagates signals via receptor regulated (R) SMAD2 and 3. These are phosphorylated by the structurally similar type I receptors ALK4, 5 and 7 (129132) In contrast, R SMAD1, 5 and 8 are specific for the BMP pathway and are induced by ALK1, 2, 3 and 6 (133) We found that the basic components for these two major SMAD pathways were present in both parental control and TM treated mutant ECs based on our findings that phosphorylation of SMAD2/3 or SMAD1/5/8 were enhanced with TGF 9 treatment, respecti vely. Interestingly, the basal level of phosphoSMAD2/3 was slightly decreased in Alk1 null ECs with or without bFGF, suggesting that the inhibitory effects of TGF in the absence of AL K1; thus Alk1 deleted ECs were more migratory and proangiogenic in response to an angiogenic cue. The most interesting finding was that BMP pathway was intact despite the absence of ALK1. In RT PCR analysis, the presence of BMP pathway implicated type II receptors including activin type II receptors (ActRII) and BMPR2 as well as other type I receptors such as ALK2, 3 and 6, was revealed, suggesting that the SMAD1/5/8regualted BMP pathway was compensated by other receptors in the ALK1 deficient mutant pulmonary ECs (Figure 2 11 B). This implied that distinctive cellular phenotypes of Alk1 null ECs might be resulted from unknown ALK1specific downstream pathway(s) rather than SMAD dependent signal transduction cascades. M any studies have demonstrated that T GF independently MAPK and PI3K pathways (39, 104) It was found that the activation of JNK MAPK (134) and p38 MAPK (135) was observed in SMADdefective cells with TGF treatment. A dditionally, it was reported that TGF induced signaling lead to phosphorylation of an intracellular component (Akt) of the PI3K pathway (136, 137) Among MAPK pathways incl uding JNK, p38, and ERK, we focused on the ERK1/2 MAPK pathway. Previously, it was


50 suggested that the activated ERK signaling by TGF phosphorylation and modulated SMAD activation (134, 138140) More importantly, a recent study showed a negative effect of caALK1 on the activation of ERK1/2 in human dermal microvascular ECs, resulting in inhibition of EC migration in response to wounding (120) However, result s from our Western blot analysis contradicted conclusions. Although it was not statistically significant, the basal phosphorylation level of ERK1/2 was de creased in ALK1 deficient ECs. The induction of ERK1/2 in response to stimulation by either TGF 9 with or without pro angiogenic bFGF was even furthe r reduced in these ECs. Since activated ERK induces the SMAD phosphorylation and subsequent activation, it indicated that TGF ht be more inhibited upon sti mulation in ALK1 defiecnt ECs. Thus, its negative effects on ECs were more repressed, leading to further elevated proangiogenic propert ies. Notably, although the difference was not significant, in the absence of ALK1, the reduction in phosphorylated ERK1/2 was greater by TGF 1 than BMP 9 treatments. Therefore, it appears that BMP 9induced inhibitory effects on the ERK MAPK pathway were reduced in Alk1 null ECs In other words, BMP 9 may be more dependent on the ALK1 signaling than TGF ECs. However, we could not rule out the possibility that immortalization by the SV40 T antigen transfection could affect many signaling pathways in our ECs. In summary, we established a pulmonary EC line containing a genetical ly modified Alk1 allele from the R26+/CreER; Alk12f/1f mouse. After a 2 day TM treatment, we could obtain Alk1 nul l ECs from these parental ECs. As compared to Alk12f/1f control ECs, the peculiar cellular phenotypes of Alk1 null ECs were increased migratio n and endothelial outgrowth or sprouting in response to the proangiogenic challenge, suggesting that the Alk1 deletion resulted in elevated angiogenic responses in lung ECs Additionally, disorganized and dilated tube like structure s


51 developed exclusively due to Alk1 deficiency in ECs. Furthermore, this in vitro tube formation assay suggests that ALK1 may be more specific or the pertinent receptor for BMP 9, rather than TGF However, since we used a much higher concentration of BMP 9 than its known phys iological concentration (approximately 5 ng/ml), it will be necessary to perform the same experiments with this concentration. Although we could not find a ALK1specific intracellular downstream pathway, our results from Western blotting demonstrated that ALK1 signaling in ECs may be SMAD independent and BMP 9 may be the more rele vant ligand for its signaling. There are several other path ways such as JNK MAPK, p38 MAPK and PI3K which has been suggested to be regulated by the TGF Most recently, a n in vivo study suggested functional redundancy between p38 and Smad4, the intracellular effector transducing the signal from all SMAD dependent pathways into the nucleus, in mediating TGF during t ooth and palate development (141) Therefore, it will be interesting to investigate how the p38 MAPK as well as JNK MAPK and PI3K pathways are differentially modulated in our in vitro systems. Moreover, we have developed another parental EC line carrying the R26+/CreER; Alk12f/2f transgene Using this EC line, we can compare cellular and molecular phenotypes of Alk12f/2f equivalent to Alk1 wildtype and Alk11f/1f equivalent to homozygote pulmonary ECs. The findings from this study will further broaden our insight into the EC specific ALK1 signaling.


52 Figure 21. Experimental scheme for cellular and molecular characterization of Alk1 heterozygous ( Alk12 f/1f) and Alk1 homozygous ( Alk11f/1f) pulmonary ECs. Parental pulmonary ECs w ere isolated from the R26+/CreER; Alk12f/1f mouse and immortalized by SV40T antigen tranfection. A pure EC population was obtained by FACS. Endothelial properties were further examined by their morphology and EC specific marker genes expression. Based on the desired characteristics, clone #28 was selected. Subsequently, Alk1 null ECs were derived from the parental cell line with a 2day TM treatment. After another FACS, both Alk12f/1f control and Alk11f/1f mutant EC lines were established. Finally, they w ere cellularly and molecularly investigated.


53 Figure 22. Establishment of R26+/CreER; Alk12f/1f ( equivalent to Alk1 null heterozygote; Alk1+/ -) parental pulmonary endothelial c ell l ine. A) After DioAcLDL labeled FACS, representative data showed the cul ture containing about 98% purity for the EC population. Note that TM treated R26+/CreER; Alk11f/1f ( equivalent to Alk1 null homozygote; Alk1/-) ECs showed the similar pure population, indicating that TM did not affect endothelial properties of ECs. B) In m orphological evaluation, isolated and immortalized ECs displayed endothelial specific polygonal shapes and cell cell contact inhibition at confluency. C) Amongst several isolated clones, the clone #28 showed the highest Alk1 and Endothelin (Edn) expressio n at the RNA level. Expression for EC specific marker genes including Flk1, Tie2, and Eng was further confirmed by the RT PCR analysis. The expression of Smad1 was examined to see whether this clone contained its representative downstream molecule. Actin w as used as the loading control.


54 Figure 23. Derivation of R26+/CreER; Alk11f/1f ( equivalent to Alk1 null h omozygote; Alk1/-) ECs from parental EC l ine. A) Schematic diagram of the Alk1 wild type, Alk12loxP (2f) and Alk11loxP (1f) alleles. Exons a nd loxP sequences are denoted by boxes and arrowheads, respecti vely. L ocations of primer pairs detecting specific regions containing loxP sequences for the genomic PCR analysis and sites of probes recognizing the 5 region of the Alk1 gene which was used f or the genomic Southern blot analysis are indicated. Note tha t since exon 5 encodes the transmembrane domain (TD) of the Alk1 gene, the deletion of exons 4 to 6 results in a null allele. B) After a 2day 1 TM treatment, the 2f Alk1 cKO al lele was undetectable due to Cre recombinasemediated deletion. T he resulting 1f Alk1 null allele was maintained for 7 days without further TM addition. Note that for genomic DNA PCR analysis genomic DNAs were collected at 2 days post TM treatment and everyday for 7 days after transferring cells in normal ECM. C) Genetic modification of Alk1 null homozygote by a 2day TM treatment was confirmed by the genomic Southern blotting. D) RT PCR analysis verified the deletion of the Alk1 gene at the RNA level. Alk1 transcripts were undetectable in the TMtreated Alk11f/1f ECs. Negative control indicates no RT PCR reaction Note that PCR primers used for RT PCR analysis detect exons 4 and 5 of the Alk1 gene.


55 Figure 24. Alk1 null pulmonary ECs displayed elevated migratory index upon angiogenic factor challenge. A) After control and mutant ECs wounding, the closing of wounds in response to bFGF (50 ng/ml) stimulation was photographed every 4 hours. Mutant cells migrated much faster compared to control cells. At 12 hours post wounding, mutant cells almost completely closed the wound, while the wound was still present in the control culture. This observation was confirmed by the statistical analysis. B) and C) In the 2D wound induced migration assay, Alk11f/1f mutant ECs were significantly more migratory than Alk12f/1f control ECs in ECM containing 20% FBS. D) and E) The same results were obtained in chemically defined growth factor and serum free ECM, indicating no FBS eff ects on the data in B and C. F) and G) In the 3D modified Boyden chamber assay, 48 hours after seeding (5 102 cells), significantly more number of migrating Alk11f/1f mutant ECs were counted in six randomly chosen fields as compared to Alk12f/1f control ECs F) with 20% FBS and G) without FBS. Note that all data represent means from three independent experiments. Error bars show standard errors.


56 Figure 24. Continued.


57 Figure 25. in v itro Matrigel tube formation a ssay. Alk1 heterozygous ( Alk12f/1f) control and Alk1 homozygous ( Alk11f/1f) mutant pulmonary ECs (1 105) were diluted into chemically defined growth factor and serum free ECM and seeded onto the Matrigel. To monitor their cellular traits in the tube formation dur ing in vitro angiogenesis, pictures were taken every 3 hours (50X ; height width = 2.1 mm 2.8 mm ) A ) Tube like structures were formed by control or mutant cells in response to bFGF (50 ng/ml). B) In addition to bFGF, one known putative physiological li gand for ALK1, TGF plemented with culture medium. C) A nothe r possible ligand for ALK1, BMP 9 (20 ng/ml) was added into medium containing bFGF. Note that all pictures are representatives of three independent experiments. Alk1 deleted ECs di sp layed an irregular, excessive and dilated capillary like network formation as compared to control cultures


58 Figure 26. Alk1 null ECs formed thicker tube like structures that are resistant to regression. A), C) and E) Original images of the control cu ltures at 6, 12, 24 hours after seeding (50X). All of original images were processed by the MatLab imaging program. Matrigel background is shown as black, while tube like structures are presented as light gray. The processed images were used to calculate coverage area of capillary like structures. The red solid line delineated boundaries of these structures. B), D) and F) Corresponding original, processed and coverage area measured images of the mutant culture. G) Coverage areas of capillary like structures formed by control or mutant cells in response to bFGF (50 ng/ml) were calculated as percentages of the whole field. Results from statistical analysis are shown. Note that pictures are representatives of three independent cultures. All data are means of th ree distinctive experiments. Error bars indicate SEM ( + p<0.05, ++ p<0.01 and +++ p<0.001).


59 Figure 26. Continued.


60 Figure 27. ALK1 deficiency resulted in an increase in length of tubes and sprouting of ECs. A) Proce ssed images of control cells were used to calculate the total length of tubes Tubular lines are shown in green solid lines. B) Corresponding images of mutant cells were used to measure the total length of tubes. C) Total length of endothelial tubes of the control and mutant were m easured and statistically analyzed. D and E) Control and mutant ECs sprouts or outgrowth from nodules are represented as red dots on tubular networks which are shown in A and B, respectively. F) The number of endothelial sprouting from each nodule was counted. The difference in sprouting properties of control and mutant cells was examined by comparing total num ber of their outgrowth per field and statistically evaluated. Note that images are representatives of three distinctive cultures. All data are means of three independent experim ents. Error bars indicate SEM (+ p<0.05, ++ p<0.01 and +++ p<0.001).


61 Figure 27. Continued.


62 Figure 28. Inhibitory effect of BMP 9 on angiogenesis was diminished whereas that of TGF in Alk1 null ECs Pictures of control and mutant ECs cultures with bFGF (50 ng/ml) in combination with either A), C) and E) TGF F) BMP 9 (20 ng/ml) taken at various time points were processed as described in Fi gures 6 and 7. Each parameter was measured by using processed images. A) and B) Surface area of capillary like networks from the control and mutant was calculated and statistically analyzed. C) and D) The total length of these networks was statistically ev aluated. E) and F) The total number of EC sprouting from nodules were counted and statistically analyzed. Note that all of calculations were performed as described in Figures 6 and 7. I mages are representatives of three distinctive cultures. All data are m eans of three independent experiments. Error bars indicate SEM (+ p<0.05, ++ p<0.01 and +++ p<0.001).


63 Figure 28. Continued.


64 Figure 29. in v ivo Alk1 deletion also resulted in higher migratory and proa ngiogenic properties of ECs. Whole mount Matrigel plugs from A) and G) control R26+/+; Alk12f/2f; Flk1+/lacZ and B) and H) mutant R26+/CreER; Alk12f/2f; Flk1+/lacZ mice were stained with X gal Representative histological sections of Matrigel plugs in the TMtreated control and mutant skin are shown in C, E G and D, F, H, respectively. A) M acroscopic examination revealed all vessels were formed between the skin layer and Matrigel plug in the c ontrol (n=6). B) However, in the mutants, several nascent vessels invaded t he Matrigel plug (n=5). Newly formed vessels are indicated by black arrows. In H & E staining, C) migrating control cells were confined to the edges of Matrigel plugs whereas D) migrating mutant cells were found much more broadly. E) and F) Similar findin gs in C and D were observed by trichrome staining, which showed the Mat rigel in blue. D) and F) Several disorganized blood vessel structures, denoted by red arrowheads, were only shown in Matrigels from the mutants. G) N uclear fast red (NFR) staining follo wing whole mount X gal staining showed the appearance of few and small nascent blood vessels in the controls H) Meanwhile, mutants displayed enlarged lumen and irregular shapes. Red arrows indicat e X gal positive vascular ECs. C ) H) Margins of Matrigel plugs are represented by blue arrows.


65 Figure 210. ALK1 signaling in pulmona ry ECs was not s pecific for SMAD dependent nor the ERK MAPK pathways After overnight serum starvation, both Alk12f/1f and Alk11f/1f ECs were incubated with either TGF ml) or BMP 9 (20 ng/ml) in the absence A ) C ) and E) or presence B ) D ) and F) of bFGF (50 ng/ml) for 30 minutes. P rotein lysates were resolved on a 8% SDS polyacrylamide gel, transferred on a blotting membrane and then immunoblotted with various antibodi es. The level of induced phosphorylation was calculated by the ratio of phosphorylated proteins to total proteins. A) and B) Activation of TGF specific SMAD2/3 pathway was detected by phosphorylatedSMAD2/3. C) and D) Stimulation of BMP 9specific SMAD1/5/8 pathway was revealed by phosphorylated SMAD1/5/8. E) and F) Induction of ERK1/2 MAPK pathway was shown by its phosphorylation. Note that all Western blots are representatives of three independent experiments. GAPDH was used as a loading control. All total and phosphorylated protein amounts were normalized to the amount of GAPDH. After normalization, the level of induced phosphorylation was calculated. Data in all graphs represent means of values measured by densitometry from three separa te blots. Error bars are SEM (+ p<0.05, ++ p<0.01 and +++ p<0.001).


66 Figure 210. Continued.


67 Figure 211. BMP specific SMAD1 /5/8 pathway for ALK1 signaling was compensated by other TGF ulmonary ECs. A) Diversity and complexity in TGF superfamily signaling through combinations among its type II, type I receptors and SMADs Black arrows in solid line show a traditional view for the ALK1 signaling. The new insight in EC specific ALK1 signal transduction mediated by BMP specific pathway is indicated by the red dotted arrows. B) RT PCR analysis confirmed the presence of compensatory pathways for the BMP pathw ay by the existence of TGF type II (ActRIIa and Bmpr2) and type I receptors (Alk 2, 3 and 6). Note that total RNAs were extracted from two different cultures of each EC line. Negative control indicated the absence of RT reaction Actin was used as a control showing that the same amount s of RNAs from each culture were used for the RT reaction.


68 CHAPTER 3 NOVEL PHYSIOLOGICAL EFFECT(S) OF ARTERIOVENOUS MALFORMATIONS (AVMS) ON TUMOR VASCULATURE Background Tumor Angiogenesis In solid tumor biology, new blood vessel formation during tumor growth, also known as tumor angiogenesis, is an important proce ss not only for understanding tumorigenesis but also designing anti tumor therapies The progression and metastasis of various solid tumors largely depend on their own vascular network. To support actively proliferating cells, tumors require their own blood supply that provides nutrients and oxygen as well as removes metabolic wastes (142) Therefore, as they grow beyond limited size, tumors turn on an angiogenic s witch (142, 143) In contrast to physiological angiogenesis in which the balance between proand anti angiogenic factors is strictly regulated and nascent vessels become rapidly stabilized, tumor induced pathological angiogenes is causes locally unbalanced proangiogenic molecules to be overproduced and as a consequence new vessels are steadily formed (143) The constantly in creased vascularity without stabilization leads to abnormal tumo r vasculature in which blood vessels are irregular in shape, dilated and sinuous (143, 144) Furthermore, these tumor blood vessels exhibit disorganization of arterio les, venules and capillaries, are very leaky and hemorrhagic (143) Blood flow within the tumor vasculature is nei ther even nor unidirectional (144) In a recent review, it is argued th at despite many differences between tumor and normal blood vessels there has been little attention on this matter (144) T he tumor vasculature is categorized into six distinct subty pes during tumor angiogenesis. Based on its developing order, structure and function, the six different types are mother vessels (MV), capillaries, glomeruloid microvascular proliferations (GMP), vascular malformat ions (VM), feeder arteries (FA) and draining veins (DV).


69 Antiangiogenic Strategies Targeting Cancers In the early 1970s, the concept of antiangiogenic therapy to treat solid tumors arose b ased on the necessity of angiogenesis for tumor growth, thus, inhibition of angiogenesis would be antitumorigenic by impeding progression of tumors (145) Since then, i nnumerable studies have been conducted to understand the underlying molecular mechanisms of angiogenesis in efforts to find positive and negative angiogenic factors. T he vascular endothelial growth factor (VEGF) receptors and ligands family is the most cha racterized activator. Researchers have learned much about the central role of the VEGF family in developmental, physiological, and pathological angiogenes is. The inhibition of the VEGF tyrosine kinase signaling pathway is the most tested and substantiated angiogenesis based anticancer strategy (146 148) Th e first FDA approved antiangiogenic drug in 2004 was a humanized anti VEGF monoclonal antibody (bevacizumab) for the treatment of metastatic colorectal cancer in combination with chemotherapy. This has been followed by various approaches designed to block the VEGF pathway including anti VEGF receptor monoclonal antibodies, chimeric soluble VEGF receptors (VEGF trap), and VEGF receptor tyrosine kinase inhibitors (TKIs) (23, 149, 150) Lessons from Numerous VEGF/VEGFR Inhibitor Based Clinical Trials A great number of preclinical and clini cal studies have proved that angiogenesis is an important therapeutic target for several types of solid tumors. In some cases, such advances provided new insight about the clinical application of anti VEGF therapy. It was or iginally anticipated that blockade of tumor angiogenesis would inhibit the blood supply to tumors, thereby starving and cause them to shrink. However, based on clinical trials, the antitumor effects of anti VEGF strategy are most likely due to the normaliz ation of the tumor vasculature It appears that suppression of VEGF signaling causes trimming of the abnormal tumor vasculature followed by r emodeling of the remaining vasculature (23, 149, 151) T he tumor


70 microenvironment becomes more accessible to efficient delivery of chemotherapeutic agents and oxygenated, heightening sensitivity to ra diation therapy. Thus the anti VEGF based monotherapy was not as successful as expected in many clinical studies however, its antitumor efficacy was considerably improved when used in combination with a conventional therapy such as chemotherapy and radiation. Like other cancer drugs, concerns a bout resistance to antiangiogenic therapies have emerged from some preclin ical and clinical trials Recurrent tumor growth was reported a fter initial tumor regression during long term treatment (152, 153) It was reported that established tumor blood vessels became resistant to anti VEGF mediated angiogenesis suppression due to the heterogeneity of tumor vasculature (154, 155) and made them intensely aggressive (143) Consequently, these suggest that angiogenic pathways in the tumor vasculature are very complicated and can not be clearly explained by any one given pathway. Thus, res earchers and clinicians realize that the most successful strategies must be combination t herapy targeting multiple pro angiogenic factors as well as different signaling pathways (23, 143, 148, 149, 151) Emerging Novel Antiangiogenic Targets Many genetic studies in mice have suggested that normal embryonic vascular development depends not only on the VEGF signaling pathway, but other signaling pathways as well. Since embryonic and adult (physiologica l and pathological) angiogenesis uses similar mechanism s it seems logical that these other signaling pathways may also play important roles during tumor angiogenesis. M embers of fibr oblast growth factor (FGF) family are commonly studied proangiogenic proteins that are seen as potential antiangiogneic therapeutic targets (35, 156) Recently, the implication of axonal guidance receptors and ligands in angiogenesis has been emphasized. Examples include the Roundabout receptor (Robo4)/Slit ligand, Ephrin receptor/Ephrin ligand (EphA2 or EphB4), Unc5 receptor (Unc5B)/Netrin ligand (Netrin1) and


71 Notch receptor/Delta like ligand 4 (Dll4) families (13, 51) They were initially identified as important molecules in neuronal guidance during development (13, 51, 148) but there i s evidence that dysregulation of these may be involved in tumor angiogenesis I nhibition of Robo4 (157) the ephrins (158) Dll4 (58, 159) and activation of Unc5B receptor le a ds to a decrease in the experimental tumor growth and angiogenesis. Interestingly the results from studies in which Dll4 induced Notch signaling was blocked suggested a new concept f or the angiogenesis based tumor therapy (58, 159) It was found tha t when Dll4 was activated by VEGF signaling, it acted as a negative regulator to block VEGF induced tumor angiogenesis (58) Thus, there i s an inverse relationship between the Dll4/Notch activity and tumor vascularity. This was consistent with the previous conception of anti VEGF therapy that in order to induce antitumor effects, tumor angiogenesis mus t be blocked. However, it was shown that blocking Dll4 resulted in i ncreased tumor vascular density consequently, reducing tumor growth (58, 159) Such an enigma was explained by the notion that unregulated angiogenesis due to blockade of Dll4 led to the formation of nonproductive or functional blood vessels in the tumor vasculature (58, 159) Furthermore, th is approach was also antitumorigenic even in anti VEGF therapy resistant tumors (58) These findings suggest a new angiogenesis based therapeutic approach to tumor therapy in which induction of disorganized and nonproductive blood vessels in the tumor vasculature would be an excellent alternative of anti VEGF mediated tumor therapies. A Novel Anti Tumor Effect of Alk1 Deletion Induced AVMs on Tumor Vasculature Activin receptor like kinase 1 (ALK1) is one of the type I receptor members from the TGF (89, 90) It is a transmembrane receptor containing a serine/threonine kinase activity. ALK 1 is primarily expressed in vascular endothelial cells (ECs) (105107) During murine embryonic and neonatal stages, it is predomina ntly expressed in arterial ECs, but its


72 expression i s diminished in most blood vessels during a dulthood. E xpression of Alk1 i s continuous in the lungs and initiated by woundingor tumorigenesis induced angiogenesis in adult s (106) There have been numerous studies showing that ALK1 plays an important role in normal embryonic vascular development and postnatal vascul ar maintenance and remodeling. Homozygous germline null mutation of the Alk1 gene was embryonic lethal between E10.511.5 due to severe vascular abnormalities (97) In humans, heterozygous mutations cause hereditary hemorrhagic telangiectasia (HHT), a dominantly inherited vascular disease Major symptoms of HHT patients are epistaxis (spontaneous and recurrent nosebleeds), mucocutaneous tela ngiectases (focal dilation of blood vessels) and arteriovenous malformations (AVMs) in internal organs including the brain, lung, liver and intestine, which result from the fragility of disorganized peripheral microvessels and direct connections between arteries and veins without proper capillary beds (38, 44, 65, 66) HHT is a unique vascular disease in which mechanisms of adult vascular maintenance and pathological or physiological functions of AVMs can be investigated. Recently, we developed a new conditional knockout (cKO) mouse model using the Cre/LoxP system specifically utilizing the CreER system In this model global homozygous Alk1 deletion could be induced by tamoxifen (TM) treatment at any given adult stage. W e found t hat upon homozygous Alk1 deletion, blood vessels that formed in the wounding area were disorganized, dilated, tortuous and showed AVMs. Furtherm ore, these vascular lesions developed only in site s where both Alk1 was deleted and ongoing angioge nesis was present. Another previous study described Alk1 deficient vascula r endothelial cells formed excessive, irregular, and enlarged abnormal vascular structures in the Matrigel plug.


73 The formation of AVMs can be problematic because 1) the lack of capillaries prevents the proper exchange of appropriate nutrients and removal of wastes between the blood and surrounding tissue normally required, and 2) the dilated vessel walls are fragile and prone to rupturing. Based on such respects, we speculated that inducing these types of abnormal blood vessels within the tumor feeding vasc ulature would result in insufficient supplies for tumor growth, thereby inhibiting their progression. To test this possibility, we injected Lewis lung carcinoma (LLC) cells into TM treated control and mutant mice and examined how Alk1 deletion affects tumo rigenesis. First in subcutaneous tumor cell injections after a TM treatment, none of Alk1 deleted mutant mice developed tumors, whe reas solid tumors formed under the skin of all control mice. Seco nd, the growth of established solid tumors in the thigh of mice was significantly inhibited in the mutant mice after TM injection, as compared to those in the control mice. By latex dye injection into the systemic circulation of the mice, we found the formation of AVMs in the tu mor feeder peripheral vessels. F urth er histological analyses revealed that tu mors from the mutant mice displayed disrupted blood vessel surrounded by many necrot ic tumor cells. Thus, creation of abnormal blood vessels during tumor angiogenesis resulted in the prevention of initiation and progression of tumorigenesis. This was most likely due to the lack of capillaries, AV shunts within AVMs and disrupt ion of tumor feeding blood vessels, suggesting that inhibition of the ALK1 signaling might be a new angiogenesis based therapeutic approach to target the tumor vasculature. Results Initiation of Tumorigenesis Was Suppressed in Alk1 Deleted Adult Mutant Mice Hyperactive tumor angiogenesis is largely responsible for the growth of solid tumors. Thus, considerable angiogenesis is initiated in the tiss ues bearing tumor masses. For the appropriate functions of a microvascular system, the local vascular network should be highly


74 organiz ed into arterioles, capillaries and venules. B lood vessels with in AVM lesions are disorganized, fragile and lack capillar ies, eventually developing AV shunts. Consequently, arterial blood flows directly into veins, bypassing opportunities for feeding to take place. Based on such facts and our preliminary data, we hypothesized that in conditions of tumor induced angiogenesis blocking or deleting ALK1 signaling would lead to the development of abnormal blood vessels that would not adequately feed and allow progression of a tumor. To test this hypothesis, we employed the same mouse model which was used in the preliminary study. In this model, to circumvent lethality of homozygous Alk1 deletion but still delete the Alk1 gene at a desired adult stage, we applied the Cre/LoxP system in combination with the CreER system In this conditional knockout (cKO) mouse ( R26+/CreER; Alk12f/2 f), the transmembrane domain (exons 4 to 6) of the Alk1 gene on both allele s i s flanked by LoxP sequences, denoted as 2f and CreER cDNA i s i ntroduced into the ROSA26 locus, which allows ubiquitous expression of Cre recombinase. Upon TM treatment, Cre is a ctive, recognizes the LoxP sites and deletes the transmembrane domain, resulting in 1f null alleles As controls, R26+/+; Alk12f/2f mice were used in which the Alk1 cKO alleles are unaffected by TM because Cre is not present. Hereafter, mice containing R2 6+/+; Alk12f/2f will be indicated as controls, while R26+/CreER; Alk12f/2f mi ce will be denoted as mutants. All in vivo experiments w ere conducted in mice between 2 to 3 months of age. First, we investigated how Alk1 deficiency affects the early stage of solid tum or development. Since globally Alk1 null mice died between 9 to 21 days after TM injection, the in vivo tumorigenesis studies needed to be performed within this time frame Previously, we confirmed that a single intraperitoneal TM (2.5 mg/25 g bodyweight) injection was sufficient and efficient for deleting the Alk1 gene in R26+/CreER; Alk12f/2f mice On what is designated day zero,


75 both control (n=5) and mutant (n= 7) mice were injected with TM (Figure 3 1A). On the following day, Lewis lung carcinom a (LLC) cells (2 105 cells) were subcutaneously injected into their dorsal region and then tumor formation was monitored daily. B etween four to five days later, tumors were apparent in almost all control mice, whereas no obvious tumor mass was observed in mutant mice even more than a week later On the ninth day post TM administration, mice were sacrificed and LLC implanted skins were collected and macroscopically examined (Figure 3 1A). Initially, t here was no significant difference in eating behaviors and bodyweights between control and mutant mice during the study. However, between eight to nine days after TM injection, Alk1 deleted mutant mice started to show distinctive appearance. Their skin and extremities were pale implying poor blood circulation in the peripheral vasculature Additionally, the feces were very dark in color indicating possible internal bleedings A s expected an autopsy confirmed some hemorrhagic sites in the lungs and GI tracts in mutant mi ce. S urprisingly, only control mice devel oped apparent solid tumors under the skin (Figure 3 1B G). The t umors developed several feeding blood vessels surrounding as well as within them (Figure 3 1B D). However, in the mutants, less, discontinuous, irregular, thick blood vessels formed and hemorr hages were found in the area where tumor cells were inoculated (Figure 31E G). Therefore, these data demonstrated that the initiation of tumor growth was suppressed by Alk1 deletion in the adult mice most likely due to abnormal tumor angiogenesis by the A lk1 deficiency. Progression of Tumorigenesis Was Significantly Inhibited in Alk1 Deleted Adult Mutant Mice Next, we tested whether Alk1 deletion delayed the progression of already established tumors. To induce the establishment of solid tumors, we injecte d L LC cells (1 105 cells) into


76 the right thigh of control (n=4) and mutant (n=8) mice (Figure 3 2A). Before injection, diameters of both right and left thighs were measured as around 8 m m. Twelve days later, diameter of tumor injected right legs reached around 10 mm (Figure 3 2B), while the size of uninjected left legs barely changed. On that day, TM was administered into the mice and the size of tumors was measured daily (Figure 3 2B). By two days post TM treatment, the growth of tumors in both control a nd mutant mice continued but with no difference in their size. On the day three, the increase in the size of tumo rs within mutants was lower than that of controls. Until the seventh day after TM injection, the tumor gr owth in control mice steadily increase d, while that of mutant mice relatively slowed down Between seven to eight days, the tumor growth in mutants plateau ed. However, in controls, the tumors continued to enlarge and due to the enormous size of their tumors, the study had to be terminated. In the statistical analysis, the rate of tumor growth was significantly higher in control mice (0.78 mm 0.034) as compared to that of mutant mice (0.46 mm 0.024) ( P <0.0001) (Figure 3 2B). In addition to the diameter measurement size of tumors was further evaluated three dimensionally (width, thickness and height) during the autopsy. Tumors from control mice (Figure 3 2C E) were larger than those of mutant mice (Figure 3 2F H) in all measurements. As a note, there was no difference in size of the uninjecte d left thighs between mice. AVMs Were Resulted From Alk1 Deficiency in Peripheral Tumor Feeding Blood Vessels To visualize microcirculation of tumor feeding blood vessels of superficial muscular tumors we injected the latex dye into the arterial blood flow via th e left ventricle of the heart right before the autopsy. Through this method only arteries are visible because the dye is too viscous to pass through the capillaries. I n the control s single lined arteries were well branched or sprouted into smaller arterioles, but not capillaries nor veins of tumor feeding vasculature visualized. In terms of vascular morphology, a thick artery of the control became narrowed down


77 until their ends, and then diverged into several smaller vessels (Figure 3 3B). However, in the mutant, excessive, irregular, and disorganized vessels were observed around the tumor mass in the right thigh, and the end of an artery was blunt and looked discontinuous (Figure 3 3C). And the overall appearance of numerous blood vessels was abnor mal, appearing entangled and convoluted (Figure 3 3C). It is noteworthy that we did not observe such abnormal vessels in counterpart tumor uninjected left legs of the same Alk1 deleted mice (Figure 3 3A). This data supported our speculation that ALK1defic iency would specifically target the tumor vasculature where there is active angiogenesis. Importantly, when we further evaluated these tumor feeding microvessels w ith high magnification, the existence of AVMs was apparent (Figure 33D). The ends of arterie s were connected directly to veins through a thin vessel. Moreover, veins were remarkably dilated and wavy (Figure 33D). Such observations indicate that within the microcirculation of the tumor vasculature, arterial blood enters into veins through the AVM s, thus veins become enlarged and tortuous due to the high arterial blood pressure. Consequently, without properly functioning capillary beds, the arterial blood is directly intermingled with venule blood flows (arterio venous shunts; AV shunts), rather th an deliveri ng supplies into tumor masses. This leads to an imp ediment of local vascular circulation indeed starving the tumors and arresting their progression. Therefore, Alk1 deletion disrupted tumor angiogenesis and inhibited tumor growth by inducing AV Ms in established tumorfeeding vasculature. Alk1 Deletion during Tumor Angiogenesis Caused Disruption of Tumor Vascular Network The blood vessels within the tumor masses were further examined by various histological analyses. In hematoxylin and eosin (H & E) staining, tumor sections from Alk1 deleted mutant mice showed much more intact muscle tissues (Figure 3 4B), whereas in the control, healthy muscle tissues were displaced by LLC cells and barely present (Figure 34A). A closer


78 examination of only tumor s cells revealed s everal tiny well developed and maintained blood vessels in the control (Figure 3 4C) However, we found that blood cells largely dispersed in spaces between tumor cells in the mutants (Figure 3 4D). Importantly, unlike tumor cells surrounding blood vessels in the control, most tumor cells around these blood smears were necrotic. Next, we performed immunohistochemistry to evaluate vascular integrity of the tumor vasculature. Platelet endothelial cell adhesion molecule (PECAM), also known a s CD31, is a mediator for endothelial cell cell interactions. The endothelial cell layer is important for perfusion of molecules between blood vessels and local tissues. PECAM immunostaining was irregularly diffused in tumors of mutant mice (Figure 3 5B), while it delineated well the continuous inner layer of microvessels within tumors of the control mice (Figure 35A). Additionally, another widely used protein marker to examine the integrity of blood vessels is SMA). It stains the smooth muscle layer which allows vascular walls to be maintained and surrounds the endothelial layer. Consistent with the results from PECAM/CD31 staining, controls showed many well shaped tumor microvessels (Figure 3 5C), but positive staining was widely scattered throughout the sections within mutants (Figure 35D). These data indicated that the integrity of tumor vessel walls was severely disrupted by Alk1 deficiency. Consequently, due to the leakage of blood through ruptured vessel walls, microvessels within the tumor mass were not capable of proper exchange of small molecules such as nutrients, oxygen and wastes, thereby hindering the tumor growth. Thus, Alk1 deletion during tumor angiogenesis inhibited tumor progression not only by causing AVMs in tumor feeding blood vessels but also compromising vascular integrity of the established tumor vascular network.


79 Discussion We investigated whether inducing vascular abnormalities in the tumor vasculature by targeting the Alk1 gene could be a novel approac h to treat solid tumors. We found that Alk1 deletion inhibited the onset and progression of tumorigenesis by causing vascular disorganization, disruption and shunts. Heterozygous mutations of the ALK1 have been implicated in HHT, a dominantly inherited vas cular disease. Haploinsufficiency of ALK1 results in thinning of blood vessel walls and loss of capillary beds, leading to a dilated, fragile, disorganized and tortuous vascular network. Consequently, such abnormal blood vessels are prone to rupture, hemor rhagic and involve several AV shunts. In our preliminary study, it was demonstrated that both Alk1 deletion and woundinduced angiogenesis were necessary and sufficient for the development of these vascular lesions in the adult mice. Based on this promisin g data, we wanted to test whether such results could be recapitulated during tumor induced angiogenesis upon the Alk1 deletion. As expected, formation of abnormal blood vessels, characterized by dilation and appearance of excessive, tortuous vessels and AVMs, resulted from Alk1 deficient tumor angiogenesis. LLC cells injected under the skin could not establish solid tumors in Alk1 null mutant mice but showed few newly formed abnormal blood vessels and hemorrhages. However, in all control mice, they develope d tumors as well as new blood vessels. This observation indicated that suppression of the onset of tumor growth was mainly due to the failure to induce functional blood vessels. However, we could not rule out the possibility that such suppression was due t o other factors rather than impaired angiogenesis. One possibility is that as Alk1 null mutant mice died very quickly upon TM treatment, the observed inhibition of tumorigenesis could be due to a general defect such as impa ired nutrient supply rather than intrinsic resistance to tumorigenesis. Furthermore, Alk1 deletion also inhibited the progression of tumor growth. Unlike tumors


80 established in the controls persistently growing, the growth rate of tumors in the mutants was gradually decreased four days lat er and almost ceased a week after the TM treatment. Due to the considerable size of tumors in control mice, we could not continue to monitor progression or regression of tumors beyond the eighth day post TM in control and mutant mice. By latex dye injectio n, we found AVMs in the tumor feeding peripheral vasculature only in the mutant. Moreover, our immunohistochemical analysis demonstrated that vascular endothelial and smooth muscle layers of tumor microvessels in the mutant were severely disrupted and vasc ular integrity of the tumor vasculature was compromised. To support their growth and metastasis, solid tumors require new blood vessel formation for the establishment of their own vasculature, termed as tumor angiogenesis Thus, the basic concept of antian giogenic t umor therapy is that blocking tumor angiogenesis can inhibit the growth of tumors by preventing blood supplies to tumor tissues (145) T here has been evidence from numerous preclinical and clinical studies to show that angiogenesis is an important therapeutic targ et to treat many solid tumors. Currently, blocking the VEGF tyrosine kinase signaling has been the most and best tested strategy for the antiangiogene sis therapy (146 148) However, numerous clinical ap plications of anti VEGF therapy have made researchers and clinicians realize that targeting a single angiogenic pathway is not enough to achieve the desired antitumor effects. Additionally, emerging tumor vascular resistance against the anti VEGF therapy e mphasizes the urgent need of novel angiogenesis based therapeutic targets such as an giogenic activators, inhibitors as well as different angiogenic signaling pathways (23, 143, 148, 149, 151) In terms of alternative angiogenic pathways, data from recent reports show that despite increased blood vessel formation, blockade of Dll4 induced Notch signaling resulted in reduced


81 tumor growth (58, 159) In contrast to previous beliefs that there is an inverse correlation between vascular density and tumor growth, the increased tumor vascularity seen in blocking Dll4/Notch signaling was indeed antitumor igenic (58, 159) This data suggested that disorganiza tion of newly formed tumor feeding blood vessels by uncontrolled angiogenesis lead to vascular non functionality or productivity, resulting in the arrest and/or regression of tumors. Such results further support our hypothesis that abnormal blood vessels caused by Alk1 deficiency during tumor induced angiogenesis are poorly functional due to disorganization of arteries, veins and lack of capillaries, thereby inhibiting the tumor growth. Our finding that development of AV shunts via AVMs in the tumor vasculature of Alk1 deleted mutant mice suggests a new mechanism for vascular nonfunctionality or productivity. Most importantly, these AVMs were specifically formed in area of active tumor angiogenesis, suggesting that ALK1 blockage is more sensitive to activ e angiogenesis and can be targeted for tumor vasculature. It is noteworthy that a recent review proposes the necessity of new strategies that are designed to target specific subsets of tumor blood vessels (144) T he tumor vasculature was divided into six different categories, depending on their structural and functional characteristics: mother vessels (MV), capillaries, glomeruloid microvascular proliferations (GMP), vascular malformat ions (VM), feeder arteries (FA) and draining veins (DV). In a previous study, it was found that mouse tumors responded differently to anti VEGF/ VEGFR depending on their stages of tumorigenesis (160) Targeting VEGFR in endothelial cells showed efficacy for tumors in the early stages. However, it was not effective in late stage tumor s due to the presence of mature blood vessels containing pericytes. Therefore, they argued that the anti VEGF/VEGFR strategy might be effective for only tumor blood vessels appearing at the early stages of tumorigenesis


82 such as MV and GMP (144) T umor blood vessels existing at the late stages, such as FA and DV, are independ ent or resistant to the anti VEGF/VEGFR approach. Consequently, it is important and necessary to find new therapeutic targets to specifically destroy these blood vessels (144) As shown in our histological analysis, Alk1 deletion obviously caused the destruction of vascular smooth muscle layers of tumor blood vessels. In that respect, inhibition of ALK1 signaling may be a feasible targeting strategy for these tumor blood vessels to treat the late stage tumors. In summary, the ALK1 deficiency had an antitumor effect on the tumor vasculature All of findings presented strengt hen our hypothesis that blockage of the ALK1 signaling may be a new angiogenesis based therapeutic approach in target ing blood vessels that support tumor growth. First, since Alk1 induced vascular abnormalities occur only where active angiogenesis is present in adult s inhibition of ALK1 may provide a way to specifically target blood vessels around and within tumor masses undergoing elevated and excessive angiogenesis. Second, we found that the pathological effects of AVMs such as the lack of capillaries and appearance of AV shunts could be antitumorigenic by causing nonfunctionali ty in the tumor blood vessels. Lastly Alk1 deletion affected the vascular endothelial as well as the smooth muscle cell layers of blood vessels, resulting in fragile, leaky blood vessels as well as disruption of established microvessels. Thus, inhibition of the ALK1 signaling can be an effective strategy to target the tumor vascular network appearing at both early and late stages tumors.


83 Figure 31. Initiation of tum or growth was r epressed in R26+/CreER; Alk11 f/1 f mutant m ice. A) Schematic diagram of the experimental plan. Tamoxifen ( TM; 2.5 mg/25 g body weight) was administered one day before LLC cells (2 105 cells) injection. Nine days later representative picture s were taken at the autopsy. B) D ) Solid tumors formed under the ski n of all R26+/+; Alk12f/2f control mice (n =5) showed well developed tumor feeding blood vessels, indicated by black arrows, around and within them. E) G ) No tumor formation was observe d in Alk 1deleted mutant mice (n=7). Abnorma l blood vessels and hemorrhages, represented by blue and red arrows respectively, were found in the area wh ere tumor cells were injected.


84 Figure 32. Tumor growth was significantly i nhibited in Alk1 null mutant m ice. A) LLC cells (1 105 cells) were intramusculary injected into the right legs of Alk12f/2f control (n=4) and Alk11f/1f mutant (n=8) m ice and grown for twelve days. On the twelfth day after LLC cell inoculation, TM was injected into the per itoneum of both control and mutant mice. For eight days, diameters of their right thighs were daily measured. B ) Rate of tumor growth in Alk1 deleted mutant mice (0.46 mm 0.024) was significantly lower than that in control mice (0.78 mm 0.034) ( P <0.0001). At the autopsy, the size of tumor masses was evaluated by a threedimensional measurement (width, thickness and height). In all measurem ents, C) E ) tumors from the Alk12f/2f controls were considerably larger than F) H ) ones from the Alk11f/1f mutan ts


85 Figure 33. Latex day injection revealed Alk1 deletion c aused AVMs i n tumor feeding blood vasculature. A ) Left leg of a TMinjected mutant mouse maintained a highly organized vascular network in which only arteries were visualized by latex dye in t he absence of tumor angiogenesis. B) In Alk12f/2 f c ontrol mice, similar to A, only arteries a ppeared as a single branches which diverged into smaller branches at their end. C) In the right leg which was from the same Alk11f/1f mutant mouse in A, t umor f eeding blood vessels were excessive, disorganized, irregular and tortuous. D) A high magnification of C demonstrated direct connections between arteries and veins (AVMs) in the tumor vasculature of Alk1 null mice. Because of AV shunts within the AVMs, veins were dilated and convoluted, indicating the absence of intervening capillary beds. In the inset, the artery is delineated with a red solid line, while veins are lined dark blue. AVMs are represented as yellow lines. Note that in D, A = artery and V = vein.


86 Figure 34. Tumor blood vessels were destroyed by Alk1 deficiency. Sections of muscle tumors were stained with hematoxylin and eosin (H & E ). A) Tumors from control mice contained ver y little muscle tissues (50X). B) In mutant mice, intact muscle tiss ue s were broadly observed (50X). Note that healthy muscle tissues in pink are point ed by black arrows in A and B. C) Growth of LLC cells was supported by several well established micro vessels in the control (200X). D) Due to the absence of microvessels, lo cally delivered blood leaked and was dispersed into tumor tissues in the mutant (200X). Unlike tumor cells in C, ma ny of them underwent necrosis. Note that black arrows in C and D indicate blood cells within tumor tissues.


87 Figure 35. Tumor vascular integrity was d isrupted by Alk1 deletion. Sections of muscle tumors were immunostained with A ) and B) an endothelial specific marker protein PECAM/CD31 or C) and D) a smooth muscle cell SMA. A) The endothelium was well maintained in tumor vessels of control mice was detected by PECAM/CD31 positive staining in light red (200X). B) PECAM/CD31 positive signals were disseminated in mutant mice suggesting disruption of the endothelial layer of tumor vasculature (200X). C) Tumors of the controls presented nicely formed smooth muscle cell layers surroundin g vascular endothelium (200X). D) In the mutants, SMA staining implied that the smooth m uscle cell layers of tumor feeding blood vessel s were also affected (200X). Note that black arrows in A and C indicate microvessels within the tumor mass. And asterisks in B and D demonstrates site s where microvessels were disrupted, thereby vascular endothelial and smooth muscle cells were scattered.


88 CHAPTER 4 CONCLUSIONS AND FUTURE STUDIES The major pathological features of HHT are the dilation of microvessels and the formation of AVMs In that respect, HHT is a unique vascular disease th at may allow us to study underlying mechanisms for the maintenance of vascular integrity and the pathological or physiological effects of AVM s. Despite the identification of HHT disease causing genes such as ENG (7073) ALK1 (7476) and SMAD4 (77) a nd intense efforts to und erstand their signaling, there are still many unanswered questions. Previously, based upon in vivo studies, we demonstrated that perturbation in the EC specific ALK1 signaling was a major cause of HHT like vascular lesions in mice (115, 161) And we recently found that combination of Alk1 deletion and ongoing an giogenesis were necessary and sufficient for de novo formation of vascular abnormalities in the established adult vasculature. T he vasculature in the lung and GI tract were consistently affected by the Alk1 deletion in both embryonic and adult stages. It i s noteworthy that Alk1 heterozygous mice did not develop such vascular lesions during adult angiogenesis. In chapter 2, we investigated cellular and biochemical properties of ALK1 deficient pulmonary ECs in angiogenesis. The significance of this study was to determine the most pertinent molecular mechanism fo r the endothelial ALK1 signaling involved in the pathogenesis of HHT. Some questions of interest include: w hat are distinctive cellular traits including proliferation, migration and endothelial capillar ylike assembly of Alk1 null ECs in response to proangiogenic cues? Is the ALK1 receptor specific for controlling the TGF signaling? Or is it more relevant receptor for regulat ing the BMP subfamily pathway? What are downstream molecular pathw ay(s) responsible for mediating the ALK1 signaling in ECs? To address such important issues in a more manageable system, we established two new murine


89 pulmonary EC lines. Since we previously observed that the heterozygous deletion of the Alk1 gene did not affect the adult vascular network in the presence or absence of angiogenesis, we isolated ECs from the mouse carrying the R26+/CreER; Alk12f/1f transgene. B y utilizing the Cre/LoxP system, their genotypes could be switched by a single TM treatment from Alk 12f/1f ( Alk1 null heterozygote) to Alk11f/1f ( Alk1 null homozygote). In accordance with our in vivo data, there was no difference observed in cellular phenotypes including migration and tube formation between Alk12f/1f and Alk11f/1f ECs in the absence of p ro angiogenic factor(s). However, upon bFGF treatment, Alk1 null ECs displayed significantly increased migratory abilities Moreover, they developed considerably excessive, disorganized, irregular and enlarged capillary like structures on the Matrigel in response. These abnormal tube like structures were due to rapid outgrowth of ALK1deficient ECs at early time points and their resistance to regression at the later stages. To test whether antiangiogenic effects of TGF tially exerted in angiogenesis in the presence or absence of the ALK1, cells were treated with TGF 9, respectively. It was found that a capillary like network formed by Alk1 null ECs was significantly regressed by TGF from 6 hours, whereas mor e than 50% of such a network was sustained in response to BMP 9 at least up to 24 hours. This data indicated that TGF induced anti angiogenic effects were present, while such effects from BMP 9 treatment were blunted in Alk1 null ECs. Western blot analy sis further supported this data by demonstrating that the activation of ERK1/2 pa thway was decreased by TGF Ho wever, such reduction was less after B MP 9 treatment. In contrast to previous studies that suggest the ALK1 signaling is SMAD1/5/8 dependent the induction of this pathway was intact in our Alk1 null ECs. Thus, it s uggests that ALK1 signaling is SMAD -


90 independent and that other molecular pathway(s) may be mediating the physiological effects of ALK1 in pulmonary ECs. This study focused on finding a specific ligand and downstream molecular pathway for endothelial ALK1 signaling. In addition, identifying molecular target(s) of its signaling is also essential in elucidating the molecular mechanism relevant to the HHT pathogenesis. R ecently our labor atory performed a microarray analysis to investigate downstream genes which may be regulated by ALK1 By comparison of the transcript profiles between Alk12f/2f control and Alk11f/1f mutant lungs at neonatal and adult stages, we obtained a group of genes t hat were commonly upor downregulated at symptomatic conditions (e.g. vessel dilations and pulmonary hemorrhages) at both stages. The functions of these genes are known to be implicated in many pathways important for the normal vascular physiology such a s TGF signaling, vascul ar tone, vascular permeability and angiogenesis. Since the whole lungs displaying HHT like phenotypes were used, this group of genes was supposed to reflect both causative and consequent gene expressions of p athological manifestation To identify EC specific ALK1 target genes that may contribute to pathological features seen in HHT future studies should include such microarray analysis examining overlapping gene expression profiles between in vivo and in vitr o HHT like pathological condit ions upon the ALK1deficiency. F urther investigations will provide a better characterization of the EC specific ALK1 signaling. Therefore, the discovery of ALK1specific molecular mechanism(s) will facilitate the development of the potential therapeutic target(s) to prevent or cure the clinical symptoms of the HHT. In chapter 3, we tested whether creation of pathophysiological effects of HHT vascular lesions in the tumor vasculature could be a novel angiogenesis based antitumor approach to treat


91 solid tumors. T he clinical manifestation of HHT such as spontaneous, recurrent bleedings from fragile microvessels and AV shunts via AVMs, are distinctive characteristics of a non functional vascular network. T raditional angiogenesis ba sed antitumor therapy was based on the concept that blockage of angiogenic signaling pathways would inhibit tumor angiogenesis preventing tumor growth. However, a recent in vivo investigation suggested a novel concept that increased induction of nonproductive or functional tumor vascularity by disrupting Notch/Dll4 signaling could be antitumorigenic (58) implyi ng the importance of organization and functional ity for the tumor vasculature. Thus, it led us to speculate that generation of ALK1 deficiencycausing nonproductive HHT like vascular abnormalities ( paucity of capillaries and development of AV shunts ) migh t inhibit the tumor growth. We investigated effects of the Alk1 deletion on the initiation and progression of tumorigenesis induced by the LLC cell i mplantation in vivo We observed that all of R26+/+; Alk12f/2f control mice developed solid tumors and nasc ent tumor feeding blood vessels under the skin. By contrast, Alk1 null mutant mice showed no tumor formation, but did form new ab normal vessels and hemorrhages surrounding the site of tumor cell inoculation. Such observations demonstrate that in order for a solid tumor to develop, ALK1 is required for tumor angiogenesis. The ALK1 deficiency also affected the tumor growth by disrupting the established tumor vasculature Upon deletion of the Alk1 gene in mice by TM administration, the rate of tumor growth slo wed down after three days and ceased growth at around a week. However, the tumor established in control mice rapidly and continuously grew. W e examined whether such a difference was due to d evelopment of HHT like vascular lesions in the tumor microvasculat ure We found the existence of AVMs in the peripheral microcirculation of the tumor va sculature in only Alk1 deleted mutant mice. Histological and immunohistochemical examination of tumor


92 sections of the control displayed well established and maintained tumor feeder microvessels. O n the other hand, those of the mutant contained blood smears, more intact muscle tissues, and necrotic tumor cells without blood vessel structures SMA staining was diffused and scattered, indicating disruption of blood vessel wall integrity. Taken together, the onset and progression of the tumorigenesis was strongly influenced by ALK1s role in tumor angiogenesis and mai ntenance of the vasculature This study investigated the antitumor impact of the blockage ALK1 signaling by taking a genetic ablation approach. A ll findings of this study substantiated our hypothesis that vascular nonproductivity and disruption resu lting from the ALK1 deficiency in the tumor microcirculation may be antitumorigenic by compromising the blood supply to tumors. Based upon our promising in vivo data, various strategies are possible to target the ALK1 signaling for clinical applications. To avoi d adverse effects, such as spontaneous formation of vascular lesion s seen in HHT in the systemic blood vessels, pharmacological means targeting the ALK1 signaling should be local and specific. So far, as a pharmacological inhibitor for the ALK1 signaling, a ALK1 trap has been developed by Acceleron Pharma Inc (Cambridge, MA). The ALK1 trap is a soluble chimeric ALK1 receptor in which the ligandbinding domain of ALK1 is fused wit h the Fc region of an immunoglobulin. Therefore, its interaction with ALK1 lig ands interrupts the ALK1 signaling by inhibiting association between the ALK1 ligand and its physiological ALK1 receptor. Although the underlying mechanism is unknown, it has been shown that ALK1Fc can block the ALK1 signaling in a cell culture system and such blocking showed some inhibitory impacts on the tumor growth. In the future investigations, it will be interesting to compare the efficacy of the ALK1 Fc to that of our genetic deletion approach in vivo In addition, it will be important to examine


93 wh ether inhibitory effects of ALK1 Fc can cause the same vascular defects observed in our Alk1 deleted mice such as AVMs and vessel wall rupture. B etter understanding of the ALK1 signaling will provide more molecular targets and more specific method in targe ting its pathway Thus, the follow up studies to chapter 2 and 3 shoul d be considered concomitantly. The identification of the HHT pathogenesis may offer specific and feasible targets within the tumor vasculature We are hopeful that based on such a discovered molecular mechanism, various pharmacological approaches can be explored to modulate the ALK1 s ignaling. Examples of them include ALK1 ligand binding proteins such as dominant negative forms of ALK1 receptor lacking transmembrane and cytoplasmic kinas e domains (ALK1trap), diverse chemical serine/threonine kinase inhibitors inhib iting kinase activities of ALK1 and small interfering RNAs (siRNAs) blocking components of the ALK1 signaling. We hope and expect that inhibitors of ALK1 signaling will be an e xcellent alternative to current angiogenesis based antitumor reagents and will be more efficacious in combination with other anti angiogenic and/or cancer drugs.


94 CHAPTER 5 MATERIALS AND METHODS Transgenic Mice All procedures performed on mice were reviewed and approved by the University of Florida Institutional Animal Care and Use Committee. The generation of R26+/+; Alk12f/2f, R26+/CreER; Alk12f/2f, and R26+/CreER; Alk12f/1f transgenic mouse lines was described previously (115) PCR genotyping was performed as previously detailed (115) Overall Cell Culture Conditions Murine pulmonary endothelial cells (pECs) were cultured in a specifically formulated endothel ial cell medium (ECM) in which Dulbeccos modified eagle medium (DMEM; GIBCO) was supplemented with 20% fetal bovine serum (FBS; HyClone), 0.5% heparin (200 mg/ml; Sigma), 1% endothelial mitogen (10 mg/ml; Biomedical Technologies Inc), 1% nonessential amin o acids (Mediatech), 1% sodium pyruvate (100 mM; Invitrogen) and 0.4% penicillinstreptomycin (Invitrogen). All culture plates (Falcon) and flasks (Falcon) used for the pEC culture were coated with 1:5 diluted bovine fibronectin stabilized solution (1 mg/m l; Bomedical Technologies Inc) and then incubated at 37C for 30 minutes before each use. Lewis lung carcinoma (LLC) cells were maintained in minimum essential medium (MEM) supplemented with 10% FBS, 0.25% L Asparagine (Sigma), 1% L Glutamine (200 mM; Invitrogen), sodium bicarbonate (added up to pH 7.27.4; 7.5% (w/v); Invitrogen) and 1% penicillin/streptomycin (Invitrogen). Both EC and LLC cell cultures were incubated at 37C with 5% CO2. Establishment of Alk12f/1f and Alk11f/1f Pulmonary Endothelial Cells (pECs) An eight week old mouse carrying the R26+/CreER; Alk12f/1f transgene was euthanized by an overdose of 100% isofluran and follo wed by cervical dislocation. W hole lungs were removed


95 and washed in HEPES followed by another washing in DMEM. Lung tissues were finely minced using a sterile scalpel The chopped tissues were subjected to serial digestion using 2 ml of a 1X trypsin soluti on [0.25% trypsin, 0.5 M EDTA (pH 8.0) in DMEM] at 37C with frequent shaking, for three times at 8 minutes each. Trypsin digestion was inactivated by adding 6 ml of normal ECM. After a 10 minute incubation at room temperature, the supernatant was carefu lly collected and plated into 2 wells of a 6 well culture plate. For immortalization, when the culture reached 50% confluency post isolation, they were transfected with SV40 DNA (4.0 g) using Lipofectamine (Invitrogen) and following t he manufacturers pro tocol. To obtain homozygous Alk1 null ( Alk11f/1f) pECs, immortalized parental heterozygous Alk1 null (Alk12f/1f) pECs were treated with 1 M 4 hydroxytamoxifen (Sigma) for two consecutive days. Sorting pECs by Fluorescent Activated Cell Sorting (FACS) At c lose to 100% confluency, the EC culture was incubated with 10 g of DioAc LDL (200 g/ml; Biomedical Technologies Inc) diluted in ECM at 37C/5% CO2 for 4 hours. C ells were then briefly washed twice with DMEM containing 10% FBS and Hanks balanced salt so lution (HBSS, Invitrogen) once To obtain a single cell suspension, cells were trypsinized and washed in DMEM containing 10% FBS three times. The final cell suspension was diluted in the phenol red free DMEM/F12 medium (DMEM and Hams F 12, 50:50 mix, 1X; Mediatech) at the appropriate concentration (3 106 cells/ml). T he cell suspension was sorted using the BD FACSAria Cell Sorting System ( BD Biosciences) at wavelengths of 484 nm (excitation) an d 507 nm (emission). E ndothelial marker positive sorted cells were washed in DMEM with 10% FBS five times Lastly, cells were diluted in ECM supplemented with 1% endothelial cell growth supplement (ECGS; BD Biosciences) at a concentration of 2.5 104 cells/cm2.


96 Genomic DNA PCR Analysis For genotyping of parental Alk 12f/1f and TM treated Alk11f/1f pECs, genom ic DNA was extracted from each cell culture grown at 100% confluency in a w ell of a 6 well culture plate. Cells were lysed in lysis buffer [50 mM Tris (pH 8.0), 0.5% TritonX 100] containing proteinase K (1 mg/ml) at 55C for overnight. On the next day, the lysed cells were centrifuged at 12,000 rpm for 10 minutes. The PCR reaction mixture was comprised of the following components: 5 l of 5X buffer (Promega), 3 l of MgCl2 (25 mM; Promega), 0.5 l of dNTPs (25 pM; Promega), 0.5 l of each primer (25 pM; Integrated DNA Technologies Inc), 1 l of genomic DNA and 12 l of H2O. The final mixture was covered with mineral oil to prevent evaporation. Before the addition of a Taq polymerase (500U; Promega), the mixture was boiled at 94C for 10 minutes. 2 l of a Taq polymerase was added at 72C A 35 cycle PCR reaction was run consisting of the following conditions: denaturing at 94C for 45 seconds, priming at 60C for 45 seconds and e xtending at 72C for 1 minute. T he last cycle remained at 72C for 10 minutes. The PCR products were separated on a 3% Agarose gel ( Lonza). The primers used for detecting the Alk1 cKO ( 2f ) or Alk1 null ( 1f ) alleles are summarized in the Table 5 1. RTPCR Analysis To confirm the expression of e ndothelial biomarker genes, parental Alk12f/1f and TM treated Alk11f/1f pECs were cultured until 100% confluency in a 25cm2 culture flask. Total RNAs from c ultures were extracted using the NucleoSpin RNA purification kit (Clontech). All experimental steps followed the manufacturers protocol. 2 g of RNA was used for reverse transcription (RT) reaction. The cDNAs were synthesized using SuperScript III FirstStand synthesis kit (Invitrogen) according to recommendations and guidelines from the company. After the RT reactio n, 2 l of cDNA was used for the PCR analysis. The PCR reaction mixture and hot start PCR reaction were performed as detailed in the genomic DNA PCR analysis. Three


97 different PCR cycles were run For the amplification of RNAs for Alk1, Smad1, E ndothelin and Actin, a total of 27 cycles were run as the following: denaturing at 94C for 45 seconds, priming at 60C for 45 seconds and e xtending at 72C for 1 minute. T he PCR products were further processed at 72C for 10 minutes For Tie2, Flk1 and Endoglin, the PCR consisted of a denaturation at 94C for 45 seconds, annealing at 60C for 45 seconds and extension at 72C for 1 minute 30 seconds for a total of 28 steps. Lastly, cDNAs for ActRIIa, Bmpr2, Alk2, Alk3, and Alk6 were amplified by the follo wing PCR cycles: denaturing at 94C for 45 seconds, priming at 55C for 45 seconds and extending at 72C for 1 minute 30 seconds. The primers used for RT PCR analyses are summarized in the Table 5 2. Western Blotting Protein lysates extracted from pECs wer e subjected to Western blot analyses. The same number of Alk12f/1f and Alk11f/1f pECs were seeded into each well and grown in a 6 well culture plate. At 100% confluency, medium was changed from normal ECM to chemically defined growth factor and serum fre e ECM (Genlantis) for overnight serum starvation. On the next day, after removing defined ECM, the cultures were replenished with defined ECM containing bFGF (50 ng/ml; BD Biosciences) supplemented with either TGF BMP 9 (20 ng/ml; R & D Systems) and incubated for 30 minutes at 37C. This stimulation was terminated by brief ly washing the cells with sterile PBS twice. C ells were lysed by adding 80 l of 1X protein loading buffer [5X; 60 mM Tris HCl (pH 6.8), 25% glycerol, 2% SDS, 14.4 mM mercaptoethanol, 0.1% bromophenol blue in H2O] and collected in an eppendorf tube placed on ice. The whole pr otein lysates were boiled at 95 to 100C for 5 minutes and cooled down on ice. After centrifugation at 13,000 rpm for 5 minutes, protein lysates were stored at 80C until needed 15 l of each protein lysate was separated by SDS PAGE on a 8% polyacrylamide (BioRad) protein gel. The fractionated proteins were transferred to PVDF membranes. Aft er


98 blocking with a 5% nonfat milk blotting solution, membranes were incubated with a primary antibody against the protein of interest at 4C overnight. On the following day, membranes were washed with a washing buffer [0.05% Igepal (Sigma) in PBS] and inc ubated with an appropriate horseradish peroxidase linked secondary antibody at 4C for 1 hour. Lastly, the protein was detected by a chemiluminescent Western blotting detection reagent (Amersham Pharmacia Biotech Inc). The primary antibodies used for Western blot a nalyses are the followings: SMAD 1 (1:500; Clone: 2C2; Chemi con International), phosphoSMAD1/5/8 (1:500; Cat #: 9511; Cell Signaling Technology), SMAD 2 (1:500; Cat #: 3103; Cell Signaling Technology), phosphoSMAD2/3 (1:500; Cat #: 3101; Cell Signali ng Technology), p44/42 MAPK (ERK 1/2; 1:500; Cat #: 9102; Cell Signaling Technologies), phosphop44/42 MAPK (phosphoERK1/2; 1:500; Cat #: 9106; Cell Signaling Technology) and GAPDH (1:10,000; Cat #: ab8245; Abcam). The secondary antibodies used for Western blot analyses are the followings: HRP labelled anti mouse IgG (1:5,000; Cat #: NA931; Amersham) and HRP linked anti rabbit IgG (1:5,000; Cat #: NA934; Amersham). in vitro Endothelial Migration Assay For the 2D woundinduced migration assay, pECs were plated and grown i n a 6well culture plate. When cells reached 100% confluency, three wounding lines per well were created by scraping with a sterile tip Then, cells were briefly washed with 1X HBSS (Invitrogen) two times and replenished with ECM or chemi cally defined growth factor and serum free ECM (Genlantis) containing bFGF (50 ng/ml) (BD Biosciences). Closing of these wounds was photographed at every 4 hours post wound up to 12 hours. Based on a decrease in the width of each wound over time, ECs migr ation was calculated.


99 added into each we ll in a 24 well culture plate. (Biosciences) were then placed into each well. pECs (5 102 cells) were of ECM containing 2% FBS and seeded into e ach migration chamber. After 48 hours incubation, medium was removed and migration chambers were stained with crystal violet dye The chambers were allowed to air dry overnight Subsequently, stained cells in six randomly chosen fields were counted under the microscope C hemically defined growth factor and serum free ECM (Genlantis) containing bFGF (50 ng/ml) (BD Biosciences) with 10% FBS was added to the bottom chambers and 1% FBS to t he upper chambers to get rid the effects of FBS on ECs migration. in v itro Tube Formation Assay on Matrigel The p henol redfree Matrigel (BD Biosciences) was thawed overnight at 4 C 200 l of Matrigel was added into each well in a pre chilled 24well plate. Then, Matrigel was solidified by incubation at 37C for 1 hour. pECs (6 104 cells/ well ) were suspended in 500 l of chemically defined growth factor and serum free ECM (Genlantis) containing bFGF (50 ng/ml; BD Biosciences) in combination with either T GF 9 (20 ng/ml; R & D Systems) and seeded into each well The endothelial capillary like network formation was phot ographed at various time points: 3, 6, 9, 12 and 24 hours after seeding. To obtain quantitative readouts f or the statistical analysis, all pictures were processed by the MatLab imaging program. From the processed images, areas of tube like structures, total length of tubes and the number of endothelial outgrowth from each nodule were calculated. in vivo Matri gel Plug Angiogenesis Assay One day before or on the same day of the Matrigel implantation, 100 l of tamoxifen (25 mg/ml of corn oil ; Sigma) was injected into the peritoneum of both R26+/+; Alk12f/2f control and

PAGE 100

100 R26+/CreER; Alk12f/2f mutant mice On the f ollowing day, 200 l of high concentration Matrigel (9.6 mg/ml) (BD Biosciences) was thawed overnight at 4C, then (250 ng/ml) (BD Biosciences) and kept on ice For anesthesia, the recipient mice were placed into a chamber in whic h 4% isoflurane gas was applied. During the injection procedure, anesthesia was maintained at 2 to 3% isoflurane. Right before the Matrigel injection, the fur on the dorsal area of the mice was shaved. injected into the skin of dorsal region of mice Between 7 to 10 days post tamoxifen injection, mice were sacrificed and the skin bearing implanted Matrigel plugs was excised. Samples were fixed in 4% paraformaldehyde for at least 24 hours at room temperat ure. After washing in PBS two times for 15 minutes each, fixed plugs were sequentially dehydrated with an increasing concentration of ethanol (70% for 20 minutes, 95% for 1 hour and 100% for 2 hours) and incubated in an organic solvent for 20 minutes The Matrigel plugs were incubated in paraffin for 3 hours and subsequently embedded in paraffin, then were sectioned at 5 examination, sections were incubated at 60C overnight and deparaffinized by incubation in an organic so lvent (C itris olv; Fisher Scientific) for 30 minutes. S ections were sequentially rehydrated with decreasing concentration s of ethanol (100% for 10 minutes, 95% for 10 minutes, 70% for 10 minutes and H2O) and subjected to hematoxylin and eosin (H & E) staining. X Gal Sta ining For the visualization of migrating ECs or nascent vessels in Matrigel plugs from above, expression of the LacZ gene under the control of Flk1 promoter was examined by X gal staining. R26+/+; Alk12f/2f control or R26+/CreER; Alk12f/2f mutant mice carr ying the Flk1LacZKI allele were sacrificed and Matrigel plugs along with the attached skin were isolated. The isolated plugs were fixed in a fixative solution [1% formaldehyde (Fisher Scientific), 0.2% glutaraldehyde (Sigma), 2 mM MgCl2, 5 mM EGTA, and 0.02% NP 40 in PBS] at room temperature for 10 minutes.

PAGE 101

101 Fixa tion was followed by washing thrice with PBS for 5 minute each with rocking. Then, samples were placed in an X gal staining solution [5 mM K3Fe(CN)6, 5 mM K4Fe(CN)6, 2 mM MgCl2, 0.01% NaDeoxycholat e, 0.02% NP 40, 100 mM phosphate buffer (pH 7.3), 0.5 mg/ml X gal (Fisher Scientific) in distilled deionized H2O ] and incubated overnight at 37C T he next day, X gal stained Matrigel plugs were washed with PBS, subsequently examined and phot ographed unde r the microscope. After microscopic observation, plugs were fixed in 4% paraformaldehyde solution at room temperature for at least 24 hours. For the histological analysis, samples were sequentially dehydrated as described above and paraffin embedded. 5 m thick sections were deparaffinized and rehydrated as described above then counterstained with nuclear fast red (NFR). in vivo Subcutaneous Tumor Generation Tumor formation under the skin was induced by the subcut aneous injection of LLC cells. 100 l of t amoxifen (25 mg/ml; Sigma) was intraperitoneally administered into R26+/+; Alk12f/2f control and R26+/CreER; Alk12f/2f mutant mice On the next day, LLC cells were suspended in sterile PBS at a concentration of 1 105 cells/10 l The mice were anesthetiz ed as detailed in the in vivo Matrigel plug angiogenesis assay Prior to tumor cell injection, the fur of mice in the area surrounding the injection site was shaved, then 20 l of the LLC cell suspension (2 105 cells) was injected into each mouse. F or fo llowing nine days, the behaviors of the mice and tumor growth were monitored daily At day 9 post LLC cell inoculation, the mice were euthanized by an overdose of 100% isofluran and subsequent cervical dislocation. The tumor cell implanted dorsal skin was excised and fixed in 4% paraformaldehyde for at least 48 hours at room temperature. After fixation, samples were washed in PBS for 15 minutes twice and stored in PBS at 4C.

PAGE 102

102 in v ivo Intramuscular Tumor Generation To study effects of Alk1 deletion on the es tablished tumor vasculature, the in vivo intramusc ular tumor model was employed. LLC cells were prepared as described previously. 10 l of tumor cell suspension (1 105 cells) was injected into the right thigh of each R26+/+; Alk12f/2f control and R26+/Cr eER; Alk12f/2f mutant mouse. The implanted tumor cells were allowed to be grown for 12 days with daily observation. On day 12, the diameter of the right thigh of control and mutant mice was between 10 to 11 mm. After intraperitoneal administration of 100 l of tamoxifen (25 mg/ml; Sigma), tumor growth was further monitored daily for 8 days. At day 20 post LLC cell implantation, the mice were sacrificed and right legs were collected. Left legs were also harvested as an internal control. The f oot and fur were removed and then the whole thighs were fixed in 4% paraformaldehyde at room temperature for at least 48 hours. The thighs were incubated in a decalcification solution (Biochemical) for 3 days at room temperature with shaking, sliced at 1 to 2 mm thicknes s for better penetration then further decalcified in fresh solution for 3 more days. Decalcified samples were then subjected to multiple brief tap water washings followed by two 30 minute washings with tap water on a rocker. Samples were stored in PBS at 4C until needed. Latex Dye Injection T o visualize the systemic vascular circulation mice bearing intramuscular tumor were subjected to the latex dye injection (Blue latex, Catalog# BR80B, Connecticut Valley Biological Supply Co, Southampton, MA) M ice we re anesthetized by intraperitoneal injection of Ketamine/Xylazine (100 mg/15 mg per 1 kg body weight) The chest cavity of the mouse was opened to expose the heart, and then latex dye was injected into the left ventricle of the heart The hair removed whole right legs were briefly washed in PBS and fixed in 4% paraformaldehyde at room temperature for at least 3 days To obtain the whole mount imaging,

PAGE 103

103 after washing in PBS twice for 15 minutes, samples were sequentially dehydrated by increasing concentratio n of methanol (20% for 20 minutes, 50% for 20 minutes, 75% for 20 minutes, 90% for 60 minutes, 100% for 1 hour and 30 minutes) and cleared with an organic solvent (Benzyl alcohol:Benzyl benzoate = 1:1; Sigma). Histology and Immunohistochemistry After fixation in 4% paraformaldehyde solution between 2448 hours depending on the size, tumor samples were washed in PBS twice for 15 minutes each. Samples were dehydrated by a n increasing ethanol series of 70% (2 times for 30 minutes/each), 95% (two times for 50 minutes/each) and 100% (three times for 1 hour/each), and then 100% organic solvent (two times for 30 minutes/each) which was followed by incubation in paraffin for 3 hours. The paraffin embedded tissues were sectioned at 5 T he sectioned tissues were deparaffinized by heating at 60C overnight and 40 minutes incubation in an organic solvent, then rehydrated through a degraded et hanol series ( 100% 2O) and stained with hematoxylin and eosin (H & E). For immunohistochemistry, sectioned samples were processed as same as described in preparation for the H & E staining. The standard ABC method using a Vector immunodetection kit (Vector Laboratories) was performed. Briefly the rehydrated sections were treated with 3% (v/v) hydrogen peroxide (H2O2) at room temperature for 10 minutes to block the endogenous peroxidase activity. After washing in PBS twice, the sections were incubated with b locking serum corresponding to the species from which the secondar y antibody was raised a t room temperature for 1 hour. Next, the incubation with a primary antibody for the protein of interest at room temperature for 1 hour and 30 minutes was followed by washing with PBS. A biotinylated secondary antibody was applied to the samples for 3 0 minutes at room temperature. S ections were treated with the peroxidaseconjugated biotin/avidin complex for 30 minutes at room

PAGE 104

104 temperature Two PBS washes were followed by procedures for the color development via applying a DAB substrate chromogenic s olution (Vector laboratories). The primary antibodies used for the immunohistochemist ry are the following : PECAM/CD31 (1:300; Cat #: CM 303; SMA (1:800; Clone: 1A4; Sigma). Statistics Data were represented as mean SEM or SE. The differences between groups were determined by the m ixed linear regression model general linear regression model, or TwoW ay ANOVA. A value of p<0.05 was considered statistically significant.

PAGE 105

105 Table 5 1. Summary of primers used for genomic PCR a nalysis. Genes Forward Primers Reverse Primers Alk1 (2f; floxed) CAGCACCTACATCTTGGGTGGAGA ACTGTTCTTCCTCGGAGCCTTGTC Alk1 (1f; null) CAGCACCTACATCTTGGGTGGAGA TCCTCTTGTCGTATATGTCCC

PAGE 106


PAGE 107

107 LIST OF REFERENCES 1. Jain, R. K. (2003) Molecular regulation of vessel maturation. Nat Med 9, 685693 2. Risau, W., and Flamme, I. (1995) Vasculogenesis. Annu Rev Cell Dev Biol 11, 7391 3. Folkman, J. (1995) Seminars in Medicine of the Beth Israel Hospital, Boston. Clinical applications of research on angiogenesis. N Engl J Med 333, 17571763 4. Folkman, J. (1995) Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1, 2731 5. Pepper, M. S. (1996) Positive and negative regulation of angiogenesis: from cell biology to th e clinic. Vasc Med 1, 259266 6. Reynolds, L. P., Killilea, S. D., and Redmer, D. A. (1992) Angiogenesis in the female reproductive system. Faseb J 6, 886892 7. Risau, W. (1997) Mechanisms of angiogenesis. Nature 386, 671674 8. Patan, S., Haenni, B., and Burri, P. H. (1996) Implementation of intussusceptive microvascular growth in the chicken chorioallantoic membrane (CAM): 1. pillar formation by folding of the capillary wall. Microvasc Res 51, 8098 9. Pepper, M. S., Mandriota, S. J., Vassalli, J. D., Orci, L., and Montesano, R. (1996) Angiogenesis regulating cytokines: activities and interactions. Curr Top Microbiol Immunol 213 ( Pt 2), 3167 10. Pepper, M. S., Montesano, R., Mandriota, S. J., Orci, L., and Vassalli, J. D. (1996) Angiogenesis: a par adigm for balanced extracellular proteolysis during cell migration and morphogenesis. Enzyme Protein 49, 138162 11. Pepper, M. S. (1997) Transforming growth factor beta: vasculogenesis, angiogenesis, and vessel wall integrity. Cytokine Growth Factor Rev 8, 2143 12. Carmeliet, P. (2000) Mechanisms of angiogenesis and arteriogenesis. Nat Med 6, 389395 13. Bicknell, R., and Harris, A. L. (2004) Novel angiogenic signaling pathways and vascular targets. Annu Rev Pharmacol Toxicol 44, 219238 14. Lawson, N D., Scheer, N., Pham, V. N., Kim, C. H., Chitnis, A. B., Campos Ortega, J. A., and Weinstein, B. M. (2001) Notch signaling is required for arterial venous differentiation during embryonic vascular development. Development 128, 36753683

PAGE 108

108 15. Baron, M., A slam, H., Flasza, M., Fostier, M., Higgs, J. E., Mazaleyrat, S. L., and Wilkin, M. B. (2002) Multiple levels of Notch signal regulation (review). Mol Membr Biol 19, 2738 16. Iso, T., Hamamori, Y., and Kedes, L. (2003) Notch signaling in vascular developm ent. Arterioscler Thromb Vasc Biol 23, 543553 17. Ogawa, K., Pasqualini, R., Lindberg, R. A., Kain, R., Freeman, A. L., and Pasquale, E. B. (2000) The ephrinA1 ligand and its receptor, EphA2, are expressed during tumor neovascularization. Oncogene 19, 60436052 18. Cheng, N., Brantley, D. M., and Chen, J. (2002) The ephrins and Eph receptors in angiogenesis. Cytokine Growth Factor Rev 13, 7585 19. Pola, R., Ling, L. E., Silver, M., Corbley, M. J., Kearney, M., Blake Pepinsky, R., Shapiro, R., Taylor, F. R., Baker, D. P., Asahara, T., and Isner, J. M. (2001) The morphogen Sonic hedgehog is an indirect angiogenic agent upregulating two families of angiogenic growth factors. Nat Med 7, 706711 20. Lee, S. H., Schloss, D. J., Jarvis, L., Krasnow, M. A., a nd Swain, J. L. (2001) Inhibition of angiogenesis by a mouse sprouty protein. J Biol Chem 276, 41284133 21. Kidd, T., Brose, K., Mitchell, K. J., Fetter, R. D., Tessier Lavigne, M., Goodman, C. S., and Tear, G. (1998) Roundabout controls axon crossing of the CNS midline and defines a novel subfamily of evolutionarily conserved guidance receptors. Cell 92, 205215 22. Huminiecki, L., Gorn, M., Suchting, S., Poulsom, R., and Bicknell, R. (2002) Magic roundabout is a new member of the roundabout receptor fa mily that is endothelial specific and expressed at sites of active angiogenesis. Genomics 79, 547552 23. Hicklin, D. J., and Ellis, L. M. (2005) Role of the vascular endothelial growth factor pathway in tumor growth and angiogenesis. J Clin Oncol 23, 10111027 24. Shibuya, M., Yamaguchi, S., Yamane, A., Ikeda, T., Tojo, A., Matsushime, H., and Sato, M. (1990) Nucleotide sequence and expression of a novel human receptor type tyrosine kinase gene (flt) closely related to the fms family. Oncogene 5, 519524 25. Matthews, W., Jordan, C. T., Gavin, M., Jenkins, N. A., Copeland, N. G., and Lemischka, I. R. (1991) A receptor tyrosine kinase cDNA isolated from a population of enriched primitive hematopoietic cells and exhibiting close genetic linkage to c kit. P roc Natl Acad Sci U S A 88, 90269030 26. Hiratsuka, S., Minowa, O., Kuno, J., Noda, T., and Shibuya, M. (1998) Flt 1 lacking the tyrosine kinase domain is sufficient for normal development and angiogenesis in mice. Proc Natl Acad Sci U S A 95, 93499354

PAGE 109

109 27. Dunk, C., and Ahmed, A. (2001) Vascular endothelial growth factor receptor 2mediated mitogenesis is negatively regulated by vascular endothelial growth factor receptor 1 in tumor epithelial cells. Am J Pathol 158, 265273 28. Millauer, B., Wizigmann Voos, S., Schnurch, H., Martinez, R., Moller, N. P., Risau, W., and Ullrich, A. (1993) High affinity VEGF binding and developmental expression suggest Flk 1 as a major regulator of vasculogenesis and angiogenesis. Cell 72, 835846 29. Zeng, H., Dvorak, H F., and Mukhopadhyay, D. (2001) Vascular permeability factor (VPF)/vascular endothelial growth factor (VEGF) peceptor1 downmodulates VPF/VEGF receptor 2 mediated endothelial cell proliferation, but not migration, through phosphatidylinositol 3 kinased ependent pathways. J Biol Chem 276, 2696926979 30. Dvorak, H. F. (2002) Vascular permeability factor/vascular endothelial growth factor: a critical cytokine in tumor angiogenesis and a potential target for diagnosis and therapy. J Clin Oncol 20, 43684380 31. Ogawa, S., Oku, A., Sawano, A., Yamaguchi, S., Yazaki, Y., and Shibuya, M. (1998) A novel type of vascular endothelial growth factor, VEGF E (NZ7 VEGF), preferentially utilizes KDR/Flk 1 receptor and carries a potent mitotic activity without heparin binding domain. J Biol Chem 273, 3127331282 32. Kaipainen, A., Korhonen, J., Mustonen, T., van Hinsbergh, V. W., Fang, G. H., Dumont, D., Breitman, M., and Alitalo, K. (1995) Expression of the fms like tyrosine kinase 4 gene becomes restricted to lymph atic endothelium during development. Proc Natl Acad Sci U S A 92, 35663570 33. Paavonen, K., Puolakkainen, P., Jussila, L., Jahkola, T., and Alitalo, K. (2000) Vascular endothelial growth factor receptor 3 in lymphangiogenesis in wound healing. Am J Path ol 156, 14991504 34. Matsumura, K., Hirashima, M., Ogawa, M., Kubo, H., Hisatsune, H., Kondo, N., Nishikawa, S., Chiba, T., and Nishikawa, S. (2003) Modulation of VEGFR 2mediated endothelial cell activity by VEGF C/VEGFR3. Blood 101, 13671374 35. Orn itz, D. M., and Itoh, N. (2001) Fibroblast growth factors. Genome Biol 2, REVIEWS3005 36. Bottcher, R. T., and Niehrs, C. (2005) Fibroblast growth factor signaling during early vertebrate development. Endocr Rev 26, 6377 37. Cao, R., Brakenhielm, E., Pa wliuk, R., Wariaro, D., Post, M. J., Wahlberg, E., Leboulch, P., and Cao, Y. (2003) Angiogenic synergism, vascular stability and improvement of hindlimb ischemia by a combination of PDGF BB and FGF 2. Nat Med 9, 604613

PAGE 110

110 38. Marchuk, D. A., Srinivasan, S., Squire, T. L., and Zawistowski, J. S. (2003) Vascular morphogenesis: tales of two syndromes. Hum Mol Genet 12 Spec No 1, R97112 39. Derynck, R., and Zhang, Y. E. (2003) Smaddependent and Smadindependent pathways in TGF beta family signalling. Nature 425, 577584 40. Massague, J. (1998) TGF beta signal transduction. Annu Rev Biochem 67, 753791 41. Chang, H., Brown, C. W., and Matzuk, M. M. (2002) Genetic analysis of the mammalian transforming growth factor beta superfamily. Endocr Rev 23, 787823 42. Roberts, A. B., and Sporn, M. B. (1993) Physiological actions and clinical applications of transforming growth factor beta (TGF beta). Growth Factors 8, 19 43. Piek, E., Heldin, C. H., and Ten Dijke, P. (1999) Specificity, diversity, and regulation in TGF beta superfamily signaling. Faseb J 13, 21052124 44. Bertolino, P., Deckers, M., Lebrin, F., and ten Dijke, P. (2005) Transforming growth factorbeta signal transduction in angiogenesis and vascular disorders. Chest 128, 585S 590S 45. Heldin, C. H., Miyazono, K., and ten Dijke, P. (1997) TGF beta signalling from cell membrane to nucleus through SMAD proteins. Nature 390, 465471 46. Attisano, L., and Wrana, J. L. (1998) Mads and Smads in TGF beta signalling. Curr Opin Cell Biol 10, 188194 47. Kre tzschmar, M., and Massague, J. (1998) SMADs: mediators and regulators of TGF beta signaling. Curr Opin Genet Dev 8, 103111 48. ten Dijke, P., and Hill, C. S. (2004) New insights into TGF beta Smad signalling. Trends Biochem Sci 29, 265273 49. Massague, J. (2000) How cells read TGF beta signals. Nat Rev Mol Cell Biol 1, 169178 50. Itoh, S., Itoh, F., Goumans, M. J., and Ten Dijke, P. (2000) Signaling of transforming growth factor beta family members through Smad proteins. Eur J Biochem 267, 69546967 51. Sullivan, D. C., and Bicknell, R. (2003) New molecular pathways in angiogenesis. Br J Cancer 89, 228231 52. Uyttendaele, H., Ho, J., Rossant, J., and Kitajewski, J. (2001) Vascular patterning defects associated with expression of activated Notch4 in embryonic endothelium. Proc Natl Acad Sci U S A 98, 56435648

PAGE 111

111 53. Mailhos, C., Modlich, U., Lewis, J., Harris, A., Bicknell, R., and IshHorowicz, D. (2001) Delta4, an endothelial specific notch ligand expressed at sites of physiological and tumor angiogenesis. Differentiation 69, 135144 54. Shutter, J. R., Scully, S., Fan, W., Richards, W. G., Kitajewski, J., Deblandre, G. A., Kintner, C. R., and Stark, K. L. (2000) Dll4, a novel Notch ligand expressed in arterial endothelium. Genes Dev 14, 13131318 55. Zimrin, A. B., Pepper, M. S., McMahon, G. A., Nguyen, F., Montesano, R., and Maciag, T. (1996) An antisense oligonucleotide to the notch ligand jagged enhances fibroblast growth factor induced angiogenesis in vitro. J Biol Chem 271, 3249932502 56. Liu, Z. J., Shirakawa, T., Li, Y., Soma, A., Oka, M., Dotto, G. P., Fairman, R. M., Velazquez, O. C., and Herlyn, M. (2003) Regulation of Notch1 and Dll4 by vascular endothelial growth factor in arterial endothelial cells: implications for modulating arterioge nesis and angiogenesis. Mol Cell Biol 23, 1425 57. Taylor, K. L., Henderson, A. M., and Hughes, C. C. (2002) Notch activation during endothelial cell network formation in vitro targets the basic HLH transcription factor HESR 1 and downregulates VEGFR 2/K DR expression. Microvasc Res 64, 372383 58. Noguera Troise, I., Daly, C., Papadopoulos, N. J., Coetzee, S., Boland, P., Gale, N. W., Lin, H. C., Yancopoulos, G. D., and Thurston, G. (2006) Blockade of Dll4 inhibits tumour growth by promoting nonproducti ve angiogenesis. Nature 444, 10321037 59. Pandey, A., Shao, H., Marks, R. M., Polverini, P. J., and Dixit, V. M. (1995) Role of B61, the ligand for the Eck receptor tyrosine kinase, in TNF alpha induced angiogenesis. Science 268, 567569 60. Stein, E., Lane, A. A., Cerretti, D. P., Schoecklmann, H. O., Schroff, A. D., Van Etten, R. L., and Daniel, T. O. (1998) Eph receptors discriminate specific ligand oligomers to determine alternative signaling complexes, attachment, and assembly responses. Genes Dev 12, 667678 61. Brose, K., Bland, K. S., Wang, K. H., Arnott, D., Henzel, W., Goodman, C. S., Tessier Lavigne, M., and Kidd, T. (1999) Slit proteins bind Robo receptors and have an evolutionarily conserved role in repulsive axon guidance. Cell 96, 795806 62. Huminiecki, L., and Bicknell, R. (2000) In silico cloning of novel endothelial specific genes. Genome Res 10, 17961806 63. Waite, K. A., and Eng, C. (2003) From developmental disorder to heritable cancer: it's all in the BMP/TGF beta family. Nat Rev Genet 4, 763773

PAGE 112

112 64. Tsuchida, K., Nakatani, M., Uezumi, A., Murakami, T., and Cui, X. (2008) Signal transduction pathway through activin receptors as a therapeutic target of musculoskeletal diseases and cancer. Endocr J 55, 1121 65. Shovlin, C. L., Guttmacher, A. E., Buscarini, E., Faughnan, M. E., Hyland, R. H., Westermann, C. J., Kjeldsen, A. D., and Plauchu, H. (2000) Diagnostic criteria for hereditary hemorrhagic telangiectasia (Rendu Osler Weber syndrome). Am J Med Genet 91, 6667 66. Abdalla, S. A., and Letarte, M. (2006) Hereditary haemorrhagic telangiectasia: current views on genetics and mechanisms of disease. J Med Genet 43, 97110 67. Bideau, A., Plauchu, H., Brunet, G., and Robert, J. (1989) Epidemiological investigation of RenduOsler dis ease in France: its geographical distribution and prevalence. Popul 44, 322 68. Kjeldsen, A. D., Vase, P., and Green, A. (1999) Hereditary haemorrhagic telangiectasia: a population based study of prevalence and mortality in Danish patients. J Intern Med 245, 3139 69. Dakeishi, M., Shioya, T., Wada, Y., Shindo, T., Otaka, K., Manabe, M., Nozaki, J., Inoue, S., and Koizumi, A. (2002) Genetic epidemiology of hereditary hemorrhagic telangiectasia in a local community in the northern part of Japan. Hum Mutat 19, 140148 70. McAllister, K. A., Grogg, K. M., Johnson, D. W., Gallione, C. J., Baldwin, M. A., Jackson, C. E., Helmbold, E. A., Markel, D. S., McKinnon, W. C., Murrell, J., and et al. (1994) Endoglin, a TGF beta binding protein of endothelial cells, i s the gene for hereditary haemorrhagic telangiectasia type 1. Nat Genet 8, 345351 71. McDonald, M. T., Papenberg, K. A., Ghosh, S., Glatfelter, A. A., Biesecker, B. B., Helmbold, E. A., Markel, D. S., Zolotor, A., McKinnon, W. C., Vanderstoep, J. L., and et al. (1994) A disease locus for hereditary haemorrhagic telangiectasia maps to chromosome 9q3334. Nat Genet 6, 197204 72. Heutink, P., Haitjema, T., Breedveld, G. J., Janssen, B., Sandkuijl, L. A., Bontekoe, C. J., Westerman, C. J., and Oostra, B. A. (1994) Linkage of hereditary haemorrhagic telangiectasia to chromosome 9q34 and evidence for locus heterogeneity. J Med Genet 31, 933936 73. Shovlin, C. L., Hughes, J. M., Tuddenham, E. G., Temperley, I., Perembelon, Y. F., Scott, J., Seidman, C. E., and Seidman, J. G. (1994) A gene for hereditary haemorrhagic telangiectasia maps to chromosome 9q3. Nat Genet 6, 205209

PAGE 113

113 74. Johnson, D. W., Berg, J. N., Gallione, C. J., McAllister, K. A., Warner, J. P., Helmbold, E. A., Markel, D. S., Jackson, C. E., Porteous, M. E., and Marchuk, D. A. (1995) A second locus for hereditary hemorrhagic telangiectasia maps to chromosome 12. Genome Res 5, 2128 75. Vincent, P., Plauchu, H., Hazan, J., Faure, S., Weissenbach, J., and Godet, J. (1995) A third locus for heredi tary haemorrhagic telangiectasia maps to chromosome 12q. Hum Mol Genet 4, 945949 76. Johnson, D. W., Berg, J. N., Baldwin, M. A., Gallione, C. J., Marondel, I., Yoon, S. J., Stenzel, T. T., Speer, M., Pericak Vance, M. A., Diamond, A., Guttmacher, A. E., Jackson, C. E., Attisano, L., Kucherlapati, R., Porteous, M. E., and Marchuk, D. A. (1996) Mutations in the activin receptor like kinase 1 gene in hereditary haemorrhagic telangiectasia type 2. Nat Genet 13, 189195 77. Gallione, C. J., Repetto, G. M., L egius, E., Rustgi, A. K., Schelley, S. L., Tejpar, S., Mitchell, G., Drouin, E., Westermann, C. J., and Marchuk, D. A. (2004) A combined syndrome of juvenile polyposis and hereditary haemorrhagic telangiectasia associated with mutations in MADH4 (SMAD4). L ancet 363, 852859 78. Govani, F. S., and Shovlin, C. L. (2009) Hereditary haemorrhagic telangiectasia: a clinical and scientific review. Eur J Hum Genet 79. Guttmacher, A. E., Marchuk, D. A., and White, R. I., Jr. (1995) Hereditary hemorrhagic telangie ctasia. N Engl J Med 333, 918924 80. Sadick, H., Sadick, M., Gotte, K., Naim, R., Riedel, F., Bran, G., and Hormann, K. (2006) Hereditary hemorrhagic telangiectasia: an update on clinical manifestations and diagnostic measures. Wien Klin Wochenschr 118, 7280 81. Sabba, C., Gallitelli, M., Pasculli, G., Suppressa, P., Resta, F., and Tafaro, G. E. (2006) HHT: a rare disease with a broad spectrum of clinical aspects. Curr Pharm Des 12, 12171220 82. Braverman, I. M., Keh, A., and Jacobson, B. S. (1990) Ul trastructure and three dimensional organization of the telangiectases of hereditary hemorrhagic telangiectasia. J Invest Dermatol 95, 422 427 83. te Veldhuis, E. C., te Veldhuis, A. H., van Dijk, F. S., Kwee, M. L., van Hagen, J. M., Baart, J. A., and van der Waal, I. (2008) RenduOsler Weber disease: update of medical and dental considerations. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 105, e3841

PAGE 114

114 84. Sabba, C. (2005) A rare and misdiagnosed bleeding disorder: hereditary hemorrhagic telangiectas ia. J Thromb Haemost 3, 22012210 85. Pau, H., Carney, A. S., and Murty, G. E. (2001) Hereditary haemorrhagic telangiectasia (Osler Weber Rendu syndrome): otorhinolaryngological manifestations. Clin Otolaryngol Allied Sci 26, 9398 86. Cottin, V., Chinet T., Lavole, A., Corre, R., Marchand, E., ReynaudGaubert, M., Plauchu, H., and Cordier, J. F. (2007) Pulmonary arteriovenous malformations in hereditary hemorrhagic telangiectasia: a series of 126 patients. Medicine (Baltimore) 86, 117 87. Peery, W. H. (1987) Clinical spectrum of hereditary hemorrhagic telangiectasia (Osler Weber Rendu disease). Am J Med 82, 989997 88. Begbie, M. E., Wallace, G. M., and Shovlin, C. L. (2003) Hereditary haemorrhagic telangiectasia (Osler Weber Rendu syndrome): a view f rom the 21st century. Postgrad Med J 79, 1824 89. ten Dijke, P., Ichijo, H., Franzen, P., Schulz, P., Saras, J., Toyoshima, H., Heldin, C. H., and Miyazono, K. (1993) Activin receptor like kinases: a novel subclass of cell surface receptors with predicte d serine/threonine kinase activity. Oncogene 8, 28792887 90. ten Dijke, P., Yamashita, H., Ichijo, H., Franzen, P., Laiho, M., Miyazono, K., and Heldin, C. H. (1994) Characterization of type I receptors for transforming growth factor beta and activin. Sc ience 264, 101104 91. Attisano, L., Carcamo, J., Ventura, F., Weis, F. M., Massague, J., and Wrana, J. L. (1993) Identification of human activin and TGF beta type I receptors that form heteromeric kinase complexes with type II receptors. Cell 75, 671680 92. Lux, A., Attisano, L., and Marchuk, D. A. (1999) Assignment of transforming growth factor beta1 and beta3 and a third new ligand to the type I receptor ALK 1. J Biol Chem 274, 99849992 93. Scharpfenecker, M., van Dinther, M., Liu, Z., van Bezooijen, R. L., Zhao, Q., Pukac, L., Lowik, C. W., and ten Dijke, P. (2007) BMP 9 signals via ALK1 and inhibits bFGF induced endothelial cell proliferation and VEGF stimulated angiogenesis. J Cell Sci 120, 964972 94. David, L., Mallet, C., Mazerbourg, S., Feige J. J., and Bailly, S. (2007) Identification of BMP9 and BMP10 as functional activators of the orphan activin receptor like kinase 1 (ALK1) in endothelial cells. Blood 109, 19531961

PAGE 115

115 95. David, L., Mallet, C., Keramidas, M., Lamande, N., Gasc, J. M., Dupuis Girod, S., Plauchu, H., Feige, J. J., and Bailly, S. (2008) Bone morphogenetic protein9 is a circulating vascular quiescence factor. Circ Res 102, 914922 96. Chen, Y. G., and Massague, J. (1999) Smad1 recognition and activation by the ALK1 group of transforming growth factor beta family receptors. J Biol Chem 274, 36723677 97. Oh, S. P., Seki, T., Goss, K. A., Imamura, T., Yi, Y., Donahoe, P. K., Li, L., Miyazono, K., ten Dijke, P., Kim, S., and Li, E. (2000) Activin receptor like kinase 1 modulat es transforming growth factor beta 1 signaling in the regulation of angiogenesis. Proc Natl Acad Sci U S A 97, 26262631 98. Ota, T., Fujii, M., Sugizaki, T., Ishii, M., Miyazawa, K., Aburatani, H., and Miyazono, K. (2002) Targets of transcriptional regul ation by two distinct type I receptors for transforming growth factor beta in human umbilical vein endothelial cells. J Cell Physiol 193, 299318 99. Urness, L. D., Sorensen, L. K., and Li, D. Y. (2000) Arteriovenous malformations in mice lacking activin receptor like kinase1. Nat Genet 26, 328331 100. Srinivasan, S., Hanes, M. A., Dickens, T., Porteous, M. E., Oh, S. P., Hale, L. P., and Marchuk, D. A. (2003) A mouse model for hereditary hemorrhagic telangiectasia (HHT) type 2. Hum Mol Genet 12, 473482 101. Cole, S. G., Begbie, M. E., Wallace, G. M., and Shovlin, C. L. (2005) A new locus for hereditary haemorrhagic telangiectasia (HHT3) maps to chromosome 5. J Med Genet 42, 577582 102. Bayrak Toydemir, P., McDonald, J., Akarsu, N., Toydemir, R. M., Calderon, F., Tuncali, T., Tang, W., Miller, F., and Mao, R. (2006) A fourth locus for hereditary hemorrhagic telangiectasia maps to chromosome 7. Am J Med Genet A 140, 21552162 103. ten Dijke, P., and Arthur, H. M. (2007) Extracellular control of TGFbet a signalling in vascular development and disease. Nat Rev Mol Cell Biol 8, 857869 104. Rahimi, R. A., and Leof, E. B. (2007) TGF beta signaling: a tale of two responses. J Cell Biochem 102, 593608 105. Roelen, B. A., van Rooijen, M. A., and Mummery, C. L. (1997) Expression of ALK 1, a type 1 serine/threonine kinase receptor, coincides with sites of vasculogenesis and angiogenesis in early mouse development. Dev Dyn 209, 418430 106. Seki, T., Yun, J., and Oh, S. P. (2003) Arterial endothelium specific activin receptor like kinase 1 expression suggests its role in arterialization and vascular remodeling. Circ Res 93, 682689

PAGE 116

116 107. Goumans, M. J., Valdimarsdottir, G., Itoh, S., Rosendahl, A., Sideras, P., and ten Dijke, P. (2002) Balancing the activation s tate of the endothelium via two distinct TGF beta type I receptors. Embo J 21, 17431753 108. Goumans, M. J., Lebrin, F., and Valdimarsdottir, G. (2003) Controlling the angiogenic switch: a balance between two distinct TGF b receptor signaling pathways. T rends Cardiovasc Med 13, 301 307 109. Fernandez, L. A., Sanz Rodriguez, F., Blanco, F. J., Bernabeu, C., and Botella, L. M. (2006) Hereditary hemorrhagic telangiectasia, a vascular dysplasia affecting the TGF beta signaling pathway. Clin Med Res 4, 6678 110. Li, D. Y., Sorensen, L. K., Brooke, B. S., Urness, L. D., Davis, E. C., Taylor, D. G., Boak, B. B., and Wendel, D. P. (1999) Defective angiogenesis in mice lacking endoglin. Science 284, 15341537 111. Bourdeau, A., Dumont, D. J., and Letarte, M. (1999) A murine model of hereditary hemorrhagic telangiectasia. J Clin Invest 104, 13431351 112. Arthur, H. M., Ure, J., Smith, A. J., Renforth, G., Wilson, D. I., Torsney, E., Charlton, R., Parums, D. V., Jowett, T., Marchuk, D. A., Burn, J., and Diamond, A. G. (2000) Endoglin, an ancillary TGFbeta receptor, is required for extraembryonic angiogenesis and plays a key role in heart development. Dev Biol 217, 4253 113. Satomi, J., Mount, R. J., Toporsian, M., Paterson, A. D., Wallace, M. C., Harrison, R. V ., and Letarte, M. (2003) Cerebral vascular abnormalities in a murine model of hereditary hemorrhagic telangiectasia. Stroke 34, 783789 114. Torsney, E., Charlton, R., Diamond, A. G., Burn, J., Soames, J. V., and Arthur, H. M. (2003) Mouse model for here ditary hemorrhagic telangiectasia has a generalized vascular abnormality. Circulation 107, 16531657 115. Park, S. O., Lee, Y. J., Seki, T., Hong, K. H., Fliess, N., Jiang, Z., Park, A., Wu, X., Kaartinen, V., Roman, B. L., and Oh, S. P. (2008) ALK5and TGFBR2 independent role of ALK1 in the pathogenesis of hereditary hemorrhagic telangiectasia type 2. Blood 111, 633642 116. Lampugnani, M. G. (1999) Cell migration into a wounded area in vitro. Methods Mol Biol 96, 177 182 117. Staton, C. A., Stribbling, S. M., Tazzyman, S., Hughes, R., Brown, N. J., and Lewis, C. E. (2004) Current methods for assaying angiogenesis in vitro and in vivo. Int J Exp Pathol 85, 233248

PAGE 117

117 118. Davis, G. E., and Senger, D. R. (2005) Endothelial extracellular matrix: biosynthesi s, remodeling, and functions during vascular morphogenesis and neovessel stabilization. Circ Res 97, 10931107 119. Passaniti, A., Taylor, R. M., Pili, R., Guo, Y., Long, P. V., Haney, J. A., Pauly, R. R., Grant, D. S., and Martin, G. R. (1992) A simple, quantitative method for assessing angiogenesis and antiangiogenic agents using reconstituted basement membrane, heparin, and fibroblast growth factor. Lab Invest 67, 519 528 120. David, L., Mallet, C., Vailhe, B., Lamouille, S., Feige, J. J., and Bailly, S. (2007) Activin receptorlike kinase 1 inhibits human microvascular endothelial cell migration: potential roles for JNK and ERK. J Cell Physiol 213, 484489 121. Liu, D., Wang, J., Kinzel, B., Mueller, M., Mao, X., Valdez, R., Liu, Y., and Li, E. (2007) Dosage dependent requirement of BMP type II receptor for maintenance of vascular integrity. Blood 110, 15021510 122. Lamouille, S., Mallet, C., Feige, J. J., and Bailly, S. (2002) Activin receptor like kinase 1 is implicated in the maturation phase of a ngiogenesis. Blood 100, 44954501 123. Roman, B. L., Pham, V. N., Lawson, N. D., Kulik, M., Childs, S., Lekven, A. C., Garrity, D. M., Moon, R. T., Fishman, M. C., Lechleider, R. J., and Weinstein, B. M. (2002) Disruption of acvrl1 increases endothelial c ell number in zebrafish cranial vessels. Development 129, 30093019 124. Wu, X., Ma, J., Han, J. D., Wang, N., and Chen, Y. G. (2006) Distinct regulation of gene expression in human endothelial cells by TGF beta and its receptors. Microvasc Res 71, 1219 125. Carlson, T. R., Yan, Y., Wu, X., Lam, M. T., Tang, G. L., Beverly, L. J., Messina, L. M., Capobianco, A. J., Werb, Z., and Wang, R. (2005) Endothelial expression of constitutively active Notch4 elicits reversible arteriovenous malformations in adult mice. Proc Natl Acad Sci U S A 102, 98849889 126. Murphy, P. A., Lam, M. T., Wu, X., Kim, T. N., Vartanian, S. M., Bollen, A. W., Carlson, T. R., and Wang, R. A. (2008) Endothelial Notch4 signaling induces hallmarks of brain arteriovenous malformations i n mice. Proc Natl Acad Sci U S A 105, 1090110906 127. Gridley, T. (2007) Notch signaling in vascular development and physiology. Development 134, 27092718 128. Roca, C., and Adams, R. H. (2007) Regulation of vascular morphogenesis by Notch signaling. G enes Dev 21, 25112524

PAGE 118

118 129. Tsuchida, K. (2004) Activins, myostatin and related TGF beta family members as novel therapeutic targets for endocrine, metabolic and immune disorders. Curr Drug Targets Immune Endocr Metabol Disord 4, 157166 130. Harrison, C A., Gray, P. C., Vale, W. W., and Robertson, D. M. (2005) Antagonists of activin signaling: mechanisms and potential biological applications. Trends Endocrinol Metab 16, 7378 131. Reissmann, E., Jornvall, H., Blokzijl, A., Andersson, O., Chang, C., Minchiotti, G., Persico, M. G., Ibanez, C. F., and Brivanlou, A. H. (2001) The orphan receptor ALK7 and the Activin receptor ALK4 mediate signaling by Nodal proteins during vertebrate development. Genes Dev 15, 20102022 132. Tsuchida, K., Nakatani, M., Yama kawa, N., Hashimoto, O., Hasegawa, Y., and Sugino, H. (2004) Activin isoforms signal through type I receptor serine/threonine kinase ALK7. Mol Cell Endocrinol 220, 5965 133. Miyazono, K., Maeda, S., and Imamura, T. (2005) BMP receptor signaling: transcriptional targets, regulation of signals, and signaling cross talk. Cytokine Growth Factor Rev 16, 251263 134. Engel, M. E., McDonnell, M. A., Law, B. K., and Moses, H. L. (1999) Interdependent SMAD and JNK signaling in transforming growth factor betamedi ated transcription. J Biol Chem 274, 3741337420 135. Yu, L., Hebert, M. C., and Zhang, Y. E. (2002) TGF beta receptor activated p38 MAP kinase mediates Smad independent TGF beta responses. Embo J 21, 37493759 136. Bakin, A. V., Tomlinson, A. K., Bhowmi ck, N. A., Moses, H. L., and Arteaga, C. L. (2000) Phosphatidylinositol 3 kinase function is required for transforming growth factor beta mediated epithelial to mesenchymal transition and cell migration. J Biol Chem 275, 3680336810 137. Vinals, F., and P ouyssegur, J. (2001) Transforming growth factor beta1 (TGF beta1) promotes endothelial cell survival during in vitro angiogenesis via an autocrine mechanism implicating TGF alpha signaling. Mol Cell Biol 21, 72187230 138. de Caestecker, M. P., Parks, W. T., Frank, C. J., Castagnino, P., Bottaro, D. P., Roberts, A. B., and Lechleider, R. J. (1998) Smad2 transduces common signals from receptor serine threonine and tyrosine kinases. Genes Dev 12, 15871592 139. Kretzschmar, M., Doody, J., Timokhina, I., and Massague, J. (1999) A mechanism of repression of TGFbeta/ Smad signaling by oncogenic Ras. Genes Dev 13, 804816

PAGE 119

119 140. Funaba, M., Zimmerman, C. M., and Mathews, L. S. (2002) Modulation of Smad2mediated signaling by extracellular signal regulated kinase. J Biol Chem 277, 4136141368 141. Xu, X., Han, J., Ito, Y., Bringas, P., Jr., Deng, C., and Chai, Y. (2008) Ectodermal Smad4 and p38 MAPK are functionally redundant in mediating TGF beta/BMP signaling during tooth and palate development. Dev Cell 15, 322 329 142. Papetti, M., and Herman, I. M. (2002) Mechanisms of normal and tumor derived angiogenesis. Am J Physiol Cell Physiol 282, C947970 143. Bergers, G., and Benjamin, L. E. (2003) Tumorigenesis and the angiogenic switch. Nat Rev Cancer 3, 401410 144. Nagy, J. A., Chang, S. H., Dvorak, A. M., and Dvorak, H. F. (2009) Why are tumour blood vessels abnormal and why is it important to know? Br J Cancer 100, 865869 145. Folkman, J. (1971) Tumor angiogenesis: therapeutic implications. N Engl J Med 285, 11821186 146. Ferrara, N., Gerber, H. P., and LeCouter, J. (2003) The biology of VEGF and its receptors. Nat Med 9, 669676 147. Ferrara, N. (2004) Vascular endothelial growth factor: basic science and clinical progress. Endocr Rev 25, 581611 148. Fe rrara, N., and Kerbel, R. S. (2005) Angiogenesis as a therapeutic target. Nature 438, 967974 149. Jain, R. K., Duda, D. G., Clark, J. W., and Loeffler, J. S. (2006) Lessons from phase III clinical trials on anti VEGF therapy for cancer. Nat Clin Pract On col 3, 2440 150. Morabito, A., De Maio, E., Di Maio, M., Normanno, N., and Perrone, F. (2006) Tyrosine kinase inhibitors of vascular endothelial growth factor receptors in clinical trials: current status and future directions. Oncologist 11, 753764 151. Kerbel, R. S. (2006) Antiangiogenic therapy: a universal chemosensitization strategy for cancer? Science 312, 11711175 152. Klement, G., Baruchel, S., Rak, J., Man, S., Clark, K., Hicklin, D. J., Bohlen, P., and Kerbel, R. S. (2000) Continuous low dose therapy with vinblastine and VEGF receptor 2 antibody induces sustained tumor regression without overt toxicity. J Clin Invest 105, R15 24 153. Kerbel, R., and Folkman, J. (2002) Clinical translation of angiogenesis inhibitors. Nat Rev Cancer 2, 727739

PAGE 120

120 154. Yu, J. L., Rak, J. W., Coomber, B. L., Hicklin, D. J., and Kerbel, R. S. (2002) Effect of p53 status on tumor response to antiangiogenic therapy. Science 295, 15261528 155. Casanovas, O., Hicklin, D. J., Bergers, G., and Hanahan, D. (2005) Drug resi stance by evasion of antiangiogenic targeting of VEGF signaling in late stage pancreatic islet tumors. Cancer Cell 8, 299309 156. Rusnati, M., and Presta, M. (2007) Fibroblast growth factors/fibroblast growth factor receptors as targets for the development of anti angiogenesis strategies. Curr Pharm Des 13, 20252044 157. Suchting, S., Heal, P., Tahtis, K., Stewart, L. M., and Bicknell, R. (2005) Soluble Robo4 receptor inhibits in vivo angiogenesis and endothelial cell migration. Faseb J 19, 121123 158. Heroult, M., Schaffner, F., and Augustin, H. G. (2006) Eph receptor and ephrin ligandmediated interactions during angiogenesis and tumor progression. Exp Cell Res 312, 642650 159. Yan, M., and Plowman, G. D. (2007) Delta like 4/Notch signaling and its therapeutic implications. Clin Cancer Res 13, 72437246 160. Bergers, G., Song, S., Meyer Morse, N., Bergsland, E., and Hanahan, D. (2003) Benefits of targeting both pericytes and endothelial cells in the tumor vasculature with kinase inhibitors. J Clin Invest 111, 12871295 161. Seki, T., Hong, K. H., and Oh, S. P. (2006) Nonoverlapping expression patterns of ALK1 and ALK5 reveal distinct roles of each receptor in vascular development. Lab Invest 86, 116129

PAGE 121

121 BIOGRAPHICAL SKETCH Eun Jung Choi was bor n in 1976 in Seoul, the capital city of Korea. Upon graduating from high school in 1995, EunJung entered the KONKUK University in Seoul where she majored in Animal Science. During her undergraduate study, EunJung joined the laboratory of Poultry and Nutr ition Sciences under Dr. ChangWon K angs mentoring for two years. In her senior year, Eun Jung worked in National Institute of Animal Science in Suwon. And she earned her Bachelor of Science degree in Animal and Life Sciences in 2001. In the following yea r, EunJung entered the Master Program in Molecular Genetics and Microbiology at the University of Florida College of Medicine. Eun Jung joined Dr. Alfred S. Lewins laboratory and studied the ribozyme mediated gene therapy for eye diseases. She completed her master thesis titled Comparison of the Effects of a Processing Sequence and a Nuclear Export Element on Ribozyme Activity in Transfected Cells and received her Ma ster of Science degree in 2004. In the same year, to pursue her Ph.D., EunJung joined t he Interdisciplinary Program in Biomedical Sciences at the University of Florida College of Medicine. Since 2005, EunJung has been working in the laboratory of Dr. S. Paul Oh in the Department of Physi ology and Functional Genomics. Under Dr. Ohs supervis ion, she did her graduate work engaged in the in vitro and in vivo studies elucidating the role of a TGF Like Kinase 1 (ALK1). EunJung complete d her Ph.D. program in August of 2009.