<%BANNER%>

Minimal Sequences Necessary for Imprinted Expression of the Prader-Willi Locus


PAGE 1

MINIMAL SEQUENCES NECESSARY FO R IMPRINTED EXPRESSION OF THE PRADER-WILLI LOCUS By CHRISTOPHER FUTTNER 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 2007

PAGE 2

Copyright 2007 by Christopher R. Futtner

PAGE 3

iii ACKNOWLEDGMENTS I would like to thank both past and pres ent members of the lab, including Stormy Chamberlain, Edwin Peery, Jessica Walrat h, Mike Elmore, Amanda DuBose, Emily Smith, Lori Kellam, and my favorite Scotswoman, Karen Johnstone. I would also like to thank my mentor Jim Resnick, not only for his scientific expertise, but also for his excellent advise on why old school is better than new school, how to have a happy marriage, and why the Yankees are the best team in baseball. I would like to thank my parents for their love and encouragement. Without them I would be biologically impossible. I am also extremely thankful to Cami Br annan. The experiments contained within this work would not have been po ssible without her ideas and vision. Lastly, I would like to thank my lovely wife, Danielle Maatouk, whose scientific coattails I will be riding to greatness.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF FIGURES...........................................................................................................vi ABSTRACT......................................................................................................................vii CHAPTER 1 INTRODUCTION........................................................................................................1 Genomic Imprinting......................................................................................................1 Prader-Willi and A ngelman Syndromes.......................................................................4 Purpose of Imprinting...................................................................................................6 The PWS/AS Locus......................................................................................................8 2 MATERIALS AND METHODS...............................................................................17 Lambda Red Recombineering....................................................................................17 Transformation of Recombineering Strains........................................................17 Transformation of Targeting Vector and Induction of Recombination...............19 BAC preparation for injection....................................................................................20 Generation of Transgenic Animals by Pronuclear Injection......................................23 Genotyping.................................................................................................................23 Identification of Transgenic Founders.................................................................23 Identification of Castaneous C7 Homozygotes...................................................23 Southern Blot..............................................................................................................24 RNA Isolation.............................................................................................................25 RT-PCR......................................................................................................................25 Preparation of High Molecular Weight Genomic DNA for Bisulfite Conversion.....26 Bisulphite Sequenci ng of Genomic DNA...................................................................27 Bisulphite Conversion.........................................................................................27 Bisulphite Polymerase Chain Reaction...............................................................27 Purification of PCR Products..............................................................................28 Cloning and Sequencing of PCR Products..........................................................28 Sequencing..................................................................................................................29 3 IMPRINTED TRANSGENE......................................................................................31

PAGE 5

v Introduction.................................................................................................................31 Results........................................................................................................................ .34 BAC Modification...............................................................................................34 Production of Transgenic Mice...........................................................................37 Analysis of 425 5 7 Transgenic Lines..............................................................38 Discussion...................................................................................................................41 4 REFINEMENT OF THE PWS-IC..............................................................................50 Introduction.................................................................................................................50 Results........................................................................................................................ .53 BAC Modification...............................................................................................53 Production of Transgenic Mice...........................................................................54 Analysis of 425 30kb Transgenic Lines.............................................................55 Discussion...................................................................................................................56 5 DEFININING THE LOCATION OF THE AS-IC.....................................................62 Introduction.................................................................................................................62 Results........................................................................................................................ .67 BAC modification...............................................................................................67 Production of Transgenic Mice...........................................................................69 Analysis of 425 U1-U3 and 425 U2/U3 Transgenic Lines..............................69 Discussion...................................................................................................................70 6 CONCLUSIIONS AND FUTURE DIRECTIONS....................................................79 LIST OF REFERENCES...................................................................................................82 BIOGRAPHICAL SKETCH.............................................................................................89

PAGE 6

vi LIST OF FIGURES Figure page 1-1 The life cycle of an imprint......................................................................................12 1-2 Mapping the Imprin ting Center (IC)........................................................................13 1-3 Molecular classes of PWS and AS...........................................................................14 1-4 Gene organization of the PWS/AS locus.................................................................15 1-5 Paternal only model of imprinting regulation..........................................................16 3-1 Schematic and expression of BAC transgenics........................................................44 3-2 425 5 7 targeting....................................................................................................45 3-3 rt-PCR analysis of 425 5 7mouse lines..................................................................46 3-4 Cast c7 breeding scheme..........................................................................................47 3-5 Bisulfite sequence analysis of maternally transmitted transgene.............................48 3-6 Bisulfite sequence analysis of paternally transmitted transgene..............................49 4-1 Schematic diagram of three nested de letions aimed at minimizing the PWS-IC.....59 4-2 rt-PCR analysis of 425 30kb mouse lines...............................................................60 4-3 Bisulfite sequence analysis of maternally transmitted 425 30kb transgene...........61 5-1 Schematic diagram of human upstream exons.........................................................74 5-2 Schematic diagram of upstream exons in the mouse...............................................75 5-3 Compensation model of AS-IC activity...................................................................76 5-4 Schematic diagram of upstream exon deletions.......................................................77 5-5 RT-PCR analysis of 425 U1-U3 mouse lines.........................................................78

PAGE 7

vii Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MINIMAL SEQUENCES NECESSARY FO R IMPRINTED EXPRESSION OF THE PRADER-WILLI LOCUS By CHRISTOPHER FUTTNER May 2007 Chair: James Resnick Major Department: Medical Sciences Genetics Prader-Willi (PWS) ands Angelman (AS) syndromes are neurodevelopmenal disorders arising from the improper expre ssion of oppositely imprinted genes located on human chromosome 15 q11-q13. Imprint regulation of this region is under the control of a bi-partite imprinting center consisting of an Angelman Imprinting Center (AS-IC) located approximately 35kb upstream of the paternally expressed Snrpn exon 1 and a Prader-Willi Imprinting Center (PWS-IC) located 5 to and including Snrpn exon 1. The PWS-IC has been shown to be a positive elem ent promoting expression of a set of genes on the paternal allele, while the AS-IC provides suppression of the PWS-IC on the maternal allele thereby suppressing expressi on of the same set of genes and allowing expression of the maternal program. Both are required for proper establishment and/or maintenance of the imprint in the germline. In the mouse, both gene order and the imprinted expression pattern have been conserve d, with the syntenic region being located

PAGE 8

viii on murine chromosome 7C. While the location of the PWS-IC has also been conserved, the position of the murine AS-IC remains unknown. We have taken a transgenic approach to locating the AS-IC and further dissecting out components of the PWS-IC. Using a r ecombineering method that utilizes the lambda phage Red genes, we have created a series of deletions within a Snrpn containing bacterial artificial chromosome (BAC) that we have shown recapitulates the imprinted expression of the endogenous locus. First we ha ve created a deletion of 30 kb just 5 to Snrpn exon 1 and show that imprinted expres sion is retained. Th is shows that the components required for establishment and/ or maintenance of the imprint on Snrpn are located within the reta ined sequence. Second, we have crea ted a series of deletions that remove several alternative upstream exons of Snrpn in various combinations and show that they contain the AS-IC element and each can act as an independent AS-IC when the other are deleted. These findings are significant because they further restrict the minimal necessary imprinting center functional unit and may suggest a possible mechanism of action.

PAGE 9

1 CHAPTER 1 INTRODUCTION Genomic Imprinting A consequence of sexual reproduction in mammals is the inheritance of both a maternal and paternal set of genes (alleles), which are virtually iden tical in sequence and for the most part expressed at similar levels However, a certain subset of genes is expressed from only one allele with the othe r being silenced. Th is phenomenon has been termed genomic imprinting. Imprinting is de fined as an epigenetic phenomenon whereby otherwise identical parental alleles are diffe rentially expressed in a parent of origin specific manner. This difference in func tionality is brought about by heritable marks placed upon the DNA without changes to gene sequence. The first evidence that the paternal and maternal genetic contributions were not functionally equal came from the labs of Solte r (McGrath and Solter, 1984) and Surani (Barton et al., 1984). Via pronucle ar transplantation experiment s, both labs independently came to the conclusion that embryos derive d from two paternal pronuclei (biparental androgenones) or two maternal pronucleii (biparental gynog enones) were incapable of completing normal embryogenesis. Zygotes de rived from two maternal genomes resulted in some embryonic growth but lacked the n ecessary extraembryonic tis sues necessary for implantation. This is similar to the forma tion of ovarian teratomas in humans, tumors derived from the spontaneous parthenogene sis of female germ cells. Like the gynogenetic zygote, ovarian teratomas devel op the various embryoni c tissues but lack extraembryonic ones (Linder et al., 1975; Oham a et al., 1985). Biparental andregenones

PAGE 10

2 on the other hand develop subs tantial extraembryonic tissu es but lack a significant embryo. The human equivalent to this is a hydatidiform mole, a conceptus with a completely paternal genome. Complete hydatidiform moles are characterized by hypertrophy of the trophoblast a nd a complete lack of embryo nic tissues (Jacobs et al., 1980). Further evidence for the inequality of the maternal and paternal genetic contribution came from the labs of Cattan ach (Cattanach, 1986; Cattanach and Kirk, 1985) and Searle (Searle and Beechey, 1978). By crossing mice heterozygous for certain Robertsonian translocations, they were able to create nondisjunction events that resulted in maternal or paternal dup lications in specific regions of individual chromosomes with the corresponding loss of the other parents a lleles. Many duplicati ons proved to have no effect as is the case with chromosomes 1, 4, 5, 9, 13, 14, and 15. However, duplications on several other chromosomes resulted in embryonic lethality or other developmental abnormalities. This showed that the genomic parental nonequivalancy was a regional phenomenon rather than a global one and suggested a role for imprinted genes. Central to the topic of genomic imprinting is the idea of an epigenetic mark. This mark contains information about the expression status of a particular allele. Importantly however it does not involve a ch ange in sequence between the maternal and paternal allele. While the eff ects of imprinting can be seen in gene expression, the actual imprint or mark that designates a gene to be imprinted is still unknown. What is known however is that the imprint has three phases in its life cycle: 1. establishment The mark must be placed w ithin cells of the developing germline in order to identify the allele as expressed or silenced. 2. maintenance The mark must be stable through mitosis so that it is propagated throughout all the somatic cells of the growing embryo.

PAGE 11

3 3. erasure The mark must be initially er ased as cells of the germline are first developing so that a new mark may be established according to the sex of the growing embryo (Figure 1-1). Current evidence suggests that chemical m odifications such as DNA methylation and histone modifications are responsible for dist inguishing maternal and paternal alleles. DNA methylation possesses these features ma king it an attractive candidate for the imprint mark (Reik and Dean, 2001). First, me thylation at imprinte d loci is typically restricted to the allele of only one pare nt. These methylation marks are typically established during gametogenesis and surviv e the genome wide wave of demethylation that occurs during the first cleavage stag es of embryogenesis. This differential methylation is maintained during the globa l de-novo methylati on that occurs post implantation. Second, DNA replication results in unmethylated CpG dinucleotides on the daughter strand. These hemimethylated sites qui ckly become fully methylated by specific DNA methyltransferases (DNMTs ) thereby retaining the mark during DNA synthesis and restricting it to previously me thylated sites. Third, methyl ation of imprinted loci is erased during gametogenesis and subsequent ly re-established during maturation of the oocyte and spermatocyte, resetting the methyl ation mark to that of the sex of the developing embryo. Evidence supporting methyla tion as the mark comes in the form of Dnmt1 (Howell et al., 2001) and Dnmt3L (Hata et al., 2002) knoc kouts that result in biallelic expression of prev iously imprinted genes It has been suggested that histone mo difications may also be involved in maintenance and establishment of differential expression of imprinted genes. Differential methylation patterns of hist ones associated with the PW S-IC and SNRPN have been reported (Xin et al., 2001) and histone methy ltransferases (Xin et al., 2003) have been shown to be required for maintenance of DN A methylation of the PWS-IC CpG island in

PAGE 12

4 cell culture. This type of mark could be both erasable and heritable fulfilling the requirements of an imprinting mark. Prader-Willi and Angelman Syndromes Prader-Willi (PWS) and Angelman (AS) syndromes are both neuro-developmental disorders linked to the same region of ch romosome 15. Both occur at a frequency of approximately 1 in 15,000 live births. Init ially, typical PWS patients exhibit neonatal hypotonia, poor suckle, failure to thrive, hypogonadism, and cryptorchidism. After the first few years of life the hypotonia and failu re to thrive give way to obesity and hyperphagia (a strong desire to eat). PWS is also characterized by moderate mental retardation and obsessive-compulsive diso rder (Holm et al., 1993) (Cassidy and Ledbetter, 1989). Obesity, complicated by a decreased caloric requirement and hyperphagia, is a major cause of morbidity am ong PWS patients. AS on the other hand is characterized by severe mental retardation, an ataxic gait, inappr opriate laughter, happy affect, and almost absent speech (Clayton-Smith and Pembrey, 1992). While the two disorders are clinically distin ct, both can be linked to disruptions in gene expression from chro mosome 15q11-q13 (Knoll et al ., 1989b). Most often PWS and AS can be attributed to large deletions of 3 4 mb encompassing this locus (category I). This region contains th e paternally expressed genes MKRN3, MAGEL2, NDN, SNURF/SNRPN several sno RNAs, and the UBE3A antisense transcript. It also contains the maternally expressed genes UBE3A and ATP10C Prader-Willi Syndrome is due to a paternal deletion of this region (Butler a nd Palmer, 1983) (Ledbetter et al., 1981) while Angelman Syndrome is due to a maternal de letion (Magenis et al ., 1987) (Knoll et al., 1989a). These de-novo deletions are belie ved to be facilitated by homologous

PAGE 13

5 recombination between large genomic duplica tions of the gene HERC2 that flank 15q11q13 (Amos-Landgraf et al., 1999). A second category of mutation that creates a functional loss is uniparental disomy (UPD) of chromosome 15 (category II). Pate rnal duplication results in AS (Malcolm et al., 1991) (Knoll et al., 1991) while maternal duplication results in PWS (Nicholls et al., 1989). UPD is thought to occur due to a non-disjunction event followed by reduction within the zygote to remain viable. Since both chromosome fifteens are inherited from one parent they both act in that manner and imprinted expression is lost from the other parents allele. A third class of mutations are only found in AS. These patients inherit an intact chromosome 15 from each parent but have mu tations within the UBE3A gene (Kishino et al., 1997) (Sutcliffe et al., 1997) or lesions of unknown origin (category IV and V). Thus, functional loss of UBE3A appears to be sufficient to cause the main clinical features of AS. Currently no PWS cases have been describe d which can be attributed to defects in a single gene. Rather it is believed that Prader-Willi is a contiguous gene syndrome in which multiple genes within 15q11-q13 must be lost to cause PWS. A fourth less frequent but important class of PWS and AS are imprinting center mutations and microdeletions (category III) (Horsthemke et al., 1997). The imprinting center (IC) is a cis acting regul atory region that controls the parental specific expression of genes in the locus. In this category both a maternal and paternal chromosome 15 are inherited, however loss of a functional IC resu lts in both behaving maternally in the case of PWS or paternally in th e case of AS. Mapping of microdeletions in PWS and AS patients has led to the identification of two shortest regions of overlap (SRO) indicating

PAGE 14

6 the boundaries of the PWS-IC and AS-IC (Figur e 1-2). The first, a 4.3kb region that is located just 5 to and includes the first e xon of SNRPN, is associated with Prader-Willi syndrome (PWS-IC) (Buiting et al., 1995) (S aitoh et al., 1996). The second, an 880bp region located approximately 35 kb upstream of SNRPN, is associated with Angleman syndrome (AS-IC) (Buiting et al., 1995) (Buiti ng et al., 1999). Evidence suggests that the PWS-IC is a positive elemen t that drives expression of th e paternal genes. Physical or functional loss of this element results in lo ss of the expression from the paternal allele. The AS-IC, on the other hand, seems to be a ne gative element silenci ng the activity of the PWS-IC on the maternal chromosome. Loss of this element results in loss of expression from the maternal allele (Figure 1-3). The severity of phenotype within AS is dependent upon the molecular defect inherited (Lossie et al., 2001) (Butler et al., 2004). Categor y I deletions produce the most severe phenotypes, showing high incidence of early onset seizures, microcephaly, and hypopigmentation while patients with UPD or IC mutations are much less likely to present with these symptoms. Patients with class IV and V mutations fall somewhere in between. The severity of category I is t hought to be due to th e physical loss of nonimprinted genes that fall between the breakpoi nts of the deletion but are outside of the imprinting locus. Function of Imprinting It is generally understood that in mamma ls the diploid state has the benefit of conferring protection against deleterious rece ssive mutations. What then could be the purpose of genomic imprinting, which in effect results in a haploid st ate at certain loci? Numerous hypothesis have been put forth to try to explai n the possible benefits of imprinting. Four are discussed below.

PAGE 15

7 The rheostat model was first suggested by the lab of Arthur Beaudet. It suggests that silencing a single allele in an individual shelters it away from the effects of natural selection (Beaudet and Jiang, 2002). This te mporary and reversible state of haploidy allows mutations that may be beneficial to the population to quick ly accumulate while those that are deleterious are hidden. Lethal mutations that are silenced may eventually mutate again and become advantageous. Bea udet suggests that this model allows for relatively rapid adaptations to environmental ch anges, however this is difficult to prove. A second hypothesis, termed the ovarian time bomb hypothesis, suggests that imprinting has evolved as a means of pr otection for the mother against ovarian trophoblastic disease (Varmuza and Mann, 1994). Those genes that are silenced within the maternal genome are suggested to be im portant in development of the trophectoderm. Without expression of these genes ovarian te ratomas that develop parthenogenetically within an ovary lose their invasiveness and remain relatively benign. While this hypothesis holds true for some male gene s it does not explain the presence of imprinting in all cases and does not explain th e presence of silenced paternal alleles. A third hypothesis is that imprinting of vital genes ensures sexual reproduction and therefore genetic diversity through hybrid vigo r (Driscoll, 1994). By requiring a genetic contribution from two parents, this ensures protection from the risk of homozygosity for deleterious recessive mutations. If parthenogen esis were possible in humans, the risk of acquiring these recessive alleles would increas e, decreasing the fitness of the species. The hypothesis that has gained the most traction within the imprinting community was put forth by David Haig in 1991. His Ki nship theory suggests a genomic tug-ofwar between maternal and paternal interest s (Moore and Haig, 1991). Here paternally

PAGE 16

8 expressed genes favor embryonic and neonatal growth as well as other traits that maximize survivability at the expense of the mother and future offspring. In order to protect herself and future offspring, the mate rnal genome silences those genes within the growing fetus that compromise the mothers re productive fitness and the survival of future pregnancies and expresses those that are growth inhibiting. Of the suggested models, the kinship th eory is supported best by observable evidence. First, the only animals in which imprinting has been observed are placental mammals. This is an important argument to th e kinship theory in th at female eutherians are more resource indebted to their young th an egg laying species. The second line of evidence comes from analysis of those genes that are imprinted. Indeed many of them are involved in fetal and pr enatal growth and behavior. None of the current explanations for imprinting can perfectly account for its existence. Many imprinted genes may in fact be hitch-hikers, genes that just happened to fall within an imprinted locis range of in fluence. Also not all genes may be imprinted for the same reasons. It seems likely that a ho st of factors probably played a role in the evolution of imprinting. The PWS/AS Locus The PWS/AS locus is highly conserved between man and mouse, both in gene order and imprinted expression making the mous e an excellent system in which to study the imprinting mechanism (Leff et al., 1992) (Lee et al., 2000) (MacDonald and Wevrick, 1997) (de los Santos et al., 2000) (Jong et al., 1999a) (Chamberlain and Brannan, 2001). The locus consists of two clusters of genes on chromosome 15q11-q13, an upstream cluster and a downstream cluste r in relation to the imprintin g center (IC). The syntenic region in the mouse is located on chromosome 7C (Figure 1-4).

PAGE 17

9 In humans and mice, the upstream cluster contains three paternally expressed intronless genes, MAGEL2/Magel2 (Boccaccio et al., 1999), and NDN/Ndn (MacDonald and Wevrick, 1997) both Mage family genes, and MKRN3/Mkrn3 (Jong et al., 1999a; Jong et al., 1999b) a putative zinc finger protein. Both NDN/Ndn and MAGEL2/Magel2 contain differentially methylated CpG islands within their 5 promoter regions as is frequently seen in imprinted genes. Both ar e hypomethylated on the paternal allele and hypermethylated on the maternal allele. MKRN3/Mkrn3 also contains a 5 CpG island, however it has yet to be shown to be differentia lly methylated. It is also associated with an anti-sense transcript of unknown function. The mouse contains a fourth paternally expressed intronless gene, Frat3 which is differentially methylated as well. Frat3 seems to have been acquired by the locus du e to a relatively recent L1 mediated retrotransposition event (Chai et al., 2001). Upon insertion into the locus, Frat3 appears to have become an innocent bystander having adopted the imprinted expression pattern of its neighboring genes. The downstream cluster consists of a single 460kb long transcript from which multiple paternally expressed gene products are spliced (Le Meur et al., 2005). Most proximal to the promoter is the bicistronic gene SNRPN/Snrpn that encodes both SmN a component of the spliceosome (Shemer et al., 1997), and SNURF/Snurf SNRPN upstream reading frame, a protein of unknown function (Gray et al., 1999). Additionally, this transcript encodes for several different snoRNAs; HBII-436/MBII-436, HBII13/MBII-13, HBII-437, HBII-438A and B and tandem repeat arrays of HBII-52/MBII-52 and HBII-85/MBII-85 (Runte et al., 2001). These snoR NAs are processed from introns contained in the long transcript. Mo st distally encoded is anti-sense UBE3A which is

PAGE 18

10 believed to be important for s ilencing the paternal copy of UBE3A (Runte et al., 2001). Two maternally expressed genes are also contained within the downstream cluster, UBE3A/Ube3a and ATP10C/Atp10c Ube3A is implicated as being the Angelman syndrome gene as deletion of UBE3A is sufficient to cause the disorder. Important to the region is the Imprinting Center (IC). The purpose of the imprinting center is to regulate mono-allelic expressi on within the locus. ICs are often found in imprinted loci however their mechanism of ac tion is not always the same. As discussed earlier, the IC is bipartite in nature consis ting of a PWS-IC and an AS-IC. The PWS-IC is located just 5 to and includes exon 1 of SNRPN/Snrpn. Paternal deletion of this region in mice leads to the loss of expression of the paternal set of genes and concurrent biallelic expression of UBE3A/Ube3A (Yang et al., 1998) (Chamberlain and Brannan, 2001). The AS-IC is approximately 35kb ups tream of the PWS-IC in humans. Microdeletions lead to bialleli c expression of the paternal ge nes and loss of expression of UBE3A/Ube3a. The AS-IC has yet to be loca ted in the mouse and is one focus of this work. Our lab has suggested a model (B rannan and Bartolomei, 1999) whereby the PWS-IC is a positive force, driving expression of the upstream and downstream paternal genes on the paternal allele by some unknown mechanism. The AS-IC on the other hand is a negative force, silencing the PWS-IC on the maternal allele and therefore silencing expression of the paternally expressed genes th ere (Figure 1-5). Several observations lead to this model. First, pate rnal inheritance of an AS-IC deletion is beni gn suggesting its function is only on the maternal allele. Li kewise maternal inheritance of a PWS-IC deletion is also benign sugges ting that it functions only on th e paternal allele. A third observation is that PWS-IC deletions th at include the AS-IC cause PWS whether

PAGE 19

11 maternally or paternally inherited. This shows that the function of the AS-IC is secondary to that of the PWS-IC. Indeed, if the function of the AS-IC is only to silence the PWS-IC, loss of both would have no effect on the maternal allele since there is no PWS-IC to drive expression of the set of pate rnal genes on both the maternal and paternal alleles.

PAGE 20

12 Figure 1-1 The life cycle of an imprint. There are three phases in the life of an imprint. First the imprint must be established in the germline. This imprinting mark can be distinguished between the matern al and paternal alleles. Second the mark must be maintained within the so matic tissues throughout the life of the animal. Third the imprint must be erased during gametogenesis of a developing embryo so that it can be re set according to the sex of the embryo.

PAGE 21

13 Figure 1-2. Mapping the Imprinti ng Center (IC). Mapping in humans using Shortest Region of Overlap (SRO) shows that the IC is bi-partite in structure. The PWS-IC is located within 4.3 kb interval that includes exon 1 of SNRPN. The AS-IC is located 35kb upstream of SN RPN exon 1 within a .88 kb interval

PAGE 22

14 Figure 1-3. Molecular classes of PWS and AS. There are four different molecular classes that can result in PWS and AS. First, large deletions of 3 4 Mb result in loss of the all genes within the locus. This deletion occurs on the paternal allele in the case of PWS and the maternal a llele in the case of AS. Second, uniparental disomy results in two alleles that act in the same manner. Maternal uniparental disomy results in PWS and Pa ternal uniparental disomy results in AS. A third class, single gene mutation, only occurs in the case of AS. This occurs when the gene UBE3A is mutated or deleted. The fact that this class does not exist in the case of PWS suggests that it is a contiguous gene disorder, requiring the loss of multiple genes to occur. A fourth class, IC microdeletions, occurs when the imprin ting control elements are missing or mutated.

PAGE 23

15 Figure 1-4. Gene organization of the PWS/AS locus. The PWS/AS locus, located on human chromosome 15q11-q13, consists of several genes upstream of the IC and a single long transcri pt originating from the SNRPN gene downstream of the IC. The syntenic region in the m ouse exists on mouse chromosome 7C and is conserved both in gene orde r and imprinted expression pattern.

PAGE 24

16 Figure 1-5. Paternal only model of imprinti ng regulation. The pate rnal only model of regulation suggests that th e PWS-IC is active only on the paternal allele driving expression of the paternal ge nes. The AS-IC on the other hand is active only on the maternal allele an d acts to silence the PWS-IC on the maternal there.

PAGE 25

17 CHAPTER 2 MATERIALS AND METHODS Lambda Red Recombineering Transformation of Recombineering Strains In order to create targeted deletions within the BAC of interest, it must first be transformed into the appropriate recombineeri ng E-coli strain. Three strains are available. DY380 is a modified DH10B strain containi ng the defective lambda prophage required for recombineering. EL350 is a modified DH10B strain containing the defective lambda prophage and an arabinose inducible cre gene used for cre /lox recombination. EL250 is a modified DH10B strain cont aining the defective lambda prophage and an arabinose inducible flpe gene used for flpe /frt recombination. Fresh BAC is prepared by alkaline lysa te method. Briefly, single colonies of DH10B containing the BAC of interest are pi cked and grown overnight in 3 mL Luria Broth (LB) with 12.5 ug/mL chloramphenicol at 32oC, 175 rpm. The next day cultures are pelleted in a microcentr ifuge at 7000 rpm, resuspended by vortexing in 100 uL solution I (50 mM dextrose, 10mM EDTA, 25mM Tris pH 8 .0) and allowed to sit at room temperature for 3 5 minutes. Next, 200 uL of solution II (.2M NaOH, 1% SDS) is added, the samples are mixed by inversion and then placed on ice. After 5 minutes 150 uL of solution III (29.4g KOAc, 11.5 ml. conc. HOAc, H20 to 100ml.) is added, the samples are shaken and then placed back on ice. 400 uL of prepared phenol chloroform is added to the tubes and the samples are shak en to mix and then centrifuged at 13000 rpm for 5 minutes. After centrifugation the aqueous la yer is transferred to a fresh tube and 1

PAGE 26

18 mL of 100% EtOH is added. The samples are then shaken and centrifuged again at 13000 rpm. The supernatant is poured off and 1ml of 70% EtOH is added. The samples are shaken and centrifuged again at 13000 rpm. Th e resulting DNA pellet is resuspended in 50ul of sterile H2O. In order to transform the BAC into the DY380 E.coli, a single colony of the appropriate recombineering strain is grown up overnight in 3 mL LB at 32oC, 175 rpm. The next day 700ul of culture is diluted in 70 mL of LB with 12.5ug/mL chloramphenicol and incubated 3 5 hours at 32oC, 175 rpm, until the culture density reaches an OD600 of .5 .6. The culture is next transferred in 10 mL aliquots to 15 mL falcon tubes and then pelleted by cold (~5oC) centrifugation at 5000 rpm. S upernatant is poured off and the pellet is resuspended in 10 mL ice cold 10% glycerol. The tubes are spun again at 5000 rpm, 5oC, for 5 minutes and the supernatant poured off. The pellet is resuspended this time in 1 mL 10%glycerol and transferred to a 1.5 mL eppendorf tube. The tubes are spun at 13000 rpm in a tabletop micro centrifuge for 30 seconds and the supernatant poured off. The pellets are again resuspended in 1 mL 10% glycerol and spun again at 13000 rpm. This time all of the supernatant is carefully drawn off with a pipette and the pellet is resuspended in 80ul of 10% glycerol. The tubes containing the now electrocompetent cells are stored on ice until use. Fresh BAC DNA is mixed with the electrocompetent cells to give a final volume of 50ul. Usually DNA volumes of 2, 5, and 10 uL mixed with 48, 45, and 40 uL competent cells will give enough of a range to produce good results. The DNA, cell mixture is transferred to a pre-cooled .1 cm Bio-Rad electroporation cuvette; catalog number 165 2089. Electroporation is perfor med using a Bio-Rad Gene Pulser set at 1.8kV, 25uF, with the pulse cont roller set at 200 ohms The electroporated

PAGE 27

19 cells are then immediately diluted with 1 mL of LB and allowed to incubate at 32oC, 175 rpm, for 1 hour. After incuba tion, the cells are pelleted by centrifugation at 7000 rpm and the supernatant drawn off with a pipette, leaving approximate ly 100ul total volume in the tube. The remaining volume is plated to LB agar plates containing 12.5 ug/mL chloramphenicol and incubated overnight at 32oC. Resulting colonies are screened by digesting with numerous different restriction enzymes and comparing them to the original BAC. Transformation of Targeting Vector and Induction of Recombination Recombination within the BAC is achieved by electroporating a targeting construct into the BAC containing recombineering stra in. The targeting cons truct should contain a selectable marker such as neomycin resist ance and should be linearized by restriction digest or amplified by polymerase chain react ion (pcr) prior to transformation. The night before transformation a single colony of the BAC containing strain is grown up overnight in 3 mL LB at 32oC, 175 rpm. The next day 700ul of cu lture is diluted in 70 mL of LB with 12.5ug/mL chloramphenicol and incubated 3 5 hours at 32oC, 175 rpm, until the culture density reaches an OD600 of .5 .6. At this point 10 mL of the culture should be drawn off and reserved on ice to serve as an uninduced control. The remaining 60 mL is induced to express the lambda RED genes by shaking in a 42oC water bath for 15 minutes. After induction the flask is transferred to an ice slurry bath and swirled for 10 minutes to turn off the defective lambda phage. The induced and uninduced control cultures are then made to be electrocom petent in the same way in which the recombineering strains were made to accept the BAC, by a series of cold 10% glycerol washes. Targeting vector DNA is mixed with the electrocompetent cells to give a final

PAGE 28

20 volume of 50ul. Usually DNA volumes of 2, 5, and 10 uL mixed with 48, 45, and 40 uL competent cells will give enough of a ra nge to produce good results. The DNA, cell mixture is transferred to a pre-cooled .1 cm Bio-Rad el ectroporation cuvette; catalog number 165 2089. Electroporation is performed usi ng a Bio-Rad Gene Pulser set at 1.8kV, 25uF, with the pulse controller set at 200 ohms. The electroporat ed cells are then immediately diluted with 1 mL of LB and allowed to incubate at 32oC, 175 rpm, for 1 hour. After incubation, the cells are pelle ted by centrifugation at 7000 rpm and the supernatant drawn off with a pipette, leavi ng approximately 100ul total volume in the tube. The remaining volume is plated to LB agar plates containing the appropriate antibiotic and incubated overnight at 32oC. The resulting coloni es are screened by Southern blot. BAC preparation for injection In order to inject BACs as transgenes, the DNA must be prepared in a way that removes all traces of impurities and leaves th e BAC as intact supercoils if possible. The DNA may be linearized by restriction digest, bu t this requires an additional purification step. BAC preparation by cesium prep and linearization are described below. Two nights before the DNA prep, a si ngle bacterial colony containing the appropriate BAC is picked and cultured overnight in 3 mL LB with 12.5 ug/mL chloramphenicol at 32oC, 175 rpm. The next day, 1 mL of this starter culture is inoculated into 3 1000ml flasks co ntaining 500 mL LB with 12.5 ug/mL chloramphenicol. These cultures are incubate d overnight at 32oC, 175 rpm. The next day, the cultures are poured into 500ml bottles and spun in a centrifuge at 5000 rpm. The resulting pellets are resuspended in 25 mL solution I each (50 mM dextrose, 10 mM

PAGE 29

21 EDTA, 25mM Tris pH 8.0) and allowe d to sit at room temperature for 3 5 minutes. 50 mL solution II (.2M NaOH, 1% SDS) is adde d to each bottle and th en inverted to mix. After 4 minutes 37.5 mL of solution III is added (29.4 g KOAc, 11.5ml glacial acetic acid, H2O to 100 ml), the bottles are mixed by i nversion and placed on ice for 10 minutes. In order to remove as much cellular debris as possible the mixture is poured into clean bottles through a layer of ga uze and then spun in a centrifuge at 5000 rpm for 5 minutes. The now cleared solution is additionally fi ltered through medium porosity filter paper (Fisher catalog number 09 802 1a). Each bottle now shoul d contain approximately 112 mL of BAC DNA solution. In order to precipit ate the DNA the solution is transferred to 3 250 mL bottles which are then filled to the neck with 100% EtOH. The EtOH/DNA mixture is spun in a centrifuge at 10,000 rp m for 20 minutes. The supernatant is poured off, the bottles filled with 70% EtOH and then spun again at 10,000 rpm for 10 minutes. The ethanol is poured off and any residual ethanol is removed from the bottles using a pipette. The pellets are then air dried for approximately 20 minutes and resuspended in a total of 8 mL TE with RNase when combined. To perform a gradient the density of the solution must be brought up to 1.55 g/mL by adding the proper amount of cesium. 1.05 gr ams of cesium are added per each mL of solution to achieve this density. Transfer the solution to a 15 mL conical tube and to this tube add 8.4 g of cesium. Mix the solution by inversion until all the cesium dissolves. The density should be checked by weighing 1 mL of solution on an analytical balance. The weight should equal between 1.55 1.57 g. The cesium/DNA mixture is then divided between 2 ultracentrifuge tube s. One hundred uL ethidium bromide is added and the volume brought up to the top of the tube us ing pre-made 1.55 g/mL TE-RNase solution

PAGE 30

22 stored in the freezer, making sure to keep the 2 tubes at an equal we ight. Seal the tubes spin them overnight at 55,000 rpm usi ng an NVT-90 fixed angle rotor in an ultracentrifuge. The next day, remove the tubes from the rotor and extract the DNA, which should be present as a visible red band within the tube, using an 18 gauge needle and syringe. Combine the DNA from the 2 tube s into one and bring up the volume using the pre-made 1.55g/mL TE-RNase solution. Seal the tube and spin again for 4 hours at 78,000 rpm using the same rotor. Again, extr act the DNA with an 18 gauge needle and syringe. It should be present now as a thicke r red band, and transfer to a 15 mL conical tube. There should be a total volume of 1 1.5 ml. Salt water saturated butanol is used to extract the ethidium from the DNA solution. To make the butanol solution, in a 250 mL bo ttle add NaCl to 50 mL of sterile water until the solution becomes completely saturated a nd the salt no longer dissolves. One hundred and fifty mL butanol is added to the bottle and mixed by shaking. The butanol is allowed to completely separated from the aqueous phase before using the solution. Two mL butanol is added to the BAC DNA solution and mixed by inverting 10 times. Allow the aqueous layer to resolve and then remove the upper butanol layer. Continue to repeat the addition and removal of butanol until the lo wer DNA solution layer is completely clear when looked at in front of a white background. The ethidium free BAC DNA solution is next transferred into a 4 inch long piece of Spectra/Por dialysis tubing (MWCO 12,000 14,000 daltons, cat.# 132676) for dialysis against the DNA injection buffer. The tightly sealed di alysis bag is allo wed to float in a bucket of 4 liters of injection buffer (10mM tris pH7.5, .1mM EDTA, 100mM NaCl) overnight at a temperature of 4oC with a slowly spinning stirbar. The next day the DNA

PAGE 31

23 solution is removed from the dialysis tubing and the con centration is measured on a spectrophotometer. Generation of Transgenic Anim als by Pronuclear Injection Injections were expertly done by the Chri s Futtner Injection Core (CFIC) at The University of Florida. Briefly DNA was micr oinjected into the pronucleus of fertilized mouse zygotes derived from superovulate d, 5 week old FVB female mice at a concentration of 2.5 3 ng/ul. Injected zygotes were allo wed to mature overnight to the two cell stage at which point they were transferred into the infundibula of pseudo pregnant (B6D2)F1 female mice. Genotyping Identification of Transgenic Founders PCR genotyping was used to identify founder animals carrying the various transgene constructs. At 2 weeks of age pups born to (B6D2)F1 foster moms were ear punched and tail clipped for screening. Ge nomic DNA was prepared from each tail piece by digestion with 10ug/mL protienase K in tail lysis buffer (100mM Tris pH 8.5, 5mM EDTA, .2% SDS, 200mM NaCl). PCR was perfo rmed to screen for the T7 BAC arm and adjoining mouse sequence us ing the following primers: T7F 5-GTA ATA CGA CTC ACT ATA GGC-3 and 425T7R1 5CTC CAA TCA TGT TCA ACT GTC-3. Founders were further screened by Southern blot, probing for Snrpn. Identification of Castaneous C7 Homozygotes PCR genotyping followed by restriction endonuclease digestion was used to identify transgenic animals that were also homozygous for Castaneous C7 at the PWS locus. This was done by screening for a polymorphic AvaII site between Mus musculus castaneous and Mus musculus domesticus that results from a cytosine nucleotide rather

PAGE 32

24 than a thymine nucleotide at position 117 of the Ndn gene. AvaII cuts the domesticus but not the castaneous allele due to the presen ce of the cytosine. PCR primers flanking the polymorphic site with the following sequen ce were used: NdnpolyF 5-ACA AAG TAA GGA CCT GAG CGA CC-3 a nd NdnpolyR 5-CAA CAT CTT CTA TCC GTT CTT CG-3. The amplified PCR product was gel pur ified on a 2% agarose gel, extracted from the gel using the Promega Wizard Prep kit, and digested with AvaII. The digested products were then electrophoresed on a 4.8% agarose gel (2:1 low-melt agarose: agarose). Southern Blot Southern blotting was performed as de scribed (Sambrook et al., 1989). Agarose gels were placed on a ultra vi olet (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 fo r transfer of the DNA from the gel to the nylon membrane. Th e 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 G ilbert, 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 hybr idization 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.

PAGE 33

25 RNA Isolation RT-PCR was performed to detect the expres sion of the transgenic Snrpn minigene. Total RNA was isolated from complete brai ns obtained from neonatal mice. RNA was extracted using RNA-zol (Tel-Test, Inc). Br iefly, tissue samples we re homogenized with 4 ml. of RNA-zol using a Polytr on homogenizer. To this homogenate .4ml of chloroform is added, the samples are shaken vigorously for 15 seconds and then put on ice for 5 minutes. After 5 minutes the tubes are th en centrifuged for 15 minutes at 12,000g after which the homogenate forms two distinct la yers: a lower blue phenol chloroform phase and an upper colorless aqueous phase containi ng the RNA. The a queous phase is added to a new tube and to it an equal volume of isopropanol is added and the samples vortexed then place on ice for 15 minutes. After 15 minutes the samples are centrifuged for 15 minutes at 12,000g. The resulting white RNA pe llet is then washed with 75%ethanol and centrifuged again at 12,000 g for 10 minutes. Af ter removal of the ethanol the pellet is allowed to air dry and then re suspended in .2 mL of DEPC H2O. RT-PCR RNA isolated from neonatal mouse brains was analyzed for transgene expression by RT-PCR. Five ug RNA is mixed with 1ul of 500 ug/mL random primers and dH2O to reach a final volume of 25.8ul. This mixture is then heated to 68oC for 3 minutes and then placed on ice. To each tube is added 3.2ul dNTP mixture, 8ul of reverse transcriptase buffer, 2ul 100mM DTT, 1ul RNase inhibitor a nd 1ul SuperscriptII and then incubated at 37oC for 60 minutes. Two uL of this r eaction is then used to seed a PCR reaction using the following primers: N2.1F 5-CCC CGA GTA TTA AGG ATC TTG-3 and N6.2R 5-GCA ACA GTG CCT CTT CCC TG-3. PCR amplification condition

PAGE 34

26 were 95oC for 5 min., followed by 30 cycles of 95oC for 30 sec., 57oC for 30 sec., 72oC for 1 min. The cycle was followed by an extension step of 72oC for 10 min. Preparation of High Molecular Weight Genomic DNA for Bisulfite Conversion High molecular weight brai n DNA was isolated from neonatal mouse brain to be used for bisulfite PCR. Two and one half mL of homogenizing solution (1X SSC, 1% SDS, .25mg/mL Pronase E) was added to a dounce homogenizer along with a single frozen neonatal mouse brain. Tissue was ho mogenized by gently moving the pestle up and down within the dounce. The homogena te was collected in a 15 mL polyproplene tube and the homogenizer washed with an additional 2.5 mL of solution that was then combined with the homogenate. The homoge nate was vortexed and then incubated for 1 hour in a 37oC water bath. After 1 hour, 5 mL of phenol chloroform was added and the mixture was vortexed and then spun in a centrifuge at 2500 rp m for 5 min to extract the DNA. The top aqueous layer was collected to a new 15 mL tube and the extraction was repeated twice more as above, first with phe nol chloroform and then with chloroform alone. After the aqueous layer was removed from the final chloroform extraction, 0.5 mL RNase A was added and then incubated for 30 min. at 37oC. One half mL of pronase E was then added and the mixture was incubated an addition 30 min at 37oC. The DNA was extracted again as above, first with phe nol chloroform and then with chloroform only. After extraction the DNA was precipita ted by the addition of 12 mL 95% ethanol followed by vortexing. The now visible DNA st rands were spooled onto a glass rod and placed into a 1.5 mL tube to dry. Once the DNA appeared to be mostly dry, 200 mL of water was added to resuspend the DNA which was then stored at oC.

PAGE 35

27 Bisulphite Sequenc ing 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 con centration is 0.3 M in 42 l) and incubated for 37oC for 30 minutes. Freshly prepared 2M sodium metabisulphite and 10 mM hydroquinone (final concentrati ons 1.55 M and 0.5 mM respectively) was added to the denatured DNA to a final volume of 240 l. Th e bisulphite reaction was inefficient on double stranded DNA. Samples were th en 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 bisu lphite ion was removed by passing the sample through a desalting column (Prome ga 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 a nd 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 dr ied 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 cy tosines (now thymines), predominantly at the 3' end. Primers were designed in regi ons in which no CpG dinucleotid es were present, however if a primer did include one or two CpGs a degenerate nucleotide (c ytosine or thymine)

PAGE 36

28 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 met hods. Primer sequences are listed in Table A-1. PCR was performed on bisulphite treated brain DNA preparations with HotStar Taq (Qiagen) using the following PCR conditi ons (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 brain DNA preparations were not quantitated for DNA concentration and 2 l was used per 25 l r eaction. PCR reactions were performed using the following cycling conditions: An initial denaturation at 95oC for 15 minutes was required for activation of HotStar Ta q 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 re peated 35 times followed by a final 72oC elongation for 10 minutes. Purification of PCR Products PCR products were purified using the Quia gen Quiaquik PCR cleanup kit as per kit directions. Once the PCR product was purified and resuspended in 45 ul, 10 uL was taken and run out on a 2% agarose gel to check for quality before cloning. Cloning and Sequencing of PCR Products PCR products were cloned using TA cloni ng with pGEM-T Easy Vector Systems (Promega). Ligations were performed as s uggested by the kit protoc ol. Ligations were transformed into chemically competent, subcloning efficiency XL-1 Blue (Stratagene)

PAGE 37

29 E.coli cells. Transformations were plated onto Luria-Bertani (LB) plates (1% tryptone, 0.5% yeast extract, 1% sodium chloride a nd 1.5% agar) containing 50 g/mL ampicillin and supplemented with 100 l of 100 mM IP TG and 40 l of 40 mg/mL X-Gal per plate (Gold Biotechnology, Inc) for blue-white co lony 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% tryp tone, 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 using the standard plas mid sequencing methods (described below). Sequences were analyzed using Sequ encher 4.2 (Gene Codes Corporation). Sequencing DNA sequencing was carried out usi ng 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 usi ng Sequencher 4.2 (Gene Codes Corporation). 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 (a pproximately 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 Pe rforma DTR Gel Filtration Columns (Edge Biosytems) and dried by vacuum centrifuga tion. Alternatively, sequencing reactions were purified by Ethanol pr ecipitation where 2 volumes of 100% ethanol and 1/2

PAGE 38

30 volumes of 7.5 mM ammonium acetate were ad ded to the sequencing reactions. Samples were washed with 70% ethanol a nd dried by vacuum centrifugation.

PAGE 39

31 CHAPTER 3 IMPRINTED TRANSGENE Introduction Imprinted loci are often or ganized as clusters of genes controlled by cis-acting regulatory elements called imprinting cente rs (IC) (Brannan and Bartolomei, 1999). These ICs can often be found large distances away from the genes they control making it difficult to localize and understand their mechan isms of action. Transgenes are excellent tools for the study of imprinted lo ci in that they serve to isol ate genes and their regulatory elements and relocate them to ectopic loca tions in the genome allowing for the study of those sequences without the influence of othe r adjacent elements. BAC transgenes have an additional advantage over conventional plasmid based tr ansgenes. Smaller plasmid transgenes are often influenced by the chroma tin into which they randomly insert. These insertional effects often result in inconsistencies in expression between individual transgenic mouse lines. The inherent greater size of BAC transgenes make them much less susceptible to position effects most likely due to the fact that they contain essential regulatory elements necessary for maintaining an open chromatin structure. For the study of imprinted loci, this greater size also incr eases the chance of inclusion of ICs and other necessary sequences required for proper expression patterns. BAC transgenes have been successfully used in the past for the study of imprinting. In one example Yevtodiyenko showed that a 178 kb BAC transgene expressed the Genetrap locus 2 (Gtl2) gene only upon maternal inheritance within the mouse embryo and placenta (Schmidt et al., 2000; Yevtodiyenko et al., 2004). This is similar to the

PAGE 40

32 endogenous locus on distal mouse chromosome 12 and human 14q32 suggesting that the 178 kb interval contains all the elements necessary to imprint Gtl2 They were also able to show that levels of expression were sim ilar to endogenous levels only in the brain but not other regions, localizing the enhancer elem ents necessary for brain expression but not other tissues to within the BAC. Lastly they showed that a second gene present within the BAC, Delta-like1 (Dlk1), was not expressed in an imprinted fashion, as is the endogenous locus, indicating that additiona l elements are required for imprinting Dlk1 The Surani lab reports using a 120 kb BAC transgene to localize imprinting elements required for paternal expression of the Peg3 gene which encodes a C2H2 type zinc-finger protein involved in maternal behavior (Li et al., 2000; Szeto et al., 2004). In this study they showed that imprinting is linked to diffe rential methylation of a region upstream of the gene. A second maternally expressed gene Zim1 located within the same cluster and present on the BAC was not imprinted correctly indicating that othe r cis acting elements are required for imprint regulation of the locus. The Surani lab also used BAC transgenes to locate dispersed cis-regulatory elemen ts necessary for paternal expression of Nueronatin (Nnat ) (John et al., 2001). By modifying the BAC they showed that these elements were located within an 80kb interval mostly upstream to Nnat As other labs have shown, BAC transgenes can be a powerful tool for localization of IC elements. In the case of the PWS locu s, current understanding of IC function in the mouse is limited to a 35 kb region in which th e PWS-IC is located. The AS-IC in the mouse has yet to be identified. Our lab has previously produced seve ral transgenic lines aimed at further elucidating the imprinting m achinery of the PWS locus (Figure 1). The first of these transgenic lines was devel oped from mouse and human phage (P1) clones

PAGE 41

33 carrying 85 and 76 kb of genomic sequence respectively (Blaydes et al., 1999). The mouse transgene contained 33 kb of 5 and 30 kb of 3 flanking DNA while the human clone carried 45kb of 5 and 7 kb of 3 flanki ng sequence. Five lines were produced from the human P1 clone, four of which were single copy and the fifth present as five copies. This clone was chosen due to the fact that it contained the complete interval in which both the PWS-IC and the AS-IC were located. All lines were expressed biallelically suggesting that the human PWS-IC, or at least the Snrpn promoter, is functional in the mouse, while the AS-IC is not. Two lines were derived from the mouse phage clone, a line containing a single copy of the transgen e and a line containing two copies. The single copy line was biallelic ally expressed while the two-copy line was correctly imprinted as it was expressed only upon pa ternal inheritance. The most likely explanation for this two-copy effect is that the mouse clone was large enough to contain the entire PWS-IC since it was capable of driving expression of Snrpn However the ASIC probably lies outside of this interval since the singl e copy line was unable to silence Snrpn expression upon maternal inheritance. The presence of a second copy somehow overcomes this deficiency, either by acting as a buffer to the insertion site effects of integration or it adds back some needed seque nces lost to the singl e copy line. This is similar to the effect seen in BAC transgenic studies of the maternally expressed H19 gene where a 14 kb transgene requires multiple copies to imprint, however a 130kb BAC transgene imprints in single copy (Bartolome i et al., 1993; Pfeifer et al., 1996). More recently the lab has created multiple transgenic lines using BAC transgenes which are much greater in size than the P1 clones. The Roswell Park RPCI-23 murine BAC library was screened for clones containing Snrpn Three were identified that carried

PAGE 42

34 sufficient sequence upstream of Snrpn ; 215A9 which extends 120 kb 5 and 20 kb 3, 380J10 which extends 8 kb 5 and 140 kb 3 and 425D18 which extends approximately 100 kb 5 and 65kb 3. These three BACS were injected as transgenes and were then tested for imprinted expression patterns (Figure3-1). The 380J10 transgene was expressed upon both maternal and paternal inheritance suggesting that the complete PWS-IC was present but the AS-IC was l acking. The 425 and 215 transgenes on the other hand showed correct imprinted expressi on. They were paternally expressed and maternally silenced. Intere stingly the 215B line suffered fr om a truncation that leaves only approximately 40 kb of sequence 5 to Snrpn This establishes the minimal genomic sequence necessary to establish imprinted e xpression in the locus as approximately 40 kb upstream and 20 kb downstream of Snrpn This project attempts to further delineat e the sequences necess ary to confer an imprinted expression pattern within the PW S locus. To do this a modified BAC transgene was developed in or der to simplify expression anal ysis. Later chapters will discuss additional modifications that were used to minimize the PWS-IC interval and locate the AS-IC. Results BAC Modification Previously, analysis of transgene expre ssion required a complex breeding scheme in order to distinguish the tr ansgenic allele from the endoge nous alleles. Transgenic founders were mated to mice carryin g a targeted deletion of exons 5 7 of Snrpn ( SmN) that results in a larger fusion transcript that include exons 1 4 and the neomycin resistance gene (Yang et al., 1998). In th is way the transgenic allele could be

PAGE 43

35 distinguished as a smaller tr anscript than the endogenous one. In order to simplify expression analysis, we decided to mark th e transgene in a way that it could be distinguished immediately. This was done us ing a technique called lambda red mediated homologous recombination developed by Dai guan Yu and Don Court at the National Cancer Institute (Lee et al., 2001). This met hod takes advantage of a defective lambda prophage from which the genes allowing entry into a lytic life cycle have been removed. What is left are only the minimal elements required to express the lambda Red genes exo, beta and gam. Exo and bet are required for recombining a double stranded (ds) linear DNA targeting construct into BAC D NA via homologous recombination. Exo encodes a 5-3 exonuclease that acts upon th e 5 ends of linear dsDNA to create 3 single stranded (ss) ends. B et encodes a pairing protein that binds the 3ssDNA ends and promotes annealing to the complementary strand on the BAC. The third gene, gam inhibits E. coli RecBCD exonuclease activity that destabilizes linear dsDNA. The defective prophage is integrated into the E. Coli DH10B (DY380) host bacterial chromosome and is under the control of the strong PL promoter. The promot er in turn is under the control of the temperature sensitive lambda cI857 repres sor. Transiently sh ifting the culture temperature from 32 C to 42 C results in expression of thes e three genes and subsequent homologous recombination between the lin ear targeting construct and BAC DNA. The 425D18 BAC was chosen for modificati on since a good amount of both 5 and 3 sequence was present. Since the prev ious SmN deletion had shown no effects on expression it was decided to create the same deletion on the BAC (Figure 3-2). Exons 5 7 of Snrpn were replaced with a floxed neomycin resistance cassette using a targeting construct (170 floxed neo) containing a 2 kb 5 arm of homology and a 6.2 kb 3 arm of

PAGE 44

36 homology. The targeting constr uct was linearized with Not I and then electroporated into induced DY380 recombination competent cells. Integration of the targeting construct was selected for by plating the transformed ce lls onto LB agar plates supplemented with neomycin and incubation overnight at 32oC. The next day 12 neomycin resistant colonies were picked and grown overnight in 3 mL of LB broth supplemented with neomycin at 32oC. After sufficient growth the cultures were subjected to alkaline lysis to extract the BAC DNA. BAC DNA was then cut with SacI restriction endonuc lease and Southern blotted. The blot was probed with a 2.2kb EcoRI/EcoRV genomic DNA fragment obtained from a restriction dige st of lab stock genomic DNA Snrpn clones. Owing to the extreme efficiency of this recombineering me thod, all of the clones picked were positive for the exon 5 7 deletion (Figure 3-2). Following successful targeting of Snrpn, the floxed neomycin cassette was to be removed by expression of Cre recombinase. To do this, the targeted 425 BAC (425 5 7neo) was transferred to an E. Coli host strain carrying the Cre recombinase gene under the co ntrol of an arabinos e inducible promoter (EL350). However, after Cre had been i nduced, a severe undesir ed truncation event occurred, making the BAC unusable (data not shown). Investigation into the composition of the BAC vector (pBACe3.6) into which the genomic DNA is cloned revealed a previously unknown third lox-p site. Induction of Cre result ed in recombination between one or both of the neomycin cassette flanking lox-ps with the third lox-p resulting in massive rearrangements. To deal with this a second targeting c onstruct, designed to replace the third lox-p site with a blas ticydin resistance cassette, was built. Recombination with this construct was induced as before and efficient removal of lox-p number three was achieved (data not shown). The doubly modified 425 5 7 BAC was

PAGE 45

37 again subjected to Cre mediated recombin ation. Successful deletion of the floxed neomycin cassette was determined by restric tion digest with EcoR I and SacI followed by Southern blot. The Southern blot was th en probed with the same 2.2kb EcoRI/EcoRV genomic DNA fragment as before (Figure). Production of Transgenic Mice Transgenic mice were produced to test of the suitability of the 425 5 7 BAC as a model system for the study of imprinting in th e PWS locus. If the modified BAC was to be a reliable system, the truncated Snrpn tran script it contains must be expressed only upon paternal inheritance. When inherited from the mother it should be silenced. BAC DNA was prepared for injection by large-scale alkaline lysis followed by cesium chloride banding. Both nicked and supercoiled DNA we re collected from the cesium prep and dialyzed against 4 liters of injection buffer (1.0 mM Tris pH7.5, .1 mM EDTA, 100 mM NaCl) for approximately 16 hours. After dial ysis, the BAC DNA was diluted to 2.5 ng/ul and injected into the male pronucleus of fe rtilized oocytes derived from superovulated FVB/N female mice. The injected oocytes were then implanted into pseudopregnant B6D2 F1 female mice. The resulting pups were allowed to mature to 3 weeks of age at which time they were ear punched for identi fication and tail clipped to obtain DNA for genotyping. Genomic tail DNA was then digested with AvaII and Southern blotted. The blot was then probed with a 1.5 kb genomic DNA HindII restriction digest fragment to both screen for the presence of the transgene and estimate copy number. Ten founder animals were obtained. Line A and D appeared to be single copy when compared to the endogenous band. Lines B, E, F, G1, H, I, and J appeared to be present in multiple copies. Line J was discarded, as it appear ed to contain an immediately observable

PAGE 46

38 rearrangement. Lines A, B, D, E, H, and I we re retained and mated as they seemed to contain the lowest numbers of copies. Li nes D and E failed to mate so they were discarded as well. The remaining four lines were analyzed for expression and methylation status. Analysis of 425 5 7 Transgenic Lines Since the modified BAC transgene coul d be easily distinguished from the endogenous locus, expression was easily test able after a single generation. Founder animals were initially mated to wild type FV B/N males and females to establish the lines and test for animals that may be chimeric. Chimerism results from late integration of the transgene into the reci pient zygotes genome after the firs t cleavage has occurred. If the cell that receives the late in tegration does not contribute to the founder animals germ line, the transgene will not be transmitted to fu ture generations and that line is worthless. All four lines transmitted the transgene (data not shown). Transgenic males and females from each line were next set up in matings with FVB/N males and females to test for expression of the transgene upon both maternal and paternal inheritance. Brains were collected from the resulting F2 day old ne onatal pups. RNA was extracted from the brains using RNAzol. The RNA was used as a template to produce cDNA that was then used to program pcr reacti ons that spanned exons 4 8 of Snrpn Since the modification to the 425 5 7 BAC removes exons 5 7 the resulting transgenic transcript appears as a band of approximately 350 base pairs (bp). This includes 242 bp of exonic sequence plus the additional lox-p left over from target ing. The endogenous band spans 513 bp. In all four lines transgenic Snrpn was expressed upon paternal in heritance and silenced upon maternal inheritance (Figure 3-3). This is in accord with expression of the endogenous

PAGE 47

39 alleles in which only paternal Snrpn is expressed. Interestingly, a faint signal was amplified from the transgene upon maternal inhe ritance. This will be discussed later on in this chapter. As a second readout of the imprinted status of the 425 5 7 transgene, the methylation status of Snrpn was analyzed using sodium bisulfite genomic sequencing, hereafter referred to as the bisulfite sequenc ing. This method uses sodium bisulfite to convert cytosines to uracils within genomic DNA. Methylat ed cytosines are protected from this conversion and remain as cytosi nes. Converted DNA is PCR amplified across the region of interest and the resulting products cloned and sequenced. When the sequence is read, methylated cytosines re main as cytosines, however unprotected cytosines appear as thymines. In this way the methylation status across an entire CpG island can be determined. Bisulfite sequencing was used to examine a CpG island present within the promoter region of Snrpn This region has been consistently shown to be hypomethylated on the paternal allele and hypermethylated on the maternal allele (Brannan and Bartolomei, 1999; Hershko et al., 2001). In order to distinguish between the transgenic and endogenous alleles, the 425 5 7 transgene was crossed onto a B6.Cast.c7 (Cast c7) background in order to take advantage of a single base pair polymorphism within Snrpn that can be differentiated during sequence analys is (Figure3-4). The Cast.c7 strain is a congenic line that contains a region of Mus musculus castaneous chromosome 7 on a Mus musculus domesticus C57BL/6J background. 425 5 7 transgenic males from the A and H lines were mated to Cast c7 females and their offspring screened for transmission of the transgene. Male and female N1, transgen e positive animals were then subjected to a

PAGE 48

40 second round of matings to the Cast c7 line. Brains from P1 neonatal offspring were collected from these matings and tail tips from each were then screened for both the transgene and Cast c7 homozygosity. Cast c7 homozygosity was determined by PCR amplification of the Necdin (Ndn) gene using the following primers: NdnpolyF 5-ACA AAG TAA GGA CCT GAG CGA CC3 A and NdnpolyR 5-CAA CAT CTT CTA TCC GTT CTT CG-3. The product was subseque ntly digested with the AvaII restriction endonuclease. Ndn contains a polymorphic cyto sine at residue 117 of the domesticus allele that results in recognition a nd subsequent cutting by AvaII. The castaneous allele has a thymine residue at position 117 and doe s not cut. Cast c7 homozygosity is therefore recognizable by a single castaneous band when electrophoresed on a 4.8% agarose gel. Four animals from the H line were identified that contained the 425 5 7 transgene and were homozygous for the Cast c7 allele. Brain DNA from two of these animals, one with a paternally transmitted transgene and the other with a maternally transmitted transgene, was subjected to bisulf ite conversion. Primers flanking the Cast c7 polymorphism were used to PCR amplify a 364 bp amplicon spanning the Snrpn promoter CpG island. Primer sequences for this region were as follows: W18 5-GTA GTA GGA ATG TTT AAG TA T TTT TTT TGG-3 and W1 9 5-CCA ATT CTC AAA AAT AAA AAT ATC TAA ATT-3 The products of this reaction were cloned and sequenced producing data for 15 clones repres enting a maternally tr ansmitted transgene (Figure 3-5) and 11 clones representing a pate rnally transmitted transgene (Figure3-6). Transgenic domesticus alleles were identified by a G at position 69 versus a T present in the endogenous castaneous alleles. In both cases 14 endogenous allele clones were sequenced as well. In summary, 11 of fi fteen maternal allele clones showed greater

PAGE 49

41 than 93% methylation while three were comple tely unmethylated. In contrast nine of eleven paternal allele clones were less th an 7% methylated while two clones were 35% and 93% methylated. The endogenous clones contained a mixture of both methylated and unmethylated clones as both maternal and paternal alleles were represented. Discussion As will be discussed in later chapters much work has gone into trying to understand the mechanism by which the PWS/AS locus is imprinted. Much of the data that has come out of this work is confusing and contradictory. As a means to simplify the system, we have created a BAC transgenic mouse model that contains 90 kb of upstream sequence and 65 kb of downstream sequence in relation to the Snrpn gene. In this way we can now study imprinting of Snrpn and its long transcript in isolation from the upstream gene cluster which, while it probabl y shares some regulatory elements, most likely has a different imprint mechanism. This BAC transgene model can now be modified by recombineering to determine where specific IC elements may lay. These initial proof of principal experime nts show that this model does indeed behave as the endogenous locus. Ex pression analysis shows that the 425 5 7 transgene is expressed when paternally i nherited and silenced when mate rnally inherited. This data shows that the interval contained within th e BAC is enough to confer correct imprinted expression upon Snrpn and it can be therefore be infe rred that all the cis-regulatory sequences, the PWS-IC and AS-IC, are presen t within. Additionally, bisulfite sequence analysis of the CpG island within the Snrpn promoter region shows a hypermethylated maternal allele and a hypomethyl ated paternal allele. This is also in agreement with what is currently known about the locus.

PAGE 50

42 Interestingly, analysis of expression from the transgene reveals a very low level of expression when maternally inherited. Bisulf ite analysis backs this up as 3 of the 15 clones sequenced were completely unmethylat ed and one can presume, expressed. This at first was concerning since it suggested that th e BAC did not contain enough sequence to silence the maternally inherited transgene at all times. However, previous data from our lab suggests that this may be the case for the endogenous locus as well. Analysis of endogenous expression of Snrpn on a PW S-IC deletion background shows leaky expression from the maternal allele that can be detected when the PWS-IC deletion is inherited from the father (Chamberlain et al ., 2004). It is important to note that this deletion also removed the Snrpn promoter, therefore expression could only be coming from the maternal allele. The bisulfite data is also important in this observation. The majority of maternally inherited clones were completely methylated and four others were at least 85%. The three unmethylated clone s were unlikely to be due to experimental artifact, the result of incomp lete conversion, since all other cytosines within the sequence were converted to thymines. This suggest s that low level leaky expression may be coming from cells that do not maintain the origin al imprint. Whether this is of biological significance or simply the result of sloppy main tenance within isolat ed cells is unknown. It is also possible that an as yet to be discovered small populat ion of cells within the brain may express Snrpn biallelically. However that is beyond the scope of this work. In conclusion, we have created a BAC tr ansgenic model that recapitulates the endogenous imprinted expression pattern of Snrpn This BAC contains all the regulatory elements required to establish and maintain th e imprint within its in terval. The next two

PAGE 51

43 chapters will describe modifications to the BAC that were used to further refine the location of the PWS-IC and locate the AS-IC.

PAGE 52

44 Figure 3-1. Schematic and expres sion of BAC transgenics. (A) Schematic representation of various BAC transgenic lines a nd their location in relationship to Snrpn .. (B) Northern blot analysis of BAC transgene expression. The 425A and 215B BAC transgenic lines express Snrpn when inherited from the father (C) but not when inherited from the mother. The 380 BAC transgene is expressed when maternally inherited and therefor e not imprinted. These lines are all derived from unmodified BACs.

PAGE 53

45 B. Figure 3-2. 425 5 7 targeting. (A) The relationship of the 425 5 7 targeting construct to Snrpn. A 2.2kb probe recognizes the 3e nd of a SacI restriction fragment. The unmodified BAC appears as a 16.8kb band. Replacement of exons with a neomycin resistance casse tte results in an additi onal SacI site and a 15.2kb band. Cre-lox removal of the neo ca ssette results in a 10.5kb band. (B) Southern blot shows efficient targeting of 12 clones with the 5 7 construct. A.

PAGE 54

46 Figure 3-3. Rt-PCR analysis of 425 5 7mouse lines. rt-PCR analysis of four transgenic mouse lines derived with the 425 5 7 BAC reveals proper imprinted expression. The Snrpn minigene cont ained on the transgene is expressed when paternally inherited and silenced when maternal ly inherited and is seen as a smaller band when compared to the endogenous band found in all lanes..

PAGE 55

47 Figure 3-4. Cast c7 breeding scheme. In order to distinguish the transgene from endogenous alleles in bisulfite analysis, the 425 5 7 transgene was crossed onto a Cast c7 background. This allows the two alleles to be distinguished by a single base pair s ubstitution within the Snrpn promoter region CpG island.

PAGE 56

48 Figure 3-5. Bisulfite sequence analysis of ma ternally transmitted transgene. Sequence analysis of genomic DNA converted by sodium bisulfite reveals that the majority of transgenic alleles are methylated at the Snrpn promoter CpG island when inherited from the moth er. A minority of alleles escape methylation, perhaps accounting for th e leaky expression seen from both the transgene and endogenous alleles.

PAGE 57

49 Figure 3-6. Bisulfite sequence analysis of pa ternally transmitted transgene. Sequence analysis of genomic DNA converted by sodium bisulfite reveals that the majority of transgenic alleles are unmethylated at the Snrpn promoter CpG island when inherited from the father.

PAGE 58

50 CHAPTER 4 REFINEMENT OF THE PWS-IC Introduction Considerable effort has been made to understand the cis elements that make up the PWS-IC. Much of this work is in the c ontext of targeted deletions of the region surrounding exon 1 of Snrpn. Initially the location of the PWS-IC was mapped to a region in humans that includes both the promoter and first exon of SNRPN This was accomplished by fine mapping of paternal microd eletions that resulted in PWS. The smallest region of overlap (SRO) between th ese deletions defined the location of the PWS-IC within humans (Saitoh et al., 1996). Deletion of this region is associated with absent expression of all the paternally expressed genes with in the locus. In 1998, our lab published a murine PWS-IC deletion model. This model contained a 35kb deletion of the syntenic region w ithin the mouse on ch romosome 7 centered around exon 1 of Snrpn (Yang et al., 1998) encompassing 16 kb of sequence 5 and 19 kb 3 to exon 1. Mice which inherit this deletion from the fath er lack expression of the paternal genes Mkrn3 Ndn and Ipw as well as the Snrpn long transcript, and exhibit several phenotypes common to PWS infants, in cluding poor suckle, failure to thrive, and lethality after approximately seven days. In contrast, mice inheriting a smaller intergenic Snrpn deletion that removed the 3 part of exon 5, exon 6 and the 5 part of exon 7 had no obvious phenotypic abnormalities when inherited fr om either the father or the mother. These data suggested that bot h the location and f unction of the PWS-IC were conserved

PAGE 59

51 between mouse and man and that the mouse c ould serve as a model with which to study the mechanism of imprinting within the PWS locus. With the goal of further refining the lo cation of the PWS-IC the Beaudet lab created two smaller deletions also centered on the first exon of Snrpn (Bressler et al., 2001). The first, encompassing 0.9 kb, remove d both the promoter and a portion of the Snrpn differentially methylated region 1 (D MR1). These mice were phenotypically normal upon either maternal or paternal inheri tance. Analysis of DNA methylation at both Snrpn and Ndn revealed no disturbance of the normal pattern as measured by methylation sensitive restriction digests. This suggests that the sequence elements necessary to confer imprinted expression upon the locus do not reside within this interval. The second deletion removed a larger 4.8 kb segment also centered on exon 1. These mice demonstrated a partial or mosaic phenotype with substantial postnatal lethality (~40 50%) and growth deficiency when the dele tion was inherited from the father. In contrast, maternal inheritance resulted in no phenotype as would be expected in a PWS IC deletion. Expression analysis was limite d to the upstream genes as the entire Snrpn promoter was removed in this deletion. Ndn and Mkrn3 showed reduced levels of expression and increased DNA me thylation when the deletion was paternally inherited. Although the partial imprinting defect makes thes e results difficult to interpret, one can make the assumption that some elements of the PWS-IC do lie within the deleted interval. The Yang lab recently documented several transcription binding sites within the Snrpn promoter region (RodriguezJato et al., 2005). Usi ng in-vivo footprinting and chromatin immunoprecipitation (ChIP) analysis they were able to show that several sites

PAGE 60

52 within this region are occupie d. Three footprints were show n to be active on the paternal allele. Footprint P2 appears to be a pot ential E2F binding site. This family of transcription factors participat es in transcriptional activa tion and suppression. Footprint P5 was shown to associate with NRF-1. Th is site is conserved between mouse and human and also overlaps with a previously repo rted seven base pair element (SBE), a cis acting element shown to be required for pr omoter activity (Hershko et al., 2001). Footprint P6/M6 was shown to be active on both the paternal and maternal alleles, although the patterns were dis tinctly different. P6/M6 co rrespond to overlapping NRF-1 and CTCF binding sites however only the NR F-1 site on the paternal allele was confirmed to be occupied by ChIP. Three ot her footprints were analyzed within the promoter but do not appear to be occupied or correspond to known transcription binding sites. The Yang lab also showed that elements with in the first intron of Snrpn were preferentially bound only on the paternal al lele. These three binding sites for the transcription factors NRF-1, YY1, and SP1 are phylogenetically conserved between mouse and man. NRF-1 and YY-1 have been shown to interact with chromatin modification enzymes suggestive of their possibl e role in maintaining or establishing the imprinted expression pattern in the locus. Th is agrees with the observation that this conserved intronic region has been shown to be associated with an open, active chromatin state. How these various factors may work to coordinate an imprinted expression pattern is unknown and additional work is n eeded to understand their roles. With the location of the PWS-IC established in the mouse, understanding the epigenetic changes that occur within the IC could provide insight in to the mechanism of imprinting at the PWS/AS locus. As discu ssed earlier, differential methylation at CpG

PAGE 61

53 islands is typically found with in imprinted loci. The PWS/AS locus is no exception. Two differentially methylated regions (DMRs ) exist within the PW S-IC, one within the promoter of Snrpn (DMR1) and one within the first intron (DMR2) (Shemer et al., 1997). These DMRs are hypomethylated on the pate rnal allele and hypermethylated on the maternal allele. Methylation of CpGs within the PWS-IC of mice occurs during oogenesis, however in humans it occurs after fertilization suggesting th at this is not the primary imprint(El-Maarri et al., 2001). It has also been shown that parent specific differential methylation of hist one H3 exists in the same region (Xin et al., 2001). H3 Lys9 is methylated on the maternal copy of the PWS-IC, while H3 Lys4 is methylated on the paternal copy. H3 Lys-9 methylation is correlated with transcriptional silencing while H3 Lys4 methylation is correlated with transcriptional activity suggesting a possible epigenetic switch. Furthermore CpG methylation has been tied to the G9a H3 Lys9/Lys27 methyltransferase (Xin et al., 2003) Its been hypothesized that H3 Lys9 methylation could be the primary imprint a nd may trigger CpG methylation of the PWSIC through G9a leading to maternal silencing of the paternal genes on the maternal allele after fertilization. This chapter discusses modifications to the 425 5 7 BAC designed to further elucidate the PWS-IC. Results BAC Modification As a means to further define the minimal cis elements necessary to establish and maintain imprinted expression in the locus I ha ve created a series of 3 nested deletions within the imprinted 425 5 7 BAC (Figure 4-1). The 5 anchor point of all three

PAGE 62

54 deletions was positioned at a pproximately 30kb upstream of Snrpn exon 1. This point was selected because targeted deletions made within the lab that extend an additional 12 kb further upstream of this point had proven to have no effect on imprinting (Peery et al., manuscript in preparation) The largest deletion, 425 30kb, extends to 133 bp upstream of exon 1, removing all but the smallest minimal promoter (SMP) of Snrpn (Hershko et al., 2001). This point was selected as to allow for transcription of the Snrpn minigene reporter. The smallest deletion, 425 17kb, extends to 11.5kb upstream of exon 1 and is approximately 17kb in length. This point was selected based on reports from the Beaudet lab of a 100kb targeted deletion anchored at the 3 end at the same position (Wu et al., 2006). As will be discussed in the next chapter, this deletion had no effect on imprinting. The third deletion, 425 23kb, extends to approximately 6kb upstream of exon 1 and is 23kb in length. This was selected as a midw ay point between the largest and smallest deletions. The 425 5 7 BAC was targeted to produce these modifications using the Lambda recombineering system previously discussed. Production of Transgenic Mice The 425 30kb BAC was chosen as the first constr uct to inject as it was supposed that this deletion would have the greatest effect. Fresh 425 30kb BAC DNA was prepared as previously described and was in jected into the pronuclei of fertilized FVB oocytes at a concentration of 2.5 ng/ul. Inject ed oocytes were transp lanted at the two cell stage into the oviducts of pseudopregnant B6D2 F1 mice. The resulting pups were allowed to mature to 3 weeks of age at which time they were ear punched for identification and tail clipped to obt ain DNA for genotyping. Genomic tail DNA was used as a template for PCR amplification of the transgene to screen for founders. The

PAGE 63

55 primers T7F 5GTA ATA CGA CTC ACT AT A GGG C 3 and T7R1 5 CTC CAA TCA TGT TCA ACT GTC C 3 amplify a regi on of the BAC adjacent to the vector arms and are transgene specific. Five founders were obtained; lines C, D, E, F, and G. Of these only lines C, D, and E transmitted to their offspring. Lines F and G were discarded. Analysis of 425 30kb Transgenic Lines Founder animals were mated to FVB/N ma les or females to establish breeding colonies and test for chimerism. Lines C, D, and E transmitted through the germ line and the resulting F1 transgenic pups were allowed to mature to six weeks of age. At six weeks, F1 transgenic males and females from each line were set up in matings with FVB/N males and females to test for expres sion of the transgene upon both maternal and paternal inheritance. Brains were collected from the resulting F2 day old neonatal pups. RNA was extracted from the br ains using RNAzol. The RNA was used as a template to produce cDNA that was then used to progr am PCR reactions that spanned exons 4 8 of Snrpn Since the modification to the 425 5 7 BAC removes exons 5 7 the resulting transgenic transcript appear s as a band of approximately 350 base pairs (bp). This includes 242 bp of exonic sequence plus the a dditional lox-p left ove r from targeting. The endogenous band spans 513 bp. Unexpectedly, the 425 30kb transgene was silenced in lines D, and E when maternally inherited. Line D never transmitted the transgene paternally however line E expre ssed the Snrpn reporter gene upon paternal transmission. Line C expressed the transgene biallelically, suggesting that a linearization of the transgene had occurred at a point incompatible with imprinted expression.

PAGE 64

56 Bisulfite sequencing was performed on the C and E lines to test for methylation at the Snrpn DMR1 as was done with the 425 5 7 lines. However, the 30kb deletion made it possible to amplify DMR1 without first backcrossing the transg ene onto the Cast.c7 line of mice. The resulting bisulfite am plicon retains 10 of th e original 15 CG dinucleotides present within the CpG is land. Twenty five micrograms of DNA of genomic brain DNA from the C and E lines wa s bisulfite converted and then used to program PCR reactions with the following primer set; 425.06 bisulf F2 5GTT TTA GGA GTT RGA GGT TTT TT 3 and 425.06 bi sulf R2 5 AAC CCA AAT CTA AAA TAT TTT AAT C 3. The products fr om these reactions were cloned and sequenced. Data from 11 clones representing a paternally transmitted transgene and 9 clones representing a maternally transmitted transgene was obtained from the E line. In summary, all paternally transmitted E alleles were completely unmethylated. In contrast, five of nine maternal E alleles were comple tely methylated. Two of nine alleles were methylated at nine sites. On e of nine alleles was methylated at only site, and one allele was completely unmethylated. Discussion The lack of data from multiple lines ma kes drawing strong conclusions difficult for these experiments. However, it seems highl y unlikely that imprinted expression of the 425 30kb transgene from line E could be due to an insertional effect. Therefore some inferences can be made. First, line E expressed the Snrpn reporter in a correctly imprinted fashion and line D was silenced when maternally inherited. This suggests that all the PWS-IC cis-regulatory elements necessary for imprinting Snrpn are within the region between intron 1 and 133bp upstream of exon 1. Since imprinting was preserved

PAGE 65

57 in this, the longest deletion of the three construct built, the other two BACs were not injected. The footprint and ChIP data reported by RodriguezJato et-al demonstrate that there are conserved elements within this regi on that are actively occupied, some in a parent of origin specific manner (Rodriguez-Jato et al., 2005). These include NRF-1, YY1, and SP1 sites located with in intron 1 and an NRF-1 site just upstream of exon one that had previously been show to be nece ssary for promoter activity (Hershko et al., 2001). It can also be presumed that the AS-IC is still intact within th is construct since the transgene is silenced upon maternal transmission. A likel y location would be upstream of the deletion since such a small interval of sequence remains downstream. A second observation is that the mechanism governing the imprinted expression of Snrpn likely differs from that governing the upstream cluster of genes. As just mentioned, Bressler et-al showed the regi on encompassing the first 600 bp upstream of Snrpn exon 1 was dispensable for imprinting as judged by expression and methylation of Ndn Only when an additional 4kb of sequence upstream was deleted did they see an imprinting defect. Although expression of the upstream genes is not measurable in this transgenic system, its fairly obvious that this more proximal region is important for imprinting Snrpn since only 133bp of upstream sequen ce confer imprinted expression on the E line. It seems possible that maternal silencing of the Snrpn long transcript may simply be due to methylation of the prom oter, thereby blocking access to trans-acting factors required for transcription. Imprin ting of the upstream genes however, would require a longer range mechanism, perhaps sim ilar to the active chromatin hub seen in the beta-globin locus (Patrinos et al., 2004; Zhou et al., 2006). Looping of the upstream genes to the PWS-IC has already been de monstrated in human cell culture (Arnold

PAGE 66

58 Heggestad personal communication) The cis-regulatory elements that would control this type of mechanism have yet to be identified. A third point that can be made co mes from the observation that the 425 30kb C line expresses transgenic Snrpn in a biallelic fashion. Since the AS-IC is the element that silences the maternal allele, mapping the point at which the BAC linearized before integration may provide clues to the location of this element.

PAGE 67

59 Figure4-1. Schematic diagram of three nest ed deletions aimed at minimizing the PWSIC. Three nested deletions anchored at the 5 end 30 kb upstream of Snrpn, were designed to decrease the minima l known interval required to imprint Snrpn. The largest of the three extend s to 133bp upstream of exon 1 or just 5 to the smallest minimal promoter. The shortest deletion extends 17kb to a 3 anchor point reported by the Beaudet lab. The third deletion extends to a midpoint between the two.

PAGE 68

60 Figure 4-2. rt-PCR analysis of 425 30kb mouse lines. rt-PCR analys is of two transgenic mouse lines derived with the 425 30kb BAC reveals proper imprinted expression of the E line and bialleli c expression of the C line. Proper expression of the E line suggests all cis-regulatory elements required for imprinted expression of Snrpn are contai ned within its interval. Biallelic expression of line C may be due to linearization of the BAC that is incompatible with imprinted expression.

PAGE 69

61 Figure 4-3. Bisulfite sequence analysis of maternally transmitted 425 30kb transgene. Sequence analysis of genomic DNA convert ed by sodium bisulfite reveals that the majority of transgenic al leles are methylated at the Snrpn promoter CpG island when inherited from the moth er. A minority of alleles escape methylation, perhaps accounting for th e leaky expression seen from both the transgene and endogenous alleles. When the transgene is paternally inherited all transgenic alleles are completely demethylated.

PAGE 70

62 CHAPTER 5 DEFINING THE LOCATION OF THE MURINE AS-IC Introduction Imprinting in the PWS/AS locus is de pendent upon coordinate control by a bipartite imprinting center, consisting of a PWS-IC that drives expression from the paternal allele and an AS-IC that is responsible for silenc ing the PWS-IC on the maternal allele. While the position of the AS-IC has been localized to a region approximately 35kb upstream of Snrpn exon 1 in humans, its location in the mouse is still unknown, hampering efforts to understand its mechanism. Targeted deletions in the mouse, based on a similar distance from Snrpn in the human, failed to disturb the differential methylation and imprinted expression patterns of Snrpn and Ndn suggesting that the location is not conserved. This chapter will focus on efforts to define the murine AS-IC For years it has been postulate d that silencing of the ma ternal allele may involve expression of a series of alternative SNRP N transcripts that arise from a set of infrequently used upstream exons. The firs t evidence for this came from analysis of Angelman Syndrome patients inheriting microde letions in a region upstream of exon 1. A consequence of these deletions is a block in the paternal to maternal imprint switch during gametogenesis resulting in a paternal epigenotype on the maternal chromosome and biallelic expression of the paternal se t of genes. Several upstream alternative SNRPN exons were then isolated from this region (D ittrich et al., 1996; Fa rber et al., 1999) which extends >700kb upstream of exon 1 (Figure 51). These upstream exons share a degree of sequence similarity to each ot her and in turn are also sim ilar to exon one, implying that

PAGE 71

63 they arose from duplications of exon one. A lternative transcripts initiate from the paternal allele from two of these exons, u1A and u1B and give rise to various splice variants. All splice variants skip exon one and splice to exon two. Maternal specific methylation is also present in the regions upstream of these exons. What is most interesting about these upstream exons is that exon u5 is deleted in all AS patients with an IC deletion. Exon u5 is contained within multiple alternative transcripts so the argument could be made that these transcript s could be somehow invol ved in the paternal to maternal imprint switch. An alternative explanation is that th e location of exon u5 is merely a coincidence and the ac tual AS-IC is a cis regulatory element located in close proximity. An argument in favor of this hypothe sis is that no altern ative transcripts to date have been identified which initiate at exon u5. If transcription of u5 were necessary for AS-IC function, patients w ith deletions encompassing exons u1A or u1B should be present since these are the exons at which those tr anscripts identified to da te initiate. It is just as likely however that as yet unidentified transcripts that do initiate at u5 are present within developing gametes, tissue which has not yet been tested. Studying these issues in humans is complicated by the fact that the tissues in which the imprints are being established, developing gametes, are difficult if not impossible to acquire. As discussed earlier, gene content and the imprin ted expression pattern within the PWS/AS locus is conserved in the m ouse making it an attractive model system. However the location of the murine AS-IC is still not known. Given that the AS-IC is located at approximately 35kb upstream of SNRPN in humans our lab targeted deletions at the corresponding location in the mouse. Nested deletions of 7 and 12kb had no effect on methylation or expression of the patern ally expressed genes suggesting that the

PAGE 72

64 location of the AS-IC relative to Snrpn is not conserved in th e mouse (E. Peery personal communication). Closer examination of this region reveals that the mouse also contains several exons upstream to Snrpn that extend approximately 500kb upstream. Rt-pcr analysis reveals that these upstream exons are also involve d in alternative transcripts similar to humans (Bressler et al., 2001; Lande rs et al., 2004). Nine U exons have been identified to date. They participate in multiple splice variants that fall into two categories, those that join to Snrpn exon 2 a nd those that exclude Snrpn and splice further downstream to the non-coding Ipw exons. As is the case in humans, it appears that the upstream exons are the result of multiple duplication events as there is a high degree of homology to the region surrounding exon 1. Sequence is highly conserved among the murine exons, more so than in humans, a pproaching 90%, however there is very little conservation between human and mouse. The function that these al ternative exons play in imprinting is yet to be determined. In fact conflicting results from two labs either supports or excludes alternativ e transcripts from playing a role. Extensive work by LeMeur et-al characterized the temporal and spatial regulation of these transcripts (Le Meur et al., 2005). Tr anscripts containing various U e xons were found to be expressed within embryos 10.5 days post-coitum (dpc) a nd later exclusively within the nervous system. This group however failed to detect expression of the U exons within testis and ovary samples suggesting that alternative transcri pts are not produced within developing gametes. This would preclude a role for alte rnative transcripts in establishing imprints within the PWS/AS locus during a time wh en these marks are placed. Mapendano et-al however did detect expression of alternativ e transcripts arising from exon U2 and splicing to Snrpn exon 2 within the ovary by rt-pcr (Mapendano et al., 2006). In-situ

PAGE 73

65 hybridization, utilizing exon U2 as a probe, detected expr ession within both oocytes and their supporting granulosa cells within Graafian follicles. Stronger signals were detected within oocytes and granulosa cells of sec ondary and developing follicles. While this contradicts the observations by Le Meur et -al, it supports the id ea that alternative transcripts may be important to imprint establishment. In a recently published report, Wu et al targeted deletions to the murine region upstream of Snrpn (Wu et al., 2006). Two lines of mice were generated mutating the region 5 of the Snrpn promoter. The firs t was an anchor mutation that introduced a targeting vector 13kb upstream of Snrpn exon 1. The insertion was comprised of the 3 portion of an Hprt selectable cassette, a loxP site, a puromycin selectable marker, and a 6kb duplication of genomic sequence. This mutation is referred to as AS-ICan. The second mutation was an 80kb deletion of geno mic sequence extending upstream of the 3 anchor created by AS-ICan. Paternal inheritance of eith er mutation resulted in a normal pattern of methylation at the Snrpn and Ndn CpG islands. Maternal inheritance of the AS-ICan mutation resulted in an imprint defect with both maternal and paternal alleles being unmethylated at Snrpn. This was correla ted with inappropriate maternal expression of Snrpn. Ndn was also affected with a part ial or mosaic pattern of methylation on the maternal allele. All subsequent breeding of AS-ICan mutant mice resulted in an AS like imprinting defect, with inappropriate hypomethyl ation of the maternal allele. Maternal inheritance of the 80kb deleti on resulted in an imprinting defect with incomplete penetrance. Three groups of mice with diffe rent methylation patterns were observed. The first (3 of 8 heterozygous mice) exhi bited a normal pattern of methylation as compared to wild type litter-mates. A sec ond group (2 of 8) showed partial methylation

PAGE 74

66 of the maternal allele. The third group of three was completely unmethylated. In trying to understand the disruption of imprinting in bo th mutant lines, it is important to note that the puromycin cassette is pr esent and transcribes in the opposite orientation to Snrpn in the AS-ICan mutant. In the 80kb deletion, the purom ycin cassette is deleted by cre-lox, but the Hprt promoter is present and expre sses in the same direction as Snrpn In the context of the upstream exons, the anchor in sertion alters the distance between the putative AS-IC and the PWS-IC and perhap s the distance between the two regulatory elements is important to coordinating imprin ted expression. A second explanation is that transcription from the puromycin cassette interferes with alternate transcripts from the upstream exons. This second explanation may also be valid in the case of the 80kb deletion. Transcription towards Snrpn from the Hprt promoter may interfere with the normal transcription signal being sent from upstream. More evidence regarding the location of th e AS-IC comes from our lab in the form of imprinted transgenes (Figure 5-2). Multip le transgenic lines have been created and characterized in our lab extending variable distances upstream of Snrpn As mentioned earlier, a phage clone carryi ng 33kb of murine sequence 5 to Snrpn did not imprint correctly as a single copy transg ene suggesting that the AS-IC is not within that interval. The 425 5 7 BAC transgene however does imprint correctly suggesting that the AS-IC is somewhere within the 100kb interval upstr eam of Snrpn. The 215B BAC transgenic line also is imprinted correctly. What is intere sting about this line is that it has suffered a truncation at approximately 40kb 5 to Snrpn. The location of mouse upstream exon 1 is at 43kb 5 to Snrpn. Whether the 215B line does contain U1 is unknown and is currently being investigated.

PAGE 75

67 These observations have led to the followi ng hypothesis (Figure 5-3). First, the upstream exons or an element nearby act as the AS-IC and establishes the primary imprint upon the PWS-IC. Second, the high degree of homology between the murine upstream exons allows for compensation to occur when one is deleted. This may explain the variable penetrance seen in the 80kb deletion. Exons fu rther upstream may not be as efficient in a role as the AS-IC as the closer U1. Lastly, only deletion of all U exons will result in a complete AS like phenotype. This would prove to be difficult in a targeted deletion strategy as it may require individually targeting each of nine identical sequences over a distance greater than 500kb. The BAC transgenic model provides an excellent opportunity to ask these question as only thr ee upstream exons are present within the 425 5 7 BAC interval. In order to inves tigate the role that murine upstream Snrpn exons play in imprinted expression of the locus several modifications to the 425 5 7 BAC were made. This chapter characterizes those modifications in an attempt to understand AS-IC function. Results BAC modification The 425 5 7 BAC transgene contains 3 upstr eam exons, U1, U2, and U3 at 43kb, 75kb, and 96 kb upstream of Snrpn exon 1 respectively. In order to test their role in ASIC function, a 2499 bp region of homology in which each U exon is imbedded, was targeted for deletion in a way as to le ave the surrounding DNA intact. This 2.5kb region of homology likely represents the region of Snrpn exon 1 duplicated to form the U exons and may contain cis regulatory element requ ired for silencing the PWS-IC. Targeting vectors to be used to delete each upstr eam exon by BAC recombineering were designed

PAGE 76

68 as follows. Primers were designed to fla nk each region of homology with approximately 1kb of additional sequence in eith er direction. The primer sets used are as follows; U1 U1delF1 5GAC TTC AGT ACA TGT TC C AAC and U1delR2 5TAG ATA GTC CAG TAC CAG GAC T , U2 U 2delF 5CCA GAA GAT GTT CCA ACT GG and U2delR 5CCA CCT AAC C AA TGG ACA GA 3, U3 U3delF1 5CCA GAT AGT GTA CCT CAT GTC a nd U3delR1 5GAG GCT TCC ACA TAA GGT CTT . The products amplified with these primers were cloned into the pGEM T easy cloning vector from Promega. In order to delete the region containing each U exon, restriction endonuclease cleavage sites were identified and used to remove the portion of sequence containing the regions of homology leaving 1kb of genomic sequence in either direction intact. The U1 T-easy vector was cleaved with AflII and PacI to release a 2753bp fragment. The U2 T-easy vector was cleaved with BamHI to release a 2515bp fragment. The U3 T-easy vector was cleav ed with BssHII and Csp45I to release a 2608bp fragment. Each linearized vector was blunted with Klenow and a loxP flanked neomycin resistance cassette was ligated in. In order to avoid cro ss-reaction between the various targeted areas when multiple modifications were made to a single BAC, mutant loxP sites were used to flank the neo cassett e in the U2 and U3 targeting vectors. FRT sites were used to flank the neo cassette in the U1 targeting vector. Lambda red recombineering was used to make targeted deletions of the three U exons in the 425 5 7 BAC using these vectors. Transgene 425 U1-U3 has all three U exons deleted (Figure 5-4A). If our compensation hypothesis is correct all three exons must be deleted to see a complete imprinting defect. Transgene 425 U2/U3 has only U2 and U3 deleted (Figure 5-4B). This is important to show that e xon U1, or the region immedi ately adjacent to it,

PAGE 77

69 serves as the AS-IC. Two more constructs, 425 U1/U3 and 425 U1/U2 leave exon U2 or U3 intact in order to demonstrate compensation. Production of Transgenic Mice The 425 U1-U3 and 425 U2/U3 BACs were chosen as the first constructs to inject. Fresh BAC DNA was prepar ed as previously described and was injected into the pronuclei of fertilized FVB oocytes at a concen tration of 2.5 ng/ul. In jected oocytes were transplanted at the two cell st age into the oviducts of pseudopregnant B6D2 F1 mice. The resulting pups were allowed to mature to 3 weeks of age at which time they were ear punched for identification and tail clippe d to obtain DNA for genotyping. Genomic tail DNA was used as a template for pcr amplificatio n of the transgene to screen for founders. The primers T7F 5GTA ATA CGA CTC AC T ATA GGG C 3 and T7R1 5 CTC CAA TCA TGT TCA ACT GTC C 3 amplify a region of the BAC adjacent to the vector arms and are transgene specific. Four founders were obtained for the 425 U1-U3 line and two founders were obtained for the 425 U2/U3 line. All were found to transmit the transgene and were consequently mated to Cast.c7 males and females to establish colonies. Analysis of 425 U1-U3 and 425 U2/U3 Transgenic Lines Founder animals were mated to Castc7 ma les or females to establish breeding colonies. Lines A, B, C, and D transmitted through the germ line and the resulting N1 transgenic pups were allowed to mature to six weeks of age. At six weeks, N1 transgenic males and females from each line were set up in matings with FVB/N males and females to test for expression of the transgene upon both maternal and pate rnal inheritance. Brains were collected from the resulting N2 day old neonatal pups. RNA was extracted

PAGE 78

70 from the brains using RNAzol The RNA was used as a template to produce cDNA that was then used to program pcr reactions that spanned exons 4 8 of Snrpn Since the modification to the 425 5 7 BAC removes exons 5 7 the resulting transgenic transcript appears as a band of approximately 350 base pairs (bp). This includes 242 bp of exonic sequence plus the additional lox-p left over from targeting. The endogenous band spans 513 bp. Line A did not express the transgene upon paternal or mate rnal inheritance, possibly due to insertional eff ects or breakage of the transgene at a point incompatible to expression prior to insertion. Line D on the other hand expressed the transgene from both the paternal and maternal alleles, sugg esting that the AS-IC had been deleted by removing the upstream exons. Lines B and C have only been tested for maternal expression at this point. Line B expressed th e maternally inherited transgene while line C did not. Conclusions on these two lines ca nnot be made until expression data from the paternally inherited alleles can be analyzed Discussion Due to the overlap of human Snrpn exon u5 with the AS-SRO, it has been hypothesized that the upstream exons have some role in AS-IC function. However, testing this hypothesis in the mouse has proven to be difficu lt, possibly due to the high degree of homology among the nine murine upstream exons. Using our BAC transgenic model provides a simpler model since only three U exons are contained within its interval. Since this data set is not yet complete, no firm conclusions can be made, however biallelic expression from the 425 U1-U3 D line and maternal expression from the B line suggest that the U exons do play a role in si lencing the maternal al lele. Expression from

PAGE 79

71 the B and C paternally inherited alleles remain s to be analyzed and this will prove to be important in determining the imprinting status of this construct. Also it will be important to determine whether or not the BAC remained intact during integrati on. If the transgene is indeed intact throughout the region 5 of Snrpn, this rais es the possibility that the U exons also serve some sort of insulating purpose. All four 425 5-7 transgenic lines proved to be capable of sile ncing maternally inherited Snrpn suggesting that the BAC contains sufficient sequence to overcome th e chromatin into which it inserts. The 425 U1-U3 lines do not seem to have this capabi lity as the A line is completely silenced upon both paternal and maternal inheritance. Southern blots spa nning the region 5 to Snrpn are currently underway to determine the whether the tr ansgenes have remained intact. The 425 U2-U3 lines still remain to be analyzed. Determining their imprinting status will provide conclusive evidence a bout whether the AS-IC resides within the upstream exons. Cloning of additional BAC constructs 425 U1/U3 and 425 U1-U2 has been completed however they have not yet been injected as transgenes. Data from these lines should provide eviden ce about whether U2 and U3 can compensate for AS-IC function when U1 is deleted. The data presented in this chapter provi des preliminary evidence that the AS-IC resides within the upstream exons. However, this data does not suggest a mechanism by which the AS-IC imprints the maternal alle le. Two possibilities immediately come to mind. First and probably the most obvious is th at transcription factor binding sites lie within the region that includes U1. At this point in time however, th is would be difficult

PAGE 80

72 to test. Additional work would need to be done to further limit the interval in which the AS-IC is contained in order for an efficient search for active binding sites could occur. A second intriguing hypothesi s is that active transc ription from the upstream promoter is what triggers heterochromatin formation and subsequent silencing of the maternal allele. There is precedence for this type of mechanism in multiple systems. First, Mayer et al have shown that transcripts from intergenic spacer sequences between rRNA genes are important for establishing a nd maintaining a specific heterochromatin conformation at the promoter of a subset of rDNA arrays (Mayer et al., 2006). These 150-300 nucleotide transcripts interact with TIP5, the large subun it of the chromatin remodeling complex NoRC and guide the complex to its target via sequence complementarity to the rDNA promoter. A seco nd example comes from the analysis of a patient suffering from -thalesemia, an inherited form of anemia. This patient was found to have a deletion that resulted in th e juxtaposition a structurally normal -globin gene (HBA2) to LUC7L (Tufarelli et al., 2003) Expression of LUC7L from the opposite strand from HBA2 resulted in an anti-sense transcript that was capable of mediating methylation and silencing of HBA2 during development. Through the use of transgenic mice they were also able to show that met hylation and modification of chromatin occurs in cis. A third mechanism by which transcrip tion could result in gene silencing is siRNA mediated DNA methylation. Until recently th is mechanism had only been known to be active in plants. Kawasaki et al were able to show that mammalian cells are also capable of siRNA mediated silencing within a cell culture model (K awasaki and Taira, 2004). Vector based siRNAs targeting CpG islands within the E-cadherin and erbB2 genes were

PAGE 81

73 shown to be capable of inducing DNA methylat ion and histone H3 lysine 9 methylation resulting in repression of transcription. The work contained in this chapter is ongoing and the data presented preliminary. Expression from the paternal allele of the 425 U1-U3 lines B and C transgenes and analysis of the 425 U2-U3 transgene remains to be done. Also, determination of the breakpoints for these transgenes will be neces sary before any firm conclusions can be made.

PAGE 82

74 Figure 5-1. Schematic diagram of human upstream exons. The PWS-IC region has undergone duplication events resulti ng in the establishment of several upstream exons. Low levels of paternal transcripts arise from these exons as shown by the arrow heads. Lollipops re present the differential methylation status at specific methyl ation sensitive restriction sites. Filled circles represent methylated CpGs while em pty circles represent unmethylated CpGs. (adapted from Farber et al.)

PAGE 83

75 Figure 5-2. Schematic diagram of upstream e xons in the mouse. Upstream exons have also arisen in the mouse via duplica tions of the PWS-IC region. Brain specific expression has been characterized from the paternal allele as shown by the black arrows. Oocyte expression has also been shown as illustrated by the red arrows. Several transgenic li nes have been developed which provide evidence that the upstream exons may play a role in imprinting. The shorter P1 clone does not contain any upstream exons and does not imprint in single copy while the two larger BAC clones c ontain one or three upstream exons and do imprint.

PAGE 84

76 Figure 5-3. Compensation model of AS-IC activ ity. Alternative exons further upstream can compensate for loss of those cl oser to Snrpn. Only by deleting all upstream exons will there be an imprinting defect.

PAGE 85

77 Figure 5-4. Schematic diagram of upstream exon deletions. In order to elicit an AS-IC imprinting defect it may be necessary to delete all upstream exons to avoid compensation by those that remain.

PAGE 86

78 Figure 5-5. RT-PCR analysis of 425 U1-U3 mouse lines. Preliminary RT-PCR analysis of four transgenic mouse lines derived with the 425 U1-U3 BAC reveals improper maternal expression of Snrpn in lines B and D. Line A does not express the transgene upon either maternal or paternal inhe ritance. Line C which has only been analyzed for mate rnal expression at this point does not express from the maternally inherited allele. Initial fi ndings suggest that deletion of the upstream e xons blocks imprinting of Snrpn on the maternal allele.

PAGE 87

79 CHAPTER 6 CONCLUSIONS AND FUTURE DIRECTIONS The mechanism by which imprinted expression occurs within the Prader-Willi / Angelman locus is currently unknown. Several animal models have been put forth in order to further our understanding of this phenomenon, however the regulatory elements that coordinate parental speci fic expression of this set of genes remains a mystery. This work describes a new transgenic model, which will hopefully further our understanding of the locus. BAC transgenes have historically been shown to be good models of the genomic loci, which they represent. Our BAC tran sgenic model faithfully recapitulates the imprinted expression of the endogenous Snrpn therefore it must contain all the cis regulatory elements required for establis hment and maintenance of the imprint. The 425 30kb modified BAC shortens the inte rval known to contain the PWS-IC considerably and leaves us with a manageab le amount of sequence within which to look for the cis elements of the PWS-IC. Data fr om Rodriguez-Jato et al has already shown that conserved transcription bindi ng sites within this interval are occupied in a parent of origin specific manner and may play a ro le. Further modifications to the 425 5 7 BAC whereby these individual sites are mutated will help to identify those that are important. The inability to identify th e AS-IC in the mouse has hampered efforts to understand how it silences the PWS-IC. Observations in humans have suggested that alternative transcripts arising from exons upstream of Snrpn exon 1 may play a role, however definitive evidence supporting this hypothesis has yet to be shown. The BAC transgenic

PAGE 88

80 model makes it possible to study these upstream exons in isolation from the rest of the locus. Preliminary evidence presented here suggests that exon U1 or the sequence in within the duplicated region may serve as the AS-IC. The mechanism by which this works is still not understood. It is possible that elements within this region attract transacting factors to the locus. These factors could then create a silenced chromatin conformation which spreads outward to silen ce the PWS-IC and thereby the paternally expressed genes on the maternal allele. A second compelling hypothesis is that transcription arising from the upstream exons regulates the imprint establishment at the PWS-IC. Transcription of non-coding RNA has b een shown to regulate epigenetic states using a variety of mechanisms in both pl ant and mammalian cells. Determination of whether the AS-IC functions via attraction of transacting factors or through a transcription based mechanism is well within the capabilities of this model system. Experiments are currently under way to insert an oocyte specific promoter in front of Snrpn exon 1 within a non-imprinted BAC. It w ill be interesting to see if transcription through the PWS-IC within the context of a growing oocyte will confer imprinted status to this transgene. Other obvious directions are to take a promote r bashing approach, deleting larger and larger portions of e xon U1 to reveal the relevant sequences. The work contained within this document will certainly add to the understanding of imprinting within the PWS/AS, however unde rstanding how the locus is controlled has further implications. Mouse models for PWS and AS will be important in developing treatments for these disorders. Understandi ng the elements that make up the PWS-IC and AS-IC will be important in developing for developing these models. .Many imprinted loci are contained within the genome, but only in a small portion is the mechanism of

PAGE 89

81 parental gene regulation understood. Furthe ring the understanding of this locus may provide clues to how imprinting works at ot her regions. Also as the possibility of therapeutic cloning becomes a reality, understanding gene regulation will become essential. Disrupted imprinting has been suggest ed to play a role in the low survival rate of animals cloned by somatic cell nuclear transfer It is therefore likely that the genetic reprogramming necessary to clone therapeutic tissues may also prove to be highly error prone. Understanding how imprinting is contro lled will be importa nt to develop these therapeutic technologies. There have also been reports of disrupted imprinting in children conceived by intracytoplasmic sperm in jection (ICSI). As assisted reproduction techniques become more commonplace, unde rstanding how imprints are acquired and maintained will be important in reducing th e risks associated with these procedures.

PAGE 90

82 LIST OF REFERENCES Amos-Landgraf, J. M., Ji, Y., Gottlieb, W., Depinet, T., Wandstrat, A. E., Cassidy, S. B., Driscoll, D. J., Rogan, P. K., Schwartz, S. and Nicholls, R. D. (1999). Chromosome breakage in the Prader-Willi and Angelman syndromes involves recombination between large, transcribed re peats at proximal a nd distal breakpoints. Am J Hum Genet 65 370-86. Bartolomei, M. S., Webber, A. L., Brunkow, M. E. and Tilghman, S. M. (1993). Epigenetic mechanisms underlying the imprinting of the mouse H19 gene. Genes Dev 7 1663-73. Barton, S. C., Surani, M. A. and Norris, M. L. (1984). Role of paternal and maternal genomes in mouse development. Nature 311 374-6. Beaudet, A. L. and Jiang, Y. H. (2002). A rheostat model for a rapid and reversible form of imprinting-dependent evolution. Am J Hum Genet 70 1389-97. Blaydes, S. M., Elmore, M., Yang, T. and Brannan, C. I. (1999). Analysis of murine Snrpn and human SNRPN gene im printing in transgenic mice. Mamm Genome 10 54955. Boccaccio, I., Glatt-Deeley, H., Watrin, F., Roeckel, N., Lalande, M. and Muscatelli, F. (1999). The human MAGEL2 gene and its mo use homologue are paternally expressed and mapped to the Prader-Willi region. Hum Mol Genet 8 2497-505. Brannan, C. I. and Bartolomei, M. S. (1999). Mechanisms of genomic imprinting. Curr Opin Genet Dev 9 164-70. Bressler, J., Tsai, T. F., Wu, M. Y., Tsai, S. F., Ramirez, M. A., Armstrong, D. and Beaudet, A. L. (2001). The SNRPN promoter is not required for genomic imprinting of the Prader-Willi/Angelman domain in mice. Nat Genet 28 232-40. Buiting, K., Lich, C., Cottrell, S., Barnicoat, A. and Horsthemke, B. (1999). A 5-kb imprinting center deletion in a family with Angelman syndrome reduces the shortest region of deletion overlap to 880 bp. Hum Genet 105 665-6. Buiting, K., Saitoh, S., Gross, S., Dittrich, B., Schwartz, S., Nicholls, R. D. and Horsthemke, B. (1995). Inherited microdeletions in the Angelman and Prader-Willi syndromes define an imprinting centre on human chromosome 15. Nat Genet 9 395-400.

PAGE 91

83 Butler, M. G., Bittel, D. C., Kibiryeva, N., Talebizadeh, Z. and Thompson, T. (2004). Behavioral differences among subjects with Prader-Willi syndrome and type I or type II deletion and maternal disomy. Pediatrics 113 565-73. Butler, M. G. and Palmer, C. G. (1983). Parental origin of chromosome 15 deletion in Prader-Willi syndrome. Lancet 1 1285-6. Cassidy, S. B. and Ledbetter, D. H. (1989). Prader-Willi syndrome. Neurol Clin 7 3754. Cattanach, B. M. (1986). Parental origin effects in mice. J Embryol Exp Morphol 97 Suppl 137-50. Cattanach, B. M. and Kirk, M. (1985). Differential activ ity of maternally and paternally derived chromosome regions in mice. Nature 315 496-8. Chai, J. H., Locke, D. P., Ohta, T ., Greally, J. M. and Nicholls, R. D. (2001). Retrotransposed genes such as Frat3 in the mouse Chromosome 7C Prader-Willi syndrome region acquire the imprinted status of their insertion site. Mamm Genome 12 813-21. Chamberlain, S. J. and Brannan, C. I. (2001). The Prader-Willi syndrome imprinting center activates the paternally expressed muri ne Ube3a antisense transcript but represses paternal Ube3a. Genomics 73 316-22. Chamberlain, S. J., Johnstone, K. A., DuBose A. J., Simon, T. A., Bartolomei, M. S., Resnick, J. L. and Brannan, C. I. (2004). Evidence for genetic modifiers of postnatal lethality in PWS-IC deletion mice. Hum Mol Genet 13 2971-7. Church, G. M. and Gilbert, W. (1985). The genomic sequencing technigque. In Prog. Clin. Biol. Res. vol. 177 (ed., pp. 17-21. Clark, S. J., Harrison, J., Paul, C. L. and Frommer, M. (1994). High sensitivity mapping of methylated cytosines. In Nucleic Acids Res vol. 22 (ed., pp. 2990-7. Clayton-Smith, J. and Pembrey, M. E. (1992). Angelman syndrome. J Med Genet 29 412-5. de los Santos, T., Schweizer, J., Rees, C. A. and Francke, U. (2000). Small evolutionarily conserved RNA, resembling C/D box small nucle olar RNA, is transcribed from PWCR1, a novel imprinted gene in the Prader-Willi deletion region, which Is highly expressed in brain. Am J Hum Genet 67 1067-82.

PAGE 92

84 Dittrich, B., Buiting, K., Korn, B., Rickard, S., Buxton, J., Saitoh, S., Nicholls, R. D., Poustka, A., Winterpacht, A., Zabel, B. et al. (1996). Imprint switching on human chromosome 15 may involve alternativ e transcripts of the SNRPN gene. Nat Genet 14 163-70. Driscoll, D. J. (1994). Genomic imprinting in humans. Mol Genet Med 4 37-77. El-Maarri, O., Buiting, K., Peery, E. G., Kroisel, P. M., Balaban, B., Wagner, K., Urman, B., Heyd, J., Lich, C., Brannan, C. I. et al. (2001). Maternal methylation imprints on human chromosome 15 are es tablished during or after fertilization. Nat Genet 27 341-4. Farber, C., Dittrich, B., Buiting, K. and Horsthemke, B. (1999). The chromosome 15 imprinting centre (IC) region has undergone mu ltiple duplication events and contains an upstream exon of SNRPN that is deleted in all Angelman syndrome patients with an IC microdeletion. Hum Mol Genet 8 337-43. Gray, T. A., Saitoh, S. and Nicholls, R. D. (1999). An imprinted, mammalian bicistronic transcript encode s two independent proteins. Proc Natl Acad Sci U S A 96 5616-21. Hata, K., Okano, M., Lei, H. and Li, E. (2002). Dnmt3L cooperates with the Dnmt3 family of de novo DNA methyltransferases to establish maternal imprints in mice. Development 129 1983-93. Hershko, A. Y., Finberg, Y., Kantor, B., Shemer, R. and Razin, A. (2001). The mouse Snrpn minimal promoter and its human orthologue: activity and imprinting. Genes Cells 6 967-75. Holm, V. A., Cassidy, S. B., Butler, M. G., Hanchett, J. M., Greenswag, L. R., Whitman, B. Y. and Greenberg, F. (1993). Prader-Willi syndrome: consensus diagnostic criteria. Pediatrics 91 398-402. Horsthemke, B., Dittrich, B. and Buiting, K. (1997). Imprinting mutations on human chromosome 15. Hum Mutat 10 329-37. Howell, C. Y., Bestor, T. H., Ding, F., La tham, 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 104 829-38. Jacobs, P. A., Wilson, C. M., Sprenkle, J. A., Rosenshein, N. B. and Migeon, B. R. (1980). Mechanism of origin of complete hydatidiform moles. Nature 286 714-6.

PAGE 93

85 John, R. M., Aparicio, S. A., Ainscough, J. F., Arney, K. L., Khosla, S., Hawker, K., Hilton, K. J., Barton, S. C. and Surani, M. A. (2001). Imprinted expression of neuronatin from modified BAC transgenes reveals regulation by di stinct and distant enhancers. Dev Biol 236 387-99. Jong, M. T., Carey, A. H., Caldwell, K. A., Lau, M. H., Handel, M. A., Driscoll, D. J., Stewart, C. L., Rinchik, E. M. and Nicholls, R. D. (1999a). Imprinting of a RING zinc-finger encoding gene in the mouse chromosome region homologous to the PraderWilli syndrome genetic region. Hum Mol Genet 8 795-803. Jong, M. T., Gray, T. A., Ji, Y., Glenn, C. C., Saitoh, S., Driscoll, D. J. and Nicholls, R. D. (1999b). A novel imprinted gene, encodi ng a RING zinc-finger protein, and overlapping antisense transcript in the Prader-Willi syndrome critical region. Hum Mol Genet 8 783-93. Kawasaki, H. and Taira, K. (2004). Induction of DNA meth ylation and gene silencing by short interfering RNAs in human cells. Nature 431 211-7. Kishino, T., Lalande, M. and Wagstaff, J. (1997). UBE3A/E6-A P mutations cause Angelman syndrome. Nat Genet 15 70-3. Knoll, J. H., Glatt, K. A., Nicholls, R. D., Malcolm, S. and Lalande, M. (1991). Chromosome 15 uniparental disomy is not frequent in Angelman syndrome. Am J Hum Genet 48 16-21. Knoll, J. H., Nicholls, R. D. and Lalande, M. (1989a). On the parental origin of the deletion in Angelman syndrome. Hum Genet 83 205-7. Knoll, J. H., Nicholls, R. D., Magenis, R. E., Graham, J. M., Jr., Lalande, M. and Latt, S. A. (1989b). Angelman and Prader-Willi syndromes share a common chromosome 15 deletion but differ in parental origin of the deletion. Am J Med Genet 32 285-90. Landers, M., Bancescu, D. L., Le Meur, E., Rougeulle, C., Glatt-Deeley, H., Brannan, C., Muscatelli, F. and Lalande, M. (2004). Regulation of the large (approximately 1000 kb) imprinted murine Ube3 a antisense transcri pt by alternative exons upstream of Snurf/Snrpn. Nucleic Acids Res 32 3480-92. Le Meur, E., Watrin, F., Landers, M., Stur ny, R., Lalande, M. and Muscatelli, F. (2005). Dynamic developmental regulation of the large non-coding RNA associated with the mouse 7C imprinted chromosomal region. Dev Biol 286 587-600. Ledbetter, D. H., Riccardi, V. M., Airhart, S. D., Strobel, R. J., Keenan, B. S. and Crawford, J. D. (1981). Deletions of chromosome 15 as a cause of the Prader-Willi syndrome. N Engl J Med 304 325-9.

PAGE 94

86 Lee, E. C., Yu, D., Martinez de Velasco, J., Tessarollo, L., Swing, D. A., Court, D. L., Jenkins, N. A. and Copeland, N. G. (2001). A highly efficient Escherichia coli-based chromosome engineering system adapted for recombinogenic targeti ng and subcloning of BAC DNA. Genomics 73 56-65. Lee, S., Kozlov, S., Hernandez, L., Chamberlai n, S. J., Brannan, C. I., Stewart, C. L. and Wevrick, R. (2000). Expression and imprinting of MAGEL2 suggest a role in Prader-willi syndrome and the hom ologous murine imprinting phenotype. Hum Mol Genet 9 1813-9. Leff, S. E., Brannan, C. I., Reed, M. L., Ozcelik, T., Francke, U., Copeland, N. G. and Jenkins, N. A. (1992). Maternal imprinting of th e mouse Snrpn gene and conserved linkage homology with the human Prader-Willi syndrome region. Nat Genet 2 259-64. Li, L. L., Szeto, I. Y., Cattanach, B. M., Ishino, F. and Surani, M. A. (2000). Organization and parent-of-orig in-specific methylation of im printed Peg3 gene on mouse proximal chromosome 7. Genomics 63 333-40. Linder, D., McCaw, B. K. and Hecht, F. (1975). Parthenogenic origin of benign ovarian teratomas. N Engl J Med 292 63-6. Lossie, A. C., Whitney, M. M., Amidon, D., Dong, H. J., Chen, P., Theriaque, D., Hutson, A., Nicholls, R. D., Zori, R. T., Williams, C. A. et al. (2001). Distinct phenotypes distinguish the molecular classes of Angelman syndrome. J Med Genet 38 834-45. MacDonald, H. R. and Wevrick, R. (1997). The necdin gene is deleted in Prader-Willi syndrome and is imprinted in human and mouse. Hum Mol Genet 6 1873-8. Magenis, R. E., Brown, M. G., Lacy, D. A., Budden, S. and LaFranchi, S. (1987). Is Angelman syndrome an altern ate result of del(15)(q11q13)? Am J Med Genet 28 829-38. Malcolm, S., Clayton-Smith, J., Nichols, M., Robb, S., Webb, T., Armour, J. A., Jeffreys, A. J. and Pembrey, M. E. (1991). Uniparental paternal disomy in Angelman's syndrome. Lancet 337 694-7. Mapendano, C. K., Kishino, T., Miyazaki, K., Kondo, S., Yoshiura, K., Hishikawa, Y., Koji, T., Niikawa, N. and Ohta, T. (2006). Expression of the Snurf-Snrpn IC transcript in the oocyte and its putative role in the imprinting establishment of the mouse 7C imprinting domain. J Hum Genet 51 236-43. Mayer, C., Schmitz, K. M., Li, J., Grummt, I. and Santoro, R. (2006). Intergenic transcripts regulate the epigen etic state of rRNA genes. Mol Cell 22 351-61.

PAGE 95

87 McGrath, J. and Solter, D. (1984). Completion of mouse embryogenesis requires both the maternal and paternal genomes. Cell 37 179-83. Moore, T. and Haig, D. (1991). Genomic imprinting in mammalian development: a parental tug-of-war. Trends Genet 7 45-9. Nicholls, R. D., Knoll, J. H., Butler, M. G., Karam, S. and Lalande, M. (1989). Genetic imprinting suggested by maternal heterodisomy in nondeletion Prader-Willi syndrome. Nature 342 281-5. Ohama, K., Nomura, K., Okamoto, E., Fu kuda, Y., Ihara, T. and Fujiwara, A. (1985). Origin of immature teratoma of the ovary. Am J Obstet Gynecol 152 896-900. Patrinos, G. P., de Krom, M., de Boer, E., Langeveld, A., Imam, A. M., Strouboulis, J., de Laat, W. and Grosveld, F. G. (2004). Multiple interact ions between regulatory regions are required to stab ilize an active chromatin hub. Genes Dev 18 1495-509. Pfeifer, K., Leighton, P. A. and Tilghman, S. M. (1996). The structural H19 gene is required for transgene imprinting. Proc Natl Acad Sci U S A 93 13876-83. Reik, W. and Dean, W. (2001). DNA methylation and mammalian epigenetics. Electrophoresis 22 2838-43. Rodriguez-Jato, S., Nicholls, R. D., Driscoll, D. J. and Yang, T. P. (2005). Characterization of cisand trans-acting elements in the imprinted human SNURFSNRPN locus. Nucleic Acids Res 33 4740-53. Runte, M., Huttenhofer, A., Gross, S., Kie fmann, M., Horsthemke, B. and Buiting, K. (2001). The IC-SNURF-SNRPN transcript serves as a host for multiple small nucleolar RNA species and as an antisense RNA for UBE3A. Hum Mol Genet 10 2687700. Saitoh, S., Buiting, K., Rogan, P. K., Buxton, J. L., Driscoll, D. J., Arnemann, J., Konig, R., Malcolm, S., Horsthemke, B. and Nicholls, R. D. (1996). Minimal definition of the imprinting center and fi xation of chromosome 15q11-q13 epigenotype by imprinting mutations. Proc Natl Acad Sci U S A 93 7811-5. Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Cl oning, (ed.: Cold Spring Harbor Laboratory Press. Schmidt, J. V., Matteson, P. G., Jones, B. K., Guan, X. J. and Tilghman, S. M. (2000). The Dlk1 and Gtl2 genes are linked and reciprocally imprinted. Genes Dev 14 1997-2002.

PAGE 96

88 Searle, A. G. and Beechey, C. V. (1978). Complementation studies with mouse translocations. Cytogenet Cell Genet 20 282-303. Shemer, R., Birger, Y., Ri ggs, A. D. and Razin, A. (1997). Structure of the imprinted mouse Snrpn gene and establishment of its parental-specific methylation pattern. Proc Natl Acad Sci U S A 94 10267-72. Sutcliffe, J. S., Jiang, Y. H., Galijaard, R. J., Matsuura, T., Fang, P., Kubota, T., Christian, S. L., Bressler, J., Catta nach, B., Ledbetter, D. H. et al. (1997). The E6-Ap ubiquitin-protein ligase (UBE3A) gene is localized within a narrowed Angelman syndrome critical region. Genome Res 7 368-77. Szeto, I. Y., Barton, S. C., Keverne, E. B. and Surani, A. M. (2004). Analysis of imprinted murine Peg3 locus in transgenic mice. Mamm Genome 15 284-95. Tufarelli, C., Stanley, J. A., Garrick, D., Sharpe, J. A., Ayyub, H., Wood, W. G. and Higgs, D. R. (2003). Transcription of antisense RNA leading to gene silencing and methylation as a novel cause of human genetic disease. Nat Genet 34 157-65. Varmuza, S. and Mann, M. (1994). Genomic imprinting-defusing the ovarian time bomb. Trends Genet 10 118-23. Wu, M. Y., Chen, K. S., Bressler, J., Hou, A., Tsai, T. F. and Beaudet, A. L. (2006). Mouse imprinting defect mutations that model Angelman syndrome. Genesis 44 12-22. Xin, Z., Allis, C. D. and Wagstaff, J. (2001). Parent-specific co mplementary patterns of histone H3 lysine 9 and H3 lysine 4 met hylation at the Prader-Willi syndrome imprinting center. Am J Hum Genet 69 1389-94. Xin, Z., Tachibana, M., Guggiari, M., He ard, E., Shinkai, Y. and Wagstaff, J. (2003). Role of histone methyltransferase G9 a in CpG methylation of the Prader-Willi syndrome imprinting center. J Biol Chem 278 14996-5000. Yang, T., Adamson, T. E., Resnick, J. L., Le ff, S., Wevrick, R., Francke, U., Jenkins, N. A., Copeland, N. G. and Brannan, C. I. (1998). A mouse model for Prader-Willi syndrome imprinting-centre mutations. Nat Genet 19 25-31. Yevtodiyenko, A., Steshina, E. Y., Farner, S. C., Levorse, J. M. and Schmidt, J. V. (2004). A 178-kb BAC transgene imprints the mouse Gtl2 gene and localizes tissuespecific regulatory elements. Genomics 84 277-87. Zhou, G. L., Xin, L., Song, W., Di, L. J., Liu, G., Wu, X. S., Liu, D. P. and Liang, C. C. (2006). Active chromatin hub of the mouse al pha-globin locus forms in a transcription factory of clustered housekeeping genes. Mol Cell Biol 26 5096-105.

PAGE 97

89 BIOGRAPHICAL SKETCH Chris Futtner was born in Hartford Conn ecticut in 1971. He graduated with honors from Eastbury Elementary School at the age of ten; it was then that he became certain that he would become a famous baseball play er. Unfortunately, baseball didnt work out for him, so he moved to Ft. Collins, CO, were he received a BS in microbiology at Colorado State University in the year 1998. Settling down and earning large sums of money did not appeal to Chris so he decided to return to school and pursue a PhD at the University of Florida which he received in May of 2007. With his degree, Chris hopes one day to become a contestant on the Wheel of Fortune, and then parlay his winnings to become a cattle baron.


Permanent Link: http://ufdc.ufl.edu/UFE0018082/00001

Material Information

Title: Minimal Sequences Necessary for Imprinted Expression of the Prader-Willi Locus
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0018082:00001

Permanent Link: http://ufdc.ufl.edu/UFE0018082/00001

Material Information

Title: Minimal Sequences Necessary for Imprinted Expression of the Prader-Willi Locus
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0018082:00001


This item has the following downloads:


Full Text











MINIMAL SEQUENCES NECESSARY FOR IMPRINTED EXPRESSION OF THE
PRADER-WILLI LOCUS














By

CHRISTOPHER FUTTNER


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2007
































Copyright 2007

by

Christopher R. Futtner
















ACKNOWLEDGMENTS

I would like to thank both past and present members of the lab, including Stormy

Chamberlain, Edwin Peery, Jessica Walrath, Mike Elmore, Amanda DuBose, Emily

Smith, Lori Kellam, and my favorite Scotswoman, Karen Johnstone.

I would also like to thank my mentor Jim Resnick, not only for his scientific

expertise, but also for his excellent advise on why old school is better than new school,

how to have a happy marriage, and why the Yankees are the best team in baseball.

I would like to thank my parents for their love and encouragement. Without them I

would be biologically impossible.

I am also extremely thankful to Cami Brannan. The experiments contained within

this work would not have been possible without her ideas and vision.

Lastly, I would like to thank my lovely wife, Danielle Maatouk, whose scientific

coattails I will be riding to greatness.




















TABLE OF CONTENTS


page

ACKNOWLEDGMENT S ................. ................. iii........ ....


LIST OF FIGURES .............. ....................vi


AB STRAC T ................ .............. vii


CHAPTER


1 INTRODUCTION ................. ...............1.......... ......


Genomic Imprinting............... .. ..............
Prader-Willi and Angelman Syndromes ................. ...............4................
Purpose of Imprinting ................. ...............6............ ....
The PWS/AS Locus ................. ...............8.......__ ....


2 MATERIALS AND METHODS .............. ...............17....


Lambda Red Recombineering .............. .... ...............17..
Transformation of Recombineering Strains ................... ........ ...........1
Transformation of Targeting Vector and Induction of Recombination ........._.....19
BAC preparation for inj section ........._.. ........ .. ...............20..
Generation of Transgenic Animals by Pronuclear Inj section ........._.._.. ......._.._.....23
Genotyping ................. .. .. ...............2
Identification of Transgenic Founders ....__. ................. ........__. ........23
Identification of Castaneous C7 Homozygotes .................. ................2
Southern BI ot ................. ...............24....... ......
RNA Isolation ............ ............ ...............25....
R T -PC R ................. ........ .. .... .. ... ..... .. ... ...... .........2
Preparation of High Molecular Weight Genomic DNA for Bisulfite Conversion .....26
Bisulphite Sequencing of Genomic DNA............... ...............27..
Bisulphite Conversion .............. .. ...............27...
Bisulphite Polymerase Chain Reaction .............. ...............27....
Purification of PCR Products ............__............ .....___............2
Cloning and Sequencing of PCR Products ....._____ ............ .............. .28
Sequencing............... ...............2

3 IMPRINTED TRANSGENE............... ...............3













Introducti on ................. ...............3.. 1..............
Re sults ................. ........... ...............34.......
BAC Modification ................. ...............34........... ....

Production of Transgenic Mice .............. ...............37....

Analysis of 425A5-7 Transgenic Lines ................ ............. .................38
Discussion ........._.__....... .__ ...............41....


4 REFINEMENT OF THE PWS-IC............... ...............50.


Introducti on ........._.__....... .__ ...............50....
Re sults........._._.... ......_ __ ...............53.....
BAC Modification ........._.__....... .__. ...............53....

Production of Transgenic Mice .............. ...............54....

Analysis of 425A30Okb Transgenic Lines .....__.___ ...... .._. __ ......_..........5 5
Discussion ........._... ...... ._ ._ ...............56....


5 DEFININING THE LOCATION OF THE AS-IC ........................... ...............62


Introducti on ................. ...............62.................
Re sults ................ ............ ...............67.......
BAC modification ............... ...............67....

Production of Transgenic Mice ............... ..... .... ...........6

Analysis of 425AUl-U3 and 425AU2/U3 Transgenic Lines ........._.._... .............69
Discussion ............. ...... ._ ...............70....


6 CONCLUSIIONS AND FUTURE DIRECTIONS ................ .........................79


LIST OF REFERENCES .....__ ................. ...............82......


BIOGRAPHICAL SKETCH .............. ...............89....


















LIST OF FIGURES


Figure pg

1-1 The life cycle of an imprint. .............. ...............12....

1-2 Mapping the Imprinting Center (IC). ............. ...............13.....

1-3 Molecular classes of PWS and AS. ................ ...............14........... ..

1-4 Gene organization of the PWS/AS locus.. ........._.. ......_. ......_. ............15

1-5 Paternal only model of imprinting regulation.. ........._.. ...._.. ......._.......16

3-1 Schematic and expression of BAC transgenics ................. ......... ................44

3-2 425A5-7 targeting. ............. ...............45.....

3-3 rt-PCR analysis s of 425A5-7mouse lines ................. ...............46........... ..

3-4 Cast c7 breeding scheme.. ............ ...............47.....

3-5 Bisulfite sequence analysis of maternally transmitted transgene. ................... .........48

3-6 Bisulfite sequence analysis of paternally transmitted transgene ................... ...........49

4-1 Schematic diagram of three nested deletions aimed at minimizing the PWS-IC.....59

4-2 rt-PCR analysis of 425A30Okb mouse lines. ....._____ .... ......... ..........__......6

4-3 Bisulfite sequence analysis of maternally transmitted 425A30kb transgene. ..........61

5-1 Schematic diagram of human upstream exons ...._.._.._ .... ... ..._. ........_.._.....74

5-2 Schematic diagram of upstream exons in the mouse. ........._.._.. ........__. ........75

5-3 Compensation model of AS-IC activity. ............. ...............76.....

5-4 Schematic diagram of upstream exon deletions. ........._.._.. ......._ ........._.....77

5-5 RT-PCR analysis of 425AUl-U3 mouse lines. ....._____ ............ ................78
















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

MINIMAL SEQUENCES NECESSARY FOR IMPRINTED EXPRESSION OF THE
PRADER-WILLI LOCUS

By

CHRISTOPHER FUTTNER

May 2007

Chair: James Resnick
Major Department: Medical Sciences~enetics

Prader-Willi (PWS) ands Angelman (AS) syndromes are neurodevelopmenal

disorders arising from the improper expression of oppositely imprinted genes located on

human chromosome 15 qll1-ql3. Imprint regulation of this region is under the control of

a bi-partite imprinting center consisting of an Angelman Imprinting Center (AS-IC)

located approximately 35kb upstream of the paternally expressed Snrpn exon 1 and a

Prader-Willi Imprinting Center (PWS-IC) located 5' to and including Snrpn exon 1. The

PWS-IC has been shown to be a positive element promoting expression of a set of genes

on the paternal allele, while the AS-IC provides suppression of the PWS-IC on the

maternal allele thereby suppressing expression of the same set of genes and allowing

expression of the maternal program. Both are required for proper establishment and/or

maintenance of the imprint in the germline. In the mouse, both gene order and the

imprinted expression pattern have been conserved, with the syntenic region being located










on murine chromosome 7C. While the location of the PWS-IC has also been conserved,

the position of the murine AS-IC remains unknown.

We have taken a transgenic approach to locating the AS-IC and further dissecting

out components of the PWS-IC. Using a recombineering method that utilizes the lambda

phage Red genes, we have created a series of deletions within a Snrpn containing

bacterial artificial chromosome (BAC) that we have shown recapitulates the imprinted

expression of the endogenous locus. First we have created a deletion of 30 kb just 5' to

Snrpn exon 1 and show that imprinted expression is retained. This shows that the

components required for establishment and/ or maintenance of the imprint on Snrpn are

located within the retained sequence. Second, we have created a series of deletions that

remove several alternative upstream exons of Snrpn in various combinations and show

that they contain the AS-IC element and each can act as an independent AS-IC when the

other are deleted. These findings are significant because they further restrict the minimal

necessary imprinting center functional unit and may suggest a possible mechanism of

action.















CHAPTER 1
INTTRODUCTION

Genomic Imprinting

A consequence of sexual reproduction in mammals is the inheritance of both a

maternal and paternal set of genes alleless), which are virtually identical in sequence and

for the most part expressed at similar levels. However, a certain subset of genes is

expressed from only one allele with the other being silenced. This phenomenon has been

termed genomic imprinting. Imprinting is defined as an epigenetic phenomenon whereby

otherwise identical parental alleles are differentially expressed in a parent of origin

specific manner. This difference in functionality is brought about by heritable marks

placed upon the DNA without changes to gene sequence.

The first evidence that the paternal and maternal genetic contributions were not

functionally equal came from the labs of Solter (McGrath and Solter, 1984) and Surani

(Barton et al., 1984). Via pronuclear transplantation experiments, both labs independently

came to the conclusion that embryos derived from two paternal pronuclei (biparental

androgenones) or two maternal pronucleii (biparental gynogenones) were incapable of

completing normal embryogenesis. Zygotes derived from two maternal genomes resulted

in some embryonic growth but lacked the necessary extraembryonic tissues necessary for

implantation. This is similar to the formation of ovarian teratomas in humans, tumors

derived from the spontaneous parthenogenesis of female germ cells. Like the

gynogenetic zygote, ovarian teratomas develop the various embryonic tissues but lack

extraembryonic ones (Linder et al., 1975; Ohama et al., 1985). Biparental andregenones









on the other hand develop substantial extraembryonic tissues but lack a significant

embryo. The human equivalent to this is a hydatidiform mole, a concepts with a

completely paternal genome. Complete hydatidiform moles are characterized by

hypertrophy of the trophoblast and a complete lack of embryonic tissues (Jacobs et al.,

1980).

Further evidence for the inequality of the maternal and paternal genetic

contribution came from the labs of Cattanach (Cattanach, 1986; Cattanach and Kirk,

1985) and Searle (Searle and Beechey, 1978). By crossing mice heterozygous for certain

Robertsonian translocations, they were able to create nondisjunction events that resulted

in maternal or paternal duplications in specific regions of individual chromosomes with

the corresponding loss of the other parents' alleles. Many duplications proved to have no

effect as is the case with chromosomes 1, 4, 5, 9, 13, 14, and 15. However, duplications

on several other chromosomes resulted in embryonic lethality or other developmental

abnormalities. This showed that the genomic parental nonequivalancy was a regional

phenomenon rather than a global one and suggested a role for imprinted genes.

Central to the topic of genomic imprinting is the idea of an "epigenetic mark."

This mark contains information about the expression status of a particular allele.

Importantly however it does not involve a change in sequence between the maternal and

paternal allele. While the effects of imprinting can be seen in gene expression, the actual

imprint or mark that designates a gene to be imprinted is still unknown. What is known

however is that the imprint has three phases in its life cycle:

1. establishment The mark must be placed within cells of the developing germline in
order to identify the allele as expressed or silenced.

2. maintenance The mark must be stable through mitosis so that it is propagated
throughout all the somatic cells of the growing embryo.









3. erasure The mark must be initially erased as cells of the germline are first
developing so that a new mark may be established according to the sex of the
growing embryo (Figure 1-1).

Current evidence suggests that chemical modifications such as DNA methylation and

histone modifications are responsible for distinguishing maternal and paternal alleles.

DNA methylation possesses these features making it an attractive candidate for the

imprint mark (Reik and Dean, 2001). First, methylation at imprinted loci is typically

restricted to the allele of only one parent. These methylation marks are typically

established during gametogenesis and survive the genome wide wave of demethylation

that occurs during the first cleavage stages of embryogenesis. This differential

methylation is maintained during the global de-novo methylation that occurs post

implantation. Second, DNA replication results in unmethylated CpG dinucleotides on the

daughter strand. These hemimethylated sites quickly become fully methylated by specific

DNA methyltransferases (DNMTs) thereby retaining the mark during DNA synthesis and

restricting it to previously methylated sites. Third, methylation of imprinted loci is

erased during gametogenesis and subsequently re-established during maturation of the

oocyte and spermatocyte, resetting the methylation mark to that of the sex of the

developing embryo. Evidence supporting methylation as the mark comes in the form of

Dnmt1 (Howell et al., 2001) and Dnmt3L (Hata et al., 2002) knockouts that result in

biallelic expression of previously imprinted genes .

It has been suggested that histone modifications may also be involved in

maintenance and establishment of differential expression of imprinted genes. Differential

methylation patterns of histones associated with the PWS-IC and SNRPN have been

reported (Xin et al., 2001) and histone methyltransferases (Xin et al., 2003) have been

shown to be required for maintenance of DNA methylation of the PWS-IC CpG island in









cell culture. This type of mark could be both erasable and heritable fulfilling the

requirements of an imprinting mark.

Prader-Willi and Angelman Syndromes

Prader-Willi (PWS) and Angelman (AS) syndromes are both neuro-developmental

disorders linked to the same region of chromosome 15. Both occur at a frequency of

approximately 1 in 15,000 live births. Initially, typical PWS patients exhibit neonatal

hypotonia, poor suckle, failure to thrive, hypogonadism, and cryptorchidism. After the

first few years of life the hypotonia and failure to thrive give way to obesity and

hyperphagia (a strong desire to eat). PWS is also characterized by moderate mental

retardation and ob sessive-compulsive di sorder (Holm et al., 1993) (Cassidy and

Ledbetter, 1989). Obesity, complicated by a decreased caloric requirement and

hyperphagia, is a maj or cause of morbidity among PWS patients. AS on the other hand is

characterized by severe mental retardation, an ataxic gait, inappropriate laughter, happy

affect, and almost absent speech (Clayton-Smith and Pembrey, 1992).

While the two disorders are clinically distinct, both can be linked to disruptions in

gene expression from chromosome 15qll-ql3 (Knoll et al., 1989b). Most often PWS

and AS can be attributed to large deletions of 3-4 mb encompassing this locus (category

I). This region contains the paternally expressed genesM2KRN3, M4rGEL2, NDN,

SNURF/SNRPN, several sno RNAs, and the UBE3A antisense transcript. It also contains

the maternally expressed genes UBE3A and A TP10C. Prader-Willi Syndrome is due to a

paternal deletion of this region (Butler and Palmer, 1983) (Ledbetter et al., 1981) while

Angelman Syndrome is due to a maternal deletion (Magenis et al., 1987) (Knoll et al.,

1989a). These de-novo deletions are believed to be facilitated by homologous









recombination between large genomic duplications of the gene HERC2 that flank 15qll1-

ql3 (Amos-Landgraf et al., 1999).

A second category of mutation that creates a functional loss is uniparental disomy

(UPD) of chromosome 15 (category II). Paternal duplication results in AS (Malcolm et

al., 1991) (Knoll et al., 1991) while maternal duplication results in PWS (Nicholls et al.,

1989). UPD is thought to occur due to a non-disjunction event followed by reduction

within the zygote to remain viable. Since both chromosome fifteens are inherited from

one parent they both act in that manner and imprinted expression is lost from the other

parents' allele.

A third class of mutations are only found in AS. These patients inherit an intact

chromosome 15 from each parent but have mutations within the UBE3A gene (Kishino et

al., 1997) (Sutcliffe et al., 1997) or lesions of unknown origin (category IV and V). Thus,

functional loss of UBE3A appears to be sufficient to cause the main clinical features of

AS. Currently no PWS cases have been described which can be attributed to defects in a

single gene. Rather it is believed that Prader-Willi is a contiguous gene syndrome in

which multiple genes within 15qll-ql3 must be lost to cause PWS.

A fourth less frequent but important class of PWS and AS are imprinting center

mutations and microdeletions (category III) (Horsthemke et al., 1997). The imprinting

center (IC) is a cis acting regulatory region that controls the parental specific expression

of genes in the locus. In this category both a maternal and paternal chromosome 15 are

inherited, however loss of a functional IC results in both behaving maternally in the case

of PWS or paternally in the case of AS. Mapping of microdeletions in PWS and AS

patients has led to the identification of two shortest regions of overlap (SRO) indicating









the boundaries of the PWS-IC and AS-IC (Figure 1-2). The first, a 4.3kb region that is

located just 5' to and includes the first exon of SNRPN, is associated with Prader-Willi

syndrome (PWS-IC) (Buiting et al., 1995) (Saitoh et al., 1996). The second, an 880bp

region located approximately 35 kb upstream of SNRPN, is associated with Angleman

syndrome (AS-IC) (Buiting et al., 1995) (Buiting et al., 1999). Evidence suggests that

the PWS-IC is a positive element that drives expression of the paternal genes. Physical

or functional loss of this element results in loss of the expression from the paternal allele.

The AS-IC, on the other hand, seems to be a negative element silencing the activity of the

PWS-IC on the maternal chromosome. Loss of this element results in loss of expression

from the maternal allele (Figure 1-3).

The severity of phenotype within AS is dependent upon the molecular defect

inherited (Lossie et al., 2001) (Butler et al., 2004). Category I deletions produce the most

severe phenotypes, showing high incidence of early onset seizures, microcephaly, and

hypopigmentation while patients with UPD or IC mutations are much less likely to

present with these symptoms. Patients with class IV and V mutations fall somewhere in

between. The severity of category I is thought to be due to the physical loss of non-

imprinted genes that fall between the breakpoints of the deletion but are outside of the

imprinting locus.

Function of Imprinting

It is generally understood that in mammals the diploid state has the benefit of

conferring protection against deleterious recessive mutations. What then could be the

purpose of genomic imprinting, which in effect results in a haploid state at certain loci?

Numerous hypothesis have been put forth to try to explain the possible benefits of

imprinting. Four are discussed below.









The rheostat model was first suggested by the lab of Arthur Beaudet. It suggests

that silencing a single allele in an individual shelters it away from the effects of natural

selection (Beaudet and Jiang, 2002). This temporary and reversible state of haploidy

allows mutations that may be beneficial to the population to quickly accumulate while

those that are deleterious are hidden. Lethal mutations that are silenced may eventually

mutate again and become advantageous. Beaudet suggests that this model allows for

relatively rapid adaptations to environmental changes, however this is difficult to prove.

A second hypothesis, termed the ovarian time bomb hypothesis, suggests that

imprinting has evolved as a means of protection for the mother against ovarian

trophoblastic disease (Varmuza and Mann, 1994). Those genes that are silenced within

the maternal genome are suggested to be important in development of the trophectoderm.

Without expression of these genes ovarian teratomas that develop parthenogenetically

within an ovary lose their invasiveness and remain relatively benign. While this

hypothesis holds true for some "male" genes it does not explain the presence of

imprinting in all cases and does not explain the presence of silenced paternal alleles.

A third hypothesis is that imprinting of vital genes ensures sexual reproduction and

therefore genetic diversity through hybrid vigor (Driscoll, 1994). By requiring a genetic

contribution from two parents, this ensures protection from the risk of homozygosity for

deleterious recessive mutations. If parthenogenesis were possible in humans, the risk of

acquiring these recessive alleles would increase, decreasing the fitness of the species.

The hypothesis that has gained the most traction within the imprinting community

was put forth by David Haig in 1991. His Kinship theory suggests a genomic "tug-of-

war" between maternal and paternal interests (Moore and Haig, 1991). Here paternally










expressed genes favor embryonic and neonatal growth as well as other traits that

maximize survivability at the expense of the mother and future offspring. In order to

protect herself and future offspring, the maternal genome silences those genes within the

growing fetus that compromise the mothers reproductive fitness and the survival of future

pregnancies and expresses those that are growth inhibiting.

Of the suggested models, the kinship theory is supported best by observable

evidence. First, the only animals in which imprinting has been observed are placental

mammals. This is an important argument to the kinship theory in that female eutherians

are more resource indebted to their young than egg laying species. The second line of

evidence comes from analysis of those genes that are imprinted. Indeed many of them

are involved in fetal and prenatal growth and behavior.

None of the current explanations for imprinting can perfectly account for its

existence. Many imprinted genes may in fact be "hitch-hikers," genes that just happened

to fall within an imprinted loci's range of influence. Also not all genes may be imprinted

for the same reasons. It seems likely that a host of factors probably played a role in the

evolution of imprinting.

The PWS/AS Locus

The PWS/AS locus is highly conserved between man and mouse, both in gene

order and imprinted expression making the mouse an excellent system in which to study

the imprinting mechanism (Leff et al., 1992) (Lee et al., 2000) (MacDonald and Wevrick,

1997) (de los Santos et al., 2000) (Jong et al., 1999a) (Chamberlain and Brannan, 2001).

The locus consists of two clusters of genes on chromosome 15qll1-ql3, an upstream

cluster and a downstream cluster in relation to the imprinting center (IC). The syntenic

region in the mouse is located on chromosome 7C (Figure 1-4).









In humans and mice, the upstream cluster contains three paternally expressed

intronless genes, M4rGEL2 Magel2 (Boccaccio et al., 1999), and NDN/Ndn (McDonald

and Wevrick, 199 7), both Mage family genes, and M~KRN3 Mrnm3 (Jong et al., 1999a;

Jong et al., 1999b), a putative zinc finger protein. Both NDN/Ndn and M~AGEL2 Magel2

contain differentially methylated CpG islands within their 5' promoter regions as is

frequently seen in imprinted genes. Both are hypomethylated on the paternal allele and

hypermethylated on the maternal allele. M~KRN3 Mrnm3 also contains a 5' CpG island,

however it has yet to be shown to be differentially methylated. It is also associated with

an anti-sense transcript of unknown function. The mouse contains a fourth paternally

expressed intronless gene, Frat3, which is differentially methylated as well. Frat3 seems

to have been acquired by the locus due to a relatively recent L1 mediated

retrotransposition event (Chai et al., 2001). Upon insertion into the locus, Frat3 appears

to have become an innocent bystander having adopted the imprinted expression pattern of

its neighboring genes.

The downstream cluster consists of a single 460kb long transcript from which

multiple paternally expressed gene products are spliced (Le Meur et al., 2005). Most

proximal to the promoter is the bicistronic gene SNRPN Snrpn that encodes both SmN, a

component of the spliceosome (Shemer et al., 1997), and SNURE SnurJ; SNRPN

upstream reading frame, a protein of unknown function (Gray et al., 1999). Additionally,

this transcript encodes for several different snoRNAs; HBHI-436 M\~BH-436, HBHI

13 M~BHI-13 HBHI-43 7 HBHI-438A and B, and tandem repeat arrays of HBHI-52 M~BH--52

and HBHI-85 M~BH-85 (Runte et al., 2001). These snoRNAs are processed from introns

contained in the long transcript. Most distally encoded is anti-sense UBE3A which is









believed to be important for silencing the paternal copy of UBE3A (Runte et al., 2001).

Two maternally expressed genes are also contained within the downstream cluster,

UBE3A/Ube3a, and ATP10C/AtplOc. Ube3A is implicated as being the Angelman

syndrome gene as deletion of UBE3A is sufficient to cause the disorder.

Important to the region is the Imprinting Center (IC). The purpose of the imprinting

center is to regulate mono-allelic expression within the locus. ICs are often found in

imprinted loci however their mechanism of action is not always the same. As discussed

earlier, the IC is bipartite in nature consisting of a PWS-IC and an AS-IC. The PWS-IC

is located just 5' to and includes exon 1 of SNRPN/Snrpn. Paternal deletion of this

region in mice leads to the loss of expression of the paternal set of genes and concurrent

biallelic expression of UBE3A/Ube3A (Yang et al., 1998) (Chamberlain and Brannan,

2001). The AS-IC is approximately 35kb upstream of the PWS-IC in humans.

Microdeletions lead to biallelic expression of the paternal genes and loss of expression of

UBE3A/Ube3a. The AS-IC has yet to be located in the mouse and is one focus of this

work. Our lab has suggested a model (Brannan and Bartolomei, 1999) whereby the

PWS-IC is a positive force, driving expression of the upstream and downstream paternal

genes on the paternal allele by some unknown mechanism. The AS-IC on the other hand

is a negative force, silencing the PWS-IC on the maternal allele and therefore silencing

expression of the paternally expressed genes there (Figure 1-5). Several observations lead

to this model. First, paternal inheritance of an AS-IC deletion is benign suggesting its

function is only on the maternal allele. Likewise maternal inheritance of a PWS-IC

deletion is also benign suggesting that it functions only on the paternal allele. A third

observation is that PWS-IC deletions that include the AS-IC cause PWS whether










maternally or paternally inherited. This shows that the function of the AS-IC is

secondary to that of the PWS-IC. Indeed, if the function of the AS-IC is only to silence

the PWS-IC, loss of both would have no effect on the maternal allele since there is no

PWS-IC to drive expression of the set of paternal genes on both the maternal and paternal

alleles.




















Maintenance


Zygote ~

















Estab lish ment


Maintenance


B astocyst


Erasure


Primordial Germ Cells


Figure 1-1 The life cycle of an imprint. There are three phases in the life of an imprint.
First the imprint must be established in the germline. This imprinting mark
can be distinguished between the maternal and paternal alleles. Second the
mark must be maintained within the somatic tissues throughout the life of the
animal. Third the imprint must be erased during gametogenesis of a
developing embryo so that it can be reset according to the sex of the embryo.
























I I_ I _______


4.3 kb,


AS-IC SRO


PHYS-IC SRO


*-(35 kbi)


Figure 1-2. Mapping the Imprinting Center (IC). Mapping in humans using Shortest
Region of Overlap (SRO) shows that the IC is bi-partite in structure. The
PWS-IC is located within 4.3 kb interval that includes exon 1 of SNRPN. The
AS-IC is located 35kb upstream of SNRPN exon 1 within a .88 kb interval


0.88 kb
















Functional loss of paternal genle expression from chromosom e 1 5 q11-ql 3




..PWS



Normal Large UPD IC micro-deletion
Deletion 75% or mutation 3%
7 3%
Functional loss of m atemnal gene expression from chrom osom e 15 q11-13




AS



Normal Large PDSingle Gene IC micro-deletion
Deletion 2% Or or mutation 2%
73%P Unknown


Figure 1-3. Molecular classes of PWS and AS. There are four different molecular classes
that can result in PWS and AS. First, large deletions of 3-4 Mb result in loss
of the all genes within the locus. This deletion occurs on the paternal allele in
the case of PWS and the maternal allele in the case of AS. Second,
uniparental disomy results in two alleles that act in the same manner. Maternal
uniparental disomy results in PWS and Paternal uniparental disomy results in
AS. A third class, single gene mutation, only occurs in the case of AS. This
occurs when the gene UBE3A is mutated or deleted. The fact that this class
does not exist in the case of PWS suggests that it is a contiguous gene
disorder, requiring the loss of multiple genes to occur. A fourth class, IC
microdeletions, occurs when the imprinting control elements are missing or
mutated.
















































Figure 1-4. Gene organization of the PWS/AS locus. The PWS/AS locus, located on
human chromosome 15qll1-ql3, consists of several genes upstream of the IC
and a single long transcript originating from the SNRPN gene downstream of
the IC. The syntenic region in the mouse exists on mouse chromosome 7C
and is conserved both in gene order and imprinted expression pattern.


~4~~p~;~P~)~y


II .. 1 I I


___


I


1


I


Humann 15ql1-ql3


-$ ,G~1~
F~c' sY iP:~5~


4
~a9~


III


.)Ub-n ~~ iL


Mousea 7C


C


~


I-, ~

i i


C1~*


P



































AS-IC PW5-IC


16




















,-" ,,
c-"""-^; ~9~- i`"t
a"
i
~a


Figure 1-5. Paternal only model of imprinting regulation. The paternal only model of
regulation suggests that the PWS-IC is active only on the paternal allele
driving expression of the paternal genes. The AS-IC on the other hand is
active only on the maternal allele and acts to silence the PWS-IC on the
maternal there.















CHAPTER 2
MATERIALS AND METHODS

Lambda Red Recombineering

Transformation of Recombineering Strains

In order to create targeted deletions within the BAC of interest, it must first be

transformed into the appropriate recombineering E-coli strain. Three strains are available.

DY3 80 is a modified DH10B strain containing the defective lambda prophage required

for recombineering. EL3 50 is a modified DH10B strain containing the defective lambda

prophage and an arabinose inducible cre gene used for cre/lox recombination. EL250 is a

modified DH10B strain containing the defective lambda prophage and an arabinose

inducible flpe gene used for flpe/f h recombination.

Fresh BAC is prepared by alkaline lysate method. Briefly, single colonies of

DH10B containing the BAC of interest are picked and grown overnight in 3 mL Luria

Broth (LB) with 12.5 ug/mL chloramphenicol at 32oC, 175 rpm. The next day cultures

are pelleted in a microcentrifuge at 7000 rpm, resuspended by vortexing in 100 uL

solution I (50 mM dextrose, 10mM EDTA, 25mM Tris pH 8.0) and allowed to sit at

room temperature for 3-5 minutes. Next, 200 uL of solution II (.2M NaOH, 1% SDS) is

added, the samples are mixed by inversion and then placed on ice. After 5 minutes 150

uL of solution III (29.4g KOAc, 11.5 ml. conc. HOAc, H20 to 100ml.) is added, the

samples are shaken and then placed back on ice. 400 uL of prepared phenol chloroform is

added to the tubes and the samples are shaken to mix and then centrifuged at 13000 rpm

for 5 minutes. After centrifugation the aqueous layer is transferred to a fresh tube and 1









mL of 100% EtOH is added. The samples are then shaken and centrifuged again at 13000

rpm. The supernatant is poured off and Iml of 70% EtOH is added. The samples are

shaken and centrifuged again at 13000 rpm. The resulting DNA pellet is resuspended in

50ul of sterile H20.

In order to transform the BAC into the DY3 80 E. coli, a single colony of the

appropriate recombineering strain is grown up overnight in 3 mL LB at 32oC, 175 rpm.

The next day 700ul of culture is diluted in 70 mL of LB with 12.5ug/mL chloramphenicol

and incubated 3-5 hours at 320C, 175 rpm, until the culture density reaches an OD600 Of

.5-.6. The culture is next transferred in 10 mL aliquots to 15 mL falcon tubes and then

pelleted by cold (~5oC) centrifugation at 5000 rpm. Supernatant is poured off and the

pellet is resuspended in 10 mL ice cold 10% glycerol. The tubes are spun again at 5000

rpm, SoC, for 5 minutes and the supernatant poured off. The pellet is resuspended this

time in 1 mL 10%glycerol and transferred to a 1.5 mL eppendorf tube. The tubes are spun

at 13000 rpm in a tabletop microcentrifuge for 30 seconds and the supernatant poured off.

The pellets are again resuspended in 1 mL 10% glycerol and spun again at 13000 rpm.

This time all of the supernatant is carefully drawn off with a pipette and the pellet is

resuspended in 80ul of 10% glycerol. The tubes containing the now electrocompetent

cells are stored on ice until use. Fresh BAC DNA is mixed with the electrocompetent

cells to give a final volume of 50ul. Usually DNA volumes of 2, 5, and 10 uL mixed with

48, 45, and 40 uL competent cells will give enough of a range to produce good results.

The DNA, cell mixture is transferred to a pre-cooled .1 cm Bio-Rad electroporation

cuvette; catalog number 165-2089. Electroporation is performed using a Bio-Rad Gene

Pulser set at 1.8kV, 25uF, with the pulse controller set at 200 ohms. The electroporated









cells are then immediately diluted with 1 mL of LB and allowed to incubate at 32oC, 175

rpm, for 1 hour. After incubation, the cells are pelleted by centrifugation at 7000 rpm and

the supernatant drawn off with a pipette, leaving approximately 100ul total volume in the

tube. The remaining volume is plated to LB agar plates containing 12.5 ug/mL

chloramphenicol and incubated overnight at 32oC. Resulting colonies are screened by

digesting with numerous different restriction enzymes and comparing them to the original

BAC.

Transformation of Targeting Vector and Induction of Recombination

Recombination within the BAC is achieved by electroporating a targeting construct

into the BAC containing recombineering strain. The targeting construct should contain a

selectable marker such as neomycin resistance and should be linearized by restriction

digest or amplified by polymerase chain reaction (per) prior to transformation. The night

before transformation a single colony of the BAC containing strain is grown up overnight

in 3 mL LB at 32oC, 175 rpm. The next day 700ul of culture is diluted in 70 mL of LB

with 12.5ug/mL chloramphenicol and incubated 3-5 hours at 32oC, 175 rpm, until the

culture density reaches an OD600 Of .5-.6. At this point 10 mL of the culture should be

drawn off and reserved on ice to serve as an uninduced control. The remaining 60 mL is

induced to express the lambda RED genes by shaking in a 42oC water bath for 15

minutes. After induction the flask is transferred to an ice slurry bath and swirled for 10

minutes to turn off the defective lambda phage. The induced and uninduced control

cultures are then made to be electrocompetent in the same way in which the

recombineering strains were made to accept the BAC, by a series of cold 10% glycerol

washes. Targeting vector DNA is mixed with the electrocompetent cells to give a final









volume of 50ul. Usually DNA volumes of 2, 5, and 10 uL mixed with 48, 45, and 40 uL

competent cells will give enough of a range to produce good results. The DNA, cell

mixture is transferred to a pre-cooled .1 cm Bio-Rad electroporation cuvette; catalog

number 165-2089. Electroporation is performed using a Bio-Rad Gene Pulser set at

1.8kV, 25uF, with the pulse controller set at 200 ohms. The electroporated cells are then

immediately diluted with 1 mL of LB and allowed to incubate at 32oC, 175 rpm, for 1

hour. After incubation, the cells are pelleted by centrifugation at 7000 rpm and the

supernatant drawn off with a pipette, leaving approximately 100ul total volume in the

tube. The remaining volume is plated to LB agar plates containing the appropriate

antibiotic and incubated overnight at 32oC. The resulting colonies are screened by

Southern blot.

BAC preparation for injection

In order to inj ect BAC's as transgenes, the DNA must be prepared in a way that

removes all traces of impurities and leaves the BAC as intact supercoils if possible. The

DNA may be linearized by restriction digest, but this requires an additional purification

step. BAC preparation by cesium prep and linearization are described below.

Two nights before the DNA prep, a single bacterial colony containing the

appropriate BAC is picked and cultured overnight in 3 mL LB with 12.5 ug/mL

chloramphenicol at 32oC, 175 rpm. The next day, 1 mL of this starter culture is

inoculated into 3 1000ml flasks containing 500 mL LB with 12.5 ug/mL

chloramphenicol. These cultures are incubated overnight at 320C, 175 rpm. The next day,

the cultures are poured into 500ml bottles and spun in a centrifuge at 5000 rpm. The

resulting pellets are resuspended in 25 mL solution I each (50 mM dextrose, 10 mM










EDTA, 25mM Tris pH 8.0) and allowed to sit at room temperature for 3-5 minutes. 50

mL solution II (.2M NaOH, 1% SDS) is added to each bottle and then inverted to mix.

After 4 minutes 37.5 mL of solution III is added (29.4 g KOAc, 11.5ml glacial acetic

acid, H20 to 100 ml), the bottles are mixed by inversion and placed on ice for 10 minutes.

In order to remove as much cellular debris as possible the mixture is poured into clean

bottles through a layer of gauze and then spun in a centrifuge at 5000 rpm for 5 minutes.

The now cleared solution is additionally filtered through medium porosity filter paper

(Fisher catalog number 09-802-la). Each bottle now should contain approximately 112

mL of BAC DNA solution. In order to precipitate the DNA the solution is transferred to 3

250 mL bottles which are then filled to the neck with 100% EtOH. The EtOH/DNA

mixture is spun in a centrifuge at 10,000 rpm for 20 minutes. The supernatant is poured

off, the bottles filled with 70% EtOH and then spun again at 10,000 rpm for 10 minutes.

The ethanol is poured off and any residual ethanol is removed from the bottles using a

pipette. The pellets are then air dried for approximately 20 minutes and resuspended in a

total of 8 mL TE with RNase when combined.

To perform a gradient the density of the solution must be brought up to 1.55 g/mL

by adding the proper amount of cesium. 1.05 grams of cesium are added per each mL of

solution to achieve this density. Transfer the solution to a 15 mL conical tube and to this

tube add 8.4 g of cesium. Mix the solution by inversion until all the cesium dissolves.

The density should be checked by weighing 1 mL of solution on an analytical balance.

The weight should equal between 1.55-1.57 g. The cesium/DNA mixture is then divided

between 2 ultracentrifuge tubes. One hundred uL ethidium bromide is added and the

volume brought up to the top of the tube using pre-made 1.55 g/mL TE-RNase solution









stored in the freezer, making sure to keep the 2 tubes at an equal weight. Seal the tubes

spin them overnight at 55,000 rpm using an NVT-90 fixed angle rotor in an

ultracentrifuge. The next day, remove the tubes from the rotor and extract the DNA,

which should be present as a visible red band within the tube, using an 18 gauge needle

and syringe. Combine the DNA from the 2 tubes into one and bring up the volume using

the pre-made 1.55g/mL TE-RNase solution. Seal the tube and spin again for 4 hours at

78,000 rpm using the same rotor. Again, extract the DNA with an 18 gauge needle and

syringe. It should be present now as a thicker red band, and transfer to a 15 mL conical

tube. There should be a total volume of 1-1.5 ml.

Salt water saturated butanol is used to extract the ethidium from the DNA solution.

To make the butanol solution, in a 250 mL bottle add NaCl to 50 mL of sterile water until

the solution becomes completely saturated and the salt no longer dissolves. One hundred

and fifty mL butanol is added to the bottle and mixed by shaking. The butanol is allowed

to completely separated from the aqueous phase before using the solution. Two mL

butanol is added to the BAC DNA solution and mixed by inverting 10 times. Allow the

aqueous layer to resolve and then remove the upper butanol layer. Continue to repeat the

addition and removal of butanol until the lower DNA solution layer is completely clear

when looked at in front of a white background.

The ethidium free BAC DNA solution is next transferred into a 4 inch long piece of

Spectra/Por dialysis tubing (MWCO 12,000-14,000 daltons, cat.# 132676) for dialysis

against the DNA inj section buffer. The tightly sealed dialysis bag is allowed to float in a

bucket of 4 liters of inj section buffer (10mM tris pH7.5, .1mM EDTA, 100mM NaC1)

overnight at a temperature of 4oC with a slowly spinning stirbar. The next day the DNA









solution is removed from the dialysis tubing and the concentration is measured on a

spectrophotometer.

Generation of Transgenic Animals by Pronuclear Injection

Inj sections were expertly done by the Chris Futtner Inj section Core (CFIC) at The

University of Florida. Briefly DNA was microinj ected into the pronucleus of fertilized

mouse zygotes derived from superovulated, 5 week old FVB female mice at a

concentration of 2.5-3 ng/ul. Inj ected zygotes were allowed to mature overnight to the

two cell stage at which point they were transferred into the infundibula of pseudo

pregnant (B6D2)F 1 female mice.

Genotyping

Identification of Transgenic Founders

PCR genotyping was used to identify founder animals carrying the various

transgene constructs. At 2 weeks of age pups born to (B6D2)F 1 foster moms were ear

punched and tail clipped for screening. Genomic DNA was prepared from each tail piece

by digestion with 10ug/mL protienase K in tail lysis buffer (100mM Tris pH 8.5, 5mM

EDTA, .2% SDS, 200mM NaC1). PCR was performed to screen for the T7 BAC arm and

adjoining mouse sequence using the following primers: T7F 5'-GTA ATA CGA CTC

ACT ATA GGC-3' and 425T7R1 5'- CTC CAA TCA TGT TCA ACT GTC-3'.

Founders were further screened by Southern blot, probing for Snrpn.

Identification of Castaneous C7 Homozygotes

PCR genotyping followed by restriction endonuclease digestion was used to

identify transgenic animals that were also homozygous for Castaneous C7 at the PWS

locus. This was done by screening for a polymorphic Avall site between M~us musculus

ca~staneous and M~us musculus domesticus that results from a cytosine nucleotide rather









than a thymine nucleotide at position 1 17 of the Ndn gene. Avall cuts the domesticus but

not the castaneous allele due to the presence of the cytosine. PCR primers flanking the

polymorphic site with the following sequence were used: NdnpolyF 5'-ACA AAG TAA

GGA CCT GAG CGA CC-3' and NdnpolyR 5'-CAA CAT CTT CTA TCC GTT CTT

CG-3'. The amplified PCR product was gel purified on a 2% agarose gel, extracted from

the gel using the Promega Wizard Prep kit, and digested with Avall. The digested

products were then electrophoresed on a 4.8% agarose gel (2:1 low-melt agarose:

agarose).

Southern Blot

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.









RNA Isolation

RT-PCR was performed to detect the expression of the transgenic Snrpn minigene.

Total RNA was isolated from complete brains obtained from neonatal mice. RNA was

extracted using RNA-zol (Tel-Test, Inc). Briefly, tissue samples were homogenized with

4 ml. of RNA-zol using a Polytron homogenizer. To this homogenate .4ml of chloroform

is added, the samples are shaken vigorously for 15 seconds and then put on ice for 5

minutes. After 5 minutes the tubes are then centrifuged for 15 minutes at 12,000g after

which the homogenate forms two distinct layers: a lower blue phenol chloroform phase

and an upper colorless aqueous phase containing the RNA. The aqueous phase is added

to a new tube and to it an equal volume of isopropanol is added and the samples vortexed

then place on ice for 15 minutes. After 15 minutes the samples are centrifuged for 15

minutes at 12,000g. The resulting white RNA pellet is then washed with 75%ethanol and

centrifuged again at 12,000 g for 10 minutes. After removal of the ethanol the pellet is

allowed to air dry and then resuspended in .2 mL of DEPC H20.

RT-PCR

RNA isolated from neonatal mouse brains was analyzed for transgene expression

by RT-PCR. Five ug RNA is mixed with lul of 500 ug/mL random primers and dH20 to

reach a final volume of 25.8ul. This mixture is then heated to 68oC for 3 minutes and

then placed on ice. To each tube is added 3.2ul dNTP mixture, 8ul of reverse

transcriptase buffer, 2ul 100mM DTT, lul RNase inhibitor and lul SuperscriptlI and then

incubated at 37oC for 60 minutes. Two uL of this reaction is then used to seed a PCR

reaction using the following primers: N2. 1F 5'-CCC CGA GTA TTA AGG ATC TTG-3'

and N6.2R 5'-GCA ACA GTG CCT CTT CCC TG-3'. PCR amplification condition









were 95oC for 5 min., followed by 30 cycles of 95oC for 30 sec., 57oC for 30 sec., 72oC

for 1 min. The cycle was followed by an extension step of 720C for 10 min.

Preparation of High Molecular Weight Genomic DNA for Bisulfite Conversion

High molecular weight brain DNA was isolated from neonatal mouse brain to be

used for bisulfite PCR. Two and one half mL of homogenizing solution (lX SSC, 1%

SDS, .25mg/mL Pronase E) was added to a dounce homogenizer along with a single

frozen neonatal mouse brain. Tissue was homogenized by gently moving the pestle up

and down within the dounce. The homogenate was collected in a 15 mL polyproplene

tube and the homogenizer washed with an additional 2.5 mL of solution that was then

combined with the homogenate. The homogenate was vortexed and then incubated for 1

hour in a 37oC water bath. After 1 hour, 5 mL of phenol chloroform was added and the

mixture was vortexed and then spun in a centrifuge at 2500 rpm for 5 min to extract the

DNA. The top aqueous layer was collected to a new 15 mL tube and the extraction was

repeated twice more as above, first with phenol chloroform and then with chloroform

alone. After the aqueous layer was removed from the final chloroform extraction, 0.5 mL

RNase A was added and then incubated for 30 min. at 37oC. One half mL of pronase E

was then added and the mixture was incubated an addition 30 min at 37oC. The DNA

was extracted again as above, first with phenol chloroform and then with chloroform

only. After extraction the DNA was precipitated by the addition of 12 mL 95% ethanol

followed by vortexing. The now visible DNA strands were spooled onto a glass rod and

placed into a 1.5 mL tube to dry. Once the DNA appeared to be mostly dry, 200 mL of

water was added to resuspend the DNA which was then stored at -20oC.









Bisulphite Sequencing of Genomic DNA

Bisulphite Conversion

Bisulphite conversion was carried out as described (Clark et al., 1994). DNA was

denatured with 4.2 Cll of 3 M NaOH (final concentration is 0.3 M in 42 CIl) 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 CIl. 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 Cll 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 Cll 7.5M Ammonium Acetate and 200 Cll 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 Cll 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)









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.

PCR was performed on bisulphite treated brain DNA preparations with HotStar

Taq (Qiagen) using the following PCR conditions (final): lX PCR buffer (with 15 mM

MgCl2), 200 CIM each dNTP, 1 CIM each primer, and 1.5 U HotStar Taq DNA

polymerase. Bisulphite converted brain DNA preparations were not quantitated for DNA

concentration and 2 Cll was used per 25 Cl 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.

Purification of PCR Products

PCR products were purified using the Quiagen Quiaquik PCR cleanup kit as per kit

directions. Once the PCR product was purified and resuspended in 45 ul, 10 uL was taken

and run out on a 2% agarose gel to check for quality before cloning.

Cloning and Sequencing of PCR Products

PCR products were cloned using TA cloning with pGEM-T Easy Vector Systems

(Promega). Ligations were performed as suggested by the kit protocol. Ligations were

transformed into chemically competent, subcloning efficiency XL-1 Blue (Stratagene)









E.coli cells. Transformations were plated onto Luria-Bertani (LB) plates (1% tryptone,

0.5% yeast extract, 1% sodium chloride and 1.5% agar) containing 50 Clg/mL ampicillin

and supplemented with 100 Cll of 100 mM IPTG and 40 Cll 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 using the standard plasmid sequencing methods (described below).

Sequences were analyzed using Sequencher 4.2 (Gene Codes Corporation).

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).

Sequencing samples were prepared as suggested by the manufacturer

(PerkinElmer): 2 Cll BigDye Terminator, 2 Cll 5X sequencing buffer, 1 Cll 3.2 pmole/CIl

primer, 2 Cll water and 3 Cll plasmid DNA (approximately 3 Gig). 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 1/2






30


volumes of 7.5 mM ammonium acetate were added to the sequencing reactions. Samples

were washed with 70% ethanol and dried by vacuum centrifugation.















CHAPTER 3
IMPRINTED TRANSGENE

Introduction

Imprinted loci are often organized as clusters of genes controlled by cis-acting

regulatory elements called imprinting centers (IC) (Brannan and Bartolomei, 1999).

These IC's can often be found large distances away from the genes they control making it

difficult to localize and understand their mechanisms of action. Transgenes are excellent

tools for the study of imprinted loci in that they serve to isolate genes and their regulatory

elements and relocate them to ectopic locations in the genome allowing for the study of

those sequences without the influence of other adj acent elements. BAC transgenes have

an additional advantage over conventional plasmid based transgenes. Smaller plasmid

transgenes are often influenced by the chromatin into which they randomly insert. These

insertional effects often result in inconsistencies in expression between individual

transgenic mouse lines. The inherent greater size of BAC transgenes make them much

less susceptible to position effects most likely due to the fact that they contain essential

regulatory elements necessary for maintaining an open chromatin structure. For the study

of imprinted loci, this greater size also increases the chance of inclusion of IC's and other

necessary sequences required for proper expression patterns.

BAC transgenes have been successfully used in the past for the study of imprinting.

In one example Yevtodiyenko showed that a 178 kb BAC transgene expressed the Gene-

trap locus 2 (Gtl2) gene only upon maternal inheritance within the mouse embryo and

placenta (Schmidt et al., 2000; Yevtodiyenko et al., 2004). This is similar to the









endogenous locus on distal mouse chromosome 12 and human 14q32 suggesting that the

178 kb interval contains all the elements necessary to imprint Gtl2. They were also able

to show that levels of expression were similar to endogenous levels only in the brain but

not other regions, localizing the enhancer elements necessary for brain expression but not

other tissues to within the BAC. Lastly they showed that a second gene present within

the BAC, Delta-likeltt~~~tt~~~ttt~~ (Dlkl), was not expressed in an imprinted fashion, as is the

endogenous locus, indicating that additional elements are required for imprinting Dlkl.

The Surani lab reports using a 120 kb BAC transgene to localize imprinting elements

required for paternal expression of the Peg3 gene which encodes a C2H2 type zinc-finger

protein involved in maternal behavior (Li et al., 2000; Szeto et al., 2004). In this study

they showed that imprinting is linked to differential methylation of a region upstream of

the gene. A second maternally expressed gene Ziml, located within the same cluster and

present on the BAC was not imprinted correctly indicating that other cis acting elements

are required for imprint regulation of the locus. The Surani lab also used BAC transgenes

to locate dispersed cis-regulatory elements necessary for paternal expression of

Nueronatin (Nnat) (John et al., 2001). By modifying the BAC they showed that these

elements were located within an 80kb interval mostly upstream to Nnat.

As other labs have shown, BAC transgenes can be a powerful tool for localization

of IC elements. In the case of the PWS locus, current understanding of IC function in the

mouse is limited to a 35 kb region in which the PWS-IC is located. The AS-IC in the

mouse has yet to be identified. Our lab has previously produced several transgenic lines

aimed at further elucidating the imprinting machinery of the PWS locus (Figure 1). The

first of these transgenic lines was developed from mouse and human phage (P l) clones










carrying 85 and 76 kb of genomic sequence respectively (Blaydes et al., 1999). The

mouse transgene contained 33 kb of 5' and 30 kb of 3' flanking DNA while the human

clone carried 45kb of 5' and 7 kb of 3' flanking sequence. Five lines were produced from

the human Pl clone, four of which were single copy and the fifth present as Hyve copies.

This clone was chosen due to the fact that it contained the complete interval in which

both the PWS-IC and the AS-IC were located. All lines were expressed biallelically

suggesting that the human PWS-IC, or at least the Snrpn promoter, is functional in the

mouse, while the AS-IC is not. Two lines were derived from the mouse phage clone, a

line containing a single copy of the transgene and a line containing two copies. The

single copy line was biallelically expressed while the two-copy line was correctly

imprinted as it was expressed only upon paternal inheritance. The most likely

explanation for this two-copy effect is that the mouse clone was large enough to contain

the entire PWS-IC since it was capable of driving expression of Snrpn. However the AS-

IC probably lies outside of this interval since the single copy line was unable to silence

Snrpn expression upon maternal inheritance. The presence of a second copy somehow

overcomes this deficiency, either by acting as a buffer to the insertion site effects of

integration or it adds back some needed sequences lost to the single copy line. This is

similar to the effect seen in BAC transgenic studies of the maternally expressed H19 gene

where a 14 kb transgene requires multiple copies to imprint, however a 130kb BAC

transgene imprints in single copy (Bartolomei et al., 1993; Pfeifer et al., 1996).

More recently the lab has created multiple transgenic lines using BAC transgenes

which are much greater in size than the Pl clones. The Roswell Park RPCI-23 murine

BAC library was screened for clones containing Snrpn. Three were identified that carried










sufficient sequence upstream of Snrpn; 215A9 which extends 120 kb 5' and 20 kb 3',

380J10 which extends 8 kb 5' and 140 kb 3', and 425D18 which extends approximately

100 kb 5' and 65kb 3'. These three BACS were inj ected as transgenes and were then

tested for imprinted expression patterns (Figure3-1). The 380J10 transgene was

expressed upon both maternal and paternal inheritance suggesting that the complete

PWS-IC was present but the AS-IC was lacking. The 425 and 215 transgenes on the

other hand showed correct imprinted expression. They were paternally expressed and

maternally silenced. Interestingly the 215B line suffered from a truncation that leaves

only approximately 40 kb of sequence 5' to Snrpn. This establishes the minimal genomic

sequence necessary to establish imprinted expression in the locus as approximately 40 kb

upstream and 20 kb downstream of Snrpn.

This proj ect attempts to further delineate the sequences necessary to confer an

imprinted expression pattern within the PWS locus. To do this a modified BAC

transgene was developed in order to simplify expression analysis. Later chapters will

discuss additional modifications that were used to minimize the PWS-IC interval and

locate the AS-IC.

Results

BAC Modification

Previously, analysis of transgene expression required a complex breeding scheme

in order to distinguish the transgenic allele from the endogenous alleles. Transgenic

founders were mated to mice carrying a targeted deletion of exons 5-7 of Snrpn (ASmN)

that results in a larger fusion transcript that include exons 1-4 and the neomycin

resistance gene (Yang et al., 1998). In this way the transgenic allele could be









distinguished as a smaller transcript than the endogenous one. In order to simplify

expression analysis, we decided to mark the transgene in a way that it could be

distinguished immediately. This was done using a technique called lambda red mediated

homologous recombination developed by Daiguan Yu and Don Court at the National

Cancer Institute (Lee et al., 2001). This method takes advantage of a defective lambda

prophage from which the genes allowing entry into a lytic life cycle have been removed.

What is left are only the minimal elements required to express the lambda Red genes exo,

beta and gam. Exo and bet are required for recombining a double stranded (ds) linear

DNA targeting construct into BAC DNA via homologous recombination. Exo encodes a

5'-3' exonuclease that acts upon the 5' ends of linear dsDNA to create 3' single stranded

(ss) ends. Bet encodes a pairing protein that binds the 3'ssDNA ends and promotes

annealing to the complementary strand on the BAC. The third gene, gamn, inhibits E. coli

RecBCD exonuclease activity that destabilizes linear dsDNA. The defective prophage is

integrated into the E. Coli DH10B (DY3 80) host bacterial chromosome and is under the

control of the strong PL promoter. The promoter in turn is under the control of the

temperature sensitive lambda cl857 repressor. Transiently shifting the culture

temperature from 32oC to 42oC results in expression of these three genes and subsequent

homologous recombination between the linear targeting construct and BAC DNA.

The 425D18 BAC was chosen for modification since a good amount of both 5' and

3' sequence was present. Since the previous SmN deletion had shown no effects on

expression it was decided to create the same deletion on the BAC (Figure 3-2). Exons

5-7 of Snrpn were replaced with a flexed neomycin resistance cassette using a targeting

construct (170 flexed neo) containing a 2 kb 5' arm of homology and a 6.2 kb 3' arm of









homology. The targeting construct was linearized with Not I and then electroporated into

induced DY380 recombination competent cells. Integration of the targeting construct

was selected for by plating the transformed cells onto LB agar plates supplemented with

neomycin and incubation overnight at 32oC. The next day 12 neomycin resistant colonies

were picked and grown overnight in 3 mL of LB broth supplemented with neomycin at

32oC. After sufficient growth the cultures were subj ected to alkaline lysis to extract the

BAC DNA. BAC DNA was then cut with SacI restriction endonuclease and Southern

blotted. The blot was probed with a 2.2kb EcoRI/EcoRV genomic DNA fragment

obtained from a restriction digest of lab stock genomic DNA Snrpn clones. Owing to the

extreme efficiency of this recombineering method, all of the clones picked were positive

for the exon 5-7 deletion (Figure 3-2). Following successful targeting of Snrpn, the

flexed neomycin cassette was to be removed by expression of Cre recombinase. To do

this, the targeted 425 BAC (425A5-7neo) was transferred to an E. Coli host strain

carrying the Cre recombinase gene under the control of an arabinose inducible promoter

(EL350). However, after Cre had been induced, a severe undesired truncation event

occurred, making the BAC unusable (data not shown). Investigation into the composition

of the BAC vector (pBACe3.6) into which the genomic DNA is cloned revealed a

previously unknown third lox-p site. Induction of Cre resulted in recombination between

one or both of the neomycin cassette flanking lox-ps with the third lox-p resulting in

massive rearrangements. To deal with this, a second targeting construct, designed to

replace the third lox-p site with a blasticydin resistance cassette, was built.

Recombination with this construct was induced as before and efficient removal of lox-p

number three was achieved (data not shown). The doubly modified 425A5-7 BAC was









again subj ected to Cre mediated recombination. Successful deletion of the flexed

neomycin cassette was determined by restriction digest with EcoRI and SacI followed by

Southern blot. The Southern blot was then probed with the same 2.2kb EcoRI/EcoRV

genomic DNA fragment as before (Figure).

Production of Transgenic Mice

Transgenic mice were produced to test of the suitability of the 425A5-7 BAC as a

model system for the study of imprinting in the PWS locus. If the modified BAC was to

be a reliable system, the truncated Snrpn transcript it contains must be expressed only

upon paternal inheritance. When inherited from the mother it should be silenced. BAC

DNA was prepared for inj section by large-scale alkaline lysis followed by cesium chloride

banding. Both nicked and supercoiled DNA were collected from the cesium prep and

dialyzed against 4 liters of inj section buffer (1.0 mM Tris pH7.5, .1 mM EDTA, 100 mM

NaC1) for approximately 16 hours. After dialysis, the BAC DNA was diluted to 2.5 ng/ul

and inj ected into the male pronucleus of fertilized oocytes derived from superovulated

FVB/N female mice. The injected oocytes were then implanted into pseudopregnant

B6D2 Fl female mice. The resulting pups were allowed to mature to 3 weeks of age at

which time they were ear punched for identification and tail clipped to obtain DNA for

genotyping. Genomic tail DNA was then digested with Avall and Southern blotted. The

blot was then probed with a 1.5 kb genomic DNA HindlI restriction digest fragment to

both screen for the presence of the transgene and estimate copy number. Ten founder

animals were obtained. Line A and D appeared to be single copy when compared to the

endogenous band. Lines B, E, F, Gl, H, I, and J appeared to be present in multiple

copies. Line J was discarded, as it appeared to contain an immediately observable









rearrangement. Lines A, B, D, E, H, and I were retained and mated as they seemed to

contain the lowest numbers of copies. Lines D and E failed to mate so they were

discarded as well. The remaining four lines were analyzed for expression and

methylation status.

Analysis of 425A5-7 Transgenic Lines

Since the modified BAC transgene could be easily distinguished from the

endogenous locus, expression was easily testable after a single generation. Founder

animals were initially mated to wild type FVB/N males and females to establish the lines

and test for animals that may be chimeric. Chimerism results from late integration of the

transgene into the recipient zygotes genome after the first cleavage has occurred. If the

cell that receives the late integration does not contribute to the founder animal's germ

line, the transgene will not be transmitted to future generations and that line is worthless.

All four lines transmitted the transgene (data not shown). Transgenic males and females

from each line were next set up in matings with FVB/N males and females to test for

expression of the transgene upon both maternal and paternal inheritance. Brains were

collected from the resulting F2 day old neonatal pups. RNA was extracted from the

brains using RNAzol. The RNA was used as a template to produce cDNA that was then

used to program per reactions that spanned exons 4-8 of Snrpn. Since the modification

to the 425A5-7 BAC removes exons 5-7 the resulting transgenic transcript appears as a

band of approximately 350 base pairs (bp). This includes 242 bp of exonic sequence plus

the additional lox-p left over from targeting. The endogenous band spans 513 bp. In all

four lines transgenic Snrpn was expressed upon paternal inheritance and silenced upon

maternal inheritance (Figure 3-3). This is in accord with expression of the endogenous









alleles in which only paternal Snrpn is expressed. Interestingly, a faint signal was

amplified from the transgene upon maternal inheritance. This will be discussed later on

in this chapter.

As a second readout of the imprinted status of the 425A5-7 transgene, the

methylation status of Snrpn was analyzed using sodium bisulfite genomic sequencing,

hereafter referred to as the bisulfite sequencing. This method uses sodium bisulfite to

convert cytosines to uracils within genomic DNA. Methylated cytosines are protected

from this conversion and remain as cytosines. Converted DNA is PCR amplified across

the region of interest and the resulting products cloned and sequenced. When the

sequence is read, methylated cytosines remain as cytosines, however unprotected

cytosines appear as thymines. In this way the methylation status across an entire CpG

island can be determined.

Bisulfite sequencing was used to examine a CpG island present within the promoter

region of Snrpn. This region has been consistently shown to be hypomethylated on the

paternal allele and hypermethylated on the maternal allele (Brannan and Bartolomei,

1999; Hershko et al., 2001). In order to distinguish between the transgenic and

endogenous alleles, the 425A5-7 transgene was crossed onto a B6.Cast.c7 (Cast c7)

background in order to take advantage of a single base pair polymorphism within Snrpn

that can be differentiated during sequence analysis (Figure3-4). The Cast.c7 strain is a

congenic line that contains a region of2~us musculus ca~staneous chromosome 7 on a M~us

musculus domesticus C57BL/6J background. 425A5-7 transgenic males from the A and

H lines were mated to Cast c7 females and their offspring screened for transmission of

the transgene. Male and female Nl, transgene positive animals were then subj ected to a









second round of matings to the Cast c7 line. Brains from Pl neonatal offspring were

collected from these matings and tail tips from each were then screened for both the

transgene and Cast c7 homozygosity. Cast c7 homozygosity was determined by PCR

amplification of the Necdin (Ndn) gene using the following primers: NdnpolyF 5'-ACA

AAG TAA GGA CCT GAG CGA CC-3' A and NdnpolyR 5'-CAA CAT CTT CTA

TCC GTT CTT CG-3'. The product was subsequently digested with the Avall restriction

endonuclease. Ndn contains a polymorphic cytosine at residue 117 of the domesticus

allele that results in recognition and subsequent cutting by Avall. The ca~staneous allele

has a thymine residue at position 117 and does not cut. Cast c7 homozygosity is

therefore recognizable by a single ca~staneous band when electrophoresed on a 4.8%

agarose gel. Four animals from the H line were identified that contained the 425A5-7

transgene and were homozygous for the Cast c7 allele. Brain DNA from two of these

animals, one with a paternally transmitted transgene and the other with a maternally

transmitted transgene, was subjected to bisulfite conversion. Primers flanking the Cast c7

polymorphism were used to PCR amplify a 364 bp amplicon spanning the Snrpn

promoter CpG island. Primer sequences for this region were as follows: W18 5'-GTA

GTA GGA ATG TTT AAG TAT TTT TTT TGG-3' and W19 5'-CCA ATT CTC AAA

AAT AAA AAT ATC TAA ATT-3'. The products of this reaction were cloned and

sequenced producing data for 15 clones representing a maternally transmitted transgene

(Figure 3-5) and 11 clones representing a paternally transmitted transgene (Figure3-6).

Transgenic domesticus alleles were identified by a "G" at position 69 versus a "T"

present in the endogenous castaneous alleles. In both cases 14 endogenous allele clones

were sequenced as well. In summary, 11 of fifteen maternal allele clones showed greater









than 93% methylation while three were completely unmethylated. In contrast nine of

eleven paternal allele clones were less than 7% methylated while two clones were 35%

and 93% methylated. The endogenous clones contained a mixture of both methylated

and unmethylated clones as both maternal and paternal alleles were represented.

Discussion

As will be discussed in later chapters, much work has gone into trying to

understand the mechanism by which the PWS/AS locus is imprinted. Much of the data

that has come out of this work is confusing and contradictory. As a means to simplify the

system, we have created a BAC transgenic mouse model that contains 90 kb of upstream

sequence and 65 kb of downstream sequence in relation to the Snrpn gene. In this way

we can now study imprinting of Snrpn and its long transcript in isolation from the

upstream gene cluster which, while it probably shares some regulatory elements, most

likely has a different imprint mechanism. This BAC transgene model can now be

modified by recombineering to determine where specific IC elements may lay.

These initial proof of principal experiments show that this model does indeed

behave as the endogenous locus. Expression analysis shows that the 425A5-7 transgene

is expressed when paternally inherited and silenced when maternally inherited. This data

shows that the interval contained within the BAC is enough to confer correct imprinted

expression upon Snrpn, and it can be therefore be inferred that all the cis-regulatory

sequences, the PWS-IC and AS-IC, are present within. Additionally, bisulfite sequence

analysis of the CpG island within the Snrpn promoter region shows a hypermethylated

maternal allele and a hypomethylated paternal allele. This is also in agreement with what

is currently known about the locus.









Interestingly, analysis of expression from the transgene reveals a very low level of

expression when maternally inherited. Bisulfite analysis backs this up as 3 of the 15

clones sequenced were completely unmethylated and one can presume, expressed. This

at first was concerning since it suggested that the BAC did not contain enough sequence

to silence the maternally inherited transgene at all times. However, previous data from

our lab suggests that this may be the case for the endogenous locus as well. Analysis of

endogenous expression of Snrpn on a PWS-IC deletion background shows leaky

expression from the maternal allele that can be detected when the PWS-IC deletion is

inherited from the father (Chamberlain et al., 2004). It is important to note that this

deletion also removed the Snrpn promoter, therefore expression could only be coming

from the maternal allele. The bisulfite data is also important in this observation. The

maj ority of maternally inherited clones were completely methylated and four others were

at least 85%. The three unmethylated clones were unlikely to be due to experimental

artifact, the result of incomplete conversion, since all other cytosines within the sequence

were converted to thymines. This suggests that low level leaky expression may be

coming from cells that do not maintain the original imprint. Whether this is of biological

significance or simply the result of sloppy maintenance within isolated cells is unknown.

It is also possible that an as yet to be discovered small population of cells within the brain

may express Snrpn biallelically. However that is beyond the scope of this work.

In conclusion, we have created a BAC transgenic model that recapitulates the

endogenous imprinted expression pattern of Snrpn. This BAC contains all the regulatory

elements required to establish and maintain the imprint within its interval. The next two






43


chapters will describe modifications to the BAC that were used to further refine the

location of the PWS-IC and locate the AS-IC.




















'~LC
H~L


_ _


rh Ll~b
~d~ ~C
~~ c


~~L~kC


r~ gs~


4~~P, 4"-~8 2i~h 11 58


Irbcll~nrrcs


~~rse ~YX


Figure 3-1. Schematic and expression of BAC transgenics. (A) Schematic representation
of various BAC transgenic lines and their location in relationship to Snrpn..
(B) Northern blot analysis of BAC transgene expression. The 425A and 215B
BAC transgenic lines express Snrpn when inherited from the father (C) but
not when inherited from the mother. The 380 BAC transgene is expressed
when maternally inherited and therefore not imprinted. These lines are all
derived from unmodified BACs.


;1'T ~5E 21~A
j~i ~iiil~Wli

9

II1IIBI~~1We ~LI












































Fl
g
B


12~


I


erl er~ er3 ~4 ~5~6 ~7er8 er~L(I


i r "Ihcn~CT -Tr


85~


I


erF er~ er3 ~4 ~5 erBFa
u u u I u u.lken~n, uu


1[351~


I I
I


~L CI C3 ~4 ~57~8


Figure 3-2. 425A5-7 targeting. (A) The relationship of the 425A5-7 targeting construct
to Snrpn. A 2.2kb probe recognizes the 3'end of a SacI restriction fragment.
The unmodified BAC appears as a 16.8kb band. Replacement of exons with a
neomycin resistance cassette results in an additional SacI site and a 15.2kb
band. Cre-lox removal of the neo cassette results in a 10.5kb band. (B)
Southern blot shows efficient targeting of 12 clones with the A5-7 construct.
















T g(+) T g(-)

Tmns~Imissn Tmnsmissb~~n Tensrmissn Tmn~Ir~anus














I--


Figure 3-3. Rt-PCR analysis of 425A5-7mouse lines. rt-PCR analysis of four transgenic
mouse lines derived with the 425A5-7 BAC reveals proper imprinted
expression. The Snrpn minigene contained on the transgene is expressed
when paternally inherited and silenced when maternally inherited and is seen
as a smaller band when compared to the endogenous band found in all lanes..











































TG/-











Figure 3-4. Cast c7 breeding scheme. In order to distinguish the transgene from
endogenous alleles in bisulfite analysis, the 425A5-7 transgene was crossed
onto a Cast c7 background. This allows the two alleles to be distinguished by
a single base pair substitution within the Snrpn promoter region CpG island.



















11 L1
C" ~"
rr rr
rr L~


48







425A5-7 Maternally Transmitted Transgene


Transgenic Alleles
(8) 1 s ,
(1) 7
(1) ~ CI-I


Endogenous Alleles













Figure 3-5. Bisulfite sequence analysis of maternally transmitted transgene. Sequence
analysis of genomic DNA converted by sodium bisulfite reveals that the
maj ority of transgenic alleles are methylated at the Snrpn promoter CpG
island when inherited from the mother. A minority of alleles escape
methylation, perhaps accounting for the leaky expression seen from both the
transgene and endogenous alleles.
















425A5-7 Paternally Transmitted Transgene

Transgenic Alleles







Endogenous Alleles




12 vI V VV~c~v vvT



Figure 3-6. Biuft seuec anlyi ofaeral trnmte trngn.Squnc







Figue 3majorulity o transenic alleless patre l unmethylated a theSrpgne prmoereCpG

island when inherited from the father.















CHAPTER 4
REFINEMENT OF THE PWS-IC

Introduction

Considerable effort has been made to understand the cis elements that make up the

PWS-IC. Much of this work is in the context of targeted deletions of the region

surrounding exon 1 of Snrpn. Initially the location of the PWS-IC was mapped to a

region in humans that includes both the promoter and first exon ofSNRPN. This was

accomplished by fine mapping of paternal microdeletions that resulted in PWS. The

smallest region of overlap (SRO) between these deletions defined the location of the

PWS-IC within humans (Saitoh et al., 1996). Deletion of this region is associated with

absent expression of all the paternally expressed genes within the locus.

In 1998, our lab published a murine PWS-IC deletion model. This model contained

a 3 5kb deletion of the syntenic region within the mouse on chromosome 7 centered

around exon 1 of Snrpn (Yang et al., 1998) encompassing 16 kb of sequence 5' and 19 kb

3' to exon 1. Mice which inherit this deletion from the father lack expression of the

paternal genes Mkrn3, Ndn, and Ipw, as well as the Snrpn long transcript, and exhibit

several phenotypes common to PWS infants, including poor suckle, failure to thrive, and

lethality after approximately seven days. In contrast, mice inheriting a smaller intergenic

Snrpn deletion that removed the 3' part of exon 5, exon 6 and the 5' part of exon 7 had no

obvious phenotypic abnormalities when inherited from either the father or the mother.

These data suggested that both the location and function of the PWS-IC were conserved









between mouse and man and that the mouse could serve as a model with which to study

the mechanism of imprinting within the PWS locus.

With the goal of further refining the location of the PWS-IC the Beaudet lab

created two smaller deletions also centered on the first exon of Snrpn (Bressler et al.,

2001). The first, encompassing 0.9 kb, removed both the promoter and a portion of the

Snrpn differentially methylated region 1 (DMR1). These mice were phenotypically

normal upon either maternal or paternal inheritance. Analysis of DNA methylation at

both Snrpn and Ndn revealed no disturbance of the normal pattern as measured by

methylation sensitive restriction digests. This suggests that the sequence elements

necessary to confer imprinted expression upon the locus do not reside within this interval.

The second deletion removed a larger 4.8 kb segment also centered on exon 1. These

mice demonstrated a partial or mosaic phenotype with substantial postnatal lethality

(~40-50%) and growth deficiency when the deletion was inherited from the father. In

contrast, maternal inheritance resulted in no phenotype as would be expected in a PWS

IC deletion. Expression analysis was limited to the upstream genes as the entire Snrpn

promoter was removed in this deletion. Ndn and ~krn3 showed reduced levels of

expression and increased DNA methylation when the deletion was paternally inherited.

Although the partial imprinting defect makes these results difficult to interpret, one can

make the assumption that some elements of the PWS-IC do lie within the deleted

interval .

The Yang lab recently documented several transcription binding sites within the

Snrpn promoter region (Rodriguez-Jato et al., 2005). Using in-vivo footprinting and

chromatin immunoprecipitation (ChIP) analysis they were able to show that several sites










within this region are occupied. Three footprints were shown to be active on the paternal

allele. Footprint P2 appears to be a potential E2F binding site. This family of

transcription factors participates in transcriptional activation and suppression. Footprint

P5 was shown to associate with NRF-1. This site is conserved between mouse and

human and also overlaps with a previously reported seven base pair element (SBE), a cis

acting element shown to be required for promoter activity (Hershko et al., 2001).

Footprint P6/M6 was shown to be active on both the paternal and maternal alleles,

although the patterns were distinctly different. P6/M6 correspond to overlapping NRF-1

and CTCF binding sites however only the NRF-1 site on the paternal allele was

confirmed to be occupied by ChlP. Three other footprints were analyzed within the

promoter but do not appear to be occupied or correspond to known transcription binding

sites. The Yang lab also showed that elements with in the first intron of Snrpn were

preferentially bound only on the paternal allele. These three binding sites for the

transcription factors NRF-1, YY1, and SP1 are phylogenetically conserved between

mouse and man. NRF-1 and YY-1 have been shown to interact with chromatin

modification enzymes suggestive of their possible role in maintaining or establishing the

imprinted expression pattern in the locus. This agrees with the observation that this

conserved intronic region has been shown to be associated with an open, active chromatin

state. How these various factors may work to coordinate an imprinted expression pattern

is unknown and additional work is needed to understand their roles.

With the location of the PWS-IC established in the mouse, understanding the

epigenetic changes that occur within the IC could provide insight into the mechanism of

imprinting at the PWS/AS locus. As discussed earlier, differential methylation at CpG









islands is typically found within imprinted loci. The PWS/AS locus is no exception.

Two differentially methylated regions (DMR' s) exist within the PWS-IC, one within the

promoter of Snrpn (DMR1) and one within the first intron (DMR2) (Shemer et al., 1997).

These DMR' s are hypomethylated on the paternal allele and hypermethylated on the

maternal allele. Methylation of CpG's within the PWS-IC of mice occurs during

oogenesis, however in humans it occurs after fertilization suggesting that this is not the

primary imprint(El-Maarri et al., 2001). It has also been shown that parent specific

differential methylation of histone H13 exists in the same region (Xin et al., 2001). H13

Lys9 is methylated on the maternal copy of the PWS-IC, while H13 Lys4 is methylated on

the paternal copy. H13 Lys-9 methylation is correlated with transcriptional silencing

while H13 Lys4 methylation is correlated with transcriptional activity suggesting a

possible epigenetic switch. Furthermore CpG methylation has been tied to the G9a H13

Lys9/Lys27 methyltransferase (Xin et al., 2003). It' s been hypothesized that H13 Lys9

methylation could be the primary imprint and may trigger CpG methylation of the PWS-

IC through G9a leading to maternal silencing of the paternal genes on the maternal allele

after fertilization.

This chapter discusses modifications to the 425A5-7 BAC designed to further

elucidate the PWS-IC.

Results

BAC Modification

As a means to further define the minimal cis elements necessary to establish and

maintain imprinted expression in the locus I have created a series of 3 nested deletions

within the imprinted 425A5-7 BAC (Figure 4-1). The 5' anchor point of all three









deletions was positioned at approximately 30kb upstream of Snrpn exon 1. This point

was selected because targeted deletions made within the lab that extend an additional 12

kb further upstream of this point had proven to have no effect on imprinting (Peery et al.,

manuscript in preparation). The largest deletion, 425A30kb, extends to 133 bp upstream

of exon 1, removing all but the smallest minimal promoter (SMP) of Snrpn (Hershko et

al., 2001). This point was selected as to allow for transcription of the Snrpn minigene

reporter. The smallest deletion, 425Al7kb, extends to 1 1.5kb upstream of exon 1 and is

approximately 17kb in length. This point was selected based on reports from the Beaudet

lab of a 100kb targeted deletion anchored at the 3' end at the same position (Wu et al.,

2006). As will be discussed in the next chapter, this deletion had no effect on imprinting.

The third deletion, 425A23kb, extends to approximately 6kb upstream of exon 1 and is

23kb in length. This was selected as a midway point between the largest and smallest

deletions. The 425A5-7 BAC was targeted to produce these modifications using the

Lambda recombineering system previously discussed.

Production of Transgenic Mice

The 425A30kb BAC was chosen as the first construct to inj ect as it was supposed

that this deletion would have the greatest effect. Fresh 425A30kb BAC DNA was

prepared as previously described and was inj ected into the pronuclei of fertilized FVB

oocytes at a concentration of 2.5 ng/ul. Injected oocytes were transplanted at the two cell

stage into the oviducts of pseudopregnant B6D2 Fl mice. The resulting pups were

allowed to mature to 3 weeks of age at which time they were ear punched for

identification and tail clipped to obtain DNA for genotyping. Genomic tail DNA was

used as a template for PCR amplification of the transgene to screen for founders. The










primers T7F 5'- GTA ATA CGA CTC ACT ATA GGG C 3' and T7R1 5' CTC CAA

TCA TGT TCA ACT GTC C 3' amplify a region of the BAC adj acent to the vector

arms and are transgene specific. Five founders were obtained; lines C, D, E, F, and G.

Of these only lines C, D, and E transmitted to their offspring. Lines F and G were

discarded.

Analysis of 425A30kb Transgenic Lines

Founder animals were mated to FVB/N males or females to establish breeding

colonies and test for chimerism. Lines C, D, and E transmitted through the germ line and

the resulting Fl transgenic pups were allowed to mature to six weeks of age. At six

weeks, Fl transgenic males and females from each line were set up in matings with

FVB/N males and females to test for expression of the transgene upon both maternal and

paternal inheritance. Brains were collected from the resulting F2 day old neonatal pups.

RNA was extracted from the brains using RNAzol. The RNA was used as a template to

produce cDNA that was then used to program PCR reactions that spanned exons 4-8 of

Snrpn. Since the modification to the 425A5-7 BAC removes exons 5-7 the resulting

transgenic transcript appears as a band of approximately 350 base pairs (bp). This

includes 242 bp of exonic sequence plus the additional lox-p left over from targeting.

The endogenous band spans 513 bp. Unexpectedly, the 425A30kb transgene was

silenced in lines D, and E when maternally inherited. Line D never transmitted the

transgene paternally however line E expressed the Snrpn reporter gene upon paternal

transmission. Line C expressed the transgene biallelically, suggesting that a linearization

of the transgene had occurred at a point incompatible with imprinted expression.










Bisulfite sequencing was performed on the C and E lines to test for methylation at

the Snrpn DMR1 as was done with the 425A5-7 lines. However, the 30kb deletion made

it possible to amplify DMR1 without first backcrossing the transgene onto the Cast.c7

line of mice. The resulting bisulfite amplicon retains 10 of the original 15 CG

dinucleotides present within the CpG island. Twenty fiye micrograms of DNA of

genomic brain DNA from the C and E lines was bisulfite converted and then used to

program PCR reactions with the following primer set; 425.06 bisulf F2 5'- GTT TTA

GGA GTT RGA GGT TTT TT 3' and 425.06 bisulf R2 5' AAC CCA AAT CTA

AAA TAT TTT AAT C 3'. The products from these reactions were cloned and

sequenced. Data from 11 clones representing a paternally transmitted transgene and 9

clones representing a maternally transmitted transgene was obtained from the E line. In

summary, all paternally transmitted E alleles were completely unmethylated. In contrast,

five of nine maternal E alleles were completely methylated. Two of nine alleles were

methylated at nine sites. One of nine alleles was methylated at only site, and one allele

was completely unmethylated.

Discussion

The lack of data from multiple lines makes drawing strong conclusions difficult for

these experiments. However, it seems highly unlikely that imprinted expression of the

425A30kb transgene from line E could be due to an insertional effect. Therefore some

inferences can be made. First, line E expressed the Snrpn reporter in a correctly

imprinted fashion and line D was silenced when maternally inherited. This suggests that

all the PWS-IC cis-regulatory elements necessary for imprinting Snrpn are within the

region between intron 1 and 133bp upstream of exon 1. Since imprinting was preserved









in this, the longest deletion of the three construct built, the other two BACs were not

inj ected. The footprint and ChlP data reported by Rodriguez-Jato et-al demonstrate that

there are conserved elements within this region that are actively occupied, some in a

parent of origin specific manner (Rodriguez-Jato et al., 2005). These include NRF-1,

YY1, and SP1 sites located within intron 1 and an NRF-1 site just upstream of exon one

that had previously been show to be necessary for promoter activity (Hershko et al.,

2001). It can also be presumed that the AS-IC is still intact within this construct since the

transgene is silenced upon maternal transmission. A likely location would be upstream of

the deletion since such a small interval of sequence remains downstream.

A second observation is that the mechanism governing the imprinted expression of

Snrpn likely differs from that governing the upstream cluster of genes. As just

mentioned, Bressler et-al showed the region encompassing the first 600 bp upstream of

Snrpn exon 1 was dispensable for imprinting as judged by expression and methylation of

Ndn. Only when an additional 4kb of sequence upstream was deleted did they see an

imprinting defect. Although expression of the upstream genes is not measurable in this

transgenic system, it's fairly obvious that this more proximal region is important for

imprinting Snrpn since only 133bp of upstream sequence confer imprinted expression on

the E line. It seems possible that maternal silencing of the Snrpn long transcript may

simply be due to methylation of the promoter, thereby blocking access to trans-acting

factors required for transcription. Imprinting of the upstream genes however, would

require a longer range mechanism, perhaps similar to the active chromatin hub seen in the

beta-globin locus (Patrinos et al., 2004; Zhou et al., 2006). Looping of the upstream

genes to the PWS-IC has already been demonstrated in human cell culture (Arnold










Heggestad personal communication). The cis-regulatory elements that would control this

type of mechanism have yet to be identified.

A third point that can be made comes from the observation that the 425A30kb C

line expresses transgenic Snrpn in a biallelic fashion. Since the AS-IC is the element that

silences the maternal allele, mapping the point at which the BAC linearized before

integration may provide clues to the location of this element.






59












17 kb

r I
30 kb 11 5kb




Figure4-1. Schematic diagram of three nested deletions aimed at minimizing the PWS-
IC. Three nested deletions anchored at the 5' end 30 kb upstream of Snrpn,
were designed to decrease the minimal known interval required to imprint
Snrpn. The largest of the three extends to 133bp upstream of exon 1 or just 5'
to the smallest minimal promoter. The shortest deletion extends 17kb to a 3'
anchor point reported by the Beaudet lab. The third deletion extends to a mid-
point between the two.













9
il~a~-splirsion


d' O
7trn~misrinrr ~w~wji~s~iaPL


d"
h7lns~ksis~ian


~"""

rp


Figure 4-2. rt-PCR analysis of 425A30kb mouse lines. rt-PCR analysis of two transgenic
mouse lines derived with the 425A30kb BAC reveals proper imprinted
expression of the E line and biallelic expression of the C line. Proper
expression of the E line suggests all cis-regulatory elements required for
imprinted expression of Snrpn are contained within its interval. Biallelic
expression of line C may be due to linearization of the BAC that is
incompatible with imprinted expression.


T g (+)


T g (-)














425A30 kb Paternally Transmitted Transgene







425A30 kb Maternally Transmitted Transgene









Figure 4-3. Bisulfite sequence analysis of maternally transmitted 425A30kb transgene.
Sequence analysis of genomic DNA converted by sodium bisulfite reveals that
the maj ority of transgenic alleles are methylated at the Snrpn promoter CpG
island when inherited from the mother. A minority of alleles escape
methylation, perhaps accounting for the leaky expression seen from both the
transgene and endogenous alleles. When the transgene is paternally inherited
all transgenic alleles are completely demethylated.















CHAPTER 5
DEFINING THE LOCATION OF THE MUJRINE AS-IC

Introduction

Imprinting in the PWS/AS locus is dependent upon coordinate control by a

bipartite imprinting center, consisting of a PWS-IC that drives expression from the

paternal allele and an AS-IC that is responsible for silencing the PWS-IC on the maternal

allele. While the position of the AS-IC has been localized to a region approximately

3 5kb upstream of Snrpn exon 1 in humans, its location in the mouse is still unknown,

hampering efforts to understand its mechanism. Targeted deletions in the mouse, based

on a similar distance from Snrpn in the human, failed to disturb the differential

methylation and imprinted expression patterns of Snrpn and Ndn suggesting that the

location is not conserved. This chapter will focus on efforts to define the murine AS-IC

For years it has been postulated that silencing of the maternal allele may involve

expression of a series of alternative SNRPN transcripts that arise from a set of

infrequently used upstream exons. The first evidence for this came from analysis of

Angelman Syndrome patients inheriting microdeletions in a region upstream of exon 1.

A consequence of these deletions is a block in the paternal to maternal imprint switch

during gametogenesis resulting in a paternal epigenotype on the maternal chromosome

and biallelic expression of the paternal set of genes. Several upstream alternative SNRPN

exons were then isolated from this region (Dittrich et al., 1996; Farber et al., 1999) which

extends >700kb upstream of exon 1 (Figure 5-1). These upstream exons share a degree

of sequence similarity to each other and in tumn are also similar to exon one, implying that










they arose from duplications of exon one. Alternative transcripts initiate from the

paternal allele from two of these exons, ulA and ulB and give rise to various splice

variants. All splice variants skip exon one and splice to exon two. Maternal specific

methylation is also present in the regions upstream of these exons. What is most

interesting about these upstream exons is that exon u5 is deleted in all AS patients with

an IC deletion. Exon u5 is contained within multiple alternative transcripts so the

argument could be made that these transcripts could be somehow involved in the paternal

to maternal imprint switch. An alternative explanation is that the location of exon u5 is

merely a coincidence and the actual AS-IC is a cis regulatory element located in close

proximity. An argument in favor of this hypothesis is that no alternative transcripts to

date have been identified which initiate at exon u5. If transcription of u5 were necessary

for AS-IC function, patients with deletions encompassing exons ulA or ulB should be

present since these are the exons at which those transcripts identified to date initiate. It is

just as likely however that as yet unidentified transcripts that do initiate at u5 are present

within developing gametes, tissue which has not yet been tested.

Studying these issues in humans is complicated by the fact that the tissues in which

the imprints are being established, developing gametes, are difficult if not impossible to

acquire. As discussed earlier, gene content and the imprinted expression pattern within

the PWS/AS locus is conserved in the mouse making it an attractive model system.

However the location of the murine AS-IC is still not known. Given that the AS-IC is

located at approximately 3 5kb upstream of SNRPN in humans our lab targeted deletions

at the corresponding location in the mouse. Nested deletions of 7 and 12kb had no effect

on methylation or expression of the paternally expressed genes suggesting that the









location of the AS-IC relative to Snrpn is not conserved in the mouse (E. Peery personal

communication). Closer examination of this region reveals that the mouse also contains

several exons upstream to Snrpn that extend approximately 500kb upstream. Rt-per

analysis reveals that these upstream exons are also involved in alternative transcripts

similar to humans (Bressler et al., 2001; Landers et al., 2004). Nine U exons have been

identified to date. They participate in multiple splice variants that fall into two

categories, those that j oin to Snrpn exon 2 and those that exclude Snrpn and splice further

downstream to the non-coding Ipw exons. As is the case in humans, it appears that the

upstream exons are the result of multiple duplication events as there is a high degree of

homology to the region surrounding exon 1. Sequence is highly conserved among the

murine exons, more so than in humans, approaching 90%, however there is very little

conservation between human and mouse. The function that these alternative exons play

in imprinting is yet to be determined. In fact conflicting results from two labs either

supports or excludes alternative transcripts from playing a role. Extensive work by

LeMeur et-al characterized the temporal and spatial regulation of these transcripts (Le

Meur et al., 2005). Transcripts containing various U exons were found to be expressed

within embryos 10.5 days post-coitum (dpc) and later exclusively within the nervous

system. This group however failed to detect expression of the U exons within testis and

ovary samples suggesting that alternative transcripts are not produced within developing

gametes. This would preclude a role for alternative transcripts in establishing imprints

within the PWS/AS locus during a time when these marks are placed. Mapendano et-al

however did detect expression of alternative transcripts arising from exon U2 and

splicing to Snrpn exon 2 within the ovary by rt-per (Mapendano et al., 2006). In-situ










hybridization, utilizing exon U2 as a probe, detected expression within both oocytes and

their supporting granulosa cells within Graafian follicles. Stronger signals were detected

within oocytes and granulosa cells of secondary and developing follicles. While this

contradicts the observations by Le Meur et-al, it supports the idea that alternative

transcripts may be important to imprint establishment.

In a recently published report, Wu et al targeted deletions to the murine region

upstream of Snrpn (Wu et al., 2006). Two lines of mice were generated mutating the

region 5' of the Snrpn promoter. The first was an anchor mutation that introduced a

targeting vector 13kb upstream of Snrpn exon 1. The insertion was comprised of the 3'

portion of an Hprt selectable cassette, a loxP site, a puromycin selectable marker, and a

6kb duplication of genomic sequence. This mutation is referred to as AS-ICa". The

second mutation was an 80kb deletion of genomic sequence extending upstream of the 3'

anchor created by AS-IC". Paternal inheritance of either mutation resulted in a normal

pattern of methylation at the Snrpn and Ndn CpG islands. Maternal inheritance of the

AS-IC" mutation resulted in an imprint defect with both maternal and paternal alleles

being unmethylated at Snrpn. This was correlated with inappropriate maternal expression

of Snrpn. Ndn was also affected with a partial or mosaic pattern of methylation on the

maternal allele. All subsequent breeding of AS-IC" mutant mice resulted in an AS like

imprinting defect, with inappropriate hypomethylation of the maternal allele. Maternal

inheritance of the 80kb deletion resulted in an imprinting defect with incomplete

penetrance. Three groups of mice with different methylation patterns were observed.

The first (3 of 8 heterozygous mice) exhibited a normal pattern of methylation as

compared to wild type litter-mates. A second group (2 of 8) showed partial methylation









of the maternal allele. The third group of three was completely unmethylated. In trying

to understand the disruption of imprinting in both mutant lines, it is important to note that

the puromycin cassette is present and transcribes in the opposite orientation to Snrpn in

the AS-ICan mutant. In the 80kb deletion, the puromycin cassette is deleted by cre-lox,

but the Hprt promoter is present and expresses in the same direction as Snrpn. In the

context of the upstream exons, the anchor insertion alters the distance between the

putative AS-IC and the PWS-IC and perhaps the distance between the two regulatory

elements is important to coordinating imprinted expression. A second explanation is that

transcription from the puromycin cassette interferes with alternate transcripts from the

upstream exons. This second explanation may also be valid in the case of the 80kb

deletion. Transcription towards Snrpn from the Hprt promoter may interfere with the

normal transcription signal being sent from upstream.

More evidence regarding the location of the AS-IC comes from our lab in the form

of imprinted transgenes (Figure 5-2). Multiple transgenic lines have been created and

characterized in our lab extending variable distances upstream of Snrpn. As mentioned

earlier, a phage clone carrying 33kb of murine sequence 5' to Snrpn did not imprint

correctly as a single copy transgene suggesting that the AS-IC is not within that interval.

The 425A5-7 BAC transgene however does imprint correctly suggesting that the AS-IC

is somewhere within the 100kb interval upstream of Snrpn. The 215B BAC transgenic

line also is imprinted correctly. What is interesting about this line is that it has suffered a

truncation at approximately 40kb 5' to Snrpn. The location of mouse upstream exon 1 is

at 43kb 5' to Snrpn. Whether the 215B line does contain Ul is unknown and is currently

being investigated.










These observations have led to the following hypothesis (Figure 5-3). First, the

upstream exons or an element nearby act as the AS-IC and establishes the primary

imprint upon the PWS-IC. Second, the high degree of homology between the murine

upstream exons allows for compensation to occur when one is deleted. This may explain

the variable penetrance seen in the 80kb deletion. Exons further upstream may not be as

efficient in a role as the AS-IC as the closer Ul. Lastly, only deletion of all U exons will

result in a complete AS like phenotype. This would prove to be difficult in a targeted

deletion strategy as it may require individually targeting each of nine identical sequences

over a distance greater than 500kb. The BAC transgenic model provides an excellent

opportunity to ask these question as only three upstream exons are present within the

425A5-7 BAC interval.

In order to investigate the role that murine upstream Snrpn exons play in imprinted

expression of the locus several modifications to the 425A5-7 BAC were made. This

chapter characterizes those modifications in an attempt to understand AS-IC function.

Results

BAC modification

The 425A5-7 BAC transgene contains 3 upstream exons, Ul, U2, and U3 at 43kb,

75kb, and 96 kb upstream of Snrpn exon 1 respectively. In order to test their role in AS-

IC function, a 2499 bp region of homology in which each U exon is imbedded, was

targeted for deletion in a way as to leave the surrounding DNA intact. This 2.5kb region

of homology likely represents the region of Snrpn exon 1 duplicated to form the U exons

and may contain cis regulatory element required for silencing the PWS-IC. Targeting

vectors to be used to delete each upstream exon by BAC recombineering were designed









as follows. Primers were designed to flank each region of homology with approximately

1kb of additional sequence in either direction. The primer sets used are as follows; Ul -

UldelF 1 5'- GAC TTC AGT ACA TGT TCC AAC -3' and UldelR2 5'- TAG ATA

GTC CAG TAC CAG GAC T -3', U2 U2delF 5'- CCA GAA GAT GTT CCA ACT

GG -3' and U2delR 5'- CCA CCT AAC CAA TGG ACA GA 3', U3 U3delF 1 5'-

CCA GAT AGT GTA CCT CAT GTC -3' and U3delR1 5'- GAG GCT TCC ACA TAA

GGT CTT -3'. The products amplified with these primers were cloned into the pGEM T

easy cloning vector from Promega. In order to delete the region containing each U exon,

restriction endonuclease cleavage sites were identified and used to remove the portion of

sequence containing the regions of homology leaving 1kb of genomic sequence in either

direction intact. The Ul T-easy vector was cleaved with Afll and PacI to release a

2753bp fragment. The U2 T-easy vector was cleaved with BamHI to release a 2515bp

fragment. The U3 T-easy vector was cleaved with BssHII and Csp45I to release a

2608bp fragment. Each linearized vector was blunted with Klenow and a loxP flanked

neomycin resistance cassette was ligated in. In order to avoid cross-reaction between the

various targeted areas when multiple modifications were made to a single BAC, mutant

loxP sites were used to flank the neo cassette in the U2 and U3 targeting vectors. FRT

sites were used to flank the neo cassette in the Ul targeting vector. Lambda red

recombineering was used to make targeted deletions of the three U exons in the 425A5-7

BAC using these vectors. Transgene 425AUl-U3 has all three U exons deleted (Figure

5-4A). If our compensation hypothesis is correct, all three exons must be deleted to see a

complete imprinting defect. Transgene 425AU2/U3 has only U2 and U3 deleted (Figure

5-4B). This is important to show that exon Ul, or the region immediately adj acent to it,










serves as the AS-IC. Two more constructs, 425AUl/U3 and 425AUl/U2 leave exon U2

or U3 intact in order to demonstrate compensation.

Production of Transgenic Mice

The 425AUl-U3 and 425AU2/U3 BAC's were chosen as the first constructs to

inj ect. Fresh BAC DNA was prepared as previously described and was inj ected into the

pronuclei of fertilized FVB oocytes at a concentration of 2.5 ng/ul. Injected oocytes were

transplanted at the two cell stage into the oviducts of pseudopregnant B6D2 Fl mice. The

resulting pups were allowed to mature to 3 weeks of age at which time they were ear

punched for identification and tail clipped to obtain DNA for genotyping. Genomic tail

DNA was used as a template for per amplification of the transgene to screen for founders.

The primers T7F 5'- GTA ATA CGA CTC ACT ATA GGG C 3' and T7R1 5' CTC

CAA TCA TGT TCA ACT GTC C 3' amplify a region of the BAC adjacent to the

vector arms and are transgene specific. Four founders were obtained for the 425AUl-U3

line and two founders were obtained for the 425AU2/U3 line. All were found to transmit

the transgene and were consequently mated to Cast.c7 males and females to establish

colonies.

Analysis of 425AU1-U3 and 425AU2/U3 Transgenic Lines

Founder animals were mated to Castc7 males or females to establish breeding

colonies. Lines A, B, C, and D transmitted through the germ line and the resulting Nl

transgenic pups were allowed to mature to six weeks of age. At six weeks, Nl transgenic

males and females from each line were set up in matings with FVB/N males and females

to test for expression of the transgene upon both maternal and paternal inheritance.

Brains were collected from the resulting N2 day old neonatal pups. RNA was extracted










from the brains using RNAzol. The RNA was used as a template to produce cDNA that

was then used to program per reactions that spanned exons 4-8 of Snrpn. Since the

modification to the 425A5-7 BAC removes exons 5-7 the resulting transgenic transcript

appears as a band of approximately 350 base pairs (bp). This includes 242 bp of exonic

sequence plus the additional lox-p left over from targeting. The endogenous band spans

513 bp. Line A did not express the transgene upon paternal or maternal inheritance,

possibly due to insertional effects or breakage of the transgene at a point incompatible to

expression prior to insertion. Line D on the other hand expressed the transgene from both

the paternal and maternal alleles, suggesting that the AS-IC had been deleted by

removing the upstream exons. Lines B and C have only been tested for maternal

expression at this point. Line B expressed the maternally inherited transgene while line C

did not. Conclusions on these two lines cannot be made until expression data from the

paternally inherited alleles can be analyzed

Discussion

Due to the overlap of human Snrpn exon u5 with the AS-SRO, it has been

hypothesized that the upstream exons have some role in AS-IC function. However,

testing this hypothesis in the mouse has proven to be difficult, possibly due to the high

degree of homology among the nine murine upstream exons. Using our BAC transgenic

model provides a simpler model since only three U exons are contained within its

interval .

Since this data set is not yet complete, no firm conclusions can be made, however

biallelic expression from the 425AUl-U3 D line and maternal expression from the B line

suggest that the U exons do play a role in silencing the maternal allele. Expression from









the B and C paternally inherited alleles remains to be analyzed and this will prove to be

important in determining the imprinting status of this construct. Also it will be important

to determine whether or not the BAC remained intact during integration. If the transgene

is indeed intact throughout the region 5' of Snrpn, this raises the possibility that the U

exons also serve some sort of insulating purpose. All four 425A5-7 transgenic lines

proved to be capable of silencing maternally inherited Snrpn suggesting that the BAC

contains sufficient sequence to overcome the chromatin into which it inserts. The

425AUl-U3 lines do not seem to have this capability as the A line is completely silenced

upon both paternal and maternal inheritance. Southern blots spanning the region 5' to

Snrpn are currently underway to determine the whether the transgenes have remained

intact.

The 425AU2-U3 lines still remain to be analyzed. Determining their imprinting

status will provide conclusive evidence about whether the AS-IC resides within the

upstream exons. Cloning of additional BAC constructs 425AUl/U3 and 425AUl-U2 has

been completed however they have not yet been inj ected as transgenes. Data from these

lines should provide evidence about whether U2 and U3 can compensate for AS-IC

function when Ul is deleted.

The data presented in this chapter provides preliminary evidence that the AS-IC

resides within the upstream exons. However, this data does not suggest a mechanism by

which the AS-IC imprints the maternal allele. Two possibilities immediately come to

mind. First and probably the most obvious is that transcription factor binding sites lie

within the region that includes Ul. At this point in time however, this would be difficult









to test. Additional work would need to be done to further limit the interval in which the

AS-IC is contained in order for an efficient search for active binding sites could occur.

A second intriguing hypothesis is that active transcription from the upstream

promoter is what triggers heterochromatin formation and subsequent silencing of the

maternal allele. There is precedence for this type of mechanism in multiple systems.

First, Mayer et al have shown that transcripts from intergenic spacer sequences between

rRNA genes are important for establishing and maintaining a specific heterochromatin

conformation at the promoter of a subset of rDNA arrays (Mayer et al., 2006). These

150-300 nucleotide transcripts interact with TIPS, the large subunit of the chromatin

remodeling complex NORC and guide the complex to its target via sequence

complementarity to the rDNA promoter. A second example comes from the analysis of a

patient suffering from a-thalesemia, an inherited form of anemia. This patient was found

to have a deletion that resulted in the juxtaposition a structurally normal a-globin gene

(HBA2) to LUC7L (Tufarelli et al., 2003). Expression of LUC7L from the opposite

strand from HBA2 resulted in an anti-sense transcript that was capable of mediating

methylation and silencing of HBA2 during development. Through the use of transgenic

mice they were also able to show that methylation and modification of chromatin occurs

in cis. A third mechanism by which transcription could result in gene silencing is siRNA

mediated DNA methylation. Until recently this mechanism had only been known to be

active in plants. Kawasaki et al were able to show that mammalian cells are also capable

of siRNA mediated silencing within a cell culture model (Kawasaki and Taira, 2004).

Vector based siRNA' s targeting CpG islands within the E-cadherin and erbB2 genes were










shown to be capable of inducing DNA methylation and histone H3 lysine 9 methylation

resulting in repression of transcription.

The work contained in this chapter is ongoing and the data presented preliminary.

Expression from the paternal allele of the 425AUl-U3 lines B and C transgenes and

analysis of the 425AU2-U3 transgene remains to be done. Also, determination of the

breakpoints for these transgenes will be necessary before any firm conclusions can be

made.
















AS 500a PWPS ERO3
Paternal Allele .4-P ++
ryuK uID ryulD'" ul~ulB` ulA ly1WlA LI2 Uj u4 U5 El E2 E3 E4 E5 E6 E7 E8 E9 E10






Maternal Allele




Figure 5-1. Schematic diagram of human upstream exons. The PWS-IC region has
undergone duplication events resulting in the establishment of several
upstream exons. Low levels of paternal transcripts arise from these exons as
shown by the arrow heads. Lollipops represent the differential methylation
status at specific methylation sensitive restriction sites. Filled circles
represent methylated CpG's while empty circles represent unmethylated
CpG' s. (adapted from Farber et al.)













Brain
U9 US


U7 U6


US U4B U2 U3 UI


El E2 E3 E4 ES Ed E7 Ea E9 E10


non-imprinteel Pl dlone UULICruuuuu1

imprinted trunca8ted 215B; BACJI_1~_~T~LLi_7

imprint~ed 42565-7 BAC i n




Figure 5-2. Schematic diagram of upstream exons in the mouse. Upstream exons have
also arisen in the mouse via duplications of the PWS-IC region. Brain
specific expression has been characterized from the paternal allele as shown
by the black arrows. Oocyte expression has also been shown as illustrated by
the red arrows. Several transgenic lines have been developed which provide
evidence that the upstream exons may play a role in imprinting. The shorter
Pl clone does not contain any upstream exons and does not imprint in single
copy while the two larger BAC clones contain one or three upstream exons
and do imprint.


















Sequence within or adj acent to U1 contains the AS-IC element
and 8ilences the PWES;-IC when maernally inherited







Deletion of individual upstream~ exons have nlo effect on imlprinted
expression of transgene as others can compensate



a~l





Figure 5-3. Compensation model of AS-IC activity. Alternative exons further upstream
can compensate for loss of those closer to Snrpn. Only by deleting all
upstream exons will there be an imprinting defect.

























_


~lb8


_ a~ I


C.


csonL eear2 emon3


U5 UA


L3 UL


K~~I)


LI O 0 1
Asant anon Amn3


LI L


UZ UK


~~hB


Figure 5-4. Schematic diagram of upstream exon deletions. In order to elicit an AS-IC
imprinting defect it may be necessary to delete all upstream exons to avoid
compensation by those that remain.


Ul-U3 tar-geted BAC deletion


U2-U3 targeted BAC deletion





Transmission


Figure 5-5. RT-PCR analysis of 425AUl-U3 mouse lines. Preliminary RT-PCR analysis
of four transgenic mouse lines derived with the 425AUl-U3 BAC reveals
improper maternal expression of Snrpn in lines B and D. Line A does not
express the transgene upon either maternal or paternal inheritance. Line C
which has only been analyzed for maternal expression at this point does not
express from the maternally inherited allele. Initial findings suggest that
deletion of the upstream exons blocks imprinting of Snrpn on the maternal
allele.


r~C+) Tg~~)
cl ,-

9
L'
Tr~t~missiQ1 Translissian Trans~ission















CHAPTER 6
CONCLUSIONS AND FUTURE DIRECTIONS

The mechanism by which imprinted expression occurs within the Prader-Willi /

Angelman locus is currently unknown. Several animal models have been put forth in

order to further our understanding of this phenomenon, however the regulatory elements

that coordinate parental specific expression of this set of genes remains a mystery. This

work describes a new transgenic model, which will hopefully further our understanding

of the locus.

BAC transgenes have historically been shown to be good models of the genomic

loci, which they represent. Our BAC transgenic model faithfully recapitulates the

imprinted expression of the endogenous Snrpn therefore it must contain all the cis

regulatory elements required for establishment and maintenance of the imprint.

The 425A30kb modified BAC shortens the interval known to contain the PWS-IC

considerably and leaves us with a manageable amount of sequence within which to look

for the cis elements of the PWS-IC. Data from Rodriguez-Jato et al has already shown

that conserved transcription binding sites within this interval are occupied in a parent of

origin specific manner and may play a role. Further modifications to the 425A5-7 BAC

whereby these individual sites are mutated will help to identify those that are important.

The inability to identify the AS-IC in the mouse has hampered efforts to understand

how it silences the PWS-IC. Observations in humans have suggested that alternative

transcripts arising from exons upstream of Snrpn exon 1 may play a role, however

definitive evidence supporting this hypothesis has yet to be shown. The BAC transgenic









model makes it possible to study these upstream exons in isolation from the rest of the

locus. Preliminary evidence presented here suggests that exon Ul or the sequence in

within the duplicated region may serve as the AS-IC. The mechanism by which this

works is still not understood. It is possible that elements within this region attract trans-

acting factors to the locus. These factors could then create a silenced chromatin

conformation which spreads outward to silence the PWS-IC and thereby the paternally

expressed genes on the maternal allele. A second compelling hypothesis is that

transcription arising from the upstream exons regulates the imprint establishment at the

PWS-IC. Transcription of non-coding RNA has been shown to regulate epigenetic states

using a variety of mechanisms in both plant and mammalian cells. Determination of

whether the AS-IC functions via attraction of trans-acting factors or through a

transcription based mechanism is well within the capabilities of this model system.

Experiments are currently under way to insert an oocyte specific promoter in front of

Snrpn exon 1 within a non-imprinted BAC. It will be interesting to see if transcription

through the PWS-IC within the context of a growing oocyte will confer imprinted status

to this transgene. Other obvious directions are to take a "promoter bashing" approach,

deleting larger and larger portions of exon Ul to reveal the relevant sequences.

The work contained within this document will certainly add to the understanding of

imprinting within the PWS/AS, however understanding how the locus is controlled has

further implications. Mouse models for PWS and AS will be important in developing

treatments for these disorders. Understanding the elements that make up the PWS-IC and

AS-IC will be important in developing for developing these models. .Many imprinted

loci are contained within the genome, but only in a small portion is the mechanism of










parental gene regulation understood. Furthering the understanding of this locus may

provide clues to how imprinting works at other regions. Also as the possibility of

therapeutic cloning becomes a reality, understanding gene regulation will become

essential. Disrupted imprinting has been suggested to play a role in the low survival rate

of animals cloned by somatic cell nuclear transfer. It is therefore likely that the genetic

reprogramming necessary to clone therapeutic tissues may also prove to be highly error

prone. Understanding how imprinting is controlled will be important to develop these

therapeutic technologies. There have also been reports of disrupted imprinting in

children conceived by intracytoplasmic sperm injection (ICSI). As assisted reproduction

techniques become more commonplace, understanding how imprints are acquired and

maintained will be important in reducing the risks associated with these procedures.
















LIST OF REFERENCES


Amos-Landgraf, J. M., Ji, Y., Gottlieb, W., Depinet, T., Wandstrat, A. E., Cassidy,
S. B., Driscoll, D. J., Rogan, P. K~, Schwartz, S. and Nicholls, R. D. (1999).
Chromosome breakage in the Prader-Willi and Angelman syndromes involves
recombination between large, transcribed repeats at proximal and distal breakpoints. Am
JHunt Genet 65, 370-86.

Bartolomei, M. S., Webber, A. L., Brunkow, M. E. and Tilghman, S. M. (1993).
Epigenetic mechanisms underlying the imprinting of the mouse H119 gene. Genes Dev 7,
1663-73.

Barton, S. C., Surani, M. A. and Norris, M. L. (1984). Role of paternal and maternal
genomes in mouse development. Nature 311, 374-6.

Beaudet, A. L. and Jiang, Y. H. (2002). A rheostat model for a rapid and reversible
form of imprinting-dependent evolution. Am JHunt Genet 70, 1389-97.

Blaydes, S. M., Elmore, M., Yang, T. and Brannan, C. I. (1999). Analysis of murine
Snrpn and human SNRPN gene imprinting in transgenic mice. 2Mann Genome 10, 549-
55.

Boccaccio, I., Glatt-Deeley, H., Watrin, F., Roeckel, N., Lalande, M. and Muscatelli,
F. (1999). The human MAGEL2 gene and its mouse homologue are paternally expressed
and mapped to the Prader-Willi region. Hunt Mol Genet 8, 2497-505.

Brannan, C. I. and Bartolomei, M. S. (1999). Mechanisms of genomic imprinting. Curr
Opin Genet Dev 9, 164-70.

Bressler, J., Tsai, T. F., Wu, M. Y., Tsai, S. F., Ramirez, M. A., Armstrong, D. and

Beaudet, A. L. (2001). The SNRPN promoter is not required for genomic imprinting of
the Prader-Willi/Angelman domain in mice. Nat Genet 28, 232-40.

Biting, K., Lich, C., Cottrell, S., Barnicoat, A. and Horsthemke, B. (1999). A 5-kb
imprinting center deletion in a family with Angelman syndrome reduces the shortest
region of deletion overlap to 880 bp. Hunt Genet 105, 665-6.

Biting, K., Saitoh, S., Gross, S., Dittrich, B., Schwartz, S., Nicholls, R. D. and
Horsthemke, B. (1995). Inherited microdeletions in the Angelman and Prader-Willi
syndromes define an imprinting centre on human chromosome 15. Nat Genet 9, 395-400.










Butler, M. G., Bittel, D. C., Kibiryeva, N., Talebizadeh, Z. and Thompson, T. (2004).
Behavioral differences among subjects with Prader-Willi syndrome and type I or type II
deletion and maternal disomy. Pediatrics 113, 565-73.

Butler, M. G. and Palmer, C. G. (1983). Parental origin of chromosome 15 deletion in
Prader-Willi syndrome. Lancet 1, 1285-6.

Cassidy, S. B. and Ledbetter, D. H. (1989). Prader-Willi syndrome. Neurol Clin 7, 37-
54.

Cattanach, B. M. (1986). Parental origin effects in mice. JEmbryol Exp M~orphol 97
Suppl, 137-50.

Cattanach, B. M. and Kirk, M. (1985). Differential activity of maternally and
paternally derived chromosome regions in mice. Nature 315, 496-8.

Chai, J. H., Locke, D. P., Ohta, T., Greally, J. M. and Nicholls, R. D. (2001).
Retrotransposed genes such as Frat3 in the mouse Chromosome 7C Prader-Willi
syndrome region acquire the imprinted status of their insertion site. Mamm Genome 12,
813-21.

Chamberlain, S. J. and Brannan, C. I. (2001). The Prader-Willi syndrome imprinting
center activates the paternally expressed murine Ube3a antisense transcript but represses
paternal Ube3a. Genomics 73, 316-22.

Chamberlain, S. J., Johnstone, K. A., DuBose, A. J., Simon, T. A., Bartolomei, M. S.,
Resnick, J. L. and Brannan, C. I. (2004). Evidence for genetic modifiers of postnatal
lethality in PWS-IC deletion mice. Hum Mol1 Genet 13, 2971-7.

Church, G. M. and Gilbert, W. (1985). The genomic sequencing technique. In Prog.
Clin. Biol. Res., vol. 177 (ed., pp. 17-21.

Clark, S. J., Harrison, J., Paul, C. L. and Frommer, M. (1994). High sensitivity
mapping of methylated cytosines. In Nucleic Acids Res, vol. 22 (ed., pp. 2990-7.

Clayton-Smith, J. and Pembrey, M. E. (1992). Angelman syndrome. J2~ed Genet 29,
412-5.

de los Santos, T., Schweizer, J., Rees, C. A. and Francke, U. (2000). Small
evolutionarily conserved RNA, resembling C/D box small nucleolar RNA, is transcribed
from PWCR1, a novel imprinted gene in the Prader-Willi deletion region, which Is highly
expressed in brain. Am JHum Genet 67, 1067-82.










Dittrich, B., Buiting, K., Korn, B., Rickard, S., Buxton, J., Saitoh, S., Nicholls, R. D.,
Poustka, A., Winterpacht, A., Zabel, B. et al. (1996). Imprint switching on human
chromosome 15 may involve alternative transcripts of the SNRPN gene. Nat Genet 14,
163-70.

Driscoll, D. J. (1994). Genomic imprinting in humans. Mol1 Genet Med 4, 37-77.

El-Maarri, O., Buiting, K., Peery, E. G., Kroisel, P. M., Balaban, B., Wagner, K.,
Urman, B., Heyd, J., Lich, C., Brannan, C. I. et al. (2001). Maternal methylation
imprints on human chromosome 15 are established during or after fertilization. Nat Genet
27, 341-4.

Farber, C., Dittrich, B., Buiting, K. and Horsthemke, B. (1999). The chromosome 15
imprinting centre (IC) region has undergone multiple duplication events and contains an
upstream exon of SNRPN that is deleted in all Angelman syndrome patients with an IC
microdeletion. Hunt Mol Genet 8, 33 7-43.

Gray, T. A., Saitoh, S. and Nicholls, R. D. (1999). An imprinted, mammalian
bicistronic transcript encodes two independent proteins. Proc NatlAcad Sci US A 96,
5616-21.

Hata, K., Okano, M., Lei, H. and Li, E. (2002). Dnmt3L cooperates with the Dnmt3
family of de novo DNA methyltransferases to establish maternal imprints in mice.
Development 129, 1983-93.

Hershko, A. Y., Finberg, Y., Kantor, B., Shemer, R. and Razin, A. (2001). The mouse
Snrpn minimal promoter and its human orthologue: activity and imprinting. Genes Cells
6, 967-75.

Holm, V. A., Cassidy, S. B., Butler, M. G., Hanchett, J. M., Greenswag, L. R.,
Whitman, B. Y. and Greenberg, F. (1993). Prader-Willi syndrome: consensus
diagnostic criteria. Pediatrics 91, 398-402.

Horsthemke, B., Dittrich, B. and Buiting, K. (1997). Imprinting mutations on human
chromosome 15. Hum Mutat 10, 329-37.

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, 829-38.

Jacobs, P. A., Wilson, C. M., Sprenkle, J. A., Rosenshein, N. B. and Migeon, B. R.
(1980). Mechanism of origin of complete hydatidiform moles. Nature 286, 714-6.










John, R. M., Aparicio, S. A., Ainscough, J. F., Arney, K. L., Khosla, S., Hawker, K.,
Hilton, K. J., Barton, S. C. and Surani, M. A. (2001). Imprinted expression of
neuronatin from modified BAC transgenes reveals regulation by distinct and distant
enhancers. Dev Biol 236, 387-99.

Jong, M. T., Carey, A. H., Caldwell, K. A., Lau, M. H., Handel, M. A., Driscoll, D.
J., Stewart, C. L., Rinchik, E. M. and Nicholls, R. D. (1999a). Imprinting of a RING
zinc-finger encoding gene in the mouse chromosome region homologous to the Prader-
Willi syndrome genetic region. Hunt Mol Genet 8, 795-803.

Jong, M. T., Gray, T. A., Ji, Y., Glenn, C. C., Saitoh, S., Driscoll, D. J. and Nicholls,
R. D. (1999b). A novel imprinted gene, encoding a RING zinc-finger protein, and
overlapping antisense transcript in the Prader-Willi syndrome critical region. Hunt Mol
G~enet 8, 783-93.

Kawasaki, H. and Taira, K. (2004). Induction of DNA methylation and gene silencing
by short interfering RNAs in human cells. Nature 431, 211-7.

Kishino, T., Lalande, M. and Wagstaff, J. (1997). UBE3A/E6-AP mutations cause
Angelman syndrome. Nat Genet 15, 70-3.

Knoll, J. H., Glatt, K. A., Nicholls, R. D., Malcolm, S. and Lalande, M. (1991).
Chromosome 15 uniparental disomy is not frequent in Angelman syndrome. Am JHunt
G~enet 48, 16-21.

Knoll, J. H., Nicholls, R. D. and Lalande, M. (1989a). On the parental origin of the
deletion in Angelman syndrome. Hunt Genet 83, 205-7.

Knoll, J. H., Nicholls, R. D., Magenis, R. E., Graham, J. M., Jr., Lalande, M. and
Latt, S. A. (1989b). Angelman and Prader-Willi syndromes share a common
chromosome 15 deletion but differ in parental origin of the deletion. Am J2~ed Genet 32,
285-90.

Landers, M., Bancescu, D. L., Le Meur, E., Rougeulle, C., Glatt-Deeley, H.,
Brannan, C., Muscatelli, F. and Lalande, M. (2004). Regulation of the large
(approximately 1000 kb) imprinted murine Ube3a antisense transcript by alternative
exons upstream of Snurf/Snrpn. Nucleic Acid~s Res 32, 3480-92.

Le Meur, E., Watrin, F., Landers, M., Sturny, R., Lalande, M. and Muscatelli, F.
(2005). Dynamic developmental regulation of the large non-coding RNA associated with
the mouse 7C imprinted chromosomal region. Dev Biol 286, 587-600.

Ledbetter, D. H., Riccardi, V. M., Airhart, S. D., Strobel, R. J., Keenan, B. S. and
Crawford, J. D. (1981). Deletions of chromosome 15 as a cause of the Prader-Willi
syndrome. NEngl J2~ed 304, 325-9.










Lee, E. C., Yu, D., Martinez de Velasco, J., Tessarollo, L., Swing, D. A., Court, D. L.,
Jenkins, N. A. and Copeland, N. G. (2001). A highly efficient Escherichia coli-based
chromosome engineering system adapted for recombinogenic targeting and subcloning of
BAC DNA. Genontics 73, 56-65.

Lee, S., Kozlov, S., Hernandez, L., Chamberlain, S. J., Brannan, C. I., Stewart, C. L.
and Wevrick, R. (2000). Expression and imprinting of MAGEL2 suggest a role in
Prader-willi syndrome and the homologous murine imprinting phenotype. Hunt Mol
G~enet 9, 1813-9.

Leff, S. E., Brannan, C. I., Reed, M. L., Ozcelik, T., Francke, U., Copeland, N. G.
and Jenkins, N. A. (1992). Maternal imprinting of the mouse Snrpn gene and conserved
linkage homology with the human Prader-Willi syndrome region. Nat Genet 2, 259-64.

Li, L. L., Szeto, I. Y., Cattanach, B. M., Ishino, F. and Surani, M. A. (2000).
Organization and parent-of-origin-specific methylation of imprinted Peg3 gene on mouse
proximal chromosome 7. Genontics 63, 333-40.

Linder, D., McCaw, B. K. and Hecht, F. (1975). Parthenogenic origin of benign
ovarian teratomas. NEngl J2~ed 292, 63-6.

Lossie, A. C., Whitney, M. M., Amidon, D., Dong, H. J., Chen, P., Therinque, D.,
Hutson, A., Nicholls, R. D., Zori, R. T., Williams, C. A. et al. (2001). Distinct
phenotypes distinguish the molecular classes of Angelman syndrome. J2~ed Genet 38,
834-45.

MacDonald, H. R. and Wevrick, R. (1997). The necdin gene is deleted in Prader-Willi
syndrome and is imprinted in human and mouse. Hunt Mol Genet 6, 1873-8.

Magenis, R. E., Brown, M. G., Lacy, D. A., Budden, S. and LaFranchi, S. (1987). Is
Angelman syndrome an alternate result of del(15)(q 11ql3)? Am J2~ed Genet 28, 829-38.

Malcolm, S., Clayton-Smith, J., Nichols, M., Robb, S., Webb, T., Armour, J. A.,
Jeffreys, A. J. and Pembrey, M. E. (1991). Uniparental paternal disomy in Angelman's
syndrome. Lancet 337, 694-7.

Mapendano, C. K., Kishino, T., Miyazaki, K., Kondo, S., Yoshiura, K., Hishikawa,
Y., Koji, T., Niikawa, N. and Ohta, T. (2006). Expression of the Snurf-Snrpn IC
transcript in the oocyte and its putative role in the imprinting establishment of the mouse
7C imprinting domain. JHunt Genet 51, 236-43.

Mayer, C., Schmitz, K. M., Li, J., Grummt, I. and Santoro, R. (2006). Intergenic
transcripts regulate the epigenetic state of rRNA genes. Mol1 Cell 22, 351-61.










McGrath, J. and Solter, D. (1984). Completion of mouse embryogenesis requires both
the maternal and paternal genomes. Cell 37, 179-83.

Moore, T. and Haig, D. (1991). Genomic imprinting in mammalian development: a
parental tug-of-war. Trends Genet 7, 45-9.

Nicholls, R. D., Knoll, J. H., Butler, M. G., Karam, S. and Lalande, M. (1989).
Genetic imprinting suggested by maternal heterodisomy in nondeletion Prader-Willi
syndrome. Nature 342, 281-5.

Ohama, K., Nomura, K., Okamoto, E., Fukuda, Y., Ihara, T. and Fujiwara, A.
(1985). Origin of immature teratoma of the ovary. Am J Obstet Gynecol 152, 896-900.

Patrinos, G. P., de Krom, M., de Boer, E., Langeveld, A., Imam, A. M., Strouboulis,
J., de Laat, W. and Grosveld, F. G. (2004). Multiple interactions between regulatory
regions are required to stabilize an active chromatin hub. Genes Dev 18, 1495-509.

Pfeifer, K., Leighton, P. A. and Tilghman, S. M. (1996). The structural H19 gene is
required for transgene imprinting. Proc NatlAcad Sci USA 93, 13876-83.

Reik, W. and Dean, W. (2001). DNA methylation and mammalian epigenetics.
Electrophoresis 22, 283 8-43.

Rodriguez-Jato, S., Nicholls, R. D., Driscoll, D. J. and Yang, T. P. (2005).
Characterization of cis- and trans-acting elements in the imprinted human SNURF-
SNRPN locus. Nucleic Acids Res 33, 4740-53.

Runte, M., Huttenhofer, A., Gross, S., Kiefmann, M., Horsthemke, B. and Buiting,
K. (2001). The IC-SNURF-SNRPN transcript serves as a host for multiple small
nucleolar RNA species and as an antisense RNA for UBE3A. Hum Mol1 Genet 10, 2687-
700.

Saitoh, S., Buiting, K., Rogan, P. K., Buxton, J. L., Driscoll, D. J., Arnemann, J.,
Konig, R., Malcolm, S., Horsthemke, B. and Nicholls, R. D. (1996). Minimal
definition of the imprinting center and fixation of chromosome 15qll1-ql3 epigenotype
by imprinting mutations. Proc Natl Acad Sci U S A 93, 781 1-5.

Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Cloning, (ed.: Cold
Spring Harbor Laboratory Press.

Schmidt, J. V., Matteson, P. G., Jones, B. K., Guan, X. J. and Tilghman, S. M.
(2000). The Dlk1 and Gtl2 genes are linked and reciprocally imprinted. Genes Dev 14,
1997-2002.










Searle, A. G. and Beechey, C. V. (1978). Complementation studies with mouse
translocations. Cytogenet Cell Genet 20, 282-303.

Shemer, R., Birger, Y., Riggs, A. D. and Razin, A. (1997). Structure of the imprinted
mouse Snrpn gene and establishment of its parental-specific methylation pattern. Proc
Natl AcadSci USA 94, 10267-72.

Sutcliffe, J. S., Jiang, Y. H., Galijaard, R. J., Matsuura, T., Fang, P., Kubota, T.,
Christian, S. L., Bressler, J., Cattanach, B., Ledbetter, D. H. et al. (1997). The E6-Ap
ubiquitin-protein ligase (UBE3A) gene is localized within a narrowed Angelman
syndrome critical region. Genome Res 7, 368-77.

Szeto, I. Y., Barton, S. C., Keverne, E. B. and Surani, A. M. (2004). Analysis of
imprinted murine Peg3 locus in transgenic mice. 2Mann Genome 15, 284-95.

Tufarelli, C., Stanley, J. A., Garrick, D., Sharpe, J. A., Ayyub, H., Wood, W. G. and
Higgs, D. R. (2003). Transcription of antisense RNA leading to gene silencing and
methylation as a novel cause of human genetic disease. Nat Genet 34, 157-65.

Varmuza, S. and Mann, M. (1994). Genomic imprinting--defusing the ovarian time
bomb. Trends Genet 10, 118-23.

Wu, M. Y., Chen, K. S., Bressler, J., Hou, A., Tsai, T. F. and Beaudet, A. L. (2006).
Mouse imprinting defect mutations that model Angelman syndrome. Genesis 44, 12-22.

Xin, Z., Allis, C. D. and Wagstaff, J. (2001). Parent-specific complementary patterns of
histone H3 lysine 9 and H3 lysine 4 methylation at the Prader-Willi syndrome imprinting
center. Am JHunt Genet 69, 1389-94.

Xin, Z., Tachibana, M., Guggiari, M., Heard, E., Shinkai, Y. and Wagstaff, J.
(2003). Role of histone methyltransferase G9a in CpG methylation of the Prader-Willi
syndrome imprinting center. JBiol Chent 278, 14996-5000.

Yang, T., Adamson, T. E., Resnick, J. L., Leff, S., Wevrick, R., Francke, U., Jenkins,
N. A., Copeland, N. G. and Brannan, C. I. (1998). A mouse model for Prader-Willi
syndrome imprinting-centre mutations. Nat Genet 19, 25-31.

Yevtodiyenko, A., Steshina, E. Y., Farner, S. C., Levorse, J. M. and Schmidt, J. V.
(2004). A 178-kb BAC transgene imprints the mouse Gtl2 gene and localizes tissue-
specific regulatory elements. Genontics 84, 277-87.

Zhou, G. L., Xin, L., Song, W., Di, L. J., Liu, G., Wu, X. S., Liu, D. P. and Liang, C.
C. (2006). Active chromatin hub of the mouse alpha-globin locus forms in a transcription
factory of clustered housekeeping genes. Mol1 CellBiol 26, 5096-105.
















BIOGRAPHICAL SKETCH

Chris Futtner was born in Hartford Connecticut in 1971. He graduated with honors

from Eastbury Elementary School at the age of ten; it was then that he became certain

that he would become a famous baseball player. Unfortunately, baseball didn't work out

for him, so he moved to Ft. Collins, CO, were he received a BS in microbiology at

Colorado State University in the year 1998. Settling down and earning large sums of

money did not appeal to Chris so he decided to return to school and pursue a PhD at the

University of Florida which he received in May of 2007. With his degree, Chris hopes

one day to become a contestant on the Wheel of Fortune, and then parlay his winnings to

become a cattle baron.