REGULATORY MECHANISMS CONTROLLING PRIMORDIAL GERM CELL DIFFERENTIATION WITHIN TH E MURINE FETAL GONAD By DANIELLE MAATOUK A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004
ACKNOWLEDGMENTS The author would first like to thank her parents for a lifetime of encouragement and support. She would also like to thank her sisters for contantly reminding her that no one understands what she is doing. She would like to thank all of the members of the lab including Lori Kellam, Karen Johnstone, Edwin Peery, Jessica Walrath, Chris Futtner, Amanda Dubose, Lindsey Williams and Steve Filippelli for their helpful comments and for making the lab an enjoyable place for the past years. The author would especially like to thank her mentor James Resnick for giving her an exciting project to work on and for allowing her to make her own decisions, both good and bad. She would also like to thank him for his years of ridiculing and for making her graduate career so enjoyable. ii
TABLE OF CONTENTS page ACKNOWLEDGMENTS..................................................................................................ii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES..........................................................................................................vii ABSTRACT.......................................................................................................................ix CHAPTER 1 INTRODUCTION........................................................................................................1 Origin of the Germ Cell Lineage..................................................................................2 Specification of the Germ Cell Lineage.......................................................................4 The Extraembryonic Origin of Primordial Germ Cells................................................8 Primordial Germ Cell Migration..................................................................................9 Sex Determination......................................................................................................13 Primordial Germ Cell Differentiation.........................................................................15 2 MATERIALS AND METHODS...............................................................................20 Cell Culture.................................................................................................................20 Preparing Feeder Layers for Germ Cell Culture.................................................20 Germ Cell Isolation and Culture..........................................................................20 Teratocarcinoma Cell Culture.............................................................................21 Immunocytochemical Methods..................................................................................21 Alkaline Phosphatase Staining............................................................................21 GCNA1 Staining: Cell Culture............................................................................22 GCNA1 Staining: Whole Mount........................................................................22 BrdU Labeling of Cultured Germ Cells..............................................................23 Primordial Germ Cell Isolation..................................................................................23 Preparation of Genomic DNA for Bisulphite Conversion..........................................25 Bisulphite Sequencing of Genomic DNA...................................................................26 Bisulphite Conversion.........................................................................................26 Bisulphite Polymerase Chain Reaction...............................................................26 Purification of PCR Products..............................................................................28 Cloning and Sequencing of PCR Products..........................................................28 Reverse Transcription Polymerase Chain Reaction...................................................29 iii
Isolation of Genomic Loci..........................................................................................29 Probe Preparation................................................................................................29 Screening of the Research Genetics Bacterial Artificial Clone Library..............30 Screening the Roswell Park Bacterial Artificial Chromosome Library..............31 Verification of BAC Clones................................................................................32 Southern Blotting.................................................................................................33 Generation of Genomic Subclones.............................................................................33 Generation of Subgenomic Libraries...................................................................34 Direct Subcloning into Bluescript.......................................................................40 Sequencing..................................................................................................................41 Plasmid Sequencing.............................................................................................41 BAC Sequencing.................................................................................................41 3 CELL AUTONOMOUS REGULATION OF PRIMORDIAL GERM CELL DIFFERENTIATION.................................................................................................44 Introduction.................................................................................................................44 In Vitro Culture of Primordial Germ Cells..........................................................44 Inductive Versus Autonomous Differentiation...................................................45 Results.........................................................................................................................48 Conclusions.................................................................................................................50 4 DNA DEMETHYLATION IS RATE LIMITING FOR PRIMORDIAL GERM CELL DIFFERENTIATION......................................................................................57 Introduction.................................................................................................................57 Chromatin Modifications and the Genome.........................................................57 DNA Methylation and Germ Cell Differentiation...............................................58 Results.........................................................................................................................59 Conclusions.................................................................................................................63 5 EXPRESSION OF POSTMIGRATORY GERM CELL GENES CORRELATES WITH LOSS OF PROMOTER METHYLATION....................................................71 Introduction.................................................................................................................71 DNA Methylation and Embryonic Development................................................71 DNA Methylation and Germ Cell Development.................................................72 Genomic imprinting.....................................................................................72 X chromosome reactivation.........................................................................73 Epigenetic Changes During Germ Cell Differentiation......................................74 Results.........................................................................................................................75 Conclusions.................................................................................................................78 iv
6 DNA METHYLATION REGULATES POSTMIGRATORY GERM CELL GENE EXPRESSION................................................................................................90 Introduction.................................................................................................................90 DNA Methylation and Control of Gene Expression...........................................90 DNA Methylation and Regulation of Postmigratory Gene Expression..............91 Results.........................................................................................................................92 Conclusions.................................................................................................................93 7 DISCUSSION.............................................................................................................97 Understanding the Cell Intrinsic Timing Mechanism................................................97 DNA Methylation of Primordial Germ Cells.............................................................99 DNA Demethylation.................................................................................................104 Understanding the Differentiation of Primordial Germ Cells..................................107 APPENDIX PRIMER SEQUENCES............................................................................108 LIST OF REFERENCES.................................................................................................113 BIOGRAPHICAL SKETCH...........................................................................................121 v
LIST OF TABLES Table page 3-1 Premigratory PGCs cultured in the presence of TPA..............................................53 A-1 Bisulfite Primers.....................................................................................................108 A-2 Sequencing primers for Mouse vasa homologue...................................................109 A-3 Sequencing primers for Synaptonemal complex protein 3....................................110 A-4 Sequencing primers for Deleted in azoospermia-like............................................111 A-5 Reverse transcription polymerase chain reaction primers......................................112 vi
LIST OF FIGURES Figure page 1-1 Whole mount embryo stained for alkaline phosphatase(TNAP).............................18 1-2 Primordial germ cell migration in a 9.5 dpc embryo...............................................18 1-3 Model for primordial germ cell development..........................................................19 1-4 Sex determination of the embryonic gonads............................................................19 2-1 Identification of primordial germ cells in culture....................................................43 2-2 Identification of proliferating germ cells in culture.................................................43 3-1 A cell intrinsic timing mechanism regulates PGC differentiation...........................54 3-2 Effects of altered proliferation on PGC differentiation............................................54 3-3 TPA inhibits PGC proliferation................................................................................55 3-4 Differentiation of 8.5 dpc PGCs in the presence of retinoic acid............................55 3-5 PGC differentiation in the presence of retinoic acid................................................56 4-1 Chemical structures of cytidine analogs...................................................................66 4-2 Differentiation of 8.5dpc PGCs in the presence of 5-azacytidine............................66 4-3 Differentiation of 10.5dpc PGCs in the presence of 5-azacytidine..........................67 4-4 Differentiation of purified 11.5dpc PGCs in the presence of 5-azacytidine............67 4-5 Differentiation of purified PGCs in the presence of trichostatin A..........................68 4-6 GCNA1 expression in F9 teratocarcinoma cells......................................................69 4-7 Postmigratory gene expession in 5-azacytidine treated F9 teratocarcinoma cells...69 4-8 Model for Primordial Germ Cell Differentiation.....................................................70 5-1 Methylation dynamics during embryogenesis..........................................................83 vii
5-2 Germ cell specific genes contain CpG islands.........................................................84 5-3 Bisulfite Analysis of mouse vasa homologue (mvh)...............................................85 5-4 Bisulfite analysis of synaptonemal complex protein 3 (scp3)..................................86 5-5 Bisulfite analysis of Deleted in Azoospermia-Like (dazl).......................................87 5-6 Bisulfite analysis of tissue non-specific alkaline phosphatase (Tnap).....................88 5-7 Methylation of the germ cell lineage........................................................................89 6-1 GCNA1 expression in a Dnmt1 mutant embryo......................................................96 viii
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy REGULATORY MECHANISMS CONTROLLING PRIMORDIAL GERM CELL DIFFERENTIATION WITHIN THE MURINE FETAL GONAD By Danielle Maatouk August 2004 Chair: James Resnick Major Department: Molecular Genetics and Microbiology Murine primordial germ cells (PGCs) can first be detected at 7.5 days post coitus (dpc) in the extraembryonic mesoderm. At 8.5dpc they leave their extraembryonic location and begin migrating through the hindgut until they reach the developing genital ridges between 10.5-11.5dpc. During this time period, both XX and XY PGCs are identical and have not undergone their separate sexual differentiation pathways. Our lab is interested in the common differentiation events undertaken by these sexually indifferent germ cells during the time in which they migrate and begin entering the genital ridge. These differentiation events include cessation of migration, cessation of proliferation, loss of the ability to form stem cells, erasure of genomic imprints, and changes in the expression levels of several genes. We have been studying these differentiation events by investigating the regulation of PGC-specific genes which are upregulated during this time period. Previously our lab has found that cultured premigratory 8.5dpc PGCs would differentiate to express the postmigratory germ cell ix
marker GCNA1on a similar temporal schedule as seen in vivo. Here we present evidence that this timing mechanism occurs independently of cell proliferation, and may be accelerated by DNA demethylation. These results indicate that GCNA1 induction requires neither the act of migration nor exposure to the gonad, and that induction of this marker may be controlled by a cell intrinsic program sensitive to DNA demethylation. To further investigate the role of DNA methylation in the differentiation of PGCs we have analyzed the methylation status of several PGC-specific genes during this time period. We have found that several genes, which are upregulated as germ cells enter the gonad contain CpG islands which are highly methylated at 10.5dpc and undergo a wave of demethylation by 13.5dpc. Our results indicate that upregulation of several PGC-specific genes necessary for further germ cell maturation may be attributed to loss of methylation in their promoter regions. Furthermore, embryos lacking a functional DNA demethylase enzyme show ectopic expression of a PGC specific marker gene suggesting that methylation may be the primary silencing mechanism for postmigratory gene expression in premigratory PGCs as well as somatic cells of the embryo. These results support a model in which a wave of demethylation controls the differentiation of primordial germ cells after entry into the fetal gonad. x
CHAPTER 1 INTRODUCTION Primordial germ cells (PGCs) are the initial founder population of cells which will give rise to the mature gametes, sperm and egg. Although PGCs are nullipotent, they differentiate to form two very different and highly specialized cells. When united, they form a single totipotent cell, giving rise to an entire organism and providing a mechanism for passage of genetic material from one generation to the next. Primordial germ cells were first identified in mouse embryos by Chiquoine, where by virtue of their high expression of the enzyme alkaline phosphatase, the PGCs could be clearly distinguished from the surrounding somatic cells of the embryo (Figure 1-1) (Chiquoine 1954). PGCs can first be detected 7.25 days post coitus (dpc) at the base of the allantois in the extraembryonic mesoderm. By 8.5 dpc the PGCs begin migrating into the embryo and, by 9.5 dpc, can be seen as a stream of cells extending from the base of the allantois throughout the length of the hindgut endoderm (Figure 1-2). The PGCs continue migrating through the hindgut endoderm, passing through the dorsal mesentery, and will begin colonizing the developing genital ridges between 10.5 and 11.5 dpc. During migration the PGC population expands from a population of 45 cells on 7.25 dpc, to a population of 25,000 cells on day 13.5 dpc when the cells cease their proliferation (Tam and Snow 1981). Sexual differentiation of the germ line begins at 13.5 dpc, with females entering meiosis and males entering a mitotic arrest. Prior to this point the two sexes are essentially indistinguishable. 1
2 In this section, aspects of PGC development will be discussed in detail answering several questions. Where does the germ cell lineage originate and what signaling, if any, is involved in this process? What are the factors controlling migration, and what controls the post-migratory events of germ cell development? Although none of these questions have been answered completely, the following sections will summarize the work that has contributed to our most recent understanding of the germ cell development in the mouse embryo. Origin of the Germ Cell Lineage Over 100 years ago, Weismann proposed the theory of a continuous germ line that could be followed throughout the life of an organism. It was observed in many species of insects and nematodes that the egg cytoplasm contained a region described as being rich in mitochondria and electron dense structures. Upon fertilization, the egg would divide asymmetrically such that those cells inheriting this specialized cytoplasmic material gave rise to the germ cell lineage, while those that did not gave rise only to somatic cells. Weismann suggested that this specialized cytoplasm, or germ plasm, contained the necessary components for germ line determination and was proposed to be the mechanism of germ line specification for all organisms (Nieuwkoop and Sutasurya 1979). It was not until the 1940s that Nieuwkoop's studies on amphibians would challenge Weismann's theories. Although it had long been known that the eggs of frogs possessed a germ plasm, Nieuwkoop observed that the salamander showed such characteristics only during later stages of embryonic development. He therefore concluded that in this species of amphibian, there is no predetermined germ line and later showed that inductive mechanisms are responsible for germ line determination in salamanders. Since
3 vertebrates share a common amphibian ancestor, it is possible that either mechanism could be present in mammals (Nieuwkoop and Sutasurya 1979). Species which have a predetermined germ lineage all share a characteristic germ plasm. In the mouse, a similar cytoplasmic structure termed the nuage, was observed in both rat and mouse migrating germ cells, but no such structure could be identified in earlier stages of development (Clark and Eddy 1975; Eddy and Clark 1975). Also, studies analyzing the developmental potential of individual blastomeres from a 4or 8cell mouse embryo found that all cells at both of these stages are totipotent, and no single cell solely gave rise to germ cells (Kelly 1977). The first definitive evidence for the origin of the mammalian germ line came from fate mapping experiments by Lawson and Hage. By injecting a fluorescent lineage marker into mouse embryos at 6.0 and 6.5 dpc, followed by about 40 hours of culturing the embryos in vitro, the descendents of each injected cell could be analyzed. It was found that the only cells capable of giving rise to germ cells at 6.0-6.5 dpc were those located in the most proximal region of the embryo, adjacent to the extraembryonic ectoderm. These proximal cells also contributed to extraembryonic mesoderm and cells of the allantois. Therefore the germ line precursors were restricted to this proximal population of cells, but no single cell was found that only gave rise to germ cells, indicating that germ cell determination had not occurred by 6.5 dpc (Lawson and Hage 1994). The final piece of evidence which ruled out any possibility for a predetermined germ line came from the work of Tam and Zhou. Previous fate mapping work had shown that cells of the most distal portion of the embryo would give rise to neural ectoderm,
4 while those at the most proximal portion would give rise to cells of the extraembryonic mesoderm including primordial germ cells. To answer the question of whether germ cells at this time are already determined to contribute to the germ cell lineage, or if their fate is still uncertain, small pieces of embryo were transferred from the distal portion of a donor embryo to the proximal portion of a recipient embryo. After two days of culture it was observed that cells transferred from proximal to distal did not give rise to germ cells. Of the cells transferred from distal, a population normally destined to become neuroectoderm, to proximal a small number were shown to give rise to PGCs (Tam and Zhou 1996). This was definitive proof that the germ cell lineage was not predetermined, but that inductive signals in the proximal region of the embryo were necessary to specify the germ line. Specification of the Germ Cell Lineage It is now well accepted that the mechanism of PGC determination is not predetermined, but instead is influenced by the surrounding somatic tissues. The question of what these influences are has yet to be fully answered, but recent data has revealed some of the molecular mechanisms involved in the process. The first signaling molecule to be uncovered was Bone Morphogenic Protein 4 (Bmp4). First identified to be involved in bone development, it is now known that this protein is involved in many developmental processes. In embryos lacking a functional Bmp4 gene, development does not proceed past mid-gastrulation. It was observed that the embryos completely lacked an allantois, the precursors of which are located in the same region as the PGC precursors. When the embryos were next subjected to alkaline phosphatase staining it was revealed the homozygous null embryos were also completely devoid of PGCs. Heterozygous embryos displayed a dose-dependent phenotype with reduced numbers of
5 PGCs. This was the first mutation found to disrupt germ cell specification (Lawson et al. 1999). From the fate mapping and transplantation experiments discussed above, it was clear that PGC precursors were located in the proximal region of the epiblast just below the extraembryonic tissues. Bmp4 expression begins around 5.5 dpc, and increases by 6.0 dpc in the extraembryonic ectoderm, adjacent to the location of the PGC precursors. It was further shown by explant culture, that isolated epiblasts separated from their adjoining extraembryonic tissues could not give rise to PGCs if separated at 5.5 dpc, but could give rise to PGCs if separated at 6.0 dpc. If extraembryonic tissues were removed and the culture medium supplemented with Bmp4, PGC specification could be rescued. Taken together, these results suggest that Bmp4 signaling from the extraembryonic ectoderm prior to 6.0 dpc is required for PGC formation (Yoshimizu et al. 2001). Although Bmp4 plays a critical role in PGC specification, it was clear that other signals must be involved. The lack of both PGCs and allantois in the homozygous mutant indicates that another signal must act that sends the PGCs down a separate pathway from those cells of the allantois. Further studies on members of the Bmp family identified two other members which are required for normal PGC specification. Bmp8b is expressed exclusively in the extraembryonic ectoderm during the time of PGC specification. Similar to the Bmp4 phenotype, Bmp8b mutant mice have reduced PGC numbers and defects in allantois formation, with heterozygote having no detectable defect compared to wildtype (Ying et al. 2000). Bmp2, expressed in the embryonic endoderm, was also found to influence PGC specification. Homozygous mutants have reduced PGC numbers and defects in allantois formation as well (Ying and Zhao 2001). Although all
6 three of these proteins are clearly involved in PGC specification, the signal separating the PGC population from the precursor cells of the allantois remains unidentified. Determined to uncover the signals required for PGC specification, Surani's group took a technically difficult approach to identify molecular changes in gene expression in the early PGC population (Saitou et al. 2002). The region of the embryo containing the PGCs was dissected from a 7.25 dpc embryo. This fragment contained 250-300 cells, including allantoic cells, PGCs and other extraembryonic mesoderm cells. Individual cells were picked at random and cDNA libraries from these single cells were generated. cDNA libraries were separated into groups based on their gene expression profiles. Cells which were positive for Bmp4 and Hoxb1, a somatic mesodermal marker, were categorized as somatic cells. Those cells which expressed no Bmp4 or Hoxb1, but expressed high levels of the early germ cell marker Tnap were categorized as PGCs. Representative PGCs and somatic cells were screened for genes which were differentially expressed between these two populations. From this screen two genes were identified which were exclusively expressed by the PGC population, Stella/Pgc7 and Fragilis. Stella expression begins around 7.0 dpc, around the time the PGC population first emerges. Expression continues until 15.5 dpc in the male and 13.5 dpc in the female at which time the gene is downregulated. Stella is turned on once again postnatally in immature oocytes, and its expression continues after fertilization until it is downregulated during the blastocyst stage. Targeted deletion of the Stella locus revealed that stella is a maternal effect gene required for pre-implantation development, but no defect in PGC specification or development was found (Payer et al. 2003; Saitou et al. 2002).
7 Interestingly, Fragilis is expressed throughout the epiblast at 6.0 dpc, but becomes restricted to the most proximal region of the embryo between 6.25 and 6.5 dpc (Saitou et al. 2002). This is coincident with the timing of Bmp4 and Bmp8b expression in the adjacent extraembryonic ectoderm. To determine whether fragilis expression is induced by extraembryonic signals, distal embryonic fragments (which normally do not express fragilis at 6.0 dpc) were cultured adjacent to extraembryonic fragments. While distal epiblast alone never expressed fragilis in this culture system, co-culture with extraembryonic tissues did induce fragilis expression . To assess whether Bmp4 signaling was involved in this induction, fragilis expression was analyzed in the Bmp4 knockout. Similar to effects on the initial founder PGC population, heterozygous Bmp4 mutants expressed a low level of fragilis while homozygous mutants showed no detectable fragilis expression. While fragilis is clearly a downstream target of Bmp4 it is speculated that the effect is not direct. Fragilis, a member of the interferon (IFN) inducible transmembrane protein family, contains two highly conserved IFN response elements within its regulatory sequences and is therefore likely to respond to IFN. Nonetheless, fragilis represents the first gene identified which is expressed in the precursor population induced by Bmp4 signaling. In summary, the population of cells which will go on to become both allantois and PGCs is located in the most proximal region of the embryo adjacent to the extraembryonic ectoderm (Lawson et al. 1999). Bmp4 and Bmp8b expression in the extraembryonic ectoderm and Bmp2 in the embryonic endoderm act on this precursor population of cells between 5.5 dpc and 6.0 dpc (Lawson et al. 1999; Ying et al. 2000; Ying and Zhao 2001). These inductive signals determine the initial precursors of the
8 allantois and germ cell lineage. The expression of Bmp4 leads to the indirect upregulation of fragilis in these precursor cells (Saitou et al. 2002). This population of cells moves in a posterior direction, through the primitive streak, and into the extraembryonic mesoderm (Lawson and Hage 1994). By 7.0 to 7.25 dpc Stella, Tnap and Oct4 are expressed exclusively in the PGC population, being the first genes to distinguish the allantoic and germ cell populations (Chiquoine 1954; Saitou et al. 2002; Scholer et al. 1989). The events between 6.0 and 7.0 dpc which signal the allantoic and germ cell precursors to assume separate developmental paths remain unknown. Other changes take place as the PGCs become specified and include repression of the region-specific homeobox genes Hoxa1, Hoxb1, Lim1 and Evx1, which remain on in the adjacent mesodermal somatic cells (Saitou et al. 2002). These changes in gene expression suggest that the PGC population is initially destined towards a somatic cell fate and re-attains a pluripotent state by both active downregulation of somatic genes and upregulation of genes which are associated with pluripotency. It has also been observed that expression of the housekeeping genes B-actin and Hprt are lower in PGCs, possibly reflecting changes in metabolic activity or the lengthening of their cell cycle time (Saitou et al. 2002). The events leading to PGC specification are summarized in Figure 1-3. The Extraembryonic Origin of Primordial Germ Cells Many aspects of germ cell development are highly conserved between different species. One of the most intriguing aspects of germ cell development is the location at which the germ line originates. Consistently forming outside of the embryo proper, the germ cell population must re-enter the embryo, and migrate to their final destination, the developing urogenital ridges. Although the pattern of extraembryonic location followed by migration is conserved, the initial location of the PGC population varies as does the
9 tissue they chose to migrate through. In mammals, PGCs originate in the extraembryonic mesoderm and migrate through the hindgut. Chicken PGCs originate in the crest cells around the anterior end of the embryo, between the primary ectoderm and primary endoderm. Subsequently they are incorporated into blood islands, circulate around the embryo with the blood stream, and exit in the region of the gonads by an unknown mechanism (Rogulska 1969). The reasoning behind the conserved extraembryonic origination of PGCs is not understood. The variety of the originating location between species has been rationalized by the theory that for germ cell specification to be conserved, it must be independent of developmental patterning events since organisms lacking three germ layers also separate their germ line from the embryo proper early in development (Wylie 1999). The reasoning for their separation from the rest of the embryo is more obscure. One possibility is that their extraembryonic location protects the germ cell population from signals triggering the differentiation of other epiblast cells towards somatic cell fates (Monk et al. 1987). Segregation from such signals ensures that the germ lineage can maintain its pluripotency. Although this is an attractive model, other theories suggest that the precursor germ cell population originally follows a somatic cell fate, and that this program is actively suppressed in cells which are induced to become germ cells (Saitou et al. 2002). Primordial Germ Cell Migration Migration in the mouse embryo has been studied by a combination of in vitro and in vivo techniques. PGCs which are classified as motile have a distinct morphology, being polarized or teardrop shaped, and extending filopodia several cell diameters in length. Non-motile PGCs are spherical and display no filopodia. Morphological
10 observations have been corroborated with in vivo time-lapse microscopy, using a mouse carrying a GFP transgene under control of the PGC-specific Oct4 promoter. PGCs can be explanted from different stages of embryos, plated in vitro on a feeder cell layer and migration observed in culture. Interestingly, analysis of motility in culture by time-lapse microscopy shows that the migratory behavior of PGCs in vitro is consistent with their migratory behavior in vivo (Donovan et al. 1986). The germ line is specified at 7.25 dpc in the extraembryonic mesoderm. By 8.0 dpc, the PGCs have moved into the endoderm of the developing hindgut. It was originally thought that the PGCs were carried into the embryo passively as the hindgut invaginated, although it has recently been shown that PGCs are motile at 7.5 dpc and 9.5 dpc. Because of the complex morphogenic movements of the hindgut during the intervening time, assessing the state of PGCs at 8.5 dpc was not possible. For migration into the hindgut, it seems that either mechanism, passive or active, is possible at this time (Anderson et al. 2000; Donovan et al. 1986). After entry into the hindgut, PGCs migrate as a stream towards the developing urogenital ridges. The first PGCs which begin their active migration through the hindgut do not need to travel very far. At this time the position of the hindgut is adjacent to the gonads and has not descended yet into the coelomic cavity, therefore the early migrating PGCs can migrate directly from the hindgut into the gonads. PGCs which migrate later must pass through the dorsal mesentery after leaving the hindgut before entering the urogenital ridges (Gomperts et al. 1994). Entry into the urogenital ridges begins around 10.5 dpc, with most PGCs arriving by 11.5 dpc. At this point the PGCs are no longer
11 motile, as both their morphology and behavior, in vivo and in vitro, changes (Anderson et al. 2000). The signals which control or guide migration are not fully understood. Several mouse mutants have been identified which interfere with germ cell migration. In particular, the ligand/receptor pair, c-kit and Steel Factor, were found to be important for both germ cell survival and migration during the migratory period. c-kit is expressed on PGCs during this time, while Steel is expressed in the surrounding somatic tissues. In mutants at either locus, establishment of the PGC population is normal, but by 8.0 dpc PGCs do not increase in number, nor do they migrate properly, forming clusters in the hindgut and some failing to leave the base of the allantois. Some PGCs do become localized along the length of the hindgut, although this may be due to passive migration as the hindgut invaginates. PGCs do not make it into the gonads in these mutants, and most die by apoptosis (Buehr and McLaren 1993; McCoshen and McCallion 1975; Mintz and Russell 1957). It is interesting to note that Steel and c-kit play an important role in the survival and migration of two other migratory cell populations, hematopoetic cells and neural crest derived melanocytes. Studies on germ cell migration in zebrafish embryos have recently uncovered another factor involved in germ cell migration. In a large-scale genetic screen for mutations that effect PGC development, one mutant was identified which caused defective migration. The mutant, odysseus (ody), was found to map to the gene Cxcr4b. In zebrafish, mutation of this gene does not affect the ability of germ cells to migrate, but impairs their ability to migrate towards the gonads causing migration to a variety of ectopic locations. Cxcr4b was therefore concluded to be involved in chemotaxis.
12 Stromal derived factor 1 (SDF-1) has been identified as the ligand for Cxcr4 in mice. Antisense reduction of SDF-1 results in an identical defect in directed migration as does its receptor (Knaut et al. 2003). Investigation of the role of both Cxcr4 and SDF-1 in the mouse led to slightly different results. In Cxcr4 deficient embryos, whereas germ cells in the zebrafish had no direction to their migration, mouse germ cells do initiate migration normally and no defect is found during migration to the hindgut. Some germ cells do make it into the urogenital ridges, although numbers are lower compared to controls, suggesting a defect in migration and survival once the germ cells leave the hindgut and migrate through the dorsal mesentery towards the urogenital ridges. There is no significant increase in the number of PGCs which migrate to ectopic locations (Ara et al. 2003; Molyneaux et al. 2003b). These are just two examples of genes involved in PGC development which are highly evolutionarily conserved, but whose function during development has significantly diverged. In zebrafish, the association of these proteins mediates the direction in migration, while in the mouse, the initial stages of migration occur normally, but entry into the gonads is compromised. Another example of such a gene is Nanos. In Drosophila, nanos is essential for PGC migration. However, the mouse homolog nanos3, has no role in migration, but is required for proliferation and survival of migrating PGCs (Tsuda et al. 2003). In summary, PGCs leave the base of the allantois and migrate into the hindgut endoderm on 8.5 dpc. Although the initial phase of their migration may be passive, PGCs then actively migrate along the hindgut endoderm, through the dorsal mesentery
13 and into the developing urogenital ridges. Steel Factor and c-kit are required for PGC migration and survival in the hindgut, while cxcr4 and SDF-1 are required for migration from the hindgut into the gonads. PGCs which stray from their path of migration will die by apoptosis, although occasionally these cells can give rise to teratocarcinomas (Stallock et al. 2003). Sex Determination Shortly after the PGCs enter the developing gonads, proliferation ceases and the germ cells initiate their first sexually dimorphic changes on 13.5 dpc. Prior to this point, both male and female PGCs are identical and have the potential to enter either pathway, independent of their respective sex chromosome constitution. Initial development of the gonad primordium also occurs independent of chromosomal content, although sex determination of the gonads is apparent one day before the PGCs, on 12.5 dpc. During mouse development, the gonads form on the ventral surface of the mesonephric tissue between 10.5 and 11.5 dpc. At this time the gonad is bipotential, being capable of giving rise to either ovarian or testicular cell types. Within the mesonephric tissues, both male and female ductal systems have begun to form alongside one another, one of which will regress upon the embryos decision to become a male or female. The first divergence between males and females occurs on 11.5 dpc when males initiate expression of Sry, the Y-linked sex determining gene. Sry expression begins in the pre-Sertoli cell population which subsequently secretes anti-Mullerian hormone (AMH), causing the regression of the female ductal system in the mesonephros. If no Sry is expressed, either due to the lack of a Y chromosome or genetic mutation, no AMH is secreted, the male ductal system degrades, and the gonad will form an ovary. Therefore,
14 it can be concluded that the default pathway for embryos of either sex is the female pathway, expression of Sry overrides this default program and allows for subsequent initiation of the male pathway (Capel 2000)). Likewise, the germ cells are pre-programmed to follow the female pathway. During migration some cells will stray from their correct path. Many of these end up in the adrenal gland, while others can be found in the mesonephric tissue adjacent to the gonads. Although germ cells in these locations have not been exposed to the gonadal environment, they enter meiosis on the same schedule as those which had properly entered the gonads. Furthermore, entry into meiosis occurs in ectopic germ cells in embryos of both sexes. Therefore, male germ cells, which normally enter a mitotic arrest at 13.5 dpc, enter meiosis if not exposed to the male gonadal environment (Zamboni and Upadhyay 1983). Anne McLaren expanded on these observations by asking whether or not a male germ cell exposed to the male gonad for a short amount of time could enter meiosis, and how long of an exposure is required before the germ cell can no longer enter meiosis. By removing germ cells from male and female embryos between 10.5 and 13.5 dpc and culturing them in lung cell aggregates, entry into meiosis could be analyzed. She found that germ cells removed on 10.5 or 11.5 dpc would enter meiosis regardless of their sex, but by 12.5 dpc, male germ cells have received a signal preventing them from entering meiosis. Therefore entry into meiosis is part of a cell intrinsic program which must be overridden by a signal from the male gonad in order for spermatogenesis to occur (McLaren and Southee 1997). Although this signal is presumed to emanate from the testis cords of the male gonad, the signal remains unknown.
15 In summary, both the gonad primordium and the germ cells which will inhabit them are pre-programmed to follow a female developmental pathway. The male gonad will divert from the female pathway only after expression of Sry is initiated, triggering formation of the testis cords which are clearly visible by 12.5 dpc (Figure 1-4). Once inside the gonads, female germ cells will stop dividing and enter meiosis. Now having entered the first phase of oogenesis, these cells are referred to as primary oocytes and will pass through prophase of meiosis I, arresting at the diplotene stage just before birth. Oogenesis will not be resumed until sexual maturity. Within the male gonad, a signal which originates from the testis cords at 12.5 dpc prevents the germ cells from entering meiosis. The male germ cells, now termed prospermatagonia, will remain in a mitotic arrest until after birth. Primordial Germ Cell Differentiation During embryonic development, primordial germ cells undergo several phases of differentiation during which the cells undergo obvious changes in cell morphology, behavior, and gene expression. In the first phase, cells are restricted in their developmental potency and are eventually induced to become the founding population of the germ cell lineage. One week later, male and female germ cells undergo their first sex-specific changes, females proceeding into meiosis, while males remain in a mitotic arrest. Although neither process, specification nor sex determination can be said to be well understood, there is one aspect of primordial germ cell development for which far less is known. Shortly after the germ cells enter the urogenital ridges both male and female germ cells undergo a common set of changes, independent of the events regarding sex determination. Changes in morphology occur, as germ cells transition from their
16 migratory to non-migratory states (Donovan et al. 1986) and alter their cell adhesion properties (Garcia-Castro et al. 1997). Changes in behavior occur, including the cessation of proliferation and the decreased potential to form pluripotent stem cell lines in culture (Matsui et al. 1992; Resnick et al. 1992). Germ cells of both sexes will also erase genomic imprinting marks and express imprinted genes biallelically (Hajkova et al. 2002; Lee et al. 2002; Szabo et al. 2002), as well as undergo a wave of apoptosis (Coucouvanis et al. 1993). Numerous changes occur during this time and it is not surprising that these changes are accompanied by changes in gene expression. Simplest models might suggest that entry into the gonads exposes germ cells to new signals leading to their differentiation. Although these changes all occur after the germ cells have entered the gonads, there is currently no strong evidence suggesting that the gonadal environment is required for these changes to occur. There is also no information regarding how many of these different events are linked. One possibility is that there is a single regulatory mechanism common to all of these events. Alternatively, it is also possible that there are independent mechanisms regulating each of these individual changes. Currently there is no evidence to support either of these scenarios. The focus of our lab is to understand the regulatory mechanisms controlling primordial germ cell differentiation in the fetal gonad. In the following chapters, data will be presented regarding several aspects of differentiation: 1) Is differentiation controlled by an inductive signal from the gonads, or are the germ cells preprogrammed to differentiate on a specific time schedule? 2) What is the nature of this differentiation program, and how is the program executed? 3) Are the separate differentiation events linked to a common regulatory mechanism? The results to be discussed have led to the
17 proposal of a working model whereby primordial germ cell differentiation is controlled by a cell intrinsic timing mechanism and executed by an epigenetic reprogramming of the genome. In the final chapter, an interpretation of the data will be presented followed by discussion of possible future directions which may lead to a greater understanding for the role of genomic reprogramming in the process of primordial germ cell differentiation.
18 Figure 1-1 Whole mount embryo stained for alkaline phosphatase (TNAP). The arrow points to the primordial germ cell population which expresses high levels of the enzyme (McLaren 2003). Figure 1-2 Primordial germ cell migration in a 9.5 dpc embryo. LacZ is expressed under the control of the TNAP promoter. Primordial germ cells can be seen migrating throughout the length of the hindgut and beginning to enter the urogenital ridges.
19 Figure 1-3 Model for primordial germ cell development. (a) Signals from the extraembryonic ectoderm including Bmp4 and Bmp8b (blue) induce neighboring proximal epiblast cells (open circles) to become precursors of the extraembryonic mesoderm tissues. During gastrulation these cells move (black arrow) through the primitive streak into the extraembryonic portion of the embryo (b). The cells which express fragilis and the highest levels of TNAP (dark green), will go on to become PGCs expressing Pgc7/stella (red). Somatic cells in the region express Hoxb1 (brown). (c) Once PGCs are specified they migrate to the future gonads (dashed arrow) (Hogan 2002). Figure 1-4 Sex determination of the embryonic gonads. 12.5 dpc male and female gonads with attached mesonephros. The testis cords are visible in the male gonad at this stage by the horizontal striping pattern (Schmahl 2003).
CHAPTER 2 MATERIALS AND METHODS Cell Culture Preparing Feeder Layers for Germ Cell Culture Two days prior to the day of dissection, 96 well plates are treated with 0.1% gelatin (Sigma) followed by 50 mg/ml poly-D-lysine (Sigma). The day before dissection, Sl/Sl4 m220 feeder layers (Toksoz et al. 1992) are plated at a density of 2.4 x 104 cells per well. Feeder layers are cultured in DMEM (Invitrogen) supplemented with 100 U penicillin (Invitrogen), 50 g streptomycin (Invitrogen), 2 M glutamine (Invitrogen), and 10% heat inactivated calf serum (Hyclone). All cell culture experiments were incubated at 37oC and 5% CO2. Germ Cell Isolation and Culture Timed matings were set up with B6C3F1 mice (Jackson Laboratories). The following morning, females were examined for copulatory plug. Noon of the day on which a mating plug was first visible was taken to be 0.5 dpc. Pregnant mice were sacrificed on the appropriate day by CO2 asphyxiation followed by cervical dislocation. For 8.5 dpc premigratory PGCs, the most posterior end of the embryo at the base of the allantois was removed (Donovan et al. 1986). The embryonic fragments were incubated for five minutes in 200 l trypsin-EDTA (1X;Invitrogen) followed by thorough trituration with a pipette tip to yield a single cell suspension. The germ cells are then plated onto the feeder layers in QBSF-58 (Quality Biological, Gaithersburg, MD) medium supplemented with 100 U penicillin (Invitrogen), 50 g streptomycin 20
21 (Invitrogen), 2 M L-glutamine (Invitrogen), and 1000 U LIF (ESGRO, Chemicon) per 500 ml. In some experiments, heat inactivated fetal bovine serum (Hyclone, Logan, UT) was added to 0.5% to stabilize feeder layers (Richards et al. 1999b). PGCs from 8.5 dpc embryos are plated onto mitotically inactivated (approximately 500 rads) feeder layers at 0.7-0.9 embryo equivalents per well. PGCs from 10.5 and 11.5 dpc immunomagnetic purifications (see below) were plated at 0.2-0.4 equivalents per well. All germ cell were performed in replicate, and each data point collected is the average of at least five replicates. Statistical anaylsis for all in vitro assays was performed using the Student's t-test with a significance of P<0.01 (http://www.physics.csbsju.edu/cgi-bin/stats/t-test_form.sh?nrow=10). Teratocarcinoma Cell Culture The F9 teratocarcinoma cell line was plated onto plates pretreated with 0.1 % gelatin (Sigma). Cells were cultured in DMEM (Invitrogen) medium supplemented with 100 U penicillin (Invitrogen), 50 g streptomycin (Invitrogen), 2 M glutamine (Invitrogen), and 10% heat inactivated fetal calf serum (Hyclone). Cells were incubated at 37oC in 5 % CO2. Immunocytochemical Methods Alkaline Phosphatase Staining Endogenous alkaline phosphatase was detected by several washes in 1X PBS followed by a 4% paraformaldehyde fixation for 15 minutes. After fixation, the cells were briefly washed with water followed by room temperature incubation in 1 mg/ml FastRed TR and 0.01% naphthol AS-MX phosphate. This causes a colorimetric reaction catalyzed by the alkaline phosphatase enzyme resulting in a red color change (Figure 2-1). After completion of the color reaction the cells are washed with 1X PBS several
22 times. Cultured feeder layers without germ cells were stained as negative controls. Plates can be stored at 4oC. Cells are counted before double staining for GCNA1 since the alkaline phosphatase stain is sometimes lost during the procedure. GCNA1 Staining: Cell Culture Immunocytochemistry for GCNA1 was performed using the 10D9G11 monoclonal antibody (Enders and May 1994). Cultures were fixed overnight at 4C in methanol:dimethyl sulfoxide (4:1), rehydrated by 15 minute incubations in 50% and 30% methanol, rinsed in PBS, and blocked at 4 for 1 hour in PBSMT (Phosphate buffered saline, 2% w/v nonfat dry milk(Carnation), 0.1% Triton X-100 (Sigma)) and in PBSMT containing 0.1% normal goat serum (Sigma). Cultures were then incubated overnight in 10D9G11 supernatant diluted 1:20 in PBSMT at 4oC. After two 10 to 30 minute washes in PBSMT at 4oC, and 3 washes at room temperature, alkaline phosphatase conjugated goat anti rat IgM (Pierce), diluted 1:1000 in PBSMT was added overnight at 4oC. The PBSMT washes were repeated and followed by three room temperature washes in NTMTL (0.1 M NaCl, 0.1 M Tris HCl pH 9.2, 50 mM MgCl2, 0.1% Triton X-100, 2 mM levamisole). Color development was performed with 175 g BCIP/ml: 250 g/NBT/ml (BCIP is 5-bromo-4-chloro-3-indolyl phosphate, NBT is nitroblue tetrazolium, (Boerhinger-Mannheim)) in NTMTL, resulting in a black approximately 10-20 minutes at room temperature. Cultured feeder layers without germ cells were stained as negative controls. GCNA1 Staining: Whole Mount Whole mount embryo staining for GCNA1 was carried out by an overnight fixation at 4oC in methanol:dimethyl sulfoxide (4:1), rehydrated by 15 minute incubations in 50% and 30% methanol, rinsed in PBS, and blocked at 4oC 2 times for 1 hour in PBSMT.
23 Embryos were then incubated overnight in 10D9G11 supernatant diluted 1:50 in PBSMT at 4oC. After two 20 to 60 minute washes in PBSMT at 4oC, and 3 washes at room temperature, horse radish peroxidase (HRP) conjugated mouse anti rat IgM (Zymed), diluted 1:000 in PBMST was added overnight at 4oC. The following day, embryos were washed in PBSMT twice for 20 to 60 minutes at 4oC and 3 times at room temperature. After three 5 minute washes in PBT (1X PBS, 0.2 % BSA, 0.1% Triton X-100) at room temperature, HRP detection was performed using the Liquid DAB Substrate Kit following the manufacturers instructions (Zymed). GCNA1 staining of 11.5 to 13.5 dpc embryos was used for positive control, while staining of 8.5 to 9.5 dpc embryos was used for negative controls. BrdU Labeling of Cultured Germ Cells Premigratory 8.5 dpc PGCs were plated on feeder cells at an approximate density of 0.7 embryo equivalents per 96-cell well in QBSF-58 (Quality Biological, Gaithersburg, MD) medium supplemented with 100 U penicillin, 50 g streptomycin, 2 M L-glutamine, and 1000 U LIF (ESGRO, Chemicon) per 500. Two days after plating, the culture media were replaced with fresh media containing a 1:1000 dilution of BrdU labeling reagent (Zymed, San Francisco, CA), incubated for an additional hour, fixed in ice-cold 70% ethanol, and stained for endogenous alkaline phosphatase as described above (Figure 2-2). The cultures were washed with deionized water and stained for BrdU with biotin-conjugated anti-BrdU monoclonal antibody as suggested by the supplier (Zymed). Cultured feeder layers without germ cells were stained as negative controls. Primordial Germ Cell Isolation Timed matings were set up with B6C3F1 mice obtained from Jackson Laboratories (Bar Harbor ME). The following morning, females were examined for copulatory plug.
24 Noon of the day on which a mating plug was first visible was taken to be 0.5 dpc. Pregnant mice were sacrificed on the appropriate day by CO2 asphyxiation followed by cervical dislocation. Embryos were removed from the uteri and placed in 1X PBS for dissection. For 10.5 and 11.5 dpc embryos, gonads with attached mesonephros were removed and placed in a microfuge tube on ice for the duration of the dissections. For 12.5, 13.5 and 14.5 dpc embryos, gonads were removed and sex segregated based on the presence of testis cords in the male gonad. For 13.5 and 14.5 dpc embryos the mesonephros is also removed. Eight to twelve embryos are used for each purification. When dissections are completed, excess 1X PBS is removed from the collected tissue, 500 l of trypsin-EDTA is added and the sample incubated for five minutes at 37oC. The tissue is then triturated with a pipette to break up the tissue, briefly spun at 2 krpm for 2 minutes, and the trypsin-EDTA then removed and replaced with 1 ml of PBS-DNase (1X PBS, 5 mM EDTA, 0.5% BSA, 20 g/ml DNAse). After a 5 minute room temperature incubation, 800ul was removed, and the tissue thoroughly triturated to a single cell suspension. The sample was then incubated at room temperature for 10 minutes with shaking followed by addition of 40 l of the TG-1 antibody. TG-1 is a monoclonal mouse antibody which recognizes a carbohydrate residue on the cell surface of the PGCs (Beverley et al. 1980). After a 30 minute incubation on ice with shaking the cell suspension is spun down for 2 minutes at 2 krpm and washed three times with 100 l of PBS-DNase. After the third wash the cell pellet is resuspended in 180 l of PBS-DNase and 20 l of the secondary antibody is added. The secondary antibody, rat anti-mouse IgM, has an iron bead conjugated to it (Miltenyi #473-02). After a 30 minute incubation on ice with shaking, the sample was purified.
25 The purification column (Miltenyi) was prepared by placing it on a Miltenyi Mini-MACS magnet (Miltenyi #422-01), and was prewashed with 500 l of Equilibration Buffer (1X PBS, 3% BSA (Sigma)). The cell suspension was then passed over the column two times, allowing for PGCs which are now associated with the antibody bound iron bead to bind to the column while attached to the magnet. The flow through, or immunodepleted fraction, contained the contaminating somatic cells, mostly gonadal and mesonephric somatic cells. The column was washed four times with 500l of PBS-DNase to remove any nonspecifically bound cells and then removed from the magnet to collect the purified fraction. The column was eluted with 500l of PBS-DNase by gravity, followed by another 1ml forced through the column with the supplied plunger. This 1.5 ml contains the purified PGCs. To assess purity, 150 l is centrifuged onto a CSA-100 silanated slides (CEL Associates, Inc) pretreated with poly-D-lysine using a Cytospin 2 (Shandon) at 700 rpm for 10 minutes. The slide is then stained for alkaline phosphatase to determine the percentage of PGCs versus somatic cells present. All purifications used were no less than 80% pure. Preparation of Genomic DNA for Bisulphite Conversion Following immunomagnetic purification, primordial germ cell samples are centrifuged at 4 krpm for 4 minutes and the supernatant discarded. Occasionally a cell pellet may be visible. Lysis buffer (1 mM SDS, 0.28 mg/ml proteinase K, 1.05 mg/ml glycogen in 1X PBS) was added to the samples for a total volume of 37.8l. DNA can then be stored at -80oC and shipped on dry ice. Just prior to bisulphite conversion the DNA was incubated at 37oC for 1.5-2 hours, then heated to 95oC for 10 minutes. Samples were cooled to room temperature before beginning the bisulphite conversion.
26 Bisulphite Sequencing of Genomic DNA Bisulphite Conversion Bisulphite conversion was carried out as described (Clark et al. 1994). DNA was denatured with 4.2 l of 3 M NaOH (final concentration is 0.3 M in 42 l) and incubated for 37oC for 30 minutes. Freshly prepared 2M sodium metabisulphite and 10 mM hydroquinone (final concentrations 1.55 M and 0.5 mM respectively) was added to the denatured DNA to a final volume of 240 l. The bisulphite reaction was inefficient on double stranded DNA. Samples were then incubated in the dark for 16-20 hours. A water control was also prepared and treated with the bisulphite stock for later use as a PCR control. Following overnight incubation, free bisulphite ion was removed by passing the sample through a desalting column (Promega Wizard DNA Clean-up System) and eluted in 45 l of water. Freshly prepared 3 M NaOH was added and the samples were incubated at 37oC for 15 minutes. The DNA was then precipitated overnight at -80oC by adding 25 l 7.5M Ammonium Acetate and 200 l 100% Ethanol. The following day DNA was pelleted by centrifugation at 13,000 rpm for 30 minutes at 4oC followed by a 70% Ethanol wash. The final pellet was dried by vacuum centrifugation and resuspended in 100 l of water. Bisulphite Polymerase Chain Reaction Bisulphite primers were designed against the bisulphite converted DNA. In order to avoid any amplification of unconverted DNA, primers were designed in regions which contained numerous converted cytosines (now thymines), predominantly at the 3' end. Primers were designed in regions in which no CpG dinucleotides were present, however if a primer did include one or two CpGs a degenerate nucleotide (cytosine or thymine)
27 was included at those positions. Care was also taken to include as many guanine residues as possible to increase the Tm of the primer, converted sequences have very few cytosines causing melting temperatures to be relatively low. Primer lengths were occasionally increased to account for this. Primers were synthesized by Integrated DNA Technologies (IDT) and purified using standard desalting methods . Primer sequences are listed in Table A-1. Polymerase Chain Reaction (PCR) was initially performed on bisulphite converted ES cell DNA, as well as unconverted ES cell DNA to ensure that primer sets exclusively amplified products from the converted DNA. Once PCR conditions were optimized on converted ES cell DNA, PCR was performed on bisulphite treated germ cell preparations with HotStar Taq (Qiagen) using the following PCR conditions (final): 1X PCR buffer (with 15 mM MgCl2), 200 M each dNTP, 1 M each primer, and 1.5 U HotStar Taq DNA polymerase. Bisulphite converted germ cell DNA preparations were not quantitated for DNA concentration and 10 l (approximately one embryo equivalent) was used per 50 l reaction. PCR reactions were performed using the following cycling conditions: An initial denaturation at 95oC for 15 minutes was required for activation of HotStar Taq polymerase. A denaturation at 95oC for 45 seconds was followed by a 53oC annealing for 30 seconds followed by a 72oC elongation for 1.5 minutes and all three steps repeated 35 times followed by a final 72oC elongation for 10 minutes. Bisulphite PCR amplification was performed one time on two independent germ cell purifications to test for any inconsistencies which might arise from conducting PCR on small amounts of DNA.
28 Purification of PCR Products PCR products were separated on 1.5-2.0% low melting point agarose (Invitrogen) gels, the correct bands cut out and purified (Promega Wizard DNA Clean-up System). If PCR reactions routinely contained no spurious bands, only a small sample (5-8 l) would be run on a 1.5-2.0% agarose gel and the remaining PCR product purified without gel purification (QIAquick PCR Purification Kit). Cloning and Sequencing of PCR Products PCR products were cloned using either TA cloning with pGEM-T and pGEM-T Easy Vector Systems (Promega) or the Topo TA Cloning Kit for Sequencing (Invitrogen) according to the manufacturers recommendations. Ligations were performed as suggested by the respective kit protocols. Ligations were transformed into chemically competent, subcloning efficiency DH5 (Invitrogen) (for pGEM-T), or TOP10 cells (Invitrogen) (Topo TA). Transformations were plated onto Luria-Bertani (LB) plates (1% tryptone, 0.5% yeast extract, 1% sodium chloride and 1.5% agar) containing 50 g/ml ampicillin and supplemented with 100 l of 100 mM IPTG and 40 l of 40 mg/ml X-Gal per plate (Gold Biotechnology, Inc) for blue-white colony selection (Sambrook et al. 1989). Plates were incubated overnight at 37oC and the following morning white colonies were picked into 3 ml of culture media (1% tryptone, 0.5% yeast extract, and 0.5% sodium chloride). Liquid cultures were grown overnight at 37oC and plasmid DNA extracted by the alkaline lysis method (Sambrook et al. 1989). Plasmid sequencing was carried out using SP6 primer for pGEM T-vectory and T3 for Topo vector using the standard plasmid sequencing methods (described below). Sequences were analyzed using Sequencher 4.2 (Gene Codes Corporation).
29 Reverse Transcription Polymerase Chain Reaction For RT-PCR analysis of 5-azacytidine treated F9 cells, one confluent plate of F9 teratocarcinoma cells was collected at one, two and three days after treatment with 10 M 5-azacytidine. One plate of untreated F9 cells was also collected as a control. For preparation of probes for BAC library screening, one mouse testis was collected. RNA was first prepared from tissue or RNA samples according to the manufacturers instructions (RNAzol). DNA contamination in RNA preparations was degraded by combining 10 g RNA, 1 l of 10X DNAse buffer (Invitrogen), 1 l DNAse (Invitrogen) and dH20 to a final volume of 10 l. RNA was quantitated using a spectrophotometer (BioRad) and 5 g of RNA was converted to cDNA. Briefly, 5 g RNA, 1 l of random primers (500 g/ml), and dH20 were combined to a final volume of 27.4 l. Primers were annealed by heating the reaction to 68oC for 3 minutes and then cooling on ice. Reverse transcription was then carried out by adding 1.6 l of each dNTP (2.5 mM; Sigma), 8l of 5X transcription buffer (Invitrogen), 2 l of DTT (Invitrogen), 1l of RNase inhibitor (RNasin; Promega) and 1 l of Superscript reverse transcriptase (Invitrogen) followed by a 37oC incubation for 1 hour. Another 5 g of RNA was set up in an identical reaction but no Superscript was added, these samples were used as negative controls. The reaction was stopped by addition of 1 l of 25mM EDTA (Invitrogen) and incubated for 65oC for 10 minutes. Isolation of Genomic Loci Probe Preparation Probes for screening genomic BAC libraries were generated by PCR amplification from testis cDNA. Primers DAZ-F1 and DAZ-R1, SCP3-F1 and SCP3-R1 and Mage-b4-F1 and Mage-b4-R1 were used to generate Dazl, Scp3 and Mage-b4 probes (Sequences
30 listed in Table A-5). PCR products were labeled with 32P-dCTP by random priming using the Prime-It II kit (Stratagene). Briefly, probe DNA was boiled for 5 minutes with random primers (provided in kit) and cooled on ice. For labeling, 5X reaction buffer, 25 Ci of 32P-dCTP (Amersham) and 1U of Klenow (exo-) was added to the DNA/primer complex and incubated for 15 minutes at 37oC. The probe was then purified using a nucleotide removal kit (QIAGEN) and eluted in 200 l of water. To determine specific activity, 1 l was spotted onto a filter and counts per minute calculated using the Bioscan/QC-4000 XER (Bioscan, Inc.). Screening of the Research Genetics Bacterial Artificial Clone Library The Research Genetics Bacterial Artificial Chromosome (BAC) Library (Research Genetics) was screened for clones containing the genomic loci for Dazl and Scp3. The BAC clones are derived from genomic DNA generated from male mice (129/Sv) which was partially digested into fragments averaging 130 kilobases size, and cloned into the pBeloBAC11 vector. The library contains 9 membranes each with 27,648 unique clones spotted in duplicate providing 8X coverage of the genome. In order to screen the BAC clones for the genes of interest, RT-PCR probes were generated from testis cDNA. To ensure that these probes do not hybridize to repetitive sequences (which could prevent further use of the BAC library) probes were tested by hybridization to a strip of EcoR1 digested genomic DNA, each probe only hybridized to a single band. Screening of the BAC library was carried out as specified by the manufacturer (Reserach Genetics). Membranes were prepared by washing in a large tray with 6X SSC and 0.1% SDS for 15 minutes followed by 2 rinses, 5 minutes each, with 6X SSC. The membranes were then placed in three hybridization tubes and individual membranes separated by Flow Mesh (Diversified Biotech). Prehybridization was performed at 65oC
31 in a hybridization oven for 15-30 minutes using 40 ml of Hyper Hyb (Research Genetics) hybridization buffer per membrane (120 ml per tube). Labeled probes were counted to determine specific activity, around 1-5 X 109 counts were used for each probe. Probes for each gene were combined and denatured by 5 minutes of boiling followed by cooling on ice. The denatured probe was then added directly to the pre-hybridization buffer and hybridization continued for 1 hour. The membranes were then washed three times for 15 minutes each at 65oC with 1X SSC and 0.1% SDS using 30 ml per membrane. Washing was then continued for 15 minutes at room temperature in large trays containing 1X SSC. Membranes were wrapped in plastic wrap and exposed to film (XAR, Kodak) overnight. Following exposure, positive clones were identified by the presence of duplicate spots in a preset configuration and BAC clones identified using the reference marks along the membrane edges. Ten clones were ordered from Research Genetics: 19K8, 327M2, 288N10, 498A5, 288M10, 146C13, 307J11, 34J8, 60J6, and 2N7. Screening the Roswell Park Bacterial Artificial Chromosome Library The Roswell Park BAC library (Research Genetics) was generated from three female mice (C57BL/6J) by partial digestion of pooled kidney and brain genomic DNA. Clones have an average size of 200kb into the pBACe3.6 vector. Each membrane from this library was spotted with 18,432 unique clones spotted in duplicate onto 10 membranes providing 11.2X coverage of the genome. This library was screened for the genomic locus containing the Mage-b4 gene using a PCR probe generated from testis cDNA. Library screening was performed as suggested by the manufacturer (Research Genetics). Membranes were soaked in 150 ml of pre-warmed (65oC) Church's Buffer (1 mM EDTA, 0.5 M NaHPO4 pH 7.2, 7% SDS, and 1% BSA). Five membranes separated
32 by mesh were placed into 2 hybridization tubes and an extra 25 ml of Church's Buffer added to each tube for prehybridization for 2 hours at 65oC. Labeled probe was used at a specific activity between 1-10 X 106 counts per minute. After denaturation by boiling for 5 minutes and cooling on ice, the probe was added directly to the hybridization tube and incubated overnight at 65oC. The following day membranes were washed at 65oC with Wash I (1 mM EDTA, 40 mM NaHPO4 pH 7.2, 5% SDS, and 0.5% BSA) five times for 15 minutes each. The membranes are then removed from hybridization tubes and placed in a large tray filled with Wash II (1 mM EDTA, 40 mM NaHPO4 pH=7.2, and 1% SDS), and rotated at 65oC, three times for 15 minutes each. After a briefly rinsing with water, the membranes were wrapped in plastic wrap and exposed to film overnight. Positive clones were identified according to the manufacturers instructions. Twelve BAC clones were then ordered: 467O24, 474B22, 441C13, 346H8, 346H6, 346B6, 354N3, 463H9, 451L4, 363M12, 357M7, and 337M12. A previous screen of the Roswell Park library had identified BAC 105P19 as containing the Mvh locus. Verification of BAC Clones BAC clones from both the Research Genetics and Roswell Park libraries were streaked onto LB plates containing 20 g/ml chloramphenicol. Since there is a small chance of contamination from other clones in the BAC clone stocks, several colonies were picked for each clone. Duplicate clones were grown in 3 ml of liquid culture overnight at 37oC and DNA prepared the following day by alkaline lysis. BAC DNA was then digested with EcoR1 and run on 0.6% agarose gels. Gels were Southern blotted (described below), hybridized with 32P-dCTP labeled probes and exposed to film for the
33 appropriate length of time. Films were exposed at -80oC and developed using the KODAK X-OMAT 2000A Processor (Kodak). From the Research Genetics BAC library, one positive clone was identified for Dazl (327M2) and three positive clones for Scp3 (498A5, 307J11, and 34J8). Six positive BACs were identified for Mage-b4 from the Roswell Park library (441C13, 346H8, 474B22, 467O24, 346H6, and 346B6). Southern Blotting Southern blotting was performed as described (Sambrook et al. 1989).Agarose gels were placed on a ultra violet (UV) light box to nick DNA. After 5 minutes, gels were soaked in alkali solution (1.5 M NaCl and 0.5 N NaOH) for 45 minutes followed by neutralizing solution (1.5 M NaCl and 1 M Tris pH 7.4) for 1.5 hours (Sambrook et al. 1989). Gels were then blotted using 10X SSC overnight allowing for transfer of the DNA from the gel to the nylon membrane. The following day, the membrane was rinsed in 2X SSC, baked at 80oC for several hours and hybridized with 20 ml of Church and Gilbert hybridization buffer (Church and Gilbert 1985) (2.5% BSA, 1 mM EDTA pH 8.0, 0.25 M sodium phosphate buffer pH 7.2, and 7% SDS) at 65oC for 2 hours. The prehybridization buffer was then poured off and another 5 ml of buffer, containing the denatured probe, was added to the hybridization tube and incubated at 65oC overnight. The membrane is washed three times for 15 minutes each with 2X SSC and 0.1% SDS, then wrapped in saran wrap and exposed to film. Generation of Genomic Subclones Due to their large size, BAC clones are difficult to work with. They are present at low copy numbers in bacteria and DNA preparations tend to degrade quickly, therefore it is advantageous to clone out fragments of the BAC which contain the region of interest
34 for further study. Two methods were used for subcloning BAC fragments: direct cloning into Bluescript vector (Stratagene) or generating subgenomic BAC libraries. Generation of Subgenomic Libraries Subgenomic libraries were prepared for BAC 327M2 containing the Dazl locus, and BAC 105P19 containing the Mvh locus. BAC DNA was prepared from 50 ml LB cultures by alkaline lysis. DNA was then partially digested with Sau3A1 enzyme which is a four-base cutter leaving overhanging ends which can then be cloned into BamH1 sites. After partial digesting conditions were optimized, a large digest was set up (150 l BAC DNA, 30 l Sau3A1buffer, 15 l Sau3A1 enzyme and 105 l water) which generated large overlapping fragments from the BAC. Digests were inactivated by addition of 75 l 0.5 M EDTA followed by a 20 minute incubation at 65oC. After two phenol chloroform-isoamylalcohol (PCIA) (24:1:1) extractions and one CIA extraction, the samples were ethanol precipitated, resuspended in STE. Continuous sucrose density gradients (10-40 %) were prepared by washing 14 X 19 mm Polyallomer centrifuge tubes (Beckman) with 95% ethanol and air dried. Tubes were then filled with 11.5 ml of 25% sucrose in STE buffer (10 mM Tris, 6 mM EDTA, 10 mM NaCl) and frozen at -20oC covered with parafilm. Once completely frozen, the tubes were removed from the freezer and allowed to thaw. The freeze/thaw was repeated one additional time, generating the sucrose density gradient. Once gradients were completely thawed a second time, 200l of partially digested BAC DNA (approximately 100g) was carefully pipetted on top of the sucrose. Tubes were placed in a SW41Ti (Beckman) rotor and spun overnight in an ultracentrifuge at 30,000 rpm at 20oC. The following day, the centrifuges was stopped without use of the brake. Samples were collected by using a 25 gauge needle to poke a small hole in the
35 bottom of the tube allowing 500 l fractions to be collected into microfuge tubes. From these fractions, 50 l was removed and run on a 0.8% agarose gel to verify size separation of the DNA fragments. To the remaining 450 l, 1 ml of ethanol and 45 l of 3 M sodium acetate was added and the samples stored at -20oC. The ZAP Express Predigested Vector Kit (Stratgene) was initially used to generate subgenomic libraries from the BAC clones. The ZAP Express vector is predigested with BamH1 and can hold inserts between 0-12 kb in size. The Lambda Dash II vector arms are also predigested with BamH1 and can accommodate inserts between 9 and 23 kb. Since large fragments are desired, fractions containing bands between 10-12 kb (ZAP Express) and 20-23 kb (Lambda Dash II) were centrifuged at 4oC for 30 minutes at 13,000 rpm. Samples were washed with 70% ethanol, dried by vacuum centrifugation and resuspended in a total of 15 l of water. Ligation into phage arms was set up according to the Stratagene protocol: 1 l phage arms (ZAP Express or Lambda Dash II, 1 g/l), 2.5 l insert (0.2-0.4 g), 0.5 l 10X ligation buffer (Stratagene), 0.5 l 10 mM ATP, 0.5 l ligase (Stratagene). Ligations were performed overnight at room temperature for ZAP Express, and 4oC overnight for Lambda Dash II. The following day, ligations were packaged into phage particles using the Gigapack III gold packaging kit (Stratagene) as instructed by the protocol. The packaging extract was quickly thawed and 2.5 l of the ligation added. Following a 2 hour room temperature incubation, 20 l of chloroform and 500 l of SM buffer was added to the reaction. Phage samples were then titered to determine optimal plating conditions for library screening. Phage stock dilutions in SM buffer were prepared at ratios of 1:10, 1:100 and
36 1:1000. MRF' (ZAP Express) and MRA(P2) (Lambda Dash II) bacteria were prepared by growing overnight at 30oC in 30 ml of LB media supplemented with 100 mM maltose and 100 mM MgSO4. The bacteria were then pelleted and resuspended in 20 ml of 10 mM MgSO4. In 15 ml tubes (Falcon) the phage dilutions were combined with 200 l of bacteria and incubated for 37oC for 15 minutes, allowing the phage to bind to the bacteria. Following incubation, 3 ml of warm top agarose (LB with 0.8% agarose) was added to the bacteria/phage mixture, the tube gently vortexed and quickly poured onto a prewarmed 10 cm LB plate (LB with 1.5% agarose). Once the agar has solidified, the plates were inverted and incubated overnight at 37oC. The following day plaque numbers are counted and multiplied by the dilution factor to generate the number of plaque forming units per ml (pfu/ml). Once the appropriate concentration was determined, five 15 cm LB plates were prepared as above to screen each library. To screen each phage library for clones containing promoter regions for each gene, plaque lifts were performed to transfer phage colonies to nitrocellulose filters (Schleicher and Schuell) which could then be hybridized with the appropriate probe and exposed to film. Plaque lifts were performed by chilling the library plates at 4oC for 2-3 hours to prevent the filters from sticking to the agar. The nitrocellulose filters were carefully placed on top of each plate and stabbed with a needle dipped in ink to mark orientation. The filters then floated on top of 200 ml of alkali solution (1.5 M NaCl and 0.5 M NaOH) for 2 minutes, 200 ml of neutralizing solution (1.5 M NaCl and 0.5 M Tris) for 5 minutes and 200 ml of wash solution (0. 2 M Tris and 2X SSC) for 30 seconds. Filters were briefly placed on Whatman 3M paper and then baked for 2 hours at 80oC.
37 To screen for phage containing promoter fragments, filters were hybridized with cDNA probes from exon 1 for each gene. Filters were prehybridized for 2 hours at 65oC in 100 ml of phage hybridization buffer (5X SSCP, 4X Denhardt's, 0.1% SDS, and 0.1 mg/ml sonicated salmon sperm DNA). Prehybridization solution was discarded and 5-10 ml of fresh solution is added which contained the denatured 32P-dCTP labeled probe and incubated overnight at 65oC. The following day filters were washed three times for 15-20 minutes each with 0.5X SSCP with 0.1% SDS at 65oC. Filters were then wrapped in plastic wrap and exposed to film for several hours at -80oC. Once the film had been developed, positive clones were identified. Positive plaques were picked from plates by removing a circle of agar around the plaque with the back end of a glass pipette. The agar plug was then placed in a microfuge tube containing 0.5 ml of SM buffer and a drop of chloroform and the tube vortexed to remove the phage from the plug. A high titer phage stock was then prepared by combining 20 l of phage with 0.2 ml of the appropriate bacteria for 15 min at 37oC. The mixture was then combined with 3 ml of top agarose and poured onto a 10 cm LB plate. After overnight incubation at 37oC, 2-3 ml of SM buffer was added to the plate, and the plate gently rocked for 1-2 hours. The SM buffer was then pipetted off the plate and centrifuged to remove bacteria. The supernatant was transferred to a microfuge tube and a drop of chloroform added. This was the high titer phage stock that can be used to prepare DNA. The ZAP Express vector is designed to allow easy excision of inserted DNA to form the pBK-CMv phagemid vector. The in vivo excision was carried out by M13 ExAssist helper phage which was simultaneously infected into the XLOLR strain of
38 bacteria. To excise the phagemid vector with inserted DNA, 250 l of high titer phage stock was combined with 200 l of XL1-Blue MRF' cells (OD600 of 1.0) and 1 l of ExAssist helper phage (>1X106 pfu/l) in a 15 ml tube (Falcon) and incubated at 37oC for 15 minutes. Three ml of LB media was then added to the tubes and incubation is continued for 2.5-3 hours with shaking. Falcon tubes are then heated to 65oC for 20 minutes and centrifuged for 15 minutes at 1000g to remove bacteria. The supernatant (containing the excised pBK-CMV phagemid vector) was moved to a clean Falcon tube and can be stored at this step for several months at 4oC. To plate the excised phagemids, 10 and 100 l of phage supernatant were added to 2 separate Falcon tubes containing 200 l of freshly grown XLOLR cells (OD600 of 1.0). The mixtures were incubated for 37oC for 15 minutes, 3 ml of LB media was then added and the sample incubated for another 45 minutes at 37oC. Between 100-200 l of the cell mixture was plated onto LB plates containing 50 g/ml kanamycin and incubated overnight at 37oC. The colonies which appeared the next day contained the excised pBK-CMV phagemid with the cloned insert DNA. No helper phage is present in these colonies since replication cannot occur in the XLOLR bacterial strain. The ZAP clones which were isolated contained a variety of fragments overlapping exon 1 of both Mvh and Dazl. The Mvh 2-2 clone was the most useful clone generated and was used to generate sequence 5' of exon 1 (Table A-2) and used to generate a LacZ promoter construct. The Daz 4-1 clone was used to generate sequence 5' of exon 1 (Table A-4). In order to isolate insert DNA from the Lambda Dash II vector, phage genomic DNA was first prepared, the insert digested out of the vector with the Not1 enzyme which
39 flanks the insert site, and then re-cloned into a Not1 cut Bluescript (pBSK). To make phage genomic DNA, 0.4 ml of bacteria (grown up and now in 10 mM MgSO4) was combined with 0.2 ml of the high titer phage stock and incubated for 15 minutes at 37oC. Ten ml of top agarose was added and the sample inverted several times then poured onto a 15 cm LB-agarose plate (1.5% agarose). Growing the phage on agar at this point could inhibit subsequent enzymatic reactions. The plate was incubated at 37oC overnight. The following day, 5 ml of phage diluent buffer (10 mM Tris pH 7.5, 10 mM MgSO4) was added to each plate and gently rocked for 1-2 hours. The buffer was then removed and placed in a Falcon tube and the plate washed with another 2 ml of phage diluent buffer and added to the tube. Tubes were then centrifuged at 8,000 rpm for 10 minutes to pellet bacteria. After moving the supernatant to a fresh tube, 2 l DNase I (1 mg/ml) and 2 l RNase (10 mg/ml) were added and incubated for 30 minutes at 30oC. The enzymes were then inactivated by adding 30 l of 10% SDS and 30 l of 0.5 M EDTA per ml of supernatant followed by heat inactivation at 65oC for 20 minutes. To isolate DNA from phage, 140 l of Formic Acid Solution (29.4g potassium acetate plus 5 ml 88% formic acid, water to 100 ml) was added per 1 ml of supernatant and the sample placed on ice for 20 minutes. After centrifuging at 8,000 rpm for 10 minutes, the supernatant was extracted with PCIA, centrifuged again, extracted with CIA and centrifuged at 5,000 rpm for 5 minutes. The supernatant was then ethanol precipitated at room temperature for 5 minutes and centrifuged at 8,000 rpm for 10 minutes. After washing with 70% ethanol and drying by vacuum centrifugation, the pellet was resuspended in 60 l of TE with 10 g/ml RNase A (Sigma).
40 To exise the genomic insert, phage DNA was digested with Not1 which flanks the insert site in the phage arms. Digests were first run on a 0.8% agarose gel to verify insert is present and then cloned into Not I cut Bluescript. Genomic clones for both Dazl and Mvh were successfully cloned in this manner and fragments spanning exon 1 which were around 20 kb in size were obtained. Interestingly, Mvh contains a Not I site upstream of exon 1, therefore the genomic fragments cloned for Mvh all begin or end at the endogenous Not1 site. The clone Mvh3.5 2-1 was used for generating sequence upstream of exon 1 and for generating LacZ reporter constructs. Direct Subcloning into Bluescript The 5' flanking region of Scp3 was subcloned directly into the Bluescript vector. Subcloning directly from BAC clones was also attempted Mvh and Dazl as well, but the process was very inefficient and was only successful for Scp3. The Scp3 containing BAC 307J11 was digested with a variety of enzymes, Southern blotted and then hybridized with a probe from exon 1. The Hind III enzyme was found to excise a 7-8 kb fragment which hybridized to the probe. This fragment was ligated to Hind III digested Bluescript vector and transformed into DH10B E. coli (Invitrogen) by electroporation to maximize efficiency. Eight positive clones were identified by hybridization with the exon 1 probe, two of which were verified by sequencing. One clone, now referred to as pSCP3, was selected and used to sequence the entire insert (primers listed in Table A-3). After generating a contig from the plasmid sequences generated, it was found that the insert was only 6.0 kb in size.
41 Sequencing DNA sequencing was carried out using ABI Prism BigDye terminator (PerkinElmer) which uses the AmpliTaq DNA polymerase. Once sequence reactions were completed the samples were sent to the Center for Mammalian Genetics DNA Sequence Core (Florida) where the samples are run on the ABI Prism 377XL Automated DNA Sequencer (PerkinElmer). Sequence files which were received from the sequencing core were then analyzed using Sequencher 4.2 (Gene Codes Corporation). Plasmid Sequencing Sequencing samples were prepared as suggested by the manufacturer (PerkinElmer): 2 l BigDye Terminator, 2 l 5X sequencing buffer, 1 l 3.2 pmole/l primer, 2 l water and 3 l plasmid DNA (approximately 3 g). Sequencing PCR was carried out by 24 cycles of s of 96oC for 30 seconds, 50oC for 15 seconds and 60oC for 4 minutes. Reactions were purified using Performa DTR Gel Filtration Columns (Edge Biosytems) and dried by vacuum centrifugation. Alternatively, sequencing reactions were purified by Ethanol precipitation where 2 volumes of 100% ethanol and volumes of 7.5 mM ammonium acetate were added to the sequencing reactions. Samples were washed with 70% ethanol and dried by vacuum centrifugation. BAC Sequencing Sequencing samples were prepared as suggested by the manufacturer (PerkinElmer): 4 l BigDye Terminator, 12 l 5X sequencing buffer, 0.5 l 100 M primer, 4 l 25mM MgCl2, 8 l Betaine (Sigma), 1-11 l sheared BAC DNA, and water to 40 l. BAC DNA was sheared by passing 200 l of DNA through a 25 gauge needle 10-15 times. BAC sequencing PCR was carried out by an initial 5 minute denaturation at 96oC followed by 30 cycles of 95oC for 30 seconds, 50oC for 10 seconds and 60oC for
42 4 minutes. Reactions were purified using Performa DTR Gel Filtration Columns (Edge Biosytems) and dried by vacuum centrifugation.
43 Figure 2-1. Identification of primordial germ cells in culture. Germ cells from 8.5 dpc embryos were plated onto mitotically inactivated M220 feeder layers. Staining for alkaline phosphatase results in a red color allowing for the germ cells to be distinguished from the cells of the feeder layer. Figure 2-2. Identification of proliferating germ cells in culture. Primordial germ cells from 8.5 dpc embryos were plated onto mitotically inactivated M220 feeder layers and after 3 days of culture, pulse labeled with 5-bromo-2â€™-deoxyuridine (BrdU). Cells were first stained for BrdU incorporation, followed by staining for endogenous alkaline phosphatase. Double stained cells represent proliferating germ cells.
CHAPTER 3 CELL AUTONOMOUS REGULATION OF PRIMORDIAL GERM CELL DIFFERENTIATION Introduction In Vitro Culture of Primordial Germ Cells Using an in vitro system which allows PGCs to be cultured for several days, aspects of germ cell development can be studied removed from the influences of the gonad. PGCs do not survive in culture for more than two or three days without the presence of a feeder cell layer. Two fibroblast feeder layers are primarily used, STO cells (Stott and Wylie 1986) or Sl/Sl4 m220 cells (Toksoz et al. 1992). Although all of the functions of these feeder layers are not known it is clear that expression of the transmembrane form of the Steel Factor is mandatory for germ cell survival (Dolci et al. 1991). Steel Factor is the ligand for the tyrosine kinase receptor c-kit (expressed by PGCs) and is encoded in two alternative splice forms, a soluble and a transmembrane bound form of the protein. In vitro, the soluble form of the protein is insufficient to support germ cell survival and in vivo, mice carrying the steel-dickie mutation which only produces the soluble ligand are sterile (Brannan et al. 1991). Therefore survival of PGCs in culture depends on the presence of the transmembrane form of the ligand (Dolci et al. 1991). Leukemia inhibitory factor (LIF) also promotes germ cell survival in culture, although mice deficient for the LIF receptor, gp130, show no defect in any aspect of PGC development and therefore may not play a significant role in vivo (Molyneaux et al. 44
45 2003a). Since 8.5 dpc PGCs will only survive in culture for four to five days and 11.5 dpc PGCs only survive for one or two days, there are likely to be other survival factors required at that stage of development which have not been identified. Although there are drawbacks to plating a small number of germ cells on a vast layer of feeder cells, the PGCs can easily be distinguished from contaminating somatic cells and the feeder layer by staining for the highly expressed alkaline phosphatase enzyme (Chiquoine 1954). A transgenic mouse expressing the green flourescent protein from the alkaline phosphatase promoter also allows for visualization of the PGC population (Yoshimizu 1999). Another marker for PGCs is recognized by the monoclonal antibody TG-l, which binds a cell surface carbohydrate antigen (Beverley et al. 1980). These germ cell specific markers have made it possible to study various aspects of PGC development including migration, survival, proliferation and differentiation. Therefore, although survival of PGCs in culture is not indefinite, the culture system remains a valuable technique for studying PGCs. Inductive Versus Autonomous Differentiation Cell differentiation during embryonic development is accomplished by a combination of mechanisms. A cell may be induced or instructed by a neighboring cell, or the cell may already be determined for differentiation. While primordial germ cell specification is mediated by an inductive process, sex determination is a predetermined program for females, but an inductive program for males. Shortly after the arrival of primordial germ cells in the urogenital ridges between 10.5 and 11.5 dpc, the cells undergo a wave of differentiation (reviewed in McLaren, 2003). Perhaps the simplest model to account for the spatial-temporal specificity of PGC differentiation would
46 suggest the presence of an inducing signal in the urogenital ridge, but experimental studies have led to a different conclusion. Several labs have reported results consistent with a cell intrinsic program controlling PGC differentiation. In vivo studies on PGCs in ectopic locations showed that both male and female germ cells will enter meiosis and initiate GCNA1 expression on a timing schedule similar to germ cells which properly entered the gonads. Likewise, in embryos lacking the gene fushi tarazu factor 1 (Ftz-F1) causing degeneration of the gonadal tissue by 12.5 dpc, PGCs continue to differentiate to express the postmigratory marker germ cell nuclear antigen 1 (GCNA1) (Wang et al. 1997). Several labs have also shown that embyronic stem cells, when allowed to form embryoid bodies in culture can give rise to cells which have similar gene expression profiles to PGCs. These cells, when cultured under the appropriate conditions, can further differentiate to form cells resembling oocytes (Hubner et al. 2003) as well as spermatocytes (Geijsen et al. 2004; Noce 2004). Although the presence of gonadal somatic cell types in the embryoid bodies cannot be excluded, these results also suggest that PGC differentiation can occur outside the environment of the gonad. Our lab has previously addressed the issue of whether or not the gonads secrete an inductive signal by testing the potential of premigratory germ cells to differentiate without exposure to the gonadal environment (Richards et al. 1999b). In these experiments, premigratory 8.5 dpc germ cells were plated in feeder culture. After each day in culture, the cells were assayed for two markers. The first marker, alkaline phosphatase, is expressed by germ cells throughout this time, allowing for PGCs to be counted after being plated on a feeder layer. The second marker, GCNA1, is expressed
47 only in gonadal germ cells in both sexes, and is therefore useful as a marker for cells which have differentiated. After plating 8.5 dpc germ cells and counting the numbers of cells expressing each of these markers, it was found that premigratory cells could differentiate without exposure to the somatic cells of the gonad (Figure 3-1). Interestingly, induction of the GCNA1 marker in culture also mimicked the timing of induction in vivo. Cells plated at 8.5 dpc initiated GCNA1 expression between two and three days in culture which corresponds to 10.5 and 11.5 dpc in vivo, the days which PGCs begin expressing GCNA1. Therefore, an inductive signal is most likely not the cause of differentiation, rather a cell intrinsic timing mechanism may regulate differentiation at this stage of germ cell development (Richards et al. 1999b). Together, with the results from other labs, there is much evidence in favor of a cell autonomous program controlling the events of PGC differentiation in the fetal gonad. To further investigate the nature of the cell intrinsic timing mechanism, we next asked the question: How is timing regulated? Cells can measure time by several mechanisms. 1) The founder population of PGCs could possess an inhibitor of differentiation. Upon subsequent cell divisions this inhibitor would be diluted. Once enough cell divisions have occured, the inhibitor would be diluted below a critical threshold level, allowing for cell differentiation to proceed (Figure 3-2). 2) Alternatively, the premigratory cells could possess an activator of differentiation which would be accumulated during cell divisions. Once this activator has reached a critical concentration, the cell could then initiate differentiation. Both of these timing mechanisms would be dependent on the number of cell divisions which the PGC population has undergone. We therefore hypothesized that
48 changes in cell proliferation would lead to changes in the timing of differentiation if either of these mechanisms existed in premigratory germ cells. To test if either an inhibitor or activator of differentiation may be present in premigratory PGCs, the effect of increased or decreased proliferation on the timing of differentiation was tested on cultured germ cells. Results Although several agents are known to cause increased proliferation of PGCs, only Transforming Growth Factor and activin have been shown to inhibit PGC proliferation (Richards et al. 1999a). Since these ligands are known to promote differentiation of other cell types, an alternative inhibitor of proliferation was sought. In a screen for possible proliferation inhibitors our lab had previously found that the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA; Sigma) inhibited the accumulation of PGCs in culuture. We next examined whether this decreased accumulation was a result of a decreased rate of proliferation. Premigratory 8.5 dpc PGCs were explanted from embryos and plated on a mitotically inactivated feeder layer in either control media, or media containing TPA. After 48 hours of culture, the cells were pulse labeled with 5-bromo-2-deoxyuridine (BrdU) for one hour, followed by staining for alkaline phosphatase and immunodetection for BrdU. To determine whether the observed decrease in PGC accumulation was due to decreased proliferation, the percentage of BrdU postive PGCs was compared between cells cultured in control media versus TPA containing media. While 40% of PGCs were postive for BrdU in control cultures only 20% incorporated the BrdU agent when cultured in the presence of TPA, therefore the decreased accumulation of PGCs is at least partly due to a decreased rate of cell division (Figure 3-3).
49 To determine if a slower rate of cell division would delay differentiation of premigratory germ cells, 8.5 dpc premigratory PGCs were plated onto feeder layers, allowed to adhere overnight and on the following day the media was replaced with either control media or TPA (100ng/l) containing media. PGCs were cultured for another three days at which point the number of alkaline phosphatase and GCNA1 expressing cells were counted. In control cultures the numbers of PGCs increased over four-fold during four days of culture while TPA treated cultures showed less than a two-fold increase. Although we observed decreased numbers of PGCs in TPA treated cultures, the ratio of PGCs expressing the postmigratory GCNA1 marker was consistent for both treated and untreated samples, suggesting that neither an inhibitor or activator of differentiation is present in premigratory germ cells (Table 3-1). Cultures of premigratory PGCs were next exposed to a known potent PGC mitogen, all-trans-retinoic acid (retinoic acid) (Koshimizu et al. 1995). 8.5 dpc PGCs were plated onto feeder layers and on the following day treated with control media or media containing retinoic acid (1M; Sigma). As expected, retinoic acid treated cultures showed a significant increase in PGC numbers after four days of culture. When the proportion of GCNA1 expressing cells was compared between treated and untreated cultures, no significant difference was observed (Figure 3-4). Our lab has previously reported a two to three day delay in the expression of GCNA1 in culture. Interestingly, this delay is also maintained in cultures treated with retinoic acid suggesting that the mechanism regulating the timing of differentiation is independent of cell proliferation. GCNA1 expression was next examined in purified 11.5 dpc PGCs. At this stage, most of the PGCs have colonized the urogenital ridges but are still motile. Since GCNA1
50 expression has initiated at this stage, the fraction of GCNA1 expressing cells is initially higher. When purified 11.5 dpc PGCs were treated with retinoic acid, similar to 8.5 dpc, the ratio of GCNA1 expressing cells remained proportional to control cultures (Figure 3-5). Together, these results suggest that the timing mechanism regulating diffentiation to express GCNA1 cannot be altered by changes in cell proliferation. Conclusions Previously, our lab has found that premigratory PGC differentiation in culture followed a timing pattern which mimicked in vivo differentiation (Richards et al. 1999b). Using the same cell culture system we have now investigated the nature of this timing mechanism, specifically whether the number of cell divisions may control the specific timing of cell differentiation. We found that the phorbal ester TPA significantly reduced PGC accumulation in culture. However, the reduced proliferation did not cause a delay in expression of the postmigratory marker GCNA1, suggesting that the timing of differentiation is independent of cell proliferation. Similarly, treatment of 8.5 and 11.5 dpc PGCs with the mitogen retinoic acid greatly increased PGC accumulation in culture, but no premature differentiation was observed. These results suggest that the timing mechanism regulating differentiation is independent of the number of cells divisions. Also, since differentiation could occur while proliferation was ongoing, it is unlikely that proliferation is followed by differentiation. Rather the data is consistent with proliferation being overlapped by differentiation. In summary, there is most likely no inhibitor of differentiation present in premigratory PGCs, nor an activator of differentiation being accumulated. If either of these scenarios were being used by premigratory PGCs, changes in their proliferative rate would have caused changes in the rate of differentiation.
51 Other interpretations of the data are possible which could lead to an opposite conclusion. We concluded that TPA is an inhibitor of proliferation based on the rate of incorporated BrdU. It is possible that there was indeed no decrease in proliferation, instead TPA may have been cytotoxic, while allowing for the increased proliferation of a subpopulation of cells. In this case the decreased accumulation of cells and decreased BrdU incorporation could be explained by the increased proliferation of a small subset of cells combined with more extensive cell death. Similarly, retinoic acic treatment may have the effect of inducing the synthesis or stability of an inhibitor (the opposite occurring in the case of an activator). In the case of such scenarios the presence of an inhibitor or activator of differentiation could not be exluded. However, we believe that our interpretation of the data is the most parsimonious and the simplest explanation for the observed results. Future experiments could further test our conclusion. It has recently been observed that mice deficient in the enzyme peptidyl-prolyl isomerase (Pin1) have significantly decreased numbers of germ cells leading to defects in fertility. When the cause for the deficient germ cell population was investigated it was found that PGC allocation, migration and colonization of the urogenital ridges occured normally, but the expansion of the cell population was impaired. Although there was no evidence of cell cycle arrest or increased apoptosis, PGCs in these Pin1 deficient embryos were found to have decreased incorporation of BrdU compared to embryos with a functional enzyme suggesting a lengthening of the cell cycle time (Atchison et al. 2003). This mouse model provides an in vivo system to test whether decreased proliferation would have any effect
52 on the rate of differentiation and would provide strong evidence for or against the presence of a timing mechanism dependent on cell proliferation. In conclusion, our results suggest that the cell intrinsic timing mechanism controlling PGC differentiation operates independently of the proliferative state of the cells and is most likely controlled by a stochastic mechanism where each PGC has an equal probability of undergoing differentiation.
53 Table 3-1. Premigratory PGCs cultured in the presence of TPA Sample Alkaline Phosphatase GCNA1 Ratio p value Control day 1 56.4 5.7 6.8 4.5 0.10 0.07 Control day 4 243 16.3 74.4 22.0 0.31 0.10 0.60 TPA day 4 93.4 20.5 32.2 9.3 0.34 0.17
54 Figure 3-1. A cell intrinsic timing mechanism regulates PGC differentiation. (Richards et al. 1999b). 8.5 dpc PGCs were explanted onto feeder layers and cultured for six days. Cultures were subsequently stained for alkaline phosphatase (closed circles) and GCNA1 (open circles). For each time point the average number and s.d. of expressing cells was calculated. Figure 3-2. Effects of altered proliferation on PGC differentiation. (A) The presence of an inhibitor is diluted from PGCs over subsequent cell divisions. Once diluted below a critical threshold differentiation can proceed. (B) When treated with a PGC mitogen cells proliferate faster, hence the inhibitor is diluted faster causing premature differentiation. (C) Decreasing the rate of proliferation causes slower dilution of the inhibitor and delayed differentiation
55 Figure 3-3. TPA inhibits PGC proliferation. PGCs from 8.5 dpc embryos were plated on feeder layers in control media, or media containing 100ng TPA per ml. Forty-eight hours later, the cultures were refed control media supplemented with BrdU labeling mixture for one hour. The percentage and s.d. of PGCs which incorporated BrdU are shown. The values were significantly different by Student's t test at P<0.01. Figure 3-4. Differentiation of 8.5 dpc PGCs in the presence of retinoic acid. PGCs from 8.5 dpc embryos were plated onto feeder layers and the following morning refed either control media (A) or media containing 1uM retinoic acid (RA) (B). Open bars: Mean and s.d. of alkaline phosphatase positive cells per well. Closed bars: Mean and s.d. of GCNA1 positive cells per well. (C) The mean ratio and s.d. of the number of GCNA1 and alkaline phosphatase positive cells in each cell well.
56 Figure 3-5. PGC differentiation in the presence of retinoic acid. PGCs from 11.5 dpc embryos were immunomagnetically purified and plated onto feeder layers in either control media (A) or media containing 1uM retinoic acid (RA) (B). Open bars: Mean and s.d. of alkaline phosphatase positive cells per well. Closed bars: Mean and s.d. of GCNA1 positive cells per well. (C) The mean ratio and s.d. of the number of GCNA1 and alkaline phosphatase positive cells per well.
CHAPTER 4 DNA DEMETHYLATION IS RATE LIMITING FOR PRIMORDIAL GERM CELL DIFFERENTIATION Introduction Chromatin Modifications and the Genome Genetic material is organized in the nucleus by the wrapping of DNA around an octamer of histone proteins. Covalent modifications of DNA, as well as the associated histone proteins, are known to regulate the transcriptional state of genes. These modifications include methylation of DNA or aceylation, methylation, phosphorylation or ubiquitination of histones. Being heritable over many cell divisions, these modifications are referred to as epigenetic, as they are not encoded by the DNA sequence (Li 2002). DNA methylation occurs exclusively at cytosine residues of CpG dinucleotide pairs. CpG pairs are underrepresented in the mammalian genome, being present at a frequency of one CpG dinucleotide every 80 dinucleotide pairs, below the expected frequency of once every 16 dinucleotide pairs. Clusters of CpG dinucleotide pairs occurring at or above their expected frequency, or CpG islands, are occasionally found to be either methylated or unmethylated in the promoters of genes. DNA methylation is generally a characteristic of genes in which transcription is repressed and can be found both on genes of the inactive X chromosome and on the silent allele of imprinted genes. On the other hand, unmethylated CpG islands are associated with genes which are actively transcribed such as housekeeping genes (reviewed in Li, 2002). 57
58 Histone modifications occur along their amino-terminal tails at specific lysine, arginine and serine residues. While arginine residues are methylated, lysine residues can be acetylated or methylated. Histone tails which are acetylated mark regions of active chromatin, while deacetylation is associated with inactive chromatin. Histone methylation is more complex. While certain residues mark active chromatin when methylated, other methylated residues mark inactive chromatin. The combination of these various marks along the length of histone tails is referred to as the histone code, influencing the transcriptional state of the surrounding DNA (Jenuwein and Allis 2001). Transcriptional repression is accomplished by several mechanisms. DNA methylation may block the binding of regulatory factors necessary for transcription, although few examples of this type of mechanism have been reported in vivo. DNA methylation can also repress transcription by the binding of various methyl-CpG-binding proteins (MeCPs or MBDs) as loss of these binding proteins has been show to lead to gene expression in vitro. The association of methyl binding proteins to methylated CpGs has also been found to recruit histone modifying enzymes to form repressive chromatin. For instance both MeCP2 and MBD2 have been shown to associate with histone deacetylase complexes, leading to compaction of chromatin. Therefore, DNA methylation and histone modifications form interactions which lead to chromating remodeling and changes in gene expression (reviewed in Li, 2002). DNA Methylation and Germ Cell Differentiation Using a germ cell culture system, our lab has shown that differentiation of the primordial germ cell population in the fetal gonad occurs through a cell intrinsic mechanism, independent of the presence of the somatic components of the gonads (See
59 Chapter 3). This cell intrinsic program occurs on a specific time schedule which is measured independently of the proliferative state of the cells. In order to further understand the mechanism regulating cell differentiation, primordial germ cells were explanted into feeder culture and exposed to various agents which might affect cell intrinsic programs. Some agents such as aphidocolin and cytochalasin were highly toxic to PGCs, others such as forskolin, cholera toxin, and transforming growth factor effected accumulation of PGCs in culture, but had no effect on the rate of germ cell differentiation. Exposure to urogenital ridge conditioned media from 12.5 and 13.5dpc embryos, also had no effect on the rate of differentiation, precluding the existence of a secreted factor from the gonad which might induce differentiation. However, we did find that the DNA demethylating agent 5-azacytidine was able to accelerate both the rate and extent of PGC differentiation in culture. Although these results do not further our understanding of how the timing mechanism is regulated, they do give us a clue as to how the program of differentiation is executed. Results 5-azacytidine is an analog of cytidine which can incorporate into both DNA and RNA. A similar chemical, 5-aza-2-deoxycytidine incorporates only into DNA (Figure 4-1). Both of these agents cause demethylation by irreversibly binding to the DNA methyltransferase 1 (Dnmt1) enzyme which is responsible for maintenance methylation during DNA replication. Therefore, inhibition of this enzyme causes demethylation in a replication dependent manner. To investigate a role for DNA methylation in PGC differentiation, premigratory 8.5dpc germ cells were plated on mitotically inactivated feeder layers. After 24 hours of culture, the media was replaced with either control media or media containing 10 M
60 5-azacytidine. After four additional days of culture the numbers of alkaline phosphatase and GCNA1 expressing cells were compared. Changes in the ratio of alkaline phosphatase to GCNA1 postive cells were already apparent after three days of culture with fewer numbers of alkaline phosphatase cells present and increased numbers of GCNA1 expressing cells in response to 5-azacytidine treatment. By day five of culture the GCNA1 population greatly increased, while there were very few alkaline phosphatase cells present in azacytidine treated cultures (Figure 4-2). Similarly, primordial germ cells from 10.5dpc embryos cultured in the presence of 5-azacytidine for four days also showed the GCNA1 expressing cells greatly exceeding the numbers of alkaline phosphatase expressing cells. (Figure 4-3). These results suggest that DNA demethylation can greatly accelerate the rate of GCNA1 expression as well as increase the numbers of cells which initiate GCNA1 expression and therefore, demethylation may be rate-limiting for differentiation to express this marker. In the presence of 5-azacytidine we frequently observed the numbers of GCNA1 expressing cells greatly outnumbering the alkaline phosphatase expressing cells. One explanation for this is that under the conditions of 5-azacytidine treatment, somatic cells might initiate GCNA1 expression. Since we never observed GCNA1 positive cells in 5-azacytidine treated feeder layers, we suspected that embryonic somatic cells could be contributing to the significant increase in GCNA1 expressing cells in the 5-azacytidine treated germ cell cultures. To determine if GCNA1 expressing cells in 5-azacytidine treated cultures were indeed primordial germ cells we plated immunomagnetically purified germ cells from and 11.5dpc embryos. Since germ cells plated at this time survive for only two or three
61 days in culture, these experiments were carried out for shorter lengths of time, and the germ cells were plated directly into control media or media containing azacytidine. After two days of culture, exposure to 5-azacytidine again showed a significant increase in GCNA1 expressing cells. Although these results do not exclude the possibility that embryonic somatic cells may contribute to the GCNA1 expressing population under these conditions, these results do suggest that primordial germ cells are at least part of the population of cells which show an increased rate and extent of GCNA1 expression in the presence of the DNA demethylater 5-azacytidine (Figure 4-4). Changes in DNA methylation are often associated with changes in chromatin structure. To determine whether changes in chromatin structure may also play a role in differentiation, primordial germ cells were cultured in the presence of the histone deacetylase inhibitor, trichostatin A. Premigratory 8.5dpc germ cells were plated in feeder culture and the following day media was replaced with either control media or media containing 0.05M trichostatin A. Trichostatin A was found to be toxic to the feeder layer, therefore the germ cells were cultured only for an additional two days. When the number of alkaline phosphatase expressing cells was compared to GCNA1 numbers it was observed that after only three days of culture, the GCNA1 numbers already exceeded alkaline phosphatase numbers (Figure 4-5). Together these results suggest that DNA demethylation, as well as chromatin remodeling can increase both the rate and extent of germ cell differentiation to express GCNA1. In vitro we have used GCNA1 as a marker for differntiation. Since differentiation to express GCNA1 occurs at the same time as other postmigratory differentiation events, it is possible that common regulatory mechanisms control these processes. Therefore,
62 DNA demethylation may be rate limiting for many of the changes which occur in the fetal gonad. Alternatively, DNA demethylation may specifically play a role in GCNA1 expression. To determine if DNA demethylation may be rate limiting for other changes occuring in the gonads at this time, we examined the expression of other postmigratory marker genes in cells after 5-azacytidine treatment. Since the feeder culture system does not easily permit studies on gene expression due to the overwhelming proportion of feeder cells compared to germ cells, a teratocarcinoma cell line was examined for expression of other postmigratory genes. Teratocarcinomas are tumors which are derived from primordial germ cells and related pluripotent cell types, and therefore share some similarities. Since these cells can be grown in culture long term without any need for a feeder layer we first investigated whether treatment with 5-azacytidine would cause an increase in GCNA1 expression as was observed in primordial germ cells. F9 teratocarcinoma cells were grown in presence of control media or media containing 10M 5-azacytidine and after three days of culture examined for GCNA1 expression. While untreated F9 teratocarcinoma cells do express low levels of GCNA1, treatment with a DNA demethylating agent greatly increased the proportion of expessing cells, as did primordial germ cells in culture (Figure 4-6). We next investigated whether other postmigratory marker genes would similarly be upregulated in repsonse to DNA demethylation. We chose to examine several markers with expression limited to postmigratory germ cells of either sex. After three days of culture in the presence of azacytidine, F9 cells were collected and cDNA prepared. Semiquantitave reverse transcriptase polymerase chain reaction (RT-PCR) was then performed to detect expression of three postmigratory genes; mouse vasa homologue
63 (mvh), deleted in azoopermia-like (dazl) and melanoma antigen germline expressed-b4 (mage-b4). Semiquantitative RT-PCRs were performed with minimal amplification cycles to prevent saturation therefore low levels of product were generated. To help visualize these RT-PCR products Southern blotting was performed and hybridized with 32P-dCTP labeled probes for each of the PCR products for signal amplification. Figure 4-7 shows that little or no expression was detected in untreated F9 teratocarcinoma cells, but exposure to 5-azacytidine caused increased expression for all three genes over the course of three days. These results suggest that other postmigratory genes are upregulated after exposure to a DNA demethylating agent, therefore it is likely that DNA demethylation may effect other changes in primordial germ cells and may not be restricted to increasing the rate of GCNA1 expression. Conclusions In an attempt to understand the mechanisms regulating primordial germ cell differentiation in the fetal gonad our lab has used a germ cell culture system which we previously showed could recapitulate the process of germ cell differentiation in vitro. To further our understanding of these regulatory mechanisms, germ cells in culture were treated with various agents which might alter a cell intrinsic program. While most of these agents had no effect on differentiation to express GCNA1, treatment with the DNA demethylater 5-azacytidine, as well as the histone deacetylase inhibitor trichostatin A, caused a significant increase in the rate and extent of differentiation to express this postmigratory marker. In cultures treated with 5-azacytidine or trichostatin A we observed a decrease in the number of cells expressing alkaline phosphatase accompanied by an increase in the number of GCNA1 expressing cells. Several explanations might account for these
64 observations. Although cultures of highly purified 11.5dpc germ cells treated with azacytidine demonstrate that PGCs are at least part of the GCNA1 expressing population, we cannot exclude the possibility that embryonic somatic cells might also initiate GCNA1 expression under these conditions. It is also possible that treatment with azacytidine causes GCNA1 to be a more sensitive probe for germ cells than alkaline phosphatase under these culture conditions. In addition, since 5-azacytidine and trichostatin A are cytotoxic it is possible that GCNA1 expression may be detected in cells which have died, or that toxicity leads to preferential loss of undifferentiated cells. Alternatively, treatment with 5-azacytidine or trichostatin A may accelerate the program of primordial germ cell differentiation. As germ cells downregulate alkaline phosphatase around 14.5dpc in vivo, it is possible that the decreased numbers of cells expressing this marker represents the normal loss of this early germ cell marker. This would allow GCNA1 to be detected in germ cells which no longer expressed the alkaline phosphatase marker. F9 teratocarcinoma cells treated with 5-azacytidine also causes upregulation of several postmigratory markers including GCNA1, mvh, dazl and mage-b4. Together these results suggest that DNA demethylation may effect numerous differentiation events which occur in germ cells at this time. It is interesting that the agent 5-azacytidine has the ability to effect primordial germ cell differentiation. The mechanism of action for this agent involves irreversible binding of the Dnmt1 enzyme to 5-azacytidine as it incorporates into the DNA. As Dnmt1 is responsible for maintaining methylation patterns during replication, this irreversible binding causes a loss of methylation during subsequent cell divisions. Therefore, it seems likely that certain loci are methylated in premigratory germ cells and that loss of
65 methylation at these loci results in gene expression changes which then allow cell differentiation to occur. Although this is a logical assumption, the germ cell lineage was previously thought to be unmethylated during embryogenesis with the exception of genes methylated on the inactive X chromosome, as well as imprinted genes which often contain methylation on the allele which remains silent (these issues will be fully discussed in chapter 5). Since there is no direct data supporting the notion of a germ line devoid of methylation, we proposed that methylation may be present at certain loci during this stage of germ line development, but these loci have simply not been detected due to the difficulty in assessing methylation states on the small number of cells present at this time. We therefore propose a working model where premigratory and migratory germ cells possess methylation blocking the binding of necessary transcription factors and preventing the germ cells from differentiating. After the passage of a predetermined amount of time, a wave of demethlyation is initiated allowing for binding of transcription factors causing activation of postmigratory genes and subsequent differentiation (Figure 4-8). Whether or not the necessary transcription factors are present in premigratory germ cells has not yet been determined, nor has the identity of these factors. An alternative explanation is that the transcription factors themselves are regulated by methylation, therefore demethylation leads to expression of these transcription factors allowing for postmigratory gene activation and primordial germ cell differentiation. The next chapters will address this model and present data in support of a role for DNA demethylation controlling differentiation.
66 Figure 4-1. Chemical structures of cytidine analogs.Both 5-Azacytidine and 5-Aza-2-deoxyctyidine incorporate into DNA. Irreversible binding of the Dnmt1 enzyme to these cytidine analogs causes replication dependent DNA demethylation. Figure 4-2. Differentiation of 8.5dpc PGCs in the presence of 5-azacytidine. PGCs from 8.5dpc embryos were plated on feeder layers and the following morning refed either control media (A) or media containing 10M 5-azacytidine (B). Open bars: Mean and s.d. of alkaline phosphatase positive cells per well. Closed bars: Mean and s.d. of GCNA1 positive cells per well.
67 Figure 4-3. Differentiation of 10.5dpc PGCs in the presence of 5-azacytidine. PGCs from 10.5dpc embryos were plated onto feeder layers in either control media (A) or media containing 10M 5-azacytidine (B). Open bars: Mean and s.d. of alkaline phosphatase positive cells per well. Closed bars: Mean and s.d. of GCNA1 positive cells per well. Figure 4-4. Differentiation of purified 11.5dpc PGCs in the presence of 5-azacytidine. PGCs from 11.5dpc embryos were immunomagnetically purified and plated onto feeder layers in either control media or media containing 10M 5-azacytidine. Open bars: Mean and s.d. of alkaline phosphatase positive cells per well. Closed bars: Mean and s.d. of GCNA1 positive cells per well.
68 Figure 4-5. Differentiation of purified PGCs in the presence of trichostatin A. PGCs from 8.5dpc embryos were plated onto feeder layers and the following morning refed either control media or media containing 0.05M trichostatin A. Open bars: Mean and s.d. of alkaline phosphatase positive cells per well. Closed bars: Mean and s.d. of GCNA1 positive cells per well.
69 A B Figure 4-6. GCNA1 expression in F9 teratocarcinoma cells. (A) F9 cells express a low level of GCNA1 expression. (B) GCNA1 is upregulated in cultures of F9 cells treated with 10M 5-azacytidine. Figure 4-7. Postmigratory gene expession in 5-azacytidine treated F9 teratocarcinoma cells. RT-PCR samples for dazh, mage-b4 and mvh were run on an agarose gel, Southern blotted and hybridized with radiolabeled cDNA probes. Expression of these genes is upregulated after three days of 5-azacytidine exposure. Testis cDNA was used as a postive control for PCR amplification. HPRT was used as loading control.
70 Figure 4-8. Model for Primordial Germ Cell Differentiation. Migratory PGCs possess methylation and repressive chromatin structure preventing expression of postmigratory germ cell genes. Whether or not the necessary transcription factors are present at this time is not known, however if present, methylation may block their ability to bind target sequences. After a predetermined amount of time, a wave of demethylation occurs accompanied by chromatin decondensation. Theses changes allow for the binding of transcription factors which are now present leading to expression of postmigratory germ cell genes.
CHAPTER 5 EXPRESSION OF POSTMIGRATORY GERM CELL GENES CORRELATES WITH LOSS OF PROMOTER METHYLATION Introduction DNA Methylation and Embryonic Development During embryogenesis there are dynamic changes in the methylation state of the genome. While the sperm and egg contain high levels of methylation, following fertilization there is a wave of genome-wide demethylation. Both genomes undergo demethylation at this time, however the paternally and maternally derived genomes use separate mechanisms to demethylate their DNA (Kafri et al. 1993). The paternal genome undergoes a very rapid loss of methylation between three to six hours postfertilization, prior to pronuclear fusion. Since this demethylation occurs in a very short amount of time and is independent of DNA replication it has been concluded that this demethylation is mediated by an active demethylating enzyme, although no such enzyme has yet been identified. Alternatively, the maternal genome undergoes passive demethylation in which the lack of a methyltransferase enzyme results in replication dependent loss of methylation (Mayer et al. 2000). By the blastocyst stage of development, almost all methylation of both genomes has been lost with the exception of genes located on the inactive X chromosome and the silenced alleles of imprinted genes (Kafri et al. 1993; Mayer et al. 2000). Between the stages of implantation and gastrulation de novo methylation of the genome is initiated thereby establishing the somatic methylation patterns which are 71
72 maintained throughout the life of an organism. During somatic cell differentiation certain genes will undergo another round of demethylation allowing for tissue specific gene expression patterns (Jaenisch 1997). Remethylation of the genome is carried out by at least two enzymes, Dnmt3a and Dnmt3b, which have overlapping functions leading to the establishment of somatic methylation patterns (Okano et al. 1999). During this time the extraembryonic lineages maintain a significantly lower level of methylation than the embryonic tissues of the embryo (Chapman et al. 1984; Monk et al. 1987; Rossant et al. 1986). DNA Methylation and Germ Cell Development After migration to the urogenital ridges primordial germ cells undergo a wave of differentiation. During this time germ cells mediate two epigenetic changes; erasure of genomic imprints and reactivation of the silent X chromosome. These two phenomena have been the focus of the majority of studies on the methylation status of germ cells at this stage of their development. Genomic imprinting Genomic imprinting is the mechanism by which certain genes are expressed monoallelically depending on their parent of origin. Within their promoter regions many imprinted genes contain CpG islands which are methylated on the silenced allele and unmethylated on the active allele. These genomic imprinting marks are established during gametogenesis and retained after fertilization as the genomic reprogramming event which erases the majority of genomic methylation avoids imprinted loci (reviewed in Jaenisch, 1997). In order for these marks to be passed on from generation to generation, males and females must reset the imprinting mark such that a male will pass on only a paternally marked genome, and a female will pass on only a maternally marked
73 genome. This resetting process is carried out partially within primordial germ cells as marks are erased such as differential methylation leaving each allele indistinguishable and coincidentally allowing for biallelic expression (Szabo et al. 2002; Szabo and Mann 1995). Subsequently, males will re-establish the mark by placing a paternal methylation pattern on both alleles during the fetal stages of spermatogenesis, while females will replace imprints with those indicating a maternal origin during the postnatal stages of oogenesis (Reviewed in McLaren, 2003). Although there was no information regarding the methylation state of the germline during the premigratory and migratory stages, it is understood that imprinted genes do not undergo genome wide demethylation after fertilization. Therefore primordial germ cells are likely to possess monoallelically expressed genes at imprinted loci. Experiments assessing the allelic expression during gametogenesis established that embryos at 6.5dpc exhibit monoallelic expression of imprinted genes, but by 11.5dpc when primordial germ cells have entered the urogenital ridges, imprinted genes are biallelically expressed (Szabo and Mann 1995). Therefore, it is likely that imprinted genes retain their monoallelic expression patterns in early germ cells, but shortly after entry into the developing gonads, differential marks are lost and imprinted genes are biallelically expressed. X chromosome reactivation X chromosome inactivation is the mechanism used by mammals to equalize the amount of gene expression from the X chromosome (reviewed in Monk, 1992). Males, chromosomally XY, have half the dosage of X chromosome gene products than females, XX. Therefore, early in embryonic development females will randomly inactivate one of their X chromosomes to achieve dosage compensation. The inactive X chromosome
74 contains CpG methylation as well as repressive chromatin structure which is stable throughout subsequent cell divisions. However, the inactive X must be reactivated to ensure that each developing oocyte contains an active X chromosome. This process is carried out within primordial germ cells as they undergo their differentiation within the fetal gonad (Monk 1992). While the mechanisms controlling this reactivation process are unknown, experimental evidence has shown that X reactivation is independent of the sex of the gonad, as XXY germ cells in the testis also reactivate the silent X (Mroz et al. 1999). Reactivation in XX germ cells occurs by 13.5dpc, but germ cells which remain outside of the gonads in ectopic locations do not reactivate the inactive X, suggesting that the gonadal environment plays a role in the reactivation process (Tam et al. 1994). Epigenetic Changes During Germ Cell Differentiation Our lab has proposed a model for primordial germ cell differentiation in which a wave of demethylation causes changes in gene expression and ultimately cell differentiation. Due to the technical difficulty in obtaining pure populations of primordial germ cells prior to entry into the developing gonads, little is known about the methylation state of genes within migratory germ cells. Experiments which analyzed the methylation of bulk genomic DNA have shown that following fertilization there is very little methylation present in the genome and that the blastocyst stage represents the lowest state of methylation levels during development. Subsequently the genome undergoes de novo methylation as methylation levels in the epiblast begin to increase until reaching the levels found in somatic cells while the extraembryonic tissues remain hypomethylated. Studies on primordial germ cells have shown that at 12.5 and 14.5dpc methylation levels are very low, similar to that seen in blastocysts as well as extraembryonic tissues (Monk et al. 1987). Since primordial germ
75 cells are moved outside of the embryo during the time when de novo methylation is initiated, it was postulated that this isolation from the somatic cells of the embryo allowed germ cells to escape de novo methylation, and to remain hypomethylated throughout this time (Figure 5-1) (Jaenisch 1997). The concept of a germ line protected from waves of methylation and differentiating signals within the epiblast is inconsistent with our model of a germ line which is methylated early and subsequently undergoes a wave of demethylation prior to the differentiation events which occur within the fetal gonad. Since there was no data on the methylation state of migratory germ cells, and our model is based on the presence of methylation, therefore we next investigated whether genes expressed exclusively in postmigratory germ cells (similar to GCNA1) contain CpG islands and if so, are these CpG islands methylated in migratory germ cells and unmethylated in postmigratory germ cells. Because the gene associated with the germ cell marker GCNA1 is unknown, we chose to examine genes with a similar expression pattern to GCNA1. Other genes which are also exclusively expressed in primordial germ cells include mouse vasa homologue (Mvh), synaptonemal complex protein 3 (Scp3) and deleted in azoospermia-like (Dazl) (Cooke et al. 1996; Di Carlo et al. 2000; Toyooka 2000). The methylation state of a gene expressed in the opposite pattern, i.e. becoming downregulated following gonadal colonization, was also investigated to determine if methylation plays a role in its regulation; tissue non-specific alkaline phosphatase (Tnap) (Hahnel et al. 1990). Results To investigate the methylation state of postmigratory germ cell genes, genomic sequence was first obtained and screened for the presence of CpG islands. At the time this project was initiated, complete sequencing of the mouse genome had not been
76 completed. Therefore, Bacterial Artificial Clone (BAC) libraries were screened for clones containing genomic loci for Mvh, Scp3 and Dazl using probes generated from testis cDNA. Due to the large size of BACs (often over 200kb), smaller genomic fragments were isolated from these clones and used to generate sequence of the region upstream of the first exon. The sequence was then screened for CpG islands, defined as having a GC content over 50% and a ratio of observed to expected CpG dinucleotides equal or greater than 0.6 (Gardiner-Garden and Frommer 1987). All three genes contained CpG islands upstream or overlapping their transcriptional start sites (Figure 5-2). There are several methods to analyze the methylation state of a gene. As methods based on methylation sensitive restriction sites are limited to screening a select few CpG sites and often require large amounts of DNA, a PCR (polymerase chain reaction) based method was chosen to overcome these limitations. The bisulphite sequencing method (Clark et al. 1994)involves the chemical conversion of genomic DNA, such that all unmethylated cytosine residues are converted to uracils. A cytosine which is methylated does not permit the chemical reaction to occur and will remain a cytosine. The region containing the CpG island is then amplified by PCR, during which uracils are amplified as thymines, the PCR product is then cloned and sequenced. This technique allows for the simultaneous analysis of many CpG residues from a few hundred cells. Genomic DNA was isolated from immunomagnetically purified primordial germ cells isolated from 10.5 and 13.5 dpc embryos and then subjected to bisulphite conversion. Regions containing CpG islands were then amplified with primers specific to bisulphite converted DNA (Table A-1). For each gene analyzed, purified 10.5dpc
77 primordial germ cells showed high levels of methylation throughout the regions analyzed. This contradicts the notion that migratory germ cells are devoid of methylation outside of imprinted genes and genes on the inactive X chromosome (Jaenisch 1997). When the methylation status of purified 13.5dpc primordial germ cells was compared, there was a significant loss of methylation for all three genes, with many clones showing a complete loss of methylation at these loci (Figures 5-3, 5-4 and 5-5). The methylation status of somatic cell types was also investigated. For each of the three genes, somatic cells from the gonad and mesonephros always showed high levels of methylation at both 10.5 and 13.5dpc (Immunodepleted fractions; Figures 5-3, 5-4 and 5-5). Somatic cells from heart, kidney, liver and lung tissues also showed high levels of methylation at 13.5dpc for each gene (data not shown). These results suggest that a wave of demethylation specific to the germ line occurs between 10.5 dpc and 13.5 dpc, and may be involved in allowing for expression of postmigratory germ cell genes. We next investigated whether genes which are downregulated during this time would exhibit changes in methylation. For the Tnap gene, purified 10.5 and 13.5dpc primordial germ cells showed significant methylation. There was a slight increase in the incidence at methylation in 13.5dpc germ cells from female embryos. If there were to be increased methylation due to gene silencing, 13.5 dpc may be too early to detect it as Tnap does not become downregulated until 14.5 to 15.5 dpc(Hahnel et al. 1990). Therefore germ cells were purified at 14.5dpc and analyzed for methylation at the Tnap locus. No significant change in methylation was detected at 14.5d pc compared to 10.5 dpc, therefore suggesting that changes in DNA methylation in this region of the Tnap locus do not correlate with changes in gene expression (Figure 5-6).
78 Conclusions Treatment of premigratory germ cells in culture with the DNA demethylating agent 5-azacytidine causes an increase in the rate and extent of germ cell differentiation to express the postmigratory marker GCNA1 (Chapter 4). Although primordial germ cells were initially presumed to be unmethylated during this time, it seemed unlikely that 5-azacytidine could elicit such changes if there was no methylation present. Since methylation is present at imprinted loci, as well as regions of the inactive X chromosome, the proteins required to maintain methylation during this time should also be present. Therefore, it is possible that other regions may be methylated in germ cells at this time. Primordial germ cell differentiation in the fetal gonad is accompanied by changes in the expression of many genes. DNA demethylation was able to increase the expression of the postmigratory marker GCNA1 in cultured primordial germ cells. GCNA1 and other postmigratory genes such as Mvh, Dazl and Mage-b4 showed increased expression when F9 teratocarcinoma cells were similarly treated with a demethylating agent. Therefore, it is likely that regulation of postmigratory gene expression may be controlled by a DNA demethylation event. To investigate whether initiation of postmigratory gene expression is accompanied by a loss of methylation, bisulphite sequence analysis was performed on immunomagnetically purified 10.5 and 13.5 dpc primordial germ cells. CpG islands were identified for Mvh, Dazl and Scp3 and were subsequently analyzed for changes in methylation during this stage of development. Primordial germ cells purified from 10.5dpc embryos showed high levels of methylation for Mvh, Scp3 and Dazl. Clones which completely lacked methylation were
79 not detected at this time. Somatic cells from the immunodepleted fraction of the germ cell purification also showed high levels of methylation similar to 10. 5dpc germ cells. By 13.5 dpc, there was a significant decrease in methylation with many clones being completely unmethylated. There are several explanations for the presence of methylated clones at this time. Since there are somatic cells present in these populations the methylation may represent contaminating somatic cells. Alternatively, it is possible that primordial germ cells do not initiate differentiation at the same time, but that some may begin before others. It is known that although GCNA1 expression begins at 11.5 dpc, some germ cells will initiate expression of these markers later than others (Enders and May 1994). Therefore, it is possible that some of the methylated clones represent germ cells which differentiate or demethylate on a slower time schedule. This is also supported by the presence of clones which are partially methylated, possibly representing germ cells in the process of undergoing demethylation. It is likely that both somatic cells and late demethylating germ cells are contributing to this population of methylated alleles at 13.5 dpc. The presence of testis cords in the fetal testis allows for male and female embryos to be distinguished at 13.5 dpc. Bisulphite analysis of male and female germ cells from 13.5 dpc embryos did not show any significant differences in the rate or extent of demethylation for any of the genes analyzed. This is not surprising since the demethylation event occurs at a time when male and female germ cells are essentially identical. Following germ cell colonization of the gonads, Mvh, Dazl and Scp3 undergo a demethylation event which correlates with transcription initiation. We next investigated
80 if changes in DNA methylation might be associated with a gene which is downregulated at this time, Tnap. Bisulfite analysis of the Tnap gene showed no change in methylation between 10.5 and 13.5 dpc suggesting that changes in methylation do not control expression of this gene. This contradicts earlier findings which suggested that Tnap expression is regulated by DNA methylation as the gene was found to be methylated in non-expressing cells (brain and STO fibroblasts) and unmethylated in expressing cells (placenta and embryonic stem cells) (Escalante-Alcalde et al. 1996). The work by Escalante-Alcalde analyzed methylation status by digesting genomic DNA with methylation specific transcription factors followed by Southern blotting. Although their analysis was restricted to analyzing a few Hpa II sites, their anaylsis covered a larger region than the region investigated in our study. Therefore it is possible that CpG sites outside of the region covered by bisulphite sequencing may be methylated in non-expressing cell types. Further investigation into the methylation of surrounding CpG sites might shed light on whether methylation plays a role in silencing of the Tnap gene. Several lines of evidence suggest that demethylation of gonadal germ cells is an active process, possibly mediated by a demethylating enzyme. Passive demethylation of DNA is caused by failure of the Dnmt1 enzyme to methylate newly replicated strands during DNA synthesis. This results in one strand which retains methylation, and a newly synthesized strand devoid of methylation. If such a mechanism was occurring in primordial germ cells, bisulphite analysis should reveal clones which are highly methylated and clones which completely lack methylation. While we observed both types of clones in 13.5 dpc germ cell preparations, we also observed clones which were partially methylated, having a pattern distinctly less methylated than that of somatic cells
81 and of 10.5 dpc germ cells, but not yet completely unmethylated. This is inconsistent with passive demethylation and suggests that demethylation is an active process. Other studies also suggest an active demethylation process as the demethylation of certain imprinted loci occurs very rapidly, sometimes within 24 hours (about 1.5 cell divisions). This rapid loss of methylation is also inconsistent with passive demethylation (Hajkova et al. 2002; Lee et al. 2002). Lastly, culture of primordial germ cells in the presence of the germ cell mitogen retinoic acid did not affect the rate of germ cell differentiation, suggesting that differentiation occurs independently of cell divisions . If differentiation to express GCNA1 in culture is mediated by a demethylation event, the demethylation event is not likely to be replication dependent. It should also be noted that the Dnmt1 enzyme is present at very high levels in the nucleus of 12.5 dpc germ cells (Hajkova et al. 2002). This is in contrast to the situation in the one-cell embryo during which there is no nuclear methyltransferase present, and passive demethylation of the maternal genome occurs (Howell et al. 2001; Kafri et al. 1993; Mayer et al. 2000). The presence of methylation in 10.5 dpc germ cells contradicts the notion of a germ line devoid of methylation (Figure 5-7). Several labs have recently reported data pertaining to the methylation state of germ cells at this time. Bisulphite analysis of 10.5 through 13.5 dpc purified germ cells confirmed that differential methylation of imprinted genes is maintained in migratory germ cells, and that imprinting marks are erased between 10.5 and 12.5 dpc (Hajkova et al. 2002; Lee et al. 2002). Nuclear transplantation experiments and investigations of RNA expression from imprinted loci have also confirmed that reprogramming of imprinted loci occurs as germ cells enter the developing gonads (Lee et al. 2002; Szabo et al. 2002). In addition to the postmigratory
82 germ cell genes which have been analyzed, several other nonimprinted genes have been investigated; alpha actin, mylC, and Xist also show demethylation during this time (Hajkova et al. 2002). In addition, repetitive elements, which are the site of the majority of genomic methylation, also show demethylation as IAP and Line1 repeat elements are also demethylated during this time (Hajkova et al. 2002). Together, these results indicate that there is a genome wide demethylation event occurring in primordial germ cells shortly after entry into the urogenital ridges which extends beyond erasure at imprinted loci and regions of the reactivating X chromosome. Although the observed demethylation event correlates with the timing of postmigratory gene expression, this does not prove that demethylation leads to the expression of these genes. Outside of genes which exhibit monoallelic expression there is no strong correlation between DNA methylation and transcriptional repression. As Bestor (1999) pointed out, in order for this correlation to be made, a methylation pattern seen in a non-expressing cell must be shown to be repressive in a cell which is normally expressing (and presumably unmethylated). Since most studies correlating methylation with transcription fail to do this, the conclusion that DNA methylation leads to transcriptional silencing (and vice versa) cannot be made (Walsh 1999). Further studies are necessary to investigate whether or not this wave of demethylation in primordial germ cells is responsible for expression of postmigratory genes, as well as germ cell differentiation within the fetal gonad.
83 Figure 5-1. Methylation dynamics during embryogenesis. Spermatocytes initially exhibit higher methylation levels than oocytes. Following fertilization, maternal DNA undergoes a transient wave of de novo methylation. Prior to pronuclear fusion, paternal DNA undergoes a very rapid demethylation carried out by an active demethylase enzyme. The maternal genome undergo replication-dependent passive methylation. By the blastocyst stage there is little methylation of the genome outside of imprinted regions and genes subject to X inactivation. Between implantation and gastrulation somatic lineages undergo de novo methylation while the germ line is sequestered in the extraembryonic tissues which remain hypomethylated. While PGCs are known to be hypomethylated by 12.5dpc, it is assumed that PGCs avoid de novo methylation due to their extraembryonic location (adapted from Jaenisch 1997).
84 A B C D Figure 5-2. Germ cell specific genes contain CpG islands. Based on the density of CpG dinucleotides, CpG islands were identified for Mvh (a), Scp3 (b), Dazl (c) and Tnap (d). The region analyzed indicates the region analyzed for methylation by bisulfite sequencing.
85 Figure 5-3. Bisulfite Analysis of mouse vasa homologue (mvh). Bisulfite sequencing analysis was performed on 10.5 and 13.5 dpc purified primordial germ cells and somatic cells from the immunodepleted fraction. At 13.5 dpc male and female germ cells were purified separately based on the presence of testis cords in the male. Each line represents one clone which was sequenced. Open circles are unmethylated CpG dinucleotides. Closed circles are methylated dinucleotides. Numbers in parenthesis indicate the number of clones with the observed methylation pattern.
86 Figure 5-4. Bisulfite analysis of synaptonemal complex protein 3 (scp3). Bisulfite sequencing analysis was performed on 10.5 and 13.5 dpc purified primordial germ cells and somatic cells from the immunodepleted fraction. At 13.5 dpc male and female germ cells were purified separately based on the presence of testis cords in the male. Each line represents one clone which was sequenced. Open circles are unmethylated CpG dinucleotides. Closed circles are methylated dinucleotides. Numbers in parenthesis indicate the number of clones with the observed methylation pattern.
87 Figure 5-5. Bisulfite analysis of Deleted in Azoospermia-Like (dazl). Bisulfite sequencing analysis was performed on 10.5 and 13.5 dpc purified primordial germ cells and somatic cells from the immunodepleted fraction. At 13.5 dpc male and female germ cells were purified separately based on the presence of testis cords in the male, however data was combined due to low number of clones sequenced. Each line represents one clone which was sequenced. Open circles are unmethylated CpG dinucleotides. Closed circles are methylated dinucleotides. Numbers in parenthesis indicate the number of clones with the observed methylation pattern.
88 Figure 5-6. Bisulfite analysis of tissue non-specific alkaline phosphatase (Tnap). Bisulfite sequencing analysis was performed on 10.5, 13.5 and 14.5 dpc purified primordial germ cells and somatic cells from the immunodepleted fraction. At 13.5 and 14.5 dpc male and female germ cells were purified separately based on the presence of testis cords in the male. Each line represents one clone which was sequenced. Open circles are unmethylated CpG dinucleotides. Closed circles are methylated dinucleotides. Numbers in parenthesis indicate the number of clones with the observed methylation pattern.
89 Figure 5-7. Methylation of the germ cell lineage. See figure 5-1 for full description. Previously it was thought that the germ cell lineage escaped de novo methylation by moving outside of the embryo into the extraembryonic mesoderm. It is now clear that primordial germ cells undergo a wave of demethylation after entry into the developing gonads, although the timing of when they acquire their methylation is still unknown. It is possible that methylation is acquired at the same time as somatic cells undergo de novo methylation, alternatively, methylation may be acquired at a later stage of their development (adapted from Jaenisch 1997).
CHAPTER 6 DNA METHYLATION REGULATES POSTMIGRATORY GERM CELL GENE EXPRESSION Introduction DNA Methylation and Control of Gene Expression The methyltransferase enzyme Dnmt1 is the main enzyme responsible for maintenance methylation. Having a high affinity for hemimethylated DNA, the Dnmt1 enzyme replaces methylation on the newly synthesized strands of DNA following replication. In mouse embryos, the full length somatic form of the protein begins being expressed throughout the embryo by 7.5 dpc, replacing the maternally stored oocyte specific form of the protein which is present during preimplantation development (Mertineit et al. 1998; Trasler et al. 1996). Embryos which lack a functional Dnmt1 enzyme show significantly decreased cytosine methylation causing developmental delay and embryonic lethality by 10.5 dpc. This decrease in genomic methylation levels results in activation of the randomly inactivated X chromosome (Sado et al. 2000), loss of differential expression patterns at several imprinted loci (Li et al. 1993) and activation of intracisternal A particle (IAP) retroviruses (Walsh et al. 1998). However, promiscuous activation of transcriptionally silenced genes has not been observed (Walsh 1999). Transcriptional silencing is mediated by the interaction of DNA methylation with methyl-CpG-binding proteins (MeCPs or MBDs). By recruiting histone modifying enzymes, these proteins mediate changes in chromatin structure and gene expression. 90
91 Association of these proteins with DNA methylation is repressive as loss of these binding proteins has been show to lead to transcriptional activation in vitro (Reviewed in Li 2003). In humans, mutation of MeCP2 has been linked to 80% of Rett syndrome patients (Amir et al. 1999). A mouse model for the disease lacking a functional MeCP2 gene has similar phenotypes to human disease patients, displaying severe neurological disorders (Chen et al. 2001; Guy et al. 2001; Shahbazian et al. 2002). Although the disease phenotype was presumed to be caused by aberrant gene expression, microarray analysis comparing normal versus diseased tissues from the mouse model have found no significant changes in gene expression, suggesting either a secondary role independent of gene silencing for the protein, or compensation by other methyl binding proteins. There have been many reports of genes which are regulated by changes in methylation based on in vitro studies and/or simple observations of methylation patterns in expressing and nonexpressing tissues. However these instances fail to demonstrate that methylation is the cause of changes in gene expression (Futscher et al. 2002; Kudo and Fukuda 1995; Salvatore et al. 1998; Walsh 1999). While decreased levels of methylation commonly cause gene activation in vitro, mouse embryos with severely decreased methylation levels due to Dnmt1 mutation and human patients with MeCP2 mutations do exhibit changes in gene expression. DNA Methylation and Regulation of Postmigratory Gene Expression Primordial germ cells undergo a wave of differentiation shortly after entry into the developing gonads. While investigating the mechanisms regulating these differentiation events our lab has found that in vitro, germ cell differentiation can be accelerated by treatment with DNA demethylating agents. Furthermore, several postmigratory genes undergo a demethylation event shortly after entry into the urogenital ridges. While it is
92 clear that an epigenetic reprogramming event mediates erasure of genomic imprints and reactivation of the inactive X chromosome, it is possible that other changes which occur at this time are also linked to this demethylation event. The demethylation of postmigratory germ cell genes suggests a role for an epigenetic mechanism regulating these changes in gene expression. However, simply observing promoter methylation while a gene is silenced does not infer that the methylation is the cause of the silencing. To further investigate the role of a genome wide demethylation event in the onset of gene expression in primordial germ cells, the expression of postmigratory germ cell genes was examined in the context of embryos deficient in methyltransferase activity. Results To investigate the potential role for DNA methylation in the silencing of postmigratory germ cell genes, expression of GCNA1 was examined in embryos carrying mutant Dnmt1 alleles. GCNA1 expression is normally initiated between 10.5 and 11.5 dpc (Enders and May 1994). As expected, immunohistochemical staining of 8.5 dpc embryos homozygous for the wildtype Dnmt1 allele (+/+) showed no detectable expression of GCNA1. However, 8.5 dpc embryos homozygous for the mutant allele (n/n) showed ectopic expression of the postmigratory germ cell marker throughout the embryo (Figure 6-1). While ectopic expression is detected in the anterior most tissues of the embryo, including the neural folds, there is a higher concentration of positive cells at the posterior portion of the embryo. Although the presence of germ cells has not been verified in these embryos, these results indicate that DNA methylation is likely the primary silencing mechanism for the postmigratory marker GCNA1 in numerous somatic cell lineages and possibly premigratory germ cells.
93 Conclusions Here we report the first known example of a gene which is ectopically expressed in response to reduced methylation levels in vivo. While genes which are monoallelically silenced show biallelic expression (as opposed to ectopic expression) in the Dnmt1 mutant (Li et al. 1993), there have previously been no reports of gene expression being initiated in nonexpressing tissues. Two recent studies have proposed a model where expression of brain-derived neurotropic factor (BDNF) is controlled by changes in promoter methylation and binding of Mecp2 (Chen et al. 2003; Martinowich et al. 2003). Examination of embryos which specifically lack Dnmt1 in the central nervous system showed increased expression of BDNF in brain tissue, however no data was presented on expression in any normally nonexpressing tissues (Martinowich et al. 2003). There are several explanations to account for the ectopic expression of the postmigratory germ cell marker GCNA1 in the somatic cells 8.5 dpc embryos. First, while the gene associated with the GCNA1 antigen is unknown, it is possible that GCNA1 may be an imprinted gene which is only expressed in the germ line. While this would explain the sensitivity to DNA demethylation, the presence of the protein in both oocytes and spermatocytes suggests that it is not imprinted. Second, the GCNA1 positive cells may be aberrantly migrating germ cells. This would not account for the temporally ectopic expression, although it could account for presence of individual GCNA1 expressing cells throughout the embryo. The germ cell population normally arises at the most posterior portion of the embryo, at the base of the allantois. While the population is migratory at 8.5 dpc, it seems unlikely that a germ cell could migrate across the length of an embryo within one day of being specified (migration to the urogenital ridge normally
94 requiring between two to three days). Since Dnmt1 expression is not initiated until 7.5 dpc, there is not likely to be a defect in germ cell specification. A third explanation suggests that GCNA1 itself is not directly regulated by methylation, but the transcription factors which allow for GCNA1 expression are regulated by a demethylation event. This would suggest that transcription factors are not present in cells which do not express GCNA1 such as premigratory and migratory germ cells as well as somatic cells. Demethylation of the genes encoding these factors would allow for synthesis of these transcription factors and subsequent gene activation. Lastly, the transcription factors required for expression of GCNA1 may be present in somatic cells and germ cells but methylation of the GCNA1 promoter causes the gene to be silenced. Demethylation caused by a mutation of a methyltransferase gene, exposure to demethylating agents such as 5-azacytidine (see chapter 4) or a naturally occurring wave of genome wide demethylation would then lead to expression of the GCNA1 gene. Several important questions are raised by the observed ectopic expression in the Dnmt1 mutant embryo. It is interesting that expression of GCNA1 in the mutant embryo is restricted to individual cells throughout the embryo. If the loss of demethylation causes GCNA1 to be expressed, why is expression not seen in every cell? One possible explanation is the nature of the Dnmt1 mutation. The original mutation (termed Dnmtn for N-terminal disruption) was generated by deletion of a part of the first coding exon which allowed for synthesis of a truncated protein (Li et al. 1992). While methylation levels are decreased to approximately 30% to that of normal levels, it is possible that cells which initiate GCNA1 expression in this mutant represent cells having lower enzyme levels than others. If this reasoning were correct, then a more severe mutant should show
95 increased expression of GCNA1. Investigation into GCNA1 expression in a mouse line harboring a null mutation (Dnmtc, for catalytic mutation) could answer this question (Lei et al. 1996). While GCNA1 is widely expressed in the Dnmt1 mutant, the presence of germ cells has not been confirmed. No group of cells was observed located at the base of the allantois, although the germ cells may have begun migrating and making them difficult to identify amongst the other GCNA1 expressing cells. Staining with early germ cell markers such as alkaline phosphatase and stage specific antigen 1 (SSEA1) could assist with identifying a germ cell population in these embryos. Several postmigratory germ cell specific genes undergo a wave of demethylation shortly after entry into the developing gonads. This loss of methylation temporally correlates with the onset of transcription for these genes. While these genes have a similar expression pattern to GCNA1, it has been proposed that demethylation also is required for expression of the postmigratory genes including Mvh, Scp3 and Dazl. Analysis of the expression of these genes in Dnmt1 mutant embryos will provide evidence for or against a role for demethylation in postmigratory gene expression. These future studies will further elucidate the role of a wave of global demethylation in the process of primordial germ cell differentiation.
96 Figure 6-1. GCNA1 expression in a Dnmt1 mutant embryo. GC NA1 is not expressed in an 8.5 dpc Dnmt1+/+ embryo (bottom). A mutant Dnmt1n/n littermate (top) shows extensive ectopic expression of GCNA1 in numerous somatic cell lineages.
CHAPTER 7 DISCUSSION Primordial germ cells are specified in the extraembryonic mesoderm and subsequently migrate through the hindgut endoderm, colonizing the urogenital ridges between 10.5 and 11.5 dpc. Shortly after arriving at the gonads, male and female germ cells undergo a common set of differentiation events independent of the changes associated with sex determination. These sex common differentiation events include cessation of migration and proliferation (Donovan et al. 1986), changes in cell adhesion (Garcia-Castro et al. 1997), decreased ability to form pluripotent stem cells(Matsui et al. 1992; Resnick et al. 1992), erasure of genomic imprinting marks (Hajkova et al. 2002; Lee et al. 2002; Szabo et al. 2002), and changes in gene expression (reviewed in McLaren 2003). The focus of our lab is to understand the mechanisms regulating these events. Understanding the Cell Intrinsic Timing Mechanism Previously our lab has shown that by culturing premigratory germ cells removed from the gonadal environment, primordial germ cells can differentiate to express the postmigratory marker GCNA1 on a timing pattern similar to germ cells in vivo (Richards et al. 1999b). Data further investigating this timing mechanism has been presented here (Chapter 3). Premigratory primordial germ cells exposed to agents which increased or decreased the rate of cell proliferation had no effect on the rate of differentiation to express the postmigratory marker GCNA1. These results argue that mechanisms, such as the dilution of a differentiation inhibitor, are most likely not present in premigratory germ cells. Patterns of primordial germ cell differentiation are consistent with a stochastic 97
98 mechanism in which cells differentiate at a specific time, each individual cell having an equal probability of differentiating at any given time. As these conclusions are based on in vitro experiments, it would be interesting to address the issue of timing in vivo. A recently described mouse model carrying a mutation for the peptidyl-prolyl isomerase (Pin1) gene causes a defect in primordial germ cell proliferation by decreasing the time it takes to complete each cell cycle (Atchison et al. 2003). As germ cells in the mutant divide slower than normal, this mouse model would provide an in vivo system to further test the hypothesis that germ cell differentiation occurs independently of cell proliferation. In an attempt to further understand the cell intrinsic timing mechanism, we next attempted to alter the rate of differentiation in culture and found that the DNA demethylating agent 5-azacytidine significantly increased the numbers of cells expressing GCNA1 and increased the rate at which these cells appeared in culture (Chapter 4). This suggests that DNA demethylation is rate limiting for germ cell differentiation. From these results we have presented a model in which premigratory germ cells possess DNA methylation at critical loci preventing differentiation from occurring. A wave of demethylation then occurs, removing these critical methylations and allowing for germ cell differentiation to occur. Although this hypothesis addresses the mechanism which causes differentiation, it does not further address how the timing mechanism is controlled. It seems likely that if a demethylation event controls differentiation, the timing mechanism should trigger this demethylation event to occur. Unfortunately, the mechanisms mediating DNA demethylation, specifically active demethylation, are not understood and no demethylating enzyme has of yet been identified. Therefore until the
99 mechanics of DNA demethylation are understood, it will be difficult to understand what initiates the wave of demethylation on a specific time schedule. DNA Methylation of Primordial Germ Cells To further characterize the role of DNA demethylation in primordial germ cell differentiation, we investigated the methylation state of genes which are upregulated following migration to the developing gonads. The model presented in Chapter 4 proposed the presence of methylation in premigratory germ cells. However it was originally theorized that the germ line was unmethylated throughout its development. Much recent data contradicts this original theory. In Chapter 5, bisulphite sequencing data for several postmigratory germ cell genes show a loss of methylation between 10.5 and 13.5 dpc. Other labs have also shown that repetetive elements, which make up 90% of the methylated genome, are highly methylated in migratory germ cells. Imprinted loci also contain monoallelic methylation during this stage of germ cell development (Hajkova et al. 2002; Lee et al. 2002). It seems clear now that primordial germ cells are highly methylated and undergo a wave of demethylation shortly after entry into the gonads. Several epigenetic processes occur within primordial germ cells following gonadal colonization including erasure of genomic imprintsand reactivation of the inactive X chromosome in females. Methylation plays a critical role in maintaining monoallelic gene silencing as embryos lacking a functional Dnmt1 gene show biallelic expression at these loci. Our data suggests that this genomic reprogramming event goes beyond mediating these epigenetic processes, causing broad changes in gene expression in postmigratory germ cells and allows for cellular differentiation. Methylation of DNA is thought to play a role in tissue specific gene expression patterns. Many genes contain
100 CpG islands which are methylated in non expressing tissues, and unmethylated in expressing tissues. Such genes would be expected to show ectopic expression in methylation deficient embryos. However, there have been no reports of any gene being misexpressed by the Dnmt1 mutant with the exception of the Bdnf gene, which shows increased, but not ectopic, expression in the brain (Chen et al. 2003; Martinowich et al. 2003). This suggests that for the majority of genes, methylation is not the cause of gene silencing and is probably a consequence of other silencing mechanisms (possibly histone chromatin modifications). Interestingly, we have found that a postmigratory germ cell gene, GCNA1 is misexpressed in the methyltransferase mutant both spatially and temporally (Chapter 6). GCNA1, which is normally expressed exclusively in germ cells at 11.5 dpc, is found in numerous somatic lineages in the mutant at 8.5 dpc. These results suggest that for GCNA1, and possibly for other postmigratory germ cell genes such as Mvh, Scp3 and Dazl, which undergo a wave of demethylation at this time, DNA methylation plays a unique role in regulating gene expression patterns. Methylation present in premigratory germ cells and somatic cells maintains these genes in a transcriptionally repressed state, while a wave of demethylation, which occurs specifically in germ cells following gonadal colonization, allows for transcriptional activation. The ectopic expression of GCNA1 in Dnmt1 mutant embryos also suggests that the transcription factors necessary for postmigratory germ cell gene expression are present at 8.5 dpc in somatic cells, as well as premigratory germ cells and that these genes are not regulated by a set of germ cell specific transcription factors. If germ cell specific transcription factors were required for postmigratory gene expression, our results suggest
101 that the transcription factors would also be regulated by loss of DNA methylation. However, it is more likely that the transcription factors are not regulated by methylation as few genes have been identified which show deregulation in a Dnmt1 mutant. Since we have shown that at least one postmigratory germ cell gene is regulated by methylation, and three other genes show a loss of methylation prior to their time of expression, it seems more likely that it is the genes themselves, which are directly regulated by changes in DNA methylation. Investigation into the identity of these transcription factors will allow us to further address these issues. Our lab has proposed a mechanism for primordial germ cell differentiation, which is mediated by DNA demethylation. As one postmigratory gene has shown inappropriate activation when DNA demethylation occurs earlier, we have hypothesized that other postmigratory genes will behave similarly. However, further experiments are necessary to validate the proposed model for differentiation. First, the ectopic expression of GCNA1 in embryos deficient in the Dnmt1 enzyme is clearly visible in somatic cells, but the presence of primordial germ cells in the embryos has not been confirmed. In order to conclude that their is promiscuous activation in primordial germ cells, double staining of these mutants must be carried out to demonstrate that cells are present which express markers of early germ cells (such as Tnap or SSEA-1) and GCNA1 concomitantly. This will confirm the presence of germ cells which prematurely express GCNA1, supporting a role for methylation as a silencing mechanism in premigratory germ cells. Second, there are many cells in the mutant embryo which do not ectopically express GCNA1. This may be due to low levels of the Dnmt1 enzyme present in embryos of the N-terminal deletion (Dnmtn/n). Low levels of Dnmt1 expression allow for methylation at 30% of
102 normal levels (Li et al. 1992). It is possible that some cells express lower levels of the enzyme than others and therefore cells expressing GCNA1 may represent cells with very low levels of the enzyme, while other cells with higher enzyme levels do not initiate GCNA1 expression. Investigation of the extent of GCNA1 expression in the C-terminal deletion mutant (Dnmt1c/c), which shows no detectable Dnmt1 methyltransferase would address this issue (Lei et al. 1996). If this idea is correct, in Dnmt1c/c mutant embryos, the ectopic expression of GCNA1 will be more extensive. Another issue which must be addressed is the connection between DNA demethylation and primordial germ cell differentiation. Investigating whether or not other differentiation events (aside from gene expression changes) occur in the methyltransferase mutant embryos would provide evidence that DNA demethylation is leading to germ cell differentiation. As mutations in the Dnmt1 gene cause embryonic lethality at 9.5 to 10.5 dpc, the differentiation events which occur upon gonadal colonization cannot be studied. It would be interesting to generate a mouse model in which the Dnmt1 gene is specifically deleted in the germ line. Since a mouse carrying a conditional mutation of the Dnmt1 gene has been generated, as well as mice expressing Cre recombinase (Cre) from the Tnap promoter (Tnap-Cre) (Lomeli et al. 2000), it is possible to generate a germ cell specific deletion which would likely not be an embryonic lethal. However, there are several problems with the Tnap-Cre mouse line. Although Tnap is highly expressed in primordial germ cells at 7.5 dpc, there are low levels of Tnap expression throughout the epiblast at earlier developmental stages including the blastocyst stage where the gene is expressed in every cell. This ectopic expression can lead to recombination in every cell
103 in some strains (C57Bl/6), although in other strains the ectopic Cre activation is less severe (129/SvJ) (personal communications, H. Scholer). Another potential problem with the Tnap-Cre mouse is that when crossed to a reporter strain, recombination is not detected until 9.5 dpc. If deletion of the Dnmt1 gene were to occur at 9.5 dpc, DNA demethylation may not occur before the normal wave of demethylation occurs. Therefore other germ cell specific Cre expressing mouse lines should be generated to avoid these problems. Two possible gene candidates for this are fragilis and stella. Fragilis is expressed at 6.5 dpc in the population of cells which give rise to germ cells and allantois (Saitou et al. 2002). This is advantageous because it allows more time for the deletion to occur before the normal wave of demethylation is initiated. Since fragilis is also expressed at low levels throughout the epiblast at earlier stages, using an inducible form of Cre recombinase (CreERT2) (Hayashi and McMahon 2002) would prevent the Cre from causing recombination until the mice were treated with the inducer Tamoxifen. This would alleviate problems with early ectopic expression. Stella, on the other hand, does not show early ectopic expression (aside from maternally stored protein) and is expressed by 7.5 dpc (Saitou et al. 2002). Therefore fragilis and stella are good candidates for producing mouse lines expressing Cre specifically in the germ line and would allow for germ cell differentiation to be studied in the context of premature DNA demethylation. An alternative approach to studying the role of DNA methylation on germ cell differentiation is to delay the demethylation event. Although the mechanism of demethylation is unknown, it would be informative to overexpress a de novo methyltransferase in premigratory germ cells. Investigation of 12.5 dpc germ cells by
104 immunofluorescence has shown that while Dnmt1 is highly expressed in the nucleus, no de novo methyltransferase is expressed. The Dnmt3b isoform is detected in the cytoplasm of 12.5 dpc germ cells, and the Dnmt3a isoform is not detectable by immunofluoresence. If DNA demethylation requires the exclusion of de novo methyltransferase from germ cell nuclei, then overexpression of such an enzyme may prevent demethylation from occurring, or cause demethylation to occur more slowly. In this case, the observation that germ cell differentiation is also delayed would strengthen the interpretation that DNA methylation is rate limiting for differentiation. DNA Demethylation Genome wide DNA demethylation occurs only two times during development, in both cases within germ cells. The first wave of demethylation occurs shortly after fertilization. The paternally contributed pronucleus undergoes a very rapid demethylation most likely mediated by factors in the egg cytoplasm as it remodels the highly condensed chromatin of the sperm derived pronucleus. Passive demethylation of the maternal pronucleus then occurs during subsequent cleavage stages (Kafri et al. 1993). Demethylation in primordial germ cells occurs shortly after entry into the developing gonads and is probably mediated by an active process (Chapter 5). As the mechanisms controlling these demethylation events are unknown, it is interesting to compare the two events and speculate whether or not they may be mediated by the same factors or conditions. DNA demethylation at different stages of development differs with respect to imprinted loci as several imprinted genes are known to avoid demethylation in the early embryo (Kafri et al. 1993), but become demethylated in primordial germ cells (Hajkova et al. 2002; Lee et al. 2002). The majority of genes subject to imprinting are expressed
105 from the paternal allele and exhibit methylation on the maternal allele, only a few genes having been identified with the opposite imprinting pattern (Reik and Walter 2001). The ability to avoid this wave of demethylation, therefore occurs for most imprinted genes during passive demethylation of the maternal genome, and only a few imprinted genes must avoid active demethylation of the male pronucleus. All imprinted genes so far investigated have been observed to undergo demethylation in primordial germ cells, thought to be mediated by an active process. While the reasoning for the skewing of imprinted genes towards paternal expression is not understood, it is possible that there is a higher chance of imprints being lost when subject to active demethylation. Alternatively, it is also possible that imprinted loci are associated with a factor during preimplantation development which protects them from the DNA demethylation process, and that this factor is not associated with imprinted loci in primordial germ cells. Similarly, there may be differences in histone modifications associated with imprinted loci during these two stages of development which could potentially be involved in mediating demethylation, or the ability to avoid demethylation. An aspect of DNA demethylation which is similar during the two stages of development is the methyltransferase enzymes which are present. In both cases of active demethylation, there are high levels of the Dnmt1 enzyme present. While the somatic form of the protein is expressed highly (relative to neighboring somatic cells) in germ cells, the maternally stored oocyte form of the protein, Dnmt1o, is highly expressed in the oocyte and during the subsequent stages of preimplantation development (Mertineit et al. 1998). While the paternal pronuclei undergoes DNA demethylation prior to nuclear envelope formation, it is likely to be exposed to the maintenance methyltransferase while
106 undergoing demethylation. This is not the case for the passively demethylated maternally derived genome, as the Dnmt1o enzyme is sequestered cytoplasmically during the subsequent cleavage stages. Interestingly, the Dnmt1o protein transits to the nucleus during the 8-cell stage where it plays a role in maintaining genomic imprinting, and is again located cytoplasmically by the 16-cell stage (Howell et al. 2001). At all stages when DNA demethylation is occurring, active or passive, there is no de novo methyltransferase present. From these observations, it seems reasonable to predict that while active demethylation can occur in the presence of the Dnmt1 enzyme, passive demethylation occurs only when there is no methyltransferase enzyme present. This is supported by several observations: 1) The lack of cytoplasmic methyltransferase enzyme during maternal genome demethylation, although there is nuclear localized Dnmt1o present for one cell division during the 8 cell stage. 2) Attempts to generate cloned bovine embryos by somatic nuclear transfer using donor cells with lower levels of the Dnmt1 enzyme resulted in embryos with methylation levels closer to that of normal preimplantation embryos. It has been hypothesized that generating cloned embryos with global methyaltion levels resembling that of normal embryos may improve the developmental potential of these clones (personal communications, K. Moore). As somatic nuclei donors bring along significant amounts of nuclear Dnmt1 enzyme, this transfer of enzyme may interfere with proper genomic reprogramming. 3) Lastly, the lack of a de novo methyltransferase during both waves of demethylation suggests that both passive and active demethylation may only occur in the absence of such enzymes. This could be
107 tested by transgenic expression of the Dnmt3 genes during either of these stages of development. Understanding the Differentiation of Primordial Germ Cells Further studies into the mechanism regulating germ cell differentiation within the fetal gonad will contribute to our understanding of several aspects of development. As errors in germ cell development often lead to infertility, understanding the mechanisms controlling proper germ cell migration, proliferation and survival of the germ cells in the gonad will provide a better understanding of the causes of infertility. Similarly, abnormal germ cell differentiation can lead to the formation of an aggressive germ cell tumor or teratocarcinoma. Therefore, understanding the mechanisms which cause germ cells to cease proliferating and migrating, as well as understanding how apoptosis in the fetal gonad is regulated, will further our understanding of how developmental errors can lead to tumorogenesis. Primordial germ cells are unique as they are the only cell type in which genomic imprints can be remodeled. Recently, several labs have successfully generated cells strikingly similar to germ cells from the differentiation of embryonic stem (ES) cells (Geijsen et al. 2004; Hubner et al. 2003; Noce 2004). As the generation of germ cells, which could appropriately be used for stem cell therapy, would require imprinting to be reset properly, it is important that we understand how this reprogramming is normally carried out by the germ line. Therefore, further studies on the mechanisms controlling primordial germ cell differentiation in the fetal gonad will contribute to our understanding of various clinical problems such as infertility and cancer as well as provide insight into the generation of germ cells for stem cell therapy.
APPENDIX PRIMER SEQUENCES Table A-1. Bisulfite Primers Primer Name Primer Sequence (5'-3') B-MVH-F1 TGAATGAATATAATGGAATTGATGAGTT B-MVH-R1 AAAACAACAAATAACATCAAA B-SCP3-F1 GAATGAGGATTTATGAGTAAAGATGGTT B-SCP3-R1 CCCCCATCTCCTTAACCTCAA B-DAZ-F1 TGAGAGTTAGATGGTTTTGGGTTTGTTT B-DAZ-R1 CAAAACAACTTAACTACCACTAACCATACAA B-TNAP-F2 GTTGTATTTTGAAGTTAGGATGAG B-TNAP-R4 ACCAAAAGRCCCCCTACCRAAAA1 1 The nucleotide R represents a cytosine (C) or a thymine (T). 108
109 Table A-2. Sequencing primers for Mouse vasa homologue Primer Name Primer Sequence (5'-3') MVH-AR38 CACATGCAGATAAACACTGAAACAGGC MVH-AR39 GCTATCTTTCAACTATGGGGC MVH-AR41 CCACGTGCAACTACTAAAGG MVH-AR45 TTCTGCCCGAAAGGATTCCC MVH-R50 GGTGCTGTCACGCACGACATC MVH-R51 AAACCGAAGAGAGGAAGA MVH-R52 TAAAGGCAGGAGGATCAGG MVH-R53 TTCATTGAGTGCAGGTCAGG MVH-R54 TCAAGGTTTCTATTCCAGG MVH-F45 AATCCTTTCGGGCAGAACC
110 Table A-3. Sequencing primers for Synaptonemal complex protein 3 Primer Name Primer Sequence (5'-3') SCP3-F1 TGTGGGGACAGCGACAGCTC SCP3-F2 CCAATCAGCAGAGAGCTTGG SCP3-F3 AAAGATGGCCAGGTCAGGTGGG SCP3-F4 TGTCTGTCTGTCTATCAGCGC pSCP3-F5 AGTAAGCAGCATCCATCCATGC pSCP3-F6 TTGAAGGTTTTGTGGGTGGC pSCP3-F7 AAGAGGAGGGGCTAATGG pSCP3-F8 ACGCTCCAGTGTATCTAG pSCP3-F9 TAGAGGGCAGGCTAAACAACTC pSCP3-F10 GCATGTTGTGGTGACCCACAA SCP3-R4 CCATGGTAATCTTGCCATAGC SCP3-R5 TGCCTGAGCATTGCAATAACG SCP3-R6 TGGCACACACTTTTAATCCC SCP3-R7 TGGCACACAATTTTAATCCCG pSCP3-R8 TCACACTACACAACAGACCTC pSCP3-R9 TCCTTGGCCCTTTTAGACTC pSCP3-R10 TTGTGGGTCACCACAACATGC pSCP3-R11 AATGATCTGAGCTCCTGCCTG
111 Table A-4. Sequencing primers for Deleted in azoospermia-like Primer Name Primer Sequence (5'-3') DAZ-F3 CTGCCACAACTTCTGAGGCTC DAZ-F4 ACTCAACCTTCTCAATGTGGCT DAZ-R3 TGCAAGGTGGAGCAGAATCC DAZ-R5 TCACTGACAGACAGATGGACC DAZ-R6 TGCTACAGCCAATAGGCGGCA DAZ-R7 ACATTGAGAAGGTTGAGTCCGG
112 Table A-5. Reverse transcription polymerase chain reaction primers Primer Name Primer Sequence (5'-3') MVH-F8 TTTGGCTCATATGATGCGGG MVH-117 ACACCCTTGTACTATCTGTCGAACTGAATGACC MAGE-B4-F2 TTGTTAGCAAGTTAGATCCC MAGE-B4-R1 CTCCTCATAATGGGTTGGGAA DAZH-F1 TTCTGCTCCACCTTCGAGGTT DAZH-R1 CTATCTTCTGCACATCCCAGTCATTA HPRT-F GCTGGTGAAAAGGACCTCT HPRT-R CACAGGACTAGAACACCTGC
LIST OF REFERENCES Amir, R. E., Van den Veyver, I. B., Wan, M., Tran, C. Q., Francke, U., and Zoghbi, H. Y. (1999). Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet 23, 185-188. Anderson, R., Copeland, T. K., Scholer, H., Heasman, J., and Wylie, C. (2000). The onset of germ cell migration in the mouse embryo. Mechanisms Of Development 91, 61. Ara, T., Nakamura, Y., Egawa, T., Sugiyama, T., Abe, K., Kishimoto, T., Matsui, Y., and Nagasawa, T. (2003). Impaired colonization of the gonads by primordial germ cells in mice lacking a chemokine, stromal cell-derived factor-1 (SDF-1). Proc Natl Acad Sci U S A. Atchison, F. W., Capel, B., and Means, A. R. (2003). Pin1 regulates the timing of mammalian primordial germ cell proliferation. Development 130, 3579-3586. Beverley, P. C., Linch, D., and Delia, D. (1980). Isolation of human haematopoietic progenitor cells using monoclonal antibodies. Nature 287, 332-333. Brannan, C. I., Lyman, S. D., Williams, D. E., Eisenman, J., Anderson, D. M., Cosman, D., Bedell, M. A., Jenkins, N. A., and Copeland, N. G. (1991). Steel-Dickie mutation encodes a c-kit ligand lacking transmembrane and cytoplasmic domains. Proc Natl Acad Sci U S A 88, 4671-4674. Buehr, M., and McLaren, A. (1993). Isolation and culture of primordial germ cells. Methods Enzymol 225, 58-77. Capel, B. (2000). The battle of the sexes. Mechanisms of Development 92, 89. Chapman, V., Forrester, L., Sanford, J., Hastie, N., and Rossant, J. (1984). Cell lineage-specific undermethylation of mouse repetitive DNA. Nature 307, 284-286. Chen, R. Z., Akbarian, S., Tudor, M., and Jaenisch, R. (2001). Deficiency of methyl-CpG binding protein-2 in CNS neurons results in a Rett-like phenotype in mice. Nat Genet 27, 327-331. Chen, W. G., Chang, Q., Lin, Y., Meissner, A., West, A. E., Griffith, E. C., Jaenisch, R., and Greenberg, M. E. (2003). Derepression of BDNF transcription involves calcium-dependent phosphorylation of MeCP2. Science 302, 885-889. 113
114 Chiquoine, A. D. (1954). The identification, origin, and migration of primordial germ cells in the mouse embryo. Anat Rec 118, 135-146. Church, G. M., and Gilbert, W. (1985). The genomic sequencing technigque. Prog Clin Biol Res 177, 17-21. Clark, J. M., and Eddy, E. M. (1975). Fine structural observations on the origin and associations of primordial germ cells of the mouse. Dev Biol 47, 136-155. Clark, S. J., Harrison, J., Paul, C. L., and Frommer, M. (1994). High sensitivity mapping of methylated cytosines. Nucleic Acids Res 22, 2990-2997. Cooke, H. J., Lee, M., Kerr, S., and Ruggiu, M. (1996). A murine homologue of the human DAZ gene is autosomal and expressed only in male and female gonads. Hum-Mol-Genet 5, 513-516 issn: 0964-6906. Coucouvanis, E. C., Sherwood, S. W., Carswell Crumpton, C., Spack, E. G., and Jones, P. P. (1993). Evidence that the mechanism of prenatal germ cell death in the mouse is apoptosis. Exp-Cell-Res 209, 238-247. Di Carlo, A., Travia, G., and De Felici, M. (2000). The meiotic specific synaptonemal complex protein SCP3 is expressed by female and male primordial germ cells of the mouse embryo. International Journal Of Developmental Biology 44, 241. Dolci, S., Williams, D. E., Ernst, M. K., Resnick, J. L., Brannan, C. I., Lock, L. F., Lyman, S. D., Boswell, H. S., and Donovan, P. J. (1991). Requirement for mast cell growth factor for primordial germ cell survival in culture. Nature 352, 809-811. Donovan, P. J., Stott, D., Cairns, L. A., Heasman, J., and Wylie, C. C. (1986). Migratory and postmigratory mouse primordial germ cells behave differently in culture. Cell 44, 831-838. Eddy, E. M., and Clark, J. M. (1975). Electron microscopic study of migrating primordial germ cells in the rat. pp. 151-67. In: Hess M, ed Electron microscopic concepts of secretion New York, Wiley 101. Enders, G. C., and May, J. J. (1994). Developmentally regulated expression of a mouse germ cell nuclear antigen examined from embryonic day 11 to adult in male and female mice. Dev-Biol 163, 331-340. Escalante-Alcalde, D., Recillas-Targa, F., Hernandez-Garcia, D., Castro-Obregon, S., Terao, M., Garattini, E., and Covarrubias, L. (1996). Retinoic acid and methylation cis-regulatory elements control the mouse tissue non-specific alkaline phosphatase gene expression. Mech Dev 57, 21-32. Futscher, B. W., Oshiro, M. M., Wozniak, R. J., Holtan, N., Hanigan, C. L., Duan, H., and Domann, F. E. (2002). Role for DNA methylation in the control of cell type specific maspin expression. Nat Genet 31, 175-179.
115 Garcia-Castro, M. I., Anderson, R., Heasman, J., and Wylie, C. (1997). Interactions between germ cells and extracellular matrix glycoproteins during migration and gonad assembly in the mouse embryo. J Cell Biol 138, 471-480. Gardiner-Garden, M., and Frommer, M. (1987). CpG islands in vertebrate genomes. J Mol Biol 196, 261-282. Geijsen, N., Horoschak, M., Kim, K., Gribnau, J., Eggan, K., and Daley, G. Q. (2004). Derivation of embryonic germ cells and male gametes from embryonic stem cells. Nature 427, 148-154. Gomperts, M., Garcia Castro, M., Wylie, C., and Heasman, J. (1994). Interactions between primordial germ cells play a role in their migration in mouse embryos. Development 120, 135-141 issn: 0950-1991. Guy, J., Hendrich, B., Holmes, M., Martin, J. E., and Bird, A. (2001). A mouse Mecp2-null mutation causes neurological symptoms that mimic Rett syndrome. Nat Genet 27, 322-326. Hahnel, A. C., Rappolee, D. A., Millan, J. L., Manes, T., Ziomek, C. A., Theodosiou, N. G., Werb, Z., Pedersen, R. A., and Schultz, G. A. (1990). Two alkaline phosphatase genes are expressed during early development in the mouse embryo. Development 110, 555-564. Hajkova, P., Erhardt, S., Lane, N., Haaf, T., El-Maarri, O., Reik, W., Walter, J., and Surani, M. (2002). Epigenetic reprogramming in mouse primordial germ cells. Mech Dev 117, 15. Hayashi, S., and McMahon, A. P. (2002). Efficient recombination in diverse tissues by a tamoxifen-inducible form of Cre: a tool for temporally regulated gene activation/inactivation in the mouse. Dev Biol 244, 305-318. Howell, C. Y., Bestor, T. H., Ding, F., Latham, K. E., Mertineit, C., Trasler, J. M., and Chaillet, J. R. (2001). Genomic imprinting disrupted by a maternal effect mutation in the Dnmt1 gene. Cell 104, 829-838. Hubner, K., Fuhrmann, G., Christenson, L. K., Kehler, J., Reinbold, R., De La Fuente, R., Wood, J., Strauss, J. F., 3rd, Boiani, M., and Scholer, H. R. (2003). Derivation of oocytes from mouse embryonic stem cells. Science 300, 1251-1256. Jaenisch, R. (1997). DNA methylation and imprinting: Why bother? Trends Genet 13, 323-329. Jenuwein, T., and Allis, C. D. (2001). Translating the histone code. Science 293, 1074-1080.
116 Kafri, T., Gao, X., and Razin, A. (1993). Mechanistic aspects of genome-wide demethylation in the preimplantation mouse embryo. Proc Natl Acad Sci U S A 90, 10558-10562. Kelly, S. J. (1977). Studies of the Developmental Potential of 4and 8-Cell Stage Mouse Blastomeres. Journal of Experimental Zoology 200, 365-376. Knaut, H., Werz, C., Geisler, R., and Nusslein-Volhard, C. (2003). A zebrafish homologue of the chemokine receptor Cxcr4 is a germ-cell guidance receptor. Nature 421, 279-282. Koshimizu, U., Watanabe, M., and Nakatsuji, N. (1995). Retinoic acid is a potent growth activator of mouse primordial germ cells in vitro. Dev-Biol 168, 683-685. Kudo, S., and Fukuda, M. (1995). Tissue specific transcriptional regulation of human leukosialin (CD43) gene is achieved by DNA methylation. Journal of Biological Chemistry 270, 13298-13302. Lawson, K. A., Dunn, N. R., Roelen, B. A. J., Zeinstra, L. M., Davis, A. M., Wright, C. V. E., Korving, J. P. W. F. M., and Hogan, B. L. M. (1999). Bmp4 is required for the generation of primordial germ cells in the mouse embryo. Genes & Development 13, 424. Lawson, K. A., and Hage, W. J. (1994). Clonal analysis of the origin of primordial germ cells in the mouse. Ciba Found Symp 182, 68-84. Lee, J., Inoue, K., Ono, R., Ogonuki, N., Kohda, T., Kaneko-Ishino, T., Ogura, A., and Ishino, F. (2002). Erasing genomic imprinting memory in mouse clone embryos produced from day 11.5 primordial germ cells. Development 129, 1807-1817. Lei, H., Oh, S. P., Okano, M., Juttermann, R., Goss, K. A., Jaenisch, R., and Li, E. (1996). De novo DNA cytosine methyltransferase activities in mouse embryonic stem cells. Development 122, 3195-3205. Li, E. (2002). Chromatin modification and epigenetic reprogramming in mammalian development. Nat Rev Genet 3, 662-673. Li, E., Beard, C., and Jaenisch, R. (1993). Role for DNA methylation in genomic imprinting. Nature 366, 362-365. Li, E., Bestor, T., and Jaenisch, R. (1992). Targeted Mutation of the DNA Methylatrasferase Gene Results in Embryonic Lethality. Cell 69, 915-926. Lomeli, H., Ramos-Mejia, V., Gertsenstein, M., Lobe, C. G., and Nagy, A. (2000). Targeted insertion of Cre recombinase into the TNAP gene: excision in primordial germ cells. Genesis 26, 116-117.
117 Martinowich, K., Hattori, D., Wu, H., Fouse, S., He, F., Hu, Y., Fan, G., and Sun, Y. E. (2003). DNA methylation-related chromatin remodeling in activity-dependent BDNF gene regulation. Science 302, 890-893. Matsui, Y., Zsebo, K., and Hogan, B. L. (1992). Derivation of pluripotential embryonic stem cells from murine primordial germ cells in culture. Cell 70, 841-847. Mayer, W., Niveleau, A., Walter, J., Fundele, R., and Haaf, T. (2000). Demethylation of the zygotic paternal genome. Nature 403, 501-502. McCoshen, J. A., and McCallion, D. J. (1975). A study of the primordial germ cells during their migratory phase in Steel mutant mice. Experientia 31, 589-590 issn: 0014-4754. McLaren, A., and Southee, D. (1997). Entry of mouse embryonic germ cells into meiosis. Dev Biol 187, 107-113. Mertineit, C., Yoder, J. A., Taketo, T., Laird, D. W., Trasler, J. M., and Bestor, T. H. (1998). Sex-specific exons control DNA methyltransferase in mammalian germ cells. Development 125, 889-897. Mintz, B., and Russell, E. S. (1957). Gene-induced embryological modifications of primordial germ cells in the mouse. J Exp Zool 134, 207-234. Molyneaux, K. A., Schaible, K., and Wylie, C. (2003a). GP130, the shared receptor for the LIF/IL6 cytokine family in the mouse, is not required for early germ cell differentiation, but is required cell-autonomously in oocytes for ovulation. Development 130, 4287-4294. Molyneaux, K. A., Zinszner, H., Kunwar, P. S., Schaible, K., Stebler, J., Sunshine, M. J., O'Brien, W., Raz, E., Littman, D., Wylie, C., and Lehmann, R. (2003b). The chemokine SDF1/CXCL12 and its receptor CXCR4 regulate mouse germ cell migration and survival. Development 130, 4279-4286. Monk, M. (1992). The X chromosome in development in mouse and man. J Inherit Metab Dis 15, 499-513. Monk, M., Boubelik, M., and Lehnert, S. (1987). Temporal and regional changes in DNA methylation in the embryonic, extraembryonic and germ cell lineages during mouse embryo development. Development 99, 371-382. Mroz, K., Carrel, L., and Hunt, P. A. (1999). Germ cell development in the XXY mouse: Evidence that X chromosome reactivation is independent of sexual differentiation. Developmental Biology 207, 229. Nieuwkoop, P. D., and Sutasurya, L. A. (1979). Primordial Germ Cells in the Chordates (Cambridge, London, Cambridge University Press).
118 Noce, T. (2004). Derivation of gametes from embryonic stem cells in culture. Tanpakushitsu Kakusan Koso 49, 616-624. Okano, M., Bell, D. W., Haber, D. A., and Li, E. (1999). DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Cell 99, 247-257. Payer, B., Saitou, M., Barton, S. C., Thresher, R., Dixon, J. P., Zahn, D., Colledge, W. H., Carlton, M. B., Nakano, T., and Surani, M. A. (2003). Stella is a maternal effect gene required for normal early development in mice. Curr Biol 13, 2110-2117. Reik, W., and Walter, J. (2001). Genomic imprinting: parental influence on the genome. Nat Rev Genet 2, 21-32. Resnick, J. L., Bixler, L. S., Cheng, L., and Donovan, P. J. (1992). Long-term proliferation of mouse primordial germ cells in culture. Nature 359, 550-551. Richards, A. J., Enders, G. C., and Resnick, J. L. (1999a). Activin and TGFbeta limit murine primordial germ cell proliferation. Dev Biol 207, 470-475. Richards, A. J., Enders, G. C., and Resnick, J. L. (1999b). Differentiation of murine premigratory primordial germ cells in culture. Biol Reprod 61, 1146-1151. Rogulska, T. (1969). Migration of the chick primordial germ cells from the intracoelomically transplanted germinal crescent into the genital ridge. Experientia 25, 631-632. Rossant, J., Sanford, J. P., Chapman, V. M., and Andrews, G. K. (1986). Undermethylation of structural gene sequences in extraembryonic lineages of the mouse. Dev Biol 117, 567-573. Sado, T., Fenner, M. H., Tan, S. S., Tam, P., Shioda, T., and Li, E. (2000). X inactivation in the mouse embryo deficient for Dnmt1: distinct effect of hypomethylation on imprinted and random X inactivation. Dev Biol 225, 294-303. Saitou, M., Barton, S. C., and Surani, M. A. (2002). A molecular programme for the specification of germ cell fate in mice. Nature 418, 293-300. Salvatore, P., Benvenuto, G., Caporaso, M., Bruni, C. B., and Chiariotti, L. (1998). High resolution methylation analysis of the galectin-1 gene promoter region in expressing and nonexpressing tissues. FEBS Letters 421, 152-158. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). Molecular Cloning, second edition edn, Cold Spring Harbor Laboratory Press).
119 Scholer, H. R., Hatzopoulos, A. K., Balling, R., Suzuki, N., and Gruss, P. (1989). A family of octamer-specific proteins present during mouse embryogenesis: evidence for germline-specific expression of an Oct factor. EMBO-J 8, 2543-2550 issn: 0261-4189. Shahbazian, M., Young, J., Yuva-Paylor, L., Spencer, C., Antalffy, B., Noebels, J., Armstrong, D., Paylor, R., and Zoghbi, H. (2002). Mice with truncated MeCP2 recapitulate many Rett syndrome features and display hyperacetylation of histone H3. Neuron 35, 243-254. Stallock, J., Molyneaux, K., Schaible, K., Knudson, C. M., and Wylie, C. (2003). The pro-apoptotic gene Bax is required for the death of ectopic primordial germ cells during their migration in the mouse embryo. Development 130, 6589-6597. Stott, D., and Wylie, C. C. (1986). Invasive behaviour of mouse primordial germ cells in vitro. J Cell Sci 86, 133-144. Szabo, P. E., Hubner, K., Scholer, H., and Mann, J. R. (2002). Allele-specific expression of imprinted genes in mouse migratory primordial germ cells. Mech Dev 115, 157-160. Szabo, P. E., and Mann, J. R. (1995). Allele-specific expression and total expression levels of imprinted genes during early mouse development: Implications for imprinting mechanisms. Gene Develop 9, 3097-3108. Tam, P. P., and Snow, M. H. (1981). Proliferation and migration of primordial germ cells during compensatory growth in mouse embryos. J Embryol Exp Morphol 64, 133-147. Tam, P. P., and Zhou, S. X. (1996). The allocation of epiblast cells to ectodermal and germ-line lineages is influenced by the position of the cells in the gastrulating mouse embryo. Dev Biol 178, 124-132. Tam, P. P., Zhou, S. X., and Tan, S. S. (1994). X-chromosome activity of the mouse primordial germ cells revealed by the expression of an X-linked lacZ transgene. Development 120, 2925-2932 issn: 0950-1991. Toksoz, D., Zsebo, K. M., Smith, K. A., Hu, S., Brankow, D., Suggs, S. V., Martin, F. H., and Williams, D. A. (1992). Support of human hematopoiesis in log-term bone marrow cultures by murine stromal cells expressing the membrane-bound and secreted forms of the human homolog of the steel gene product, stem cell factor. Proc Natl Acad Sci USA 89, 7350-7354. Toyooka, Y. T. N. T. Y. M. Y. S. M. N. T. (2000). Expression and intracellular localization of mouse Vasa-homologue protein during germ cell development. Mechanisms Of Development 93, 139.
120 Trasler, J. M., Trasler, D. G., Bestor, T. H., Li, E., and Ghibu, F. (1996). DNA methyltransferase in normal and Dnmtn/Dnmtn mouse embryos. Dev Dyn 206, 239-247. Tsuda, M., Sasoaka, Y., Kiso, M., Abe, K., Haraguchi, S., Kobayashi, S., and Saga, Y. (2003). Conserved Role of nanos Proteins in Germ Cell Development. Science 301, 1239-1241. Walsh, C. P., Chaillet, J. R., and Bestor, T. H. (1998). Transcription of IAP endogenous retroviruses is constrained by cytosine methylation. Nat Genet 20, 116-117. Walsh, C. P. B. T. H. (1999). Cytosine methylation and mammalian development. Genes & Development 13, 26. Wang, D. H., Ikeda, Y., Parker, K. L., and Enders, G. C. (1997). Germ cell nuclear antigen (GCNA1) expression does not require a gonadal environment or steroidogenic factor 1: Examination of GCNA1 in ectopic germ cells and in Ftz-F1 null mice. Mol Reprod Dev 48, 154-158. Wylie, C. (1999). Germ cells. Cell 96, 165. Ying, Y., Liu, X. M., Marble, A., Lawson, K. A., and Zhao, G. Q. (2000). Requirement of Bmp8b for the generation of primordial germ cells in the mouse. Mol Endocrinol 14, 1053-1063. Ying, Y., and Zhao, G. Q. (2001). Cooperation of endoderm-derived BMP2 and extraembryonic ectodermderived BMP4 in primordial germ cell generation in the mouse. Dev Biol 232, 484-492. Yoshimizu, T., Obinata, M., and Matsui, Y. (2001). Stage-specific tissue and cell interactions play key roles in mouse germ cell specification. Development 128, 481-490. Yoshimizu, T. S. N. D. F. M. I. Y. Y. O. K. M. K. O. M. A. K. S. H. R. M. Y. (1999). Germline-specific expression of the Oct-4/green fluorescent protein (GFP) transgene in mice. Development Growth & Differentiation 41, 675. Zamboni, L., and Upadhyay, S. (1983). Germ cell differentiation in mouse adrenal glands. J Exp Zool 228, 173-193.
BIOGRAPHICAL SKETCH Danielle Maatouk grew up Roslyn, New York. After moving to Florida she attended Tarpon Springs High School and gradutated in 1995. She attended the University of Florida for undergraduate studies in Microbiology and Cell Science, receiving her Bachelor of Science degree in 1999. She continued at the University of Florida entering the Interdisciplinary Program in Biomedical Sciences in 1999, joining the department of Molecular Genetics and Microbiology. 121