In Vitro and in Vivo Analysis of the Establishment and Maintenance of Beta-Globin Locus Chromatin Structure

Material Information

In Vitro and in Vivo Analysis of the Establishment and Maintenance of Beta-Globin Locus Chromatin Structure
LEVINGS, PADRIAC P. ( Author, Primary )
Copyright Date:


Subjects / Keywords:
Cells ( jstor )
Chromatin ( jstor )
DNA ( jstor )
Gene expression ( jstor )
Genes ( jstor )
Genetic loci ( jstor )
Histones ( jstor )
Mice ( jstor )
Plasmids ( jstor )
Transgenes ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Padriac P. Levings. 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:
Resource Identifier:
74810509 ( OCLC )


This item is only available as the following downloads:

Full Text




This work is dedicated to my father, Thomas J. Levings. He has been the most important force in my life.


iii ACKNOWLEDGMENTS There are many people I wish to thank fo r contributing to my completion of my doctoral degree. Without them I would have never found fruition. The primary motivation and balance in this endeavor have been my parents. I would like to thank my mother for preaching patience and giving me comfort ion times when it was needed. My father has been the most important figure in my life. He is the reason I am the person I am today and when I could not find a r eason for going on, his advice concerning my setbacks and accolades of my success was more than enough motivation. I would also like to thank my mentor and friend Jrg Bungert. I think if I would have chosen another place to call home to my studies I would never have survived the past several years. He has been not only a me ntor, but a role model as well. I appreciate his patience with my idiosyncrasies. He was ab le to direct my exube rance into something creative, and allowed me to go to the gym in the middle of the day. I am certain I will always look to him for guidance in the future. My lab mates during the past seven years deserve a large degree of credit for my success, specifically, Kelly Leach, Karen Viei ra, and Valerie Crusselle. Kelly Leach was my best friend for the first five years of graduate school until her graduation. She was a source of constructive criticism as well as praise. Karen Vieira was a stable presence in the lab and also a very close friend. She had a strong personality and a sharp, focused sense of purpose. She was a good friend not only in the lab, but also at home as we were housemates for more than a year. I would like to thank Valerie Crusselle for just being


iv Val. She has put up with my teasing for the last year and will have to do so for the next. It is always a pleasure to be in the lab when sh e is there; her outlook and voice are always a refreshing change from my own. I am forever indebted to the members of my thesis committee, Drs. Laipis, Lewin, Muczycka, and Yang. They are the most intellig ent men I have ever met. I never felt that a committee meeting was an onerous task; I al ways felt that it was a positive experience. Their approval of my research has been one of the most gratifying experiences of my life. I do not think I could find a more worthy group of peers. I would also like to thank Thomas Conlon and Leah Cochran, the two people whom I spend the most time with outside of the lab. Their constructive and often unsolicited criticisms were often the kick in the right pl ace I needed to get me off the couch and back to work. They have been inspirational. Lastly I would like to tha nk Athena (my Siberian hus ky), Ogre, Thor, Achilles, Starburst, Lemonade, Sunfire, and Tigger (the Australian bearded dragons), and Turbo (the African spade foot tortoise), for the we lcome distractions they have provided during the construction of this work.


v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF FIGURES..........................................................................................................vii ABSTRACT....................................................................................................................... ix CHAPTER 1 INTRODUCTION........................................................................................................1 Background and Significance.......................................................................................1 Chromatin Structure a nd Gene Regulation...................................................................3 Development and Hematopoiesis.................................................................................6 In vitro Differentiation of Embryonic Stem Cells........................................................9 Globin Gene Organization..........................................................................................11 Globin Gene Proximal Regulatory Elements and Transcription Factors Involved in Erythropoiesis.....................................................................................................13 Locus Control Regions and Hemoglobin Gene Switching.........................................17 Summation..................................................................................................................22 2 MATERIALS AND METHODS...............................................................................26 3 ESTABLISHMENT OF PATTERNS OF HISTONE MODIFICATION AND FACTOR RECRUITMENT WITHIN THE GLOBIN LOCUS DURING INVITRO DIFFERENTIATION OF MURI NE EMBRYONIC STEM CELLs.............36 Introduction.................................................................................................................36 Results........................................................................................................................ .39 Discussion...................................................................................................................45 4 COMPOSITE -GLOBIN LOCUS TRANSGENES EXHIBITING POSITIONINDEPENDENT EXPRESSION IN TRANSGENIC MICE.....................................50 Introduction.................................................................................................................50 Results........................................................................................................................ .53 Discussion...................................................................................................................59


vi 5 SITE-SPECIFIC INTEGRATION OF MODIFIED YEAST ARTIFICIAL CHROMOSOMES IN MOUSE EMBRYONIC STEM CELLS BY RECOMBINASE MEDIA TE CASSETTE EXCHANGE.........................................64 Introduction.................................................................................................................64 Results and Discussion...............................................................................................70 6 SUMMATION OF RESULTS...................................................................................78 Functional Significance of the -globin Locus Control Region.................................78 In Vitro Differentiation of Murine Embryonic Stem Cells.........................................79 Site-Specific Integration and Positi on-Independent Expression of Human -globin Transgenes..............................................................................................................80 LIST OF REFERENCES...................................................................................................83 BIOGRAPHICAL SKETCH.............................................................................................93


vii LIST OF FIGURES Figure page 1-1 The human -globin locus........................................................................................13 3-1 PCR analysis of DNase I treated and reverse-transcribed total RNA extracted from differentiating embryonic stem cel ls at the indicated time points...................40 3-2 Interaction of transcription fact ors and RNA polymerase II with the -globin locus.........................................................................................................................4 1 3-3 Quantitative analysis of Pol II binding and histone modifications within the globin locus at “Day 0” in murine ES cells..............................................................43 3-4 Quantitative analysis of Pol II binding and histone modifications within the globin locus at “Day12” in di fferentaited murine ES cells......................................45 3-5 Fold change in factor binding and histone modifications at the murine -globin locus during in vitro differentiation.........................................................................47 3-6 Recruitment of transcription complexes and chromatin structure alterations at the -globin locus in undifferentiated pr ecursors and defini tive erythrocytes.........49 4-1 Structure and expression of the 432 4 transgene.....................................................54 4-2 Location of the human S1 integr ation site in the mouse genome............................56 4-3 Expression analysis of a -globin construct containi ng insulator sequences as well as LCR elements HS2 and 3 with their flanking DNA in transgenic mice......57 4-4 Analysis of the integration site, -globin gene expression, and DNase I HS sites in transgenic line 43f2 4B.......................................................................................59 4-5 Relative expression levels of two -globin transgenic constructs..........................61 5-1 Principle of RMCE...................................................................................................67 5-2 Structure of gene targeting cassettes for RMCE in ES cells....................................71 5-3 YAC constructs for RMCE......................................................................................72


viii 5-4 LoxP sequences used in target ing and exchange cassettes..........................................73 5-5 Schematic representation of the structure of integrated plasmids used to generate YAC exchange cassettes..........................................................................................74 5-6 Southern analysis of yeast transformants.................................................................75 5-7 Southern Analysis of human -globin locus exchange constructs...........................76


ix Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy IN VITRO AND IN VIVO ANALYSIS OF THE ESTABLISHMENT AND MAINTENANCE OF -GLOBIN LOCUS CHROMATIN STRUCTURE By Padraic P. Levings May 2005 Chair: Jrg Bungert Major Department: Biochemi stry and Molecular Biology Thalassemias are mutations in one or mo re of the globin genes causing defective hemoglobin synthesis and can result in severe anemia. They are the most common single gene disorders and affect thousands of pe ople worldwide. Current therapies for these disorders require regular red cell transfusi ons coupled with iron chelation therapy to reduce the effects of iron overload that can cause organ failure. Bone marrow transplantation is the only currently availabl e cure but is limited to those patients with matched siblings and carries an 80% mort ality rate. Whereas, extensive knowledge concerning the transcripti onal regulation of both the and -globin gene loci has been accumulated, the cis-acting regulatory elements required to achieve therapeutic levels of gene expression in a therapeutic context have yet to be fully elucidated. To this end we have investigated the a ssociation of various transcription factors and chromatin modifiers at the mouse -globin locus during the process of in vitro embryonic stem (ES) cell differentiation. From this we hope to gain insight into the


x elements/factors required for the establishmen t and maintenance of a transcriptionally active chromatin domain. Secondly, we have generated several lines of transgenic mice harboring various regulatory elements from the human -globin locus to test their ability to imbue position-independent, copy number-dependent expression characteristics on a cis -linked -globin gene. Results from the in vitro analysis of the association of transcription factors and histone modification during differentiation of murine embryonic stem cells indicates that the locus control region acts as the primary s ite of recruitment of chromatin modifying activities and transcription complexes prio r to gene activation. These activities are already present at this elem ent in undifferentiated cells and do not appear at gene promoters later in development. Our in vivo studies show that the inclusion of sequences flanking the core regions of HS2 and 3 of th e human locus effectively enhance transgene expression above that of cons tructs lacking these elements , even when integrated to generally repressive chromosomal regions. Fu rthermore, inclusion of boundary elements from the chicken locus may shield these constr ucts from the influence of chromatin at the site of integration. These resu lts not only provide in sight into the mechanisms involved in the activation of hematopoietic genes but may serve to provide more effective means of treatment of diseases associat ed with globin gene expression.


1 CHAPTER 1 INTRODUCTION Background and Significance Hemoglobin is a tetrameric protein composed of two and two like subunits ( that functions to transport the oxygen in al l vertebrate species from the lungs, gills, and skin to that of the capillaries for use in respiration. Hemoglobin was one of the first proteins whose molecular mass was accurately determined, whose existence was associated with a physiological function and the first in which a single point mutation was shown to result in an amino acid cha nge (1). Whereas the hemoglobin protein has served as a model for cooperati ve interactions and allosteric mechanisms of regulation, it is the complex program of developmental st ageand tissue specifi c regulation of the genes that make up the hemoglobin tetramer that has received the most focus in the past two decades. Perturbations of this transcripti onal program can result in a form of severe anemia occurring early in life and associated with splenomegaly and characteristic bone marrow changes termed thalassemia (2). Thal assemias are classified as a heterogeneous group of inherited disorders of hemoglobin sy nthesis classified by reduced or absent synthesis of one or more of the globin pep tides of hemoglobin and are the most common single gene disorders in the world (3). Homozygousthalassemia is a condition in whic h there is a lack or severely reduced synthesis of like globin chains ultimately re sulting in erythroid cell death. Current treatment of this disorder includes re gular red cell transfusi ons that often lead to


2 iron overload and require chelation therapy. The observation that elevated levels of fetal hemoglobin synthesis beyond the perinatal pe riod in some patients results in an amelioration of the severity of the anemia has led to the exploration of alternative treatments that utilize pharmacological inte rvention (3). These treatments involve the administration of various chemotherapeut ic agents such as 5-azacytidine (4), hydroxyurea, and butyric acid analogues (5) ai med at increasing the amount of fetal hemoglobin production. However, concerns of ca rcinogenesis, myelotoxi city, and lack of effectiveness have made these treatments less than satisfactory for treatment of these conditions. Current experimental procedures involve the use of viral vectors to transduce corrected forms of the gene in to stem cells and bone marrow engraftment (6). While gene therapy appears promising we have yet to fully understand the means for ensuring that the transferred gene is properly expressed. In order to do this we must fully elucidate the mechanisms and sequences required to prope rly regulate globin ge ne expression through detailed studies. In humans the genes encoding the and globin loci are located on chromosomes 16 and 11, respectively. Both the and globin gene clusters are multigene loci that are spatially arranged in a linear fashion along the chromosome and the order of this arrangement refl ects the order of their expre ssion in erythroid cells during development. As the developing embryo progres ses from the embryonic to fetal and from the fetal to adult stages of development th ere is a corresponding sw itch in the type of globin mRNA synthesized, resulting in ch anges of the peptide composition of the hemoglobin molecule. These sequential change s in gene expression that occur during development are defined as hemoglobin switchi ng and have been shown to be regulated


3 primarily at the transcriptional level (3). Understanding how these transcriptional switches are regulated and the minimal genetic el ements required to recapitulate them in a therapeutic context would allow for the de velopment of more effective treatments for the thalassemic diseases. Chromatin Structure and Gene Regulation In order for eukaryotic organisms to f it the vast amounts of DNA encoding their genomes into a nucleus only a few microns in diameter it is required that the DNA be condensed and packaged into nucleoprotein structures composed of histone and nonhistone proteins. These structures form the dynamic chromatin polymers that make up chromosomes. Nucleosomes are the basic units of chromatin and c onsist of ~147 bp of DNA wrapped around two subunits each of the hi ghly conserved H2A, H2B, H3 and H4 core proteins. These nucleosomal units are compacted into increasingly more condensed structures ranging from the 10nm filament to th at of chromosomes. W ithin the eukaryotic genomes the chromosomes themselves contain areas differing in their level of compaction and distinguished by varying degrees of sensitivity to nucleases. Nuclease sensitivity is thought to give an indication of the “openness” or degree of decondensation of chromatin. Areas that are nuclease sensitiv e are enriched for transcriptionally active sequences and called euchromatic, whereas ar eas that are comparatively resistant to nuclease digestion are usually transcriptionally silent and heterochromatic. Nucleosomes are stable structures under physiological conditions, are able to selfassociate, and are extremely resistant to physic al perturbations. Furthermore, in order to follow the left-handed spiral formed by th e histone fold domains the DNA must be severely distorted resulting in topological constraints on the accessibility of nucleosome associated DNA sequences to various transacting factors. A fact made evident by the


4 repressive nature of histones to transcription both in vitro and in vivo . A central question arising from these observations is how such a transcriptionaly inhospitable structure could act as the physiological template of genetic information in eukaryotes. Are nucleosomes static structures that are functionally irrelevant to gene expression, simply functioning to package DNA into more manageable dimensions? It was not until Littau and colleagues noted an association between histone acetylation and transcription in eukaryotes that the idea of the passive na ture of chromatin was challenged and its relevance to regulating gene expression explored (7). In addition to the histone fold domains are external tail dom ains found at the Ntermini of all four core proteins. Thes e domains are not required for nucleosomal assembly or positioning although they are re quired for the formation of higher order chromatin structures and may mediate internucleosomal cont acts (8). An extensive body of literature has been amassed studying the a ffects of various histone tail modifications on transcriptional regulation, replication, repair, recombin ation, cell cycle progression, and chromosome segregation. These modifica tions include acetyl ation, methylation, phosphorylation, ubiquitinylation, ADP-ribos ylation, and sumoylation, the various combinations and permutations of which c onstitutes a language termed the “histone code” (9). This code of covalent modificati ons has been shown to be an integral and increasingly complex aspect of DNA meta bolism particularly in terms of gene expression. It is predicated on the hypothesis that specific modificati ons of histone tail residues would result in the modul ation of the affinities of chromatin-associated proteins for their targets. It also states that modifi cations may be interdependent, such that one modification can affect another and that inte rdependent modifications need not be on the


5 same tail. Lastly, it predicts that domains of higher order chroma tin structure may be directly regulated by the local concentration of differentia lly modified nucleosomes (10, 11). These ideas have been thoroughly integrat ed into research studying changes in gene expression that occur without alterations in DNA sequence, or epigenetic changes. The best characterized epigenetic modificat ions are that of ly sine acetylation and methylation as well as that of serine phosphorylation as the systems catalyzing these modifications have been identified (12, 13). Lysine acetylation and deacetylation of core histones is catalyzed by histone acetyltransf erases (HATs) and deacetylases (HDACs), respectively and has been shown to be causa lly linked to changes in transcriptional activity (12, 14). These modifications occur on sp ecific lysine residues in the tails of both histones H3 and H4 and it is t hought that acetylation of thes e lysine residues decreases the overall positive charge of the tail domain. This causes a decrease in its affinity for DNA and thus increases the accessibility of other factors to their cognate binding sequences. Additional research has also show n that various transc ription factors and chromatin remodeling complexes recognize spec ific acetylation marks and bind to these regions using a bromodomain. Two such factors are PCAF and TAFII250 both of which contain HAT activity and that act as activators of transc ription (15, 16) Histone methylation, similar to acetylation, oc curs on specific lysine residues in the amino terminal tails of the core histon es. Until recently it was thought that histone methylation was irreversible, as no histone demthylase had been characterized. A report by Shi et al. identifying the histone demthylase, LS D1 showed that there exist enzymes capable of removing this modification. The fact that it appears to specific for mono and dimethylated lysine four of histone H3 raises many question about its role in


6 transcriptional regulation ( 17). Histone methylation is catalyzed by histone methyl transferases (HMTs). It is thought that in some cases methyla tion of histones is associated with heterochromatin formation as a recent report showed that the Drosophila protein Su(var)3-9 localized to hetero chromatin and that human and S. pombe homologues of this protein are histone H3 lysine 9 specific HMTs in vitro (18, 19). In addition, research shows that two heterochromatin associat ed proteins, HP1 and Swi6, contain a chromodomain that mediates preferentia l binding to methylated H3 lysine 9 in vitro (20, 21). In contrast to H3 lysine 9 methyla tion, methylation of H3 lysine 4 has been associated with transcriptional activation ( 20, 21). These data support the idea that it is not only the type of modifica tion but also its location and the state of modification of neighboring residues of the same and adj acent core histones that dictate the transcriptional state of a genomic locus. Development and Hematopoiesis The use of pluripotent embryoand adult tissue-derived cells for the therapeutic treatment of damaged or dysfunctional tissues has become the focus of a considerable amount of research. In order for stem cell therapy to become a realistic means of treatment and to accurately manipulate lineage choice and differentiation in vitro a detailed description of the molecular and ce llular events involved must be developed. Development is a complex and exquisitely regulated process that requires the proper spatio-temporal regulation of ce ll growth and signaling. Around the time of implantation the embryo is composed of thr ee distinct cell types: the trophectoderm, the primitive endoderm, and the inner cell mass (ICM) . The trophectoderm gives rise to the extraembryonic tissues of the placenta, while the primitive endoderm forms both visceral and parietal endoderm that lines the yolk sac (22). The ICM forms the epiblast that will


7 ultimately give rise to the embryo and other extra embryonic tissues (23). Gastrulation of the embryo results in the formation of the th ree primary germ layers of the embryo from cells of the epiblast. These cells are the proge nitors of all the tissues of the fetus. One of the earliest systems to be established during mammalian development is the hematopoietic system. In the mouse hematopoietic development initiates at approximately 7-7.5 days post coitus (d.p.c.) fr om cells of mesodermal origin that have migrated through the primitive streak (24). Thes e cells are then allo cated into discrete structures known as blood isla nds that are composed of a central focus of developing hematopoietic cells surrounded by primitive angi oblasts. The close spatial and temporal development of these lineages has led to th e proposal that they arise from a common progenitor, the “hemangioblast” (25). This id ea is supported by the fact that these two cell lineages share a number of expresse d genes including CD34, flk-1, flt-1, TIE2, scl/tal-1, GATA-2, and PECAM-1. In addition, se veral genes shown to be involved in the regulation of hematopoietic development (sc l/tal-1, GATA-2, rbtn2) are expressed prior to the establishment of yolk sac hematopoi esis. This fact indi cates that certain populations of early mesodermal precursor ce lls may have already initiated the genetic program that predisposes them to the hema topoietic lineage at a time shortly after gastrulation (26). At approximately 8.5 to 9 d.p.c. the vasculature forms and the heart begins to beat. At this time hematopoietic cells begin to circulate and seed the fetal liver, which becomes the predominant site of he matopoiesis until the bone marrow assumes this role around the time of birth. In the mouse primitive hematopoeitic cells are first detected at about 7.5 days postcoitum (d.p.c.) in the blood islands. Prim itive hematopoiesis is restricted to the


8 production of primitive erythrocytes and m acrophage. Primitive erythrocytes differ from those of definitive origin in that they expre ss different isoforms of globin, are larger, and remain nucleated. This population of cells is only generated in the yolk sac during this stage of development. The exact duration of yolk sac erythropoeisis is not known; production of primitive erythroid progenitors is detectable until day 9 of gestation indicating that they are a tr ansient population existing for no more than 48 hours (27, 28). The primitive hematopoietic system appears to arise from a progenitor population distinct from that of the definitive system. This id ea is supported by a series of experiments involving the identification of tr anscription factors required for all hematopoietic lineages except primitive erythrocytes (29, 30). The onl y other hematopoietic cells present in the yolk sac are macrophages (31). These cells a ppear distinct from those found at later stages due to the fact that they mature mo re rapidly and express certain genes at lower levels (25). By day 10-11 of gestation yolk s ac hematopoiesis declines and the fetal liver becomes the dominant site of hematopoies is although other intraembryonic sites have been identified. A region composed of sp lanchnopleural mesoderm, referred to as the AGM (aorta-gonad-mesenephros) was shown to have hemogenic activity via a number of engraftment studies as well as in vitro culture studies (32-35). These studies showed that only cells isolated from the AGM were cap able of generating a ll adult hematopoeitic lineages supporting the idea that yolk sac and adult hematopoietic stem cells differ in nature. It is believed that yolk sac HSCs ar e devoted to a transient burst of primitive erythropoiesis and that long term repopulati ng stem cells (LTRSCs) required for the production of definitive lineages do not arise until hematopoiesis becomes active within the embryo proper. It is however, important to note that all of the above studies rely on ex


9 vivo culture on isolated embryonic explants and various culture conditions may affect the developmental potential of the engrafted cells. In fact Yoder et al . observed that in the mouse LTRSCs could be found in day 9 yol k sacs and that the AGM may simply represent a site of maturation for yolk sac de rived LTRSCs as little evidence exists that cells of the AGM generates committed precursors (27, 36). In vitro Differentiation of Embryonic Stem Cells As stated earlier the hemat opoietic system is one of th e earliest systems to arise during mammalian development and it is now well established that the yolk sac represents the earliest site of both hematopoi etic and endothelial cel l development. Little is known about the mechanisms underlying the regulation of the even ts leading to the commitment and maturation of cells of mesode rmal origin into cells of these lineages. The minute size of the embryo and number of cells available have made it untenable for the various molecular, cellular and biochemical assays required fo r full understanding of the processes underlying the developmental ev ents following gastrulation. In order to circumvent these problems a number of groups have exploited both the availability and pluripotency of embryonic stem (ES) cells to develop in vitro systems to recapitulate the processes of in vivo development. One system that can be used to study hemat opoiesis exploits the ability of ES cells to form complex three-dimensional structur es that contain developing precursor cells from multiple lineages called embryoid bodies (EBs). When cultured under appropriate conditions, EBs can be used to generate cells of both primitive and definitive hematopoiesis as well as that of the vascular system (37). The ability of this system to recapitulate the events of early embryoni c development has been demonstrated by a number of studies (38). Analysis of the te mporal development of various lineages has


10 shown that primitive erythroid and myeloid lin eages are the first to develop followed by definitive lineages in a fashion similar to that observed in utero . Furthermore, lineage specific expression of marker genes occurs in a pattern consistent with that observed in the developing embryo (39). A second system used to study hematopoiesis is that of the ES/OP9 system. This is a two-dimensional system that relies on a feed er layer of OP9 cells. OP9 cells are stromal cells isolated from the calvaria of newborn op/op mice. Stromal cells are derived from bone marrow mesenchymal cells and are capable of supporting the growth and differentiation of hematopoietic cells, however the rapid proliferation of macrophage often interferes with th e examination of other he matopoietic lineages. The op/op mice have a mutation in the coding region of M-CSF gene that is essential for the differentiation of osteoclasts and formation of the bone marrow cavity (40). These mice suffer from osteoporosis due to defective os teoclast formation and the OP9 cell line was established from these mice to avoid macropha ge proliferation When ES cells are cocultured on stromal cell lines from normal mice the resulting cells are almost entirely macrophage. The OP9 cell line however, has been shown to be capable of supporting the differentiation of both myeloid and B lymphoid lineages in vitro (41). Although both the EB and ES/OP9 systems can effectively recapitulate the events of in vivo hematopoiesis there are se veral differences between the two: The EB system is structurally more complex as it generates cel ls from all three germ layers simultaneously in a three-dimensional sphere. The OP9 system is two-dimensional and is more amenable to observation as one cannot see what is occurring within the embryoid body. EB formation results in the production of hema topoietic progenitors, whereas; the OP9


11 system produces both progenitors and mature blood cells. The OP9 method allows for the production of both megakaryocytes and B cell lineages. The EB method is not capable of generating these cells. Lastly, the EB method is restricted to the ES cell lines that efficiently produce embryoid bodies. No such re strictions exist for the OP9 method as all cells tested to date have efficiently been used in this system (42). Though the two systems have their differen ces, both are useful for the observation and analysis of development of the hematopoietic system. In addition, these two systems have been used in conjunction to generate tr ansplantable hematopoietic stem cells (43). Globin Gene Organization All vertebrate species utilize hemoglobi n as a means of transporting the oxygen critical for respiration. The proteins that compose the hemoglobin tetramer are members of a small family of proteins encoded by the and like globin gene loci. The two loci are located on separate chromosomes in mammals and avians and are composed of several genes that are regulat ed in a tissueand developm ental stage-specific fashion, such that the peptide composition of the he moglobin tetramer varies dependent upon the developmental state of the organism. The -globin gene clusters re side in a linear array along the chromosome, the order of which refl ects their order of expression during red blood cell ontogeny; the more 5’ or embryonic genes are expressed first followed by their simultaneous silencing and activation of feta l and adult specific genes as development progresses. These developmental stage-specif ic changes in activation and suppression of gene expression has been termed hemoglobi n switching and has made the globin loci a paradigm system for the study of how multi-gene loci are regulated. The central focus of globin gene research has been focused upon how the exquisitely regula ted pattern of gene


12 expression is achieved, with a particular emphasis upon the like globins as understanding the molecular events involved in their regulation may provide new insights into treatment of globin-associated hemoglobinopathies. The human -globin locus resides on chromosome 11 and is composed of five genes. The 5’ most embryonic gene ( is expressed first when the primary site of hematopoiesis is within the blood islands of the embryonic yolk sac. Shortly thereafter the embryonic gene is silenced and the fetal globin genes (A and G are activated as the primary site of hematopoeisis switches from that of the yolk sac to the fetal liver. A second switch occurs shortly before birth a nd the two fetal genes are silenced and the adult and globin genes are activated and remain the predominant forms of globin expressed in the adult organism. Approximately six to twenty kilobases upstr eam of the embryonic globin gene is a region of DNA containing a series of elements exhibiting a high degree of sensitivity to the nuclease DNase I. These elements were termed hypersensitive sites (HSs) and the region was dubbed the Locus Control Region (L CR) due to the fact that a number of naturally occurring deletions of this region can result in the disruption of globin gene expression. There exists a wealth of data indicating that cooperation between gene proximal promoter and enhancer elements al ong with that of those found within the LCR is required to achieve proper sp atial and temporal expression of globin genes (44-48).


13 LCR G A HS 5 4 3 2 1 A 3’E 3’E 3’HS1 >100 Kbp >60 Kbp Yolk Sac Fetal Liver Bone Marrow Figure 1-1. The human -globin locus. The five globin genes are depicted as colored boxes with labels above. The arrows indi cate the direction of transcription. LCR hypersensitive sites and 3’ enha ncer elements are labeled below. Horizontal bars indicate the stag e at which the gene is expressed. Globin Gene Proximal Regulatory Elements and Transcription Factors Involved in Erythropoiesis The promoter regions and gene proximal regulatory elements of the -globin locus contain binding sites for both ubiquitous and tissue specific factors. These include TATA boxes, Initiator sequences, GATA, CCAAT, and CACCC motifs (49). GATA binding sites are recognized by members of the GATA family of transcription factors denoted by a novel zi nc finger DNA binding domain, and in the case of the globin loci are most likely bound by the hematopoietic lineag e specific factors GATA-1 and GATA-2 (50, 51). GATA binding s ites are found in numerous erythroid gene promoters and in all LCR elements (51). The importance of these factors in hematopoiesis and erythroid differentiation ha s been shown by studies in which the genes were disrupted in ES cells. GATA-1 null ES cells are unable to contribute to erythropoiesis and GATA-2 mu tant embryos die around 9.5 d.p.c. of severe anemia (52, 53). Because GATA-1 null ES cells show growth a rrest at the proerythroblast stage, it has


14 been proposed that GATA-1 has a role in re gulating the genes required for erythroid cell maturation and commitment (54). Indeed, in vitro hematopoietic differentiation in GATA-1 ablated ES cells revealed a tota l loss of primitive erythropoiesis although a number of putative GATA-1 target gene s, including globins, SCL, erythropoietin receptor, and EKLF, are still expressed (55). Th is may be due to the ability of one of the other GATA proteins to compensate for lo ss of GATA-1. GATA-2 expression increases as much as 50-fold in the absence of GATA-1, suggesting that it is negatively regulated by GATA-1. GATA-2 appears to be necessary for progenitor cell expansion and survival. When GATA-2 null ES cells were introdu ced into normal mouse blastocysts, the resulting chimeric animals revealed that th e GATA-2 mutant cells fa iled to contribute to the erythroid lineage, particularly that of the definitive stage (53). In addition, forced expression of GATA-2 in avian erythrocytes leads to an arrest in differentiation and promotes the proliferation of primitive er ythroid progenitors, indicting that its suppression my be required for progression thro ugh later stages of development (56). It has also been shown that globin gene activation in multipot ential hematopoietic progenitors may be dependent upon prot ein complexes nucleated by GATA-1 and GATA-2 (57). GATA binding site s are found in close proxim ity to canonical Sp1 sites (51) and Merika and colleagues were able to show that GATA-1 was able to physically interact with Sp1 and EKLF via their zinc finger domain s (58). Furthermore, GATA-1 has been shown to physically interact w ith the histone acetyltransferase CBP linking GATA function with that of ch romatin structure alterations that are intrinsic to gene expression (59).


15 NF-E2 is a heterodimeric protein composed of 45 and 18 kd subunits and recognizes an AP-1 like sequence (TGA[C/G] TCA) found in many erythroid promoter and enhancer elements, including that of the and globin loci. NF-E2 motifs often exist in close juxtaposition to that of GAT A motifs and the two have been shown to act synergistically to activate elements of the globin LCR (60). p45 is an erythroid specific protein containing a basic region le ucine zipper motif (b-zip) and a DNA binding domain related to the Drosophila protein cap n’ collar (cnc). The smaller p18 subunit also contains a b-zip domain but is ubiquitous ly distributed and related to the vmaf oncogene. Although it lacks a transactiva tion domain the p18 subunit can still dimerize and bind DNA independent of p45. Using a p45 NF-E2 nu ll immortalized cel l line (CB3) and a dominant negative form of p18 expressed in MEL cells; Kotow and Orkin were able to show that the p45/p18 heterodimer is th e active NF-E2 complex (61). Johnson and colleagues proceeded to show that p45NFE2 was required for histone hyperacetylation and RNA polymerase II recruitment at the adult globin gene promoter (62). Furthermore, homodimers of p45 are incapable of binding NF-E2 sites although homodimers of small maf proteins retain this ability, indicating that changes in dimer composition may play a role in regulating se ts of erythroid gene s (63, 64). Research involving p45 NF-E2 null mice indicate that the p45 peptide may be dispensable for erythropoiesis because only subtle pertur bations in red cell development such as microcytosis and decreased hemoglobin conten t have been observed (65). Thus, although p45 NF-E2 may be required for optimal globin gene expression, other AP-1 like proteins may be able to substitute for the transcri ptional activitie s provided by p45. Interestingly, p45 NF-E2 null mice do suffer from a comple te absence of platelets as well as


16 hemorrhaging resulting in extreme morbidit y during the neonatal pe riod, indicating that p45 is indispensable for certain hematopoietic functions. Erythroid Krppel-like factor (EKLF) is an erythroid and mast cell specific transcription factor containing three TFIII A-like zinc fingers homologous to other Krppel-like factors. EKLF is expressed in both primitive and definitive erythroid lineages and its promoter contains a GATA motif shown to be required for its activation. EKLF binds to elements c ontaining a CCACACCT sequence su ch as that found in the globin gene promoters and LCR cores. Naturally occurring mutations of this element in the adult globin gene promoter lead to thalassemia in humans (66). Although other zinc finger proteins such as Sp1 are al so capable of binding CACC sequences, the in vivo role of EKLF has been more accurately defi ned using gene-targeting experiments. EKLF null embryos die of anemia during the fetal stage due to severe thalassemia (67). EKLF activity appears to be rest ricted to adult lineages as both embryonic and fetal hematopoiesis is unaffected, however, it may play some role in the feta l to adult switch in globin gene expression as EKLF null mice carrying a human globin locus transgene display a delayed fetal to adu lt switch (68). Over-expression of EKLF in fetal erythroid cells enhances globin gene expression more than a 1000-fold and mutation of the globin gene CACCC element abrogates this affect. Homozygous knock out of EKLF in mice results in persiste nt expression of human -globin transgenes in adult erythroid cells (68, 69). It appears that the globin gene is principally re gulated in an autonomous fashion and that the adult globin gene is active at basal le vels during the yolk sac stage, suggesting that the role of EKLF in fetal to adult globin gene switching is not primary.


17 An additional factor that has been shown to play a critical role in hematopoietic development is stem cell leukemia factor (S CL). It is a basic helix-loop-helix (b-HLH) protein expressed in several hematopoietic lineages. Abnormal expression of SCL is associated with acute T-cell leukemia (70). Many b-HLH prot eins can act as transcriptional activators and bind DNA speci fically at E-box elements (CANNTG) found in a variety of erythroid promoter a nd enhancer elements (71). Gene targeting experiments in mice have shown that SCL is critical for red blood cell development and SCL anti-sense RNA blocks differentiation of MEL cells (72). SCL mutant mice die around 8.5-9.5 d.p.c., exhibiting a total abse nce of nucleated red blood cells and mimicking the defects caused by loss of GAT A-1. The SCL gene itself may be regulated by one or more GATA factors as one of its two promoters contains a GATA site. Locus Control Regions and Hemoglobin Gene Switching The globin locus control region LCR is a distal regulatory element located approximately 6-22 kb(73) upstream of the globin gene. The LCR is characterized by a series of elements that are extremely accessibl e to the nuclease DNase I in erythroid cells. The core regions of these hypersensitive sites are roughly 200-400 bp and are separated by one to two kilobases of flanking sequen ces. The cores contain a high density of binding sites for both ubiquitous and tissue-sp ecific transcription factors. The overall structure of these domains is conserved in several species, suggesting a functional significance (74). The motifs that are most conserved are maf recognition elements (MAREs) and GATA sequences in HS2, 3, and 4, KLF sites in HS2 and 3, as well as an E-box motif in HS2. The GATA sites are mo st likely bound by GATA-1 or GATA-2, as these are the only known GATA factors expr essed in erythroid cells. The MARE sequences are most likely bound by heterodimers of small maf proteins and other b-zip


18 family members such as p45 NF-E2 and Bach 1. The CACCC elements in HS2 and 3 are most likely bound by EKLF in vivo as transgenic mice lacking EKLF show a specific decrease in hypersensitivity at HS3 as s hown by Lee and colleagues (73). E-box motifs are bound by helix-loop-helix protei ns such as USF and SCL and in vitro studies indicate that the E-box in HS2 is bound by both th ese proteins, alth ough the functional significance of these interacti ons have yet to be elucidated (73). Protein-protein interactions most likely play a critical role in LCR function as most of the proteins associated with HS elements have the potenti al to interact with one or more partners, perhaps to mediate the formation of higher or der structures that are involved in changes in gene expression and chromatin struct ure observed throughout the locus (46). Locus control regions are defined as elements capable of conferring positionindependent and copy number-dependent expression on cis -linked genes irrespective of chromosomal position. They are capable of directing tissue-specific expression at physiological levels and may also be involved in the replication timi ng of gene loci. The founding member of this cla ss of regulatory elements is that of the mammalian globin locus and accordingly an extensive amount of research has been aimed at elucidating its role in globin gene regu lation. However, the exact in vivo activities of the LCR are still a matter of much debate, the reasons for which are manifold. It is generally agreed now that the LCR is required for high-level expression of all th e genes at all developmental stages. Whether or not the LCR simply act s as a classical, a lthough somewhat more complex, enhancer is one of the fundament al questions yet to be answered. The importance of the LCR was first realized th rough the study of a series of naturally occurring deletions of the region, which result in the cl inical manifestation of


19 thalassemia. The smallest of these deletions (Hispanic Thalassemia) removes just 35 kb of sequence upstream of the globin gene but results in the silencing of all the genes and a loss of general DNase I sensitivity th roughout the remainder of the locus. This observation indicated that the LCR must contain some positive cis -acting elements required for the activa tion of the locus and expression of all the globin genes (47, 75). Transgenic studies found that globin transgen es were expressed at variable levels and only in a small portion of animals, indicat ing the constructs were subject to position effects. However when linked in cis to an LCR or combinations of various HS sites, the transgenes were expressed at physiological levels in a position-independent and copy number-dependent manner (76). These resu lts indicated that not only could the globin LCR act as a strong enhancer but also har bored a dominant chromatin opening activity. Although all of the above data suggest a prominent role for the LCR in globin locus activation and resistance to the nature of chromosomal position effects other research argues against the same function at endogenous loci. The individual HS sites themselves have been shown to contain both unique as well as redundant activities. In one group of e xperiments deletion of a 375 bp core element from HS2 in the context of an otherwise comp lete yeast artificial chromosome containing the human globin locus resulted in cat astrophic reductions in gl obin gene expression as well impaired HS site formation(77). Re placement of HS2 by HS3 resulted in the restoration of HS formation and thus chromatin opening, but was unable to fully compensate for the transcriptional enhancement activities of HS2 (78). In a similar series of studies by the same group it was found that deletion of HS sites 3 and 4 resulted in a drastic decrease in expression of all the gl obin genes at all stages . When HS3 replaced


20 the deleted HS4, proper developmental stage expr ession at physiological ly relevant levels was restored. However, when HS4 was used to replace HS3 gene expression levels was attenuated at all stages (76-78). In a parallel group of experiments Peterson et al ., created larger deletions of HS2 and 3 encompassing not only the cores but the flanking sequences as well(79). These deletions resulted in a decrease in globin expression in the case of HS3 and a minor decrease in andglobin gene expression when HS2 was deleted, contrasting the results of deletion of the core HS elements alone (79). These results provide some insight in to the discrete natu re of LCR elements and suggest that a certain functional synergism is required for fu ll activity. They also suggest that certain sequences act as structural components for the formation of a higher ordered structure such as that postulated in the holocompl ex model of LCR activity (68, 77). The hypothesis is that the LCR HS sites physica lly interact to form a higher ordered chromatin structure and that this interacti on is mediated by protei n-protein and proteinDNA interactions (see below). Studies of LCR function at the endo genous human and mouse loci provided somewhat contradictory results. In one such study Reik and colleagues used homologous recombination and the DT40 shuttl e system to generate human globin LCR mutations and study their effects in MEL cells. The authors found that de letion of the LCR in this system resulted in the loss of globin gene e xpression but that the ch romatin structure of the remainder of the locus was essentially unchanged. They concluded that for the human locus the LCR was necessary for high-level gene expression, but not for the maintenance of chromatin structure (80). A similar dele tion of the mouse LCR resulted in a drastic decrease in the expression of all the gene s although temporal regulation was maintained


21 and HS site formation at the promoters wa s unperturbed (81, 82). These results suggest two possibilities: that the L CR simply acts as an enhancer ensuring that the genes are expressed at high levels, or that some re dundant chromatin opening activities 5’ to the currently defined LCR exist. In both mouse and humans the globin locus is embedded in inactive odorant receptor loci and HS sites outside the current boundaries of the defined LCR have been shown to have so me impact on globin gene expression and formation of higher order chromatin structur es associated with developmental switching. Interestingly a recent report by Farrell and colleagues observed that an 11 kb deletion of HS sites 5’ to the LCR sequences had little or no effect on HS site formation and gene expression that may shift the focus of current research to sequences even farther upstream (83). The fact that the function of the LCR, when integrated at ectopic sites, appears to differ from that of the locus in the context of its endogenous location within the genome has confounded attempts to fully integrat e the entire body of research into a comprehensive model for globin gene switching. Consequently several models have been formulated describing this phenomena, the two most prominent ar e the “linking” and “looping” models of gene activation. The linking model states that stage specific activities bound at the LCR are transmitted via chromatin associated facilitator proteins that interact with factors in the transcriptionally active domain (84). The looping model posits that the LCR holocomplex interacts physically with sp ecific gene promoter and proximal regulatory elements by looping of th e intervening DNA (89). This interaction allows for the transfer of activities first recruited to the LCR to be deposited at the genes at the proper time. Two recent reports indicate that not only do sequences within the LCR


22 and active globin genes reside in close proxim ity in the nucleus but that the conformation of such interactions is dynamic and depe ndent upon the developmental state (44, 48). Whereas both models are consistent with curr ent data there is no irrefutable evidence supporting either, although the looping model has gained support in recent years. Summation The complex pattern of gene expressi on and dynamic changes in chromatin structure that are associat ed with expression of the -like globin genes have made hemoglobin gene switching a paradigm for th e study of how coordinated regulation of multi-gene loci can be achieved. Although a complete understanding of the events leading to the activation and sile ncing of particular genes has yet to be obtained they can conceptually be divided into several steps: generation of a highly accessible holocomplex, recruitment of transcription f actors and chromatin modifiers to the LCR, activation of stage specific globin sub-domai ns, and transfer of transcription complexes to the appropriate gene promoters. The initial event of globin gene expre ssion must be an opening of the domain rendering is accessible to various transacting factors; an event that may occur prior to commitment to the erythroid lineage a nd supported by the fact LCR HS sites are detectable in uncommitted progenitors (85). Hematopoieitic-specific transcription factors expressed in uncommitted progenitors could diffuse into inactive or non-permissive chromosomal domains and bind to target se quences within the locus. These initial binding events could result in a general increa se in accessibility of the locus to other factors such as HATs and ATP-dependent ch romatin remodeling complexes as well as sequestration of the locus in an active region of the nucleus. GATA factors may be


23 involved at this step as data indicates that they are expressed in multi-potential progenitor cells and GATA binding sites are found through out the locus (86). In addition, GATA-1 is known to associate with the HAT CBP, an in teraction that stimulates GATA-1 activity in transient transfection assays (87). The ge neral opening of the locus does not appear to require activities present within the LCR as even when it is deleted the remaining elements retain nuclease sensitivity and incr eased levels of histone H4 acetylation (82, 88). The increased permeability of the locus to proteins present within the nucleus and the high concentration of bindi ng sites within the core LCR HS elements most likely results in the accumulation of these factors at the LCR, perturbing the DNA. It is this perturbation that most likely results in hypersensitivity and holocomplex formation as many of the proteins involved not only c ontain DNA binding domains but protein-protein interaction domains as well. Once formed the LCR holocomplex could pot entially act as a sink for the macromolecular complexes that direct ly mediate transcription, first recruiting them to the locus and than distributing them to globin gene pr omoters in a developmental stage-specific fashion. In vitro studies indicate that RNA Pol II is present at LCR HS2 in undifferentiated ES cells prior to its appearance at promoter s (89). It is possible that recruitment of transcription complexes to th e LCR could further serve to modify the subdomains. Transcription of the LCR itself as we ll as that of intergenic regions has been observed and movement of polymerase comp lexes through these regions may facilitate the formation of transcriptionally active s ub-domains (90). These intergenic transcripts have been used to delineate developmental stage-specific chromatin domains within the locus, however their exact role is not curren tly known. They may simply be a bi-product


24 of active transcription, although arguing against this the fact that deletion of the adultspecific transcription initiation site results in a decrease in general DNase I sensitivity of the domain as well as a reduc tion in expression of the globin gene (90). Despite this lack of information concerning the role of intergenic transcripti on, the idea of stage specific sub-domains is supported by obser ved changes in the degree of DNase I sensitivity and patterns of hi stone modification that occu r during development (88, 91). These changes are most likely due to the dyna mic association of f actors involved in the regulation of chromatin st ructure including, but not lim ited, to histone acetyl transfereases and deacetylases as well as nucleosome remodeling complexes. The ultimate outcome of stage-specific reconfiguration of the globin locus would be the direct interaction of the LCR with th e appropriate gene promoter and the inhibition of promiscuous interactions with regula tory elements outside the domain being transcribed. This would result in the transf er of activities requ ired for high-level expression of the globin genes from the LCR to these promoters. Evidence supporting direct LCR-promoter communication come s from studies showing a localized hyperacetylation of histone H3 at the LCR and active gene promoter that is not observed in the absence of the LCR, an event consiste nt with the “looping” model of LCR function (88). In addition, HS2 and HS3 have been shown capable of recruiting RNA PolII in vitro and in vivo (62, 92). The in vivo data indicate that the recrui tment of PolII to the LCR and its transfer to the globin promoter requires NF-E2 as in MEL cells lacking this protein PolII can be found at the promoter but is absent from the gene. Although this is a simplified view of the events occurring at the globin locus during development it provides the framework from which to base further studies into the


25 molecular nature of developmentally re gulated changes in gene expression. A fundamental understanding of these processe s will be critical in identifying genes required for proper development and ame liorating conditions brought about by their disregulation.


26 CHAPTER 2 MATERIALS AND METHODS ES cell differentiation . Mouse ES cells were differentiated to generate cells of the hematopoietic lineage using the ES/OP9 met hod established and described by Kitajima et al .(42). Briefly, ESD3 cells (ATCC, CRL -1934) were seeded onto a confluent monolayer of mouse embryonic fibrobl asts (MEFs) at a density of 105 cells/25 cm2 in ES media (DMEM, 4.5g/l glucose, 1.5g/l sodium bicarbonate, 15% FBS, 0.1 mM 2mercaptoethanol and 106U/ml LIF), grown for two days, and then passaged (1:6) and grown for another day. An aliquot of the cells (3-4x107) was taken at this time (Day 0) and subjected to RT-PCR and ChIP analysis. The remaining day 0 cells were then seeded onto confluent OP9 stromal cells in OP9 media (-MEM with ribonucleosides and deoxyribonucleosides; 20% FBS) in the absence of LIF at a density of 104 cells/well in 6 well tissue culture dishes. At day 3 Epo or Epo and SCF was added (2U/ml and 50ng/ml, respectively) for the remainder of the course of induction. On day five of induction cells were passaged and reseeded onto fres h OP9 cultures at a density of 105 cells/well. On day 12 cells were collected and subjected to RT-PCR and ChIP analysis. RT-PCR . RNA was isolated for RT-PCR us ing the Arum Total RNA Mini Kit (Bio Rad) according to th e manufacturer’s protocol. Reverse Transcription was performed using 200 to 250 ng RNA and the iS cript cDNA synthesis Kit (Bio-Rad) as described by the manufacturers protocol. PCR amplification was performed using the Eppendorf PCR Mastermix (Eppendorf). Primer sequences specific for Flk-1, Epo-R, GATA-2, -actin, Rex-1 were obtained from Elefanty et al. (93). Additional primers used




28 5’GTGTGCACTGGGTGGGTAGA 3’; IVR3: US 5’ TGTGCTAGCCTCAAGCTCACA 3’ and DS 5’ TCCCAGCACTCAGAAG AAGGA 3’; Mouse Rex-1: US 5’AACTGCATCCTCTGCTTGTG 3’ and DS 5’ TGCGCTCTATTTCCTCCTTG3 ’. Antibodies: TFIIB sc-225, Po l II (N-20) sc-899, NF-E2 (C -19) sc-291, (all purchased from Santa Cruz Biotechnology), Pol II 05-623, hi stone H3 di-methylated at lysine 4, and acetylated histone H4 (all purchasd fromUpstate Biotech.). All antibodies were tested in Western blotting experiments us ing MEL or K562 nuclear extracts as described by Leach et al. (94). Transgene Construction. Components of the chicken and human globin locus were combined to generate the plasmid 432 4. Briefly, the core region from human HS3 was isolated as an XbaI-XhoI fragment from the plasmid HS434 (77). The HS2 core was generated by PCR from a YAC containing the entire human globin locus as a template (YACA201F4.3; (95)) using th e following primers: HS2US 5’ACCTCGAGCCCTCTATCCCTT CCAGCATCC 3’; and DS, 5’ACGATTCGAATATCACATTCTGTCTCA 3’; XhoI and EcoRI sites were included 5’ and 3’, respectively, to facilitate cloni ng. The fragments were cloned between the XbaI and EcoRI sites of the plasmid pGEM 7 to create pGEM7HS32cores. HS4 from the chicken globin locus was PCR amplified from chicken genomic DNA using primers that introduced AatII and SphI sites 5’and 3’ to the 250 bp core element (96): 5’ HS4US, 5’ ACGACGTCGAGCTCAGGGGACAGCCCCCCC 3’; DS, 5’GTGGACCCCCTATGCCCCTT TTGCATGCAC 3’. The resulting fragment was cloned into pGEM7HS32cores us ing AatII and SphI sites. From this plasmid a SacI-KpnI fragment containing all three core regions (chicken-HS4, Human-HS3 and -2 cores) was


29 isolated and cloned into pUC19 to create pU C432cores. Into this plasmid the human globin gene/3’enhancer was cloned as a 4.6kb KpnI-XbaI fragment isolated from the plasmid A/X (92)using KpnI and XbaI sites. Finally the 3’ chicken HS4 was PCR generated to contain 5’ and 3’ SalI sites: 3’HS4US, 5’ATATGTCGACCTCACGGGGACA GCC 3’; 3’ and DS, 5’ CCCGGTCGACCCCCGTATCCCCCCA 3’. The resu lting fragment was cloned into the SalI site of pUC432cores to create the plasmid 432 4pUC. The plasmid containing the -globin integration construct (p43f24) was constructed using the pNEB193 plasmid (New England Biolabs) as the backbone. A 500 bp EcoRI-PstI fragment containing the AAV2 5’ITR and p5 promoter was blunt ended on the 3’ end and cloned into the EcoRI and SmaI sites of pNEB193. A linker contai ning PacI, SpeI, Ns iI, ApaI and XbaI restriction enzyme sites were then cloned in to the PacI and XbaI sites. The 246 bp 3’ cHS4 fragment was PCR amplified using the 3’HS4 primers described above and ligated into the SalI site of the vector. Next a 3.8 kb NsiI-XbaI fragment containing the human -globin gene including the promoter and 3’ enhancer elements was cloned into the respective sites in the vector. A 6.9 kb NsiI fragment containing the human -globin LCR from HS3 through HS2 was ligated into the vector and clones with the correct orientation were identified by restriction enzyme analysis. The 5’HS4 fragment was PCR amplified using the following primers: HS4 US, 5’CCTTAATTAACTCACGGGGACAGCC-3’, HS4DS 5’CTAGTCTAGACCCCGTATCCCCCA3’ introducing PacI and XbaI sites at the 5’ and 3’ ends, respectively. This fragment was ligated into the PacI and SpeI sites of the vector. This cloning step removed 742 bp of the 5’ end of the HS3 flank HS2 LCR


30 fragment previously ligated into this vector. The complete p43f24 plasmid was purified using Qiagen’s maxiprep kit. Both plasmids were sequenced to verify the integrity of the regulatory elements. The 3.6 kb EcoRI-KpnI fr agment containing th e human AAVS1 site was ligated into pBS246 (Invitrogen). Transgenic Mouse Production. We used FvB/B6 mice (Jackson laboratories) to generate all transgenic lines. Plasmid p4324 was linearized with AatII; pAAVS1 was linearized with EcoRI. The linearized plas mids were purified from agarose gels and resuspended in injection buffer at a concentration of 2 ng/l. Transgenic mice were generated as described previously (77). Tr ansgenic founders were first identified by PCR on DNA isolated from tail clips. Copy number and integrity was analyzed by Southern blotting. AAVS1 transgenic mice were mated to generate mice homozygous for the transgene. In experiments using the -globin integration construct (p43f2 4) 1-5 ng of the supercoiled plasmid DNA was complexed on ice with a 1:5, 1:10, or 1:15 molar ratio of DNA to purified AAV2 rep68 protein and inje cted into fertilized oocytes homozygous for the AAVS1 transgene. DNA Isolation, PCR Screening, Inverse PCR, and Southern Blot Analysis. DNA was isolated from mouse tail a nd the presence of human AAVS1 or -globin locus sequences was first determined by PCR using primers against the human AAVS1 site and the flanking region between human -globin HS3 and HS2: AAVS1 US 5’ATCTGCCCGGCATTTCTGAC 3’, AAVS1 DS 5’CGCAAAATGTCGCAAAACAC 3’. The primers amplifying a region be tween HS2 and HS3 were published by Leach et al. (94). DNA from tails that contained AAVS1 and/or human -globin sequences were then subjected to Southern bl ot analysis. Approximately 10 g of tail DNA was digested


31 with restriction enzymes, size fractionate d on 1.2% agarose gels, and transferred onto Nylon membranes as descri bed (78). The membranes were then probed with DNA fragments corresponding to regions of the -globin transgene or of the human AAVS1 site. The mid probe is a 917 bp BamHI-E coRI fragment derived pA/X (92) and corresponding to the codi ng region of the human -globin gene. The 3’ -globin probe is a PstI fragment encompassing the -globin 3’enhancer and derived from pA/X. The AAVS1 probe was derived by PCR using the primers described above (AAVS1 US and DS). Copy number of transgenes was de termined by hybridizing the nylon membranes with a radioactive probe corres ponding to the murine snrp N gene. This probe is a 300bp EcoRI/SacI fragment derived from the snrp N locus which hybridizes to a 4kb EcoRI fragment in Southern blotting experiments. This probe was made available to us by Dr. Camilynn Brannan (UF). Inverse PCR was carried out as describe d by Hartl and Ochman (97). Briefly, genomic DNA from S1 transgenes was digested with SacI or MspI, ligated and subjected to PCR using the following primers, S1tg DSI: 5’CACAGCCCCAGGTGGAGAAA CT3’, S1tg DSII: 5'CCCGGGTTGGAGGAAGAA GACT3’, S1tg US: 5’TTCTCCAGGCAGGTCCCCAA3’. PCR produc ts were subcloned into the TopoII vector (Invitrogen) for sequencing. Metaphase Preparation and FISH Analysis. F1 animals containing the transgene were sacrificed and the spleens were isol ated in 2-3 ml ster ile PBS for metaphase chromosome preparations (98). Cells were isol ated from the spleen and pelleted in a total volume of 10 ml PBS at 500g for 10 min. The cells were then immediately resuspended


32 in 10 ml 0.075M KCl (prewarmed to 37C) and incubated at 37C for 20 min. 2 ml fixative (3 parts methanol: 1 part glacial acet ic acid) was immediately added to the cells and the cells were pelleted at 500 g for 10 min. The metaphase cells were washed 3 times in 10 ml fixative and stored at 4C in fixative. Metaphase cells were placed on microscope slides and allowed to air dry. The slides were aged in an 80C incubator for 1 hr and then immediately used for FISH. In the dark, 10-15 l fluorescently labeled probe was pl aced on each slide, covered with a glass slide, and then placed in a HyBrite (V ysis) apparatus overni ght where the slides were denatured at 75C for 15 min and hybridized at 37C for16 hrs. Cover glasses were then removed in the dark, a nd the slides were washed for 2 min at 75C in 0.4 X SSC, 0.3% NP-40, followed immediately by 1 wash in 2 X SSC, 1% NP-40 for 1 min at room temperature. The slides were allowed to ai r dry in the dark and counterstained with 10 l DAPI II solution (Vysis). The slides were visualized by fluorescence microscopy or stored in the dark at 4C. The entire -globin transgene (p43f2 4) or the AAVS1 plasmid was fluorescently labeled using the BioPrime kit (Invitrogen) by substituting rhodamine-tagged dUTP (Tetramethylrhodamine-5-2’-deoxy-uridine-5’triphosphate, Roche) for the biotin-tagged dUTP that comes with the kit. The ch romosome paint probes specific for mouse chromosomes 7 and 15, were obtained from ID Labs Inc. and used according to the manufacturer’s in struction. RNA Isolation and Semi-quantitative RT-PCR . F1 animals containing the globin transgene were made anemic by injec ting phenylhydrazine as described previously (77). RNA was extracted from the spleen and cDNA was prepared as described in


33 Bungert et al .(77). 10% of the RT reaction was used for semi-quantitative PCR analysis using primers against the human -globin and mouse -globin genes using primers published previously(77). We used a new mouse -globin downstream primer to span an intron in these experiments: ma DS , 5’TCCACACGCAGCTTGTGGGCATGCAG 3’. PCR samples were removed at 14, 16, and 18 cycles and size fractionated on a 10% polyacrylamide gel. The gels were stai ned with SyBr-green and quantitated by phosphorimager analysis using a storm scanner. Human -globin transgene expression level was calculated relative to the expression level of the endogenous mouse -globin gene. DNase I Hypersensitivity Analysis. Cells taken from a spleen of mice made anemic by phenylhydrazine injection (see RNA isol ation section) were washed with PBS, pelleted, and subjected to DNase I digestion as described by Kang et al. (99). 10 g of DNase I treated DNA was digested with EcoR I, size fractionated on a 0.8% agarose gel, and subjected to Southern blot analysis using the -mid probe as described above. YAC Modification. Generation of mutant human -globin locus YACs by homologous recombination in yeast was carri ed out as described previously (77). Flanking regions for recombination vectors were constructed by P CR amplification. The vectors were linearized within the 5’ or 3’ flanking regions and used to transform yeast cells containing the human -globin locus YAC (A201F4.3). Transformed yeast cells were plated onto agar ose plates lacking uracil. DNA was prepared from cells growing on uracil-deficient medium and analyzed by Southern blotting for correct integration of the loxP containing mutation. Clones that had the plasmid integrated into the homologous site of the -globin locus YAC were grown in uraci l-containing medium and plated onto


34 agarose plates containing 5'-fluoroorotic acid (FOA). Growth on FOA-containing plates indicates removal of the URA3 gene by homologous excision within the YAC, generating cells with either wild type or the desired mutant -globin locus structure. DNA was prepared from these cells and analyzed by Southern bl otting for of the inserted loxP site our selectable marker. Yeast clones bearing the mutant -globin YACs were embedded in agarose plugs for pulsed-field gel el ectrophoretic analyses or DNA isolation. Four inserstion vectors were generated fo r introducing wild-type (L1) or mutated (L2) loxP sites, as well as a puromyc in selectable marker into the human -globin YAC. Sequences for mutated and wild-type sites di ffered by a single G-A basepair mutation in the spacer region of the site ( 100). For insertion of the L2 site 5’ to the LCR a vector, 5’LCRL2, was created by PCR using sequences approximately 3 kb to the 5’ EcoR1 at the 5’ of the locus as defined by the HUMHBB sequence. The 5’ homology of this vector was approximately 500 bp and was de limited 5’ by a HindIII site and 3’ by a BamH1 site. The 3’ homology was 700 bp and contai ned SpeI and SfiI restriction sites at its 5’ and 3’ ends, respectively. The fragment s were cloned into the pBS246 loxP vector (Gibco). An L2 site was introduced as a synthetic oligonucleotide ligated into unique BamH1 and SpeI sites . In addition to the loxP site these oligos also contained a unique RsrII site that would later be ued for removal of the YAC vector arms and circularization. The cassette was cloned into th e pRS306 plasmid as a NotI fragment, linearized using a unique BtrI site that cuts in the 5’ homology and introduced via electroporation into the A201F4.3 strain of S. cerevisiea . Similar strategies were used for producing the 3’LCR L2 and 3’ -globin genes L1 (3’BGL1) construc ts. The 3’LCRL2 construct contained regions of homology corresponding to bp 13000-13900 (5’ homology) and 13900-15100


35 (3’ homology) of the HUMHBB locus and c ontained unique, PCR generated EcoRV and BamHI sites (5’homology) and SpeI and SfiI sites (3’homology). BamH1 and SpeI sites were included at the 5’ and 3’ ends to facilitate cloning of th e L2 site into the vector. The same L2 oligo was used to introduce the L2 site between the 5’and 3’ homologous segments with the following exceptions: a SmaI site was included at the 5’ end and a BamHI site was present at the 3’ end. Agai n, these fragments were sub-cloned into the pBS246 vector followed by cloning into the pRS vector as a Not I cassette. The vector was linearized in the 3’ homology using X hoI. The 3’BGL1 plasmid utilized sequences ranging from 66000 to 68100 for its 5’ and 3’ homo logies. The wild-type loxP site in this vector was derived from the pBS246 plasmid a nd the final vector was linearized in the 5’homology using SphI. All fini shed vectors were sequenced to confirm the presence of L1 and L2 sites. The puromycin gene inse rtion vector was creat ed using homologous sequences spanning bp’s 65423 to 66900 of the HUMHBB locus. PCR generated homologies were ligated into pRS306 and th e puromycin gene was subcloned as AscI fragment from the vector pKOSelectPuro (Str atagene). The vector was linearized with the enzyme BbvCI prior to electroporation. L2 Oligos: 5’LCRL2, Forward: 5’ GATCCGGTCCGATAACTTCGTATAATGTATACTATACGAAGTTATA 3’; Reverse:5’CTAGTATAACTTCGTATAGTATACATTATACGAAGTTATCGGACCG G 3’; 3’LCRL2, Forward: 5’GGGCGGACCGATAACTTCGTATAATGTATACTATACGAAGTTATAG3’ Reverse:5’GATCCATAACTTCGTATAGTATACATTATACGAAGTTATCGGTCCG CCC3’


36 CHAPTER 3 ESTABLISHMENT OF PATTERNS OF HISTONE MODIFICATION AND FACTOR RECRUITMENT WITHIN THE GLOBIN LOCUS DURING IN-VITRO DIFFERENTIATION OF MURINE EMBRYONIC STEM CELLS Introduction Multicellular organisms are com posed of a variety of cell types, all derived from a common precursor and characterized by different patterns of gene e xpression. It is the transcriptional profile of a specific cell type that determ ines its morphology and function. The establishment of the expres sion patterns of terminally diffe rentiated cells is mediated by various ubiquitously expre ssed and tissue-specific transcri ption factors and repressors, as well as nucleosome modifying and remode ling factors whose activity results in the proper spatial and temporal expression of specific subsets of genes. The sequential silencing of genes involved in maintenance of pluripotent and multi potent states and the activation of those involved in differentiation is believed to be a dominant factor in the progression from multi-lineage precursors to that of specific cell typeS. The maintenance of this transcriptional stat e following cell division depends upon not only the direct action of transacting factors, but also the heritable epigenetic status they impart. Data accumulated in recent years indicates that combinations of covalent histone modifications may constitute a “histone code” that regulates the use of genetic information (10). The manner in which the acquisition of various epigenetic states is regulated during development is only partially understood. The vertebrate globin gene family ha s provided a model system to study the molecular basis of developmentally regulated differential gene expres sion. It contains a


37 number of tissue-specific ge nes that are coordinately re gulated and whose expression changes during development of the hematopoi etic system, a process termed “hemoglobin gene switching”. Epigenetic modifications have been shown to play an important role in the expression of the like globin genes. The chicken globin locus has been shown to reside in a domain of uniform histone hyperacetylation with the active genes being acetylated on lysine 4 of histone H3 a nd inactive genes exhibiting H3 lysine 9 methylation (101, 102). Differential acetylation has also been observed in the murine globin locus. Forsberg and colleague s observed dynamic changes in histone acetylation of the globin genes during deve lopment with the LCR and active genes marked by increased H3 and H4 acetylation (91). These observations suggest epigenetic modifications may be an important factor in the maintenance of an active locus, however, how and when these patterns are established is not entirely known. Bottardi and others investigated the epigenet ic state of the human globin locus in hematopoietic progenitor cells (HPCs) and transgenic mice (103). They found that histone H3 at the promoter was hyperacetylated and dimetylated at lysine 4 in HPCs but deacetylated in mature erythroid cells. In contrast, the human promoters lacked these modifications in HPCs and transgenic fetal liver cells. These results indicate acetylation pl ays a critical role in the transcriptional potentiati on and developmental regulation of these genes in progenitor cells or cells that have yet to express th e genes at physiologically relevant levels. Chromatin structure modifications in uncom mitted progenitor cells have also been observed for the murine globin locus (85, 104). A recent study by Johnson and colleagues showed that RNA Pol II is recruited in a strictly localized fashion within the LCR and was only detected at the core regi ons. Localization of Pol II to the LCR was


38 independent of active transcriptional elonga tion; the addition of DRB did not affect recruitment (105). The observation that a gene can exhibit a chromatin structure similar to that of active loci in precusor cells has b een made at other loci, such as the lysozyme locus (106), c-fms (107), and the myelopero xidase gene (108). Understanding how epigenetic states are acquired during devel opment and how they impact globally on gene expression is a critical step in the treatment of a number of diseas es, ranging from birth defects to cancer (109). A logi cal first step in this proce ss would be to determine the mechanisms involved in this process at the le vel of individual genes. Previously we had proposed a model describing the processe s involved in the activation of the globin locus during development (46). A key aspect of this model was the initial transition in chromatin structure from a closed, inaccessibl e conformation to that of a more open one. We hypothesized that this first step could be mediated by tr anscription factors such as GATA-1, which would bind first at the LCR and perturb the st ructure of nucleosomes in this region. This would be followed by the re cruitment of various chromatin remodeling complexes and transcriptional co-activat ors as well as complexes containing RNA polymerase II (Pol II). This would result in the nucleation of a transcriptionally competent chromatin conformation that could be perpetuated to gene promoter proximal regions to activate transcription. Thus, we proposed that the LCR acts as a center of attraction at early stag es of hematopoietic development to ensure the proper activation of the genes in erythroid cells. In this study we wished to investigat e the hypothesis that chromatin structure modifications and factor recruitment occurs first at the LCR and sunsequently to the genes. We analyzed the events of Pol II recuitment and chroma tin structure alterations at


39 the murine globin locus in un-induced embryonic stem cells (day 0), mesodermal cells (day 5), as well as that of primitive and definitive erythroid cells (day 12). Using chromatin immunoprecipitation (ChIP) we demons trate that core elements of the LCR adopt a structure characteristic of transc riptionally active chro matin and recruit RNA polymerase II prior to erythroid differentia tion in murine ES cells. Real time PCR analysis indicates that the locus is first ac tivated at the LCR and that this state is perpetuated to more distal regions as the process of differentia tion proceeds. Histone modifications and factor recruitment correspond ing to a transcriptionally permissive state appear to be acquired pr ior to gene expression Results We examined the association of Pol II with the murine -globin gene locus during in vitro differentiation of murine embryonic stem (ES) cells. In these experiments we utilized the ES/OP9 cell in vitro differentiation system described by Kitayima et al . (42). The ability of these cells to generate mice was not examined so their pluripotency was not directly confirmed, howeve r they did express markers of early development such as Rex-1 and did not express any of the globin ge nes (Fig.3-1). Furthermore, we were able to generate cells of both the hematopoietic and nervous systems (data not shown). Total RNA was isolated from ES/STO and ES/OP9 cultures at the indicated points following the start of induction and treated with DNase-1 to remove genomic DNA. Reversetranscription polymerase chain reaction (RT-PCR) was used to examine the developmental progression of cell samples and primer sets span introns in all cases except for the Rex-1 gene. The appearance of transcripts was visualized by polyacrylamide gel electrophor esis and staining with SY BR green. Day 0 cells are


40 composed of ES and STO cells grown in ES media containing LIF. These cells express the Rex-1 and -actin genes but do not expres s any of the genes of the -globin locus. Actin Rex/ globin majglobinDay 0Day 8Day 12 Day 5Day 10majglobin Rex1 / globin -Actin majglobin Rex1 / globin -Actin Actin Rex/ globin majglobin Actin Rex/ globin majglobin Figure 3-1. PCR analysis of DNase I treated and reverse-transcribed total RNA extracted from differentiating embryonic stem cells at the indicated time points. All primer sets span introns with the exception of Rex-1 and size of each RTPCR product is as follows; Rex-1, ~600bp; -actin, 480 bp; globin, 400 bp; maj, 220 bp. None of the samples showed genomic DNA amplification (not shown). Upon differentiation the embr yonicand adult-specific -globin genes are activated sequentially. The gene is activated first with tr anscripts appearing as early as day 5 of the time course (Fig.3-1). Expression of the adult specific gene is first observed at low levels at day 8 and is then up regulated upon the initiation of definitive erythropoiesis (Days 10-12). Expression of Re x-1 is reduced, although still detectable;


41 this is most likely due to re sidual undifferentiated cells present in culture rather than genomic DNA contamination. HS2 maj-globinInput unspec. Pol II MeH3Lys4 AcH4 Input unspec. Pol II MeH3Lys4 AcH4 Day 0Day 12 Rex-1 HS2 maj-globinInput unspec. Pol II MeH3Lys4 AcH4 Input unspec. Pol II MeH3Lys4 AcH4 Day 0Day 12 Rex-1 Figure 3-2. Interaction of transcription fact ors and RNA polymerase II with the -globin locus. Differentiating ES cells were incubated in formaldehyde and the crosslinked chromatin was fragmente d, isolated, and precipitated with antibodies specific for with chicken anti-IgG (unspec.), RNA polymerase II (Pol II), di-methylated histone H3 lysine 4 (MeH3Lys4), and acetylated histone H4 (AcH4). DNA purified from the precipitate was analyzed by PCR with primers corresponding to regions in the murine -globin locus as indicated. This is supported by the lack of genomic bands for the other primer sets. We analyzed the interaction of Pol II and the appearance of modified histones w ithin the globin locus during the course of differentiation using th e ChIP assay (Fig.3-2). We used antibodies specific for Pol II, acetylated histone H4 (AcH4) and for di-methylated lysine 4 of histone H3 (Me2K4H3). Di-methylation of H3 at ly sine 4 is associated with regions


42 permissive for transcription (110). The Pol II-specific antibody detected both phosphorylated and un-phosphorylated forms of the protein. Each antibody was used in three independent experiments. The results show that Pol II and Me2K4H3 are present at the LCR but not at the globin gene in undifferentiated ES cells (day 0) suggesting that Pol II recruitment to the LCR occurs before activation of any of the globin genes (Fig.3-2 and 3-3). The presence of H3 di-methylated at K4 indicates that these elements are permissible to active transcription; an observation supported by the detection of transcri ption through HS2 and HS3 cores (data not shown). This lysine methyla tion is specific to the core regions of the HS sites and this mark is not detected in a region between the HS2 and 3 cores (3/2Flank) and no transcription within this regi on is detected (not shown). The -globin gene is associated with acetylated histone H4 but not H3 dimethyl-K4 suggesting that the chromatin structure has been modified but is not transcriptionally ac tive. The Rex-1 gene is associated with a chromatin structure charact eristic of an open, transcriptionally active domain. We examined both the promoter a nd the transcribed region of the Rex-1 gene (not shown). In differentiated erythroid cell samples containing both mature primitive and definitive precursors (Day 12) Pol II is associated with the LCR and the -globin gene (Fig.3-2 and 3-4). The enrichment of H3 di-m ethylated at lysine 4 in these regions is compatible with an open, transc riptionally active chromatin st ructure. In contrast, the Rex-1 gene is repressed at this stage and this is accompanied by a reduction in K4 dimethylated H3 and acetylated H4 as well as a decrease in Pol II recruitment at the promoter. These results demonstrate that pr ior to erythroid differentiation, the LCR lies


43 within a transcriptionally competent chro matin structure and is associated with components of the basal transcription machinery (Pol II). Day 00 0.05 0.1 0.15 0.2 0.25 Pol IIMeH3-K4AcH4 AntibodiesFraction Inpu t HS2 Bmajor necdin Figure 3-3. Quantitative analysis of Pol II bind ing and histone modifications within the globin locus at “Day 0” in murine ES cells. Values were calculated using standard curve titration of input samples. We next employed real-time PCR analysis to obtain a quantit ative measure of factor binding and histone modifications during differentiation (F igs.3-3 through 3-5). We hoped to show that the locus underwent ch anges in factor recruitment and chromatin structure as the genes became active. Quantitative analysis of these samples and comparison of the values obtained throughout th e time course to those obtained for the neuronal specific necdin gene show that Pol II recruitment increased as much as five-fold at the globin gene promoters and approximately 20-fold at the core regions of the LCR (Fig.3-5). The concentration of histone modi fications associated with transcriptional activation at the promoter regions increased from two to five-fold. Distinct differences between the changes at the embryonic and a dult stage-specific promoters were observed. The significance of this obser vation is not known but may reflect different mechanisms involved in the regulation of these genes. At the embryonic H1 gene promoter there was


44 a detectable increase in met hylation of histone H3 lysine 4 but not in acetylated H4. The exact opposite was observed at the major gene promoter. At HS2 both di-methylation of histone H3 lysine 4 and acetylation of H4 increased during differentiation, although the increase in histone H4 acetylation was more dramatic. Although not shown we also examined the association of these factors at time points that should contain pre-hematopoietic cells (day 5) and definitive hematopoietic precursors (day 10). We also chose to analyze on ly the adherent cells in these cultures in contrast to the previous e xperiments where both adherent and floating fractions were obtained. The major motivation for this was th e fact that a day 5 very few cells were actually floating freely in the media. We wish ed to use equal number s of cells in these assays. Using the same ES/OP9 system Suwabe et al . found that cells isolated from the floating fraction contained a grea ter abundance of mRNAs for the and major–globin genes compared to that of the adherent fr action (111). The adherent fraction of these cultures contained approximately 10 times mo re erythroidand granulocyte-monocyte colony-forming units than the floating fracti on (111). We hypothesize d that the adherent fraction may contain cells whose chroma tin structure was indicative of a less differentiated cell type than that of thos e floating freely, a conclusion supported by the presence of more precursor cells in these samp les. In day 5 cells we observed low levels of transcription of the embryonic -globin gene and little or no detectable transcription of the adult specific major-globin gene. By day 10 both genes were active. Consistent with previous experiments RNA Pol II was pres ent at the LCR in significant quantities, however very little was observed at the gene promoters and levels did not appear to increase as before. Similar results we re obtained for MeH3K4 and AcH4 antibody


45 fractions. Most importantly we were unable to observe the formation of developmental stage –specific subdomains. Our results agreed with the analysis performed by Suwabe and colleagues using a similar ES/OP9 cells culture system. They showed that the adherent cells present in ES/OP9 cultures we re enriched for precursor cells and that the structure of the -globin locus was similar to that f ound in precursor cells isolated from other systems (105). Further analysis of fact or recruitment and histone modifications at these points is ongoing. Day 120 0.1 0.2 0.3 0.4 0.5 0.6 pol-IImeH3-K4AcH4 AntibodiesFraction Inpu t HS2 Bmajor necdin Figure 3-4. Quantitative analysis of Pol II bind ing and histone modifications within the globin locus at “Day12” in differentaited murine ES cells. Values were calculated using standard curv e titration of input samples Discussion The commitment of pluripotent stem cells to successively less plastic progenitors and finally, differentiated cells exhibiting stable expression pa tterns is thought to involve the reorganization of the chromatin envir onment of many lineage-specific genes. The timing of these changes, in many cases, has been shown to precede gene transcription (103, 105, 107). In the present study we have asse ssed the temporal natu re and extent of


46 covalent histone modifications and association of transcri tion complexes at the murine globin locus during the in vitro differentiation of murine em bryonic stem cells. We have observed that elements of the -globin LCR are capable of recruiting RNA polymerase II and histone modifications comp atible with transcription pr ior to lineage sp ecification. We also observe transcrip tion of HS elements of the LCR in undifferentiated ES cells. These results suggest that the -globin locus may already exist, in part, in a transcriptionally active state very early during deve lopment. It appears that in th e context of this system it remains so in a number of pre-hematopoietic precursor cell populations and undergoes a number of alterations in chro matin structure and factor recruitment as these cells progress towards hematopoietic commitment. Quantitativ e analysis shows that recruitment of transcription complexes and hi stone modifications are presen t in greater abundance at the LCR compared to the gene promoters. This is consistent with the idea that the LCR may be activated in a number of hematopoietic a nd pre-hematopoietic cell types, whereas the activation of the genes is restrict ed to that of the erythroid li neage. Whether or not this is a requirement for the proper stage-specific activ ation of the genes is not known. It may be that these observations are a bi-product of the generally non-repressive chromatin environment of ES cells and that the locus l acks these marks in a number of cells that are still capable of expressing th e globin genes. It would be of interest to study the conformation of other gene loci that contain LCRs to examin e the possibility that a high degree of accessibility in multipotential precurs ors and ES cells is an intrinsic quality of these dominant regulatory elements.


47 Fold Change During Differentiation0 5 10 15 20 25 30 35 HS2IVR3BH1Bmajor3/2flank Amplified RegionFold Change POL II MeH3K4 AcH4 Figure 3-5. Fold change in factor binding and histone modifications at the murine globin locus during in vitro differentiation. Values are shown on the Y-axis as the fold difference compared to the values obtained for the necdin promoter. However, from this data we can construc t a very basic picture describing the two observed states of the locus in our studies (F ig. 3-6). In “Day 0” precursor cells the LCR already exhibits charac teristics of a transcritpionally active region. These include histone modifications and recruitment of transcrip tion complexes containing RNA Pol II. These activities may be recruited to the LCR so that they may be deposited at the gene promoter regions to activate transcripti on at the proper developmental stage. Transcription through the LCR core regions may facilitate these inte ractions. In differentia ted cells that express the globin genes these marks are present at both the LCR and the gene promoters. How these activities are transferred is not known. It may be throug h direct interaction of the LCR with gene proximal elements (looping) or by the movement of transcription complexes and associated chromatin modifi ers along the DNA (linking). Whatever the mechanism it is clear that the LCR plays an im portant role in the in itial reorganization of the locus prior to high-level expression of the globin genes in erythroid cells. Recently a


48 study by Szutorisz and others produced sim ilar observations for the B-cell specific VpreB1 and genes (112). They characterize a cis -acting element in this locus marked by H3 acetylation, H3 lysine 4 di-methylati on, and Pol II recruitment in ES cells and show that these marks occur independently of the recruitment of any lineage-specific transcription factors such as PU.1. Furthermor e, they observe the presence of components of the TFIID complex (TAF 10 and TBP) to th is element in ES cells. They label these marks collectively as the early transcripti on competence mark (ETCM) and substantiate its importance by making light of the fact that subsequent, similar modifications appear to spread outward in both directions to the genes it controls. This is id entical to the observed appearance of these marks at the LCR of th e globin locus in ES cells followed by the genes in our ES/OP9 cultures. These results call into question the notion of differences between the pluripotent and multipotent states . These results suggest that the progression from a progenitor cell to that of a terminally differentiated and f unctional cell cannot be explained by a simple cascade of gene re pression and activation, since many tissuespecific genes appear to be poised for activation prior to commitment.


49 Figure 3-6. Recruitment of transcription comple xes and chromatin structure alterations at the -globin locus in undifferentiated pr ecursors and defini tive erythrocytes. Transcription complexes recruited to the LCR in uncommitted progenitors could be delivered to gene promot ers directlt through looping of the intervening DNA or via a tracking mechanism. ? ? Da y 0 Day 12 LCR LCR g lobin RNA Pol II Differentiation


50 CHAPTER 4 COMPOSITE -GLOBIN LOCUS TRANSGENES EXHIBITING POSITIONINDEPENDENT EXPRESSION IN TRANSGENIC MICE Introduction Transgenic mice have proven to be an excellent system for the in vivo analysis of how the regulatory elements involved in ge ne expression function. Fr om such studies we have identified elements critical for the expr ession of a large number of genes including HOX genes as well as those of the growth hor mone and globin loci. As informative as such studies are the results gleaned from an alyzing the behavior of genes at ectopic locations within the genome is often confounded by the influence of the surrounding chromatin environment, a phenomena known as position effect variegation (PEV) (113). Position effects can often cause disparate leve ls of gene expression between independent lines harboring identical transgenes, an importa nt fact when consider ed in the context of gene therapy. In order to circumvent th e difficulties associated with chromatin environment many researchers have sought to identify dominant regula tory elements able to protect genes from position effects in tran sgenic assays. One class of elements capable of such a feat is locus co ntrol regions (LCRs). LCRs ar e often composite elements containing multiple core regions that exhib it heightened sensitivity to nucleases in specific cell types (114-116). Th ese HS sites can be clustere d or spread throughout a gene locus (115, 117). Exactly how LCRs func tion is not entire ly known, although many studies have shown that the LCR HS sites of the human -globin gene locus synergize to confer high-level and position-independent expression (77, 118-121). Results from


51 genetic as well as conformational studie s suggest that the HS sites of the -globin LCR interact with each other and w ith the genes they activate at a particular developmental stage (45, 77, 78, 122). These interactions could establish a configuratio n that protects the genes from negative effects exerted by nei ghboring chromatin (48). Another group of regulatory DNA elements that protect genes from position of integration effects are the so-called boundary elements. These elements were first discovered as sequences that protect transgenes from pos ition effect variegation in drosophila (113). Similar elements were discovered in higher eukaryotic cells . For example matrix attachment regions (MARs) flanking the chicken lysozyme gene locus protect the gene from position effects in transgenic mice (123). Likewise, elements flanking the chicken -globin gene locus, cHS4, are able to protect re porter genes in transgenic assays (96, 124). Chicken HS4 harbors two distinguishable ac tivities; it blocks the functi on of enhancers on activating promoters, and it establishes a boundary be tween open and closed chromatin (124). While enhancer blocking activity of cHS4 is mediated by CTCF (125), proteins mediating the boundary function remain to be determined, although USF proteins have recently been shown to be involved through binding the element and recruitment of histone modifying activities ( 126). Results also suggest th at some insulators may be tethered to nuclear compartments, e.g. nuclear pore complex or nucleolus, via interactions with proteins known to reside in these structures (127, 128). A promising strategy for avoiding position of integration effects is to direct the transgene into a specific site in the genome . Several strategies have been applied to directing transgenes into sp ecific genomic sites (100, 129, 130). In this study we utilized components of the adeno associated virus ( AAV) integration machinery with the goal to


52 direct a human -globin expression construc t into a specific site in the mouse genome. AAV is a small human virus that is able to establish latent infec tion by integrating its DNA into a specific site on human chromosome 19, called the AAVS1 site (131, 132). Integration is mediated by DNA sequences pres ent in the viral genom e as well as by the AAV encoded rep protein, which contains AT Pase, helicase, and DNA-nicking activities (133-135). We generated transgenic mice cont aining the human AAVS1 integration site. The-globin expression construc t contained sequences from AAV that have previously been shown to be critical or to enhance integration into the AAVS1 site (136). This construct was incubated with recombinant AAV rep protein and the mixture injected into the pronuclei of fertilized muri ne oocytes transgenic for th e AAVS1 site. Despite using different conditions and rep/ DNA molar ratios none of the transgenic mice had the transgene integrated into the human AAVS1 site. During the course of this study we have analyzed integration sites and expression levels of two different -globin gene constructs. The fi rst construct contained LCR core elements HS2 and HS3, the -globin gene, and the -globin gene 3’enhancer, flanked by insulator elements from the chicken -globin gene locus (cHS4). The second construct differed from the first one in that we includ ed the HS2/HS3 flanking sequences. Both of these constructs expressed the -globin gene from different positions in the murine genome. However, the construct containing the HS2/HS3 flanking sequence consistently revealed higher -globin expression levels, even when integrated in or close to a centromere. This suggests that the HS2/3 flanking region facilit ates the activation by LCR core elements.


53 Results We began our studies by generating and analyzing transgenic mice harboring an expression construct of the human -globin gene that is small enough to be packaged into recombinant AAV (rAAV). This construct cont ained the core sequenc es of LCR HS sites 2 and 3, the human -globin gene, and the -globin 3’enhancer. These elements were flanked on either site with a single copy of an insulator sequence derived from the chicken -globin gene locus (cHS4). The rationa le for the inclusion of the regulatory elements was the following: HS2 has been shown to have strong enhancer activity when linked to globin or other re porter gene constructs (137) and HS3 has been shown to harbor both enhancer and chromatin opening ac tivity (138). We did not include human LCR element HS4 because results from previous studies suggest that it does not contribute unique activities for -globin gene activation (77, 119, 139). Chicken HS4 (cHS4) exhibits both enhancer blocking as we ll as boundary activitie s (124). Finally the -globin 3’enhancer has been show n to be important for high level -globin gene expression in the context of the complete locus in -globin yeast artificial chromosome (-globin YAC) transgenic mice (140). We hypothesized that HS2 and HS3 would open the chromatin regardless of the transgene inte gration site and that the presence of cHS4 would protect the -globin gene expression construct fr om any negative effect exerted by surrounding chromatin at the site of integration. We have generated and analyzed four transgenic lines with this construct. The copy number of the transgene was determined by southern blotting experime nts in which a single copy -globin YAC transgene was used as a standard and the murine snrp N gene as an internal cont rol (data not shown). The single copy line cont ained the entire human -globin locus in th e context of a YAC (141). -globin gene expression in the tran sgenic lines was analyzed by semi-


54 quantitative RT-PCR using pairs of primers specific for human and murine -globin cDNAs (77). Murine -globin expression levels were us ed as an internal control and globin gene expression (/-globin) was calculated as per cent expression of that of the single copy -globin YAC transgenic lin e (set at 100%). Figure 4-1. Structure and expression of the 4324 transgene. A) Structure of the transgene showing cHS4 in sulators, hypersensitive site cores 3 and 2 (HS3, 2) from the human -globin locus and a -globin gene/3’enhancer. B) PCR analysis of cDNA reverse transcribe d from RNA isolated from anemic spleens of transgenic mice. Human -globin and endogenous mouse -globin are shown (shown are the signals fo r 14, 16, 18, and 20 PCR cycles). C) Summary of expression leve ls and copy numbers in th e individual transgenic lines shown in Panel B. RT-P CR signals were quantitated by phosphorimaging. Expression levels were calculated based on expression of the mouse -globin gene and presented as % with expression levels in the human -globin YAC transgenic line set as 100%. -globin gene expression is s hown as expression per copy or to tal. The data show that expression per copy is low in these transgenic mice, demonstrating that the combination of regulatory elements present in the 4324 expression construct is not sufficient to confer high-level -globin gene expression (Fig.4-1). The fact that all of the lines do 100 1 WT 4 25 3 20 Expression/ copy (% of WT)Copy Number Line 100 1 432 4A Expression/ copy (% of WT)Copy Number Line 100 100 60 Expression(% of WT) 100 Expression total (% of WT) 432 4B 432 4C 432 4D 15<1<15 12448Ch m WT 432 4A 432 4B 432 4C 432 4D B -globin -globin cHS4 globin gene/3'enhancer HS2 cHS4 HS3 -A


55 express the -globin gene indicates that the cons truct is likely pr otected from position effects due to the presence of the cHS4 insulator elements. Because the inclusion of additional regulat ory elements would render the transgene too large for packaging into AAV, we decided to pursue an alternative strategy that, if successful, would allow us to in tegrate the transgene into a defined site into the mouse genome and which would not be limited by th e size of the DNA construct. Wild-type AAV integrates its DNA into a specific si te on human chromosome 19, called AAVS1 (131). Although the mechanism of integration is not entirely clear, the process requires cis -acting elements present in the viral genome as well as the function of the AAV encoded rep protein. The AAVS1 sequence wa s ligated as a 3.6 kb fragment into the EcoRI/KpnI restriction sites of the vector pB S246 leaving a single loxp site at the 3’ end of the AAVS1 site. The presence of a single loxP site would allow us to eventually reduce the copy-number by Cre mediated r ecombination (142). The construct was linearized with EcoRI and used to genera te transgenic mice. We generated two transgenic lines, however, only one of th ese lines continued to transmit the AAVS1 sequence. Southern blotting experiments showed that this line contained the AAVS1 plasmid integrated in 5 tandem copies (dat a not shown). We ma pped the position of the transgene by inverse PCR a nd DNA FISH analysis. The AAVS1 transgene integrated within the intron of a gene pr edicted to encode a metallopr otease and located near the telomere of chromosome 15 (Fig. 4-2). We next generated a larger -globin expression construct th at is similar to the one described in Fig.1 but contains in addition the flanking region of the HS2 and HS3 core enhancers (43f24, Fig. 3A). Previous work has s hown that inclusion of DNA flanking


56 the core regions allows the LCR HS sites to synergistically activate globin gene expression (120). The final construct is a bout 11 kb in size and contains DNA sequence elements derived from the AAV genome, the 5’ inverted terminal re peat (ITR) and the p5 promoter region. Both of these sequences ha ve been implicated in the integration of wild-type AAV into the S1 site (136). These elements were placed outside of the 5’cHS4 sequence. The supercoiled plasmid DNA was incubated on ice with recombinant rep68 and the mixture injected into the pronuclei of fertilized oocytes homozygous for the AAVS1 integration site. The offspring was fi rst analyzed by PCR for the presence of the -globin transgene. We generated f our transgenic lines with the second -globin gene Figure 4-2. Location of the human S1 integrat ion site in the mouse genome. A)Inverse PCR was used to obtain sequence inform ation from the site of transgene integration. This sequence was blaste d against the mouse genomic database ( A perfect match to the sequence is located close to the telomeric end of chromosome 15. B) C onfirmation of the integration site by DNA FISH. Metaphase spreads of spleen cells from transgenic mice were hybridized using a fluorescent probe speci fic for the human AAVS1 site (red) and a chromosome paint specific for chromosome 15 (green). B C 100 1 WTA 7 13 43f2 4C 40 7 43f2 4B 10 30 43f2 4A Expression/ copy (% of WT) Copy Number Line 100 97 280 300 Expression total (% of WT) 43f2 4A h 43f2 4C h flanking region HS3HS2 cHS4cHS4 5’ITR/P5 -globin gene flanking region HS3HS2 cHS4cHS4 -globin gene/3’enhancerA A Red = AAVS1 probe Green = Chr15 paintB AAV S1


57 construct (43f24); the first line did not transmit the tr ansgene. The three remaining lines were bred with AAVS1 transgenic mice to generate mice heterozygous for the -globin expression construct. Determination of copy number and integrity as well as expression analysis was performed as described for the first construct. We also analyzed the integration patterns of these mice using DNA FISH. The expression analysis in these mice is summarized in Fig. 4-3C. All three lines expressed the -globin gene at higher levels than those harboring the smaller expre ssion construct, supporti ng previous findings that the flanking sequences allow the HS core sites to synergistically activate -globin gene expression (120). Figure 4-3. Expression analysis of a -globin construct containi ng insulator sequences as well as LCR elements HS2 and 3 with their flanking DNA in transgenic mice. A) Structure of the 43f2 4 plasmid. B) Integration pattern of 43f2 4 transgenic lines (43f2 4 A and C) analyzed by DNA FISH. C) Summary of human globin gene expression and copy nu mber in transgenic mice harboring the 43f2 4 construct (43f2 4 A to C). Analysis performed as described in Fig.4-1. We next used DNA FISH to determine the s ite of integration of these constructs within the mouse genome. The re sults demonstrate that the human -globin gene 100 1 WTA 7 13 43f2 4C 40 7 43f2 4B 10 30 43f2 4A Expression/ copy (% of WT) Copy Number Line 100 97 280 300 Expression total (% of WT) 43f2 4A h 43f2 4C h B C flanking region HS3HS2 cHS4cHS4 5’ITR/P5 -globin gene flanking region HS3HS2 cHS4cHS4 -globin gene/3’enhancerA


58 construct did not integrate in to the transgenic human AA VS1 site on chromosome 15 (Fig.4-3 and 4). Recently, a mouse orthol og of the human AAV S1 site has been identified. This sequence is located in the peri-centromeric region of chromosome 7. Because the 43f24B transgene integrated into or close to a murine centromere we examined whether the murine ortholog of th e human AAVS1 site has been targeted. However, southern blotting experiments and DNA FISH with a chromosome 7 paint revealed that it did not integrate into the mouse S1 site (data not shown). Analysis showed that average expression of the -globin gene in the context of 43f24 construct was roughly four fold hi gher per copy than that of the 432 4 construct, and total expression was five fold higher in 43f24 transgenic lines (Fig.4-5). The line 43f24B expressed the human -globin gene at 40% of le vels found in a single copy globin YAC transgenic mouse. This was the hi ghest level of expression per copy in all of the transgenic lines analyzed in this study. Th e FISH analysis reveal ed that the transgene integrated into a region located at the centromeric end of a mouse chromosome as indicated by the intense DAPI staining (Fig.4-4). The expres sion analysis as well as the mapping of DNase I hypersensitive sites dem onstrates that the transgene is open and active in this location.


59 A DNaseI HS2 p HS EcoRI HS3 EcoRI HS2 43f2 4B -YAC Cycle 141618 h -globin m -globin -YAC -YAC M 43f2 4B 43f2 4BC43f2 4B h B Figure 4-4. Analysis of the integration site, -globin gene expression, and DNase I HS sites in transgenic line 43f2 4B. A) Analysis of the 43f2 4B integration site by DNA FISH. Metaphase spreads of spleen cells from transgenic mice were hybridized to fluorescent probes specific for the plasmid 43f2 4 (red) and chromosome 15 (green), stained with DAPI and visualized by fluorescence microscopy. The insert is a magnifica tion of the region containing the 43f2 4 integration site. B) Analysis of DNase I HS sites in the 43f2 4 transgene. Spleen nuclei of anemic transgenic mice were digested with increasing concentrations of DNase I. The genomic DNA, isolated from these samples, was digested with EcoRI, size-fra ctionated by gel-electrophoresis, and transferred to a nylon membrane. The DNA was hybridized to a radioactive probe corresponding to a region just 5’ of the downstream EcoRI site as indicated. The arrows indicate the pos ition of HS sites associated with LCR element HS2 and the -globin promoter. Lane M re presents a 1 Kb ladder. C) Analysis of -globin gene expression by RT-PCR in 43f2 4B and bglobin YAC (WT) transgenic mice. Expression of the human and mouse globin genes was analyzed as describe d in Figure 4-1 Panel B. Shown are the signals for 14, 16, and 18 PCR cycles. Discussion Hemoglobinopathies are among the most comm on inherited diseases in the human population (143). Due to problems associated with current treatments, alternative therapies are highly sought after. Ideally one would like to use gene therapy to deliver corrected copies of the mutated gene into ce lls of the hematopoietic system. For example,


60 investigators have successfully used len tiviral vectors to deliver therapeutic -globin expressing constructs into hematopoietic cel ls of mice carrying mutations in the globin locus (121, 144). However, the use of lentiv iral vectors may not be without problems because the viral DNA is integrated more or less randomly into the genome (145). Adeno-associated virus is considered to be safe because it does not cause strong immune reactions and DNA of recombinant viruses us ed in gene therapy experiments often remain episomal (146), although some data suggest that AAV serotype 2 integrates preferentially into transcriptionally active re gions in the nucleus (147). The disadvantage of AAV is its low packaging capacity. This is particularly problematic in terms of the globin locus as it appears to require a comp lex set of regulatory elements to achieve physiologically relevant levels of gene expres sion. We wished to a void this drawback by exploiting elements of the AAV system invol ved in site-specific integration without actually packaging the DNA into a virus. Despite numerous attempts, using a number of different conditions, this strategy proved unsuccessful. The components used in ou r studies have been shown sufficient for AAV integration into the S1 site (136). Howe ver since the normal means of delivery is through viral infection critical steps may be bypassed in our injection procedure. The chromatin environment surrounding the integration site of our S1 transgene may also play a role as it appears to be integrated with in a telomeric region of chromosome 15. This however, is unlikely as transgenes are cap able of integrating randomly throughout the genome with no apparent bias for chromatin structure. Further i nvestigation into the conditions required for rep-medi ated integration into our transgenic AAV S1 site is ongoing.


61 Figure 4-5. Relative expression levels of two -globin transgenic constructs. Average expression levels of 432 and 432f transgenic constructs as compared to that of a single copy F4.3 YAC containing and intact human -globin locus. The total expression and e xpression per copy is shown. Our initial experiments were aimed at inve stigating the ability of a construct that was capable of being packaged into th e AAV virus to exhibit position-independent expression at physiologically rele vant levels. We included what we considered to be the minimal genetic elements from the globin locus in combin ation with a known boundary element in these constructs. The data show that inclusion of these elements is not sufficient to confer high-level -globin gene expressi on. The levels of -globin gene transcription observed could be due to the fact that the regulatory elements fail to independently establish an active chromatin domain permissible for transcription. The second construct containing the HS2 and HS3 flanking DNA led to higher expression levels in all the transgenes analyzed supporting previous conclusions that the cores function better in the presence of flanking regi ons (120). This is in line with the LCR holocomplex model according to which the HS sites interact with each other to activate globin gene expression (68, 77) . It may be that the flanking sequences serve an 0% 50% 100% 150% 200% 250% 432B443F2B4F4.3WTYAC Per copy Total


62 architectural role allowing the HS cores to interact with one another. Recent conformational studies suggest th at the LCR HS sites are in close proximity in erythroid cells and that the genes that are expressed at a specific developmental stage contact the LCR holocomplex (48). Formation of the LCR holocomplex and its association with other HS sites in the globin locus, a comp lex termed the “chromatin hub”, could establish an architecture that is resist ant to negative influences exer ted by neighboring chromatin at the site of transgene in tegration. In this respect, it is pos sible that the larg er construct is able to establish such an architecture or micro-domain, whereas the smaller one, due to the absence of the HS core flanking DNA, is unable to do so. Alternatively, the HS 2/3 core flanking region may harbor additional regulatory elements that cooperate with the core HS sites to enhance globin gene tran scription a notion suppor ted by the fact that evolutionarily conserved seque nce elements exist within in these regions (74, 120). There is also the possibility that the AAV ITR/p5 promoter may influence transgene expression although this is most likely not the case as it lies outside the cHS4 elements that have been shown to be highly effici ent in enhancer blocking (96). The highest level of -globin gene expression was obser ved in a transgenic line in which the construct integrat ed into a centromeric regi on. Centromeres are usually incompatible with transcription, although it has been reported that genes can be expressed within functional centromeres (148). We do not yet know whether the human -globin construct integrated into he terochromatic centromeric repe ats, but the DNA FISH result shows that the transgene is located at the very tip of the chromosome in a region of intense DAPI staining. The data thus demonstrate that HS2 and HS3 in combination with insulator sequences are able to c onfer high-level expression to the -globin gene in a


63 centromeric location. It shoul d also be noted that this mouse line carries the lowest number of transgenic copies among all the tr ansgenes analyzed here. Copy-number itself can influence the expression levels of transg enes. Previous work has shown that a high number of copies can repress transgene expr ession, possibly due to the fact that highly repetitive DNA tends to fold into a heterochromatic structure (142). In summary, our data demonstrate that the activity of LCR HS sites is enhanced in the presence of their flanking DNA in transgen ic mice. The combination of specific LCR HS sites and boundary elements can provide high-level expression to a -globin transgene even when integrated in a centromeric location.


64 CHAPTER 5 SITE-SPECIFIC INTEGRATION OF MODIFIED YEAST ARTIFICIAL CHROMOSOMES IN MOUSE EMBRYONIC STEM CELLS BY RECOMBINASE MEDIATE CASSETTE EXCHANGE Introduction The best way of examining the function of genetic regulatory elements is in the context of an intact organism. For the study of mammalian gene regulation the transgenic mouse has proven to be an extrem ely useful animal model due to its rapid proliferation and development as well as extensive genomic sequence information. However, the generation of transgenic mice or any transgenic organism is not without its pitfalls. Position of integration effects (PEV) first observed in drosophila often complicate transgenic studies concerning the roles of regulatory elements by altering the context in which they are observed. Positi on effects describe the influence of the chromatin environment around the position that a transgene integrates; a transcriptionally permissive environment can have drastically different effects on transgene expression than one that is by nature repressive, such as peri-centromeric regions. Furthermore, the orientation of integration may impact upon transgene expression. Feng and colleagues demonstrated that position of integration effects can, in some instances be highly dependent upon the orientation of the construc t within the integrat ed locus, with one orientation being repressive and the other permissive to expression (149). The random manner in which transgenes integrate into th e genome when introduced into fertilized oocytes often precludes comparative assessm ent of transgene behavior. Moreover, in many cases one would like to study the beha vior of gene and co rresponding regulatory


65 elements in a specific context, say that of constitutive heterochromatin, or juxtaposed to an inactive gene to test for promiscuous interactions. Homologous gene targeting strategies (HR) are commonly applied in order to introduce mutations or genes of interest into defined locations, however HR is highly in efficient often occurring at frequencies as low as 1:1000 when compared to that of random insertion (150). Random insertion can result in the introduction of multiple copies of a transgene, an event with a number of known consequences on gene expression, such as that of repeat induced silencing and copy-number dependent expression patterns ( 142). In order to circumvent these difficulties many investigators have employed th e use of site-specific recombinases or integrases to introduce tran sgenes into defined genomic locations (151, 152). Enzymatically catalyzed integration by site-spe cific recombinases is more efficient than random integration and allows for the introduc tion of a single copy of the transgene. There are several, well defined systems used for this purpose the most widely studied being that of the Cre (causes recombination) isolated from the bacteriophage P1, and the S. cerevisiae Flp (flippase) recombinase. Both act in a sequence-specific fashion recognizing short target sequences a nd share a common mechanism of DNA recombination that involves strand cleavage, exchange and ligation (153). Both the Cre and Flp systems use two 13 bp inverted repeat sequences recognized by the enzyme that are separated by an 8 bp spacer sequence, lox P sites for Cre, frt sites for Flp. In the presence of two target sites recombinas e monomers bound to the inverted repeats promote DNA synaptic complex formation and recombination between the two sites (152). Because it is the length of the spacer and not its sequence that is critical for recombination, substitution mutations within th is region will not affect the ability of the


66 site to act as a substrate fo r the enzyme (154). Sequence muta tions will affect the ability of the site to efficiently recombine with a wi ld type site, such that sites with identical spacer sequences will recombine much more r eadily with one another than a wild type site (155). It is this charac teristic of these systems that is the basis for recombinase mediated cassette exchange or RMCE. Use of these enzymes in mammalian cells initially involved the integration of a single lox P or FRT site followed by trapping of rare integration events. This approach was hindered by the fact that they are inefficient, the persistence of the enzyme often results in th e re-excision of the “f loxed” or “flrted” allele, which is always more efficient in th is context than the integration. Secondly, the entire plasmid integrates and often leaves a positive selection marker behind. Sequences of prokaryotic origin or co -expressed genes can severely perturb the expression of neighboring genes as well as the gene of in terest (156). By employing variant target sites that recombine inefficiently or not at all it is possible to introduce selected sequences site-specifically and st ably into the genome. The first step in the process is to integrate a selectable marker flanked by hete rotypic target sites into the genome. These recognition site-containing cassettes can be site-specifically targeted via homologous recombination to any location in the genome, allowing for the examination of chromatin environment prior to the integration of the gene of interest. Once characterized these target cassettes can act as substrates for the exchange reaction i nvolving a circularized exchange cassette containing the gene of intere st flanked by the same sites present in the genomically integrated targeting cassette. The exchange plasmid can then integrate via a two-step mechanism into the site, exchangi ng the original selectable marker for the desired transgenic sequences (Fig.5-1). Si nce only homotypic target sites and not


67 heterotypic ones are capable of efficient recombination, the integration event is stable. Furthermore, due to the nature of the inte gration subsequent targ etings to the same location can occur, allowing for a comparison of the behavior of a number of transgenes from the same location. Figure 5-1. Principle of RMCE. The system is based on the single copy integration of cassette 1 containing a positive selection marker flanked by heterotypic lox P sites. Cotransfection of a plasmid cont aining cassette 2 with a Cre expression plasmid can result in two outcomes, integration followed by excision or integration followed by resolution resulting in the stable integration of cassette 2 (adapted from ref. 149). We sought to exploit the Cre based system of RMCE to introduce human -globin locus transgenes into the inactive X chromo some in murine embryonic stem cells through a combination of homologous recombination and RMCE, and to subsequently generate transgenic mice from these clon es. The central aims of this study are to examine the role Gene L1L2 Neo L1L2 Gene L1L2 Neo L1L2 L1XL1 L1XL1 L2XL2 excision resolution Gene chromosome plasmid chromosome L1L2 Neo L1L2 chromosome Targeted integration of cassette 1


68 of the LCR in resisting the repressive ev ents that occur durin g the process of Xinactivation and secondly, to establish a system for reproducibly introducing large chromosomal domains, those in excess of 50 K b, into defined locati ons within the mouse genome. The first step in this process is to introduce the target cassette containing a neomycin or other selectable marker vi a homologous recombination into the X-linked HPRT locus in murine ES cells, creating a ta rget site for exchange cassettes. We would then use yeast artificial chromosome s (YACs) bearing an intact human -globin locus that we have modified to contain heterotypic loxP sites as well as a puromycin selectable marker to act as exchange cassettes for RMCE . Co-transfection of our circularized YACs with a Cre expression plasmid would allow for the stable integration of the globin transgenes into the targeted locus. By em ploying an elegant breed ing strategy involving the use of Searle’s mice in which X-inactiv ation is not random we could analyzed the chromatin structure and expression characteristic of littermates bearing the locus on either inactive or active X chromosome (Sear l’s mice harbor a T(X;16)16H translocation and it is the wild type X chromosome that is preferentially in-activated) (157). X chromosome inactivation is an exte nsively studied phenomenon whereby in female mammals a “random choice” mech anism decides which of the two X chromosomes will be inactivated. This inactivation results in the transcriptional silencing of most of the genes and the condensation of the chromosome into the heterochromatic Barr body. The purpose of this process is to ensu re equal levels of expression of X-linked genes in XX females and XY males. In eutherian mammals this is accomplished by inactivating one of the female X chromosome s early in embryogenesis (158). Studies have revealed that a specific region of th e X-chromosome, the X inactivation center (Xic)


69 is required in cis for an X chromosome to be inactivated (159). The Xic is a complex element believed to be involved not only in counting how many, but choosing which X chromosome will be inactivated (160). Furtherm ore, a cell must contain at least two Xics for inactivation to occur. F ound within the Xic is the Xist transcript, a non-coding RNA that coats the chromosome in cis and triggers its silencing. A second critical element to this process is the Xce or X controlling element found within the Xic. It is believed that the choice of which chromosome to be inactivated occurs at this element. This assumption is based on the discovery that diffe rences in Xce alleles can result in skewed patterns of X-inactivation (161). Once the choice is made the to-be inactivated chromosome undergoes a number of changes to make it distinct from that of the active one. These include the up-regulation of the Xist transcript located in the Xic, replication at later point in S-phase (162, 163), differen tial methylation of selected CpG islands (164), and characteristic changes in histone modification patterns, including hypoacetylation of histones H3 and H4, th e appearance of histone modifications associated with gene silenci ng including di-methylation of H3 lys-9 and tri-methylation of H3 lys-27 (165). It is this unique and well defined process of gene silencing and heterochromatization that led us to choose th is environment to test the ability of the human -globin locus control region (LCR) to act as a dominant regulatory element capable of resisting the repressive nature of the inactive X chromosome. As defined previously the LCR is capable of conferring position-independent expression on cis linked genes. However, whether or not this is the case for all locations is still a matter of contention. Similar studies examining the abil ity of another LCR containing locus, the


70 lysozyme locus to escape X-inactivation associ ated silencing revealed that the elements included in the domain were unable to protect from gene silencing (166). We believe intrinsic differences in -globin and lysozyme locus structure and regulation may result in a different outcome for globin locus transgen es subjected to the same conditions. We wish to examine the chromatin structure and expression characteristics of the globin genes in the pr esence or absence of the LCR by creating two constructs in which either the entire locu s (“floxed locus”) or th e genes alone (“floxed genes”) are contained within th e boundaries of hetero typic loxP sites. From this we hope to discern the activities provi ded by the LCR versus those f ound intrinsically within the gene proximal regulatory elements. If the intact locus does indeed remain active when integrated into the Xi we would also like to study active domain formation and spreading during the developmental process using an in vitro differentiation system. Lastly, if successful these studies will be the first exam ple of a system capable of site specificintegration of large chromosomal domains (> 100Kb) into defined pos itions within the mouse genome. Results and Discussion As stated, the most importan t aspect of site-specific RM CE is to introduce a target cassette into the desired location within the genome. For this purpos e we have generated two DNA constructs for targeting the murine Hp rt gene locus (Fig.5-2A and B). In the first construct we have embedded a neomycin resistance gene and two heterospecific lox P sites within Hprt homologous sequences deri ved from a vector Del 19.2 acquired from the laboratory of Allan Bradly and shown to be capable of targeting the locus. These sequences were cloned into the pKO916 gene targeting vector (Str atagene). The second


71 construct is somewhat more complex and cont ains; in addition to the Hprt homologous sequences extracted from Del 19.2, neomycin resistance gene, and the two heterotypic loxP sites, an intact -globin LCR. The larger constr uct was generated by ligation of Hprt sequences flanking LCR homologous sequence elements derived from Figure 5-2. Structure of gene targeting ca ssettes for RMCE in ES cells. A) Small targeting construct cont aining only “floxed” neomycin gene and Hprt homologous sequences. B) Larger construct containing the human -globin LCR in addition to Hprt homologous sequences and “floxed” selectable marker. the 5’ and 3’ end of the LCR. The neomycin resistance gene and the two lox P sites were ligated 3’ to the LCR homologous sequences. The yeast shuttle vector pRS316 was used as the backbone of this constr uct as it is capable of being replicated in both bacteria and as an episome in yeast. It also contains the ura3 gene. The strain of S. cerevisiae used for these experiments is a heterotrophic mutant for components of lysine, tryptophan, and uracil metabolism. The lysine and trypt ophan mutations are complemented by the presence of these genes in the vector arms of the A2014.3 YAC that contains the human -globin locus. The YAC is maintained via growth in medium lacking lysine and tryptophan. Upon linearization of the pRS316 plasmid between selected regions of globin locus homology, the plasmid is introduced vi a electroporation into the A201F4.3 strain. 5’Hprt Neo 3’Hprt 5’Hprt LCR 3’Hprt Neo A BL2L1 L2L1


72 The transformed cells are then grown in medium lacking uracil as well lysine and tryptophan (lys-,trp-,uramedi a). The surviving colonies are those that have undergone the desired recombination event. This pro cess will be discussed further in the next section. Plasmids that have incorporated the entire LCR were shuttled into bacteria, amplified, purified, and used to transform ES cells. Human -globin locus cassette exchange cons tructs were made using homologous recombination in yeast to retrofit the YAC with heterotypic loxP sites and a puromycin selectable marker. Four recombination vectors were used to create two human -globin locus exchange casse ttes (Fig.5-3). Figure 5-3. YAC constructs for RMCE. F4.3 human -globin locus yeast artificial chromosomes were modified via homologous recombination to create exchange cassettes for the integration of domains spaning th e entire locus or the genes alone. OriS/Puro, a bacteria l origin of replication from the F factor and a puromycin selectable marker.One construct spans the entire locus (“floxed locus”) and the sec ond contains only the five -globin genes, lacking an LCR (“floxed genes”). The vectors were created by PCR generation of homologous sequences to the regions flanking the site to which the loxP mutation was directed. LoxP sites were generated from synthetic oligos (L2 sites) or derived from the pBS246 plasmid (L1 sites). Based on earlier studies by Bouha ssira et al. we generated ca ssettes in which the loxP G A LCR G A loxPSites (L2)loxPSites (L1)FloxedLocus FloxedGenesRsrII RsrII RsrII RsrII OriS/Puro


73 sites differed in a single G-A mutation within the spacer region and were termed L1 and L2 (100). The sequences are shown in Fig.5-4. L1:ATAACTTCGTATAATGTAT G CTATACGAAGTTAT L2: ATAACTTCGTATAATGTAT A CTATACGAAGTTAT Figure 5-4 LoxP sequences used in targeting and exchange cassettes. G-A mutation is underlined These sites recombine at low efficiencies, as little as 1-5% in the absence of selection (167), and under selective pressure the desired in tegrants can be found at levels of 100% (100). The four target vectors in cluded 5’LCRL2, 5’GenesL2, 3’GenesL1, and 3’Genes Puro. These vectors would allow us to construct to modi fied YACs capable of integrating via RMCE into defined loci w ithin the genome and exchange a neomycin selection marker for that of a puromycin marker, allowing us to enrich for populations that had undergone legitimate site-specific in tegration. Transformation is carried out by linearizing the vectors in the 5’ or 3’ homology and electropora tion into yeast that are in mid log-phase growth. Recombination between the 5’ homologies results in the integration of the vector as shown in Fig5-5. Similar to the prs316 shuttle vector the backbone of these constructs (pRS306) contai ns a ura3 gene allowing transformants to grow in lys-,trp-,uramedia. The surviving colonies are analyzed by PCR and at least two different southern blots (Fi g.5-6). Following confirmation of correct targeting selected clones are grown for 2 days in non-selectiv e media then on solid media containing 5fluor-orotic acid (5-FOA). This step is referre d to as the “pop-out” step, and results in either the excision of entire construct or only the pRS vector sequences due to the duplication of the homologous sequences upon in tegration. Excised sequences are lost due to the lack of an autonomous re plication sequence (ARS). 5-FOA is


74 Figure 5-5. Schematic representation of the st ructure of integrated plasmids used to generate YAC exchange cassettes. Rest riction sites and fragments generated in Southern analysis of transformant s are shown. Homologies are shown as horizontal bars labeled 5’ and 3’. Ura 3, uracil biosynthesis gene contained in plasmid backbone and used to select integrants. metabolized by the ura3 gene product into a toxic metabolite and se lects against those cells that have retained the vector after gr owth in non-selective media. This selects for those clones that have lost the plasmid sequences and allows for multiple mutations to be introduced. The unique RsrII site within the vectors will allow for removal of all “unfloxed” DNA as well as circularizati on of the YAC for replication in the DH10B strain of E. coli. To date we have inserted all of the required lox P sites, puromycin markers and shown the RsrII sites to be f unctional (Fig.5-7). However, we have yet to successfully shuttle them into bacteria. We believe that this is due to our attempts to use the bacterial origin of replication found in the pRS plasmids. The larg e (>50 kb) constructs are most likely too large for this origin. We are curren tly constructing a vector to incorperate an Ori S from the fertility factor into the cassette. This origin allows for the replication of much larger plasmids (300 Kb). 5’3’ 5’ 5’ 3’ 5’ Ura3 3’ 5’ 5 ’5 ’l o x P3 ’U r a 3Bam HI SpeI SpeI Bam HI 8.0Kb 9.4Kb


75 Figure 5-6. Southern analysis of yeast tran sformants. Clones were isolated following growth on selective media and DNA wa s extracted, purified, and digested. The restriction enzyme used and the e xpected fragment size are shown above each blot. Positive clones are shown with an asterisk. ES cells were transformed via electroporation using 25-30 g of target construct. Hprt targeted ES cells should be neomycin re sistant and insensitive to 6-thioguanine (due to disruption of the Hprt gene). We have isolated several such clones from transformations with the two constructs. Although the clones were positive for the transfected DNA, so far we were not successful in targeting the Hprt gene as indicated by southern and PCR analysis. However, as revealed by fluorescence in situ hybridization (FISH) several clones appear to have integrat ed into telomeric or centromeric regions of the genome (data not shown). We will use th ese clones to first establish the second step of targeting, which is the Cre-mediated si te-specific recombination of globin locus constructs into the heterospecific loxp site s. At the same time we will modify our targeting constructs so as to substitute the current segments of Hprt homology, which were derived from Balbc mice with those isol ated from a 129 strain line library. The Control #1* #2 * #3 #4 * #5 #6 #7 #8 #9 1Kb ladder #10 #11 #12 #13 #14 #15 * #16 #17 * #18 1Kb ladderBam HI digest -8.0 Kb 8.0kbSpeI digest -9.4Kb 1 Kb ladder #1 * #2 * #3 #4 * #5 control #15 * #17 * 9.4kb


76 reason being that our lack of success in targeting the desired locus may be due to the fact that the homology incorporated in to our target vector is non-i sogenic as our ES cells are of 129 strain origin. We intend to redesign th e targeting vector to include regions of homology to Hprt derived from a 129 library. RsrII(+)RsrII(-)FGFLWTFGFLWT53Kb 71Kb 150Kb Figure 5-7. Southern Analysis of human -globin locus exchange constructs. Isolated YAC DNA was digested with RsrII, run of a pulse field gel and analyzed by southern protocol using a probe specific for the -globin gene. The unmodified A201F4.3 human -globin locus YAC is approximately 150 Kb. The “floxed locus” and “floxed genes” cassettes should be approximately 71 and 52 Kb. following digestion. Digested and undigested samples are shown. Despite preliminary difficulties in our ta rgeting and exchange strategy we believe the subtle modifications suggested will prove successful. The use of RMCE for the sitespecific integration of larg e chromosomal domains has several merits. As stated, understanding the chromatin envi ronment at the site of in tegration is critical in understanding the behavior of any transgene. Being able to choose the context in which you can examine the impact of various regul atory elements on transcription has obvious advantages. The application of this technique to large YAC or BAC transgenes will allow for the inclusion of more distal elements involved in gene regulation without assembling them in a completely artificial constr uct. We propose integrating the human -globin


77 locus into the inactive X-chromosome, but on ce the target cassette has been properly recombined into the Hprt locus any properly modified genetic construct could be studied in this context. A recent publication by Ad ams and colleagues has shown that cassette exchange is possible in ES cells, thus e liminating one possible variable and supporting the continuation of the project (168).


78 CHAPTER6 SUMMATION OF RESULTS Functional Significance of the -globin Locus Control Region Since its discovery several decades ago much research and speculation concerning the role of the LCR in -globin locus gene regulation has been put forth. Even today a complete picture of its function has not fully been described. There is however a large body of work describing LCR function in both na tural and artificial contexts (143). At its endogenous position it appears that the LCR may only be required for high-level expression of the globin genes wh ereas; transgenic studies ha ve shown that globin genes lacking an intact LCR or critical combina tions of LCR elements were expressed at variable levels or not at a ll (76, 80). These observations and others like them led authors to conclude that perhaps the LCR contained two, dissociable activit ies; a tissue-specific enhancing activity and the ability to domina ntly open chromatin structure and ectopic sites within the genome. This over-simplif ied description of LCR function belies the vastly complex nature of events occurring at this el ement during erythroid ontogeny. During the establishment of the hematopoi etic system a number of complex and icompletely defined series of protein:DNA and protein:pr otein interaction occur and result in the exqusitely regulated series of transcriptional switches collectively known as hemoglobin gene switching. A second facet of th e LCR that has gained much attention is its ability to ensure expression of cislinked genes in a transgenic context. This ability has been attributed to a dominant chromatin opening activity present with elements of the LCR.


79 Although neither the LCR’s role in gene switching and the means by which it is able to remodel the structure of chromatin at ectopic sites has been fully elucidated we wished to dissect certain aspects of these phenomena in this work. Using an in vitro system that allowed us to observe some of the earliest events in globin locus activation we examined gene expression and chromatin structure changes in uncommitted progenitors and definitive eryt hrocytes. Secondly, using transgenic mice we attempted to define the minimal genetic elements of -globin locus LCRs from various vertebrate species required to establish independently regulated chromosomal domains and hence, position-independent patterns of gene expression. Lastly, we attempted to use components of the Adeno Associated Virus integration machinery to deliver human globin transgenes in a site-specific fashion. In Vitro Differentiation of Murine Embryonic Stem Cells The ES/OP9 in vitro differentiation system provided an excellent means of studying the initial events occurring at the globin locus prior to the onset of gene expression. Our goal was to test the hypothesis that during the developmental process the progressive activation of the locus would first be imitiated by chromatin structure changes and factor recruitment at the L CR. Our results have shown that histone modifications such as acetyla tion of histone H4 and di-methyl ation of lysine of histone H3 were present at the LCR preceding detectab le expression of the globin genes. We also observed recruitment of transcription comp lexes containing RNA polymerase II to the LCR but not to the promoters of the globin ge nes in these cells. At later time points these marks and factors were detectable at actively transcribed genes as well as that the core regions of the LCR HS sites.


80 Analysis of intermediate time points such as that corresponding to the onset of embryonic hematopoiesis were inconclusive as we were unable to detect the establishment of an embryonic-sp ecific chromatin domain as described in our model (46). Despite this, we were able to establish that the LCR appears to be the initial site of activation and recruitment of transcription-sp ecific activities to the locus. The means by which these activities are transferred to the gene promoters is not known and a number of models have been proposed (169). These results clearly challenge the role of the LCR as acting as a simple enhancer. We wish to expand upon data obtained in these studies in several ways. We first wished to examine other time points during the different iation experiments. Of particular interest is day 5 because this is when the embryonic bu t not adult genes appear to be active. We would like to determine if the region spa nning the embryonic genes exhibits a chromatin structure different from that the adult genes as suggested by Gribnau et al .(90). The inclusion of additional antibodies, such as TAF 10 would also be if interest. The possibility that transcription complexes recr uited to the LCR prior to gene expression posses different subunit compositions could sugge st a role for Pol II in remodeling of the LCR and holocomplex formation. Lastly, a FACS analysis to determine the fraction of cells in our assays that are actually he matopoietic and/or er ythroid could provide evidence that the LCR is active in a number of cell types. Site-Specific Integration and Positi on-Independent Expression of Human -globin Transgenes Many attempts have been made to ascertain the exact role of the various regulatory elements found within both gene proximal el ements and that of the more distal locus control region of the -globin locus. Many of these studies have been performed using


81 transgenic models, most notab ly that of the mouse. Confounding the results from these studies has been the influence of chromatin stru cture at the site of integration of many of these transgenes. These “position effects” can cause variable levels of gene expression between independent lines b earing the same transgenic construct. One method of avoiding this problem has been to include dominant regulatory sequences or boundary elements that can resist the influences of neighboring chromatin. A second strategy used to avoid position effects has been to direct transgenes to specific locations within the genome and a number of systems have been used for this purpose (100, 129, 130). We sought to create a human -globin containing transgene containing the minimal genetic elements from vertebrate globin gene loci that would ensure position-independent expression at therapeutically re levant levels as well as direct our gene, using the AAV virus, to a specific location within the m ouse genome using a mouse transgenic human AAVS1 site. This construct proved to exhi bit a position independe nt expression pattern as human -globin gene transcripts we re detectable in four, independent lines. However in all cases expression levels were low indicating that additional regulatory elements were required. We therefore included the sequences flanking the core regions of the HS2 and 3 normally found in the context if an int act locus. Inclusion of flanking sequence had previously been shown to enhance expre ssion of linked genes (121, 144). This rendered the construct too large for pack aging of our constr uct into AAV. We then modified our strategy and included elements from the AAV genome critical for integration (136). The larger construct was incubated with various molar concentrations of the Rep 68 protein and injected into fertilized S1 transgenic oocytes. Whereas, expressi on levels were higher we failed to site-specifically integrate the construct into our target site. We can only


82 speculate that either our DNA:protein ratios need adjustme nt our that some aspect of the viral cycle required for integration of the genome was bypassed by non-viral delivery of the DNA. Despite this, one of the lines genera ted was of particular interest. There were two reasons for this. The first being that is expressed at levels 40% of that of a singlecopy, intact wild type -globin YAC. Secondly, upon FISH analysis this transgene appeared to have integrated within or n ear the centromere of the chromosome. It is known that certain centromeres are transcript ionally permissible, but in general these regions are highly heterochromatic and transcriptionally repressive (148). Hypersensitive sites were also detectable in the promoter and in HS2. Currently we are focused and identifying the exact site of integration of this transgene and will characterize the chromatin structure within th e transgene as well as the regions flanking the site of integration. We also intend to continue to optimize the c onditions required for Repmediated, site-specific integr ation of globin constructs. The results of these studies support the hypothe sis that LCR acts as the primary site of recruitment for transcription and coactivator complexes during development. Furthermore, we have shown that synergistic interactions between HS sites of the LCR are critical to high-level e xpression of the globin genes.


83 LIST OF REFERENCES 1. Pauling, L., Itano, H. A. (1949) Science 110, 543-8. 2. Weatherall, D. J. (2004) Nature Genetics 5, 625-31. 3. Stamatoyannopoulos, G. N., A.W. Mj erus, P.W. Varmus, H. (1994) The Molecular Basis of Blood Diseases (W. B. Saunders, Philadelphia). 4. Olivieri, N. F., Brittenham, G. M., Mc Laren, C. E., Templeton, D. M., Cameron, R. G., McClelland, R. A., Burt, A. D. & Fleming, K. A. (1998) N Engl J Med 339, 417-23. 5. Blau, C. A., Constantoulakis, P., al-K hatti, A., Spadaccino, E., Goldwasser, E., Papayannopoulou, T. & Stamatoyannopoulos, G. (1993) Blood 81, 227-33. 6. Ellis, J. & Pannell, D. (2001) Clin. Genet. 59, 17-24. 7. Littau, V. C., Burdick, C. J., Allfrey, V. G. & Mirsky, S. A. (1965) Proc. Natl. Acad. Sci. U S A 54, 1204-12. 8. Hayes, J. J., Tullius, T. D. & Wolffe, A. P. (1990) Proc. Natl. Acad. Sci. U S A 87, 7405-9. 9. Strahl, B. D. & Allis, C. D. (2000) Nature 403, 41-5. 10. Jenuwein, T. & Allis, C. D. (2001) Science 293, 1074-80. 11. Turner, B. M. (2000) Bioessays 22, 836-45. 12. Roth, S. Y., Denu, J. M. & Allis, C. D. (2001) Annu. Rev. Biochem. 70, 81-120. 13. Hsu, J. Y., Sun, Z. W., Li, X., Reuben, M., Tatchell, K., Bishop, D. K., Grushcow, J. M., Brame, C. J., Caldwell, J. A., Hunt, D. F., Lin, R., Smith, M. M. & Allis, C. D. (2000) Cell 102, 279-91. 14. Sterner, D. E. & Berger, S. L. (2000) Microbiol. Mol. Biol. Rev. 64, 435-59. 15. Dhalluin, C., Carlson, J. E., Zeng, L., He, C., Aggarwal, A. K. & Zhou, M. M. (1999) Nature 399, 491-6. 16. Jacobson, R. H., Ladurner, A. G., King, D. S. & Tjian, R. (2000) Science 288, 1422-5.


84 17. Shi, Y., Lan, F., Matson, C., Mulligan, P., Whetstine, J. R., Cole, P. A. & Casero, R. A. (2004) Cell 119, 941-53. 18. Rea, S., Eisenhaber, F., O'Carroll, D., Strahl, B. D., Sun, Z. W., Schmid, M., Opravil, S., Mechtler, K., Ponting, C. P., Allis, C. D. & Jenuwein, T. (2000) Nature 406, 593-9. 19. Jenuwein, T., Laible, G., Dorn, R. & Reuter, G. (1998) Cell. Mol. Life Sci. 54, 8093. 20. Lachner, M., O'Carroll, D., Rea, S ., Mechtler, K. & Jenuwein, T. (2001) Nature 410, 116-20. 21. Bannister, A. J., Zegerman, P., Partridge, J. F., Miska, E. A., Thomas, J. O., Allshire, R. C. & Kouzarides, T. (2001) Nature 410, 120-4. 22. Roach, M. L. & McNeish, J. D. (2002) Methods Mol. Biol. 185, 1-16. 23. Loebel, D. A., Watson, C. M., De Young, R. A. & Tam, P. P. (2003) Dev. Biol. 264, 1-14. 24. Kinder, S. J., Tsang, T. E., Quinlan, G. A., Hadjantonakis, A. K., Nagy, A. & Tam, P. P. (1999) Development 126, 4691-701. 25. Keller, G., Lacaud, G. & Robertson, S. (1999) Exp. Hematol. 27, 777-87. 26. Silver, L. & Palis, J. (1997) Blood 89, 1154-64. 27. Wong, P. M., Chung, S. W., Chui, D. H. & Eaves, C. J. (1986) Proc. Natl. Acad. Sci. U S A 83, 3851-4. 28. Palis, J. & Segel, G. B. (1998) Blood Rev. 12, 106-14. 29. Okuda, T., van Deursen, J., Hiebert, S. W., Grosveld, G. & Downing, J. R. (1996) Cell 84, 321-30. 30. Wang, Q., Stacy, T., Binder, M., Marin-Padi lla, M., Sharpe, A. H. & Speck, N. A. (1996) Proc. Natl. Acad. Sci. U S A 93, 3444-9. 31. Cline, M. J. & Moore, M. A. (1972) Blood 39, 842-9. 32. Medvinsky, A. & Dzierzak, E. (1996) Cell 86, 897-906. 33. Muller, A. M., Medvinsky, A., Strouboulis, J., Grosveld, F. & Dzierzak, E. (1994) Immunity 1, 291-301. 34. Godin, I. & Cumano, A. (2002) Nat. Rev. Immunol. 2, 593-604. 35. Cumano, A., Dieterlen-Liev re, F. & Godin, I. (1996) Cell 86, 907-16.


85 36. Yoder, M. C., Hiatt, K., Dutt, P., Mukherj ee, P., Bodine, D. M. & Orlic, D. (1997) Immunity 7, 335-44. 37. Keller, G. M. (1995) Curr. Opin. Cell Biol. 7, 862-9. 38. Robertson, S., Kennedy, M. & Keller, G. (1999) Ann. N. Y. Acad. Sci. 872, 9-15; discussion 15-6. 39. Keller, G., Kennedy, M., Papayannopoulou, T. & Wiles, M. V. (1993) Mol. Cell. Biol. 13, 473-86. 40. Yoshida, H., Hayashi, S., Kunisada, T., Ogawa, M., Nishikawa, S., Okamura, H., Sudo, T. & Shultz, L. D. (1990) Nature 345, 442-4. 41. Kodama, H., Nose, M., Niida, S. & Nishikawa, S. (1994) Exp. Hematol. 22, 97984. 42. Kitajima, K., Tanaka, M., Zheng, J., Sakai-Ogawa, E. & Nakano, T. (2003) Methods Enzymol. 365, 72-83. 43. Kyba, M., Perlingeiro, R. C. & Daley, G. Q. (2002) Cell 109, 29-37. 44. Palstra, R. J., Tolhuis, B., Splinter, E., Nijmeijer, R., Grosveld, F. & de Laat, W. (2003) Nat. Genet. 35, 190-4. 45. Tolhuis, B., Palstra, R. J., Splinter, E., Grosveld, F. & de Laat, W. (2002) Mol. Cell 10, 1453-65. 46. Levings, P. P. & Bungert, J. (2002) Eur. J. Biochem. 269, 1589-99. 47. Forrester, W. C., Epner, E., Drisco ll, M. C., Enver, T., Brice, M., Papayannopoulou, T. & Groudine, M. (1990) Genes. Dev. 4, 1637-49. 48. de Laat, W. & Grosveld, F. (2003) Chromosome Res. 11, 447-59. 49. Hardison, R., Chao, K. M., Schwartz, S ., Stojanovic, N., Ganetsky, M. & Miller, W. (1994) Genomics 21, 344-53. 50. Bungert, J. & Engel, J. D. (1996) Ann. Med. 28, 47-55. 51. Orkin, S. H. (1995) J. Biol. Chem. 270, 4955-8. 52. Pevny, L., Simon, M. C., Robertson, E., Klein, W. H., Tsai, S. F., D'Agati, V., Orkin, S. H. & Costantini, F. (1991) Nature 349, 257-60. 53. Tsai, F. Y., Keller, G., Kuo, F. C., Weiss, M., Chen, J., Rosenblatt, M., Alt, F. W. & Orkin, S. H. (1994) Nature 371, 221-6.


86 54. Zon, L. I., Mather, C., Burgess, S., Bolce, M. E., Harland, R. M. & Orkin, S. H. (1991) Proc. Natl. Acad. Sci. U S A 88, 10642-6. 55. Weiss, M. J., Keller, G. & Orkin, S. H. (1994) Genes. Dev. 8, 1184-97. 56. Briegel, K., Lim, K. C., Plank, C., Be ug, H., Engel, J. D. & Zenke, M. (1993) Genes. Dev. 7, 1097-109. 57. Anguita, E., Hughes, J., Heyworth, C., Bl obel, G. A., Wood, W. G. & Higgs, D. R. (2004) Embo. J. 23, 2841-52. 58. Merika, M. & Orkin, S. H. (1995) Mol. Cell. Biol . 15, 2437-47. 59. Blobel, G. A., Nakajima, T., Eckner, R., Montminy, M. & Orkin, S. H. (1998) Proc. Natl. Acad. Sci. U S A 95, 2061-6. 60. Stamatoyannopoulos, J. A., Goodwin, A., Joyce, T. & Lowrey, C. H. (1995) Embo. J. 14, 106-16. 61. Kotkow, K. J. & Orkin, S. H. (1996) Proc. Natl. Acad. Sci. U S A 93, 3514-8. 62. Johnson, K. D., Christensen, H. M., Zhao, B. & Bresnick, E. H. (2001) Mol. Cell. 8, 465-71. 63. Engel, J. D. (1994) Nature 367, 516-7. 64. Igarashi, K., Kataoka, K., Itoh, K., Hayash i, N., Nishizawa, M. & Yamamoto, M. (1994) Nature 367, 568-72. 65. Shivdasani, R. A., Rosenblatt, M. F., Zucker-Franklin, D., Ja ckson, C. W., Hunt, P., Saris, C. J. & Orkin, S. H. (1995) Cell 81, 695-704. 66. Crossley, M., Tsang, A. P., Biek er, J. J. & Orkin, S. H. (1994) J. Biol. Chem. 269, 15440-4. 67. Perkins, A. C., Sharpe, A. H. & Orkin, S. H. (1995) Nature 375, 318-22. 68. Wijgerde, M., Grosveld, F. & Fraser, P. (1995) Nature 377, 209-13. 69. Donze, D., Townes, T. M. & Bieker, J. J. (1995) J. Biol. Chem. 270, 1955-9. 70. Mouthon, M. A., Bernard, O., Mitjavila, M. T., Romeo, P. H., Vainchenker, W. & Mathieu-Mahul, D. (1993) Blood 81, 647-55. 71. Bresnick, E. H. & Felsenfeld, G. (1993) J. Biol. Chem. 268, 18824-34. 72. Aplan, P. D., Nakahara, K., Orki n, S. H. & Kirsch, I. R. (1992) Embo. J. 11, 4073-81.


87 73. Lee, J. S., Lee, C. H. & Chung, J. H. (1999) Proc. Natl. Acad. Sci. U S A 96, 10051-5. 74. Hardison, R. (1998) J. Exp. Biol. 201, 1099-117. 75. Forrester, W. C., Takegawa, S., Papayannopoulou, T., Stamatoyannopoulos, G. & Groudine, M. (1987) Nucleic Acids Res. 15, 10159-77. 76. Grosveld, F., van Assendelft, G. B., Greaves, D. R. & Kollias, G. (1987) Cell 51, 975-85. 77. Bungert, J., Dave, U., Lim, K. C., Lieuw, K. H., Shavit, J. A., Liu, Q. & Engel, J. D. (1995) Genes Dev. 9, 3083-96. 78. Bungert, J., Tanimoto, K., Patel, S., Li u, Q., Fear, M. & Engel, J. D. (1999) Mol. Cell Biol. 19, 3062-72. 79. Peterson, K. R., Clegg, C. H., Navas, P. A., Norton, E. J., Kimbrough, T. G. & Stamatoyannopoulos, G. (1996) Proc. Natl. Acad. Sci. U S A 93, 6605-9. 80. Reik, A., Telling, A., Zitnik, G., Cimbora, D., Epner, E. & Groudine, M. (1998) Mol. Cell. Biol. 18, 5992-6000. 81. Epner, E., Reik, A., Cimbora, D., Telli ng, A., Bender, M. A., Fiering, S., Enver, T., Martin, D. I., Kennedy, M., Ke ller, G. & Groudine, M. (1998) Mol. Cell. 2, 447-55. 82. Bender, M. A., Bulger, M., Close, J. & Groudine, M. (2000) Mol. Cell. 5, 387-93. 83. Farrell, C. M., Grinberg, A., Huang, S. P ., Chen, D., Pichel, J. G., Westphal, H. & Felsenfeld, G. (2000) Proc. Natl. Acad. Sci. U S A 97, 14554-9. 84. Bulger, M. & Groudine, M. (1999) Genes Dev. 13, 2465-77. 85. Jimenez, G., Griffiths, S. D., Ford, A. M., Greaves, M. F. & Enver, T. (1992) Proc. Natl. Acad. Sci. U S A 89, 10618-22. 86. Orkin, S. H. (1995) Eur. J. Biochem. 231, 271-81. 87. Blobel, G. A. (2000) Blood 95, 745-55. 88. Schubeler, D., Francastel, C., Cimbora, D. M., Reik, A., Martin, D. I. & Groudine, M. (2000) Genes Dev. 14, 940-50. 89. Vieira, K. F., Levings, P. P., Hill, M. A., Crusselle, V. J., Kang, S. H., Engel, J. D. & Bungert, J. (2004) J. Biol. Chem. 279, 50350-7. 90. Gribnau, J., Diderich, K., Pruzina, S. , Calzolari, R. & Fraser, P. (2000) Mol. Cell. 5, 377-86.


88 91. Forsberg, E. C., Downs, K. M., Chri stensen, H. M., Im, H., Nuzzi, P. A. & Bresnick, E. H. (2000) Proc. Natl. Acad. Sci. U S A 97, 14494-9. 92. Leach, K. M., Nightingale, K., Igarashi, K., Levings, P. P., Engel, J. D., Becker, P. B. & Bungert, J. (2001) Mol. Cell Biol. 21, 2629-40. 93. Elefanty, A. G., Robb, L., Birn er, R. & Begley, C. G. (1997) Blood 90, 1435-47. 94. Leach, K. M., Vieira, K. F., Kang, S. H ., Aslanian, A., Teichmann, M., Roeder, R. G. & Bungert, J. (2003) Nucleic Acids Res. 31, 1292-301. 95. Gaensler, K. M., Burmeister, M., Browns tein, B. H., Taillon-Miller, P. & Myers, R. M. (1991) Genomics 10, 976-84. 96. Chung, J. H., Bell, A. C. & Felsenfeld, G. (1997) Proc. Natl. Acad. Sci. U S A 94, 575-80. 97. Hartl, D. L. & Ochman, H. (1996) Methods Mol. Biol. 58, 293-301. 98. Trask, B. J. (1991) Curr. Opin. Genet. Dev. 1, 82-7. 99. Kang, S. H., Vieira, K. & Bungert, J. (2002) Nucleic Acids. Res. 30, e44. 100. Bouhassira, E. E., Westerma n, K. & Leboulch, P. (1997) Blood 90, 3332-44. 101. Litt, M. D., Simpson, M., Gaszner, M ., Allis, C. D. & Felsenfeld, G. (2001) Science 293, 2453-5. 102. Hebbes, T. R., Clayton, A. L., Thor ne, A. W. & Crane-Robinson, C. (1994) Embo. J. 13, 1823-30. 103. Bottardi, S., Aumont, A., Grosveld, F. & Milot, E. (2003) Blood 102, 3989-97. 104. Delassus, S., Titley, I. & Enver, T. (1999) Blood 94, 1495-503. 105. Johnson, K. D., Grass, J. A., Park, C., Im , H., Choi, K. & Bres nick, E. H. (2003) Mol. Cell Biol. 23, 6484-93. 106. Kontaraki, J., Chen, H. H., Riggs, A. & Bonifer, C. (2000) Genes Dev. 14, 210622. 107. Tagoh, H., Himes, R., Clarke, D., Leen en, P. J., Riggs, A. D., Hume, D. & Bonifer, C. (2002) Genes. Dev. 16, 1721-37. 108. Ford, A. M., Bennett, C. A., Healy, L. E ., Towatari, M., Greaves, M. F. & Enver, T. (1996) Proc. Natl. Acad. Sci. U S A 93, 10838-43. 109. Lund, A. H. & van Lohuizen, M. (2004) Genes Dev. 18, 2315-35.


89 110. Strahl, B. D., Ohba, R., Cook, R. G. & Allis, C. D. (1999) Proc. Natl. Acad. Sci. U S A 96, 14967-72. 111. Suwabe, N., Takahashi, S., Nakano, T. & Yamamoto, M. (1998) Blood 92, 410818. 112. Szutorisz, H., Canzonetta, C., Georgiou, A., Chow, C. M., Tora, L. & Dillon, N. (2005) Mol. Cell. Biol. 25, 1804-20. 113. Kellum, R. & Schedl, P. (1991) Cell 64, 941-50. 114. Forrester, W. C., Novak, U., Ge linas, R. & Groudine, M. (1989) Proc. Natl. Acad. Sci. U S A 86, 5439-43. 115. Bonifer, C. (2000) Trends Genet. 16, 310-5. 116. Tuan, D., Solomon, W., Li, Q. & London, I. M. (1985) Proc. Natl. Acad. Sci. U S A 82, 6384-8. 117. Li, Q., Peterson, K. R., Fang, X. & Stamatoyannopoulos, G. (2002) Blood 100, 3077-86. 118. Li, G., Lim, K. C., Engel, J. D. & Bungert, J. (1998) Genes Cells 3, 415-29. 119. Milot, E., Fraser, P. & Grosveld, F. (1996) Trends Genet. 12, 123-6. 120. Molete, J. M., Petrykowska, H., Bouha ssira, E. E., Feng, Y. Q., Miller, W. & Hardison, R. C. (2001) Mol. Cell. Biol. 21, 2969-80. 121. May, C., Rivella, S., Callegari, J., Helle r, G., Gaensler, K. M., Luzzatto, L. & Sadelain, M. (2000) Nature 406, 82-6. 122. Carter, D., Chakalova, L., Osborne, C. S., Dai, Y. F. & Fraser, P. (2002) Nat. Genet. 32, 623-6. 123. Stief, A., Winter, D. M., Strat ling, W. H. & Sippel, A. E. (1989) Nature 341, 3435. 124. West, A. G., Gaszner, M. & Felsenfeld, G. (2002) Genes. Dev. 16, 271-88. 125. Bell, A. C., West, A. G. & Felsenfeld, G. (1999) Cell 98, 387-96. 126. West, A. G., Huang, S., Gaszner, M., Litt, M. D. & Felsenfeld, G. (2004) Mol. Cell 16, 453-63. 127. Ishii, K., Arib, G., Lin, C., Van H ouwe, G. & Laemmli, U. K. (2002) Cell 109, 551-62.

PAGE 100

90 128. Yusufzai, T. M. & Felsenfeld, G. (2004) Proc. Natl. Acad. Sci. U S A 101, 86204. 129. Porteus, M. H., Cathomen, T., Weitz man, M. D. & Baltimore, D. (2003) Mol. Cell. Biol. 23, 3558-65. 130. Belteki, G., Gertsenstein, M., Ow, D. W. & Nagy, A. (2003) Nat. Biotechnol. 21, 321-4. 131. Kotin, R. M., Linden, R. M. & Berns, K. I. (1992) Embo. J. 11, 5071-8. 132. Linden, R. M., Ward, P., Giraud, C., Winocour, E. & Berns, K. I. (1996) Proc. Natl. Acad. Sci. U S A 93, 11288-94. 133. Ryan, J. H., Zolotukhin, S. & Muzyczka, N. (1996) J. Virol. 70, 1542-53. 134. Weitzman, M. D., Kyostio, S. R., Kotin, R. M. & Owens, R. A. (1994) Proc. Natl. Acad. Sci. U S A 91, 5808-12. 135. Zhou, X., Zolotukhin, I., Im, D. S. & Muzyczka, N. (1999) J. Viro.l 73, 1580-90. 136. Philpott, N. J., Gomos, J., Berns, K. I. & Falck-Pedersen, E. (2002) Proc. Natl. Acad. Sci. U S A 99, 12381-5. 137. Ellis, J., Talbot, D., Dillon, N. & Grosveld, F. (1993) Embo. J . 12, 127-34. 138. Ellis, J., Tan-Un, K. C., Harper, A., Michalovich, D., Yannoutsos, N., Philipsen, S. & Grosveld, F. (1996) Embo. J. 15, 562-8. 139. Navas, P. A., Peterson, K. R., Li, Q., Sk arpidi, E., Rohde, A., Shaw, S. E., Clegg, C. H., Asano, H. & Stamatoyannopoulos, G. (1998) Mol. Cell. Biol. 18, 4188-96. 140. Liu, Q., Bungert, J. & Engel, J. D. (1997) Proc. Natl. Acad. Sci. U S A 94, 169-74. 141. Tanimoto, K., Liu, Q., Bungert , J. & Engel, J. D. (1999) Nature 398, 344-8. 142. Garrick, D., Fiering, S., Marti n, D. I. & Whitelaw, E. (1998) Nat. Genet. 18, 56-9. 143. Higgs, D. R., Thein, S.L. & Wood, W.G. (2001) The Thalassaemia Syndromes , 133-191. (Blackwell Science, Oxford) 144. Pawliuk, R., Westerman, K. A., Fabry, M. E., Payen, E., Tighe, R., Bouhassira, E. E., Acharya, S. A., Ellis, J., London, I. M., Eaves, C. J., Humphries, R. K., Beuzard, Y., Nagel, R. L. & Leboulch, P. (2001) Science 294, 2368-71. 145. Marshall, E. (2002) Science 298, 510-1. 146. Flotte, T. R. (2004) Gene. Ther. 11, 805-10.

PAGE 101

91 147. Nakai, H., Montini, E., Fuess, S., Stor m, T. A., Grompe, M. & Kay, M. A. (2003) Nat. Genet. 34, 297-302. 148. Saffery, R., Sumer, H., Hassan, S., Wong, L. H., Craig, J. M., Todokoro, K., Anderson, M., Stafford, A. & Choo, K. H. (2003) Mol. Cell. 12, 509-16. 149. Feng, Y. Q., Lorincz, M. C., Fiering, S., Greally, J. M. & Bouhassira, E. E. (2001) Mol. Cell. Biol. 21, 298-309. 150. Bode, J., Schlake, T., Iber, M., Schu beler, D., Seibler, J., Snezhkov, E. & Nikolaev, L. (2000) Biol. Chem. 381, 801-13. 151. Farley, F. W., Soriano, P., Steffe n, L. S. & Dymecki, S. M. (2000) Genesis 28, 106-10. 152. Stark, W. M., Boocock, M. R. & Sherratt, D. J. (1992) Trends Genet. 8, 432-9. 153. Sadowski, P. D. (1995) Prog. Nucleic. Acid. Res. Mol. Biol. 51, 53-91. 154. Hoess, R. H., Wierzbicki, A. & Abremski, K. (1986) Nucleic Acids Res. 14, 2287300. 155. Baer, A. & Bode, J. (2001) Curr. Opin. Biotechnol. 12, 473-80. 156. Branda, C. S. & Dymecki, S. M. (2004) Dev. Cell 6, 7-28. 157. Lyon, M. F., Searle, A. G., Ford, C. E. & Ohno, S. (1964) Cytogenetics 15, 30623. 158. Lyon, M. F. (1961) Nature 190, 372-3. 159. Rastan, S. & Brown, S. D. (1990) Genet. Res. 56, 99-106. 160. Heard, E. (2004) Curr. Opin. Cell Biol. 16, 247-55. 161. Cattanach, B. M. & Papworth, D. (1981) Genet. Res. 38, 57-70. 162. Tada, T. & Takagi, N. (1992) Mol. Reprod. Dev. 31, 20-7. 163. Keohane, A. M., O'Neill L, P., Belyaev, N. D., Lavender, J. S. & Turner, B. M. (1996) Dev. Biol. 180, 618-30. 164. Miller, O. J., Schnedl, W., Alle n, J. & Erlanger, B. F. (1974) Nature 251, 636-7. 165. Silva, J., Mak, W., Zvetkova, I., Appana h, R., Nesterova, T. B., Webster, Z., Peters, A. H., Jenuwein, T., Otte, A. P. & Brockdorff, N. (2003) Dev. Cell 4, 48195.

PAGE 102

92 166. Chong, S., Kontaraki, J., Bonifer, C. & Riggs, A. D. (2002) Mol. Cell. Biol. 22, 4667-76. 167. Feng, Y. Q., Seibler, J., Alami, R., Eisen, A., Westerman, K. A., Leboulch, P., Fiering, S. & Bouhassira, E. E. (1999) J. Mol. Biol. 292, 779-85. 168. Adams, L. D., Choi, L., Xian, H. Q., Ya ng, A., Sauer, B., Wei, L. & Gottlieb, D. I. (2003) Brain. Res. Mol. Brain Res. 110, 220-33. 169. Engel, J. D. & Tanimoto, K. (2000) Cell 100, 499-502.

PAGE 103

93 BIOGRAPHICAL SKETCH Padraic Levings is one of nine childr en and was born on Long Island to Thomas and Miriam Levings. He attended the Univer sity at Stony Brook wh ere he played three seasons of division IA lacrosse. He was also a member of the and Gold Key honor societies. He graduated cum laude in 1998 w ith a bachelorÂ’s degree in biochemistry.