1 DEVELOPMENT OF A NOVEL MOUSE MODEL TO STUDY THE MECHANISM OF HHT PATHOGENESIS By EUNJI LEE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008
2 2008 Eunji Lee
3 To my parents, Young-Min Lee and Yi-Soon Kim
4 ACKNOWLEDGMENTS I would like to thank m y mentor, Dr. S. Paul Oh, for all of his support for my graduate career. And I would like to expres s my gratitude to my committ ee members, Dr. Naohiro Terada and Dr. Gregory Schultz, for thei r support during my studies. I also wish to thank the members of Dr. Oh laboratory for their help and advice. I would also like to give my special thanks to my parents, Young-Min Lee and Yi-Soon Kim for their sincere support and encouragement.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........6 LIST OF FIGURES.........................................................................................................................7 ABSTRACT.....................................................................................................................................8 CHAP TER 1 INTRODUCTION..................................................................................................................11 Clinical Manifestations of HHT............................................................................................. 11 Genetics of HHT................................................................................................................ .....12 Signal Transduction of TGFFa mily Proteins.....................................................................14 Arterial Endothelium-Sp ecific Alk1 Expression .................................................................... 15 Role of Alk1 in Vascular Remodeling.................................................................................... 16 The Cre/loxP Site-Specific Recombination System...............................................................18 Transgenic Mouse Strains for Expressing Cre Recombinase in ECs..................................... 18 Experimental Scheme using Alk1+/GFPCre Mice....................................................................... 20 2 MATERIALS AND METHODS........................................................................................... 23 Generation and analysis of Alk1+/GFPCre m ice................................................................... 23 Mouse Breeding......................................................................................................................23 X-gal staining................................................................................................................. .........24 Histology and Immuno histochem istry.................................................................................... 24 PCR.........................................................................................................................................25 Genomic Southern............................................................................................................... ...26 3 RESULTS...............................................................................................................................29 Screening of Alk1+/GFPCre mouse strain................................................................................... 29 Characterization of the Cre Transgene Expression in the Alk1+/GFPCr Line............................29 Investigation of Alk1+/GFPCre Line as a Potential Endothelial Cre Line for Conditional KO Approaches...................................................................................................................33 4 DISCUSSION.........................................................................................................................46 LIST OF REFERENCES...............................................................................................................52 BIOGRAPHICAL SKETCH.........................................................................................................57
6 LIST OF TABLES Table page 2-1 Primers used in PCR........................................................................................................ ..28
7 LIST OF FIGURES Figure page 1-1 Signal transduction of TGFsuperfam ily........................................................................ 211-2 The Cre/LoxP Recombination System...............................................................................223-1 Procedures for the generation of Alk1+/GFPCre construct..................................................... 363-2 The lacZ expression in the blood vessel of Alk1+/GFPCre; R26R embryos............................ 373-3 Expression pattern of LacZ during embryonic stage E10.5 in Alk1+/GFPCre; R26R and Tg( Alk1 GFPCre);R26R embryos.......................................................................................383-4 The lacZ expression of tissues derived from Alk1+/GFPCre; R26R embryo at E15.5............ 393-5 The Cre recombinase is active in Alk1+/GFPCre; R26R newborn endothelium of blood vessel......................................................................................................................... .........413-6 Conditioanl deletion of Tgfbr2 gene by Alk1+/GFPCre line showed noticeable hemorrhage in the cranial region at E11.5......................................................................... 423-7 Histological analysis of Alk1+/GFPCre mediated conditional deletion of Tgfbr2 gene in the brain and heart at E11.5............................................................................................... 433-8 More severe but identical pattern of a hemorrhage was observed in E13.5 embryos....... 443-9 Comparision of the phenotype of brain, placenta and heart between Alk1+/GFPCre; Tgfbr2f/f embryos and control one in the histological analysis......................45
8 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science DEVELOPMENT OF A NOVEL MOUSE MODEL TO STUDY THE MECHANISM OF HHT PATHOGENESIS By Eunji Lee May 2008 Chair: S. Paul Oh Major: Medical Sciences Hereditary hemorrhagic telangiectasia (HHT) or Rendu-Osler-Weber (OWR) syndrome is an autosomal-dominant vascular disorder w ith an incidence around 1/10 000. It is characterized primarily by epistaxis (spontane ous and recurrent nosebleeds), telangiectases and arteriovenous malformations (AVM). Arteriovenous malformations are characterized by direct communication of arterioles to venuoles wit hout intervening capillaries. Bl ood vessels in AVM are dilated, convoluted, and tortuous, and the vascular wall of these vessels often devoid of smooth muscle layer. Telangiectasia refers to small AVMs th at usually form in the mucosal layer of skin. Pulmonary, cerebral, and gastro intestinal AVMs can cause significant morbidity and mortality. The molecular pathogenetic mechanisms underlying AVM formation remain elusive. More than 300 heterozygous mutations have been recognized in ENG (Endoglin) and ALK1 (activin receptor-like kinase1 or ACVRL1 ) genes as the cause of HHT1 and HHT2, respectively. Both of these genes are plasma membrane proteins involved in the signal transduction of TGF(Transforming Growth Factor) superfamily proteins. In addition, mutations in SMAD4 a common mediator of TGFsignaling, are seen in patients with the combined syndrome of juvenile polyposis (JP) and HHT (JP-HHT). Si nce all three known HHT genes are involved in
9 TGFfamily signaling, it has been po stulated that impaired TGFfamily signaling underlie the pathogenesis of HHT. Alk1-deficient mice die in utero with severe vascular abnormalities, characterized by hyperdilation of large blood vessels and extreme fusion of the capillary plexus into cavernous vessels. Arterial endothelium-specific ALK1 expr ession patterns and thin, vein-like vessels in Alk1 -deficient embryos suggested that ALK1 may pl ay an essential role in arteriogenesis. However, the specific function of ALK1 in this biological process remains to be determined. Our laboratory has recently generated a novel mouse lin e in which GFPCre gene a fusion protein of GFP (Green Fluorescent Protein) and Cr e recombinasewas inserted into the Alk1 gene locus ( Alk1+/GFPCre) by the gene targeting method. These mice will enable us to distinguish ECs expressing ALK1 from ECs that do not express ALK1 in vivo using GFP-mediated cell sorting, and thereby we can identify the downstream ta rgets of ALK1 signaling. Another important utility of this mouse strain is to spatiote mporal expression of ALK1 during development, postnatal stages, and pathological processes. The first step toward these goals is to carefully characterize expression of GFPC re gene during development in order to establish strong foundation of this mouse strain for further use. Since our laboratory has previously generated additional mouse lines which express bacterial -galactosidase gene (lacZ) under the endogenous Alk1 promoter ( Alk1lacZ), comparison of expression pattern of transgene between Alk1+/GFPCre and Alk1lacZ will provide complete information regard ing ALK1 expression in vivo. To monitor GFPCre expression, I monitored Cre activities by crossing the Alk1+/GFPCre mice with R26R mice (a Cre reporter). The lacZ expression pattern in Alk1+/ GFPCre; R26R bigenic mice should reflect Cre activities, and show cell types expressed ALK 1. I found that lacZ-positive cells are mostly vascular endothelial cells. Unlike the pattern in Alk1lacZ, where lacZ-positive cells were
10 predominantly localized in arteri al endothelium, venous endothelial cells were also lacZ-positive in these bigenic lines, indicati ng that Alk1 is indeed expressed in venous endothelium at a lower level than arterial endothelium. I also found that the mesenchy mal cells of the atrioventricular (AV) cushions during heart development were also lacZ-positive, suggesting that Alk1+/GFPCremediated DNA recombination occurred prior to endothelial-mesenchymal transformation during AV cushion formation. In terms of the tempor al expression, lacZ-positive cells were first detected in embryonic day (E) 9.5 in a patch stai ning pattern, but become uniformly detected in vascular endothelial cells by E10.5. This is one or two days delay compare to endogenous ALK1 expression observed by Alk1lacZ mice. Accounting for the f act that the Cre-medicated recombination requires time, these data show that GFPCre expression in this novel mouse strain is consistent with endogenous ALK1 expr ession in a spatiotemporal manner. I also present data showing th at this novel mouse line can be useful as anot her endothelial Cre line by characterizing mice in which Tgfbr2 (TGFtype II receptor) gene is conditionally deleted by Alk1+/GFPCre mice. Although due to limited time, I was unable to reach the ultimate goal of this project, my results demonstrate that Alk1+/GFPCre line is an important resource with which one can study ALK1 expression patterns in various pathogenic conditions, conditionally delete genes in the ALK1 expressing cel ls, and sort out cells expressing ALK1.
11 CHAPTER 1 INTRODUCTION Clinical Manifestations of HHT The hereditary hem orrhagic telangiect asia (HHT) is an under-diagnosed disease. Important reason for this is due to the variability of the disease symptoms in terms of the onset, location, severity, and types of clinical manifest ations. Epistaxis (recurr ent nose bleeds) is the most common symptom of HHT [2 ]. More than 90% of HHT pa tients are reported to have epistaxis in their life time. Telangiectasias (f ocal dilation of postcapill ary venule) in varying sizes which usually appear in the mucosal layer of nasal and oral cavity and nail beds are another common sign of HHT. More than 60% of HHT patients possess more than one arteriovenous malformations (AVMdirect connections between ar teries and veins) in their brain, lung, liver, and gastrointestinal tract (GI). HHT is a genetic disease, a nd thus most HHT patients have family history of these symptoms. HHT is diagnosed, if one has three of the four common symptoms--epistaxis, telangiectasia, visceral AVM, family history, AVMs are associated with high morbidity and mortality in HHT patients. Cerebral AVMs can cause severe headache and stroke, pulm onary AVMs can lead to cerebral abscess, liver AVMs can result in cirrhosis, and GI AVM s often leads to chronic hemorrhages. Many cases AVMs are benign, but having AVMs in a body is like to have a ti me bomb. Often HHT patients suffer from anemia, blood transfusions due to chronic hemorrhages. Embolism is the best treatment option for CAVM and PAVM when only a few AVMs formed, or transplantations are required for severe cases. Unfortunately, no therapeutic options are available for HHT patients to date. Fortunately, a few genes have been identified as HHT-causing (or predisposing) genes, including Alk1 as details below. Investigating the function of these genes, therefore, will provide important clues for development of therapeutic options for HHT patients.
12 Genetics of HHT HHT is a genetically heterogeneous di sorder that has been m apped to areas 9q33-q34.1 on chromosome 9 (HHT1) and 12q11-q14 on chro mosome 12 (HHT2). HHT 1 turned out to be Eng ( Endoglin ) whereas HHT 2 Alk1 (activin receptor-like kinase 1 or ACVRL1) . Both ENG and ALK1 are plasma membrane protei n involved in signal transduction of TGFsuperfamily signals. Haploinsufficiency (diminished levels of functional protein) appears to be the principal mechanism. Recen tly, it has been reported th at a subgroup of people having juvenile polyposis with SMAD4 mutations also displayed HHT symptoms . Furthermore, a new locus for HHT (HHT3) has been mapped to chromosome 5, and a fourth locus (HHT4) to chromosome 7. As many as 156 different ENG mutations have been reported. Of 14 coding exons, mutations have been observed in exons 112 coding for the extra cellular domain, but no mutations have been identified in exon13 and 14 coding transmembrane and cytoplasmic domains . All types of muta tions including deletions, missens e, splice sites, insertions, nonsense, and insertion+deletion have been found. A large number of ENG mutations would lead to structural unsteadiness, unsuccessful intracellular trafficki ng and loss of function, suggesting that haploinsufficiency as a mechanism of disease. The Alk1 gene covers more than 15kb of genomic DNA and the cDNA encodes a protein of 503 amino acids. ALK1 prot ein has the characteristics of a TGFtype I serinethreonine kinase receptor. Like other type I receptors, it ha s a relatively high extent of similarity in serine/threonine kinase subdomains a glycine/serine-rich (GS) region heading the intracellular kinase domain, a short C-terminal tail  and an N-terminal extracellular domain with 10 conserved cysteine resi dues. The coding region include s 9 exons, consisting of the start codon in exon 2 and the termination codon in exon 10. As many as 123 mutations in the
13 Alk1 gene have been reported. Nonsense, deleti ons, insertions, inserti on+deletion and splice site mutations were also identified in ALK1 mutations (the number of frequency is 16, 24, 13, 1 and 4, respectively). Unlike ENG, missense s ubstitution consist the highest portion of the mutations (53%) connected to ALK1. The extrac ellular domain occupied 20% mutations (n=25), intracellular kinase domain 75% (n=92) and tran smembrane 5% (n=6). 65% of all identified mutations were confirmed in exons 3, 7 and 8 . Because the majority of HHT patients have been found to have mutations in ENG or ALK1, it is probable that only a small fraction of the remaining cases will be caused by mutations in these additional gene s. Gallione et al screened 30 patients from unselected group of HHT patients, who had been referred for DNA based testing for HHT and veri fied to be negative for mutations in either ENG or ALK1, but had mutations in the SMAD4 gene, a key downstream effector of TGFsignaling transduction and present in all cell types. The region of the SMAD4 gene coding for the highly conserved carboxyl terminus accounts for these mutations. The presence of both juvenile polyposis and HHT in an affected individual defines the syndrome of juvenile polyposis and HHT (JPHT). Juvenile polyposis is distinguished by the presence of five or more om inous gastrointestinal po lyps, and was shown to be associated with mutations in either SMAD4 or BMPRIA. In a previously described HHT family no mu tations were identified in the ENG, ALK1, or SMAD4 genes. The putative HHT3 gene was recently linked to chromosome 5, and was mapped to regions 5q31.3q32. As all three iden tified HHT genes encoded proteins are involved in TGFsignaling pathway, the gene responsible for HHT3 is also anticipated to encode a protein involve d in SMAD dependent TGFsignaling. Furthermore, the presence of another gene causing HHT4 has been reported r ecently. A family displaye d classic features of
14 HHT, but with milder epistaxis and telangiec tasia phenotypes. A comp lete genome linkage investigation and fine mappi ng suggested this HHT4 gene was on a 7 Mb region on the chromosome 7 (7p14) . Signal Transduction of TGFFamily Proteins As described above, all three ge nes linked to HHT are involved in TG Fsignal transduction (Figure 1-1). TGFsuperfamily members are a large group of pleiotrophic secreted proteins, which can be grouped into several subfamilies, such as TGF, bone morphogenetic protein (BMP), activins/inhibin, a nd growth and differentiation factor (GDF). Each member is capable of effecting a diverse range of cellular processes including cell proliferation, migration, differentiation, lineage determin ation, pattern formation, adhe sion, and apoptosis. TGFfamily cytokines control gene expression by bind ing to heteromeric complexes of two types of transmembrane serine/threonine ki nase receptors. Five type II and seven type I receptors are known in mammals. The type I receptor and ty pe II receptor have the same fundamental structural elements: a cysteine-rich N-terminal extracellular domain, which is associated with ligand binding; a single transmembrane domain; and a C-terminal cytoplasmic kinase domain. Type I receptors have distinctive amino acid se quence at its cytoplasmic juxtamembrane area called the GS (glycine/serine-rich domain) domain . TGFsignaling is commenced by the binding of TGFligands to the type II receptors, which then recruit type I receptors and transphosp horylate at the GS domain of type I receptors. The activated type I receptors send out signals through phosphorylating receptor-regulated SMAD (R-SMAD) proteins, that is, SMAD1, SMAD2, SMAD3, SMAD5 and SMAD8. Phosphorylated R-SMADs form heteromeric complex with the coregulatory SMAD (SMAD4 in mammals), and this complex translocates into the nucleus, where they regulate target gene transcription by interaction with trascriptinal co-activators or co-repressors . A few
15 conserved amino acid residues in th e L45 loop of the kinase domain of the type I receptor and the L3 loop of the MH2 domain of R-SMADs de termines specific communication between the type I receptors and the R-SMADs [19, 21]. De pending on which R-SMAD is utilized the TGFfamily signals can be divided into two path ways. Generally, BMP si gnals activate SMAD1, 5, and 8, whereas TGFand Activins signal through SMAD2 a nd 3. There are also inhibitory ISMADs (SMAD6 and SMAD7), which negatively alter TGFsignaling by inhibiting the interaction of R-SMADs with SMAD4 or their corresponding type I re ceptors. [17, 20]. Two type III receptors, TGFBR3 (a lso called betaglycan) and ENG are known to be involved in TGFsignaling. The precise roles of th ese type III receptors are unclea r, but generally believed that thy act as coreceptors for facilitating (or inhibi ting) interactions between ligands and type II receptors. Arterial Endothelium-Specific Alk1 Expression ALK1 is one of the seven type I receptors for TGFfa mily proteins. The ligand specificity of the type I receptor has been define d by the ability to bind to a given ligand and to stimulate specific downstream genes in the pr esence of corresponding type II receptors. ALK1 can interact with TGF -1 or activins in the existence of either TGFBR2 or activin type II receptors (ACVR2 or ACVR2B), respectively. In addition, recent studies showed that BMP9 and its close subfamily member BMP10 can tr ansduce its signal through either BMPRII or ActRIIA . Therefore it is unclear which lig and is the most relevant ALK1 ligand in vivo, especially for pathogenesis of HHT. Our laboratory previously generated a lacZ reporter line ( Alk1lacZ), in which the bacterial -galactosidase gene is inserted into the Alk1 locus. Using this reporter line, it was demonstrated that ALK1 is primarily expressed in developing arterial endothelium . ALK1deficient mice die in mid-gestation (E10.5), exhibiting severe vascular abnormalities such as
16 hyperdilation of large vessels, fusion of capillary plexus into major vessels, and AVMs [24, 55]. Furthermore, ALK1-deficient embryos showed severe defects in the development of vascular smooth muscle cells, demonstrating that ALK1 signaling is necessary for proper differentiation and recruitment of vascular smooth muscle cells . Taken together, th ese data suggest that ALK1 may play a pivotal role in the development of arterial structure (arterialization), and stabilization and maturation of blood vessels . Role of Alk1 in Vascular Remodeling The vascular system consists of two larg ely distinct and separate networks of arterial and venous blood vessels. Arteries and veins are morphologically, functionally and molecularly distinguishable. Recent evidence has revealed that molecular differences do exist between arterial and venous endothelial cells, even be fore blood vessels are formed and that complex genetic paths are responsible fo r these early differences. Identification of several molecular signaling offers information regarding mol ecular regulators of arteriovenous boundaries, configuration, and cell fate. Notch and its ligands have a ke y function in identifying artery versus vein, as well as determining cell des tiny. Notch signaling is suggested to have a significant role in vasculogenesis and angiogenesis. Notch1 and Notc h4 are expressed in vascular endothelium and deficient mice in various genes involved in Notch signali ng exhibited defects in blood vessel formation [26-28] Ephrin-B2, an Eph family transmembrane ligand, denotes the destiny of arterial but not venous endothelial ce lls in the beginning of angiogenesis. Conversely, Eph-B4, a receptor for ephrin-B2, accounts for the fate of veins but not arteries. Notch signaling regulates the differential expression of E phrin and Eph genes in the vasculature, while Ephrin/Eph signaling helps delineate the borders between arteries a nd veins, assisting to define the arterial cellular compartment. In addition, the vascular endoth elial growth factor (VEGF), an
17 indispensable vascular signaling factor, also emer ged as a critical trigger of Notch pathway in developing vasculature in zebrafish and mice . The fact that ALK1 is specifically expre ssed in arterial endothel ium implies that ALK1 signaling may play an important ro le in arteriogenesis. As a TGFreceptor in vascular endothelium, ALK1 signaling may required for the production of perivascul ar matrix proteins and the induction of the differen tiation and recruitment of periendothelial cells, such as pericytes and smooth muscle cells. In additi on, the fact that ALK1 signaling participates in the maturation of arterial vessels accounts for the reason why ALK1 expression is confined to arteries only during the course of fast growth stages. Meanwhile, vascular function and structure are also defined by the profound effects of hemodynamic stim uli. The distinctive characteristics of arteries and veins are destined to satisfy the physiological ne cessities, such as shear stress originated from blood flow or cyclic circumferential strain from pressure. The response to hemodynamic changes, including the creation of extr acellular matrix proteins and proliferation of vascular smooth muscle cells, are exerted by these factors. ALK1 is believed to have a function in the signaling of the vascular remode ling of arteries, which was suggested by Alk1 expression in the mesenteric arteries in accord ance with increased blood flow, as well as, in already existing arteries apart from wound and tumor during wound reparation and tumorigenesis . Increased pulmonary blood flow stimulates upregulated TGFand ALK1 expression in pulmonary blood vessel , in addition, endothelial cells disclosed to high level of shear stress induce the expression of TGF1. Taken together, these data suggest that ALK1 is perhaps a protein that responds to arteri al hemodynamic milieu and play s an essential role for the remodeling arteries.
18 The Cre/loxP Site-Specific Recombination System The Cre/lox P site-specific recombination system is a powerful system to delete a gene of interest in a specific cell type s or stage, and most widely utilized among several conditional knockout systems. Cre is a 38 kDa recombinase pr otein from bacteriophage P1 which mediates excisive or inversional site specific recombination between loxP sequences. A loxP sequence is made up of two 13 bp inverted repeats separate d by an 8 bp asymmetric spacer area (Figure 12a). Recombination can take place in the asymmetiric spacer area. Cre inverts the intervening part of DNA when two loxP sequences are in an opposite orientation to each other, while it excises the intervening DNA between the sites leaving one loxP s ite afterward when two sites are in a direct orientation (Figure 1-2b). When a convent ional knockout line results in embryonic lethality, this Cre/loxP system can make it possible to overcome embryonic lethality and explore the role of a gene in later stages of development or postnatal life. This conditional approach is also widely used to determine the pr imary cells (or tissues) where the gene of interst plays . Transgenic Mouse Strains for Exp ressing Cre Recombinase in ECs Several transgenic (Tg) and knockin (KI) line s have been developed to express the Cre recombinase in the vascular endothelium. These include; Tg( Tie1-Cre) and Tg( Tie2-Cre), Tg( Flk1-Cre), Flk1+/Cre and Tg( VECad -Cre) [34-37]. Among them, the Tg( Tie2Cre) line has been used most widely for conditional knockout in vascular endothelium . The Cre is expressed in endothelial cells of arteries and veins by the promot er and enhancer regions of the mouse Tie2 gene . In addition to such a pan-endothelial Cre expression, Tie2Cre-mediated recombination (lacZ activities) was found in the endocardium and the mesenchymal cells of the atrioventricular (AV) cushions . In recen t studies it has been shown that conditional elimination of several genes using the Tg( Tie2Cre) line lead to defects in the valve formation as
19 well as minor or detrimental phenotypes in bl ood vessels [40-42]. In these studies, the Tg( Tie2Cre) line was suggested to be a practical model to investigate the function of a gene in the whole ECs or to study cell fate-mapping, but is not suitable for studying functions of genes in the ECs of specific vascular beds. Tg( Flk1 -Cre) lines, were generated using regulatory sequence of the mouse Flk1 gene adequate for endothelial cell-sp ecific expression of the lacZ reporter gene. In these mice, intensive endothelium-specific staining of most vascular beds was detected at E11.5 to E13.5, but they also demonstrated lacZ expression in muscle lineages. Furthermore, Cre-mediated lineage tracing study using Flk1 +/Cre mice have shown that both vascular endothelium and cardiac muscle arise from Flk1 -positive mesodermal progenitors during development . Tg( Tie1 -Cre) mice were directed to endothelial cells by the mouse Tie1 promoter. In Tie1 Cre line intercrossing with ROSA26R reporter mice, in which lacZ reporter gene is activated by the Cre-medicated excision, lacZ staining was obs erved in almost all cells of the forming vasculature in early stage of va scular formation (E8.5-E9.5) and in most endothelial cells within the embryo between E10.5 and birth, suggesting that Tie1 -Cre transgenic strain can proficiently direct to the deletion of floxed genes in endothelial cells in vivo. However, like the Tg(Tie2Cre) line, the Tg( Tie1Cre) line also displays the Cre-me diated DNA excision in the endocardial cells, mesenchymal cells in the AV cushion and EC s [34, 44], indicating th at before epithelialmesenchymal transition (EMT) during AV cushion formation is complete, the Tie1Cremediated DNA recombination might happen . The VE-Cadherin (vascular endothelial cadherin)-Cre line exhibits standardized expr ession in the endothelium of developing and
20 dormant vessels of all organs, however, it is also expressed within a small compartment of hematopoietic cells. Experimental Scheme using Alk1+/GFPCre Mice Primary goal of my thesis project is to characterize Alk1+/GFPCre mouse line from which spatiotemporal expression of the tran sgene can be compared with that in Alk1+/lacZ line. To analyze the Cre expression, the Alk1+/ GFPCre mice were intercrossed with the ROSA26 reporter line ( R26R ). Embryos at various gesta tional stages from this cross were then subjected to whole mount X-gal staining followed by histochemical analysis. In addition, I was also involved in characterization of a transgenic mouse line, Tg( Alk1 GFPCre), in which GFPCre is expressed under previously identified 9.2-kb Alk1 regulatory fragment. Both mouse lines are valuable to characterizing the pattern of Alk1 expression by reciprocally co mpensating the limitations of each line. As described above, to date, an exclusiv e endothelial specific Cr e line with constitutive expression is yet to be developed. Theref ore, as the second approach to verify Alk1+/ GFPCre line, we examined the activity of Cre recombinase in Alk1+/ GFPCre line by Cre-mediated conditional deletion of Tgf r 2gene to compare this line with alrea dy established Cre lines, which expression are driven by other endothelial cell-specific promoters, such as Tie2Cre Tie1Cre Flk1Cre, and Flk +/cre lines.
21 Figure 1-1. Signal transduction of TGFsuperfamily. More than 30 known ligands are known to be contained in the TGFsuperfamily. TGFfamily members include TGFs, bone morphogenic proteins(BMPs), activins/ inhibin, and growth and differentiation factors (GDFs). Binding of the ligand to its type II receptor activates the receptor, which combines with and phosphorylates the type I receptor. The activated type I receptor subsequently phosphorylates a receptor-regulated SMAD (R-SMAD), permitting this protein to associate with SMAD4 and translocate into the nucleus. This complex binds to specific enhancers in the nucleus of target genes, activating transcription. TGFfamily members are classified into seven type I, five type II and two type III receptors. The type II receptor consists of TGFtype II receptor (T RII), activin type II receptors (ActRIIA and ActRIIB), bone morphogenetic protein type II receptor (BMPR2), and Mlle rian inhibiting substa nce type II receptor (MISR2). The type I receptors are comprise d of activin receptor-like kinase (ALK) 17. The type III receptors include -glycan and endoglin. At present, eight SMAD proteins have been identified in mammals. III p p T R-II ActRIIA ActRIIB BMPRII MIS-II TGFs BMPs Activins GDFs ALK1-7 Endoglin -glycan R-SMAD R-SMAD SMAD4 P P Target gene R-SMAD SMAD4
22 Figure 1-2. The Cre/LoxP Recombina tion System (a) A detailed structure of LoxP site. LoxP is a site on the Bacteriophage P1 consisting of 34bp. There exists an asymmetric 8bp sequence between two sets of palindromic 13bp sequences flanking it. (b) A model experiment using the Cre/LoxP system. Tw o mouse lines are required for conditional gene deletion. First, a conve ntional transgenic mouse line with Cre targeted to a specific tissue or cell type, a nd secondly a mouse strain th at embodies a target gene flanked by two loxP sites in a direct or ientation (floxed gene). Recombination, which is an excision and consequently inactivation of target gene, occurs only in those cells expressing Cre recombinase.
23 CHAPTER 2 MATERIALS AND METHODS Generation and analysis of Alk1+/GFPCre mice An Alk1 -eGFPCre-knock-in ( Al k1+/GFPCre) vector was constructed. SV40 splice donor and acceptor signals (SD/SA), internal ribosomal entry sequence (IRES) and poly(A)signal (pA) were sequentially subcloned into pBluescriptIISK( +), of which XhoI site was deleted, and the eGFPCre fusion gene was subcloned between IRES and pA. This SD/SA-IRES-eGFPCre-pA fragment was replaced with SD/SA-IRES-lacZ fragment in the Alk1 -SIBN2 knock-in vector previously generated . Afte r electroporation of the linear lized knock-in vector, 365 G418resistant clones were screened by Southern hybridization analysis. Southern hybridization analysis was done by digestion of G418-resist ant ES cell DNA with Ec oRI, separated on 0.8% agarose gels, transferred to Hybond-XL membrane (Amersham Bi osciences) and probed with a [32P]dCTP labeled 0.4-kb HindIII-HindIII fragme nt to analyze the homologous recombination. Three of positive ES clones obtained from 365 of G418. Positive ES cells were injected into C57BL/6J (B6) blastocysts to ge nerate chimeric mice. Mating of chimeric mice with B6 females was performed to establish and maintain the Alk1+/GFPCre line on a mixed 129/B6 hybrid background. Mouse Breeding All procedures perform ed on animals were re viewed and approved by the University of Florida Institutional Animal Care and Use Committ ee. To investigate the expression pattern of ALK1 in embryos, male Alk1+/ GFPCre mice were crossed with female ROSA26 reporter line. When female mice had a vaginal plug, the stage of the embryo was judged as at embryonic days (E) 0.5. The pregnant female mice were dissecte d at various embryonic stages such as E9.5, 10.5, 11.5, 12.5, 15.5, and 17.5 embryonic stages and postnatal stage. At E9.5, 10.5 and 11.5,
24 embryos were harvested, fixed and stained with an X-gal solution. To examine the expression pattern of ALK1 at E12.5, 15.5, 17.5 and postnatal stage, various organs such as lung, liver, heart, intestine, eyes, brain, kidney, and sternum were isolated and used for X-gal staining. At every stage, PCR was performed for genotyping using yolk sac (E9.5 to 11.5) or small piece of tail (E12.5 to postnatal stage). X-gal staining To investigate AL K1 expression pattern in the Alk1+/ GFPCre; R26R mice, pregnant R26R female crossed with Alk1+/GFPCre male mice were euthanized by cervi cal dislocation and embryos were removed from uteri. For the embryos older than E11.5, internal organs were isolated. For whole mount staining, dissected embryos of E9.5 to E11.5 were subjected to fixation with a fixative solution containing 1% formal dehyde, 0.2% glutaraldehyde, 2mM MgCl2, 5mM EGTA, and 0.02% NP-40 for 15 minutes at room temperat ure. Fixed embryos were then washed with two times 3 minutes each with PBS on a rocker. Then, staining step was performed with solution containing 5mM K3Fe(CN)6, 5mM K4Fe(CN)6, 2mM MgCl2, 0.01% Na-Deoxycholate, and 0.5 mg/ml X-gal. Embryos were incubated in staining solution overnight at 37 Stained embryos were washed with PBS, fixed in 4% formalde hyde at room temperature for 1 to 2 hours and washed with PBS. To observe the lacZ expre ssion in organs at E12.5 to 17.5 and postnatal stages, each isolated organ was removed, sliced to 1-2 mm thickness and stained with the same X-gal staining solution just as the embryos. Histology and Immunohistochemistry Histological analysis wa s carried out by using standard m ethods. Embryos and pieces of organs from embryos and postnatal mice were dehydr ated with serial con centrations of ethanol (70 % 95 % 100 %) for 15 minutes each and then kept in Citrosolv for 10 minutes, followed
25 by incubating in paraffin for 3 hours. Paraffin embedded samples were subsequently sectioned into 5 thickness, placed on glass slides. Afte r deparaffinazing, the sections were counterstained with nuclear fast red (NFR) to determine the X-gal positive cells. Particularly, sternum piece from newborn was serially dehydrated in methanol (25% 50% 75%100%) for 10 minutes each and then cleared with organic solvent including benzyl benzoate and benzoic acid with 1:1 ratio. For immunohistochemistry, the standard ABC method was used with a Vector staining kit (Vector laboratories, Inc ., CA). Briefly, sections were hydrated and endogenous peroxidase activity wa s blocked by treating them with 3% (v/v) hydrogen peroxide (H2O2) at room temperature for 10 minutes. After tw o times of washes with PBS, the sections were incubated with blocking serums corresponding to their secondary an tibodies species for 1 hour at room temperature, followed by an incuba tion of primary antibodies. Biotinated secondary antibodies were incubated for thirty minutes af ter being washed with PBS. After secondary antibody incubation, sections were treated with peroxidase-conjuga ted avidin/biotin complex for thirty minutes followed by two PBS washes. Colo r development was carried out with a DAB + substrate chromogenic solution (Vector laboratories, Inc., CA). The antibody used for immunohistochemistry is SMA (clone: 1A4; Sigma, 1:800). PCR A s mall piece of the yolk sac or tail was used for the genotyping of embryonic and postnatal stages. Tissues were lysed in lysis buffer consisting of 50mM Tris (pH 8.0) and 0.5% TritonX-100 with proteinase K (1 mg/ml) for overnight at 55 The lysed samples were centrifuged at 10000 rpm for 10 minutes to prec ipitate tissue fragment. The PCR reaction mixture was made up of the following: 2.5l of 10X buffer, 3l of MgCl2 (25mM), 0.5l of dNTPs (25pM), 0.5l of each primer, 1.8l of genomic DNA, and 15l of H2O. To prevent the
26 PCR mixture from evaporation Mineral oil was overlaid. A hot start PCR reaction was set under the following guidance: the mixture wa s stayed at 94C for 10 minutes then the temperature was lowered to 72C.Once th e temperature reached 72C, 2l of a Taq polymerase was added. After adding the Taq polymerase, the PCR cycle was proceeded in the following way: 35 cycles at 94C for 45 seconds, 1 cycle at 60C for 45 seconds, and 1cycle at 72C for 1 minute. The primers used for genotyping are su mmarized in the Table1. PCR products were analyzed on 2% agarose gel containing ethidium bromide for about 20min at 105V. DNA bands were visualized under UV light and photogr aphed. The primers used for genotyping are summarized in the Table2-1. Genomic Southern DNA was extracted fro m ES cells by a dding 250l of lysis buffer containing 100mM Tris (pH8.0), 5mM EDTA, 0.2% SDS, 200mM NaCl with proteinase K. The plates were wrapped thoroughly and incubated at 55C over night with shaki ng at about 210rpm. The next day the plates were cooled down at room temperature and then 250l of isopropanol was added and mixed. DNA was carefully picked up by for ceps and excess isopropanol was removed and then resuspended in 25 l of deinoized water. The tubes with DNA mixture was then incubated at 55C overnight for complete melting. Digest ion 10l of genomic DNA was performed with 2l of EcoR and H buffer and 6l distilled water, in 20l total volume overnight at 37C. Next day, 2l loading dye per one digest sample was added and samples loaded into wells of 0.8% agarose gel, and electrophorese was carried out overnight at about 23voltage until the dye passes off the end of the gel. Gel was, then photographe d with a ruler and placed in glass dish followed by shaking gently in the series of solutions for the specified times; 1. 0.25N HCl (depurination solution) : 30minute, two times 2. Distilled water ( for rinse) 3. 0.4N NaOH (denaturation solution) : 30minute, two times
27 4. 20X SSC (transfer buffer for soaking) The Hybond-XL membrane (Amersham Bioscien ces), which is positively charged, and blotter papers were prewet in deionized wate r for 5 minutes, then in 0.4N NaOH, followed by putting on the gel/membrane blotting stack dissemb led in the order. The blot was stayed overnight for transfer. The memb rane was hybridized with a [32P]dCTP labeled 0.4-kb HindIIIHindIII fragment to analyze the homologous recombination.
28 Table 2-1. Primers used in PCR Genes Cre LacZ GFP Tgfbr2 (floxed) Forward primers GCTAAACATGCTTCATCGTCGGTC GTCGTTTTACAACGTCGTGACT CGATGGGGGTGTTCTGCTGGTAGT TAAACAAGGTCCGGAGCCCA Reverse primers CAGATTACGTATATCCTGGCAGCG GATGGGCGCATCGTAACCGTGC ATGGTGAGCAAGGGCGAGGAG ACTTCTGCAAGAGGTCCCCT
29 CHAPTER 3 RESULTS Screening of Alk1+/GFPCre mouse strain The Alk1+/GFPCre mouse line was generated by inserting the GFPCre fusion gene into the Alk1 locus by homologous recombination in order to express GFPCre in cells where endogenous ALK1 is expressed. The Alk1 -eGFPCre-knock-in ( Alk1+/GFPCre ) targeting vector consists of SV40 splice donor and acceptor signals (SD/SA), internal ribosomal entry sequence (IRES), eGFPCre fusion gene, and poly(A)signal (pA) (Figur e 3-1), essentially identical (except for the GFPCre portion) to previous Alk1+/lacZ targeting vector . Characterization of the Cre Transgene Expression in the Alk1+/GFPCr Line Alk1 is Expressed Primarily in Endothelium of Arteries, and also Detectable in that of Veins, Endocardium and Mesechymal Cell in Artiroventircular Cushions : To monitor the expression of functional Cre activities during development, Alk1+/GFPCre mice were intercrossed with the ROSA26 reporter line, referred to as R26R, in which lacZ reporter gene is activated under the ubiquitous ROSA promoter, only when Cre excises the STOP cassette flancked by loxP sequences. Therefore, lacZ expression in Alk1+/GFPCre;R26R mice represents the cells where Cre-mediated recombination had been occurred. The lacZ expression in various tissues/organs from various stages of embryos (from E9.5 to new born) was analyzed by X-gal staining. The same stages of embryos from Tg( Alk1 GFPCre) mice were also paralleled for comparison purposes. Littermates of Alk1+/GFPCre;R26R(-) and Alk1+/+;R26R(+) were used as a control for endogenous lacZ activit y at all stages. Consistent to our previous data from Alk1+/ lacZ mice, the lacZ activity was primarily detected in endothelium of developing blood vessels of Alk1+/CreGFP;R26R embryos throughout development.
30 In E9.5 embryos, lacZ positive cells were observed in the developing heart, dorsal aorta, head vascular system, however, they appeared to be patchy and spotty at the vascular tree (Figure 3-2a). However, positive lacZ expression was increased throughout development (Figure 3-2a and 2b). In whole mount E10.5 embryos, Xgal staining displayed vascular trees more uniformly (Figure 3-2b, 2e and 2f), whereas these vessel-specific staining was not detectable in the control (Figure 3-2c). At this stage, lacZ expression appeared earlier in artery than in vein, especially in umbilical vessels (Figure 3-2g). Comp arable lacZ expression pattern in vasculature was observed in Tg( Alk1GFPCre);R26R embryos, though the homogeneous vascular lacZ expression pattern appeared little earlier in th e Tg embryos than the knockin embryos (Figure 32d and 2h). This difference may be due to hi gher level of transgene expression in the Tg embryos compared to the knockin embryos. In E11.5 embryos, the pattern of X-gal positivity was similar with that of E10.5, but more specifica lly detected in capillaries surrounding neural tube and brain (Figure 3-2i and 2j). As the Alk1+/GFPCre line is supposed to be expressed both GFP and Cre protein, we also examined the GFP expression under the fluorescent microscope, confirming its expression in the lungs of Tg( Alk1 -GFPCre) newborn mice (Figure 3-2k and 2l). Taken together, these results further confirm th at the GFPCre was knocked into the Alk1 locus, and indicate that expressed Cr e becomes active at around E9.5. Histological analysis of the whole-mount Xgal stained embryos revealed that X-galpositive cells are mostly vascular endothel ial cells (ECs) (Figure 3-3a). Unlike Alk1+/ lacZ embryos in which venous ECs were lacZ-negative, venous as well as arterial ECs were X-gal positive in Alk1+/GFPCre mice, although X-gal staining intens ity in venous ECs was weaker than that in arterial ECs (Figure 3-3b to 3d). This result indicates that a lower level of ALK1 is indeed expressed in venous ECs. Consistent with this view, as described above, when umbilical
31 vessels were stained in X-gal only for a few hours, umbilical artery but not vein turned to blue (X-gal-positive) (Figure 3-2g). When these vessel s were incubated in the X-gal staining solution for overnight, however, both umbilical artery and veins became X-gal positive (data not shown). I found that non-ECs were also X-gal posit ive. As shown in Figure 3-3e to 3g, endocardium and mesenchymal cells in the AV cushions were X-gal positive. Similar results were also observed in Tg( Alk1 -GFPCre):R26R embryos (Figure 3-3h to 3j). Since endogenous ALK1 is not expressed in the AV cushion cells, th is data indicate that Cre-medicated activation of the lacZ reporter gene occu rred before the epithelium-mese nchymal-transition (EMT) in the artrioventricular region, resulting mesenchymal ce lls to show positive reporter gene expression (Figure 3-3e to 3g). Immunosta ining with anti-smooth muscle -actin ( -SMA) antibodies for vascular smooth muscle layers further confirmed that Alk1 e xpression were limited in ECs but not in smooth muscle layers of blood vessels (Figure. 3-3k). Strong lacZ expression was detected in the va sculature of most organs at E12.5 to E17.5. At these stages, we dissected out organs includi ng the lungs, liver, kidney, intestine and dorsal aorta. At E15.5, noticeable arteries and capilla ry network of the lung was obviously stained, while the epithelium of the bronc hioli and smooth muscle layers of the airways and blood vessels was negative (Figure 3-4a and 4b). Also, the en dothelium of dorsal aorta show positive lacZ staining (Figure 3-4d and 4e). The vessels of developing kidney glomeruli show positive lacZ expression (Figure 3-4g and 4h). In liver, sinusoidal endothelium looked negative, and ECs in large blood vessels rarely showed positive X-gal (F ig. 3-4j and 4k). Interestingly, lacZ activity was also detected in the submucosal layer of sma ll intestine (Fig. 3-4m a nd 4n). Not all of cells showing positive X-gal staining in intestinal subm ucosal layer appear to be ECs, implying the inclusion of mesenchymal cells. I expect that it can be verifi ed by using cell type specific-
32 antibodies to examine in terms of a protein level, or by using in situ hybridization regarding of a RNA level. So I observed two areas where non-e ndothelial cells express lacZ, the mesenchymal cells of AVC and intestinal submucosal layer. Endogenous lacZ activities were detected in the bone, kidney and GI tracts. However, these bac kground staining did not inte rfere our analysis as shown in Figure 3-4c, 4f, 4i, 4l, and 4o. At newborn mice, endothelium-specific lacZ activity was still dete cted in most organ system including hematopoietic organ such as th e liver, in which distinguishable lacZ activity was not examined in embryonic stages. In the liver lacZ activity was det ectable in endothelium of large veins and in some dispersed cells that may be possibly originated from hematopoietic cells (Figure 3-5a to 5c). Positive lacZ in ve nous endothelium may resu lt from ALK1 expression in the process of a hemotopoiesis in liver around E13.5, as I suspect. Intensive lacZ reaction was noted in the lung (Figure 3-5d). Histological examination revealed positive labeli ng in arteries and in the capillaries. No staini ng was noted in the epithelia of br onchiolar units (Figure 3-5e and 5f). The lacZ was also contiguous in the endothelium of arteries a nd in the capillaries as well as glomeruli of the kidney (Figure 3-5g to 5i). In the intestine, mesenteric vessel was positively stained with X-gal (Figure 3-5j) a nd histological section revealed that lacZ was expressed in the mucosa and submucosa layers (Figure 3-5k and 5l). The strong LacZ expression was still observed in endothelium of dorsal aorta (Figure 3-5m and 5n). We also examined the eye having a very sensitive vessel capillar y network, which showed positive staining in the endothelium of vessels around ciliary body, lens and optic nerve (Figure 3-5o and 5p). lacZ expression is also shown in the blood vessels on the thoracic wall (F igure 3-5q and 5r). Figure 3-5r shows X-gal stained internal arteries, but not veins. In a ddition, the brain vasculat ure showed positive and specific staining in all vessels ob served (Figure 3-5s and 5t).
33 Investigation of Alk1+/GFPCre Line as a Potential Endothelial Cre Line for Conditional KO Approaches.. Angiogenesis during development as well as pathological conditions such as diabetes retinopathy, tumor angiogenesis, and ischemic insu lts receive much attention in modern medical sciences. Since endothelium is the most fundame ntal structural and functional unit of the blood vessels, function of a gene in ECs has been heavily investigated both in vitro and in vivo. The Cre/loxP system has been the most powerful appr oach to delineate EC-specific functions of a gene. For this reason, numerous transgenic a nd knock-in mouse lines expressing Cre in ECs have been generated and characterized. Recent studies have shown heterogeneity among ECs, e.g. arterial vs. venous or microvessel vs. large ve ssels. Spatiotemporal expressions as well as tightness of EC-restricted expression vary among these EC-specific promoters. As shown in chapter 3, Alk1+/GFPCre mice express Cre mostly in vascular ECs, but the Cre expression appeared relatively late compared to other EC-specific Cre lines. In order to compare the Cre activity of the Alk1+/GFPCre line with that of a known EC-Cre deleter line, I have investigated Alk1+/GFPCre-mediated Tgfbr2 -deleted mice. Tgfbr2 -floxed mice were available in our laboratory, and Tg( Tie1 -Cre)and Tg( Tie2 -Cre)-mediated Tgfbr2-conditional knockout studies have been reported. Interestingly, Tg(Tie1 -Cre);Tgfbr2f/f embryos exhibited abnormalities in the vasculature of the yolk sac at E9.5 and die around E10.5 w ith almost identical lethal phenotype with Tgfbr2-/embryos which is the lacked ne tworks of vessels retardation, pericardial effusion in heart , whereas Tg( Tie2 -Cre); Tgfbr2f/f embryos die around E13.5 with idiopathic cardiovascular defects . I crossed Alk1+/GFPCre; Tgfbr2+/f male mice with Tbfbr2f/f females to generate Alk1+/GFPCre; Tgfbr2f/f mice. The resulting embryos of this cross were Alk1+/GFPCre; Tgfbr2+/f, Alk1+/GFPCre; Tgfbr2f/f Tgfbr2+/f, Tgfbr2f/f (Figure3-6a). Three embryos having noticeable
34 hemorrhage in the cranial region among the litt ermate (Figure3-6c and 6d), whose genotype turned to be Alk1+/GFPCre; Tgfbr2f/f (Figure3-6a lane2, 4, and 6). And the rest of embryos without a hemorrhage (Figure3-6b) were Alk1+/GFPCre; Tgfbr2+/f Tgfbr2+/f or Tgfbr2f/f( Figure3-6a lane1, 3, 5, 7 and 8). Therefore, only mutant embryos in which loxP sequence in both alleles of Tgfbr2 gene was deleted by Alk1+/GFPCre displayed the obvious hemorrhag e in the specific region of the head. The histological analysis of this area showed that red blood cells leaked from damaged blood vessels were observed to migrate into the ep ithelium layer of neural tube (Figure3-7a and 7b), whereas the control mice displayed no hemorrha ge sign in the area (Figure3-7c). Concerned with the heart, there was the slight difference in terms of the thickness of the ventricular septum, the density of trabeculation in ve ntricles between mutant and cont rol mice (Figure3-7d and 7e). The identical phenotype was observed in Alk1+/GFPCre; Tgfbr2f/f embryo at E13.5 (Figure3-8). More severe hemorrhage than that of E11.5 in the same cranial region was found (Figure 3-8b and 8c), whereas control embryos of Tgfbr2f/f still showed the absence of the hemorrhage (Figure 3-8a). Red blood cells in this hemorrhage area pe netrating into the epithelium layer of neural tube were confirmed again in the histological anal ysis (Figure 3-9a and 9b). I also examined the placenta of both mutant and control embryos, in which no significant discrepancy between mutant placenta and that of control was detected (Figure 3-9d to 9f). In addition, hearts of both mutant and control were distingu ished in a view of density of the ventricular trabeculation, in which denser and less organized trabecular was displayed in mutant heart (Figure3-9g and 9h). I expanded the stage to examine the consistency of the phenotype. At E15.5, six embryos were generated from the crossing of Alk1+/GFPCre; Tgfbr2+/f mice with Tbfbr2f/f females. Two embryos with normal phenotype turned to be Alk1+/GFPCre; Tgfbr2+/f and Tgfbr2+/f. One of embryo was discovered to die about one da y before, whose genotype was Alk1+/GFPCre; Tgfbr2f/f, demonstrating
35 mutant embryos of Alk1+/GFPCre; Tgfbr2f/f displayed the embryonic lethality around E14.5. The rest of three embryos appeared to be dead at earlier stage, but its genotypes were not determined due to their resolved DNA.
36 Figure 3-1. Procedures for the generation of Alk1+/GFPCre construct. Constructed Alk1 -eGFPCre( Alk1+/GFPCre) knock-in vector contains SV40 splice donor and acceptor signals (SD/SA), internal ribosomal entry sequen ce (IRES) and poly(A)signal (pA) that are sequentially subcloned into pBluescriptIISK( +), of which XhoI site was deleted, and the eGFPCre fusion gene was subcloned between IRES and pA.
37 Figure 3-2 .The LacZ expression in the blood vessel of Alk1+/GFPCre; R26R embryos.(a) At E9.5, lacZ was detected in developi ng vasculature, but at this time expression was appeared as spotty pattern. (b) Whole mount E10.5 embryo show s positive lacZ in most embryonic vascular system. (c) Cre-negativ e littermate also stained for lacZ, however, no lacZ activity was detected. (d,h) LacZ expression in Tg( Alk1 GFPCre);R26R mice at E10.5 comparable to that of Alk1+/ CreGFP;R26R (e) At E10.5, lacZ expression was found in devel oping capillaries covering brain. (f) Intersomatic arteries and capillaries on neural tube and limb of E10.5 embryo show positive lacZ staining. (g) E10.5 umbilical vesse l stained with X-gal. Note that lacZ was detected only in umbilical artery soon after X-gal staining. (i ) Dorsal view of E11.5 embryo; lacZ-positive capillary-like vesse ls in the developing neural tube. (j) Higher magnification of the cephalic vasc ular tree at E11.5. The staining was specifically associated with blood vessels. (k and l) Fluorescent (k) and bright field image (l) of lungs of Tg( AlkGFPCre) newborn mouse.
38 Figure 3-3 Expression pattern of lacZ during embryonic stage E10.5 in Alk1+/GFPCre; R26R and Tg( Alk1 GFPCre);R26R embryos. Transverse sections of whole mount X-gal stained embryo of Alk1+/GFPCre; R26R ;(a to g), Tg(Alk1 GFPCre);R26R ;(h to j). Embryo was counterstained with NFR (a to j) or immunstained with SMA antibody (k). (a) Note that lacZ expression was detected in endothe lium of dorsal aorta, cardinal vein, heart endocardium and in the AV cushions of Alk1+/GFPCre; R26R embryo. (b to d) The endothelium of dorsal aorta(c) and cardinal vein(d) show positive lacZ staining. (e to g) mesenchyme of AV cushion (e and f) and endocardium (g) are obviously lacZpositive. (h to j) In Tg( Alk1 GFPCre);R26R embyro at E10.5, the identical pattern of lacZ expression was observed in dorsal ao rta, cardinal vein, endocardium and AV cushion, which were comparable to that of Alk1+/ CreGFP;R26R embryo. (k) LacZ positive cells were localized in endothelium of dorsal aorta and cardinal vein. DA; dorsal aorta CV; cardinal ve in ED; endocarium AVC; atrioventricular cushion
39 Figure 3-4. The lacZ expressi on of tissues derived from Alk1+/GFPCre; R26R embryo at E15.5 (a and b) LacZ staining was detected in alveolar capillarie s and arterial endothelium of lung while the epithelium of broncholi are devoid of staining (b) A high magnification view of LacZ stained pulmona ry arteries. (d) Endothelium of dorsal aorta show positive staining. (e) Higher magnification of positively stained dorsal aorta (g) Vessels of developing kidney gl omeruli show positive lacZ expression. (h) High magnification view of st ained glomeruli and endothilum of kidney vessel. (j) Endothelium of liver large blood vessel show negative, lacZ staining. (k) Higher magnified liver blood vessel with devoid of lacZ staining. (m) In the submucosal layer of small intestine, lacZ expressi on was observed. (n) High magnification view
40 of mucosal layers show strong lacZ activity. (c,f, i, l, and o) No lacZ expression was shown in all tissues of control.
41 Figure 3-5. The Cre recombinase is active in Alk1+/GFPCre; R26R newborn endothelium of blood vessel. X-gal stained newborn organs (a, d, g, j, q, and s) were then counterstained with nuclear fast red (NFR) (b, e, h, k, m, o). (a) Liver (b) Histological section of liver (c) High magnification of the liver with lacZ expre ssion in the endothelium of large veins, in the sinusoidal endothelium with dot-like fashion and in some hematopoietic cells. (d) Lung. (e) Lung hi stology demonstrating lacZ-positive pulmonary endothelium. (f) Positive lacZ st aining in arterial endothelium, while no staining in the epithelia of bronchiolar units. (g) Ki ndey. (h) Kindey histological section (i) Capillaries in glomeruli are lacZ positive (j) Mesenter ic vessels in small intestine show positive lacZ staining. (k) Histol ogical section of the intestine. (l) High magnification of intestine vessels revealed lacZ-positive capillary of submucosal area. (m) Histological section of dorsal aorta. (n) Higher magnified view showed strong lacZ expression in endothelial layer of dorsal aorta. (o) Hi stological section of eye (p) Eye showed capillary endothelial staining. (q) Blood vessels on the thoracic wall. (r) Magnified view of thoracic wall. Arrows indicate artery (A) showing positive staining and vein (V) showing no lacZ staining. (s) Brain. (t) Brain vasculature showed positive and specific staining in all vessels.
42 Figure3-6 Conditioanl deletion of Tgfbr2 gene by Alk1+/GFPCre line showed noticeable hemorrhage in the cranial region at E11.5. (a) The genotype of embryos from the crossing of Alk1+/GFPCre; Tgfbr2+/f mice with Tbfbr2f/f females. The resulting embryos were Alk1+/GFPCre; Tgfbr2+/f, Alk1+/GFPCre; Tgfbr2f/f Tgfbr2+/f, Tgfbr2f/f (b) The control embryo of Alk1+/GFPCre; tgfbr2+/f shows the absence of hemorrhage. (c) The embryos of Alk1+/GFPCre; Tgfbr2f/f displayed a noticeable hemorrhage in the cranial region of head. (d) High magnification view of the cranial region of head in Alk1+ /GFPCre; Tgfbr2f/f embryos. Red arrow indicates the hemorrhage sign.
43 Figure3-7 Histological analysis of Alk1+/GFPCre mediated conditional deletion of Tgfbr2 gene in the brain and heart at E11.5.(a) The transverse section of the cranial region of head having noticeable hemorrhage in Alk1+/GFPCre; Tgfbr2f/f embryos. (b) High magnification view shows that red blood cel ls escaped from damaged blood vessels migrate into the epithelium layer of ne ural tube. (c) Control mice displayed no hemorrhage sign in the area. (d and e) The tr ansverse section of both mutant (d) and control heart (e). The mutant heart displa yed thicker ventricular septum and higher density of trabeculation in ventricles (d) compared to th at of control heart (e).
44 Figure 3-8 More severe but iden tial pattern of a hemorrhage wa s observed in E13.5 embryos. (a) Tgfbr2f/f control embryo displayed no hemorrhage. (b) Alk1+/GFPCre; Tgfbr2f/f embryo showed severe hemorrhage sign in the cranial region of h ead, consistant with that observed in E11.5. (c) High magnified vi ew showed the hemorrhage spot.
45 Figure 3-9. Comparision of the phenotype of brain, placenta and heart between Alk1+/GFPCre; Tgfbr f/f embryos and control one in the histologi cal analysis. (a and b) The transverse section of head with the hemorrhage area. (a) The red blood cells were found in the epithelium layer of neural tube. (b) RBCs we re observed to penetrate into the neural tube area in high magnification view. (c). The hemorrhage sign was absent in the control embryo brain. (d to f) The transverse section of the mutant placenta (d and e) and control placenta(f). There was no noticeable discrepancy observed between mutant placenta and that of control. (g and h) There was somehow difference in the density of the ventricular trabeculation betw een mutant heart (g) a nd control heart (h). Note that denser and less organized tr abecular was displayed in mutant heart.
46 CHAPTER 4 DISCUSSION In a previous study using a null mutant m ouse line Alk1+/lacZ in our laboratory showed a dynamic spatiotemporal expression pattern of ALK1; This ALK! expression is prevalent in developing arterial endothelium and tumor angi ogenesis. The strongest signal is observed in preexisting arteries and newly de veloped arteries, during wound hea ling, but its considerably reduced in adult arteries. These re sults demonstrated a necessary ro le of ALK1 in arterialization and remodeling of arteries. However, explor ing the complete mechanism of ALK1 function in arteriogenesis is ongoing, and we have consider ed various approaches to this topic. One approach was to delete ALK1 in arteries ( knockout or conditional knoc kout), then compare ALK1-deleted ECs to normal ECs, and finally an alyze their ALK1 functi on with microarray. To achieve this, the deletion of ALK1 can be implemen ted in cultured ECs, but it is not guaranteed that in vitro condition can reflect in vivo environment of ALK1 expression. A better approach might be to distinguish ALK1-expressi ng ECs from ALK1-nonexpressing cells in vivo. Establishment of mouse line directs to the pur pose to label positive-ALK1 endothelium with GFP reporter gene by insertion of GFP gene in Alk1 gene locus or by inducing GFP expression driven by Alk1 promoter in transgenic mice, which is Alk1+/GFPCre or Tg( Alk1 GFPCre) line, respectively. Once sorting of AL K1-expressing ECs are complete, we can determine distinctive gene expression profiles of ALK1-expressing ECs and identify the downstream targets of ALK1 signaling. The ultimate goal of this thesis is to elucidate molecular mechanism by which ALK1 regulates arteriogenesis. Sin ce HHT results from a heterozygous mutation in ALK1, the verification of the mechanism that ALK1 func tions in arteriogenesis would demonstrate underlying pathogenesis for HHT and to expand a range of th erapeutic design for these disorders.
47 The first approach to this object wa s to generate and characterize the Alk1+/GFPCre mouse line, which reflected endogenous ALK1 expression by inserting the fusion gene of GFP and Cre recombinase into Alk1 gene locus. We observed that ALK1 is mainly expressed in developing arterial endothelium, but was also expresse d in venous endothelium, endocardial cells, and mesenchymal cells in the AV cushion. This was de tected by the lacZ expression in endothelium of cardinal vein, endocardium lining the chamber of the heart, and the mesenchymal layer in a heart cushion as well as arterial endothelium in Alk1+/GFPCre; R26R line. This expression pattern is different from previously defined arterial endothelium specific ALK1 expression in Alk1+/lacZ or Tg( Alk1 -lacZ) line. In fact, this distinctive result is regard ed as another achievement obtained from Alk1+/GFPCre, since this line is sufficient to compen sate limitations to characterize ALK1 expression in Alk1+/lacZ or Tg( Alk1 -lacZ) line. Continuous arterial endothelium specific Alk1 expression and lacZ staining observed in Alk1+/lacZ or Tg(Alk1-lacZ) line demonstrates a transcriptional activity of the Alk1 promoter at a given time poin t. As lacZ is expressed by endogenous Alk1 promoter in this line, lacZ can be undetectable, if the promoter is insufficient to visualize the intense expression of the reporter gene, even though given ce lls or tissues actually express ALK1. Otherwise, in the Alk1+/lacZ line, once a relatively high level of lacZ expression can be observed in the cells or tissues where ALK1 is expressed, it implies that ALK1 is concurrently expressed in that regi on. That is, lacZ expression in Alk1+/lacZ is detected in a quantitative manner, so that it is difficult to determ ine whether the endothelium in an examined tissue actually expresses AL K1 or not. In contrast, lacZ expression in Alk1+/GFPCre;R26R mice, which is derived from crossing a R26R reporter line showing strong promoter activity, would constitutively show the lacZ activity, because once Cre-mediated recombination occurs and
48 activates lacZ expression, resulting recombined lacZ gene would be inherited throughout the cellular lineage. Therefor e, ALK1 expression in Alk1+/GFPCre might be visualized in a relatively qualitative fashion compared to that in Alk1+/lacZ. This is also why lacZ expression in an Alk1+/GFPCre; R26R line cannot be considered to guarant ee the concurrent Cre activity. However, due to inserted GFP gene expression, Alk1+/GFPCre provides an advantage to detect current ALK1 expression as well, which makes this line possible to exert more enhanced detection ability than that of Alk1+/lacZ line. As described above, for labeling ALK1-expressing ECs with GFP reporter gene, our laboratory also created a Tg( Alk1 GFPCre) as well as an Alk1+/GFPCre line. Tg( Alk1 GFPCre) mouse was generated from a previously characterized 9.2kb fragment of the Alk1 gene, including a 2.7kb promoter region and the whole homol ogous sequence in intron 2, which precisely recapitulated the endogenous ALK1 expression. The e nhancer elements in a conserved region of intron 2 of this transgenic lin e regulate arterial endothelium -specific expression . Cre expression in this Tg( Alk1 GFPCre), however, may be affected by the circumferential sequence toward the regulatory element, such as a repr essor for ALK1 expression in venous ECs or an activator for ALK1 expression in arterial ECs. Consequently, we need a scheme to verify the possible ectopic expression in Tg( Alk1 GFPCre). Alk1+/GFPCre might be a device to compensate for this limitation observed in transgenic line, since it reflects endogenous ALK1 expression. Histological analysis has shown that Alk1+/GFPCr e-mediated lacZ activity was found in the endocardium, the mesenchyme of the atrioventricu lar (AV) cushions as well as vascular ECs (Figure 3-3), suggesting that Alk1+/GFPCre-mediated DNA recombination occurred prior to endothelial-mesenchymal transition during th e AV cushion formation. The onset of Cremediated recombination in the ECs and the mesenchyme of AV cushions of the
49 Alk1+/GFPCre;R26R mice was observed at E10.5. During early heart valve development, a subset of ECs which specifies the cushion-forming re gions divide and enter the cardiac jelly, a gathering of connective tissue, where they subsequently proliferate and fulfill their differentiation into mesenchymal cells [47, 48]. Positive lacZ activity in Alk1+/GFPCre; R26R line suggest that it possibly deletes a gene in th e AV cushion as well as in ECs, which then conceivably causes defects in cardiac valve deve lopment. Consistent with this expectation, widely used pan-endothe lial Cre line, Tg(Tie2 -Cr e), also deletes a gene of interest in the AV cushion as well as in the ECs. In recent studies elimination of some genes using the Tg(Tie2 Cr e) line lead to defects in the valve formation as well as slight or dete riorative phenotypes in the blood vessel [40, 41]. This study proposes the cornerstone to identif y genes regulated by ALK1 signaling in ECs and in advance, to characterize downstream targets of ALK1 signaling. The lethality of Alk1 deletion in mice implies its essential role in arteriogenesis, but exact roles are still largely unknown. To understand the function of ALK1 in arte riogenesis, it is nece ssary to elucidate its activities in angiogenesis, cell mi gration, adhesion, vasculature, cytoskeleton establishment and gene regulation. The studies related to this field were performed to determine the role of ALK1 in ECs by using a constitutively active ALK1 adenovirus to infect human umbilical vein endothelial cells (HUVEC), or th e human micro-vascular endotheli al cell line HMEC-1 [49-51]. HHT is a vascular disorder of impaired angiog enesis caused by mutations in elements of TGFsignal transduction, such as AL K1, ENG and SMAD4. It implies that ALK1 signaling regulates angiogenesis linked to certain genes. Actually, several recent studies support this theory by identifying genes involved in angiogenesis and arteriogenesis. These genes include IL-8 (interleukin-8) , ET-1(e ndothelin-1) , ID 1 (Inhibitor of DN A binding 1) ,
50 HPTP (human protein-tyrosine phosphatase ) , TEAD4(TEA domain family member4) , SMAD6/7, STAT1, SMAD1, CXCR4, Ephrin-A 1 . The deletion of some of these genes caused embryonic lethality in mice with defective vascular development, which shared a similar phenotype seen in HHT mice models . ID proteins are helix-lo op-helix proteins that stimulate endothelial cell prolif eration and migration and are i nduced to be expressed by TGF1 . In interruption of both ID1 and ID3 gene s, vessels were dilated and lacked branching capillaries , which was comparable to vascul ar abnormalities in ALK1-deficient mice. These data suggest that ID1-3 may play a critical role in regulating endothelial cells and possibly vascular functions as downstream mediators of ALK1 [47, 50]. Mi croarray analysis has enabled us to accomplish a comprehensive investigation of many responsive genes to Alk1 in ECs. This approach can reveal transcriptional targets and their regulation by ALK1. In addition, we presented the Alk1+/GFPCre line as another useful EC -Cre line compared to other known pan-EC specific Cre lines, in which Tgfbr2 gene is conditionally deleted by Alk1+/GFPCre line. Alk1+/GFPCre; Tgfbr2f/f embryos at E11.5 and E13.5 displayed the localized hemorrhage in the cranial region of the head on both sides. The severity appeared to be increased throughout development between E11.5 and E13.5. Ot her than the cranial area, there was no hemorrhage observed. The mechanism by which the hemorrhage in the specific area of the brain appears is still under investigation, but we su spect the rupture of blood vessels may cause hemorrhage as an indirect effect of the deletion of Tgfbr2 gene in vessel development or heart function, because these localized vascular defects were limited to only specific areas of the brain, not leading to systemic vascular abnormality. We will continue monitoring the phenotype of Alk1+/GFPCre-mediated conditional deletion of Tgfbr2 gene at later developmental stages.
51 Originally, our lab established the Alk1+/GFPCre mouse line to disti nguish and characterize the ALK1-expressing endothelium by labeling positive-ALK1 endothelium with GFP reporter gene. Once characterization of ALK1 -expressing ECs is complete, we can assess distinctive gene expression profil es of ALK1-expressing ECs, which would enable us to identify the downstream targets of ALK1 signaling. Initially, we characterized the ALK1 expression pattern in Alk1+/GFPCre mice line. The cumulative developmental and biological data from ALK1-regulated genes suggest that ALK1 is part of an independent network in vascular homeostasis, which might explain the vascular defects observed in HHT patients. Many genes regulated by ALK1 may be potenti al candidates of study in clini cal research of HHT patients. The long-term goal of this thesis is to eluc idate the molecular mechanisms by which ALK1 regulates arteriogenesis. Since HHT results from a heterozygous mutation in ALK1, verification of these mechanisms would provide valuable in sight to the underlying pathogenesis of HHT as well as contribute to therapeutic appr oaches to this vascular disease.
52 LIST OF REFERENCES 1. Berg, J.N., Guttmacher A.E., Marchuk D.A., Porteous M.E. Clinical heterogeneity in hereditary haem orrhagic tela ngiectasia: are pulmonary arte riovenous malformations more common in families linked to endoglin? J Med Genet, 1996. 33(3): p. 256-257. 2. Abdalla, S.A., Geisthoff U.W., Bonneau D., Plauchu H., McDonald J., Kennedy S., Faughnan M.E., Letarte M. Visceral manifestati ons in hereditary haem orrhagic telangiectasia type 2. J Med Genet, 2003. 40(7): p. 494-502. 3. Morgan T., McDonald J., Anderson C., Ismail M., Miller F., Mao R., Madan A., Barnes P., Hudgins L., Manning M. Intracranial hemorrhage in inf ants and children with hereditary hemorrhagic telangiectasia (Osl er-Weber-Rendu syndrome). Pediatrics, 2002. 109(1): p. E12. 4. Haitjema T., Disch F., Overtoom T.T., Westermann C.J., Lammers J.W. Screening fa mily members of patients with heredita ry hemorrhagic telangiectasia. Am J Med, 1995. 99(5): p. 519-24. 5. Ference, B.A., Shannon T.M., White RI Jr, Zawin M., Burdge C.M. Life-threatening pulm onary hemorrhage with pulmonary arteriovenous malformations and hereditary hemorrhagic telangiectasia Chest, 1994. 106(5): p. 1387-90. 6. Martini, G.A., The liver in hereditary haemorrhagic telean giectasia: an inborn error of vascular structure with multiple manifesta tions: a reappraisal. Gut, 1978. 19(6): p. 531-7. 7. Garcia-Tsao G., Korzenik J.R., Young L., Henderson K.J., Jain D., Byrd B., Pollak J.S., White RI Jr Liver disease in patients with he reditary hem orrhagic telangiectasia. N Engl J Med, 2000. 343(13): p. 931-6. 8. Vase P. and Grove O, Gastrointestinal lesions in hereditary hemorrhagic telangiectasia. Gastroenterology, 1986. 91(5): p. 1079-83. 9. Shovlin C.L., Hughes J.M., Tuddenham E.G., Temperley I., Perembelon Y.F., Scott J., Seidman C.E., Seidman J.G. A gene for hereditary haem orrhagic telangiectasia maps to chromosome 9q3. Nat Genet, 1994. 6(2): p. 205-9. 10. Abdalla, S.A. and Letarte M. Hereditary ha emorrhagic telangiect asia: current views on genetics and mechanisms of diseas e. J Med Genet, 2006. 43(2): p. 97-110. 11. Gallione C.J., Richards J.A., Letteboer T.G., Rushlow D., Prigoda N.L., Leedom T.P., Ganguly A., Castells A., Ploos van Amstel J.K., Westermann C.J., Pyeritz R.E., Marchuk D.A. SMAD4 m utations found in unselected HHT patients. J Med Genet, 2006. 43(10): p. 793-7. 12. Bayrak-Toydemir, P., McDonald J., Akarsu N., Toydemir R.M., Calderon F., Tuncali T., Tang W., Miller F., Mao R. A fourth locus for hered ita ry hemorrhagic telangiectasia maps to chromosome 7. Am J Med Genet A, 2006. 140(20): p. 2155-62.
53 13. Berg J.N., Gallione C.J., Stenzel T.T., Johnson D.W., Allen W.P., Schwartz C.E., Jackson C.E., Porteous M.E., Marchuk D.A. The activin re ceptor-like kinase 1 gene: genom ic structure and mutations in heredita ry hemorrhagic telangiectasia type 2. Am J Hum Genet, 1997. 61(1): p. 60-7. 14. Abdalla S.A., Cymerman U., Johnson R.M., Deber C.M., Letarte M. Disease-associated mutations in conserved residue s of ALK-1 kinase dom ain. Eur J Hum Genet, 2003. 11(4): p. 279-87.. 15. Cole S.G., Begbie M.E., Wallace G.M., Shovlin C.L. A new locus for hereditary haem orrhagic telangiectasia (HHT3) maps to chromosome 5. J Med Genet, 2005. 42(7): p. 577-82. 16. Gallione C.J., Repetto G.M., Legius E., Rustgi A.K., Schelley S.L., Tejpar S., Mitchell G., Drouin E., Westermann C.J., Marchuk D.A. A combined syndrome of juvenile polyposis and hereditary haem orrhagic tela ngiectasia associated with mutations in MADH4 (SMAD4). Lancet, 2004. 363(9412): p. 852-9. 17. Massague J., TGF-beta signal transduction. Annu Rev Biochem, 1998. 67: p. 753-91. 18. Seki T., J. Yun, and S.P. Oh, Arterial endot helium-specific activin receptor-like kinase 1 expression suggests its role in arterializa tion and vascular remodeling. Circ Res, 2003. 93(7): p. 682-9. 19. Huse M., Chen Y.G., Massagu J., Kuriyan J. Crystal structure of the cytoplasmic dom ain of the type I TGF beta receptor in complex with FKBP12. Cell, 1999. 96(3): p. 425-36. 20. David L., Mallet C., Mazerbourg S., Feige J.J., Bailly S. Identification of BMP9 and BMP10 as functional activators of th e orphan activin receptor-like kinase 1 (ALK1) in endothelial cells. Bl ood, 2007. 109(5): p. 1953-61. 21. Lo R.S., Chen Y.G., Shi Y., Pavletich N.P., Massagu J. The L3 loop: a structural motif determ ining specific interactions between SMAD proteins and TGF-beta receptors. Embo J, 1998. 17(4): p. 996-1005. 22. Seki T., K.H. Hong, and S.P. Oh, Nonoverl apping expression patterns of ALK1 and ALK5 reveal distinct roles of each receptor in vascular development. Lab Invest, 2006. 86(2): p. 116-29. 23. FernandezL.A., Garrido-Martin E.M., Sanz-Rodriguez F., Pericacho M., RodriguezBarbero A., Eleno N., Lopez-Novoa J.M., Dwell A., Vega M.A., Bernabeu C., Botella L.M. Gene expression fingerprinting for hu m an hereditary hemorrhagic telangiectasia. Hum Mol Genet, 2007. 16(13): p. 1515-33. 24. Oh S.P., Seki T., Goss K.A., Imamura T., Yi Y., Donahoe P.K., Li L., Miyazono K., ten Dijke P., Kim S., Li E. Activin receptor-like kina se 1 m odulates transforming growth
54 factor-beta 1 signaling in the regulation of angiogenesis. Proc Natl Acad Sci U S A, 2000. 97(6): p. 2626-31. 25. Torres-Vazquez J., Kamei M., and Weinst ein B.M., Molecular distinction between arteries and veins. Cell Tissue Res, 2003. 314(1): p. 43-59. 26. Shutter J.R., Scully S., Fan W., Richards W.G., Kitajewski J., Deblandre G.A., Kintner C.R., Stark K.L. Dll4, a novel Notch ligand ex pressed in arterial endothelium Genes Dev, 2000. 14(11): p. 1313-8. 27. Liu Z.J., Shirakawa T., Li Y., Soma A., Oka M., Dotto G.P., Fairman R.M., Velazquez O.C., Herlyn M. Regulation of Notch1 and Dll4 by vascular endothelial growth factor in arterial endothelial cells: im plications for modulating ar teriogenesis and angiogenesis. 2003. ;23(1): p. 14-25. 28. Williams C.K., Li J.L., Murga M., Harris A.L., Tosato G. Up-regulation of the Notch ligand Delta-like 4 inhibits VEGF-indu ced endothelial cell function. 2006.;107(3): p. 931-9. 29. Wang H.U., Chen Z.F., and Anderson D.J. Molecular distinction and angiogenic interaction between embryonic arteries and ve ins revealed by ephrin-B2 and its receptor Eph-B4. Cell, 1998. 93(5): p. 741-53. 30. Ohno M., Cooke J.P., Dzau V.J., Gibbons G.H. Fluid shear stress induces endothelial transform ing growth factor beta-1 transcription and pr oduction. Modulation by potassium channel blockade. J Clin Invest, 1995. 95(3): p. 1363-9. 31. Mata-Greenwood E., Meyrick B., Steinhorn R.H., Fineman J.R., Black S.M. Alterations in TGF-beta1 expression in lam bs with in creased pulmonary blood flow and pulmonary hypertension. Am J Physiol Lung Cell Mol Physiol, 2003. 285(1): p. L209-21. 32. Sauer B. and Henderson N., Site-specifi c DNA recombination in mammalian cells by the Cre recombinase of bacteriophage P1. Proc Natl Acad Sci U S A, 1988. 85(14): p. 516670. 33. Camenisch T.D., Spicer A.P., Brehm-Gibson T., Biesterfeldt J., Augustine M.L., Calabro A Jr, Kubalak S., Klewer S.E., McDonald J.A. Disruption of hyaluronan synthase-2 abrogates norm al cardiac morphogenesis a nd hyaluronan-mediated transformation of epithelium to mesenchyme. J C lin Invest, 2000. 106(3): p. 349-60. 34. Gustafsson E., Brakebusch C., Hietanen K., Fssler R. Tie-1-dire cted expression of Cre recombinase in endothelial cells of embryoi d bodies and transgenic mice. J Cell Sci, 2001. 114(Pt 4): p. 671-6. 35. Kisanuki Y.Y., Hammer R.E., Miyazaki J., Williams S.C., Richardson J.A., Yanagisawa M. Tie2-Cre transgenic m ice: a new model fo r endothelial cell-lineag e analysis in vivo. Dev Biol, 2001. 230(2): p. 230-42.
55 36. Licht, A.H., Raab S., Hofmann U., Breier G. Endothelium-specific Cre recombinase activity in flk-1-Cre transgenic m ice. Dev Dyn, 2004. 229(2): p. 312-8. 37. Alva, J.A., Zovein A.C., Monvoisin A., Murphy T., Salazar A., Harvey N.L., Carmeliet P., Iruela-Arispe M.L. VE-Cadherin-Cre-recombinase transgenic mouse: a tool for lineage analysis and gene deletion in e ndothelial cells. Dev Dyn, 2006. 235(3): p. 759-67. 38. Schlaeger T.M., Bartunkova S., Lawitts J.A., Teichmann G., Risau W., Deutsch U., Sato T.N. Unifor m vascular-endothelial-cell-specific gene expression in both embryonic and adult transgenic mice. Proc Natl Acad Sci U S A, 1997. 94(7): p. 3058-63. 39. de Lange F.J., Moorman A.F., Anderson R.H., Mnner J., Soufan A.T., de Gier-de Vries C., Schneider M.D., Webb S., van den Hoff M.J., Christoffels V.M. Lineage and morphogenetic analysis of the cardiac valves. Circ Res, 2004. 95(6): p. 645-54. 40. Park C., Lavine K., Mishina Y., Deng C.X., Ornitz D.M., Choi K. Bone morphogenetic protein receptor 1A signaling is dispensable for hem atopoie tic development but essential for vessel and atrioventricular endocardi al cushion formation. Development, 2006. 133(17): p. 3473-84. 41. Nanba D., Kinugasa Y., Morimoto C., Koizumi M., Yamamura H., Takahashi K., Takakura N., Mekada E., Hashimoto K., Higashiyama S. Loss of HB-EGF in smooth muscle or endothelial cell lineages causes heart malform ation. Biochem Biophys Res Commun, 2006. 350(2): p. 315-21. 42. Lincoln J., Kist R., Scherer G., Yutzey K.E. Sox9 is required for precursor cell expansion and extracellular m atrix organiza tion during mouse heart valve development. Dev Biol, 2007. 305(1): p. 120-32. 43. Motoike T., Markham D.W., Rossant J., Sato T.N. Evidence for novel fate of Flk1+ progenitor: contribution to m uscle li neage. Genesis, 2003. 35(3): p. 153-9. 44. Song L., Fssler R., Mishina Y., Jiao K., Baldwin H.S. Essential functions of Alk3 during AV cushion m orphogenesis in mouse embryonic hearts. Dev Biol, 2007. 301(1): p. 276-86. 45. Schlaeger T.M., Bartunkova S., Lawitts J.A., Teichmann G., Risau W., Deutsch U., Sato T.N. Unifor m vascular-endothelial-cell-specific gene expression in both embryonic and adult transgenic mice. Proc Natl Acad Sci U S A, 1997. 94(7): p. 3058-63. 46. Seki T., Hong K.H., Yun J., Kim S.J., Oh S.P. Isolation of a regulatory region of activin recep tor-like kinase 1 gene sufficient for arterial endothelium-sp ecific expression. Circ Res, 2004. 94(8): p. e72-7. 47. Armstrong, E.J. and Bischoff J. Heart valv e development: endothelial cell signaling and differentiation. Circ Res, 2004. 95(5): p. 459-70.
56 48. Garry, D.J. and Olson E.N. A common proge nitor at the heart of development. Cell, 2006. 127(6): p. 1101-4. 49. Lamouille S., Mallet C., Feige J.J., Bailly S. Activin receptorlik e kinase 1 is implicated in the maturation phase of a ngiogenesis. Blood, 2002. 100(13): p. 4495-501. 50. Lux A., Salway F., Dressman H.K., Krner-Lux G., Hafner M., Day P.J., Marchuk D.A., Garland J. ALK1 signalling analysis identi fies angiogenesis rela ted genes and reveals disparity between TGF-beta and con stitutively active receptor induced gene expression. BMC Cardiovasc Disord, 2006. 6: p. 13. 51. Ota T., Fujii M., Sugizaki T., Ishii M., Miyazawa K., Aburatani H., Miyazono K. Targets of transcriptional regulation by two distin ct type I receptors for transforming growth factor-beta in human umbilical vein endothe lial cells. J Cell Physiol, 2002. 193(3): p. 299-318. 52. Belperio J.A., Keane M.P., Arenberg D.A., Addison C.L., Ehlert J.E., Burdick M.D., Strieter R.M. CXC chemokines in angi ogenesis. J Leukoc Biol, 2000. 68(1): p. 1-8. 53. Bagnato, A. and Spinella F., Emerging role of endothelin-1 in tumor angiogenesis. Trends Endocrinol Metab, 2003. 14(1): p. 44-50. 54. Sakurai D., Tsuchiya N., Yamaguchi A., Okaji Y., Tsuno N.H., Kobata T., Takahashi K., Tokunaga K. Crucial role of inhibitor of DNA b inding/differentiation in the vascular endothelial growth facto r-induced activation and angiogenic processes of human endothelial cells. J I mmunol, 2004. 173(9): p. 5801-9. 55. Takahashi T., Takahashi K., St John P.L., Fleming P.A., Tomemori T., Watanabe T., Abrahamson D.R., Drake C.J., Shirasawa T., Daniel T.O. A mutant receptor tyrosine phosphatase, CD148, causes defects in vascular developm ent. Mol Cell Biol, 2003. 23(5): p. 1817-31. 56. Shie J.L., Wu G., Wu J., Liu F.F., Laham R.J., Oettgen P., Li J. RTEF-1, a novel transcriptional stim ulator of vascular endothelial growth factor in hypoxic endothelial cells. J Biol Chem, 2004. 279(24): p. 25010-6. 57. Urness L.D., Sorensen L.K., Li D.Y. Arteriovenous malformations in mice lacking activin receptor-like kinase1. Nat G enet, 2000. 26: p. 328 331. 58. Carvalho R.L., Itoh F., Goumans M.J., Lebrin F., Kato M., Takahashi S., Ema M., Itoh S., van Rooijen M., Bertolino P., Ten Dijke P., Mummery C.L. Compensatory signalling induced in the yolk sac vasculature by de letion of TGFbeta receptors in mice. J Cell Sci. 2007. 120(Pt 24): p. 4269-77. 59. Jiao K., Langworthy M., Batts L., Brown C.B., Moses H.L., Baldwin H.S. Tgf signaling is required for atri oventricular cushion m esenchy me remodeling during in vivo cardiac development. 2006. 133(22): p. 4585-93.
BIOGRAPHICAL SKETCH Eunji Lee graduated the Chung-Ang Universi ty in South Korea where she earned a Bachelor of Science degree in pharmacy in 2005. In 2006, she joined the Departm ent of Physiology and Functional Genomics at the Univer sity of Florida to achieve her masters education under the mentorship of Dr. S. Paul Oh.